diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,12669 @@
+{
+  "best_metric": 1.143676996231079,
+  "best_model_checkpoint": "./SOLAR_NEW_DATASET_SLEEP_fc_out/checkpoint-1800",
+  "epoch": 1.8828451882845187,
+  "eval_steps": 300,
+  "global_step": 1800,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.1021552085876465,
+      "learning_rate": 1e-06,
+      "loss": 2.2775,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": NaN,
+      "learning_rate": 1e-06,
+      "loss": 2.2251,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.894376277923584,
+      "learning_rate": 2e-06,
+      "loss": 2.5432,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.441768646240234,
+      "learning_rate": 2.9999999999999997e-06,
+      "loss": 2.4134,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.577650785446167,
+      "learning_rate": 4e-06,
+      "loss": 1.7783,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6100335121154785,
+      "learning_rate": 4.9999999999999996e-06,
+      "loss": 1.9091,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.566389560699463,
+      "learning_rate": 5.999999999999999e-06,
+      "loss": 1.6825,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.842864513397217,
+      "learning_rate": 7e-06,
+      "loss": 2.0507,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.9386281967163086,
+      "learning_rate": 8e-06,
+      "loss": 1.743,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.832657814025879,
+      "learning_rate": 8.999999999999999e-06,
+      "loss": 2.1379,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7397637367248535,
+      "learning_rate": 9.999999999999999e-06,
+      "loss": 1.1938,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1158039569854736,
+      "learning_rate": 1.1e-05,
+      "loss": 1.6535,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9051222801208496,
+      "learning_rate": 1.1999999999999999e-05,
+      "loss": 1.5113,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.472515106201172,
+      "learning_rate": 1.3e-05,
+      "loss": 1.7566,
+      "step": 14
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.4902000427246094,
+      "learning_rate": 1.4e-05,
+      "loss": 2.5336,
+      "step": 15
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.6271560192108154,
+      "learning_rate": 1.4999999999999999e-05,
+      "loss": 1.9697,
+      "step": 16
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.09641695022583,
+      "learning_rate": 1.6e-05,
+      "loss": 2.0609,
+      "step": 17
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0167582035064697,
+      "learning_rate": 1.6999999999999996e-05,
+      "loss": 1.8016,
+      "step": 18
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.549201011657715,
+      "learning_rate": 1.7999999999999997e-05,
+      "loss": 2.3704,
+      "step": 19
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9748847484588623,
+      "learning_rate": 1.9e-05,
+      "loss": 1.811,
+      "step": 20
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.768181324005127,
+      "learning_rate": 1.9999999999999998e-05,
+      "loss": 1.5366,
+      "step": 21
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.290782928466797,
+      "learning_rate": 2.1e-05,
+      "loss": 2.4972,
+      "step": 22
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7193872928619385,
+      "learning_rate": 2.2e-05,
+      "loss": 1.7859,
+      "step": 23
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.833155632019043,
+      "learning_rate": 2.2999999999999997e-05,
+      "loss": 2.0601,
+      "step": 24
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.297412157058716,
+      "learning_rate": 2.3999999999999997e-05,
+      "loss": 1.3346,
+      "step": 25
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.314594268798828,
+      "learning_rate": 2.4999999999999998e-05,
+      "loss": 1.9146,
+      "step": 26
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7510643005371094,
+      "learning_rate": 2.6e-05,
+      "loss": 1.8501,
+      "step": 27
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2888293266296387,
+      "learning_rate": 2.6999999999999996e-05,
+      "loss": 1.8301,
+      "step": 28
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.5635504722595215,
+      "learning_rate": 2.8e-05,
+      "loss": 2.1873,
+      "step": 29
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.103892803192139,
+      "learning_rate": 2.8999999999999997e-05,
+      "loss": 1.9607,
+      "step": 30
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.598682880401611,
+      "learning_rate": 2.9999999999999997e-05,
+      "loss": 1.7272,
+      "step": 31
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7731786966323853,
+      "learning_rate": 3.0999999999999995e-05,
+      "loss": 1.8001,
+      "step": 32
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.651439905166626,
+      "learning_rate": 3.2e-05,
+      "loss": 1.6389,
+      "step": 33
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4542276859283447,
+      "learning_rate": 3.2999999999999996e-05,
+      "loss": 1.7928,
+      "step": 34
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9039721488952637,
+      "learning_rate": 3.399999999999999e-05,
+      "loss": 1.9144,
+      "step": 35
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4144136905670166,
+      "learning_rate": 3.5e-05,
+      "loss": 1.7467,
+      "step": 36
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.434959411621094,
+      "learning_rate": 3.5999999999999994e-05,
+      "loss": 2.1261,
+      "step": 37
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.9881653785705566,
+      "learning_rate": 3.7e-05,
+      "loss": 1.7411,
+      "step": 38
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.064974308013916,
+      "learning_rate": 3.8e-05,
+      "loss": 1.6445,
+      "step": 39
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.88478946685791,
+      "learning_rate": 3.9e-05,
+      "loss": 1.7376,
+      "step": 40
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7643675804138184,
+      "learning_rate": 3.9999999999999996e-05,
+      "loss": 1.6438,
+      "step": 41
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1387925148010254,
+      "learning_rate": 4.0999999999999994e-05,
+      "loss": 1.4881,
+      "step": 42
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5704689025878906,
+      "learning_rate": 4.2e-05,
+      "loss": 1.4022,
+      "step": 43
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.6913387775421143,
+      "learning_rate": 4.3e-05,
+      "loss": 1.6571,
+      "step": 44
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2858211994171143,
+      "learning_rate": 4.4e-05,
+      "loss": 1.7759,
+      "step": 45
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.481564998626709,
+      "learning_rate": 4.4999999999999996e-05,
+      "loss": 1.6854,
+      "step": 46
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.004473924636841,
+      "learning_rate": 4.599999999999999e-05,
+      "loss": 1.5408,
+      "step": 47
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.49001145362854,
+      "learning_rate": 4.7e-05,
+      "loss": 1.6673,
+      "step": 48
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1351754665374756,
+      "learning_rate": 4.7999999999999994e-05,
+      "loss": 1.3816,
+      "step": 49
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.705871820449829,
+      "learning_rate": 4.899999999999999e-05,
+      "loss": 1.8553,
+      "step": 50
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0498299598693848,
+      "learning_rate": 4.9999999999999996e-05,
+      "loss": 1.4696,
+      "step": 51
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.111211061477661,
+      "learning_rate": 5.1e-05,
+      "loss": 1.6067,
+      "step": 52
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0597739219665527,
+      "learning_rate": 5.2e-05,
+      "loss": 1.334,
+      "step": 53
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0325064659118652,
+      "learning_rate": 5.2999999999999994e-05,
+      "loss": 1.2886,
+      "step": 54
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.371424436569214,
+      "learning_rate": 5.399999999999999e-05,
+      "loss": 1.5398,
+      "step": 55
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.390995025634766,
+      "learning_rate": 5.499999999999999e-05,
+      "loss": 1.9422,
+      "step": 56
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.701133728027344,
+      "learning_rate": 5.6e-05,
+      "loss": 1.9618,
+      "step": 57
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4953861236572266,
+      "learning_rate": 5.6999999999999996e-05,
+      "loss": 1.6602,
+      "step": 58
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.652651071548462,
+      "learning_rate": 5.7999999999999994e-05,
+      "loss": 1.5143,
+      "step": 59
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.5872697830200195,
+      "learning_rate": 5.899999999999999e-05,
+      "loss": 1.2444,
+      "step": 60
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.623898506164551,
+      "learning_rate": 5.9999999999999995e-05,
+      "loss": 1.7179,
+      "step": 61
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.985961437225342,
+      "learning_rate": 6.1e-05,
+      "loss": 1.4591,
+      "step": 62
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.351639986038208,
+      "learning_rate": 6.199999999999999e-05,
+      "loss": 1.4416,
+      "step": 63
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.620721817016602,
+      "learning_rate": 6.299999999999999e-05,
+      "loss": 1.258,
+      "step": 64
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.6299662590026855,
+      "learning_rate": 6.4e-05,
+      "loss": 1.2641,
+      "step": 65
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.620664596557617,
+      "learning_rate": 6.5e-05,
+      "loss": 1.469,
+      "step": 66
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.9535703659057617,
+      "learning_rate": 6.599999999999999e-05,
+      "loss": 1.4236,
+      "step": 67
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.4887490272521973,
+      "learning_rate": 6.699999999999999e-05,
+      "loss": 1.322,
+      "step": 68
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.64335823059082,
+      "learning_rate": 6.799999999999999e-05,
+      "loss": 1.438,
+      "step": 69
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.370468854904175,
+      "learning_rate": 6.9e-05,
+      "loss": 1.2633,
+      "step": 70
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.171403408050537,
+      "learning_rate": 7e-05,
+      "loss": 1.6355,
+      "step": 71
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2457258701324463,
+      "learning_rate": 7.099999999999999e-05,
+      "loss": 1.5638,
+      "step": 72
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.745772361755371,
+      "learning_rate": 7.199999999999999e-05,
+      "loss": 1.3061,
+      "step": 73
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8581624031066895,
+      "learning_rate": 7.3e-05,
+      "loss": 1.0935,
+      "step": 74
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.144169807434082,
+      "learning_rate": 7.4e-05,
+      "loss": 1.4205,
+      "step": 75
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.5456461906433105,
+      "learning_rate": 7.5e-05,
+      "loss": 1.2544,
+      "step": 76
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.343386650085449,
+      "learning_rate": 7.6e-05,
+      "loss": 1.7345,
+      "step": 77
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.298736095428467,
+      "learning_rate": 7.699999999999999e-05,
+      "loss": 1.4268,
+      "step": 78
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.657853126525879,
+      "learning_rate": 7.8e-05,
+      "loss": 1.1615,
+      "step": 79
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.102127552032471,
+      "learning_rate": 7.899999999999998e-05,
+      "loss": 1.3367,
+      "step": 80
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.519240140914917,
+      "learning_rate": 7.999999999999999e-05,
+      "loss": 1.2521,
+      "step": 81
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7445943355560303,
+      "learning_rate": 8.1e-05,
+      "loss": 1.0568,
+      "step": 82
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.702052116394043,
+      "learning_rate": 8.199999999999999e-05,
+      "loss": 1.2667,
+      "step": 83
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.428670883178711,
+      "learning_rate": 8.3e-05,
+      "loss": 1.1848,
+      "step": 84
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.7245421409606934,
+      "learning_rate": 8.4e-05,
+      "loss": 1.5863,
+      "step": 85
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.550326824188232,
+      "learning_rate": 8.499999999999999e-05,
+      "loss": 1.0869,
+      "step": 86
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.588808059692383,
+      "learning_rate": 8.6e-05,
+      "loss": 0.9643,
+      "step": 87
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.284501552581787,
+      "learning_rate": 8.699999999999999e-05,
+      "loss": 1.1781,
+      "step": 88
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.207043647766113,
+      "learning_rate": 8.8e-05,
+      "loss": 1.3532,
+      "step": 89
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.409078359603882,
+      "learning_rate": 8.9e-05,
+      "loss": 1.1546,
+      "step": 90
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.021597385406494,
+      "learning_rate": 8.999999999999999e-05,
+      "loss": 0.8727,
+      "step": 91
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.004073619842529,
+      "learning_rate": 9.099999999999999e-05,
+      "loss": 0.8301,
+      "step": 92
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.221350193023682,
+      "learning_rate": 9.199999999999999e-05,
+      "loss": 1.1282,
+      "step": 93
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.712344646453857,
+      "learning_rate": 9.3e-05,
+      "loss": 1.5645,
+      "step": 94
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.151594638824463,
+      "learning_rate": 9.4e-05,
+      "loss": 1.1636,
+      "step": 95
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.9005351066589355,
+      "learning_rate": 9.499999999999999e-05,
+      "loss": 0.9978,
+      "step": 96
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.770527839660645,
+      "learning_rate": 9.599999999999999e-05,
+      "loss": 0.7176,
+      "step": 97
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.691393852233887,
+      "learning_rate": 9.699999999999999e-05,
+      "loss": 0.9424,
+      "step": 98
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.191307544708252,
+      "learning_rate": 9.799999999999998e-05,
+      "loss": 0.8837,
+      "step": 99
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.723273277282715,
+      "learning_rate": 9.9e-05,
+      "loss": 1.2063,
+      "step": 100
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1191282272338867,
+      "learning_rate": 9.999999999999999e-05,
+      "loss": 0.8371,
+      "step": 101
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.945955276489258,
+      "learning_rate": 0.00010099999999999999,
+      "loss": 0.9069,
+      "step": 102
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 8.733388900756836,
+      "learning_rate": 0.000102,
+      "loss": 1.319,
+      "step": 103
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.735355854034424,
+      "learning_rate": 0.00010299999999999998,
+      "loss": 1.3797,
+      "step": 104
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.036331653594971,
+      "learning_rate": 0.000104,
+      "loss": 0.5183,
+      "step": 105
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.886319637298584,
+      "learning_rate": 0.00010499999999999999,
+      "loss": 1.5606,
+      "step": 106
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.50234317779541,
+      "learning_rate": 0.00010599999999999999,
+      "loss": 1.3502,
+      "step": 107
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.697140216827393,
+      "learning_rate": 0.000107,
+      "loss": 0.9047,
+      "step": 108
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.2634806632995605,
+      "learning_rate": 0.00010799999999999998,
+      "loss": 1.7172,
+      "step": 109
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.540546894073486,
+      "learning_rate": 0.00010899999999999999,
+      "loss": 1.4783,
+      "step": 110
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.533005714416504,
+      "learning_rate": 0.00010999999999999998,
+      "loss": 1.3668,
+      "step": 111
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.5528564453125,
+      "learning_rate": 0.00011099999999999999,
+      "loss": 1.6343,
+      "step": 112
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.9506545066833496,
+      "learning_rate": 0.000112,
+      "loss": 0.5561,
+      "step": 113
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.126518726348877,
+      "learning_rate": 0.00011299999999999998,
+      "loss": 1.574,
+      "step": 114
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.16034460067749,
+      "learning_rate": 0.00011399999999999999,
+      "loss": 1.6017,
+      "step": 115
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.636258602142334,
+      "learning_rate": 0.000115,
+      "loss": 0.4447,
+      "step": 116
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.657654762268066,
+      "learning_rate": 0.00011599999999999999,
+      "loss": 1.683,
+      "step": 117
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.899839401245117,
+      "learning_rate": 0.000117,
+      "loss": 1.5159,
+      "step": 118
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.1795125007629395,
+      "learning_rate": 0.00011799999999999998,
+      "loss": 0.984,
+      "step": 119
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.968451976776123,
+      "learning_rate": 0.00011899999999999999,
+      "loss": 1.6149,
+      "step": 120
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.363707542419434,
+      "learning_rate": 0.00011999999999999999,
+      "loss": 1.68,
+      "step": 121
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.613768577575684,
+      "learning_rate": 0.00012099999999999999,
+      "loss": 0.9822,
+      "step": 122
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9544668197631836,
+      "learning_rate": 0.000122,
+      "loss": 1.1721,
+      "step": 123
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.6496357917785645,
+      "learning_rate": 0.00012299999999999998,
+      "loss": 1.0635,
+      "step": 124
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.220294952392578,
+      "learning_rate": 0.00012399999999999998,
+      "loss": 0.3894,
+      "step": 125
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.641209602355957,
+      "learning_rate": 0.000125,
+      "loss": 1.2102,
+      "step": 126
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1986618041992188,
+      "learning_rate": 0.00012599999999999997,
+      "loss": 0.32,
+      "step": 127
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.084183931350708,
+      "learning_rate": 0.000127,
+      "loss": 1.5444,
+      "step": 128
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7426083087921143,
+      "learning_rate": 0.000128,
+      "loss": 1.2618,
+      "step": 129
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.764894008636475,
+      "learning_rate": 0.000129,
+      "loss": 1.2171,
+      "step": 130
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.4803237915039062,
+      "learning_rate": 0.00013,
+      "loss": 0.3114,
+      "step": 131
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9172024726867676,
+      "learning_rate": 0.00013099999999999999,
+      "loss": 0.4047,
+      "step": 132
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.119025468826294,
+      "learning_rate": 0.00013199999999999998,
+      "loss": 1.1797,
+      "step": 133
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.320390701293945,
+      "learning_rate": 0.000133,
+      "loss": 1.2692,
+      "step": 134
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.118578910827637,
+      "learning_rate": 0.00013399999999999998,
+      "loss": 1.579,
+      "step": 135
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.933561086654663,
+      "learning_rate": 0.000135,
+      "loss": 1.1655,
+      "step": 136
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4326107501983643,
+      "learning_rate": 0.00013599999999999997,
+      "loss": 1.3859,
+      "step": 137
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.371403217315674,
+      "learning_rate": 0.000137,
+      "loss": 1.2995,
+      "step": 138
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.300549507141113,
+      "learning_rate": 0.000138,
+      "loss": 1.6417,
+      "step": 139
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.252517223358154,
+      "learning_rate": 0.000139,
+      "loss": 1.1399,
+      "step": 140
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.720141410827637,
+      "learning_rate": 0.00014,
+      "loss": 1.935,
+      "step": 141
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.964791297912598,
+      "learning_rate": 0.00014099999999999998,
+      "loss": 0.9864,
+      "step": 142
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.073893070220947,
+      "learning_rate": 0.00014199999999999998,
+      "loss": 0.3407,
+      "step": 143
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5245461463928223,
+      "learning_rate": 0.00014299999999999998,
+      "loss": 1.1783,
+      "step": 144
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.986094951629639,
+      "learning_rate": 0.00014399999999999998,
+      "loss": 1.0759,
+      "step": 145
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.874466896057129,
+      "learning_rate": 0.000145,
+      "loss": 1.5019,
+      "step": 146
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.829094409942627,
+      "learning_rate": 0.000146,
+      "loss": 1.8141,
+      "step": 147
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.950973033905029,
+      "learning_rate": 0.000147,
+      "loss": 1.6488,
+      "step": 148
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.917398691177368,
+      "learning_rate": 0.000148,
+      "loss": 0.2216,
+      "step": 149
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.3399293422698975,
+      "learning_rate": 0.000149,
+      "loss": 1.1146,
+      "step": 150
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.7433464527130127,
+      "learning_rate": 0.00015,
+      "loss": 1.1838,
+      "step": 151
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8234732151031494,
+      "learning_rate": 0.00015099999999999998,
+      "loss": 1.7287,
+      "step": 152
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.2081990242004395,
+      "learning_rate": 0.000152,
+      "loss": 1.5875,
+      "step": 153
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.479569673538208,
+      "learning_rate": 0.00015299999999999998,
+      "loss": 1.3674,
+      "step": 154
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.263027191162109,
+      "learning_rate": 0.00015399999999999998,
+      "loss": 1.0861,
+      "step": 155
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.20069694519043,
+      "learning_rate": 0.000155,
+      "loss": 0.8604,
+      "step": 156
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7267985343933105,
+      "learning_rate": 0.000156,
+      "loss": 1.6038,
+      "step": 157
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.7871580123901367,
+      "learning_rate": 0.000157,
+      "loss": 1.0676,
+      "step": 158
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.6624743938446045,
+      "learning_rate": 0.00015799999999999996,
+      "loss": 0.9559,
+      "step": 159
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.26881742477417,
+      "learning_rate": 0.000159,
+      "loss": 1.6486,
+      "step": 160
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.106972694396973,
+      "learning_rate": 0.00015999999999999999,
+      "loss": 1.0407,
+      "step": 161
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 12.234991073608398,
+      "learning_rate": 0.00016099999999999998,
+      "loss": 0.788,
+      "step": 162
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.54979133605957,
+      "learning_rate": 0.000162,
+      "loss": 1.2009,
+      "step": 163
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.871726036071777,
+      "learning_rate": 0.00016299999999999998,
+      "loss": 1.3877,
+      "step": 164
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.019041538238525,
+      "learning_rate": 0.00016399999999999997,
+      "loss": 1.5193,
+      "step": 165
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.0723676681518555,
+      "learning_rate": 0.000165,
+      "loss": 1.2119,
+      "step": 166
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.5908708572387695,
+      "learning_rate": 0.000166,
+      "loss": 0.9027,
+      "step": 167
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.6804304122924805,
+      "learning_rate": 0.00016699999999999997,
+      "loss": 0.7778,
+      "step": 168
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.805787563323975,
+      "learning_rate": 0.000168,
+      "loss": 0.2573,
+      "step": 169
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.424121141433716,
+      "learning_rate": 0.000169,
+      "loss": 0.9506,
+      "step": 170
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.17886209487915,
+      "learning_rate": 0.00016999999999999999,
+      "loss": 1.0797,
+      "step": 171
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.611898899078369,
+      "learning_rate": 0.00017099999999999998,
+      "loss": 0.7596,
+      "step": 172
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.827244281768799,
+      "learning_rate": 0.000172,
+      "loss": 1.5111,
+      "step": 173
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.5306131839752197,
+      "learning_rate": 0.00017299999999999998,
+      "loss": 1.7616,
+      "step": 174
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.013223171234131,
+      "learning_rate": 0.00017399999999999997,
+      "loss": 1.2749,
+      "step": 175
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.9415531158447266,
+      "learning_rate": 0.000175,
+      "loss": 1.3843,
+      "step": 176
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.582239151000977,
+      "learning_rate": 0.000176,
+      "loss": 1.2561,
+      "step": 177
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.296367645263672,
+      "learning_rate": 0.00017699999999999997,
+      "loss": 1.1076,
+      "step": 178
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.854572296142578,
+      "learning_rate": 0.000178,
+      "loss": 1.3671,
+      "step": 179
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.11190128326416,
+      "learning_rate": 0.000179,
+      "loss": 1.1844,
+      "step": 180
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.326961040496826,
+      "learning_rate": 0.00017999999999999998,
+      "loss": 0.9527,
+      "step": 181
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.8623571395874023,
+      "learning_rate": 0.000181,
+      "loss": 0.8668,
+      "step": 182
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3031466007232666,
+      "learning_rate": 0.00018199999999999998,
+      "loss": 1.6397,
+      "step": 183
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3300487995147705,
+      "learning_rate": 0.00018299999999999998,
+      "loss": 1.5656,
+      "step": 184
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.852734088897705,
+      "learning_rate": 0.00018399999999999997,
+      "loss": 1.0274,
+      "step": 185
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.535395860671997,
+      "learning_rate": 0.000185,
+      "loss": 0.9264,
+      "step": 186
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.4967265129089355,
+      "learning_rate": 0.000186,
+      "loss": 0.7599,
+      "step": 187
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.382897138595581,
+      "learning_rate": 0.00018699999999999996,
+      "loss": 0.7655,
+      "step": 188
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.327773094177246,
+      "learning_rate": 0.000188,
+      "loss": 0.2743,
+      "step": 189
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.7694613933563232,
+      "learning_rate": 0.00018899999999999999,
+      "loss": 1.0972,
+      "step": 190
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.895078182220459,
+      "learning_rate": 0.00018999999999999998,
+      "loss": 0.5177,
+      "step": 191
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.051892280578613,
+      "learning_rate": 0.000191,
+      "loss": 1.4296,
+      "step": 192
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4651572704315186,
+      "learning_rate": 0.00019199999999999998,
+      "loss": 0.2123,
+      "step": 193
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.485482692718506,
+      "learning_rate": 0.00019299999999999997,
+      "loss": 1.8226,
+      "step": 194
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.438916206359863,
+      "learning_rate": 0.00019399999999999997,
+      "loss": 0.7194,
+      "step": 195
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.396063804626465,
+      "learning_rate": 0.000195,
+      "loss": 1.6922,
+      "step": 196
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.508432388305664,
+      "learning_rate": 0.00019599999999999997,
+      "loss": 1.6448,
+      "step": 197
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.487405776977539,
+      "learning_rate": 0.00019699999999999996,
+      "loss": 1.1225,
+      "step": 198
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8554043769836426,
+      "learning_rate": 0.000198,
+      "loss": 1.4758,
+      "step": 199
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.9215590953826904,
+      "learning_rate": 0.00019899999999999999,
+      "loss": 0.2644,
+      "step": 200
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.809658408164978,
+      "learning_rate": 0.00019999999999999998,
+      "loss": 0.6342,
+      "step": 201
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.948272705078125,
+      "learning_rate": 0.000201,
+      "loss": 1.3171,
+      "step": 202
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.659675598144531,
+      "learning_rate": 0.00020199999999999998,
+      "loss": 1.3682,
+      "step": 203
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1061782836914062,
+      "learning_rate": 0.00020299999999999997,
+      "loss": 0.6586,
+      "step": 204
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.067229270935059,
+      "learning_rate": 0.000204,
+      "loss": 1.4402,
+      "step": 205
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.9608683586120605,
+      "learning_rate": 0.000205,
+      "loss": 1.5721,
+      "step": 206
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.231966972351074,
+      "learning_rate": 0.00020599999999999997,
+      "loss": 1.4125,
+      "step": 207
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.415351629257202,
+      "learning_rate": 0.00020699999999999996,
+      "loss": 1.234,
+      "step": 208
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.1476757526397705,
+      "learning_rate": 0.000208,
+      "loss": 0.8833,
+      "step": 209
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.153148651123047,
+      "learning_rate": 0.00020899999999999998,
+      "loss": 1.5821,
+      "step": 210
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.2117390632629395,
+      "learning_rate": 0.00020999999999999998,
+      "loss": 1.1968,
+      "step": 211
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.106515884399414,
+      "learning_rate": 0.00021099999999999998,
+      "loss": 1.1919,
+      "step": 212
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.5760037899017334,
+      "learning_rate": 0.00021199999999999998,
+      "loss": 1.6354,
+      "step": 213
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4257736206054688,
+      "learning_rate": 0.00021299999999999997,
+      "loss": 0.6762,
+      "step": 214
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.381577968597412,
+      "learning_rate": 0.000214,
+      "loss": 1.3986,
+      "step": 215
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.177124500274658,
+      "learning_rate": 0.000215,
+      "loss": 0.9631,
+      "step": 216
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.468573570251465,
+      "learning_rate": 0.00021599999999999996,
+      "loss": 1.4267,
+      "step": 217
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.499484062194824,
+      "learning_rate": 0.000217,
+      "loss": 1.0607,
+      "step": 218
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8307218551635742,
+      "learning_rate": 0.00021799999999999999,
+      "loss": 0.5877,
+      "step": 219
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.825871229171753,
+      "learning_rate": 0.00021899999999999998,
+      "loss": 0.6927,
+      "step": 220
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.552651047706604,
+      "learning_rate": 0.00021999999999999995,
+      "loss": 0.2026,
+      "step": 221
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.360166311264038,
+      "learning_rate": 0.00022099999999999998,
+      "loss": 1.4072,
+      "step": 222
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.88201642036438,
+      "learning_rate": 0.00022199999999999998,
+      "loss": 0.6932,
+      "step": 223
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.5754194259643555,
+      "learning_rate": 0.00022299999999999997,
+      "loss": 1.1474,
+      "step": 224
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.174070835113525,
+      "learning_rate": 0.000224,
+      "loss": 1.6259,
+      "step": 225
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.317915916442871,
+      "learning_rate": 0.000225,
+      "loss": 1.5691,
+      "step": 226
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.2685751914978027,
+      "learning_rate": 0.00022599999999999996,
+      "loss": 0.7069,
+      "step": 227
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4414174556732178,
+      "learning_rate": 0.000227,
+      "loss": 1.4426,
+      "step": 228
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.9098880290985107,
+      "learning_rate": 0.00022799999999999999,
+      "loss": 1.4848,
+      "step": 229
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.94317364692688,
+      "learning_rate": 0.00022899999999999998,
+      "loss": 1.5398,
+      "step": 230
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.696044445037842,
+      "learning_rate": 0.00023,
+      "loss": 1.0508,
+      "step": 231
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.5637283325195312,
+      "learning_rate": 0.00023099999999999998,
+      "loss": 1.293,
+      "step": 232
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.252113342285156,
+      "learning_rate": 0.00023199999999999997,
+      "loss": 1.0238,
+      "step": 233
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.8909826278686523,
+      "learning_rate": 0.00023299999999999997,
+      "loss": 0.8286,
+      "step": 234
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0664350986480713,
+      "learning_rate": 0.000234,
+      "loss": 1.049,
+      "step": 235
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.636462926864624,
+      "learning_rate": 0.00023499999999999997,
