diff --git "a/epoch_9/checkpoint-1100/trainer_state.json" "b/epoch_9/checkpoint-1100/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/epoch_9/checkpoint-1100/trainer_state.json"
@@ -0,0 +1,7734 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.25,
+  "eval_steps": 500,
+  "global_step": 1100,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00022727272727272727,
+      "grad_norm": 2.8016968868818073,
+      "learning_rate": 1.25e-06,
+      "loss": 0.1255,
+      "step": 1
+    },
+    {
+      "epoch": 0.00045454545454545455,
+      "grad_norm": 4.112971111605968,
+      "learning_rate": 1.249999840689502e-06,
+      "loss": 0.1102,
+      "step": 2
+    },
+    {
+      "epoch": 0.0006818181818181819,
+      "grad_norm": 2.284279232933533,
+      "learning_rate": 1.2499993627580887e-06,
+      "loss": 0.0922,
+      "step": 3
+    },
+    {
+      "epoch": 0.0009090909090909091,
+      "grad_norm": 3.302173277719191,
+      "learning_rate": 1.2499985662060041e-06,
+      "loss": 0.1276,
+      "step": 4
+    },
+    {
+      "epoch": 0.0011363636363636363,
+      "grad_norm": 2.0594478547564146,
+      "learning_rate": 1.249997451033654e-06,
+      "loss": 0.1148,
+      "step": 5
+    },
+    {
+      "epoch": 0.0013636363636363637,
+      "grad_norm": 3.286789038531438,
+      "learning_rate": 1.249996017241607e-06,
+      "loss": 0.1415,
+      "step": 6
+    },
+    {
+      "epoch": 0.001590909090909091,
+      "grad_norm": 4.263909691089189,
+      "learning_rate": 1.249994264830594e-06,
+      "loss": 0.1536,
+      "step": 7
+    },
+    {
+      "epoch": 0.0018181818181818182,
+      "grad_norm": 3.1395386531922522,
+      "learning_rate": 1.2499921938015086e-06,
+      "loss": 0.1243,
+      "step": 8
+    },
+    {
+      "epoch": 0.0020454545454545456,
+      "grad_norm": 1.50090695643886,
+      "learning_rate": 1.2499898041554066e-06,
+      "loss": 0.0773,
+      "step": 9
+    },
+    {
+      "epoch": 0.0022727272727272726,
+      "grad_norm": 2.187826181591493,
+      "learning_rate": 1.2499870958935056e-06,
+      "loss": 0.1215,
+      "step": 10
+    },
+    {
+      "epoch": 0.0025,
+      "grad_norm": 2.5804134571029755,
+      "learning_rate": 1.2499840690171872e-06,
+      "loss": 0.1421,
+      "step": 11
+    },
+    {
+      "epoch": 0.0027272727272727275,
+      "grad_norm": 2.3184080898064656,
+      "learning_rate": 1.2499807235279937e-06,
+      "loss": 0.0589,
+      "step": 12
+    },
+    {
+      "epoch": 0.0029545454545454545,
+      "grad_norm": 1.537495944168739,
+      "learning_rate": 1.249977059427631e-06,
+      "loss": 0.1289,
+      "step": 13
+    },
+    {
+      "epoch": 0.003181818181818182,
+      "grad_norm": 2.4564551204506397,
+      "learning_rate": 1.2499730767179668e-06,
+      "loss": 0.1534,
+      "step": 14
+    },
+    {
+      "epoch": 0.003409090909090909,
+      "grad_norm": 1.6481162661194857,
+      "learning_rate": 1.2499687754010318e-06,
+      "loss": 0.0966,
+      "step": 15
+    },
+    {
+      "epoch": 0.0036363636363636364,
+      "grad_norm": 2.656047394405215,
+      "learning_rate": 1.2499641554790185e-06,
+      "loss": 0.1401,
+      "step": 16
+    },
+    {
+      "epoch": 0.003863636363636364,
+      "grad_norm": 2.9301601048990182,
+      "learning_rate": 1.2499592169542823e-06,
+      "loss": 0.1214,
+      "step": 17
+    },
+    {
+      "epoch": 0.004090909090909091,
+      "grad_norm": 2.109199870348404,
+      "learning_rate": 1.2499539598293406e-06,
+      "loss": 0.069,
+      "step": 18
+    },
+    {
+      "epoch": 0.004318181818181818,