+      "loss": 1.3311,
+      "step": 236
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.8265254497528076,
+      "learning_rate": 0.00023599999999999996,
+      "loss": 1.0761,
+      "step": 237
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.1792984008789062,
+      "learning_rate": 0.000237,
+      "loss": 1.143,
+      "step": 238
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.2147650718688965,
+      "learning_rate": 0.00023799999999999998,
+      "loss": 1.0399,
+      "step": 239
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9508872032165527,
+      "learning_rate": 0.00023899999999999998,
+      "loss": 1.5802,
+      "step": 240
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6398966312408447,
+      "learning_rate": 0.00023999999999999998,
+      "loss": 1.2396,
+      "step": 241
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.227956771850586,
+      "learning_rate": 0.00024099999999999998,
+      "loss": 0.2776,
+      "step": 242
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.4774301052093506,
+      "learning_rate": 0.00024199999999999997,
+      "loss": 0.9155,
+      "step": 243
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.6463623046875,
+      "learning_rate": 0.000243,
+      "loss": 1.5446,
+      "step": 244
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.357964277267456,
+      "learning_rate": 0.000244,
+      "loss": 0.4428,
+      "step": 245
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.01023268699646,
+      "learning_rate": 0.000245,
+      "loss": 1.0256,
+      "step": 246
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.485054016113281,
+      "learning_rate": 0.00024599999999999996,
+      "loss": 0.9325,
+      "step": 247
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.258424758911133,
+      "learning_rate": 0.000247,
+      "loss": 1.1518,
+      "step": 248
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.5742106437683105,
+      "learning_rate": 0.00024799999999999996,
+      "loss": 0.9365,
+      "step": 249
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.131269931793213,
+      "learning_rate": 0.000249,
+      "loss": 1.006,
+      "step": 250
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8377223014831543,
+      "learning_rate": 0.00025,
+      "loss": 0.2142,
+      "step": 251
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.699740886688232,
+      "learning_rate": 0.000251,
+      "loss": 1.2012,
+      "step": 252
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.748816967010498,
+      "learning_rate": 0.00025199999999999995,
+      "loss": 1.2162,
+      "step": 253
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5976569652557373,
+      "learning_rate": 0.00025299999999999997,
+      "loss": 0.7342,
+      "step": 254
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.2964205741882324,
+      "learning_rate": 0.000254,
+      "loss": 0.6944,
+      "step": 255
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2986418008804321,
+      "learning_rate": 0.00025499999999999996,
+      "loss": 0.1366,
+      "step": 256
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.5206665992736816,
+      "learning_rate": 0.000256,
+      "loss": 0.8447,
+      "step": 257
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.548757791519165,
+      "learning_rate": 0.00025699999999999996,
+      "loss": 0.9743,
+      "step": 258
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.173283576965332,
+      "learning_rate": 0.000258,
+      "loss": 1.0704,
+      "step": 259
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.844398260116577,
+      "learning_rate": 0.00025899999999999995,
+      "loss": 0.9054,
+      "step": 260
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.233757495880127,
+      "learning_rate": 0.00026,
+      "loss": 1.0035,
+      "step": 261
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.585799217224121,
+      "learning_rate": 0.000261,
+      "loss": 1.0935,
+      "step": 262
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3614985942840576,
+      "learning_rate": 0.00026199999999999997,
+      "loss": 1.1227,
+      "step": 263
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.9490151405334473,
+      "learning_rate": 0.000263,
+      "loss": 1.0003,
+      "step": 264
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.104151725769043,
+      "learning_rate": 0.00026399999999999997,
+      "loss": 0.8303,
+      "step": 265
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2483747005462646,
+      "learning_rate": 0.000265,
+      "loss": 0.2044,
+      "step": 266
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9833176136016846,
+      "learning_rate": 0.000266,
+      "loss": 0.6864,
+      "step": 267
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.925781011581421,
+      "learning_rate": 0.000267,
+      "loss": 1.1738,
+      "step": 268
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0927608013153076,
+      "learning_rate": 0.00026799999999999995,
+      "loss": 0.7891,
+      "step": 269
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4451119899749756,
+      "learning_rate": 0.000269,
+      "loss": 0.7228,
+      "step": 270
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.7132692337036133,
+      "learning_rate": 0.00027,
+      "loss": 0.8993,
+      "step": 271
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.241142988204956,
+      "learning_rate": 0.000271,
+      "loss": 0.1454,
+      "step": 272
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2361118793487549,
+      "learning_rate": 0.00027199999999999994,
+      "loss": 0.2268,
+      "step": 273
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.1823840141296387,
+      "learning_rate": 0.00027299999999999997,
+      "loss": 1.3618,
+      "step": 274
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.52080237865448,
+      "learning_rate": 0.000274,
+      "loss": 0.191,
+      "step": 275
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.636223316192627,
+      "learning_rate": 0.00027499999999999996,
+      "loss": 1.4811,
+      "step": 276
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.443178176879883,
+      "learning_rate": 0.000276,
+      "loss": 1.7303,
+      "step": 277
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1942827701568604,
+      "learning_rate": 0.00027699999999999996,
+      "loss": 0.4678,
+      "step": 278
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.1649677753448486,
+      "learning_rate": 0.000278,
+      "loss": 1.8159,
+      "step": 279
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.49360728263855,
+      "learning_rate": 0.000279,
+      "loss": 1.3272,
+      "step": 280
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.574930429458618,
+      "learning_rate": 0.00028,
+      "loss": 1.1373,
+      "step": 281
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.4578397274017334,
+      "learning_rate": 0.00028099999999999995,
+      "loss": 1.3062,
+      "step": 282
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.2991280555725098,
+      "learning_rate": 0.00028199999999999997,
+      "loss": 1.0804,
+      "step": 283
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.0849409103393555,
+      "learning_rate": 0.000283,
+      "loss": 1.0897,
+      "step": 284
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.3282253742218018,
+      "learning_rate": 0.00028399999999999996,
+      "loss": 0.8288,
+      "step": 285
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.600067615509033,
+      "learning_rate": 0.000285,
+      "loss": 0.8487,
+      "step": 286
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.722736120223999,
+      "learning_rate": 0.00028599999999999996,
+      "loss": 1.7637,
+      "step": 287
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.682899236679077,
+      "learning_rate": 0.000287,
+      "loss": 1.1277,
+      "step": 288
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.891709804534912,
+      "learning_rate": 0.00028799999999999995,
+      "loss": 1.5311,
+      "step": 289
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.917855978012085,
+      "learning_rate": 0.000289,
+      "loss": 0.9244,
+      "step": 290
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.527740240097046,
+      "learning_rate": 0.00029,
+      "loss": 0.8896,
+      "step": 291
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.798820972442627,
+      "learning_rate": 0.00029099999999999997,
+      "loss": 1.4213,
+      "step": 292
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.3912644386291504,
+      "learning_rate": 0.000292,
+      "loss": 0.8839,
+      "step": 293
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.7786941528320312,
+      "learning_rate": 0.00029299999999999997,
+      "loss": 1.4677,
+      "step": 294
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2656240463256836,
+      "learning_rate": 0.000294,
+      "loss": 0.8155,
+      "step": 295
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2138984203338623,
+      "learning_rate": 0.00029499999999999996,
+      "loss": 0.453,
+      "step": 296
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.567823886871338,
+      "learning_rate": 0.000296,
+      "loss": 1.3948,
+      "step": 297
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.424060106277466,
+      "learning_rate": 0.00029699999999999996,
+      "loss": 1.2355,
+      "step": 298
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.3072586059570312,
+      "learning_rate": 0.000298,
+      "loss": 1.5173,
+      "step": 299
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.51914381980896,
+      "learning_rate": 0.000299,
+      "loss": 1.565,
+      "step": 300
+    },
+    {
+      "epoch": 0.31,
+      "eval_loss": 1.2794907093048096,
+      "eval_runtime": 1.0128,
+      "eval_samples_per_second": 4.937,
+      "eval_steps_per_second": 0.987,
+      "step": 300
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.966632843017578,
+      "learning_rate": 0.0003,
+      "loss": 1.4345,
+      "step": 301
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.790022850036621,
+      "learning_rate": 0.0002998831775700934,
+      "loss": 1.4484,
+      "step": 302
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.3552589416503906,
+      "learning_rate": 0.0002997663551401869,
+      "loss": 2.0556,
+      "step": 303
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.0574560165405273,
+      "learning_rate": 0.0002996495327102804,
+      "loss": 0.523,
+      "step": 304
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.8360860347747803,
+      "learning_rate": 0.00029953271028037383,
+      "loss": 1.4312,
+      "step": 305
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2051467895507812,
+      "learning_rate": 0.0002994158878504672,
+      "loss": 1.1874,
+      "step": 306
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.811757564544678,
+      "learning_rate": 0.00029929906542056073,
+      "loss": 1.2609,
+      "step": 307
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9196655750274658,
+      "learning_rate": 0.0002991822429906542,
+      "loss": 0.3526,
+      "step": 308
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.7059614658355713,
+      "learning_rate": 0.00029906542056074763,
+      "loss": 1.0694,
+      "step": 309
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.548222780227661,
+      "learning_rate": 0.0002989485981308411,
+      "loss": 1.4761,
+      "step": 310
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.1029458045959473,
+      "learning_rate": 0.0002988317757009346,
+      "loss": 1.0938,
+      "step": 311
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.750783920288086,
+      "learning_rate": 0.00029871495327102803,
+      "loss": 1.5409,
+      "step": 312
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.425215244293213,
+      "learning_rate": 0.0002985981308411215,
+      "loss": 1.2011,
+      "step": 313
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.5072877407073975,
+      "learning_rate": 0.00029848130841121493,
+      "loss": 0.8516,
+      "step": 314
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0297932624816895,
+      "learning_rate": 0.0002983644859813084,
+      "loss": 1.428,
+      "step": 315
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.4369113445281982,
+      "learning_rate": 0.00029824766355140183,
+      "loss": 1.5548,
+      "step": 316
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.450639247894287,
+      "learning_rate": 0.0002981308411214953,
+      "loss": 0.7257,
+      "step": 317
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8958454132080078,
+      "learning_rate": 0.0002980140186915888,
+      "loss": 1.743,
+      "step": 318
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.61169695854187,
+      "learning_rate": 0.00029789719626168224,
+      "loss": 1.2043,
+      "step": 319
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2103683948516846,
+      "learning_rate": 0.0002977803738317757,
+      "loss": 0.2433,
+      "step": 320
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.215791940689087,
+      "learning_rate": 0.00029766355140186914,
+      "loss": 0.5894,
+      "step": 321
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2916409969329834,
+      "learning_rate": 0.0002975467289719626,
+      "loss": 0.7389,
+      "step": 322
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.915350914001465,
+      "learning_rate": 0.00029742990654205604,
+      "loss": 1.776,
+      "step": 323
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9768015146255493,
+      "learning_rate": 0.0002973130841121495,
+      "loss": 0.5903,
+      "step": 324
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9784126281738281,
+      "learning_rate": 0.00029719626168224294,
+      "loss": 1.3178,
+      "step": 325
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2313179969787598,
+      "learning_rate": 0.00029707943925233644,
+      "loss": 1.4696,
+      "step": 326
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.7534449100494385,
+      "learning_rate": 0.0002969626168224299,
+      "loss": 1.0521,
+      "step": 327
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.8486061096191406,
+      "learning_rate": 0.00029684579439252334,
+      "loss": 1.1824,
+      "step": 328
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.505977153778076,
+      "learning_rate": 0.0002967289719626168,
+      "loss": 0.6442,
+      "step": 329
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.27866268157959,
+      "learning_rate": 0.00029661214953271024,
+      "loss": 1.5632,
+      "step": 330
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.004122495651245,
+      "learning_rate": 0.0002964953271028037,
+      "loss": 0.8333,
+      "step": 331
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0221564769744873,
+      "learning_rate": 0.00029637850467289714,
+      "loss": 0.2884,
+      "step": 332
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.4967293739318848,
+      "learning_rate": 0.00029626168224299065,
+      "loss": 1.805,
+      "step": 333
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.704538822174072,
+      "learning_rate": 0.0002961448598130841,
+      "loss": 1.2926,
+      "step": 334
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2548699378967285,
+      "learning_rate": 0.00029602803738317755,
+      "loss": 0.6661,
+      "step": 335
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8257749676704407,
+      "learning_rate": 0.000295911214953271,
+      "loss": 0.1868,
+      "step": 336
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.1990132331848145,
+      "learning_rate": 0.0002957943925233645,
+      "loss": 1.4626,
+      "step": 337
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.2955174446105957,
+      "learning_rate": 0.0002956775700934579,
+      "loss": 1.1332,
+      "step": 338
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0604300498962402,
+      "learning_rate": 0.00029556074766355135,
+      "loss": 0.3651,
+      "step": 339
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.6107492446899414,
+      "learning_rate": 0.00029544392523364485,
+      "loss": 0.8476,
+      "step": 340
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7882015705108643,
+      "learning_rate": 0.0002953271028037383,
+      "loss": 1.1006,
+      "step": 341
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.7657339572906494,
+      "learning_rate": 0.00029521028037383175,
+      "loss": 1.2806,
+      "step": 342
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8367657661437988,
+      "learning_rate": 0.0002950934579439252,
+      "loss": 0.2574,
+      "step": 343
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.316808223724365,
+      "learning_rate": 0.00029497663551401865,
+      "loss": 0.7867,
+      "step": 344
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.0288803577423096,
+      "learning_rate": 0.00029485981308411215,
+      "loss": 0.4766,
+      "step": 345
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.54671049118042,
+      "learning_rate": 0.00029474299065420555,
+      "loss": 0.824,
+      "step": 346
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.718745231628418,
+      "learning_rate": 0.00029462616822429905,
+      "loss": 1.3466,
+      "step": 347
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.292029857635498,
+      "learning_rate": 0.0002945093457943925,
+      "loss": 0.7201,
+      "step": 348
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1415319442749023,
+      "learning_rate": 0.00029439252336448596,
+      "loss": 1.7006,
+      "step": 349
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0505154132843018,
+      "learning_rate": 0.0002942757009345794,
+      "loss": 1.4469,
+      "step": 350
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2195465564727783,
+      "learning_rate": 0.00029415887850467286,
+      "loss": 1.4332,
+      "step": 351
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.9841485023498535,
+      "learning_rate": 0.00029404205607476636,
+      "loss": 0.7682,
+      "step": 352
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.8377418518066406,
+      "learning_rate": 0.0002939252336448598,
+      "loss": 1.0434,
+      "step": 353
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.081266403198242,
+      "learning_rate": 0.0002938084112149532,
+      "loss": 0.5904,
+      "step": 354
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4761319160461426,
+      "learning_rate": 0.0002936915887850467,
+      "loss": 1.1598,
+      "step": 355
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.700859308242798,
+      "learning_rate": 0.00029357476635514016,
+      "loss": 0.523,
+      "step": 356
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.601006031036377,
+      "learning_rate": 0.0002934579439252336,
+      "loss": 1.567,
+      "step": 357
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2836368083953857,
+      "learning_rate": 0.00029334112149532706,
+      "loss": 0.1971,
+      "step": 358
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.212724208831787,
+      "learning_rate": 0.00029322429906542056,
+      "loss": 0.5909,
+      "step": 359
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.0591113567352295,
+      "learning_rate": 0.000293107476635514,
+      "loss": 0.939,
+      "step": 360
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9243104457855225,
+      "learning_rate": 0.00029299065420560746,
+      "loss": 0.3335,
+      "step": 361
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6790910363197327,
+      "learning_rate": 0.0002928738317757009,
+      "loss": 0.2015,
+      "step": 362
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.168013095855713,
+      "learning_rate": 0.00029275700934579436,
+      "loss": 0.404,
+      "step": 363
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2125356197357178,
+      "learning_rate": 0.0002926401869158878,
+      "loss": 1.6901,
+      "step": 364
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8769881725311279,
+      "learning_rate": 0.00029252336448598126,
+      "loss": 0.4032,
+      "step": 365
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.276057243347168,
+      "learning_rate": 0.00029240654205607477,
+      "loss": 0.9828,
+      "step": 366
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7380950450897217,
+      "learning_rate": 0.0002922897196261682,
+      "loss": 1.5195,
+      "step": 367
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8719393610954285,
+      "learning_rate": 0.00029217289719626167,
+      "loss": 0.2892,
+      "step": 368
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1746935844421387,
+      "learning_rate": 0.0002920560747663551,
+      "loss": 0.1869,
+      "step": 369
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2664616107940674,
+      "learning_rate": 0.00029193925233644857,
+      "loss": 0.9131,
+      "step": 370
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.618701934814453,
+      "learning_rate": 0.000291822429906542,
+      "loss": 1.2094,
+      "step": 371
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.9570870399475098,
+      "learning_rate": 0.00029170560747663547,
+      "loss": 0.722,
+      "step": 372
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.253892183303833,
+      "learning_rate": 0.0002915887850467289,
+      "loss": 1.4645,
+      "step": 373
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.9197115898132324,
+      "learning_rate": 0.0002914719626168224,
+      "loss": 0.6279,
+      "step": 374
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.723787546157837,
+      "learning_rate": 0.00029135514018691587,
+      "loss": 0.8615,
+      "step": 375
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6628739833831787,
+      "learning_rate": 0.0002912383177570093,
+      "loss": 0.5986,
+      "step": 376
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0160231590270996,
+      "learning_rate": 0.0002911214953271028,
+      "loss": 1.4512,
+      "step": 377
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.787649631500244,
+      "learning_rate": 0.0002910046728971962,
+      "loss": 1.2608,
+      "step": 378
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7951927185058594,
+      "learning_rate": 0.0002908878504672897,
+      "loss": 1.7605,
+      "step": 379
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1774840354919434,
+      "learning_rate": 0.0002907710280373831,
+      "loss": 0.9581,
+      "step": 380
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.9759175777435303,
+      "learning_rate": 0.00029065420560747663,
+      "loss": 1.7762,
+      "step": 381
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0567970275878906,
+      "learning_rate": 0.0002905373831775701,
+      "loss": 0.4799,
+      "step": 382
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9178869128227234,
+      "learning_rate": 0.00029042056074766353,
+      "loss": 0.3698,
+      "step": 383
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.4597392082214355,
+      "learning_rate": 0.000290303738317757,
+      "loss": 1.5229,
+      "step": 384
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4791572093963623,
+      "learning_rate": 0.0002901869158878505,
+      "loss": 1.6204,
+      "step": 385
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.3275704383850098,
+      "learning_rate": 0.0002900700934579439,
+      "loss": 0.9628,
+      "step": 386
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1707335710525513,
+      "learning_rate": 0.00028995327102803733,
+      "loss": 0.7267,
+      "step": 387
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.8721630573272705,
+      "learning_rate": 0.00028983644859813083,
+      "loss": 1.1341,
+      "step": 388
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.7177517414093018,
+      "learning_rate": 0.0002897196261682243,
+      "loss": 1.4855,
+      "step": 389
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.316612720489502,
+      "learning_rate": 0.00028960280373831773,
+      "loss": 1.005,
+      "step": 390
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4785094261169434,
+      "learning_rate": 0.0002894859813084112,
+      "loss": 0.6708,
+      "step": 391
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6307642459869385,
+      "learning_rate": 0.00028936915887850463,
+      "loss": 0.8549,
+      "step": 392
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2438905239105225,
+      "learning_rate": 0.00028925233644859814,
+      "loss": 1.6076,
+      "step": 393
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3739724159240723,
+      "learning_rate": 0.00028913551401869153,
+      "loss": 1.12,
+      "step": 394
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.674501657485962,
+      "learning_rate": 0.00028901869158878504,
+      "loss": 0.5076,
+      "step": 395
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6238901615142822,
+      "learning_rate": 0.0002889018691588785,
+      "loss": 0.6966,
+      "step": 396
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.171786785125732,
+      "learning_rate": 0.00028878504672897194,
+      "loss": 1.3563,
+      "step": 397
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.3918046951293945,
+      "learning_rate": 0.0002886682242990654,
+      "loss": 1.4797,
+      "step": 398
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8573514223098755,
+      "learning_rate": 0.00028855140186915884,
+      "loss": 0.6252,
+      "step": 399
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.9090523719787598,
+      "learning_rate": 0.00028843457943925234,
+      "loss": 1.1731,
+      "step": 400
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.797912359237671,
+      "learning_rate": 0.0002883177570093458,
+      "loss": 0.5088,
+      "step": 401
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.8787248134613037,
+      "learning_rate": 0.0002882009345794392,
+      "loss": 1.8064,
+      "step": 402
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8502628803253174,
+      "learning_rate": 0.0002880841121495327,
+      "loss": 0.4324,
+      "step": 403
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2849178314208984,
+      "learning_rate": 0.00028796728971962614,
+      "loss": 1.2576,
+      "step": 404
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.455745220184326,
+      "learning_rate": 0.0002878504672897196,
+      "loss": 1.4212,
+      "step": 405
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.977170944213867,
+      "learning_rate": 0.00028773364485981304,
+      "loss": 0.9915,
+      "step": 406
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.004940032958984,
+      "learning_rate": 0.00028761682242990655,
+      "loss": 0.8094,
+      "step": 407
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.568225383758545,
+      "learning_rate": 0.0002875,
+      "loss": 0.8189,
+      "step": 408
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1345443725585938,
+      "learning_rate": 0.00028738317757009345,
+      "loss": 1.0812,
+      "step": 409
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8404977321624756,
+      "learning_rate": 0.0002872663551401869,
+      "loss": 0.5456,
+      "step": 410
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2581323385238647,
+      "learning_rate": 0.00028714953271028035,
+      "loss": 0.2504,
+      "step": 411
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7793455123901367,
+      "learning_rate": 0.0002870327102803738,
+      "loss": 1.1482,
+      "step": 412
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3713812828063965,
+      "learning_rate": 0.00028691588785046725,
+      "loss": 1.1142,
+      "step": 413
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7433762550354004,
+      "learning_rate": 0.00028679906542056075,
+      "loss": 1.1598,
+      "step": 414
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.55407977104187,
+      "learning_rate": 0.0002866822429906542,
+      "loss": 0.5134,
+      "step": 415
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.431687116622925,
+      "learning_rate": 0.00028656542056074765,
+      "loss": 1.0363,
+      "step": 416
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9685442447662354,
+      "learning_rate": 0.0002864485981308411,
+      "loss": 0.1242,
+      "step": 417
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.535361647605896,
+      "learning_rate": 0.00028633177570093455,
+      "loss": 0.735,
+      "step": 418
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.509974002838135,
+      "learning_rate": 0.000286214953271028,
+      "loss": 1.194,
+      "step": 419
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9291192293167114,
+      "learning_rate": 0.00028609813084112145,
+      "loss": 0.4654,
+      "step": 420
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.855079174041748,
+      "learning_rate": 0.0002859813084112149,
+      "loss": 1.4628,
+      "step": 421
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7849321365356445,
+      "learning_rate": 0.0002858644859813084,
+      "loss": 0.9878,
+      "step": 422
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.733687162399292,
+      "learning_rate": 0.00028574766355140185,
+      "loss": 0.2796,
+      "step": 423
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.3665411472320557,
+      "learning_rate": 0.0002856308411214953,
+      "loss": 1.2903,
+      "step": 424
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7777185440063477,
+      "learning_rate": 0.00028551401869158875,
+      "loss": 1.3224,
+      "step": 425
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3624370098114014,
+      "learning_rate": 0.0002853971962616822,
+      "loss": 0.6824,
+      "step": 426
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.8823182582855225,
+      "learning_rate": 0.00028528037383177565,
+      "loss": 1.4465,
+      "step": 427
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.512755870819092,
+      "learning_rate": 0.0002851635514018691,
+      "loss": 0.8171,
+      "step": 428
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.120779275894165,
+      "learning_rate": 0.0002850467289719626,
+      "loss": 0.6632,
+      "step": 429
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4150028228759766,
+      "learning_rate": 0.00028492990654205606,
+      "loss": 0.7185,
+      "step": 430
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5052907466888428,
+      "learning_rate": 0.0002848130841121495,
+      "loss": 0.9646,
+      "step": 431
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.989311933517456,
+      "learning_rate": 0.00028469626168224296,
+      "loss": 1.4964,
+      "step": 432
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0057992935180664,
+      "learning_rate": 0.00028457943925233646,
+      "loss": 0.5575,
+      "step": 433
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1718015670776367,
+      "learning_rate": 0.00028446261682242986,
+      "loss": 0.9122,
+      "step": 434
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.9890103340148926,
+      "learning_rate": 0.0002843457943925233,
+      "loss": 1.196,
+      "step": 435
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.763711452484131,
+      "learning_rate": 0.0002842289719626168,
+      "loss": 1.309,
+      "step": 436
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5950281620025635,
+      "learning_rate": 0.00028411214953271026,
+      "loss": 0.5255,
+      "step": 437
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1492373943328857,
+      "learning_rate": 0.0002839953271028037,
+      "loss": 0.7657,
+      "step": 438
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.520798683166504,
+      "learning_rate": 0.00028387850467289716,
+      "loss": 0.6183,
+      "step": 439
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": NaN,
+      "learning_rate": 0.00028387850467289716,
+      "loss": 1.5529,
+      "step": 440
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.6735353469848633,
+      "learning_rate": 0.0002837616822429906,
+      "loss": 1.1269,
+      "step": 441
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8713912963867188,
+      "learning_rate": 0.0002836448598130841,
+      "loss": 0.5722,
+      "step": 442
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9943745136260986,
+      "learning_rate": 0.0002835280373831775,
+      "loss": 0.3957,
+      "step": 443
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4403250217437744,
+      "learning_rate": 0.000283411214953271,
+      "loss": 0.9147,
+      "step": 444
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.170886993408203,
+      "learning_rate": 0.00028329439252336447,
+      "loss": 0.8155,
+      "step": 445
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1296136379241943,
+      "learning_rate": 0.0002831775700934579,
+      "loss": 1.2639,
+      "step": 446
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.446403980255127,
+      "learning_rate": 0.00028306074766355137,
+      "loss": 0.5806,
+      "step": 447
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.98860764503479,
+      "learning_rate": 0.0002829439252336448,
+      "loss": 1.1213,
+      "step": 448
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.848855972290039,
+      "learning_rate": 0.0002828271028037383,
+      "loss": 0.7091,
+      "step": 449
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.713719367980957,
+      "learning_rate": 0.00028271028037383177,
+      "loss": 0.7886,
+      "step": 450
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9307395219802856,
+      "learning_rate": 0.00028259345794392517,
+      "loss": 1.0976,
+      "step": 451
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1166470050811768,
+      "learning_rate": 0.00028247663551401867,
+      "loss": 0.3523,
+      "step": 452
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9235117435455322,
+      "learning_rate": 0.0002823598130841121,
+      "loss": 0.6344,
+      "step": 453
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.735736846923828,
+      "learning_rate": 0.00028224299065420557,
+      "loss": 1.0573,
+      "step": 454
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.388742446899414,
+      "learning_rate": 0.000282126168224299,
+      "loss": 1.2948,
+      "step": 455
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0969812870025635,
+      "learning_rate": 0.0002820093457943925,
+      "loss": 0.8144,
+      "step": 456
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.877164363861084,
+      "learning_rate": 0.000281892523364486,
+      "loss": 0.5874,
+      "step": 457
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.144896984100342,
+      "learning_rate": 0.0002817757009345794,
+      "loss": 0.9143,
+      "step": 458
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1770153045654297,
+      "learning_rate": 0.0002816588785046729,
+      "loss": 0.8423,
+      "step": 459
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5973880290985107,
+      "learning_rate": 0.0002815420560747663,
+      "loss": 1.3463,
+      "step": 460
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3985776901245117,
+      "learning_rate": 0.0002814252336448598,
+      "loss": 0.8316,
+      "step": 461
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.82443904876709,
+      "learning_rate": 0.0002813084112149532,
+      "loss": 0.7252,
+      "step": 462
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2212815284729004,
+      "learning_rate": 0.00028119158878504673,
+      "loss": 0.8296,
+      "step": 463
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1041300296783447,
+      "learning_rate": 0.0002810747663551402,
+      "loss": 1.4688,