+      "grad_norm": 3.6280019518536424,
+      "learning_rate": 1.2499483841068736e-06,
+      "loss": 0.1424,
+      "step": 19
+    },
+    {
+      "epoch": 0.004545454545454545,
+      "grad_norm": 2.558709531356719,
+      "learning_rate": 1.2499424897897237e-06,
+      "loss": 0.0929,
+      "step": 20
+    },
+    {
+      "epoch": 0.004772727272727273,
+      "grad_norm": 3.9209490023886957,
+      "learning_rate": 1.2499362768808958e-06,
+      "loss": 0.103,
+      "step": 21
+    },
+    {
+      "epoch": 0.005,
+      "grad_norm": 2.0035355427055377,
+      "learning_rate": 1.2499297453835574e-06,
+      "loss": 0.1737,
+      "step": 22
+    },
+    {
+      "epoch": 0.005227272727272727,
+      "grad_norm": 1.366643171398265,
+      "learning_rate": 1.2499228953010379e-06,
+      "loss": 0.0567,
+      "step": 23
+    },
+    {
+      "epoch": 0.005454545454545455,
+      "grad_norm": 2.4159855117324445,
+      "learning_rate": 1.2499157266368298e-06,
+      "loss": 0.0655,
+      "step": 24
+    },
+    {
+      "epoch": 0.005681818181818182,
+      "grad_norm": 1.2622281145845518,
+      "learning_rate": 1.2499082393945871e-06,
+      "loss": 0.0899,
+      "step": 25
+    },
+    {
+      "epoch": 0.005909090909090909,
+      "grad_norm": 5.088549448239914,
+      "learning_rate": 1.2499004335781272e-06,
+      "loss": 0.2048,
+      "step": 26
+    },
+    {
+      "epoch": 0.006136363636363636,
+      "grad_norm": 2.7766264540697447,
+      "learning_rate": 1.2498923091914293e-06,
+      "loss": 0.1138,
+      "step": 27
+    },
+    {
+      "epoch": 0.006363636363636364,
+      "grad_norm": 2.828792441583351,
+      "learning_rate": 1.249883866238635e-06,
+      "loss": 0.1093,
+      "step": 28
+    },
+    {
+      "epoch": 0.006590909090909091,
+      "grad_norm": 2.08913381466705,
+      "learning_rate": 1.2498751047240488e-06,
+      "loss": 0.1088,
+      "step": 29
+    },
+    {
+      "epoch": 0.006818181818181818,
+      "grad_norm": 0.9662882916535072,
+      "learning_rate": 1.2498660246521371e-06,
+      "loss": 0.0487,
+      "step": 30
+    },
+    {
+      "epoch": 0.007045454545454546,
+      "grad_norm": 2.851062180418608,
+      "learning_rate": 1.2498566260275289e-06,
+      "loss": 0.144,
+      "step": 31
+    },
+    {
+      "epoch": 0.007272727272727273,
+      "grad_norm": 1.315393713532606,
+      "learning_rate": 1.2498469088550156e-06,
+      "loss": 0.097,
+      "step": 32
+    },
+    {
+      "epoch": 0.0075,
+      "grad_norm": 3.8481047595563336,
+      "learning_rate": 1.2498368731395507e-06,
+      "loss": 0.1813,
+      "step": 33
+    },
+    {
+      "epoch": 0.007727272727272728,
+      "grad_norm": 2.4923456141583458,
+      "learning_rate": 1.2498265188862505e-06,
+      "loss": 0.1761,
+      "step": 34
+    },
+    {
+      "epoch": 0.007954545454545454,
+      "grad_norm": 2.787587786820959,
+      "learning_rate": 1.2498158461003935e-06,
+      "loss": 0.1797,
+      "step": 35
+    },
+    {
+      "epoch": 0.008181818181818182,
+      "grad_norm": 2.2179956854849574,
+      "learning_rate": 1.2498048547874208e-06,
+      "loss": 0.0693,
+      "step": 36
+    },
+    {
+      "epoch": 0.00840909090909091,
+      "grad_norm": 2.600913547903702,
+      "learning_rate": 1.2497935449529355e-06,
+      "loss": 0.1528,
+      "step": 37
+    },
+    {
+      "epoch": 0.008636363636363636,
+      "grad_norm": 2.380118719029788,
+      "learning_rate": 1.2497819166027035e-06,
+      "loss": 0.066,
+      "step": 38
+    },
+    {
+      "epoch": 0.008863636363636363,
+      "grad_norm": 2.9431380076793725,