+      "step": 464
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.560612916946411,
+      "learning_rate": 0.00028095794392523363,
+      "loss": 1.0018,
+      "step": 465
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8889777660369873,
+      "learning_rate": 0.0002808411214953271,
+      "loss": 1.0207,
+      "step": 466
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.116854429244995,
+      "learning_rate": 0.00028072429906542053,
+      "loss": 0.8899,
+      "step": 467
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6070220470428467,
+      "learning_rate": 0.000280607476635514,
+      "loss": 0.7297,
+      "step": 468
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0625195503234863,
+      "learning_rate": 0.00028049065420560743,
+      "loss": 0.916,
+      "step": 469
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.634510517120361,
+      "learning_rate": 0.0002803738317757009,
+      "loss": 1.059,
+      "step": 470
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.602274179458618,
+      "learning_rate": 0.0002802570093457944,
+      "loss": 1.4006,
+      "step": 471
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.149019479751587,
+      "learning_rate": 0.00028014018691588784,
+      "loss": 0.6402,
+      "step": 472
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.638101577758789,
+      "learning_rate": 0.0002800233644859813,
+      "loss": 0.301,
+      "step": 473
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0306832790374756,
+      "learning_rate": 0.00027990654205607474,
+      "loss": 1.4859,
+      "step": 474
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.660569667816162,
+      "learning_rate": 0.0002797897196261682,
+      "loss": 0.7307,
+      "step": 475
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.271596908569336,
+      "learning_rate": 0.00027967289719626164,
+      "loss": 1.2726,
+      "step": 476
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.8445537090301514,
+      "learning_rate": 0.0002795560747663551,
+      "loss": 1.0576,
+      "step": 477
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8691243529319763,
+      "learning_rate": 0.0002794392523364486,
+      "loss": 0.2533,
+      "step": 478
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.495535135269165,
+      "learning_rate": 0.00027932242990654204,
+      "loss": 1.0136,
+      "step": 479
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.485879898071289,
+      "learning_rate": 0.0002792056074766355,
+      "loss": 0.964,
+      "step": 480
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.352851629257202,
+      "learning_rate": 0.00027908878504672894,
+      "loss": 0.9433,
+      "step": 481
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3454959392547607,
+      "learning_rate": 0.00027897196261682244,
+      "loss": 0.9644,
+      "step": 482
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.8068840503692627,
+      "learning_rate": 0.00027885514018691584,
+      "loss": 1.4043,
+      "step": 483
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.597325801849365,
+      "learning_rate": 0.0002787383177570093,
+      "loss": 1.2679,
+      "step": 484
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.513976097106934,
+      "learning_rate": 0.0002786214953271028,
+      "loss": 1.3383,
+      "step": 485
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.5806922912597656,
+      "learning_rate": 0.00027850467289719624,
+      "loss": 1.1213,
+      "step": 486
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.425881862640381,
+      "learning_rate": 0.0002783878504672897,
+      "loss": 0.5837,
+      "step": 487
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.552446961402893,
+      "learning_rate": 0.00027827102803738314,
+      "loss": 0.7027,
+      "step": 488
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.1759777069091797,
+      "learning_rate": 0.0002781542056074766,
+      "loss": 0.8814,
+      "step": 489
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7307231426239014,
+      "learning_rate": 0.0002780373831775701,
+      "loss": 0.2867,
+      "step": 490
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0762507915496826,
+      "learning_rate": 0.0002779205607476635,
+      "loss": 0.4642,
+      "step": 491
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5488018989562988,
+      "learning_rate": 0.000277803738317757,
+      "loss": 0.1932,
+      "step": 492
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7389271259307861,
+      "learning_rate": 0.00027768691588785045,
+      "loss": 0.206,
+      "step": 493
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.4336817264556885,
+      "learning_rate": 0.0002775700934579439,
+      "loss": 1.0873,
+      "step": 494
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1345937252044678,
+      "learning_rate": 0.00027745327102803735,
+      "loss": 1.6094,
+      "step": 495
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7259533405303955,
+      "learning_rate": 0.0002773364485981308,
+      "loss": 0.1039,
+      "step": 496
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.417773246765137,
+      "learning_rate": 0.0002772196261682243,
+      "loss": 1.4269,
+      "step": 497
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8780008554458618,
+      "learning_rate": 0.00027710280373831775,
+      "loss": 0.6249,
+      "step": 498
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9605576992034912,
+      "learning_rate": 0.00027698598130841115,
+      "loss": 0.6416,
+      "step": 499
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7016693353652954,
+      "learning_rate": 0.00027686915887850465,
+      "loss": 0.2955,
+      "step": 500
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8871127367019653,
+      "learning_rate": 0.0002767523364485981,
+      "loss": 0.3411,
+      "step": 501
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.952772319316864,
+      "learning_rate": 0.00027663551401869155,
+      "loss": 0.4015,
+      "step": 502
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.231271982192993,
+      "learning_rate": 0.000276518691588785,
+      "loss": 1.223,
+      "step": 503
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.837109088897705,
+      "learning_rate": 0.0002764018691588785,
+      "loss": 0.8736,
+      "step": 504
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.7296323776245117,
+      "learning_rate": 0.00027628504672897196,
+      "loss": 0.568,
+      "step": 505
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1369681358337402,
+      "learning_rate": 0.0002761682242990654,
+      "loss": 0.3843,
+      "step": 506
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.714029312133789,
+      "learning_rate": 0.00027605140186915886,
+      "loss": 0.8895,
+      "step": 507
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.8727917671203613,
+      "learning_rate": 0.0002759345794392523,
+      "loss": 0.6678,
+      "step": 508
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3814454078674316,
+      "learning_rate": 0.00027581775700934576,
+      "loss": 0.4698,
+      "step": 509
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3573546409606934,
+      "learning_rate": 0.0002757009345794392,
+      "loss": 0.775,
+      "step": 510
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1534595489501953,
+      "learning_rate": 0.0002755841121495327,
+      "loss": 0.788,
+      "step": 511
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.0166823863983154,
+      "learning_rate": 0.00027546728971962616,
+      "loss": 0.3544,
+      "step": 512
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.865272879600525,
+      "learning_rate": 0.0002753504672897196,
+      "loss": 0.7045,
+      "step": 513
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.6572630405426025,
+      "learning_rate": 0.00027523364485981306,
+      "loss": 1.1575,
+      "step": 514
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.012207269668579,
+      "learning_rate": 0.0002751168224299065,
+      "loss": 1.0693,
+      "step": 515
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9867669343948364,
+      "learning_rate": 0.00027499999999999996,
+      "loss": 0.4059,
+      "step": 516
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.223353385925293,
+      "learning_rate": 0.0002748831775700934,
+      "loss": 1.9971,
+      "step": 517
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2805993556976318,
+      "learning_rate": 0.00027476635514018686,
+      "loss": 0.508,
+      "step": 518
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3479033708572388,
+      "learning_rate": 0.00027464953271028037,
+      "loss": 0.3285,
+      "step": 519
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.865412950515747,
+      "learning_rate": 0.0002745327102803738,
+      "loss": 1.4067,
+      "step": 520
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.9801173210144043,
+      "learning_rate": 0.00027441588785046727,
+      "loss": 1.2826,
+      "step": 521
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8976773619651794,
+      "learning_rate": 0.0002742990654205607,
+      "loss": 0.3444,
+      "step": 522
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.2174181938171387,
+      "learning_rate": 0.00027418224299065417,
+      "loss": 0.3907,
+      "step": 523
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3877856731414795,
+      "learning_rate": 0.0002740654205607476,
+      "loss": 1.284,
+      "step": 524
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1452724933624268,
+      "learning_rate": 0.00027394859813084107,
+      "loss": 1.33,
+      "step": 525
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.878735065460205,
+      "learning_rate": 0.00027383177570093457,
+      "loss": 1.112,
+      "step": 526
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.091610312461853,
+      "learning_rate": 0.000273714953271028,
+      "loss": 0.3708,
+      "step": 527
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5223939418792725,
+      "learning_rate": 0.00027359813084112147,
+      "loss": 1.6085,
+      "step": 528
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.6348628997802734,
+      "learning_rate": 0.0002734813084112149,
+      "loss": 0.8705,
+      "step": 529
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.8594553470611572,
+      "learning_rate": 0.0002733644859813084,
+      "loss": 0.7917,
+      "step": 530
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1773569583892822,
+      "learning_rate": 0.0002732476635514018,
+      "loss": 0.8118,
+      "step": 531
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.174500942230225,
+      "learning_rate": 0.00027313084112149527,
+      "loss": 0.5933,
+      "step": 532
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.555126428604126,
+      "learning_rate": 0.0002730140186915888,
+      "loss": 0.7971,
+      "step": 533
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.0811665058135986,
+      "learning_rate": 0.0002728971962616822,
+      "loss": 1.4148,
+      "step": 534
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.6521191596984863,
+      "learning_rate": 0.0002727803738317757,
+      "loss": 0.9921,
+      "step": 535
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9518895149230957,
+      "learning_rate": 0.0002726635514018691,
+      "loss": 1.5573,
+      "step": 536
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.6719624996185303,
+      "learning_rate": 0.0002725467289719626,
+      "loss": 0.8119,
+      "step": 537
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9216612577438354,
+      "learning_rate": 0.0002724299065420561,
+      "loss": 1.417,
+      "step": 538
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0820119380950928,
+      "learning_rate": 0.0002723130841121495,
+      "loss": 0.9631,
+      "step": 539
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7720649242401123,
+      "learning_rate": 0.000272196261682243,
+      "loss": 1.3626,
+      "step": 540
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1084446907043457,
+      "learning_rate": 0.00027207943925233643,
+      "loss": 0.8393,
+      "step": 541
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.0460463762283325,
+      "learning_rate": 0.0002719626168224299,
+      "loss": 0.1851,
+      "step": 542
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8795809745788574,
+      "learning_rate": 0.00027184579439252333,
+      "loss": 1.0474,
+      "step": 543
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6006479263305664,
+      "learning_rate": 0.0002717289719626168,
+      "loss": 1.9008,
+      "step": 544
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1522631645202637,
+      "learning_rate": 0.0002716121495327103,
+      "loss": 1.2351,
+      "step": 545
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7561458349227905,
+      "learning_rate": 0.00027149532710280373,
+      "loss": 0.8284,
+      "step": 546
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9102766513824463,
+      "learning_rate": 0.00027137850467289713,
+      "loss": 0.7709,
+      "step": 547
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.9740517139434814,
+      "learning_rate": 0.00027126168224299063,
+      "loss": 0.9322,
+      "step": 548
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.9720640182495117,
+      "learning_rate": 0.0002711448598130841,
+      "loss": 0.5249,
+      "step": 549
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.711076259613037,
+      "learning_rate": 0.00027102803738317753,
+      "loss": 0.8303,
+      "step": 550
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2298800945281982,
+      "learning_rate": 0.000270911214953271,
+      "loss": 0.1259,
+      "step": 551
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.824721336364746,
+      "learning_rate": 0.0002707943925233645,
+      "loss": 1.2054,
+      "step": 552
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.455997943878174,
+      "learning_rate": 0.00027067757009345794,
+      "loss": 0.7108,
+      "step": 553
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.940439224243164,
+      "learning_rate": 0.0002705607476635514,
+      "loss": 1.4464,
+      "step": 554
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.046644926071167,
+      "learning_rate": 0.00027044392523364484,
+      "loss": 0.3161,
+      "step": 555
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.7850840091705322,
+      "learning_rate": 0.0002703271028037383,
+      "loss": 0.4476,
+      "step": 556
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6609066724777222,
+      "learning_rate": 0.00027021028037383174,
+      "loss": 0.186,
+      "step": 557
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8339734077453613,
+      "learning_rate": 0.0002700934579439252,
+      "loss": 1.0351,
+      "step": 558
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.947851836681366,
+      "learning_rate": 0.0002699766355140187,
+      "loss": 0.4941,
+      "step": 559
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7194287776947021,
+      "learning_rate": 0.00026985981308411214,
+      "loss": 0.3142,
+      "step": 560
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0309903621673584,
+      "learning_rate": 0.0002697429906542056,
+      "loss": 0.9159,
+      "step": 561
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7473992109298706,
+      "learning_rate": 0.00026962616822429904,
+      "loss": 0.3141,
+      "step": 562
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.604888916015625,
+      "learning_rate": 0.0002695093457943925,
+      "loss": 0.2259,
+      "step": 563
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9733377695083618,
+      "learning_rate": 0.00026939252336448594,
+      "loss": 0.6048,
+      "step": 564
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3935818672180176,
+      "learning_rate": 0.0002692757009345794,
+      "loss": 1.1154,
+      "step": 565
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5244064331054688,
+      "learning_rate": 0.00026915887850467284,
+      "loss": 1.4576,
+      "step": 566
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2174429893493652,
+      "learning_rate": 0.00026904205607476635,
+      "loss": 1.3647,
+      "step": 567
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.750571370124817,
+      "learning_rate": 0.0002689252336448598,
+      "loss": 0.7841,
+      "step": 568
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.9296715259552,
+      "learning_rate": 0.00026880841121495325,
+      "loss": 0.8488,
+      "step": 569
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.912085056304932,
+      "learning_rate": 0.0002686915887850467,
+      "loss": 1.8655,
+      "step": 570
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.35219407081604,
+      "learning_rate": 0.00026857476635514015,
+      "loss": 0.6427,
+      "step": 571
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.13024640083313,
+      "learning_rate": 0.0002684579439252336,
+      "loss": 1.5838,
+      "step": 572
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.217462182044983,
+      "learning_rate": 0.00026834112149532705,
+      "loss": 0.3112,
+      "step": 573
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.472783088684082,
+      "learning_rate": 0.00026822429906542055,
+      "loss": 0.4344,
+      "step": 574
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.9528167247772217,
+      "learning_rate": 0.000268107476635514,
+      "loss": 0.8929,
+      "step": 575
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.462442398071289,
+      "learning_rate": 0.00026799065420560745,
+      "loss": 1.1492,
+      "step": 576
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3671787977218628,
+      "learning_rate": 0.0002678738317757009,
+      "loss": 0.3359,
+      "step": 577
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.802870750427246,
+      "learning_rate": 0.0002677570093457944,
+      "loss": 1.2908,
+      "step": 578
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.338261127471924,
+      "learning_rate": 0.0002676401869158878,
+      "loss": 0.8661,
+      "step": 579
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6252553462982178,
+      "learning_rate": 0.00026752336448598125,
+      "loss": 1.606,
+      "step": 580
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6777186989784241,
+      "learning_rate": 0.00026740654205607476,
+      "loss": 0.2579,
+      "step": 581
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.9596633315086365,
+      "learning_rate": 0.0002672897196261682,
+      "loss": 0.462,
+      "step": 582
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.5880768299102783,
+      "learning_rate": 0.00026717289719626166,
+      "loss": 0.9047,
+      "step": 583
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7881953716278076,
+      "learning_rate": 0.0002670560747663551,
+      "loss": 1.3014,
+      "step": 584
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.6334471702575684,
+      "learning_rate": 0.00026693925233644856,
+      "loss": 0.4504,
+      "step": 585
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4941344261169434,
+      "learning_rate": 0.00026682242990654206,
+      "loss": 0.6579,
+      "step": 586
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.56283438205719,
+      "learning_rate": 0.00026670560747663546,
+      "loss": 1.3118,
+      "step": 587
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.000277280807495,
+      "learning_rate": 0.00026658878504672896,
+      "loss": 0.7094,
+      "step": 588
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.8379180431365967,
+      "learning_rate": 0.0002664719626168224,
+      "loss": 0.9054,
+      "step": 589
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7205324172973633,
+      "learning_rate": 0.00026635514018691586,
+      "loss": 0.751,
+      "step": 590
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9362393617630005,
+      "learning_rate": 0.0002662383177570093,
+      "loss": 0.4933,
+      "step": 591
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.944587230682373,
+      "learning_rate": 0.00026612149532710276,
+      "loss": 0.3332,
+      "step": 592
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3814319372177124,
+      "learning_rate": 0.00026600467289719627,
+      "loss": 0.4593,
+      "step": 593
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5680646896362305,
+      "learning_rate": 0.0002658878504672897,
+      "loss": 1.4059,
+      "step": 594
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6818034052848816,
+      "learning_rate": 0.0002657710280373831,
+      "loss": 0.1153,
+      "step": 595
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9206401109695435,
+      "learning_rate": 0.0002656542056074766,
+      "loss": 1.141,
+      "step": 596
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1888787746429443,
+      "learning_rate": 0.00026553738317757007,
+      "loss": 1.0324,
+      "step": 597
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.396777629852295,
+      "learning_rate": 0.0002654205607476635,
+      "loss": 0.9237,
+      "step": 598
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.9854557514190674,
+      "learning_rate": 0.00026530373831775697,
+      "loss": 1.5611,
+      "step": 599
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.558633804321289,
+      "learning_rate": 0.00026518691588785047,
+      "loss": 0.6527,
+      "step": 600
+    },
+    {
+      "epoch": 0.63,
+      "eval_loss": 1.24161958694458,
+      "eval_runtime": 1.0124,
+      "eval_samples_per_second": 4.939,
+      "eval_steps_per_second": 0.988,
+      "step": 600
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5544018745422363,
+      "learning_rate": 0.0002650700934579439,
+      "loss": 0.2054,
+      "step": 601
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5441477298736572,
+      "learning_rate": 0.00026495327102803737,
+      "loss": 0.5178,
+      "step": 602
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9024848937988281,
+      "learning_rate": 0.0002648364485981308,
+      "loss": 1.1805,
+      "step": 603
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5047481060028076,
+      "learning_rate": 0.00026471962616822427,
+      "loss": 0.4179,
+      "step": 604
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.946559190750122,
+      "learning_rate": 0.0002646028037383177,
+      "loss": 0.6404,
+      "step": 605
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.839164972305298,
+      "learning_rate": 0.00026448598130841117,
+      "loss": 1.4949,
+      "step": 606
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.933732509613037,
+      "learning_rate": 0.0002643691588785047,
+      "loss": 1.1352,
+      "step": 607
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.644740581512451,
+      "learning_rate": 0.0002642523364485981,
+      "loss": 0.9017,
+      "step": 608
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.920925140380859,
+      "learning_rate": 0.0002641355140186916,
+      "loss": 0.9778,
+      "step": 609
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.580047369003296,
+      "learning_rate": 0.000264018691588785,
+      "loss": 0.3734,
+      "step": 610
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5725473165512085,
+      "learning_rate": 0.0002639018691588785,
+      "loss": 0.4728,
+      "step": 611
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.36690092086792,
+      "learning_rate": 0.0002637850467289719,
+      "loss": 0.8096,
+      "step": 612
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.609687328338623,
+      "learning_rate": 0.0002636682242990654,
+      "loss": 1.2741,
+      "step": 613
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9025781154632568,
+      "learning_rate": 0.0002635514018691588,
+      "loss": 1.2263,
+      "step": 614
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.570065975189209,
+      "learning_rate": 0.00026343457943925233,
+      "loss": 1.6307,
+      "step": 615
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7822004556655884,
+      "learning_rate": 0.0002633177570093458,
+      "loss": 0.1984,
+      "step": 616
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.2356181144714355,
+      "learning_rate": 0.00026320093457943923,
+      "loss": 0.6621,
+      "step": 617
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.7617156505584717,
+      "learning_rate": 0.0002630841121495327,
+      "loss": 1.6087,
+      "step": 618
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9174844026565552,
+      "learning_rate": 0.00026296728971962613,
+      "loss": 0.6119,
+      "step": 619
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9553015232086182,
+      "learning_rate": 0.0002628504672897196,
+      "loss": 0.6097,
+      "step": 620
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.008237838745117,
+      "learning_rate": 0.00026273364485981303,
+      "loss": 1.0158,
+      "step": 621
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.6738736629486084,
+      "learning_rate": 0.00026261682242990653,
+      "loss": 1.8953,
+      "step": 622
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6950339078903198,
+      "learning_rate": 0.0002625,
+      "loss": 0.6117,
+      "step": 623
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8285133838653564,
+      "learning_rate": 0.00026238317757009343,
+      "loss": 1.5104,
+      "step": 624
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8922878503799438,
+      "learning_rate": 0.0002622663551401869,
+      "loss": 1.0823,
+      "step": 625
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4725558757781982,
+      "learning_rate": 0.0002621495327102804,
+      "loss": 1.0039,
+      "step": 626
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8886311054229736,
+      "learning_rate": 0.00026203271028037384,
+      "loss": 0.223,
+      "step": 627
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8503968715667725,
+      "learning_rate": 0.00026191588785046723,
+      "loss": 0.3353,
+      "step": 628
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.4176552295684814,
+      "learning_rate": 0.00026179906542056074,
+      "loss": 1.174,
+      "step": 629
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.0846028327941895,
+      "learning_rate": 0.0002616822429906542,
+      "loss": 1.167,
+      "step": 630
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3127601146698,
+      "learning_rate": 0.00026156542056074764,
+      "loss": 1.3337,
+      "step": 631
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.930818796157837,
+      "learning_rate": 0.0002614485981308411,
+      "loss": 1.1385,
+      "step": 632
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.258216381072998,
+      "learning_rate": 0.00026133177570093454,
+      "loss": 0.2404,
+      "step": 633
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.8267745971679688,
+      "learning_rate": 0.00026121495327102804,
+      "loss": 1.6167,
+      "step": 634
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8974788188934326,
+      "learning_rate": 0.0002610981308411215,
+      "loss": 1.0328,
+      "step": 635
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.77346271276474,
+      "learning_rate": 0.00026098130841121494,
+      "loss": 0.3266,
+      "step": 636
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.7831459045410156,
+      "learning_rate": 0.0002608644859813084,
+      "loss": 1.4077,
+      "step": 637
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.6431167125701904,
+      "learning_rate": 0.00026074766355140184,
+      "loss": 0.6833,
+      "step": 638
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6087179780006409,
+      "learning_rate": 0.0002606308411214953,
+      "loss": 0.2766,
+      "step": 639
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9367430210113525,
+      "learning_rate": 0.00026051401869158874,
+      "loss": 0.9431,
+      "step": 640
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7643646001815796,
+      "learning_rate": 0.00026039719626168225,
+      "loss": 0.266,
+      "step": 641
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9407589435577393,
+      "learning_rate": 0.0002602803738317757,
+      "loss": 1.8693,
+      "step": 642
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.003830671310425,
+      "learning_rate": 0.00026016355140186915,
+      "loss": 1.502,
+      "step": 643
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.822230339050293,
+      "learning_rate": 0.0002600467289719626,
+      "loss": 1.3887,
+      "step": 644
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.560053586959839,
+      "learning_rate": 0.00025992990654205605,
+      "loss": 1.0752,
+      "step": 645
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.014336109161377,
+      "learning_rate": 0.0002598130841121495,
+      "loss": 1.1494,
+      "step": 646
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.923731803894043,
+      "learning_rate": 0.00025969626168224295,
+      "loss": 1.881,
+      "step": 647
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4779105186462402,
+      "learning_rate": 0.00025957943925233645,
+      "loss": 1.3292,
+      "step": 648
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.266974449157715,
+      "learning_rate": 0.0002594626168224299,
+      "loss": 0.8172,
+      "step": 649
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.011704921722412,
+      "learning_rate": 0.00025934579439252335,
+      "loss": 1.0187,
+      "step": 650
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.450458526611328,
+      "learning_rate": 0.0002592289719626168,
+      "loss": 0.7546,
+      "step": 651
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7575178146362305,
+      "learning_rate": 0.00025911214953271025,
+      "loss": 0.3802,
+      "step": 652
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.205942392349243,
+      "learning_rate": 0.0002589953271028037,
+      "loss": 0.7156,
+      "step": 653
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2393170595169067,
+      "learning_rate": 0.00025887850467289715,
+      "loss": 0.3533,
+      "step": 654
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4706906080245972,
+      "learning_rate": 0.00025876168224299066,
+      "loss": 0.213,
+      "step": 655
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9835875034332275,
+      "learning_rate": 0.0002586448598130841,
+      "loss": 0.9362,
+      "step": 656
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.087411880493164,
+      "learning_rate": 0.00025852803738317756,
+      "loss": 0.9816,
+      "step": 657
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7548537254333496,
+      "learning_rate": 0.000258411214953271,
+      "loss": 1.2555,
+      "step": 658
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9126662015914917,
+      "learning_rate": 0.00025829439252336446,
+      "loss": 0.4827,
+      "step": 659
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2377541065216064,
+      "learning_rate": 0.0002581775700934579,
+      "loss": 0.8133,
+      "step": 660
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.1898380517959595,
+      "learning_rate": 0.00025806074766355136,
+      "loss": 0.3985,
+      "step": 661
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.593562602996826,
+      "learning_rate": 0.0002579439252336448,
+      "loss": 1.1507,
+      "step": 662
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7690136432647705,
+      "learning_rate": 0.0002578271028037383,
+      "loss": 0.6489,
+      "step": 663
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.863168478012085,
+      "learning_rate": 0.00025771028037383176,
+      "loss": 1.0941,
+      "step": 664
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.265000820159912,
+      "learning_rate": 0.0002575934579439252,
+      "loss": 1.8224,
+      "step": 665
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0160512924194336,
+      "learning_rate": 0.00025747663551401866,
+      "loss": 0.3632,
+      "step": 666
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.9630157947540283,
+      "learning_rate": 0.00025735981308411216,
+      "loss": 0.8955,
+      "step": 667
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.4230360984802246,
+      "learning_rate": 0.00025724299065420556,
+      "loss": 0.9785,
+      "step": 668
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.195234775543213,
+      "learning_rate": 0.000257126168224299,
+      "loss": 1.3986,
+      "step": 669
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4523710012435913,
+      "learning_rate": 0.0002570093457943925,
+      "loss": 0.6776,
+      "step": 670
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4690325260162354,
+      "learning_rate": 0.00025689252336448597,
+      "loss": 0.249,
+      "step": 671
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.279630184173584,
+      "learning_rate": 0.0002567757009345794,
+      "loss": 0.7246,
+      "step": 672
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.060863971710205,
+      "learning_rate": 0.00025665887850467287,
+      "loss": 1.2851,
+      "step": 673
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2301100492477417,
+      "learning_rate": 0.00025654205607476637,
+      "loss": 1.0589,
+      "step": 674
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.657299757003784,
+      "learning_rate": 0.0002564252336448598,
+      "loss": 0.43,
+      "step": 675
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5239768028259277,
+      "learning_rate": 0.0002563084112149532,
+      "loss": 0.751,
+      "step": 676
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.668548583984375,
+      "learning_rate": 0.0002561915887850467,
+      "loss": 0.828,
+      "step": 677
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8801578283309937,
+      "learning_rate": 0.00025607476635514017,
+      "loss": 0.3485,
+      "step": 678
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.409930944442749,
+      "learning_rate": 0.0002559579439252336,
+      "loss": 1.4811,
+      "step": 679
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8931938409805298,
+      "learning_rate": 0.00025584112149532707,
+      "loss": 1.2942,
+      "step": 680
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7943670749664307,
+      "learning_rate": 0.0002557242990654205,
+      "loss": 0.0962,
+      "step": 681
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.3645403385162354,
+      "learning_rate": 0.000255607476635514,
+      "loss": 1.0911,
+      "step": 682
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8726574778556824,
+      "learning_rate": 0.0002554906542056075,
+      "loss": 0.4142,
+      "step": 683
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9188904762268066,
+      "learning_rate": 0.0002553738317757009,
+      "loss": 0.7505,
+      "step": 684
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5008811950683594,
+      "learning_rate": 0.0002552570093457944,
+      "loss": 0.1469,
+      "step": 685
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0353548526763916,
+      "learning_rate": 0.0002551401869158878,
+      "loss": 0.7272,
+      "step": 686
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.0340296030044556,
+      "learning_rate": 0.0002550233644859813,
+      "loss": 0.5369,
+      "step": 687
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5980859398841858,
+      "learning_rate": 0.0002549065420560747,
+      "loss": 0.2287,
+      "step": 688
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8595726490020752,
+      "learning_rate": 0.00025478971962616823,
+      "loss": 1.4049,
+      "step": 689