+      "learning_rate": 1.2497699697426523e-06,
+      "loss": 0.0855,
+      "step": 39
+    },
+    {
+      "epoch": 0.00909090909090909,
+      "grad_norm": 2.334877499989268,
+      "learning_rate": 1.2497577043788732e-06,
+      "loss": 0.0867,
+      "step": 40
+    },
+    {
+      "epoch": 0.009318181818181817,
+      "grad_norm": 2.3855494996020385,
+      "learning_rate": 1.2497451205176183e-06,
+      "loss": 0.1264,
+      "step": 41
+    },
+    {
+      "epoch": 0.009545454545454546,
+      "grad_norm": 2.1156951000772866,
+      "learning_rate": 1.2497322181653032e-06,
+      "loss": 0.1359,
+      "step": 42
+    },
+    {
+      "epoch": 0.009772727272727273,
+      "grad_norm": 2.507116569738136,
+      "learning_rate": 1.249718997328505e-06,
+      "loss": 0.1447,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4708856549531997,
+      "learning_rate": 1.2497054580139642e-06,
+      "loss": 0.0941,
+      "step": 44
+    },
+    {
+      "epoch": 0.010227272727272727,
+      "grad_norm": 3.9405320029202033,
+      "learning_rate": 1.2496916002285823e-06,
+      "loss": 0.1265,
+      "step": 45
+    },
+    {
+      "epoch": 0.010454545454545454,
+      "grad_norm": 2.63791984901363,
+      "learning_rate": 1.2496774239794246e-06,
+      "loss": 0.1248,
+      "step": 46
+    },
+    {
+      "epoch": 0.010681818181818181,
+      "grad_norm": 2.5345907881408136,
+      "learning_rate": 1.2496629292737176e-06,
+      "loss": 0.1683,
+      "step": 47
+    },
+    {
+      "epoch": 0.01090909090909091,
+      "grad_norm": 2.5907297598441246,
+      "learning_rate": 1.249648116118851e-06,
+      "loss": 0.0838,
+      "step": 48
+    },
+    {
+      "epoch": 0.011136363636363637,
+      "grad_norm": 1.657849698521552,
+      "learning_rate": 1.2496329845223759e-06,
+      "loss": 0.1263,
+      "step": 49
+    },
+    {
+      "epoch": 0.011363636363636364,
+      "grad_norm": 1.790591345862795,
+      "learning_rate": 1.2496175344920069e-06,
+      "loss": 0.0675,
+      "step": 50
+    },
+    {
+      "epoch": 0.011590909090909091,
+      "grad_norm": 2.9769824377999266,
+      "learning_rate": 1.24960176603562e-06,
+      "loss": 0.0776,
+      "step": 51
+    },
+    {
+      "epoch": 0.011818181818181818,
+      "grad_norm": 4.09324300622227,
+      "learning_rate": 1.2495856791612538e-06,
+      "loss": 0.1775,
+      "step": 52
+    },
+    {
+      "epoch": 0.012045454545454545,
+      "grad_norm": 1.9505514146615008,
+      "learning_rate": 1.2495692738771095e-06,
+      "loss": 0.1311,
+      "step": 53
+    },
+    {
+      "epoch": 0.012272727272727272,
+      "grad_norm": 3.340965635650713,
+      "learning_rate": 1.2495525501915503e-06,
+      "loss": 0.119,
+      "step": 54
+    },
+    {
+      "epoch": 0.0125,
+      "grad_norm": 2.590211579571482,
+      "learning_rate": 1.2495355081131017e-06,
+      "loss": 0.1047,
+      "step": 55
+    },
+    {
+      "epoch": 0.012727272727272728,
+      "grad_norm": 1.7126546389525177,
+      "learning_rate": 1.249518147650452e-06,
+      "loss": 0.0781,
+      "step": 56
+    },
+    {
+      "epoch": 0.012954545454545455,
+      "grad_norm": 1.6791481403257953,
+      "learning_rate": 1.249500468812451e-06,
+      "loss": 0.104,
+      "step": 57
+    },
+    {
+      "epoch": 0.013181818181818182,
+      "grad_norm": 2.772601988807807,
+      "learning_rate": 1.2494824716081117e-06,
+      "loss": 0.1044,
+      "step": 58
+    },
+    {
+      "epoch": 0.013409090909090909,
+      "grad_norm": 2.4287061769206924,
+      "learning_rate": 1.2494641560466087e-06,
+      "loss": 0.1063,
+      "step": 59
+    },
+    {
+      "epoch": 0.013636363636363636,
+      "grad_norm": 4.931772367793536,
+      "learning_rate": 1.249445522137279e-06,
+      "loss": 0.1982,
+      "step": 60
+    },
+    {
+      "epoch": 0.013863636363636364,
+      "grad_norm": 1.5683802428051412,
+      "learning_rate": 1.2494265698896224e-06,
+      "loss": 0.0687,
+      "step": 61
+    },
+    {
+      "epoch": 0.014090909090909091,
+      "grad_norm": 3.6622125210898666,
+      "learning_rate": 1.2494072993133005e-06,
+      "loss": 0.1115,
+      "step": 62
+    },
+    {
+      "epoch": 0.014318181818181818,
+      "grad_norm": 1.5981830241199848,
+      "learning_rate": 1.2493877104181373e-06,
+      "loss": 0.0918,
+      "step": 63
+    },
+    {
+      "epoch": 0.014545454545454545,
+      "grad_norm": 2.2461126255277444,
+      "learning_rate": 1.249367803214119e-06,
+      "loss": 0.0734,
+      "step": 64
+    },
+    {
+      "epoch": 0.014772727272727272,
+      "grad_norm": 2.9922052531051264,
+      "learning_rate": 1.2493475777113945e-06,
+      "loss": 0.0737,
+      "step": 65
+    },
+    {
+      "epoch": 0.015,
+      "grad_norm": 3.0304712723495446,
+      "learning_rate": 1.2493270339202742e-06,
+      "loss": 0.2087,
+      "step": 66
+    },
+    {
+      "epoch": 0.015227272727272726,
+      "grad_norm": 3.616756934691463,
+      "learning_rate": 1.2493061718512314e-06,
+      "loss": 0.0935,
+      "step": 67
+    },
+    {
+      "epoch": 0.015454545454545455,
+      "grad_norm": 3.4906124774385083,
+      "learning_rate": 1.2492849915149013e-06,
+      "loss": 0.1423,
+      "step": 68
+    },
+    {
+      "epoch": 0.015681818181818182,
+      "grad_norm": 3.038765170731685,
+      "learning_rate": 1.2492634929220817e-06,
+      "loss": 0.1127,
+      "step": 69
+    },
+    {
+      "epoch": 0.015909090909090907,
+      "grad_norm": 3.2114710879274244,
+      "learning_rate": 1.2492416760837326e-06,
+      "loss": 0.0896,
+      "step": 70
+    },
+    {
+      "epoch": 0.016136363636363636,
+      "grad_norm": 3.3772458882062177,
+      "learning_rate": 1.2492195410109757e-06,
+      "loss": 0.2259,
+      "step": 71
+    },
+    {
+      "epoch": 0.016363636363636365,
+      "grad_norm": 1.2392666587572057,
+      "learning_rate": 1.2491970877150955e-06,
+      "loss": 0.0739,
+      "step": 72
+    },
+    {
+      "epoch": 0.01659090909090909,
+      "grad_norm": 3.7352926778664997,
+      "learning_rate": 1.2491743162075384e-06,
+      "loss": 0.0809,
+      "step": 73
+    },
+    {
+      "epoch": 0.01681818181818182,
+      "grad_norm": 3.7495487663408498,
+      "learning_rate": 1.2491512264999135e-06,
+      "loss": 0.1764,
+      "step": 74
+    },
+    {
+      "epoch": 0.017045454545454544,
+      "grad_norm": 4.148932597238052,
+      "learning_rate": 1.2491278186039916e-06,
+      "loss": 0.1807,
+      "step": 75
+    },
+    {
+      "epoch": 0.017272727272727273,
+      "grad_norm": 3.382982224916496,
+      "learning_rate": 1.2491040925317057e-06,
+      "loss": 0.1111,
+      "step": 76
+    },
+    {
+      "epoch": 0.0175,
+      "grad_norm": 2.9498608287299772,
+      "learning_rate": 1.2490800482951515e-06,
+      "loss": 0.1479,
+      "step": 77
+    },
+    {
+      "epoch": 0.017727272727272727,
+      "grad_norm": 2.2171068439041144,
+      "learning_rate": 1.2490556859065865e-06,
+      "loss": 0.1084,
+      "step": 78
+    },
+    {
+      "epoch": 0.017954545454545456,
+      "grad_norm": 2.404748491524079,
+      "learning_rate": 1.2490310053784301e-06,
+      "loss": 0.1175,
+      "step": 79