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.352480173110962,
+      "learning_rate": 0.0002546728971962617,
+      "loss": 0.4344,
+      "step": 690
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.3725690841674805,
+      "learning_rate": 0.00025455607476635513,
+      "loss": 0.9403,
+      "step": 691
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.2894887924194336,
+      "learning_rate": 0.0002544392523364486,
+      "loss": 0.7964,
+      "step": 692
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6135333776474,
+      "learning_rate": 0.00025432242990654203,
+      "loss": 0.7688,
+      "step": 693
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0678396224975586,
+      "learning_rate": 0.0002542056074766355,
+      "loss": 1.6217,
+      "step": 694
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9601709842681885,
+      "learning_rate": 0.00025408878504672893,
+      "loss": 0.994,
+      "step": 695
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.180920362472534,
+      "learning_rate": 0.00025397196261682243,
+      "loss": 0.6805,
+      "step": 696
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7055142521858215,
+      "learning_rate": 0.0002538551401869159,
+      "loss": 0.2653,
+      "step": 697
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6646941900253296,
+      "learning_rate": 0.00025373831775700933,
+      "loss": 0.4899,
+      "step": 698
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7786800265312195,
+      "learning_rate": 0.0002536214953271028,
+      "loss": 0.3756,
+      "step": 699
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5901905298233032,
+      "learning_rate": 0.00025350467289719623,
+      "loss": 0.6204,
+      "step": 700
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7284669876098633,
+      "learning_rate": 0.0002533878504672897,
+      "loss": 0.4973,
+      "step": 701
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.1285464763641357,
+      "learning_rate": 0.00025327102803738313,
+      "loss": 0.7694,
+      "step": 702
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5928149223327637,
+      "learning_rate": 0.00025315420560747664,
+      "loss": 1.1837,
+      "step": 703
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4394235610961914,
+      "learning_rate": 0.0002530373831775701,
+      "loss": 0.7965,
+      "step": 704
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4615914821624756,
+      "learning_rate": 0.00025292056074766354,
+      "loss": 1.0362,
+      "step": 705
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9794842600822449,
+      "learning_rate": 0.000252803738317757,
+      "loss": 0.476,
+      "step": 706
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6198465824127197,
+      "learning_rate": 0.00025268691588785044,
+      "loss": 0.9868,
+      "step": 707
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.645184874534607,
+      "learning_rate": 0.0002525700934579439,
+      "loss": 0.6718,
+      "step": 708
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.293039321899414,
+      "learning_rate": 0.00025245327102803734,
+      "loss": 0.8782,
+      "step": 709
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.1102802753448486,
+      "learning_rate": 0.0002523364485981308,
+      "loss": 0.5605,
+      "step": 710
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.9872053861618042,
+      "learning_rate": 0.0002522196261682243,
+      "loss": 0.2716,
+      "step": 711
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.878575801849365,
+      "learning_rate": 0.00025210280373831774,
+      "loss": 1.51,
+      "step": 712
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.1407004594802856,
+      "learning_rate": 0.0002519859813084112,
+      "loss": 0.5516,
+      "step": 713
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4332990646362305,
+      "learning_rate": 0.00025186915887850464,
+      "loss": 0.9987,
+      "step": 714
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.410146713256836,
+      "learning_rate": 0.00025175233644859815,
+      "loss": 1.5265,
+      "step": 715
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5623661279678345,
+      "learning_rate": 0.00025163551401869154,
+      "loss": 0.9784,
+      "step": 716
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.8849399089813232,
+      "learning_rate": 0.000251518691588785,
+      "loss": 1.1001,
+      "step": 717
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.7529332637786865,
+      "learning_rate": 0.0002514018691588785,
+      "loss": 0.7718,
+      "step": 718
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.7470431327819824,
+      "learning_rate": 0.00025128504672897195,
+      "loss": 0.7242,
+      "step": 719
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.720810651779175,
+      "learning_rate": 0.0002511682242990654,
+      "loss": 1.4097,
+      "step": 720
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.660722255706787,
+      "learning_rate": 0.00025105140186915885,
+      "loss": 1.1479,
+      "step": 721
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.5043423175811768,
+      "learning_rate": 0.00025093457943925235,
+      "loss": 0.6971,
+      "step": 722
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6985735297203064,
+      "learning_rate": 0.0002508177570093458,
+      "loss": 0.2465,
+      "step": 723
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.065187692642212,
+      "learning_rate": 0.0002507009345794392,
+      "loss": 1.2275,
+      "step": 724
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.846989631652832,
+      "learning_rate": 0.0002505841121495327,
+      "loss": 0.584,
+      "step": 725
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5561527013778687,
+      "learning_rate": 0.00025046728971962615,
+      "loss": 0.8527,
+      "step": 726
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.3260703086853027,
+      "learning_rate": 0.0002503504672897196,
+      "loss": 1.0349,
+      "step": 727
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7104542255401611,
+      "learning_rate": 0.00025023364485981305,
+      "loss": 0.5936,
+      "step": 728
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5694267749786377,
+      "learning_rate": 0.0002501168224299065,
+      "loss": 1.0465,
+      "step": 729
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.052154302597046,
+      "learning_rate": 0.00025,
+      "loss": 1.0695,
+      "step": 730
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.267850160598755,
+      "learning_rate": 0.00024988317757009346,
+      "loss": 1.1448,
+      "step": 731
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.9715405702590942,
+      "learning_rate": 0.0002497663551401869,
+      "loss": 0.5502,
+      "step": 732
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.080495834350586,
+      "learning_rate": 0.00024964953271028036,
+      "loss": 0.8377,
+      "step": 733
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.0935518741607666,
+      "learning_rate": 0.0002495327102803738,
+      "loss": 1.665,
+      "step": 734
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.7347010970115662,
+      "learning_rate": 0.00024941588785046726,
+      "loss": 0.288,
+      "step": 735
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.334194302558899,
+      "learning_rate": 0.0002492990654205607,
+      "loss": 0.3979,
+      "step": 736
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.544018507003784,
+      "learning_rate": 0.0002491822429906542,
+      "loss": 0.8903,
+      "step": 737
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9076032638549805,
+      "learning_rate": 0.00024906542056074766,
+      "loss": 0.7581,
+      "step": 738
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0532596111297607,
+      "learning_rate": 0.0002489485981308411,
+      "loss": 0.5799,
+      "step": 739
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9696563482284546,
+      "learning_rate": 0.00024883177570093456,
+      "loss": 1.4261,
+      "step": 740
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7314748764038086,
+      "learning_rate": 0.000248714953271028,
+      "loss": 0.2073,
+      "step": 741
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2788918018341064,
+      "learning_rate": 0.00024859813084112146,
+      "loss": 0.3712,
+      "step": 742
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2528998851776123,
+      "learning_rate": 0.0002484813084112149,
+      "loss": 1.7037,
+      "step": 743
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.076824903488159,
+      "learning_rate": 0.0002483644859813084,
+      "loss": 1.1481,
+      "step": 744
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8832570314407349,
+      "learning_rate": 0.00024824766355140186,
+      "loss": 0.6691,
+      "step": 745
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.096408843994141,
+      "learning_rate": 0.0002481308411214953,
+      "loss": 1.2458,
+      "step": 746
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.913481593132019,
+      "learning_rate": 0.00024801401869158876,
+      "loss": 0.921,
+      "step": 747
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3440444469451904,
+      "learning_rate": 0.0002478971962616822,
+      "loss": 0.9811,
+      "step": 748
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1785082817077637,
+      "learning_rate": 0.00024778037383177566,
+      "loss": 0.5827,
+      "step": 749
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.120669364929199,
+      "learning_rate": 0.0002476635514018691,
+      "loss": 0.992,
+      "step": 750
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.2538716793060303,
+      "learning_rate": 0.0002475467289719626,
+      "loss": 1.2138,
+      "step": 751
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8802791833877563,
+      "learning_rate": 0.00024742990654205607,
+      "loss": 1.4769,
+      "step": 752
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9248284101486206,
+      "learning_rate": 0.0002473130841121495,
+      "loss": 1.3245,
+      "step": 753
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0179731845855713,
+      "learning_rate": 0.00024719626168224297,
+      "loss": 1.1622,
+      "step": 754
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.760973334312439,
+      "learning_rate": 0.0002470794392523364,
+      "loss": 0.6696,
+      "step": 755
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.4388394355773926,
+      "learning_rate": 0.00024696261682242987,
+      "loss": 1.0924,
+      "step": 756
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.125169277191162,
+      "learning_rate": 0.0002468457943925233,
+      "loss": 1.508,
+      "step": 757
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.1705702543258667,
+      "learning_rate": 0.00024672897196261677,
+      "loss": 0.4995,
+      "step": 758
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.662678837776184,
+      "learning_rate": 0.0002466121495327103,
+      "loss": 0.4772,
+      "step": 759
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.0994110107421875,
+      "learning_rate": 0.0002464953271028037,
+      "loss": 0.6918,
+      "step": 760
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3928942680358887,
+      "learning_rate": 0.0002463785046728972,
+      "loss": 1.7424,
+      "step": 761
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1945111751556396,
+      "learning_rate": 0.0002462616822429906,
+      "loss": 1.1678,
+      "step": 762
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6065562963485718,
+      "learning_rate": 0.00024614485981308413,
+      "loss": 0.2749,
+      "step": 763
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.162224292755127,
+      "learning_rate": 0.0002460280373831775,
+      "loss": 0.9568,
+      "step": 764
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2459497451782227,
+      "learning_rate": 0.000245911214953271,
+      "loss": 0.7103,
+      "step": 765
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7798285484313965,
+      "learning_rate": 0.0002457943925233645,
+      "loss": 0.5845,
+      "step": 766
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8737719058990479,
+      "learning_rate": 0.00024567757009345793,
+      "loss": 0.9293,
+      "step": 767
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1505696773529053,
+      "learning_rate": 0.0002455607476635514,
+      "loss": 0.5254,
+      "step": 768
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1278975009918213,
+      "learning_rate": 0.00024544392523364483,
+      "loss": 0.6887,
+      "step": 769
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0058064460754395,
+      "learning_rate": 0.00024532710280373833,
+      "loss": 1.4686,
+      "step": 770
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.39130163192749,
+      "learning_rate": 0.0002452102803738318,
+      "loss": 0.84,
+      "step": 771
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1862969398498535,
+      "learning_rate": 0.0002450934579439252,
+      "loss": 0.8424,
+      "step": 772
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3539414405822754,
+      "learning_rate": 0.0002449766355140187,
+      "loss": 0.5938,
+      "step": 773
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.706716537475586,
+      "learning_rate": 0.00024485981308411213,
+      "loss": 0.5478,
+      "step": 774
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.8543996810913086,
+      "learning_rate": 0.0002447429906542056,
+      "loss": 0.8018,
+      "step": 775
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0526278018951416,
+      "learning_rate": 0.00024462616822429903,
+      "loss": 0.6896,
+      "step": 776
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.7325849533081055,
+      "learning_rate": 0.0002445093457943925,
+      "loss": 1.2247,
+      "step": 777
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.3046183586120605,
+      "learning_rate": 0.000244392523364486,
+      "loss": 1.2575,
+      "step": 778
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.125836133956909,
+      "learning_rate": 0.00024427570093457944,
+      "loss": 0.7927,
+      "step": 779
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7354124784469604,
+      "learning_rate": 0.0002441588785046729,
+      "loss": 0.317,
+      "step": 780
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.628983736038208,
+      "learning_rate": 0.00024404205607476634,
+      "loss": 0.5974,
+      "step": 781
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.533229351043701,
+      "learning_rate": 0.0002439252336448598,
+      "loss": 1.4416,
+      "step": 782
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.980332851409912,
+      "learning_rate": 0.00024380841121495324,
+      "loss": 1.5009,
+      "step": 783
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8960031867027283,
+      "learning_rate": 0.00024369158878504671,
+      "loss": 0.2601,
+      "step": 784
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8643038272857666,
+      "learning_rate": 0.00024357476635514016,
+      "loss": 1.589,
+      "step": 785
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.299959897994995,
+      "learning_rate": 0.00024345794392523364,
+      "loss": 1.6819,
+      "step": 786
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8638948202133179,
+      "learning_rate": 0.0002433411214953271,
+      "loss": 1.3736,
+      "step": 787
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6240087747573853,
+      "learning_rate": 0.00024322429906542051,
+      "loss": 0.5334,
+      "step": 788
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.438859224319458,
+      "learning_rate": 0.000243107476635514,
+      "loss": 1.2165,
+      "step": 789
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9611045122146606,
+      "learning_rate": 0.00024299065420560744,
+      "loss": 1.4199,
+      "step": 790
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.787391185760498,
+      "learning_rate": 0.00024287383177570092,
+      "loss": 1.1579,
+      "step": 791
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4074761867523193,
+      "learning_rate": 0.00024275700934579437,
+      "loss": 1.5847,
+      "step": 792
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7211910486221313,
+      "learning_rate": 0.00024264018691588785,
+      "loss": 0.6404,
+      "step": 793
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.036355495452881,
+      "learning_rate": 0.0002425233644859813,
+      "loss": 0.4518,
+      "step": 794
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.337662696838379,
+      "learning_rate": 0.00024240654205607477,
+      "loss": 0.8751,
+      "step": 795
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0157761573791504,
+      "learning_rate": 0.0002422897196261682,
+      "loss": 0.6199,
+      "step": 796
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2106847763061523,
+      "learning_rate": 0.00024217289719626165,
+      "loss": 1.3694,
+      "step": 797
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.1246968507766724,
+      "learning_rate": 0.00024205607476635512,
+      "loss": 0.4661,
+      "step": 798
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3180928230285645,
+      "learning_rate": 0.00024193925233644857,
+      "loss": 1.1047,
+      "step": 799
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.011066913604736,
+      "learning_rate": 0.00024182242990654205,
+      "loss": 0.7958,
+      "step": 800
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.280853271484375,
+      "learning_rate": 0.0002417056074766355,
+      "loss": 1.0047,
+      "step": 801
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.310791492462158,
+      "learning_rate": 0.00024158878504672895,
+      "loss": 1.691,
+      "step": 802
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7300541400909424,
+      "learning_rate": 0.00024147196261682243,
+      "loss": 1.4916,
+      "step": 803
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.2915143966674805,
+      "learning_rate": 0.00024135514018691585,
+      "loss": 0.9508,
+      "step": 804
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.642038583755493,
+      "learning_rate": 0.00024123831775700933,
+      "loss": 0.835,
+      "step": 805
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.139812707901001,
+      "learning_rate": 0.00024112149532710278,
+      "loss": 0.8397,
+      "step": 806
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.194612979888916,
+      "learning_rate": 0.00024100467289719623,
+      "loss": 1.2641,
+      "step": 807
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.043593168258667,
+      "learning_rate": 0.0002408878504672897,
+      "loss": 1.1955,
+      "step": 808
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.5014655590057373,
+      "learning_rate": 0.00024077102803738315,
+      "loss": 1.1901,
+      "step": 809
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.0221776962280273,
+      "learning_rate": 0.00024065420560747663,
+      "loss": 1.5222,
+      "step": 810
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3791370391845703,
+      "learning_rate": 0.00024053738317757008,
+      "loss": 1.2012,
+      "step": 811
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.600904941558838,
+      "learning_rate": 0.0002404205607476635,
+      "loss": 1.5508,
+      "step": 812
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9494956731796265,
+      "learning_rate": 0.00024030373831775698,
+      "loss": 0.8289,
+      "step": 813
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.927809715270996,
+      "learning_rate": 0.00024018691588785043,
+      "loss": 1.2377,
+      "step": 814
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.0436631441116333,
+      "learning_rate": 0.0002400700934579439,
+      "loss": 0.2766,
+      "step": 815
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.220791816711426,
+      "learning_rate": 0.00023995327102803736,
+      "loss": 1.1935,
+      "step": 816
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.4562952518463135,
+      "learning_rate": 0.00023983644859813084,
+      "loss": 0.9728,
+      "step": 817
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.344489812850952,
+      "learning_rate": 0.00023971962616822429,
+      "loss": 0.9271,
+      "step": 818
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.47554874420166,
+      "learning_rate": 0.00023960280373831776,
+      "loss": 0.7369,
+      "step": 819
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6083333492279053,
+      "learning_rate": 0.00023948598130841119,
+      "loss": 0.2156,
+      "step": 820
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2916417121887207,
+      "learning_rate": 0.00023936915887850464,
+      "loss": 0.5969,
+      "step": 821
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.7046704292297363,
+      "learning_rate": 0.0002392523364485981,
+      "loss": 1.4117,
+      "step": 822
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.682845115661621,
+      "learning_rate": 0.00023913551401869156,
+      "loss": 0.7092,
+      "step": 823
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9161324501037598,
+      "learning_rate": 0.00023901869158878504,
+      "loss": 1.6658,
+      "step": 824
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.6389951705932617,
+      "learning_rate": 0.0002389018691588785,
+      "loss": 1.0424,
+      "step": 825
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.222813606262207,
+      "learning_rate": 0.00023878504672897194,
+      "loss": 1.8298,
+      "step": 826
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.769364058971405,
+      "learning_rate": 0.00023866822429906542,
+      "loss": 0.3196,
+      "step": 827
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.443896770477295,
+      "learning_rate": 0.00023855140186915884,
+      "loss": 1.2312,
+      "step": 828
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5431735515594482,
+      "learning_rate": 0.00023843457943925232,
+      "loss": 1.0854,
+      "step": 829
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.4142019748687744,
+      "learning_rate": 0.00023831775700934577,
+      "loss": 0.7535,
+      "step": 830
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.1519904136657715,
+      "learning_rate": 0.00023820093457943922,
+      "loss": 1.633,
+      "step": 831
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1021811962127686,
+      "learning_rate": 0.0002380841121495327,
+      "loss": 1.1532,
+      "step": 832
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.130340099334717,
+      "learning_rate": 0.00023796728971962615,
+      "loss": 0.9221,
+      "step": 833
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.2165324687957764,
+      "learning_rate": 0.00023785046728971962,
+      "loss": 1.0475,
+      "step": 834
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.920137882232666,
+      "learning_rate": 0.00023773364485981307,
+      "loss": 1.4406,
+      "step": 835
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.199028253555298,
+      "learning_rate": 0.0002376168224299065,
+      "loss": 0.945,
+      "step": 836
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.366101026535034,
+      "learning_rate": 0.00023749999999999997,
+      "loss": 1.2456,
+      "step": 837
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0884835720062256,
+      "learning_rate": 0.00023738317757009342,
+      "loss": 1.371,
+      "step": 838
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.546804666519165,
+      "learning_rate": 0.0002372663551401869,
+      "loss": 1.0952,
+      "step": 839
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8506627082824707,
+      "learning_rate": 0.00023714953271028035,
+      "loss": 1.1875,
+      "step": 840
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.8292133808135986,
+      "learning_rate": 0.00023703271028037383,
+      "loss": 1.2783,
+      "step": 841
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.448188781738281,
+      "learning_rate": 0.00023691588785046728,
+      "loss": 1.0957,
+      "step": 842
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.644364595413208,
+      "learning_rate": 0.00023679906542056075,
+      "loss": 0.4811,
+      "step": 843
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5675925016403198,
+      "learning_rate": 0.00023668224299065418,
+      "loss": 0.4748,
+      "step": 844
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.530189871788025,
+      "learning_rate": 0.00023656542056074763,
+      "loss": 0.6145,
+      "step": 845
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5131443738937378,
+      "learning_rate": 0.0002364485981308411,
+      "loss": 1.2528,
+      "step": 846
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.101269245147705,
+      "learning_rate": 0.00023633177570093455,
+      "loss": 0.8421,
+      "step": 847
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0407848358154297,
+      "learning_rate": 0.00023621495327102803,
+      "loss": 0.6174,
+      "step": 848
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4575300216674805,
+      "learning_rate": 0.00023609813084112148,
+      "loss": 0.5449,
+      "step": 849
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.286203145980835,
+      "learning_rate": 0.00023598130841121493,
+      "loss": 1.8737,
+      "step": 850
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9306058883666992,
+      "learning_rate": 0.0002358644859813084,
+      "loss": 0.8433,
+      "step": 851
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.401845932006836,
+      "learning_rate": 0.00023574766355140183,
+      "loss": 0.1541,
+      "step": 852
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6361005306243896,
+      "learning_rate": 0.0002356308411214953,
+      "loss": 0.7135,
+      "step": 853
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.040468215942383,
+      "learning_rate": 0.00023551401869158876,
+      "loss": 0.5627,
+      "step": 854
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.8553760051727295,
+      "learning_rate": 0.0002353971962616822,
+      "loss": 1.019,
+      "step": 855
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.2895450592041016,
+      "learning_rate": 0.00023528037383177569,
+      "loss": 0.7987,
+      "step": 856
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.958900213241577,
+      "learning_rate": 0.00023516355140186914,
+      "loss": 1.0162,
+      "step": 857
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.653841972351074,
+      "learning_rate": 0.0002350467289719626,
+      "loss": 0.9121,
+      "step": 858
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.3145627975463867,
+      "learning_rate": 0.00023492990654205606,
+      "loss": 1.6394,
+      "step": 859
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.452100992202759,
+      "learning_rate": 0.00023481308411214949,
+      "loss": 0.9281,
+      "step": 860
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.180034875869751,
+      "learning_rate": 0.00023469626168224296,
+      "loss": 0.4314,
+      "step": 861
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.1082077026367188,
+      "learning_rate": 0.0002345794392523364,
+      "loss": 0.6923,
+      "step": 862
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2554924488067627,
+      "learning_rate": 0.0002344626168224299,
+      "loss": 0.9983,
+      "step": 863
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.957667589187622,
+      "learning_rate": 0.00023434579439252334,
+      "loss": 1.4724,
+      "step": 864
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.3966641426086426,
+      "learning_rate": 0.00023422897196261682,
+      "loss": 0.8326,
+      "step": 865
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6179311275482178,
+      "learning_rate": 0.00023411214953271027,
+      "loss": 1.7635,
+      "step": 866
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6067564487457275,
+      "learning_rate": 0.00023399532710280374,
+      "loss": 0.5869,
+      "step": 867
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7425529956817627,
+      "learning_rate": 0.00023387850467289717,
+      "loss": 0.9132,
+      "step": 868
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3865866661071777,
+      "learning_rate": 0.00023376168224299062,
+      "loss": 1.1675,
+      "step": 869
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8935320377349854,
+      "learning_rate": 0.0002336448598130841,
+      "loss": 0.8528,
+      "step": 870
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8941283226013184,
+      "learning_rate": 0.00023352803738317754,
+      "loss": 0.7109,
+      "step": 871
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0906989574432373,
+      "learning_rate": 0.00023341121495327102,
+      "loss": 0.8982,
+      "step": 872
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4004254341125488,
+      "learning_rate": 0.00023329439252336447,
+      "loss": 0.637,
+      "step": 873
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8859999179840088,
+      "learning_rate": 0.00023317757009345792,
+      "loss": 0.8726,
+      "step": 874
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0776727199554443,
+      "learning_rate": 0.0002330607476635514,
+      "loss": 0.8262,
+      "step": 875
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6682260036468506,
+      "learning_rate": 0.00023294392523364482,
+      "loss": 0.4234,
+      "step": 876
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.968831181526184,
+      "learning_rate": 0.0002328271028037383,
+      "loss": 0.1693,
+      "step": 877
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7580511569976807,
+      "learning_rate": 0.00023271028037383175,
+      "loss": 0.7502,
+      "step": 878
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.454692840576172,
+      "learning_rate": 0.0002325934579439252,
+      "loss": 1.7017,
+      "step": 879
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.112985610961914,
+      "learning_rate": 0.00023247663551401868,
+      "loss": 1.0773,
+      "step": 880
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.550245761871338,
+      "learning_rate": 0.00023235981308411213,
+      "loss": 1.2831,
+      "step": 881
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.5282704830169678,
+      "learning_rate": 0.0002322429906542056,
+      "loss": 0.5417,
+      "step": 882
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.580568552017212,
+      "learning_rate": 0.00023212616822429905,
+      "loss": 1.4039,
+      "step": 883
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3084704875946045,
+      "learning_rate": 0.00023200934579439248,
+      "loss": 1.3537,
+      "step": 884
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7144612073898315,
+      "learning_rate": 0.00023189252336448595,
+      "loss": 0.2048,
+      "step": 885
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.499814510345459,
+      "learning_rate": 0.0002317757009345794,
+      "loss": 0.9366,
+      "step": 886
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.09336519241333,
+      "learning_rate": 0.00023165887850467288,
+      "loss": 1.7808,
+      "step": 887
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5693227648735046,
+      "learning_rate": 0.00023154205607476633,
+      "loss": 0.1463,
+      "step": 888
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.217259645462036,
+      "learning_rate": 0.0002314252336448598,
+      "loss": 1.1722,
+      "step": 889
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.928795337677002,
+      "learning_rate": 0.00023130841121495326,
+      "loss": 0.7112,
+      "step": 890
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.595618486404419,
+      "learning_rate": 0.00023119158878504674,
+      "loss": 0.7386,
+      "step": 891
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.523850679397583,
+      "learning_rate": 0.00023107476635514016,
+      "loss": 1.5385,
+      "step": 892
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7062773108482361,
+      "learning_rate": 0.0002309579439252336,
+      "loss": 0.3865,
+      "step": 893
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.406078577041626,
+      "learning_rate": 0.00023084112149532709,
+      "loss": 0.3898,
+      "step": 894
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.1150354146957397,
+      "learning_rate": 0.00023072429906542054,
+      "loss": 0.8224,
+      "step": 895
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0604560375213623,
+      "learning_rate": 0.000230607476635514,
+      "loss": 0.5875,
+      "step": 896
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.368139743804932,
+      "learning_rate": 0.00023049065420560746,
+      "loss": 1.3192,
+      "step": 897
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.1544618606567383,
+      "learning_rate": 0.0002303738317757009,
+      "loss": 0.7825,
+      "step": 898
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3759732246398926,
+      "learning_rate": 0.0002302570093457944,
+      "loss": 0.4662,
+      "step": 899
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9174294471740723,
+      "learning_rate": 0.0002301401869158878,
+      "loss": 1.0672,
+      "step": 900
+    },
+    {
+      "epoch": 0.94,
+      "eval_loss": 1.1630452871322632,
+      "eval_runtime": 1.0017,
+      "eval_samples_per_second": 4.992,
+      "eval_steps_per_second": 0.998,
+      "step": 900
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9549384117126465,
+      "learning_rate": 0.0002300233644859813,
+      "loss": 1.0137,
+      "step": 901
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.340488314628601,
+      "learning_rate": 0.00022990654205607474,
+      "loss": 0.5765,
+      "step": 902
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.487250804901123,
+      "learning_rate": 0.0002297897196261682,
+      "loss": 1.3398,
+      "step": 903
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.376892328262329,
+      "learning_rate": 0.00022967289719626167,
+      "loss": 1.1591,
+      "step": 904
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.779705286026001,
+      "learning_rate": 0.00022955607476635512,
+      "loss": 1.4591,
+      "step": 905
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9788891077041626,
+      "learning_rate": 0.0002294392523364486,
+      "loss": 0.9374,
+      "step": 906
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1561695337295532,
+      "learning_rate": 0.00022932242990654204,
+      "loss": 0.4868,
+      "step": 907
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8206226825714111,
+      "learning_rate": 0.00022920560747663547,
+      "loss": 0.3994,
+      "step": 908
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.6440787315368652,
+      "learning_rate": 0.00022908878504672894,
+      "loss": 1.3197,
+      "step": 909
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1878468990325928,
+      "learning_rate": 0.0002289719626168224,
+      "loss": 0.332,
+      "step": 910
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2395591735839844,
+      "learning_rate": 0.00022885514018691587,
+      "loss": 1.1612,
+      "step": 911
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8139145374298096,
+      "learning_rate": 0.00022873831775700932,
+      "loss": 1.7079,
+      "step": 912
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3331412076950073,
+      "learning_rate": 0.0002286214953271028,
+      "loss": 0.67,
+      "step": 913
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7752532958984375,
+      "learning_rate": 0.00022850467289719625,
+      "loss": 1.2638,
+      "step": 914