+    },
+    {
+      "epoch": 0.01818181818181818,
+      "grad_norm": 3.213366108512472,
+      "learning_rate": 1.249006006723265e-06,
+      "loss": 0.0861,
+      "step": 80
+    },
+    {
+      "epoch": 0.01840909090909091,
+      "grad_norm": 2.2487279541462746,
+      "learning_rate": 1.2489806899538349e-06,
+      "loss": 0.0871,
+      "step": 81
+    },
+    {
+      "epoch": 0.018636363636363635,
+      "grad_norm": 1.3990597415449715,
+      "learning_rate": 1.2489550550830462e-06,
+      "loss": 0.0725,
+      "step": 82
+    },
+    {
+      "epoch": 0.018863636363636364,
+      "grad_norm": 2.957544631567143,
+      "learning_rate": 1.2489291021239674e-06,
+      "loss": 0.1963,
+      "step": 83
+    },
+    {
+      "epoch": 0.019090909090909092,
+      "grad_norm": 2.8956582173986587,
+      "learning_rate": 1.2489028310898293e-06,
+      "loss": 0.0843,
+      "step": 84
+    },
+    {
+      "epoch": 0.019318181818181818,
+      "grad_norm": 2.6354376087607654,
+      "learning_rate": 1.2488762419940244e-06,
+      "loss": 0.1849,
+      "step": 85
+    },
+    {
+      "epoch": 0.019545454545454546,
+      "grad_norm": 3.219587107200409,
+      "learning_rate": 1.248849334850108e-06,
+      "loss": 0.1339,
+      "step": 86
+    },
+    {
+      "epoch": 0.01977272727272727,
+      "grad_norm": 1.4032569908965096,
+      "learning_rate": 1.2488221096717967e-06,
+      "loss": 0.0918,
+      "step": 87
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8863054724096378,
+      "learning_rate": 1.2487945664729703e-06,
+      "loss": 0.1386,
+      "step": 88
+    },
+    {
+      "epoch": 0.020227272727272726,
+      "grad_norm": 2.6708806807596623,
+      "learning_rate": 1.2487667052676699e-06,
+      "loss": 0.181,
+      "step": 89
+    },
+    {
+      "epoch": 0.020454545454545454,
+      "grad_norm": 2.6673794745766406,
+      "learning_rate": 1.2487385260700987e-06,
+      "loss": 0.1636,
+      "step": 90
+    },
+    {
+      "epoch": 0.020681818181818183,
+      "grad_norm": 1.6036753171548173,
+      "learning_rate": 1.2487100288946228e-06,
+      "loss": 0.1055,
+      "step": 91
+    },
+    {
+      "epoch": 0.02090909090909091,
+      "grad_norm": 2.4371313674557507,
+      "learning_rate": 1.2486812137557693e-06,
+      "loss": 0.0817,
+      "step": 92
+    },
+    {
+      "epoch": 0.021136363636363637,
+      "grad_norm": 3.7090598436546007,
+      "learning_rate": 1.2486520806682283e-06,
+      "loss": 0.1273,
+      "step": 93
+    },
+    {
+      "epoch": 0.021363636363636362,
+      "grad_norm": 2.537560508386459,
+      "learning_rate": 1.248622629646852e-06,
+      "loss": 0.114,
+      "step": 94
+    },
+    {
+      "epoch": 0.02159090909090909,
+      "grad_norm": 3.4628601759308153,
+      "learning_rate": 1.2485928607066537e-06,
+      "loss": 0.1155,
+      "step": 95
+    },
+    {
+      "epoch": 0.02181818181818182,
+      "grad_norm": 3.399846064780786,
+      "learning_rate": 1.2485627738628097e-06,
+      "loss": 0.2044,
+      "step": 96
+    },
+    {
+      "epoch": 0.022045454545454545,
+      "grad_norm": 2.0772256662629176,
+      "learning_rate": 1.248532369130658e-06,
+      "loss": 0.0768,
+      "step": 97
+    },
+    {
+      "epoch": 0.022272727272727274,
+      "grad_norm": 2.7271493410189582,
+      "learning_rate": 1.2485016465256987e-06,
+      "loss": 0.0805,
+      "step": 98
+    },
+    {
+      "epoch": 0.0225,
+      "grad_norm": 3.9454937508078767,
+      "learning_rate": 1.2484706060635945e-06,
+      "loss": 0.0976,
+      "step": 99
+    },
+    {
+      "epoch": 0.022727272727272728,
+      "grad_norm": 4.287036939763489,
+      "learning_rate": 1.248439247760169e-06,
+      "loss": 0.1484,
+      "step": 100
+    },
+    {
+      "epoch": 0.022954545454545453,
+      "grad_norm": 3.621650790974889,
+      "learning_rate": 1.2484075716314085e-06,
+      "loss": 0.0815,
+      "step": 101
+    },
+    {
+      "epoch": 0.023181818181818182,
+      "grad_norm": 4.2363895224865,
+      "learning_rate": 1.2483755776934616e-06,
+      "loss": 0.2861,
+      "step": 102
+    },
+    {
+      "epoch": 0.02340909090909091,
+      "grad_norm": 3.3518543639439966,
+      "learning_rate": 1.2483432659626384e-06,
+      "loss": 0.1313,
+      "step": 103
+    },
+    {
+      "epoch": 0.023636363636363636,
+      "grad_norm": 1.6851817813430572,
+      "learning_rate": 1.2483106364554115e-06,
+      "loss": 0.1106,
+      "step": 104
+    },
+    {
+      "epoch": 0.023863636363636365,
+      "grad_norm": 3.7558796733114153,
+      "learning_rate": 1.248277689188415e-06,
+      "loss": 0.0833,
+      "step": 105
+    },
+    {
+      "epoch": 0.02409090909090909,
+      "grad_norm": 2.5185027449534556,
+      "learning_rate": 1.248244424178445e-06,
+      "loss": 0.2037,
+      "step": 106
+    },
+    {
+      "epoch": 0.02431818181818182,
+      "grad_norm": 4.483827071898472,
+      "learning_rate": 1.2482108414424602e-06,
+      "loss": 0.0956,
+      "step": 107
+    },
+    {
+      "epoch": 0.024545454545454544,
+      "grad_norm": 2.3022194497893427,
+      "learning_rate": 1.2481769409975805e-06,
+      "loss": 0.0865,
+      "step": 108
+    },
+    {
+      "epoch": 0.024772727272727273,
+      "grad_norm": 4.509365227731544,
+      "learning_rate": 1.2481427228610881e-06,
+      "loss": 0.1261,
+      "step": 109
+    },
+    {
+      "epoch": 0.025,
+      "grad_norm": 1.9078702077884082,
+      "learning_rate": 1.2481081870504278e-06,
+      "loss": 0.0878,
+      "step": 110
+    },
+    {
+      "epoch": 0.025227272727272727,
+      "grad_norm": 3.1598412687896937,
+      "learning_rate": 1.2480733335832052e-06,
+      "loss": 0.0829,
+      "step": 111
+    },
+    {
+      "epoch": 0.025454545454545455,
+      "grad_norm": 3.6971181252606655,
+      "learning_rate": 1.2480381624771882e-06,
+      "loss": 0.0914,
+      "step": 112
+    },
+    {
+      "epoch": 0.02568181818181818,
+      "grad_norm": 2.008987440641116,
+      "learning_rate": 1.2480026737503073e-06,
+      "loss": 0.1008,
+      "step": 113
+    },
+    {
+      "epoch": 0.02590909090909091,
+      "grad_norm": 2.99919645772555,
+      "learning_rate": 1.2479668674206543e-06,
+      "loss": 0.1245,
+      "step": 114
+    },
+    {
+      "epoch": 0.026136363636363635,
+      "grad_norm": 3.8748833281129578,
+      "learning_rate": 1.2479307435064827e-06,
+      "loss": 0.0976,
+      "step": 115
+    },
+    {
+      "epoch": 0.026363636363636363,
+      "grad_norm": 2.122199281855629,
+      "learning_rate": 1.2478943020262087e-06,
+      "loss": 0.1251,
+      "step": 116
+    },
+    {
+      "epoch": 0.026590909090909092,
+      "grad_norm": 2.492267421311055,
+      "learning_rate": 1.2478575429984097e-06,
+      "loss": 0.152,
+      "step": 117
+    },
+    {
+      "epoch": 0.026818181818181817,
+      "grad_norm": 2.653566603024868,
+      "learning_rate": 1.2478204664418254e-06,
+      "loss": 0.0991,
+      "step": 118
+    },
+    {
+      "epoch": 0.027045454545454546,
+      "grad_norm": 4.198276247734506,
+      "learning_rate": 1.2477830723753567e-06,
+      "loss": 0.1579,
+      "step": 119
+    },
+    {
+      "epoch": 0.02727272727272727,