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7987313270568848,
+      "learning_rate": 0.00022838785046728973,
+      "loss": 1.0048,
+      "step": 915
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3547263145446777,
+      "learning_rate": 0.00022827102803738315,
+      "loss": 0.552,
+      "step": 916
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9099464416503906,
+      "learning_rate": 0.0002281542056074766,
+      "loss": 0.4864,
+      "step": 917
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.29815149307251,
+      "learning_rate": 0.00022803738317757008,
+      "loss": 1.5697,
+      "step": 918
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8438161611557007,
+      "learning_rate": 0.00022792056074766353,
+      "loss": 0.6215,
+      "step": 919
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.376704454421997,
+      "learning_rate": 0.000227803738317757,
+      "loss": 1.48,
+      "step": 920
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3474555015563965,
+      "learning_rate": 0.00022768691588785045,
+      "loss": 1.1367,
+      "step": 921
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.982185125350952,
+      "learning_rate": 0.0002275700934579439,
+      "loss": 1.443,
+      "step": 922
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.1197896003723145,
+      "learning_rate": 0.00022745327102803738,
+      "loss": 1.083,
+      "step": 923
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.8206117153167725,
+      "learning_rate": 0.0002273364485981308,
+      "loss": 1.0382,
+      "step": 924
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9323513507843018,
+      "learning_rate": 0.00022721962616822425,
+      "loss": 1.2924,
+      "step": 925
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5834381580352783,
+      "learning_rate": 0.00022710280373831773,
+      "loss": 0.9821,
+      "step": 926
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3738291263580322,
+      "learning_rate": 0.00022698598130841118,
+      "loss": 0.7764,
+      "step": 927
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9597241878509521,
+      "learning_rate": 0.00022686915887850466,
+      "loss": 0.8203,
+      "step": 928
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6998361349105835,
+      "learning_rate": 0.0002267523364485981,
+      "loss": 0.1453,
+      "step": 929
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9886910915374756,
+      "learning_rate": 0.00022663551401869158,
+      "loss": 1.1826,
+      "step": 930
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.020965099334717,
+      "learning_rate": 0.00022651869158878503,
+      "loss": 0.8265,
+      "step": 931
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5389176607131958,
+      "learning_rate": 0.00022640186915887846,
+      "loss": 0.5353,
+      "step": 932
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.712364912033081,
+      "learning_rate": 0.00022628504672897194,
+      "loss": 0.4676,
+      "step": 933
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.8216958045959473,
+      "learning_rate": 0.00022616822429906539,
+      "loss": 1.0984,
+      "step": 934
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.8422656059265137,
+      "learning_rate": 0.00022605140186915886,
+      "loss": 0.3279,
+      "step": 935
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1753759384155273,
+      "learning_rate": 0.0002259345794392523,
+      "loss": 0.5013,
+      "step": 936
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3775906562805176,
+      "learning_rate": 0.0002258177570093458,
+      "loss": 1.9156,
+      "step": 937
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1752231121063232,
+      "learning_rate": 0.00022570093457943924,
+      "loss": 1.2206,
+      "step": 938
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4699342250823975,
+      "learning_rate": 0.00022558411214953272,
+      "loss": 0.9297,
+      "step": 939
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.877358615398407,
+      "learning_rate": 0.00022546728971962614,
+      "loss": 0.3107,
+      "step": 940
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7622907161712646,
+      "learning_rate": 0.0002253504672897196,
+      "loss": 0.4124,
+      "step": 941
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6599836349487305,
+      "learning_rate": 0.00022523364485981307,
+      "loss": 1.7029,
+      "step": 942
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.203469753265381,
+      "learning_rate": 0.00022511682242990652,
+      "loss": 1.5104,
+      "step": 943
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.2014012336730957,
+      "learning_rate": 0.000225,
+      "loss": 1.0979,
+      "step": 944
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.75557279586792,
+      "learning_rate": 0.00022488317757009344,
+      "loss": 0.5876,
+      "step": 945
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8471612930297852,
+      "learning_rate": 0.0002247663551401869,
+      "loss": 0.2429,
+      "step": 946
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.0547866821289062,
+      "learning_rate": 0.00022464953271028037,
+      "loss": 1.34,
+      "step": 947
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2609493732452393,
+      "learning_rate": 0.00022453271028037382,
+      "loss": 0.9612,
+      "step": 948
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6535717248916626,
+      "learning_rate": 0.00022441588785046724,
+      "loss": 0.2885,
+      "step": 949
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7238208055496216,
+      "learning_rate": 0.00022429906542056072,
+      "loss": 0.5047,
+      "step": 950
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.345454216003418,
+      "learning_rate": 0.00022418224299065417,
+      "loss": 0.6963,
+      "step": 951
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6006823182106018,
+      "learning_rate": 0.00022406542056074765,
+      "loss": 0.1972,
+      "step": 952
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3403688669204712,
+      "learning_rate": 0.0002239485981308411,
+      "loss": 0.4257,
+      "step": 953
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.329044342041016,
+      "learning_rate": 0.00022383177570093458,
+      "loss": 1.3304,
+      "step": 954
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3422555923461914,
+      "learning_rate": 0.00022371495327102803,
+      "loss": 0.3354,
+      "step": 955
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.834536075592041,
+      "learning_rate": 0.0002235981308411215,
+      "loss": 1.2915,
+      "step": 956
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.039134979248047,
+      "learning_rate": 0.00022348130841121493,
+      "loss": 0.7149,
+      "step": 957
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.1390619277954102,
+      "learning_rate": 0.00022336448598130838,
+      "loss": 0.3418,
+      "step": 958
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4070141315460205,
+      "learning_rate": 0.00022324766355140185,
+      "loss": 0.3751,
+      "step": 959
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.0205116271972656,
+      "learning_rate": 0.0002231308411214953,
+      "loss": 0.9732,
+      "step": 960
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.3249778747558594,
+      "learning_rate": 0.00022301401869158878,
+      "loss": 0.3551,
+      "step": 961
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.534330129623413,
+      "learning_rate": 0.00022289719626168223,
+      "loss": 0.5207,
+      "step": 962
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.9354491233825684,
+      "learning_rate": 0.0002227803738317757,
+      "loss": 0.3616,
+      "step": 963
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.116346836090088,
+      "learning_rate": 0.00022266355140186916,
+      "loss": 1.0175,
+      "step": 964
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.099825859069824,
+      "learning_rate": 0.00022254672897196258,
+      "loss": 0.5644,
+      "step": 965
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.362391710281372,
+      "learning_rate": 0.00022242990654205606,
+      "loss": 0.9214,
+      "step": 966
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 1.5349410772323608,
+      "learning_rate": 0.0002223130841121495,
+      "loss": 0.4219,
+      "step": 967
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.8551294803619385,
+      "learning_rate": 0.00022219626168224296,
+      "loss": 0.9566,
+      "step": 968
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 3.527337074279785,
+      "learning_rate": 0.00022207943925233643,
+      "loss": 1.4504,
+      "step": 969
+    },
+    {
+      "epoch": 1.01,
+      "grad_norm": 2.6233699321746826,
+      "learning_rate": 0.00022196261682242988,
+      "loss": 0.5531,
+      "step": 970
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.0485408306121826,
+      "learning_rate": 0.00022184579439252336,
+      "loss": 0.7565,
+      "step": 971
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 3.941380262374878,
+      "learning_rate": 0.0002217289719626168,
+      "loss": 0.6388,
+      "step": 972
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.8376948833465576,
+      "learning_rate": 0.00022161214953271023,
+      "loss": 0.9422,
+      "step": 973
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.7060530185699463,
+      "learning_rate": 0.0002214953271028037,
+      "loss": 0.8072,
+      "step": 974
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.446294069290161,
+      "learning_rate": 0.00022137850467289716,
+      "loss": 0.5575,
+      "step": 975
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 1.1146581172943115,
+      "learning_rate": 0.00022126168224299064,
+      "loss": 0.2803,
+      "step": 976
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.5902735590934753,
+      "learning_rate": 0.0002211448598130841,
+      "loss": 0.2146,
+      "step": 977
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.6364985704421997,
+      "learning_rate": 0.00022102803738317757,
+      "loss": 0.2221,
+      "step": 978
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.4132468700408936,
+      "learning_rate": 0.00022091121495327102,
+      "loss": 0.7333,
+      "step": 979
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.3000097274780273,
+      "learning_rate": 0.0002207943925233645,
+      "loss": 1.6004,
+      "step": 980
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.1039211750030518,
+      "learning_rate": 0.00022067757009345792,
+      "loss": 0.7024,
+      "step": 981
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.9659446477890015,
+      "learning_rate": 0.00022056074766355137,
+      "loss": 1.5743,
+      "step": 982
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.8441948890686035,
+      "learning_rate": 0.00022044392523364484,
+      "loss": 0.6775,
+      "step": 983
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.7042272090911865,
+      "learning_rate": 0.0002203271028037383,
+      "loss": 0.5636,
+      "step": 984
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.106860637664795,
+      "learning_rate": 0.00022021028037383177,
+      "loss": 1.243,
+      "step": 985
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 1.890007734298706,
+      "learning_rate": 0.00022009345794392522,
+      "loss": 1.2065,
+      "step": 986
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 3.0000622272491455,
+      "learning_rate": 0.00021997663551401867,
+      "loss": 0.8706,
+      "step": 987
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.43005108833313,
+      "learning_rate": 0.00021985981308411215,
+      "loss": 0.5945,
+      "step": 988
+    },
+    {
+      "epoch": 1.03,
+      "grad_norm": 2.614928960800171,
+      "learning_rate": 0.00021974299065420557,
+      "loss": 0.5288,
+      "step": 989
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.0085580348968506,
+      "learning_rate": 0.00021962616822429905,
+      "loss": 0.4928,
+      "step": 990
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.5782190561294556,
+      "learning_rate": 0.0002195093457943925,
+      "loss": 0.1496,
+      "step": 991
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.675377607345581,
+      "learning_rate": 0.00021939252336448595,
+      "loss": 0.2521,
+      "step": 992
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.4882884621620178,
+      "learning_rate": 0.00021927570093457943,
+      "loss": 0.1353,
+      "step": 993
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 1.384681224822998,
+      "learning_rate": 0.00021915887850467288,
+      "loss": 0.6342,
+      "step": 994
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.8572222590446472,
+      "learning_rate": 0.00021904205607476635,
+      "loss": 0.117,
+      "step": 995
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.631561040878296,
+      "learning_rate": 0.0002189252336448598,
+      "loss": 1.2984,
+      "step": 996
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.5276076793670654,
+      "learning_rate": 0.00021880841121495323,
+      "loss": 0.7352,
+      "step": 997
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 3.206256151199341,
+      "learning_rate": 0.0002186915887850467,
+      "loss": 1.4104,
+      "step": 998
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.2699246406555176,
+      "learning_rate": 0.00021857476635514015,
+      "loss": 0.5665,
+      "step": 999
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.288726806640625,
+      "learning_rate": 0.00021845794392523363,
+      "loss": 0.696,
+      "step": 1000
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.2717671394348145,
+      "learning_rate": 0.00021834112149532708,
+      "loss": 0.9446,
+      "step": 1001
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.451627731323242,
+      "learning_rate": 0.00021822429906542056,
+      "loss": 1.362,
+      "step": 1002
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.8697036504745483,
+      "learning_rate": 0.000218107476635514,
+      "loss": 0.5908,
+      "step": 1003
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.1752212047576904,
+      "learning_rate": 0.00021799065420560748,
+      "loss": 1.2794,
+      "step": 1004
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 5.133718013763428,
+      "learning_rate": 0.0002178738317757009,
+      "loss": 0.9315,
+      "step": 1005
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.6146345138549805,
+      "learning_rate": 0.00021775700934579436,
+      "loss": 1.3513,
+      "step": 1006
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 2.957939386367798,
+      "learning_rate": 0.00021764018691588783,
+      "loss": 1.5546,
+      "step": 1007
+    },
+    {
+      "epoch": 1.05,
+      "grad_norm": 1.8134377002716064,
+      "learning_rate": 0.00021752336448598128,
+      "loss": 0.6268,
+      "step": 1008
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.2340497970581055,
+      "learning_rate": 0.00021740654205607476,
+      "loss": 0.6106,
+      "step": 1009
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.744544744491577,
+      "learning_rate": 0.0002172897196261682,
+      "loss": 0.8776,
+      "step": 1010
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.7156779766082764,
+      "learning_rate": 0.00021717289719626166,
+      "loss": 1.5869,
+      "step": 1011
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.6292368173599243,
+      "learning_rate": 0.00021705607476635514,
+      "loss": 0.2781,
+      "step": 1012
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.877001166343689,
+      "learning_rate": 0.00021693925233644856,
+      "loss": 0.8583,
+      "step": 1013
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.631164014339447,
+      "learning_rate": 0.00021682242990654204,
+      "loss": 0.1627,
+      "step": 1014
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.1797211170196533,
+      "learning_rate": 0.0002167056074766355,
+      "loss": 0.259,
+      "step": 1015
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 3.3959903717041016,
+      "learning_rate": 0.00021658878504672894,
+      "loss": 0.4309,
+      "step": 1016
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.6847167015075684,
+      "learning_rate": 0.00021647196261682242,
+      "loss": 1.2204,
+      "step": 1017
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 1.9416961669921875,
+      "learning_rate": 0.00021635514018691587,
+      "loss": 1.2186,
+      "step": 1018
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.53370201587677,
+      "learning_rate": 0.00021623831775700934,
+      "loss": 0.3904,
+      "step": 1019
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.584384799003601,
+      "learning_rate": 0.0002161214953271028,
+      "loss": 1.0917,
+      "step": 1020
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.6039652824401855,
+      "learning_rate": 0.00021600467289719622,
+      "loss": 0.7606,
+      "step": 1021
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.1853108406066895,
+      "learning_rate": 0.0002158878504672897,
+      "loss": 0.717,
+      "step": 1022
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 4.574007987976074,
+      "learning_rate": 0.00021577102803738314,
+      "loss": 1.2157,
+      "step": 1023
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.5888893604278564,
+      "learning_rate": 0.00021565420560747662,
+      "loss": 1.4272,
+      "step": 1024
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 3.850660800933838,
+      "learning_rate": 0.00021553738317757007,
+      "loss": 0.6831,
+      "step": 1025
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 2.4923202991485596,
+      "learning_rate": 0.00021542056074766355,
+      "loss": 0.5352,
+      "step": 1026
+    },
+    {
+      "epoch": 1.07,
+      "grad_norm": 1.457641839981079,
+      "learning_rate": 0.000215303738317757,
+      "loss": 0.4242,
+      "step": 1027
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.6069073677062988,
+      "learning_rate": 0.00021518691588785047,
+      "loss": 0.5674,
+      "step": 1028
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.628026008605957,
+      "learning_rate": 0.0002150700934579439,
+      "loss": 0.9033,
+      "step": 1029
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.1619964838027954,
+      "learning_rate": 0.00021495327102803735,
+      "loss": 0.1613,
+      "step": 1030
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.593367099761963,
+      "learning_rate": 0.00021483644859813082,
+      "loss": 0.4673,
+      "step": 1031
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.325135588645935,
+      "learning_rate": 0.00021471962616822427,
+      "loss": 0.3284,
+      "step": 1032
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.5256247520446777,
+      "learning_rate": 0.00021460280373831775,
+      "loss": 0.9166,
+      "step": 1033
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.4277769327163696,
+      "learning_rate": 0.0002144859813084112,
+      "loss": 0.3661,
+      "step": 1034
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.2940905094146729,
+      "learning_rate": 0.00021436915887850465,
+      "loss": 0.6463,
+      "step": 1035
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 1.9980312585830688,
+      "learning_rate": 0.00021425233644859813,
+      "loss": 1.0043,
+      "step": 1036
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.1811680793762207,
+      "learning_rate": 0.00021413551401869155,
+      "loss": 0.414,
+      "step": 1037
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.60305655002594,
+      "learning_rate": 0.00021401869158878503,
+      "loss": 0.3784,
+      "step": 1038
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.7862429618835449,
+      "learning_rate": 0.00021390186915887848,
+      "loss": 0.2915,
+      "step": 1039
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.8023910522460938,
+      "learning_rate": 0.00021378504672897193,
+      "loss": 0.6625,
+      "step": 1040
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.0663235187530518,
+      "learning_rate": 0.0002136682242990654,
+      "loss": 1.3877,
+      "step": 1041
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.6616790890693665,
+      "learning_rate": 0.00021355140186915886,
+      "loss": 0.226,
+      "step": 1042
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.5566747188568115,
+      "learning_rate": 0.00021343457943925233,
+      "loss": 1.1761,
+      "step": 1043
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 1.1111620664596558,
+      "learning_rate": 0.00021331775700934578,
+      "loss": 0.4354,
+      "step": 1044
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 0.5970983505249023,
+      "learning_rate": 0.0002132009345794392,
+      "loss": 0.165,
+      "step": 1045
+    },
+    {
+      "epoch": 1.09,
+      "grad_norm": 2.0268211364746094,
+      "learning_rate": 0.00021308411214953268,
+      "loss": 0.6597,
+      "step": 1046
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.6138622760772705,
+      "learning_rate": 0.00021296728971962613,
+      "loss": 1.1535,
+      "step": 1047
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.9670050144195557,
+      "learning_rate": 0.0002128504672897196,
+      "loss": 1.0955,
+      "step": 1048
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 6.258568286895752,
+      "learning_rate": 0.00021273364485981306,
+      "loss": 1.0739,
+      "step": 1049
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.8623459339141846,
+      "learning_rate": 0.00021261682242990654,
+      "loss": 0.8344,
+      "step": 1050
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.135180950164795,
+      "learning_rate": 0.0002125,
+      "loss": 0.6828,
+      "step": 1051
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.476372480392456,
+      "learning_rate": 0.00021238317757009347,
+      "loss": 0.9149,
+      "step": 1052
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.1945157051086426,
+      "learning_rate": 0.0002122663551401869,
+      "loss": 0.701,
+      "step": 1053
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.106142520904541,
+      "learning_rate": 0.00021214953271028034,
+      "loss": 0.7747,
+      "step": 1054
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 1.4185359477996826,
+      "learning_rate": 0.00021203271028037382,
+      "loss": 0.8606,
+      "step": 1055
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.8948013782501221,
+      "learning_rate": 0.00021191588785046727,
+      "loss": 0.3924,
+      "step": 1056
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 4.456455230712891,
+      "learning_rate": 0.00021179906542056074,
+      "loss": 0.5109,
+      "step": 1057
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.550096273422241,
+      "learning_rate": 0.0002116822429906542,
+      "loss": 0.6182,
+      "step": 1058
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.1884052753448486,
+      "learning_rate": 0.00021156542056074764,
+      "loss": 1.2729,
+      "step": 1059
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 1.9526294469833374,
+      "learning_rate": 0.00021144859813084112,
+      "loss": 0.6765,
+      "step": 1060
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.172882318496704,
+      "learning_rate": 0.00021133177570093454,
+      "loss": 1.0866,
+      "step": 1061
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.258068799972534,
+      "learning_rate": 0.00021121495327102802,
+      "loss": 1.1084,
+      "step": 1062
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.720658302307129,
+      "learning_rate": 0.00021109813084112147,
+      "loss": 0.7547,
+      "step": 1063
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 0.6227821111679077,
+      "learning_rate": 0.00021098130841121492,
+      "loss": 0.2561,
+      "step": 1064
+    },
+    {
+      "epoch": 1.11,
+      "grad_norm": 2.7355456352233887,
+      "learning_rate": 0.0002108644859813084,
+      "loss": 1.2223,
+      "step": 1065
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.6639083623886108,
+      "learning_rate": 0.00021074766355140185,
+      "loss": 0.2132,
+      "step": 1066
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.391815185546875,
+      "learning_rate": 0.00021063084112149532,
+      "loss": 0.5128,
+      "step": 1067
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.4221978187561035,
+      "learning_rate": 0.00021051401869158877,
+      "loss": 1.4154,
+      "step": 1068
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.7561087608337402,
+      "learning_rate": 0.0002103971962616822,
+      "loss": 0.6037,
+      "step": 1069
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.95412540435791,
+      "learning_rate": 0.00021028037383177567,
+      "loss": 0.9519,
+      "step": 1070
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.575559139251709,
+      "learning_rate": 0.00021016355140186912,
+      "loss": 0.5565,
+      "step": 1071
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.1348252296447754,
+      "learning_rate": 0.0002100467289719626,
+      "loss": 1.2106,
+      "step": 1072
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.2099993228912354,
+      "learning_rate": 0.00020992990654205605,
+      "loss": 0.7627,
+      "step": 1073
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 3.261626958847046,
+      "learning_rate": 0.00020981308411214953,
+      "loss": 0.7699,
+      "step": 1074
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 1.9946436882019043,
+      "learning_rate": 0.00020969626168224298,
+      "loss": 0.4004,
+      "step": 1075
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.639751672744751,
+      "learning_rate": 0.00020957943925233646,
+      "loss": 0.8801,
+      "step": 1076
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.098203659057617,
+      "learning_rate": 0.00020946261682242988,
+      "loss": 0.6697,
+      "step": 1077
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.1346704959869385,
+      "learning_rate": 0.00020934579439252333,
+      "loss": 0.4131,
+      "step": 1078
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.8900938034057617,
+      "learning_rate": 0.0002092289719626168,
+      "loss": 0.7294,
+      "step": 1079
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.871983289718628,
+      "learning_rate": 0.00020911214953271026,
+      "loss": 1.1347,
+      "step": 1080
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.3414250612258911,
+      "learning_rate": 0.00020899532710280373,
+      "loss": 0.3847,
+      "step": 1081
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.448362350463867,
+      "learning_rate": 0.00020887850467289718,
+      "loss": 1.0003,
+      "step": 1082
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 2.3160276412963867,
+      "learning_rate": 0.00020876168224299063,
+      "loss": 1.2056,
+      "step": 1083
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 0.8403595089912415,
+      "learning_rate": 0.0002086448598130841,
+      "loss": 0.2933,
+      "step": 1084
+    },
+    {
+      "epoch": 1.13,
+      "grad_norm": 1.617903709411621,
+      "learning_rate": 0.00020852803738317753,
+      "loss": 0.3853,
+      "step": 1085
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.0691823959350586,
+      "learning_rate": 0.000208411214953271,
+      "loss": 0.7975,
+      "step": 1086
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.9060134887695312,
+      "learning_rate": 0.00020829439252336446,
+      "loss": 0.4843,
+      "step": 1087
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.8235893249511719,
+      "learning_rate": 0.0002081775700934579,
+      "loss": 0.8804,
+      "step": 1088
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.7594454288482666,
+      "learning_rate": 0.0002080607476635514,
+      "loss": 0.669,
+      "step": 1089
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 3.9268202781677246,
+      "learning_rate": 0.00020794392523364484,
+      "loss": 1.1165,
+      "step": 1090
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.8990890979766846,
+      "learning_rate": 0.00020782710280373831,
+      "loss": 0.6744,
+      "step": 1091
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.999157428741455,
+      "learning_rate": 0.00020771028037383176,
+      "loss": 1.1191,
+      "step": 1092
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 2.2413084506988525,
+      "learning_rate": 0.0002075934579439252,
+      "loss": 0.8147,
+      "step": 1093
+    },
+    {
+      "epoch": 1.14,
+      "grad_norm": 1.5006706714630127,
+      "learning_rate": 0.00020747663551401867,
+      "loss": 0.351,
+      "step": 1094
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.8998149633407593,
+      "learning_rate": 0.00020735981308411212,
+      "loss": 0.5434,
+      "step": 1095
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.9295514822006226,
+      "learning_rate": 0.0002072429906542056,
+      "loss": 0.308,
+      "step": 1096
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.4306630790233612,
+      "learning_rate": 0.00020712616822429904,
+      "loss": 0.1004,
+      "step": 1097
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.194831132888794,
+      "learning_rate": 0.00020700934579439252,
+      "loss": 0.4371,
+      "step": 1098
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.3609731197357178,
+      "learning_rate": 0.00020689252336448597,
+      "loss": 1.1762,
+      "step": 1099
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.4672547578811646,
+      "learning_rate": 0.00020677570093457945,
+      "loss": 1.0477,
+      "step": 1100
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 1.824455976486206,
+      "learning_rate": 0.00020665887850467287,
+      "loss": 0.5816,
+      "step": 1101
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 2.7696213722229004,
+      "learning_rate": 0.00020654205607476632,
+      "loss": 1.3223,
+      "step": 1102
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.8341705799102783,
+      "learning_rate": 0.0002064252336448598,
+      "loss": 0.28,
+      "step": 1103
+    },
+    {
+      "epoch": 1.15,
+      "grad_norm": 0.5999470353126526,
+      "learning_rate": 0.00020630841121495325,
+      "loss": 0.1999,
+      "step": 1104
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.682039499282837,
+      "learning_rate": 0.00020619158878504672,
+      "loss": 0.6262,
+      "step": 1105
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.5219595432281494,
+      "learning_rate": 0.00020607476635514017,
+      "loss": 1.2373,
+      "step": 1106
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.5517832040786743,
+      "learning_rate": 0.00020595794392523362,
+      "loss": 0.1942,
+      "step": 1107
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.029999017715454,
+      "learning_rate": 0.0002058411214953271,
+      "loss": 0.6479,
+      "step": 1108
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.3053765296936035,
+      "learning_rate": 0.00020572429906542052,
+      "loss": 0.3246,
+      "step": 1109
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 3.070949077606201,
+      "learning_rate": 0.000205607476635514,
+      "loss": 0.8793,
+      "step": 1110
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.68199223279953,
+      "learning_rate": 0.00020549065420560745,
+      "loss": 0.2444,
+      "step": 1111
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.8713790774345398,
+      "learning_rate": 0.0002053738317757009,
+      "loss": 0.4021,
+      "step": 1112
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.6102296710014343,
+      "learning_rate": 0.00020525700934579438,
+      "loss": 0.224,
+      "step": 1113
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.1383678913116455,
+      "learning_rate": 0.00020514018691588783,
+      "loss": 1.1279,
+      "step": 1114
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.3480324745178223,
+      "learning_rate": 0.0002050233644859813,
+      "loss": 0.7442,
+      "step": 1115
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.174382209777832,
+      "learning_rate": 0.00020490654205607476,
+      "loss": 1.1191,
+      "step": 1116
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.6875200271606445,
+      "learning_rate": 0.00020478971962616818,
+      "loss": 1.2578,
+      "step": 1117
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.9592370986938477,
+      "learning_rate": 0.00020467289719626166,
+      "loss": 0.2917,
+      "step": 1118
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 2.211015462875366,
+      "learning_rate": 0.0002045560747663551,
+      "loss": 0.4775,
+      "step": 1119
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.5659094452857971,
+      "learning_rate": 0.00020443925233644858,
+      "loss": 0.241,
+      "step": 1120
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.4829139709472656,
+      "learning_rate": 0.00020432242990654203,
+      "loss": 0.7289,
+      "step": 1121
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 0.6128184199333191,
+      "learning_rate": 0.0002042056074766355,
+      "loss": 0.2248,
+      "step": 1122
+    },
+    {
+      "epoch": 1.17,
+      "grad_norm": 3.0875678062438965,
+      "learning_rate": 0.00020408878504672896,
+      "loss": 0.7469,
+      "step": 1123
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.5821774005889893,
+      "learning_rate": 0.00020397196261682244,
+      "loss": 1.7178,
+      "step": 1124
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.0887813568115234,
+      "learning_rate": 0.00020385514018691586,
+      "loss": 0.5692,
+      "step": 1125
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 5.818206787109375,
+      "learning_rate": 0.0002037383177570093,
+      "loss": 0.7365,
+      "step": 1126
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.984973430633545,
+      "learning_rate": 0.0002036214953271028,
+      "loss": 1.0391,
+      "step": 1127
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.1518402099609375,
+      "learning_rate": 0.00020350467289719624,
+      "loss": 1.0598,
+      "step": 1128
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 1.685483694076538,
+      "learning_rate": 0.00020338785046728971,
+      "loss": 0.2998,
+      "step": 1129
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.082444429397583,
+      "learning_rate": 0.00020327102803738316,
+      "loss": 1.2295,
+      "step": 1130
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.6049225330352783,
+      "learning_rate": 0.00020315420560747661,
+      "loss": 0.9492,
+      "step": 1131
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 3.3788952827453613,
+      "learning_rate": 0.0002030373831775701,
+      "loss": 1.1305,
+      "step": 1132
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.523831367492676,
+      "learning_rate": 0.00020292056074766351,
+      "loss": 0.8953,
+      "step": 1133
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.302276134490967,
+      "learning_rate": 0.000202803738317757,
+      "loss": 1.136,
+      "step": 1134
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.4932010173797607,
+      "learning_rate": 0.00020268691588785044,
+      "loss": 1.1149,
+      "step": 1135
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 0.7620978355407715,
+      "learning_rate": 0.0002025700934579439,
+      "loss": 0.2841,
+      "step": 1136
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 3.206010103225708,
+      "learning_rate": 0.00020245327102803737,
+      "loss": 1.2515,
+      "step": 1137
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.4083549976348877,
+      "learning_rate": 0.00020233644859813082,
+      "loss": 1.551,
+      "step": 1138
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.192397952079773,
+      "learning_rate": 0.0002022196261682243,
+      "loss": 0.2744,
+      "step": 1139
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.2741410732269287,
+      "learning_rate": 0.00020210280373831775,
+      "loss": 1.2151,
+      "step": 1140
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 1.7759387493133545,
+      "learning_rate": 0.00020198598130841117,
+      "loss": 0.2124,
+      "step": 1141
+    },
+    {
+      "epoch": 1.19,
+      "grad_norm": 2.2041332721710205,
+      "learning_rate": 0.00020186915887850465,
+      "loss": 1.2591,
+      "step": 1142
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.2854342460632324,
+      "learning_rate": 0.0002017523364485981,
+      "loss": 0.6741,
+      "step": 1143
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.1802997589111328,
+      "learning_rate": 0.00020163551401869157,
+      "loss": 0.6383,
+      "step": 1144
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.4487531185150146,
+      "learning_rate": 0.00020151869158878502,
+      "loss": 1.0766,
+      "step": 1145
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.068978786468506,
+      "learning_rate": 0.0002014018691588785,
+      "loss": 1.0623,
+      "step": 1146
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.6643881797790527,
+      "learning_rate": 0.00020128504672897195,
+      "loss": 1.4822,
+      "step": 1147
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.279334306716919,
+      "learning_rate": 0.00020116822429906543,
+      "loss": 0.274,
+      "step": 1148
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.5919523239135742,
+      "learning_rate": 0.00020105140186915885,
+      "loss": 0.4191,
+      "step": 1149
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.5224316120147705,
+      "learning_rate": 0.0002009345794392523,
+      "loss": 1.18,
+      "step": 1150
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 3.791248321533203,
+      "learning_rate": 0.00020081775700934578,
+      "loss": 1.0462,
+      "step": 1151
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.8050941824913025,
+      "learning_rate": 0.00020070093457943923,
+      "loss": 0.3471,
+      "step": 1152
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.283918619155884,
+      "learning_rate": 0.0002005841121495327,
+      "loss": 1.1178,
+      "step": 1153
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.0043535232543945,
+      "learning_rate": 0.00020046728971962616,
+      "loss": 0.8636,
+      "step": 1154
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.247918128967285,
+      "learning_rate": 0.0002003504672897196,
+      "loss": 1.1561,
+      "step": 1155
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 3.544013023376465,
+      "learning_rate": 0.00020023364485981308,
+      "loss": 0.7743,
+      "step": 1156
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.5524381399154663,
+      "learning_rate": 0.0002001168224299065,
+      "loss": 0.1143,
+      "step": 1157
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.6980960369110107,
+      "learning_rate": 0.00019999999999999998,
+      "loss": 1.0201,
+      "step": 1158
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 0.896748423576355,
+      "learning_rate": 0.00019988317757009343,
+      "loss": 0.4344,
+      "step": 1159
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 1.7616729736328125,
+      "learning_rate": 0.00019976635514018688,
+      "loss": 0.4877,
+      "step": 1160
+    },
+    {
+      "epoch": 1.21,
+      "grad_norm": 2.6219723224639893,
+      "learning_rate": 0.00019964953271028036,
+      "loss": 1.5149,
+      "step": 1161
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.9278208613395691,
+      "learning_rate": 0.0001995327102803738,
+      "loss": 0.3392,
+      "step": 1162
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.1442532539367676,
+      "learning_rate": 0.0001994158878504673,
+      "loss": 0.8593,
+      "step": 1163
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.008396863937378,
+      "learning_rate": 0.00019929906542056074,
+      "loss": 0.3437,
+      "step": 1164
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.7166587114334106,
+      "learning_rate": 0.00019918224299065416,
+      "loss": 0.2603,
+      "step": 1165
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.835529088973999,
+      "learning_rate": 0.00019906542056074764,
+      "loss": 0.5484,
+      "step": 1166
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 1.9410781860351562,
+      "learning_rate": 0.0001989485981308411,
+      "loss": 0.9887,
+      "step": 1167
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.6602988243103027,
+      "learning_rate": 0.00019883177570093456,
+      "loss": 1.1274,
+      "step": 1168
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.1548686027526855,
+      "learning_rate": 0.00019871495327102801,
+      "loss": 1.0373,
+      "step": 1169
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.9196207523345947,
+      "learning_rate": 0.0001985981308411215,
+      "loss": 1.0006,
+      "step": 1170
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 3.27177095413208,
+      "learning_rate": 0.00019848130841121494,
+      "loss": 0.5959,
+      "step": 1171
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.277813196182251,
+      "learning_rate": 0.00019836448598130842,
+      "loss": 0.7807,
+      "step": 1172
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.142241954803467,
+      "learning_rate": 0.00019824766355140184,
+      "loss": 0.8082,
+      "step": 1173
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.313734292984009,
+      "learning_rate": 0.0001981308411214953,
+      "loss": 1.6829,
+      "step": 1174
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 1.6561822891235352,
+      "learning_rate": 0.00019801401869158877,
+      "loss": 0.5527,
+      "step": 1175
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 3.4537465572357178,
+      "learning_rate": 0.00019789719626168222,
+      "loss": 0.3977,
+      "step": 1176
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.100102663040161,
+      "learning_rate": 0.0001977803738317757,
+      "loss": 1.1519,
+      "step": 1177
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 0.8067005276679993,
+      "learning_rate": 0.00019766355140186915,
+      "loss": 0.292,
+      "step": 1178
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.6419880390167236,
+      "learning_rate": 0.0001975467289719626,
+      "loss": 0.8597,
+      "step": 1179
+    },
+    {
+      "epoch": 1.23,
+      "grad_norm": 2.25374174118042,
+      "learning_rate": 0.00019742990654205607,
+      "loss": 1.153,
+      "step": 1180
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.927171230316162,
+      "learning_rate": 0.0001973130841121495,
+      "loss": 0.7577,
+      "step": 1181
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.136739730834961,
+      "learning_rate": 0.00019719626168224297,
+      "loss": 0.434,
+      "step": 1182
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.9915034770965576,
+      "learning_rate": 0.00019707943925233642,
+      "loss": 0.5844,
+      "step": 1183
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.9362226128578186,
+      "learning_rate": 0.00019696261682242987,
+      "loss": 0.2859,
+      "step": 1184
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.459658622741699,
+      "learning_rate": 0.00019684579439252335,
+      "loss": 0.4476,
+      "step": 1185
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.027477502822876,
+      "learning_rate": 0.0001967289719626168,
+      "loss": 0.6242,
+      "step": 1186
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.1415551900863647,
+      "learning_rate": 0.00019661214953271028,
+      "loss": 0.4766,
+      "step": 1187
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 1.7337886095046997,
+      "learning_rate": 0.00019649532710280373,
+      "loss": 1.1372,
+      "step": 1188
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 3.035874366760254,
+      "learning_rate": 0.00019637850467289715,
+      "loss": 0.9431,
+      "step": 1189
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 2.2807509899139404,
+      "learning_rate": 0.00019626168224299063,
+      "loss": 1.0572,
+      "step": 1190
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.0541138648986816,
+      "learning_rate": 0.00019614485981308408,
+      "loss": 0.9536,
+      "step": 1191
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.9543030261993408,
+      "learning_rate": 0.00019602803738317755,
+      "loss": 0.1971,
+      "step": 1192
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.7279877662658691,
+      "learning_rate": 0.000195911214953271,
+      "loss": 0.4557,
+      "step": 1193
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.7344752550125122,
+      "learning_rate": 0.00019579439252336448,
+      "loss": 0.2517,
+      "step": 1194
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.9353537559509277,
+      "learning_rate": 0.00019567757009345793,
+      "loss": 0.6948,
+      "step": 1195
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 2.699579954147339,
+      "learning_rate": 0.0001955607476635514,
+      "loss": 1.4253,
+      "step": 1196
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 3.5536420345306396,
+      "learning_rate": 0.00019544392523364483,
+      "loss": 0.7727,
+      "step": 1197
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.5282950401306152,
+      "learning_rate": 0.00019532710280373828,
+      "loss": 1.0087,
+      "step": 1198
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 1.729099154472351,
+      "learning_rate": 0.00019521028037383176,
+      "loss": 0.5094,
+      "step": 1199
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 3.1926229000091553,
+      "learning_rate": 0.0001950934579439252,
+      "loss": 1.0685,
+      "step": 1200
+    },
+    {
+      "epoch": 1.26,
+      "eval_loss": 1.1658847332000732,
+      "eval_runtime": 1.0094,
+      "eval_samples_per_second": 4.954,
+      "eval_steps_per_second": 0.991,
+      "step": 1200
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7322415113449097,
+      "learning_rate": 0.0001949766355140187,
+      "loss": 0.2591,
+      "step": 1201
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.7986508011817932,
+      "learning_rate": 0.00019485981308411214,
+      "loss": 0.2158,
+      "step": 1202
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.1974411010742188,
+      "learning_rate": 0.0001947429906542056,
+      "loss": 0.5025,
+      "step": 1203
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.495664358139038,
+      "learning_rate": 0.00019462616822429906,
+      "loss": 0.6207,
+      "step": 1204
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.8991073369979858,
+      "learning_rate": 0.0001945093457943925,
+      "loss": 0.6394,
+      "step": 1205
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.0878992080688477,
+      "learning_rate": 0.00019439252336448596,
+      "loss": 0.4923,
+      "step": 1206
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 2.0958821773529053,
+      "learning_rate": 0.00019427570093457941,
+      "loss": 1.186,
+      "step": 1207
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 4.290122032165527,
+      "learning_rate": 0.00019415887850467286,
+      "loss": 1.005,
+      "step": 1208
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 3.115603446960449,
+      "learning_rate": 0.00019404205607476634,
+      "loss": 0.8839,
+      "step": 1209
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.7535462379455566,
+      "learning_rate": 0.0001939252336448598,
+      "loss": 0.419,
+      "step": 1210
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 4.466191291809082,
+      "learning_rate": 0.00019380841121495327,
+      "loss": 1.1126,
+      "step": 1211
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.9874646663665771,
+      "learning_rate": 0.00019369158878504672,
+      "loss": 0.8667,
+      "step": 1212
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 2.7608489990234375,
+      "learning_rate": 0.00019357476635514014,
+      "loss": 1.2938,
+      "step": 1213
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.9021081328392029,
+      "learning_rate": 0.00019345794392523362,
+      "loss": 0.4965,
+      "step": 1214
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.2415839433670044,
+      "learning_rate": 0.00019334112149532707,
+      "loss": 0.5533,
+      "step": 1215
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.8637053370475769,
+      "learning_rate": 0.00019322429906542055,
+      "loss": 0.3898,
+      "step": 1216
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 1.9663678407669067,
+      "learning_rate": 0.000193107476635514,
+      "loss": 1.0766,
+      "step": 1217
+    },
+    {
+      "epoch": 1.27,
+      "grad_norm": 0.6900842189788818,
+      "learning_rate": 0.00019299065420560747,
+      "loss": 0.2224,
+      "step": 1218
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5902005434036255,
+      "learning_rate": 0.00019287383177570092,
+      "loss": 0.1778,
+      "step": 1219
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.3478646278381348,
+      "learning_rate": 0.0001927570093457944,
+      "loss": 1.0055,
+      "step": 1220
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.5951477289199829,
+      "learning_rate": 0.00019264018691588782,
+      "loss": 0.2086,
+      "step": 1221
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.806227684020996,
+      "learning_rate": 0.00019252336448598127,
+      "loss": 0.6624,
+      "step": 1222
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 1.736450433731079,
+      "learning_rate": 0.00019240654205607475,
+      "loss": 1.4586,
+      "step": 1223
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.0015785694122314,
+      "learning_rate": 0.0001922897196261682,
+      "loss": 0.7516,
+      "step": 1224
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.632293701171875,
+      "learning_rate": 0.00019217289719626168,
+      "loss": 0.6397,
+      "step": 1225
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.11899471282959,
+      "learning_rate": 0.00019205607476635513,
+      "loss": 1.7155,
+      "step": 1226
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.5453994274139404,
+      "learning_rate": 0.00019193925233644858,
+      "loss": 0.5191,
+      "step": 1227
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.255399703979492,
+      "learning_rate": 0.00019182242990654205,
+      "loss": 0.6001,
+      "step": 1228
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.749321937561035,
+      "learning_rate": 0.00019170560747663548,
+      "loss": 0.7474,
+      "step": 1229
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.781606912612915,
+      "learning_rate": 0.00019158878504672895,
+      "loss": 0.7796,
+      "step": 1230
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.328237533569336,
+      "learning_rate": 0.0001914719626168224,
+      "loss": 1.1671,
+      "step": 1231
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.4334988594055176,
+      "learning_rate": 0.00019135514018691585,
+      "loss": 0.6105,
+      "step": 1232
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.9006870985031128,
+      "learning_rate": 0.00019123831775700933,
+      "loss": 0.4648,
+      "step": 1233
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.9073917865753174,
+      "learning_rate": 0.00019112149532710278,
+      "loss": 1.2155,
+      "step": 1234
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.242873430252075,
+      "learning_rate": 0.00019100467289719626,
+      "loss": 0.6119,
+      "step": 1235
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.9779205322265625,
+      "learning_rate": 0.0001908878504672897,
+      "loss": 0.2464,
+      "step": 1236
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 2.8669989109039307,
+      "learning_rate": 0.00019077102803738313,
+      "loss": 0.7974,
+      "step": 1237
+    },
+    {
+      "epoch": 1.29,
+      "grad_norm": 1.161089539527893,
+      "learning_rate": 0.0001906542056074766,
+      "loss": 0.1048,
+      "step": 1238
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 4.102783679962158,
+      "learning_rate": 0.00019053738317757006,
+      "loss": 0.9183,
+      "step": 1239
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.924055814743042,
+      "learning_rate": 0.00019042056074766354,
+      "loss": 1.4965,
+      "step": 1240
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.143636465072632,
+      "learning_rate": 0.00019030373831775699,
+      "loss": 0.2837,
+      "step": 1241
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.9068852663040161,
+      "learning_rate": 0.00019018691588785046,
+      "loss": 0.9671,
+      "step": 1242
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.4567501544952393,
+      "learning_rate": 0.0001900700934579439,
+      "loss": 0.3306,
+      "step": 1243
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.6074962615966797,
+      "learning_rate": 0.0001899532710280374,
+      "loss": 1.601,
+      "step": 1244
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.3963537216186523,
+      "learning_rate": 0.0001898364485981308,
+      "loss": 1.1746,
+      "step": 1245
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 1.6171189546585083,
+      "learning_rate": 0.00018971962616822426,
+      "loss": 0.349,
+      "step": 1246
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.6384449005126953,
+      "learning_rate": 0.00018960280373831774,
+      "loss": 0.8075,
+      "step": 1247
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.3255858421325684,
+      "learning_rate": 0.0001894859813084112,
+      "loss": 1.1681,
+      "step": 1248
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.8400274515151978,
+      "learning_rate": 0.00018936915887850467,
+      "loss": 0.3144,
+      "step": 1249
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.320280075073242,
+      "learning_rate": 0.00018925233644859812,
+      "loss": 0.6442,
+      "step": 1250
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 0.9876735210418701,
+      "learning_rate": 0.00018913551401869157,
+      "loss": 0.1542,
+      "step": 1251
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.2251205444335938,
+      "learning_rate": 0.00018901869158878504,
+      "loss": 0.9543,
+      "step": 1252
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 1.0860331058502197,
+      "learning_rate": 0.00018890186915887847,
+      "loss": 0.6192,
+      "step": 1253
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 4.4009881019592285,
+      "learning_rate": 0.00018878504672897195,
+      "loss": 0.9997,
+      "step": 1254
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.2344605922698975,
+      "learning_rate": 0.0001886682242990654,
+      "loss": 0.5521,
+      "step": 1255
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.8071048259735107,
+      "learning_rate": 0.00018855140186915885,
+      "loss": 0.3418,
+      "step": 1256
+    },
+    {
+      "epoch": 1.31,
+      "grad_norm": 2.5023577213287354,
+      "learning_rate": 0.00018843457943925232,
+      "loss": 0.3715,
+      "step": 1257
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.644026517868042,
+      "learning_rate": 0.00018831775700934577,
+      "loss": 0.525,
+      "step": 1258
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.5025291442871094,
+      "learning_rate": 0.00018820093457943925,
+      "loss": 1.0208,
+      "step": 1259
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.4263161420822144,
+      "learning_rate": 0.0001880841121495327,
+      "loss": 0.3544,
+      "step": 1260
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.8181207180023193,
+      "learning_rate": 0.00018796728971962612,
+      "loss": 0.5881,
+      "step": 1261
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.331484079360962,
+      "learning_rate": 0.0001878504672897196,
+      "loss": 0.4943,
+      "step": 1262
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.104075312614441,
+      "learning_rate": 0.00018773364485981305,
+      "loss": 0.225,
+      "step": 1263
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.39052748680114746,
+      "learning_rate": 0.00018761682242990653,
+      "loss": 0.1284,
+      "step": 1264
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.3384759426116943,
+      "learning_rate": 0.00018749999999999998,
+      "loss": 0.291,
+      "step": 1265
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 1.8362799882888794,
+      "learning_rate": 0.00018738317757009345,
+      "loss": 0.7755,
+      "step": 1266
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.271969795227051,
+      "learning_rate": 0.0001872663551401869,
+      "loss": 0.377,
+      "step": 1267
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.036214828491211,
+      "learning_rate": 0.00018714953271028038,
+      "loss": 0.5356,
+      "step": 1268
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 4.588119029998779,
+      "learning_rate": 0.00018703271028037383,
+      "loss": 1.0096,
+      "step": 1269
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 1.5631695985794067,
+      "learning_rate": 0.00018691588785046725,
+      "loss": 0.3514,
+      "step": 1270
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.6622061729431152,
+      "learning_rate": 0.00018679906542056073,
+      "loss": 0.8358,
+      "step": 1271
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 4.314937114715576,
+      "learning_rate": 0.00018668224299065418,
+      "loss": 0.9017,
+      "step": 1272
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.548337459564209,
+      "learning_rate": 0.00018656542056074766,
+      "loss": 1.1176,
+      "step": 1273
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.879739284515381,
+      "learning_rate": 0.0001864485981308411,
+      "loss": 0.7511,
+      "step": 1274
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.590949535369873,
+      "learning_rate": 0.00018633177570093456,
+      "loss": 0.8095,
+      "step": 1275
+    },
+    {
+      "epoch": 1.33,
+      "grad_norm": 2.957411766052246,
+      "learning_rate": 0.00018621495327102804,
+      "loss": 0.7659,
+      "step": 1276
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.2279860973358154,
+      "learning_rate": 0.00018609813084112149,
+      "loss": 0.9038,
+      "step": 1277
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.5968736410140991,
+      "learning_rate": 0.00018598130841121494,
+      "loss": 0.1685,
+      "step": 1278
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 3.183213472366333,
+      "learning_rate": 0.00018586448598130839,
+      "loss": 1.0396,
+      "step": 1279
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.9355374574661255,
+      "learning_rate": 0.00018574766355140184,
+      "loss": 0.5406,
+      "step": 1280
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 5.673951625823975,
+      "learning_rate": 0.0001856308411214953,
+      "loss": 1.2144,
+      "step": 1281
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 3.6509315967559814,
+      "learning_rate": 0.00018551401869158876,
+      "loss": 1.406,
+      "step": 1282
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.6314842700958252,
+      "learning_rate": 0.00018539719626168224,
+      "loss": 0.2197,
+      "step": 1283
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.440687894821167,
+      "learning_rate": 0.0001852803738317757,
+      "loss": 0.7316,
+      "step": 1284
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 2.5698628425598145,
+      "learning_rate": 0.00018516355140186917,
+      "loss": 0.4897,
+      "step": 1285
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.361003875732422,
+      "learning_rate": 0.0001850467289719626,
+      "loss": 0.7666,
+      "step": 1286
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 3.423429489135742,
+      "learning_rate": 0.00018492990654205604,
+      "loss": 0.7807,
+      "step": 1287
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 3.4809789657592773,
+      "learning_rate": 0.00018481308411214952,
+      "loss": 0.5707,
+      "step": 1288
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.4859747886657715,
+      "learning_rate": 0.00018469626168224297,
+      "loss": 1.6028,
+      "step": 1289
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.108879566192627,
+      "learning_rate": 0.00018457943925233644,
+      "loss": 0.545,
+      "step": 1290
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.270004987716675,
+      "learning_rate": 0.0001844626168224299,
+      "loss": 0.6226,
+      "step": 1291
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.700712203979492,
+      "learning_rate": 0.00018434579439252337,
+      "loss": 0.8702,
+      "step": 1292
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.3957457542419434,
+      "learning_rate": 0.00018422897196261682,
+      "loss": 1.3282,
+      "step": 1293
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 2.5982956886291504,
+      "learning_rate": 0.00018411214953271024,
+      "loss": 0.6065,
+      "step": 1294
+    },
+    {
+      "epoch": 1.35,
+      "grad_norm": 1.1615874767303467,
+      "learning_rate": 0.00018399532710280372,
+      "loss": 0.6769,
+      "step": 1295
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.294102191925049,
+      "learning_rate": 0.00018387850467289717,
+      "loss": 0.4803,
+      "step": 1296
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.3875255584716797,
+      "learning_rate": 0.00018376168224299065,
+      "loss": 0.7764,
+      "step": 1297
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.732203483581543,
+      "learning_rate": 0.0001836448598130841,
+      "loss": 0.768,
+      "step": 1298
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.0977386236190796,
+      "learning_rate": 0.00018352803738317755,
+      "loss": 0.2692,
+      "step": 1299
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 0.6086471676826477,
+      "learning_rate": 0.00018341121495327103,
+      "loss": 0.1623,
+      "step": 1300
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.480614185333252,
+      "learning_rate": 0.00018329439252336448,
+      "loss": 1.3882,
+      "step": 1301
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 2.656018018722534,
+      "learning_rate": 0.00018317757009345793,
+      "loss": 0.8735,
+      "step": 1302
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 3.2506661415100098,
+      "learning_rate": 0.00018306074766355138,
+      "loss": 0.8242,
+      "step": 1303
+    },
+    {
+      "epoch": 1.36,
+      "grad_norm": 1.9476683139801025,
+      "learning_rate": 0.00018294392523364483,
+      "loss": 0.8056,
+      "step": 1304
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 3.344555616378784,
+      "learning_rate": 0.0001828271028037383,
+      "loss": 0.9017,
+      "step": 1305
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.0389673709869385,
+      "learning_rate": 0.00018271028037383175,
+      "loss": 0.9627,
+      "step": 1306
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.904149055480957,
+      "learning_rate": 0.00018259345794392523,
+      "loss": 0.9032,
+      "step": 1307
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.444079875946045,
+      "learning_rate": 0.00018247663551401868,
+      "loss": 1.5189,
+      "step": 1308
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 4.824917316436768,
+      "learning_rate": 0.00018235981308411216,
+      "loss": 1.1401,
+      "step": 1309
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.3599419593811035,
+      "learning_rate": 0.00018224299065420558,
+      "loss": 0.3766,
+      "step": 1310
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.9183833599090576,
+      "learning_rate": 0.00018212616822429903,
+      "loss": 0.5002,
+      "step": 1311
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 2.920369863510132,
+      "learning_rate": 0.0001820093457943925,
+      "loss": 0.6568,
+      "step": 1312
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 3.70578932762146,
+      "learning_rate": 0.00018189252336448596,
+      "loss": 1.8708,
+      "step": 1313
+    },
+    {
+      "epoch": 1.37,
+      "grad_norm": 4.370184898376465,
+      "learning_rate": 0.00018177570093457944,
+      "loss": 1.1221,
+      "step": 1314
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.540465831756592,
+      "learning_rate": 0.00018165887850467289,
+      "loss": 1.241,
+      "step": 1315
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.349395990371704,
+      "learning_rate": 0.00018154205607476636,
+      "loss": 0.754,
+      "step": 1316
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.9466333389282227,
+      "learning_rate": 0.0001814252336448598,
+      "loss": 0.5155,
+      "step": 1317
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.9178152084350586,
+      "learning_rate": 0.00018130841121495324,
+      "loss": 0.8991,
+      "step": 1318
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.0130553245544434,
+      "learning_rate": 0.0001811915887850467,
+      "loss": 1.0502,
+      "step": 1319
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.6518648862838745,
+      "learning_rate": 0.00018107476635514016,
+      "loss": 0.3683,
+      "step": 1320
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.7037773132324219,
+      "learning_rate": 0.00018095794392523364,
+      "loss": 0.6605,
+      "step": 1321
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.6347455978393555,
+      "learning_rate": 0.0001808411214953271,
+      "loss": 0.5213,
+      "step": 1322
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 3.909637451171875,
+      "learning_rate": 0.00018072429906542054,
+      "loss": 1.0715,
+      "step": 1323
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 1.7166997194290161,
+      "learning_rate": 0.00018060747663551402,
+      "loss": 0.4536,
+      "step": 1324
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.9127898216247559,
+      "learning_rate": 0.00018049065420560747,
+      "loss": 0.5231,
+      "step": 1325
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.906681537628174,
+      "learning_rate": 0.00018037383177570092,
+      "loss": 0.8225,
+      "step": 1326
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.234952926635742,
+      "learning_rate": 0.00018025700934579437,
+      "loss": 0.9737,
+      "step": 1327
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.5723302364349365,
+      "learning_rate": 0.00018014018691588782,
+      "loss": 1.3425,
+      "step": 1328
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.9902420043945312,
+      "learning_rate": 0.0001800233644859813,
+      "loss": 1.6166,
+      "step": 1329
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.5610482692718506,
+      "learning_rate": 0.00017990654205607474,
+      "loss": 0.6638,
+      "step": 1330
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 2.5049209594726562,
+      "learning_rate": 0.00017978971962616822,
+      "loss": 0.6624,
+      "step": 1331
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.2319400310516357,
+      "learning_rate": 0.00017967289719626167,
+      "loss": 0.6184,
+      "step": 1332
+    },
+    {
+      "epoch": 1.39,
+      "grad_norm": 1.0322645902633667,
+      "learning_rate": 0.00017955607476635515,
+      "loss": 0.3184,
+      "step": 1333
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.68593430519104,
+      "learning_rate": 0.00017943925233644857,
+      "loss": 1.3337,
+      "step": 1334
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 3.216642141342163,
+      "learning_rate": 0.00017932242990654202,
+      "loss": 1.3633,
+      "step": 1335
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.0757498741149902,
+      "learning_rate": 0.0001792056074766355,
+      "loss": 0.5981,
+      "step": 1336
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.787240743637085,
+      "learning_rate": 0.00017908878504672895,
+      "loss": 0.8051,
+      "step": 1337
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 4.88184928894043,
+      "learning_rate": 0.00017897196261682243,
+      "loss": 1.4034,
+      "step": 1338
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.5093235969543457,
+      "learning_rate": 0.00017885514018691588,
+      "loss": 0.151,
+      "step": 1339
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.8691405653953552,
+      "learning_rate": 0.00017873831775700935,
+      "loss": 0.3816,
+      "step": 1340
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.2686407566070557,
+      "learning_rate": 0.0001786214953271028,
+      "loss": 0.403,
+      "step": 1341
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 3.1247494220733643,
+      "learning_rate": 0.00017850467289719623,
+      "loss": 1.461,
+      "step": 1342
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 2.5943002700805664,
+      "learning_rate": 0.0001783878504672897,
+      "loss": 0.8829,
+      "step": 1343
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.44015634059906,
+      "learning_rate": 0.00017827102803738315,
+      "loss": 0.2918,
+      "step": 1344
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 1.8427207469940186,
+      "learning_rate": 0.00017815420560747663,
+      "loss": 0.5179,
+      "step": 1345
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.8338239192962646,
+      "learning_rate": 0.00017803738317757008,
+      "loss": 0.7246,
+      "step": 1346
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.8842005133628845,
+      "learning_rate": 0.00017792056074766353,
+      "loss": 0.3522,
+      "step": 1347
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.3126587867736816,
+      "learning_rate": 0.000177803738317757,
+      "loss": 1.291,
+      "step": 1348
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.49202239513397217,
+      "learning_rate": 0.00017768691588785046,
+      "loss": 0.1394,
+      "step": 1349
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.2132513523101807,
+      "learning_rate": 0.0001775700934579439,
+      "loss": 1.1189,
+      "step": 1350
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 0.8676378726959229,
+      "learning_rate": 0.00017745327102803736,
+      "loss": 0.3637,
+      "step": 1351
+    },
+    {
+      "epoch": 1.41,
+      "grad_norm": 2.5113818645477295,
+      "learning_rate": 0.0001773364485981308,
+      "loss": 0.701,
+      "step": 1352
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.5786195397377014,
+      "learning_rate": 0.00017721962616822428,
+      "loss": 0.1862,
+      "step": 1353
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.432063341140747,
+      "learning_rate": 0.00017710280373831773,
+      "loss": 0.3464,
+      "step": 1354
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 3.4058117866516113,
+      "learning_rate": 0.0001769859813084112,
+      "loss": 0.6988,
+      "step": 1355
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.8814724683761597,
+      "learning_rate": 0.00017686915887850466,
+      "loss": 0.5569,
+      "step": 1356
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.372424364089966,
+      "learning_rate": 0.00017675233644859814,
+      "loss": 0.8539,
+      "step": 1357
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.3942906856536865,
+      "learning_rate": 0.00017663551401869156,
+      "loss": 0.086,
+      "step": 1358
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 4.250904083251953,
+      "learning_rate": 0.000176518691588785,
+      "loss": 0.9614,
+      "step": 1359
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 1.135056972503662,
+      "learning_rate": 0.0001764018691588785,