+      "grad_norm": 3.899807537109214,
+      "learning_rate": 1.2477453608180673e-06,
+      "loss": 0.1107,
+      "step": 120
+    },
+    {
+      "epoch": 0.0275,
+      "grad_norm": 2.3833320390188555,
+      "learning_rate": 1.2477073317891822e-06,
+      "loss": 0.1554,
+      "step": 121
+    },
+    {
+      "epoch": 0.02772727272727273,
+      "grad_norm": 3.8212587869189245,
+      "learning_rate": 1.2476689853080883e-06,
+      "loss": 0.0865,
+      "step": 122
+    },
+    {
+      "epoch": 0.027954545454545454,
+      "grad_norm": 1.9136882810464502,
+      "learning_rate": 1.2476303213943346e-06,
+      "loss": 0.0787,
+      "step": 123
+    },
+    {
+      "epoch": 0.028181818181818183,
+      "grad_norm": 1.9864512670945367,
+      "learning_rate": 1.2475913400676314e-06,
+      "loss": 0.0679,
+      "step": 124
+    },
+    {
+      "epoch": 0.028409090909090908,
+      "grad_norm": 3.057361348148449,
+      "learning_rate": 1.2475520413478516e-06,
+      "loss": 0.1558,
+      "step": 125
+    },
+    {
+      "epoch": 0.028636363636363637,
+      "grad_norm": 2.4559864848992423,
+      "learning_rate": 1.247512425255029e-06,
+      "loss": 0.0568,
+      "step": 126
+    },
+    {
+      "epoch": 0.028863636363636362,
+      "grad_norm": 2.3064987947510125,
+      "learning_rate": 1.2474724918093594e-06,
+      "loss": 0.0874,
+      "step": 127
+    },
+    {
+      "epoch": 0.02909090909090909,
+      "grad_norm": 3.4630333466988286,
+      "learning_rate": 1.2474322410312012e-06,
+      "loss": 0.1433,
+      "step": 128
+    },
+    {
+      "epoch": 0.02931818181818182,
+      "grad_norm": 3.391975827415409,
+      "learning_rate": 1.247391672941074e-06,
+      "loss": 0.0829,
+      "step": 129
+    },
+    {
+      "epoch": 0.029545454545454545,
+      "grad_norm": 3.2248882073989957,
+      "learning_rate": 1.2473507875596586e-06,
+      "loss": 0.1843,
+      "step": 130
+    },
+    {
+      "epoch": 0.029772727272727274,
+      "grad_norm": 4.118551536256316,
+      "learning_rate": 1.2473095849077984e-06,
+      "loss": 0.1819,
+      "step": 131
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.137811432545649,
+      "learning_rate": 1.2472680650064984e-06,
+      "loss": 0.1778,
+      "step": 132
+    },
+    {
+      "epoch": 0.030227272727272728,
+      "grad_norm": 3.6247559276452463,
+      "learning_rate": 1.247226227876925e-06,
+      "loss": 0.161,
+      "step": 133
+    },
+    {
+      "epoch": 0.030454545454545453,
+      "grad_norm": 1.7953374315892745,
+      "learning_rate": 1.2471840735404066e-06,
+      "loss": 0.0475,
+      "step": 134
+    },
+    {
+      "epoch": 0.03068181818181818,
+      "grad_norm": 3.9170817361284764,
+      "learning_rate": 1.2471416020184332e-06,
+      "loss": 0.1046,
+      "step": 135
+    },
+    {
+      "epoch": 0.03090909090909091,
+      "grad_norm": 3.193892805213982,
+      "learning_rate": 1.2470988133326564e-06,
+      "loss": 0.0783,
+      "step": 136
+    },
+    {
+      "epoch": 0.031136363636363636,
+      "grad_norm": 2.392755985657823,
+      "learning_rate": 1.2470557075048897e-06,
+      "loss": 0.114,
+      "step": 137
+    },
+    {
+      "epoch": 0.031363636363636364,
+      "grad_norm": 4.884480413611041,
+      "learning_rate": 1.2470122845571081e-06,
+      "loss": 0.1408,
+      "step": 138
+    },
+    {
+      "epoch": 0.03159090909090909,
+      "grad_norm": 2.3704649582296926,
+      "learning_rate": 1.2469685445114486e-06,
+      "loss": 0.0848,
+      "step": 139
+    },
+    {
+      "epoch": 0.031818181818181815,