+      "loss": 0.516,
+      "step": 1360
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.6572325825691223,
+      "learning_rate": 0.00017628504672897194,
+      "loss": 0.1664,
+      "step": 1361
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 2.5226712226867676,
+      "learning_rate": 0.00017616822429906542,
+      "loss": 0.893,
+      "step": 1362
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.456709623336792,
+      "learning_rate": 0.00017605140186915887,
+      "loss": 0.8222,
+      "step": 1363
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.7911829352378845,
+      "learning_rate": 0.00017593457943925234,
+      "loss": 0.3,
+      "step": 1364
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 3.0613296031951904,
+      "learning_rate": 0.0001758177570093458,
+      "loss": 1.4222,
+      "step": 1365
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.1822903156280518,
+      "learning_rate": 0.00017570093457943922,
+      "loss": 0.4439,
+      "step": 1366
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.6135878562927246,
+      "learning_rate": 0.0001755841121495327,
+      "loss": 0.2818,
+      "step": 1367
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 3.2960660457611084,
+      "learning_rate": 0.00017546728971962614,
+      "loss": 0.5994,
+      "step": 1368
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 0.9349135160446167,
+      "learning_rate": 0.00017535046728971962,
+      "loss": 0.3976,
+      "step": 1369
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.733487129211426,
+      "learning_rate": 0.00017523364485981307,
+      "loss": 0.8686,
+      "step": 1370
+    },
+    {
+      "epoch": 1.43,
+      "grad_norm": 2.251721143722534,
+      "learning_rate": 0.00017511682242990652,
+      "loss": 1.0455,
+      "step": 1371
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.2353663444519043,
+      "learning_rate": 0.000175,
+      "loss": 0.9254,
+      "step": 1372
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 4.86463737487793,
+      "learning_rate": 0.00017488317757009345,
+      "loss": 1.1722,
+      "step": 1373
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.4062563478946686,
+      "learning_rate": 0.0001747663551401869,
+      "loss": 0.07,
+      "step": 1374
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 4.017411708831787,
+      "learning_rate": 0.00017464953271028035,
+      "loss": 0.6862,
+      "step": 1375
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 3.2097668647766113,
+      "learning_rate": 0.0001745327102803738,
+      "loss": 0.6583,
+      "step": 1376
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 3.701545238494873,
+      "learning_rate": 0.00017441588785046728,
+      "loss": 1.1785,
+      "step": 1377
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.956078290939331,
+      "learning_rate": 0.00017429906542056073,
+      "loss": 0.7798,
+      "step": 1378
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 4.492142677307129,
+      "learning_rate": 0.0001741822429906542,
+      "loss": 1.1939,
+      "step": 1379
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 3.10951828956604,
+      "learning_rate": 0.00017406542056074765,
+      "loss": 0.6798,
+      "step": 1380
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 3.4178178310394287,
+      "learning_rate": 0.00017394859813084113,
+      "loss": 1.7696,
+      "step": 1381
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.3063645362854004,
+      "learning_rate": 0.00017383177570093455,
+      "loss": 0.6285,
+      "step": 1382
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.3473939895629883,
+      "learning_rate": 0.000173714953271028,
+      "loss": 0.9029,
+      "step": 1383
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.47164860367774963,
+      "learning_rate": 0.00017359813084112148,
+      "loss": 0.1249,
+      "step": 1384
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.301178216934204,
+      "learning_rate": 0.00017348130841121493,
+      "loss": 1.1899,
+      "step": 1385
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.9900512099266052,
+      "learning_rate": 0.0001733644859813084,
+      "loss": 0.4836,
+      "step": 1386
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 0.7707881927490234,
+      "learning_rate": 0.00017324766355140186,
+      "loss": 0.3347,
+      "step": 1387
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.288416862487793,
+      "learning_rate": 0.00017313084112149533,
+      "loss": 1.1194,
+      "step": 1388
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 1.8858298063278198,
+      "learning_rate": 0.00017301401869158878,
+      "loss": 0.4368,
+      "step": 1389
+    },
+    {
+      "epoch": 1.45,
+      "grad_norm": 2.7885446548461914,
+      "learning_rate": 0.0001728971962616822,
+      "loss": 1.2908,
+      "step": 1390
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 3.9828786849975586,
+      "learning_rate": 0.00017278037383177568,
+      "loss": 0.9096,
+      "step": 1391
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.8363916873931885,
+      "learning_rate": 0.00017266355140186913,
+      "loss": 1.0472,
+      "step": 1392
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.291325092315674,
+      "learning_rate": 0.0001725467289719626,
+      "loss": 0.6374,
+      "step": 1393
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.0536561012268066,
+      "learning_rate": 0.00017242990654205606,
+      "loss": 0.5722,
+      "step": 1394
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.9046696424484253,
+      "learning_rate": 0.0001723130841121495,
+      "loss": 0.6212,
+      "step": 1395
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 3.9934444427490234,
+      "learning_rate": 0.000172196261682243,
+      "loss": 0.7917,
+      "step": 1396
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.6799383163452148,
+      "learning_rate": 0.00017207943925233644,
+      "loss": 0.3291,
+      "step": 1397
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.4810992181301117,
+      "learning_rate": 0.0001719626168224299,
+      "loss": 0.174,
+      "step": 1398
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.9290904998779297,
+      "learning_rate": 0.00017184579439252334,
+      "loss": 1.0489,
+      "step": 1399
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 1.0876963138580322,
+      "learning_rate": 0.0001717289719626168,
+      "loss": 0.5794,
+      "step": 1400
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.6646468043327332,
+      "learning_rate": 0.00017161214953271027,
+      "loss": 0.3333,
+      "step": 1401
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.0397571325302124,
+      "learning_rate": 0.00017149532710280372,
+      "loss": 0.5403,
+      "step": 1402
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 3.3683550357818604,
+      "learning_rate": 0.0001713785046728972,
+      "loss": 0.9768,
+      "step": 1403
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 1.7085354328155518,
+      "learning_rate": 0.00017126168224299064,
+      "loss": 0.8301,
+      "step": 1404
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 0.9951138496398926,
+      "learning_rate": 0.00017114485981308412,
+      "loss": 0.5877,
+      "step": 1405
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.1947481632232666,
+      "learning_rate": 0.00017102803738317754,
+      "loss": 1.3012,
+      "step": 1406
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.249422311782837,
+      "learning_rate": 0.000170911214953271,
+      "loss": 0.5428,
+      "step": 1407
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.176011562347412,
+      "learning_rate": 0.00017079439252336447,
+      "loss": 0.6749,
+      "step": 1408
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 2.369123935699463,
+      "learning_rate": 0.00017067757009345792,
+      "loss": 0.8806,
+      "step": 1409
+    },
+    {
+      "epoch": 1.47,
+      "grad_norm": 3.223184823989868,
+      "learning_rate": 0.0001705607476635514,
+      "loss": 0.978,
+      "step": 1410
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.7362747192382812,
+      "learning_rate": 0.00017044392523364485,
+      "loss": 0.4143,
+      "step": 1411
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.9589636325836182,
+      "learning_rate": 0.00017032710280373833,
+      "loss": 0.9767,
+      "step": 1412
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.106274127960205,
+      "learning_rate": 0.00017021028037383178,
+      "loss": 0.5638,
+      "step": 1413
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.9418315887451172,
+      "learning_rate": 0.0001700934579439252,
+      "loss": 0.9572,
+      "step": 1414
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.3618438243865967,
+      "learning_rate": 0.00016997663551401868,
+      "loss": 0.673,
+      "step": 1415
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.0478830337524414,
+      "learning_rate": 0.00016985981308411213,
+      "loss": 0.5286,
+      "step": 1416
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 3.376671314239502,
+      "learning_rate": 0.0001697429906542056,
+      "loss": 1.1043,
+      "step": 1417
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 1.0522353649139404,
+      "learning_rate": 0.00016962616822429905,
+      "loss": 0.0854,
+      "step": 1418
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.4326811730861664,
+      "learning_rate": 0.0001695093457943925,
+      "loss": 0.118,
+      "step": 1419
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 3.028013229370117,
+      "learning_rate": 0.00016939252336448598,
+      "loss": 0.7437,
+      "step": 1420
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 2.15781307220459,
+      "learning_rate": 0.00016927570093457943,
+      "loss": 0.9347,
+      "step": 1421
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 2.592618465423584,
+      "learning_rate": 0.00016915887850467288,
+      "loss": 0.9163,
+      "step": 1422
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 2.283153533935547,
+      "learning_rate": 0.00016904205607476633,
+      "loss": 1.0058,
+      "step": 1423
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.3824974298477173,
+      "learning_rate": 0.00016892523364485978,
+      "loss": 0.8266,
+      "step": 1424
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 3.0501463413238525,
+      "learning_rate": 0.00016880841121495326,
+      "loss": 1.4845,
+      "step": 1425
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.2995612621307373,
+      "learning_rate": 0.0001686915887850467,
+      "loss": 0.1369,
+      "step": 1426
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 1.060412049293518,
+      "learning_rate": 0.00016857476635514018,
+      "loss": 0.2031,
+      "step": 1427
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 2.402827262878418,
+      "learning_rate": 0.00016845794392523363,
+      "loss": 0.6165,
+      "step": 1428
+    },
+    {
+      "epoch": 1.49,
+      "grad_norm": 2.1554033756256104,
+      "learning_rate": 0.0001683411214953271,
+      "loss": 0.9243,
+      "step": 1429
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.8789975047111511,
+      "learning_rate": 0.00016822429906542053,
+      "loss": 0.3416,
+      "step": 1430
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 3.1596078872680664,
+      "learning_rate": 0.00016810747663551398,
+      "loss": 0.8469,
+      "step": 1431
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.5885155200958252,
+      "learning_rate": 0.00016799065420560746,
+      "loss": 0.4804,
+      "step": 1432
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.705204427242279,
+      "learning_rate": 0.0001678738317757009,
+      "loss": 0.2568,
+      "step": 1433
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.0723459720611572,
+      "learning_rate": 0.0001677570093457944,
+      "loss": 0.5643,
+      "step": 1434
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.1040033102035522,
+      "learning_rate": 0.00016764018691588784,
+      "loss": 0.4714,
+      "step": 1435
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 6.07669734954834,
+      "learning_rate": 0.00016752336448598132,
+      "loss": 1.2381,
+      "step": 1436
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 1.854767918586731,
+      "learning_rate": 0.00016740654205607477,
+      "loss": 0.5169,
+      "step": 1437
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.84541916847229,
+      "learning_rate": 0.0001672897196261682,
+      "loss": 0.7201,
+      "step": 1438
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.7634392976760864,
+      "learning_rate": 0.00016717289719626167,
+      "loss": 0.4629,
+      "step": 1439
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 2.0387461185455322,
+      "learning_rate": 0.00016705607476635512,
+      "loss": 0.7306,
+      "step": 1440
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.1351913213729858,
+      "learning_rate": 0.0001669392523364486,
+      "loss": 0.1392,
+      "step": 1441
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 3.165987253189087,
+      "learning_rate": 0.00016682242990654204,
+      "loss": 1.0364,
+      "step": 1442
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 0.5354011654853821,
+      "learning_rate": 0.0001667056074766355,
+      "loss": 0.1786,
+      "step": 1443
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.6701676845550537,
+      "learning_rate": 0.00016658878504672897,
+      "loss": 0.3902,
+      "step": 1444
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 2.9532558917999268,
+      "learning_rate": 0.00016647196261682242,
+      "loss": 1.308,
+      "step": 1445
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 3.9400277137756348,
+      "learning_rate": 0.00016635514018691587,
+      "loss": 0.8879,
+      "step": 1446
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 2.938701629638672,
+      "learning_rate": 0.00016623831775700932,
+      "loss": 1.1334,
+      "step": 1447
+    },
+    {
+      "epoch": 1.51,
+      "grad_norm": 1.7852617502212524,
+      "learning_rate": 0.00016612149532710277,
+      "loss": 0.5887,
+      "step": 1448
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 1.42865788936615,
+      "learning_rate": 0.00016600467289719625,
+      "loss": 0.276,
+      "step": 1449
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 2.0600006580352783,
+      "learning_rate": 0.0001658878504672897,
+      "loss": 0.6751,
+      "step": 1450
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6229619383811951,
+      "learning_rate": 0.00016577102803738317,
+      "loss": 0.1782,
+      "step": 1451
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.6343135833740234,
+      "learning_rate": 0.00016565420560747662,
+      "loss": 0.2434,
+      "step": 1452
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 3.1584959030151367,
+      "learning_rate": 0.0001655373831775701,
+      "loss": 0.738,
+      "step": 1453
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.9337627291679382,
+      "learning_rate": 0.00016542056074766352,
+      "loss": 0.4132,
+      "step": 1454
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 2.4795186519622803,
+      "learning_rate": 0.00016530373831775697,
+      "loss": 1.1011,
+      "step": 1455
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 2.3030202388763428,
+      "learning_rate": 0.00016518691588785045,
+      "loss": 0.31,
+      "step": 1456
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 3.6737964153289795,
+      "learning_rate": 0.0001650700934579439,
+      "loss": 1.5343,
+      "step": 1457
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 4.254706382751465,
+      "learning_rate": 0.00016495327102803738,
+      "loss": 0.8686,
+      "step": 1458
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.095659852027893,
+      "learning_rate": 0.00016483644859813083,
+      "loss": 0.552,
+      "step": 1459
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 2.9776525497436523,
+      "learning_rate": 0.0001647196261682243,
+      "loss": 1.247,
+      "step": 1460
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.611816167831421,
+      "learning_rate": 0.00016460280373831776,
+      "loss": 1.2273,
+      "step": 1461
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.910477876663208,
+      "learning_rate": 0.00016448598130841118,
+      "loss": 1.2431,
+      "step": 1462
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 4.306861877441406,
+      "learning_rate": 0.00016436915887850466,
+      "loss": 1.3482,
+      "step": 1463
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 3.019909381866455,
+      "learning_rate": 0.0001642523364485981,
+      "loss": 1.1731,
+      "step": 1464
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 2.6072418689727783,
+      "learning_rate": 0.00016413551401869158,
+      "loss": 0.6615,
+      "step": 1465
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 2.3710482120513916,
+      "learning_rate": 0.00016401869158878503,
+      "loss": 0.5472,
+      "step": 1466
+    },
+    {
+      "epoch": 1.53,
+      "grad_norm": 1.8100333213806152,
+      "learning_rate": 0.00016390186915887848,
+      "loss": 0.8855,
+      "step": 1467
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.7441534996032715,
+      "learning_rate": 0.00016378504672897196,
+      "loss": 1.5301,
+      "step": 1468
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 1.9324244260787964,
+      "learning_rate": 0.0001636682242990654,
+      "loss": 1.0325,
+      "step": 1469
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 4.5111565589904785,
+      "learning_rate": 0.00016355140186915886,
+      "loss": 0.9948,
+      "step": 1470
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.6447874903678894,
+      "learning_rate": 0.0001634345794392523,
+      "loss": 0.1443,
+      "step": 1471
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.870194673538208,
+      "learning_rate": 0.00016331775700934576,
+      "loss": 0.5645,
+      "step": 1472
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.656026303768158,
+      "learning_rate": 0.00016320093457943924,
+      "loss": 0.2724,
+      "step": 1473
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.9519236087799072,
+      "learning_rate": 0.0001630841121495327,
+      "loss": 0.2745,
+      "step": 1474
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.730017364025116,
+      "learning_rate": 0.00016296728971962617,
+      "loss": 0.1684,
+      "step": 1475
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.9128291606903076,
+      "learning_rate": 0.00016285046728971962,
+      "loss": 0.7495,
+      "step": 1476
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.1994237899780273,
+      "learning_rate": 0.0001627336448598131,
+      "loss": 0.6549,
+      "step": 1477
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 2.888871908187866,
+      "learning_rate": 0.00016261682242990652,
+      "loss": 0.5391,
+      "step": 1478
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 2.2752833366394043,
+      "learning_rate": 0.00016249999999999997,
+      "loss": 1.3852,
+      "step": 1479
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 2.8205981254577637,
+      "learning_rate": 0.00016238317757009344,
+      "loss": 0.5354,
+      "step": 1480
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.2935905456542969,
+      "learning_rate": 0.0001622663551401869,
+      "loss": 0.6873,
+      "step": 1481
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 3.462782621383667,
+      "learning_rate": 0.00016214953271028037,
+      "loss": 1.3113,
+      "step": 1482
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.3133735656738281,
+      "learning_rate": 0.00016203271028037382,
+      "loss": 0.9383,
+      "step": 1483
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 1.02700936794281,
+      "learning_rate": 0.0001619158878504673,
+      "loss": 0.555,
+      "step": 1484
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 3.047100305557251,
+      "learning_rate": 0.00016179906542056075,
+      "loss": 1.227,
+      "step": 1485
+    },
+    {
+      "epoch": 1.55,
+      "grad_norm": 3.2910258769989014,
+      "learning_rate": 0.00016168224299065417,
+      "loss": 0.7439,
+      "step": 1486
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 2.5884504318237305,
+      "learning_rate": 0.00016156542056074765,
+      "loss": 1.1479,
+      "step": 1487
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 2.7963638305664062,
+      "learning_rate": 0.0001614485981308411,
+      "loss": 1.2802,
+      "step": 1488
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 3.7087883949279785,
+      "learning_rate": 0.00016133177570093457,
+      "loss": 1.2185,
+      "step": 1489
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.849493682384491,
+      "learning_rate": 0.00016121495327102802,
+      "loss": 0.3358,
+      "step": 1490
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.7434322834014893,
+      "learning_rate": 0.00016109813084112147,
+      "loss": 0.2501,
+      "step": 1491
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.6131064295768738,
+      "learning_rate": 0.00016098130841121495,
+      "loss": 0.1483,
+      "step": 1492
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 1.3495182991027832,
+      "learning_rate": 0.0001608644859813084,
+      "loss": 0.4082,
+      "step": 1493
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 2.7759788036346436,
+      "learning_rate": 0.00016074766355140185,
+      "loss": 0.9269,
+      "step": 1494
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 2.0917370319366455,
+      "learning_rate": 0.0001606308411214953,
+      "loss": 0.7764,
+      "step": 1495
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 3.4610509872436523,
+      "learning_rate": 0.00016051401869158875,
+      "loss": 1.1966,
+      "step": 1496
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 3.6502034664154053,
+      "learning_rate": 0.00016039719626168223,
+      "loss": 1.1802,
+      "step": 1497
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 2.2995026111602783,
+      "learning_rate": 0.00016028037383177568,
+      "loss": 1.0207,
+      "step": 1498
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 3.921745777130127,
+      "learning_rate": 0.00016016355140186916,
+      "loss": 0.9952,
+      "step": 1499
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 3.0950217247009277,
+      "learning_rate": 0.0001600467289719626,
+      "loss": 0.4891,
+      "step": 1500
+    },
+    {
+      "epoch": 1.57,
+      "eval_loss": 1.1533477306365967,
+      "eval_runtime": 1.0124,
+      "eval_samples_per_second": 4.939,
+      "eval_steps_per_second": 0.988,
+      "step": 1500
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.7715978026390076,
+      "learning_rate": 0.00015992990654205608,
+      "loss": 0.2482,
+      "step": 1501
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 3.1252341270446777,
+      "learning_rate": 0.0001598130841121495,
+      "loss": 0.6653,
+      "step": 1502
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.6810656785964966,
+      "learning_rate": 0.00015969626168224296,
+      "loss": 0.2498,
+      "step": 1503
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 0.776508092880249,
+      "learning_rate": 0.00015957943925233643,
+      "loss": 0.3936,
+      "step": 1504
+    },
+    {
+      "epoch": 1.57,
+      "grad_norm": 2.2532429695129395,
+      "learning_rate": 0.00015946261682242988,
+      "loss": 0.8081,
+      "step": 1505
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.6081033945083618,
+      "learning_rate": 0.00015934579439252336,
+      "loss": 0.4501,
+      "step": 1506
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 2.283398389816284,
+      "learning_rate": 0.0001592289719626168,
+      "loss": 1.52,
+      "step": 1507
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 2.9111759662628174,
+      "learning_rate": 0.0001591121495327103,
+      "loss": 1.0499,
+      "step": 1508
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 2.6676039695739746,
+      "learning_rate": 0.00015899532710280374,
+      "loss": 0.6595,
+      "step": 1509
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 2.395261526107788,
+      "learning_rate": 0.00015887850467289716,
+      "loss": 0.3202,
+      "step": 1510
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 2.177774667739868,
+      "learning_rate": 0.00015876168224299064,
+      "loss": 0.3961,
+      "step": 1511
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 2.0199573040008545,
+      "learning_rate": 0.0001586448598130841,
+      "loss": 0.5494,
+      "step": 1512
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 1.7580457925796509,
+      "learning_rate": 0.00015852803738317756,
+      "loss": 0.4739,
+      "step": 1513
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 2.2193167209625244,
+      "learning_rate": 0.00015841121495327101,
+      "loss": 1.3114,
+      "step": 1514
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 3.2167983055114746,
+      "learning_rate": 0.00015829439252336446,
+      "loss": 1.0337,
+      "step": 1515
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.9968187808990479,
+      "learning_rate": 0.00015817757009345794,
+      "loss": 0.6111,
+      "step": 1516
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.3653596341609955,
+      "learning_rate": 0.0001580607476635514,
+      "loss": 0.0711,
+      "step": 1517
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.1816760301589966,
+      "learning_rate": 0.00015794392523364484,
+      "loss": 0.338,
+      "step": 1518
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 0.6894358992576599,
+      "learning_rate": 0.0001578271028037383,
+      "loss": 0.2162,
+      "step": 1519
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 1.4301586151123047,
+      "learning_rate": 0.00015771028037383174,
+      "loss": 0.88,
+      "step": 1520
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 2.3182575702667236,
+      "learning_rate": 0.00015759345794392522,
+      "loss": 1.1359,
+      "step": 1521
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 2.250000238418579,
+      "learning_rate": 0.00015747663551401867,
+      "loss": 0.8995,
+      "step": 1522
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 2.535024404525757,
+      "learning_rate": 0.00015735981308411215,
+      "loss": 0.8351,
+      "step": 1523
+    },
+    {
+      "epoch": 1.59,
+      "grad_norm": 3.6850876808166504,
+      "learning_rate": 0.0001572429906542056,
+      "loss": 1.597,
+      "step": 1524
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.4490113258361816,
+      "learning_rate": 0.00015712616822429907,
+      "loss": 0.6743,
+      "step": 1525
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.006627082824707,
+      "learning_rate": 0.0001570093457943925,
+      "loss": 0.5022,
+      "step": 1526
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.6286826133728027,
+      "learning_rate": 0.00015689252336448595,
+      "loss": 0.4875,
+      "step": 1527
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.4024123251438141,
+      "learning_rate": 0.00015677570093457942,
+      "loss": 0.0667,
+      "step": 1528
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.9532690048217773,
+      "learning_rate": 0.00015665887850467287,
+      "loss": 0.5431,
+      "step": 1529
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 3.2107231616973877,
+      "learning_rate": 0.00015654205607476635,
+      "loss": 0.8777,
+      "step": 1530
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 3.744612693786621,
+      "learning_rate": 0.0001564252336448598,
+      "loss": 1.0231,
+      "step": 1531
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.101804256439209,
+      "learning_rate": 0.00015630841121495328,
+      "loss": 0.5195,
+      "step": 1532
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.6645424365997314,
+      "learning_rate": 0.00015619158878504673,
+      "loss": 0.8324,
+      "step": 1533
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.112236738204956,
+      "learning_rate": 0.00015607476635514015,
+      "loss": 1.2638,
+      "step": 1534
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.9546703100204468,
+      "learning_rate": 0.00015595794392523363,
+      "loss": 0.6795,
+      "step": 1535
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.0169304609298706,
+      "learning_rate": 0.00015584112149532708,
+      "loss": 0.4056,
+      "step": 1536
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.9648607969284058,
+      "learning_rate": 0.00015572429906542056,
+      "loss": 0.3932,
+      "step": 1537
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 2.3926336765289307,
+      "learning_rate": 0.000155607476635514,
+      "loss": 0.8817,
+      "step": 1538
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 2.6523513793945312,
+      "learning_rate": 0.00015549065420560746,
+      "loss": 0.861,
+      "step": 1539
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 2.9970486164093018,
+      "learning_rate": 0.00015537383177570093,
+      "loss": 1.0401,
+      "step": 1540
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.2951358556747437,
+      "learning_rate": 0.00015525700934579438,
+      "loss": 0.2964,
+      "step": 1541
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 1.569375991821289,
+      "learning_rate": 0.00015514018691588783,
+      "loss": 0.3614,
+      "step": 1542
+    },
+    {
+      "epoch": 1.61,
+      "grad_norm": 2.29843807220459,
+      "learning_rate": 0.00015502336448598128,
+      "loss": 0.4918,
+      "step": 1543
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 3.2564470767974854,
+      "learning_rate": 0.00015490654205607473,
+      "loss": 0.7068,
+      "step": 1544
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 2.965322971343994,
+      "learning_rate": 0.0001547897196261682,
+      "loss": 0.861,
+      "step": 1545
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 3.3394079208374023,
+      "learning_rate": 0.00015467289719626166,
+      "loss": 1.5551,
+      "step": 1546
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 2.289705514907837,
+      "learning_rate": 0.00015455607476635514,
+      "loss": 0.9215,
+      "step": 1547
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.7418241500854492,
+      "learning_rate": 0.0001544392523364486,
+      "loss": 0.4123,
+      "step": 1548
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 2.0741488933563232,
+      "learning_rate": 0.00015432242990654206,
+      "loss": 1.2427,
+      "step": 1549
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.3914594054222107,
+      "learning_rate": 0.0001542056074766355,
+      "loss": 0.0735,
+      "step": 1550
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 3.96793532371521,
+      "learning_rate": 0.00015408878504672894,
+      "loss": 0.9863,
+      "step": 1551
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 1.1604739427566528,
+      "learning_rate": 0.00015397196261682241,
+      "loss": 0.5499,
+      "step": 1552
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 4.087449550628662,
+      "learning_rate": 0.00015385514018691586,
+      "loss": 1.1625,
+      "step": 1553
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 1.983540415763855,
+      "learning_rate": 0.00015373831775700934,
+      "loss": 1.3978,
+      "step": 1554
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 2.174452781677246,
+      "learning_rate": 0.0001536214953271028,
+      "loss": 0.613,
+      "step": 1555
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 2.5901448726654053,
+      "learning_rate": 0.00015350467289719627,
+      "loss": 0.5222,
+      "step": 1556
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 2.7491252422332764,
+      "learning_rate": 0.00015338785046728972,
+      "loss": 0.5357,
+      "step": 1557
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 2.587277889251709,
+      "learning_rate": 0.00015327102803738314,
+      "loss": 1.4659,
+      "step": 1558
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 3.2006373405456543,
+      "learning_rate": 0.00015315420560747662,
+      "loss": 1.2877,
+      "step": 1559
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 2.9988112449645996,
+      "learning_rate": 0.00015303738317757007,
+      "loss": 0.5014,
+      "step": 1560
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 3.011751890182495,
+      "learning_rate": 0.00015292056074766355,
+      "loss": 0.6075,
+      "step": 1561
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 2.0071706771850586,
+      "learning_rate": 0.000152803738317757,
+      "loss": 1.0028,
+      "step": 1562
+    },
+    {
+      "epoch": 1.63,
+      "grad_norm": 0.4963289499282837,
+      "learning_rate": 0.00015268691588785045,
+      "loss": 0.0778,
+      "step": 1563
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.5799555778503418,
+      "learning_rate": 0.00015257009345794392,
+      "loss": 0.2804,
+      "step": 1564
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 2.747706651687622,
+      "learning_rate": 0.00015245327102803737,
+      "loss": 0.3919,
+      "step": 1565
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 6.703662872314453,
+      "learning_rate": 0.00015233644859813082,
+      "loss": 0.723,
+      "step": 1566
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.7172305583953857,
+      "learning_rate": 0.00015221962616822427,
+      "loss": 0.4695,
+      "step": 1567
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.9481459856033325,
+      "learning_rate": 0.00015210280373831772,
+      "loss": 0.444,
+      "step": 1568
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.0129424333572388,
+      "learning_rate": 0.0001519859813084112,
+      "loss": 0.4649,
+      "step": 1569
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 1.9686312675476074,
+      "learning_rate": 0.00015186915887850465,
+      "loss": 0.2723,
+      "step": 1570
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 2.451975107192993,
+      "learning_rate": 0.00015175233644859813,
+      "loss": 0.7925,
+      "step": 1571
+    },
+    {
+      "epoch": 1.64,
+      "grad_norm": 3.254682779312134,
+      "learning_rate": 0.00015163551401869158,
+      "loss": 0.9014,
+      "step": 1572
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.4323906898498535,
+      "learning_rate": 0.00015151869158878506,
+      "loss": 0.3117,
+      "step": 1573
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.9285516738891602,
+      "learning_rate": 0.00015140186915887848,
+      "loss": 1.5676,
+      "step": 1574
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 3.664214849472046,
+      "learning_rate": 0.00015128504672897193,
+      "loss": 1.0707,
+      "step": 1575
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 2.9163613319396973,
+      "learning_rate": 0.0001511682242990654,
+      "loss": 0.9862,
+      "step": 1576
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 3.5014891624450684,
+      "learning_rate": 0.00015105140186915886,
+      "loss": 0.8882,
+      "step": 1577
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.6664868593215942,
+      "learning_rate": 0.00015093457943925233,
+      "loss": 0.5145,
+      "step": 1578
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 2.4006149768829346,
+      "learning_rate": 0.00015081775700934578,
+      "loss": 0.6458,
+      "step": 1579
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.898476481437683,
+      "learning_rate": 0.00015070093457943926,
+      "loss": 0.5072,
+      "step": 1580
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 1.0624420642852783,
+      "learning_rate": 0.0001505841121495327,
+      "loss": 0.2644,
+      "step": 1581
+    },
+    {
+      "epoch": 1.65,
+      "grad_norm": 2.577528476715088,
+      "learning_rate": 0.00015046728971962613,
+      "loss": 0.9336,
+      "step": 1582
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 2.376910448074341,
+      "learning_rate": 0.0001503504672897196,
+      "loss": 0.6101,
+      "step": 1583
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 3.466841697692871,
+      "learning_rate": 0.00015023364485981306,
+      "loss": 1.1847,
+      "step": 1584
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 2.308872699737549,
+      "learning_rate": 0.00015011682242990654,
+      "loss": 0.546,
+      "step": 1585
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 2.0582919120788574,
+      "learning_rate": 0.00015,
+      "loss": 0.5657,
+      "step": 1586
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 0.6894460916519165,
+      "learning_rate": 0.00014988317757009344,
+      "loss": 0.1334,
+      "step": 1587
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 2.060089349746704,
+      "learning_rate": 0.00014976635514018691,
+      "loss": 1.3325,
+      "step": 1588
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 2.185960054397583,
+      "learning_rate": 0.00014964953271028036,
+      "loss": 0.609,
+      "step": 1589
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 2.608562469482422,
+      "learning_rate": 0.00014953271028037381,
+      "loss": 1.4962,
+      "step": 1590
+    },
+    {
+      "epoch": 1.66,
+      "grad_norm": 2.8520126342773438,
+      "learning_rate": 0.0001494158878504673,
+      "loss": 1.1023,
+      "step": 1591
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.7289494276046753,
+      "learning_rate": 0.00014929906542056074,
+      "loss": 0.4568,
+      "step": 1592
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 5.0688090324401855,
+      "learning_rate": 0.0001491822429906542,
+      "loss": 1.581,
+      "step": 1593
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.531291127204895,
+      "learning_rate": 0.00014906542056074764,
+      "loss": 1.2141,
+      "step": 1594
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.8704463243484497,
+      "learning_rate": 0.00014894859813084112,
+      "loss": 0.9042,
+      "step": 1595
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 3.1523284912109375,
+      "learning_rate": 0.00014883177570093457,
+      "loss": 1.2137,
+      "step": 1596
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 3.0960710048675537,
+      "learning_rate": 0.00014871495327102802,
+      "loss": 0.6209,
+      "step": 1597
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 0.5884736776351929,
+      "learning_rate": 0.00014859813084112147,
+      "loss": 0.1863,
+      "step": 1598
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.639065146446228,
+      "learning_rate": 0.00014848130841121495,
+      "loss": 0.4886,
+      "step": 1599
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 2.418137311935425,
+      "learning_rate": 0.0001483644859813084,
+      "loss": 1.194,
+      "step": 1600
+    },
+    {
+      "epoch": 1.67,
+      "grad_norm": 1.7051281929016113,
+      "learning_rate": 0.00014824766355140185,
+      "loss": 0.7105,
+      "step": 1601
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 2.0735154151916504,
+      "learning_rate": 0.00014813084112149532,
+      "loss": 0.5942,
+      "step": 1602
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 2.791166305541992,
+      "learning_rate": 0.00014801401869158877,
+      "loss": 1.4937,
+      "step": 1603
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 1.4166151285171509,
+      "learning_rate": 0.00014789719626168225,
+      "loss": 0.5445,
+      "step": 1604
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 2.774190902709961,
+      "learning_rate": 0.00014778037383177567,
+      "loss": 0.5487,
+      "step": 1605
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 2.2081236839294434,
+      "learning_rate": 0.00014766355140186915,
+      "loss": 0.7735,
+      "step": 1606
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.9153463244438171,
+      "learning_rate": 0.0001475467289719626,
+      "loss": 0.3899,
+      "step": 1607
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 0.8370518684387207,
+      "learning_rate": 0.00014742990654205608,
+      "loss": 0.3468,
+      "step": 1608
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 3.461733102798462,
+      "learning_rate": 0.00014731308411214953,
+      "loss": 1.1843,
+      "step": 1609
+    },
+    {
+      "epoch": 1.68,
+      "grad_norm": 2.4178223609924316,
+      "learning_rate": 0.00014719626168224298,
+      "loss": 1.168,
+      "step": 1610
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 2.963109016418457,
+      "learning_rate": 0.00014707943925233643,
+      "loss": 1.0568,
+      "step": 1611
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.694535493850708,
+      "learning_rate": 0.0001469626168224299,
+      "loss": 0.2683,
+      "step": 1612
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.9057122468948364,
+      "learning_rate": 0.00014684579439252335,
+      "loss": 0.3639,
+      "step": 1613
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.505016028881073,
+      "learning_rate": 0.0001467289719626168,
+      "loss": 0.1307,
+      "step": 1614
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.8371268510818481,
+      "learning_rate": 0.00014661214953271028,
+      "loss": 0.4274,
+      "step": 1615
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 2.2967214584350586,
+      "learning_rate": 0.00014649532710280373,
+      "loss": 1.2995,
+      "step": 1616
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 1.8484244346618652,
+      "learning_rate": 0.00014637850467289718,
+      "loss": 0.4558,
+      "step": 1617
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 2.2368478775024414,
+      "learning_rate": 0.00014626168224299063,
+      "loss": 0.7785,
+      "step": 1618
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 2.3678500652313232,
+      "learning_rate": 0.0001461448598130841,
+      "loss": 1.4081,
+      "step": 1619
+    },
+    {
+      "epoch": 1.69,
+      "grad_norm": 0.6158261895179749,
+      "learning_rate": 0.00014602803738317756,
+      "loss": 0.1808,
+      "step": 1620
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.6939470767974854,
+      "learning_rate": 0.000145911214953271,
+      "loss": 0.509,
+      "step": 1621
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 2.4551703929901123,
+      "learning_rate": 0.00014579439252336446,
+      "loss": 0.473,
+      "step": 1622
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 2.0342600345611572,
+      "learning_rate": 0.00014567757009345794,
+      "loss": 0.4449,
+      "step": 1623
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.9285507202148438,
+      "learning_rate": 0.0001455607476635514,
+      "loss": 1.3971,
+      "step": 1624
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.5966708660125732,
+      "learning_rate": 0.00014544392523364484,
+      "loss": 0.2074,
+      "step": 1625
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 2.5656778812408447,
+      "learning_rate": 0.00014532710280373831,
+      "loss": 0.6084,
+      "step": 1626
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 2.896272897720337,
+      "learning_rate": 0.00014521028037383176,
+      "loss": 0.9998,
+      "step": 1627
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 2.119194746017456,
+      "learning_rate": 0.00014509345794392524,
+      "loss": 0.3859,
+      "step": 1628
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 2.5145485401153564,
+      "learning_rate": 0.00014497663551401866,
+      "loss": 0.6619,
+      "step": 1629
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 1.912062406539917,
+      "learning_rate": 0.00014485981308411214,
+      "loss": 0.8044,
+      "step": 1630
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 2.8148365020751953,
+      "learning_rate": 0.0001447429906542056,
+      "loss": 0.5945,
+      "step": 1631
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 2.518463373184204,
+      "learning_rate": 0.00014462616822429907,
+      "loss": 0.6838,
+      "step": 1632
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 3.0430805683135986,
+      "learning_rate": 0.00014450934579439252,
+      "loss": 0.8019,
+      "step": 1633
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 3.276461124420166,
+      "learning_rate": 0.00014439252336448597,
+      "loss": 0.7042,
+      "step": 1634
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 2.3962302207946777,
+      "learning_rate": 0.00014427570093457942,
+      "loss": 1.4162,
+      "step": 1635
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 0.7789445519447327,
+      "learning_rate": 0.0001441588785046729,
+      "loss": 0.291,
+      "step": 1636
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 3.588848352432251,
+      "learning_rate": 0.00014404205607476635,
+      "loss": 1.125,
+      "step": 1637
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 2.0526657104492188,
+      "learning_rate": 0.0001439252336448598,
+      "loss": 1.168,
+      "step": 1638
+    },
+    {
+      "epoch": 1.71,
+      "grad_norm": 3.957766056060791,
+      "learning_rate": 0.00014380841121495327,
+      "loss": 1.0162,
+      "step": 1639
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.8492609262466431,
+      "learning_rate": 0.00014369158878504672,
+      "loss": 0.4544,
+      "step": 1640
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 2.658005714416504,
+      "learning_rate": 0.00014357476635514017,
+      "loss": 0.6238,
+      "step": 1641
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 2.339304208755493,
+      "learning_rate": 0.00014345794392523362,
+      "loss": 1.2503,
+      "step": 1642
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 2.0792555809020996,
+      "learning_rate": 0.0001433411214953271,
+      "loss": 0.4744,
+      "step": 1643
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 2.6944262981414795,
+      "learning_rate": 0.00014322429906542055,
+      "loss": 1.2977,
+      "step": 1644
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 3.134782314300537,
+      "learning_rate": 0.000143107476635514,
+      "loss": 0.9947,
+      "step": 1645
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 4.126375198364258,
+      "learning_rate": 0.00014299065420560745,
+      "loss": 1.2147,
+      "step": 1646
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 2.5845067501068115,
+      "learning_rate": 0.00014287383177570093,
+      "loss": 1.1301,
+      "step": 1647
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 3.6140151023864746,
+      "learning_rate": 0.00014275700934579438,
+      "loss": 1.4641,
+      "step": 1648
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.5477080941200256,
+      "learning_rate": 0.00014264018691588783,
+      "loss": 0.1762,
+      "step": 1649
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 2.663865089416504,
+      "learning_rate": 0.0001425233644859813,
+      "loss": 0.7445,
+      "step": 1650
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 0.3674846887588501,
+      "learning_rate": 0.00014240654205607475,
+      "loss": 0.087,
+      "step": 1651
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.6670650243759155,
+      "learning_rate": 0.00014228971962616823,
+      "loss": 1.2636,
+      "step": 1652
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 2.501187801361084,
+      "learning_rate": 0.00014217289719626165,
+      "loss": 0.8981,
+      "step": 1653
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 2.9647672176361084,
+      "learning_rate": 0.00014205607476635513,
+      "loss": 1.5088,
+      "step": 1654
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.6598446369171143,
+      "learning_rate": 0.00014193925233644858,
+      "loss": 0.3709,
+      "step": 1655
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 2.7449827194213867,
+      "learning_rate": 0.00014182242990654206,
+      "loss": 1.3826,
+      "step": 1656
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 2.3242647647857666,
+      "learning_rate": 0.0001417056074766355,
+      "loss": 0.4957,
+      "step": 1657
+    },
+    {
+      "epoch": 1.73,
+      "grad_norm": 1.6587960720062256,
+      "learning_rate": 0.00014158878504672896,
+      "loss": 0.7291,
+      "step": 1658
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.4735907316207886,
+      "learning_rate": 0.0001414719626168224,
+      "loss": 0.48,
+      "step": 1659
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.680935263633728,
+      "learning_rate": 0.00014135514018691589,
+      "loss": 0.656,
+      "step": 1660
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.679986834526062,
+      "learning_rate": 0.00014123831775700934,
+      "loss": 0.5056,
+      "step": 1661
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.520255208015442,
+      "learning_rate": 0.00014112149532710279,
+      "loss": 0.2241,
+      "step": 1662
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.179740309715271,
+      "learning_rate": 0.00014100467289719626,
+      "loss": 0.2556,
+      "step": 1663
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 3.3339192867279053,
+      "learning_rate": 0.0001408878504672897,
+      "loss": 1.1833,
+      "step": 1664
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.6800388097763062,
+      "learning_rate": 0.00014077102803738316,
+      "loss": 0.4323,
+      "step": 1665
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 2.357207775115967,
+      "learning_rate": 0.0001406542056074766,
+      "loss": 0.7178,
+      "step": 1666
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.8271247148513794,
+      "learning_rate": 0.0001405373831775701,
+      "loss": 0.6007,
+      "step": 1667
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.7845667004585266,
+      "learning_rate": 0.00014042056074766354,
+      "loss": 0.3273,
+      "step": 1668
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 2.022867441177368,
+      "learning_rate": 0.000140303738317757,
+      "loss": 0.8424,
+      "step": 1669
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 2.6293094158172607,
+      "learning_rate": 0.00014018691588785044,
+      "loss": 0.7543,
+      "step": 1670
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 2.154860258102417,
+      "learning_rate": 0.00014007009345794392,
+      "loss": 0.8477,
+      "step": 1671
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.9960460662841797,
+      "learning_rate": 0.00013995327102803737,
+      "loss": 0.4638,
+      "step": 1672
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.6653493046760559,
+      "learning_rate": 0.00013983644859813082,
+      "loss": 0.1838,
+      "step": 1673
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 2.4807372093200684,
+      "learning_rate": 0.0001397196261682243,
+      "loss": 0.6451,
+      "step": 1674
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 1.979583978652954,
+      "learning_rate": 0.00013960280373831774,
+      "loss": 0.4743,
+      "step": 1675
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 4.964494705200195,
+      "learning_rate": 0.00013948598130841122,
+      "loss": 1.1769,
+      "step": 1676
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 2.702359199523926,
+      "learning_rate": 0.00013936915887850465,
+      "loss": 1.7705,
+      "step": 1677
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 2.9674830436706543,
+      "learning_rate": 0.00013925233644859812,
+      "loss": 0.8044,
+      "step": 1678
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6541764736175537,
+      "learning_rate": 0.00013913551401869157,
+      "loss": 0.277,
+      "step": 1679
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.7068645358085632,
+      "learning_rate": 0.00013901869158878505,
+      "loss": 0.2462,
+      "step": 1680
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.6276235580444336,
+      "learning_rate": 0.0001389018691588785,
+      "loss": 0.2266,
+      "step": 1681
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.7366113662719727,
+      "learning_rate": 0.00013878504672897195,
+      "loss": 0.6867,
+      "step": 1682
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.0275715589523315,
+      "learning_rate": 0.0001386682242990654,
+      "loss": 0.3801,
+      "step": 1683
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 2.116137981414795,
+      "learning_rate": 0.00013855140186915888,
+      "loss": 1.5116,
+      "step": 1684
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 3.340062379837036,
+      "learning_rate": 0.00013843457943925233,
+      "loss": 0.6008,
+      "step": 1685
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 1.818067193031311,
+      "learning_rate": 0.00013831775700934578,
+      "loss": 0.8285,
+      "step": 1686
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 2.169006109237671,
+      "learning_rate": 0.00013820093457943925,
+      "loss": 1.56,
+      "step": 1687
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 2.3605668544769287,
+      "learning_rate": 0.0001380841121495327,
+      "loss": 0.476,
+      "step": 1688
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.9693185091018677,
+      "learning_rate": 0.00013796728971962615,
+      "loss": 0.5195,
+      "step": 1689
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.9456527829170227,
+      "learning_rate": 0.0001378504672897196,
+      "loss": 0.4079,
+      "step": 1690
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.855286955833435,
+      "learning_rate": 0.00013773364485981308,
+      "loss": 0.5983,
+      "step": 1691
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 2.4261703491210938,
+      "learning_rate": 0.00013761682242990653,
+      "loss": 0.3252,
+      "step": 1692
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 2.379132032394409,
+      "learning_rate": 0.00013749999999999998,
+      "loss": 0.861,
+      "step": 1693
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 0.2820170521736145,
+      "learning_rate": 0.00013738317757009343,
+      "loss": 0.0576,
+      "step": 1694
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 2.008049964904785,
+      "learning_rate": 0.0001372663551401869,
+      "loss": 0.4623,
+      "step": 1695
+    },
+    {
+      "epoch": 1.77,
+      "grad_norm": 1.8559750318527222,
+      "learning_rate": 0.00013714953271028036,
+      "loss": 0.518,
+      "step": 1696
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 2.115945816040039,
+      "learning_rate": 0.0001370327102803738,
+      "loss": 0.5179,
+      "step": 1697
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.434239149093628,
+      "learning_rate": 0.00013691588785046729,
+      "loss": 0.5205,
+      "step": 1698
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 1.3366906642913818,
+      "learning_rate": 0.00013679906542056074,
+      "loss": 0.7658,
+      "step": 1699
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 3.734058141708374,
+      "learning_rate": 0.0001366822429906542,
+      "loss": 1.525,
+      "step": 1700
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 3.2385447025299072,
+      "learning_rate": 0.00013656542056074764,
+      "loss": 0.8095,
+      "step": 1701
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 3.614528179168701,
+      "learning_rate": 0.0001364485981308411,
+      "loss": 1.0613,
+      "step": 1702
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.726099967956543,
+      "learning_rate": 0.00013633177570093456,
+      "loss": 0.1841,
+      "step": 1703
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 3.171700954437256,
+      "learning_rate": 0.00013621495327102804,
+      "loss": 1.0322,
+      "step": 1704
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 2.1440253257751465,
+      "learning_rate": 0.0001360981308411215,
+      "loss": 1.0489,
+      "step": 1705
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.680216908454895,
+      "learning_rate": 0.00013598130841121494,
+      "loss": 0.2887,
+      "step": 1706
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 3.5613088607788086,
+      "learning_rate": 0.0001358644859813084,
+      "loss": 0.8109,
+      "step": 1707
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 3.2237210273742676,
+      "learning_rate": 0.00013574766355140187,
+      "loss": 1.3278,
+      "step": 1708
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 3.895890951156616,
+      "learning_rate": 0.00013563084112149532,
+      "loss": 1.1268,
+      "step": 1709
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.0976706743240356,
+      "learning_rate": 0.00013551401869158877,
+      "loss": 0.4548,
+      "step": 1710
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 2.142307758331299,
+      "learning_rate": 0.00013539719626168224,
+      "loss": 0.507,
+      "step": 1711
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 2.718857765197754,
+      "learning_rate": 0.0001352803738317757,
+      "loss": 0.4967,
+      "step": 1712
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 0.7044395208358765,
+      "learning_rate": 0.00013516355140186914,
+      "loss": 0.215,
+      "step": 1713
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 2.744619607925415,
+      "learning_rate": 0.0001350467289719626,
+      "loss": 0.7843,
+      "step": 1714
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 1.998574137687683,
+      "learning_rate": 0.00013492990654205607,
+      "loss": 1.4232,
+      "step": 1715
+    },
+    {
+      "epoch": 1.79,
+      "grad_norm": 3.0873868465423584,
+      "learning_rate": 0.00013481308411214952,
+      "loss": 0.7932,
+      "step": 1716
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 2.173910617828369,
+      "learning_rate": 0.00013469626168224297,
+      "loss": 0.7113,
+      "step": 1717
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 2.539853096008301,
+      "learning_rate": 0.00013457943925233642,
+      "loss": 0.3913,
+      "step": 1718
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.858372449874878,
+      "learning_rate": 0.0001344626168224299,
+      "loss": 0.5278,
+      "step": 1719
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.7758537530899048,
+      "learning_rate": 0.00013434579439252335,
+      "loss": 0.384,
+      "step": 1720
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.768137514591217,
+      "learning_rate": 0.0001342289719626168,
+      "loss": 0.3117,
+      "step": 1721
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.7663379311561584,
+      "learning_rate": 0.00013411214953271028,
+      "loss": 0.3133,
+      "step": 1722
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 2.0158727169036865,
+      "learning_rate": 0.00013399532710280373,
+      "loss": 0.6746,
+      "step": 1723
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 2.7553088665008545,
+      "learning_rate": 0.0001338785046728972,
+      "loss": 0.8026,
+      "step": 1724
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.7455641031265259,
+      "learning_rate": 0.00013376168224299063,
+      "loss": 0.4179,
+      "step": 1725
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.652976155281067,
+      "learning_rate": 0.0001336448598130841,
+      "loss": 0.7564,
+      "step": 1726
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.452163815498352,
+      "learning_rate": 0.00013352803738317755,
+      "loss": 0.3617,
+      "step": 1727
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 2.5906240940093994,
+      "learning_rate": 0.00013341121495327103,
+      "loss": 0.7063,
+      "step": 1728
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.8791476488113403,
+      "learning_rate": 0.00013329439252336448,
+      "loss": 0.4949,
+      "step": 1729
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 3.926201820373535,
+      "learning_rate": 0.00013317757009345793,
+      "loss": 1.5384,
+      "step": 1730
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 3.678696870803833,
+      "learning_rate": 0.00013306074766355138,
+      "loss": 1.3355,
+      "step": 1731
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 4.7318501472473145,
+      "learning_rate": 0.00013294392523364486,
+      "loss": 0.9178,
+      "step": 1732
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 3.0158326625823975,
+      "learning_rate": 0.0001328271028037383,
+      "loss": 1.0379,
+      "step": 1733
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.457908034324646,
+      "learning_rate": 0.00013271028037383176,
+      "loss": 0.3679,
+      "step": 1734
+    },
+    {
+      "epoch": 1.81,
+      "grad_norm": 1.1844749450683594,
+      "learning_rate": 0.00013259345794392524,
+      "loss": 0.4305,
+      "step": 1735
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 2.570006847381592,
+      "learning_rate": 0.00013247663551401869,
+      "loss": 0.4679,
+      "step": 1736
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 3.1339075565338135,
+      "learning_rate": 0.00013235981308411214,
+      "loss": 0.7069,
+      "step": 1737
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.4061032235622406,
+      "learning_rate": 0.00013224299065420559,
+      "loss": 0.1061,
+      "step": 1738
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 2.7448058128356934,
+      "learning_rate": 0.00013212616822429906,
+      "loss": 1.1243,
+      "step": 1739
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 2.1957473754882812,
+      "learning_rate": 0.0001320093457943925,
+      "loss": 0.3921,
+      "step": 1740
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 0.7898138761520386,
+      "learning_rate": 0.00013189252336448596,
+      "loss": 0.3201,
+      "step": 1741
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 3.37489652633667,
+      "learning_rate": 0.0001317757009345794,
+      "loss": 1.1398,
+      "step": 1742
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 4.243715286254883,
+      "learning_rate": 0.0001316588785046729,
+      "loss": 0.603,
+      "step": 1743
+    },
+    {
+      "epoch": 1.82,
+      "grad_norm": 2.359915256500244,
+      "learning_rate": 0.00013154205607476634,
+      "loss": 1.0448,
+      "step": 1744
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.555791974067688,
+      "learning_rate": 0.0001314252336448598,
+      "loss": 0.3294,
+      "step": 1745
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 1.50589120388031,
+      "learning_rate": 0.00013130841121495327,
+      "loss": 0.3309,
+      "step": 1746
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 2.945429801940918,
+      "learning_rate": 0.00013119158878504672,
+      "loss": 0.8982,
+      "step": 1747
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 3.0196831226348877,
+      "learning_rate": 0.0001310747663551402,
+      "loss": 0.9588,
+      "step": 1748
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 2.560878038406372,
+      "learning_rate": 0.00013095794392523362,
+      "loss": 0.6943,
+      "step": 1749
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 2.355393409729004,
+      "learning_rate": 0.0001308411214953271,
+      "loss": 0.6189,
+      "step": 1750
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 4.520174503326416,
+      "learning_rate": 0.00013072429906542054,
+      "loss": 0.9338,
+      "step": 1751
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 2.112287998199463,
+      "learning_rate": 0.00013060747663551402,
+      "loss": 0.4086,
+      "step": 1752
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 3.6935834884643555,
+      "learning_rate": 0.00013049065420560747,
+      "loss": 0.4796,
+      "step": 1753
+    },
+    {
+      "epoch": 1.83,
+      "grad_norm": 0.544985294342041,
+      "learning_rate": 0.00013037383177570092,
+      "loss": 0.1304,
+      "step": 1754
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 3.0136711597442627,
+      "learning_rate": 0.00013025700934579437,
+      "loss": 0.4961,
+      "step": 1755
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 4.112677097320557,
+      "learning_rate": 0.00013014018691588785,
+      "loss": 0.9081,
+      "step": 1756
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 2.986521005630493,
+      "learning_rate": 0.0001300233644859813,
+      "loss": 0.963,
+      "step": 1757
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 0.8491998910903931,
+      "learning_rate": 0.00012990654205607475,
+      "loss": 0.3462,
+      "step": 1758
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 2.529773712158203,
+      "learning_rate": 0.00012978971962616823,
+      "loss": 0.8424,
+      "step": 1759
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 1.9695286750793457,
+      "learning_rate": 0.00012967289719626168,
+      "loss": 1.0353,
+      "step": 1760
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 3.785512685775757,
+      "learning_rate": 0.00012955607476635513,
+      "loss": 1.0751,
+      "step": 1761
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 3.8414695262908936,
+      "learning_rate": 0.00012943925233644858,
+      "loss": 1.1467,
+      "step": 1762
+    },
+    {
+      "epoch": 1.84,
+      "grad_norm": 2.077259063720703,
+      "learning_rate": 0.00012932242990654205,
+      "loss": 0.7422,
+      "step": 1763
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 2.666192054748535,
+      "learning_rate": 0.0001292056074766355,
+      "loss": 0.7893,
+      "step": 1764
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 2.1345760822296143,
+      "learning_rate": 0.00012908878504672895,
+      "loss": 0.527,
+      "step": 1765
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 3.850268840789795,
+      "learning_rate": 0.0001289719626168224,
+      "loss": 0.9478,
+      "step": 1766
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 2.2270233631134033,
+      "learning_rate": 0.00012885514018691588,
+      "loss": 1.0623,
+      "step": 1767
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.8560389280319214,
+      "learning_rate": 0.00012873831775700933,
+      "loss": 0.2867,
+      "step": 1768
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 3.4770913124084473,
+      "learning_rate": 0.00012862149532710278,
+      "loss": 1.1433,
+      "step": 1769
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 4.113784313201904,
+      "learning_rate": 0.00012850467289719626,
+      "loss": 1.1744,
+      "step": 1770
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 2.4345414638519287,
+      "learning_rate": 0.0001283878504672897,
+      "loss": 1.0072,
+      "step": 1771
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 0.8015138506889343,
+      "learning_rate": 0.00012827102803738318,
+      "loss": 0.206,
+      "step": 1772
+    },
+    {
+      "epoch": 1.85,
+      "grad_norm": 1.1205389499664307,
+      "learning_rate": 0.0001281542056074766,
+      "loss": 0.6316,
+      "step": 1773
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 2.1893746852874756,
+      "learning_rate": 0.00012803738317757008,
+      "loss": 0.9665,
+      "step": 1774
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 3.2439510822296143,
+      "learning_rate": 0.00012792056074766353,
+      "loss": 1.1231,
+      "step": 1775
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 2.1773808002471924,
+      "learning_rate": 0.000127803738317757,
+      "loss": 1.114,
+      "step": 1776
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 2.366384983062744,
+      "learning_rate": 0.00012768691588785046,
+      "loss": 0.8207,
+      "step": 1777
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.9399102926254272,
+      "learning_rate": 0.0001275700934579439,
+      "loss": 0.5248,
+      "step": 1778
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.47793617844581604,
+      "learning_rate": 0.00012745327102803736,
+      "loss": 0.1291,
+      "step": 1779
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 1.048606514930725,
+      "learning_rate": 0.00012733644859813084,
+      "loss": 0.5026,
+      "step": 1780
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 2.299088478088379,
+      "learning_rate": 0.0001272196261682243,
+      "loss": 0.7212,
+      "step": 1781
+    },
+    {
+      "epoch": 1.86,
+      "grad_norm": 0.593243420124054,
+      "learning_rate": 0.00012710280373831774,
+      "loss": 0.164,
+      "step": 1782
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.7880874276161194,
+      "learning_rate": 0.00012698598130841122,
+      "loss": 0.4402,
+      "step": 1783
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 3.372981309890747,
+      "learning_rate": 0.00012686915887850467,
+      "loss": 0.9855,
+      "step": 1784
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.975346326828003,
+      "learning_rate": 0.00012675233644859812,
+      "loss": 0.9623,
+      "step": 1785
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 2.360755205154419,
+      "learning_rate": 0.00012663551401869157,
+      "loss": 0.6337,
+      "step": 1786
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 4.685933589935303,
+      "learning_rate": 0.00012651869158878504,
+      "loss": 1.1126,
+      "step": 1787
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 2.2752182483673096,
+      "learning_rate": 0.0001264018691588785,
+      "loss": 1.0364,
+      "step": 1788
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 3.0285208225250244,
+      "learning_rate": 0.00012628504672897194,
+      "loss": 0.6778,
+      "step": 1789
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 0.5671124458312988,
+      "learning_rate": 0.0001261682242990654,
+      "loss": 0.2275,
+      "step": 1790
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 1.9352588653564453,
+      "learning_rate": 0.00012605140186915887,
+      "loss": 0.4641,
+      "step": 1791
+    },
+    {
+      "epoch": 1.87,
+      "grad_norm": 4.273382663726807,
+      "learning_rate": 0.00012593457943925232,
+      "loss": 0.907,
+      "step": 1792
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.3872417211532593,
+      "learning_rate": 0.00012581775700934577,
+      "loss": 0.3541,
+      "step": 1793
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 2.6732919216156006,
+      "learning_rate": 0.00012570093457943925,
+      "loss": 1.3397,
+      "step": 1794
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.4396637678146362,
+      "learning_rate": 0.0001255841121495327,
+      "loss": 0.8367,
+      "step": 1795
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 2.2039003372192383,
+      "learning_rate": 0.00012546728971962618,
+      "loss": 0.8741,
+      "step": 1796
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 2.8381385803222656,
+      "learning_rate": 0.0001253504672897196,
+      "loss": 0.4666,
+      "step": 1797
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 1.1700317859649658,
+      "learning_rate": 0.00012523364485981308,
+      "loss": 0.5679,
+      "step": 1798
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 2.4915921688079834,
+      "learning_rate": 0.00012511682242990653,
+      "loss": 1.4566,
+      "step": 1799
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 3.598515033721924,
+      "learning_rate": 0.000125,
+      "loss": 1.4513,
+      "step": 1800
+    },
+    {
+      "epoch": 1.88,
+      "eval_loss": 1.143676996231079,
+      "eval_runtime": 1.0027,
+      "eval_samples_per_second": 4.987,
+      "eval_steps_per_second": 0.997,
+      "step": 1800
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2868,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 300,
+  "total_flos": 6.46034707094569e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}