+      "grad_norm": 2.2985560540167653,
+      "learning_rate": 1.2469244873902089e-06,
+      "loss": 0.1029,
+      "step": 140
+    },
+    {
+      "epoch": 0.032045454545454544,
+      "grad_norm": 5.319831122659349,
+      "learning_rate": 1.2468801132158499e-06,
+      "loss": 0.2105,
+      "step": 141
+    },
+    {
+      "epoch": 0.03227272727272727,
+      "grad_norm": 2.7266103565527855,
+      "learning_rate": 1.2468354220109926e-06,
+      "loss": 0.1006,
+      "step": 142
+    },
+    {
+      "epoch": 0.0325,
+      "grad_norm": 2.744775358082282,
+      "learning_rate": 1.2467904137984208e-06,
+      "loss": 0.125,
+      "step": 143
+    },
+    {
+      "epoch": 0.03272727272727273,
+      "grad_norm": 4.2252693306222975,
+      "learning_rate": 1.246745088601079e-06,
+      "loss": 0.0937,
+      "step": 144
+    },
+    {
+      "epoch": 0.03295454545454545,
+      "grad_norm": 1.8301450973591429,
+      "learning_rate": 1.246699446442074e-06,
+      "loss": 0.0974,
+      "step": 145
+    },
+    {
+      "epoch": 0.03318181818181818,
+      "grad_norm": 2.178641124699276,
+      "learning_rate": 1.2466534873446738e-06,
+      "loss": 0.1036,
+      "step": 146
+    },
+    {
+      "epoch": 0.03340909090909091,
+      "grad_norm": 3.2479690123414993,
+      "learning_rate": 1.246607211332308e-06,
+      "loss": 0.143,
+      "step": 147
+    },
+    {
+      "epoch": 0.03363636363636364,
+      "grad_norm": 2.7161801022330376,
+      "learning_rate": 1.2465606184285679e-06,
+      "loss": 0.0784,
+      "step": 148
+    },
+    {
+      "epoch": 0.03386363636363637,
+      "grad_norm": 2.946252975196419,
+      "learning_rate": 1.2465137086572057e-06,
+      "loss": 0.1087,
+      "step": 149
+    },
+    {
+      "epoch": 0.03409090909090909,
+      "grad_norm": 3.7646888424049143,
+      "learning_rate": 1.2464664820421365e-06,
+      "loss": 0.2077,
+      "step": 150
+    },
+    {
+      "epoch": 0.03431818181818182,
+      "grad_norm": 2.0046468126086787,
+      "learning_rate": 1.246418938607436e-06,
+      "loss": 0.1261,
+      "step": 151
+    },
+    {
+      "epoch": 0.034545454545454546,
+      "grad_norm": 1.4259818662884742,
+      "learning_rate": 1.246371078377341e-06,
+      "loss": 0.0542,
+      "step": 152
+    },
+    {
+      "epoch": 0.034772727272727275,
+      "grad_norm": 2.8762613449347647,
+      "learning_rate": 1.246322901376251e-06,
+      "loss": 0.1723,
+      "step": 153
+    },
+    {
+      "epoch": 0.035,
+      "grad_norm": 3.9290196607015493,
+      "learning_rate": 1.2462744076287257e-06,
+      "loss": 0.1381,
+      "step": 154
+    },
+    {
+      "epoch": 0.035227272727272725,
+      "grad_norm": 3.164945939472189,
+      "learning_rate": 1.2462255971594874e-06,
+      "loss": 0.0871,
+      "step": 155
+    },
+    {
+      "epoch": 0.035454545454545454,
+      "grad_norm": 2.3422239909264024,
+      "learning_rate": 1.2461764699934192e-06,
+      "loss": 0.0733,
+      "step": 156
+    },
+    {
+      "epoch": 0.03568181818181818,
+      "grad_norm": 3.1067031772787477,
+      "learning_rate": 1.2461270261555657e-06,
+      "loss": 0.1047,
+      "step": 157
+    },
+    {
+      "epoch": 0.03590909090909091,
+      "grad_norm": 4.2162583523635115,
+      "learning_rate": 1.246077265671133e-06,
+      "loss": 0.1415,
+      "step": 158
+    },
+    {
+      "epoch": 0.03613636363636363,
+      "grad_norm": 2.7951422230559486,
+      "learning_rate": 1.2460271885654891e-06,
+      "loss": 0.0885,
+      "step": 159
+    },
+    {
+      "epoch": 0.03636363636363636,
+      "grad_norm": 2.048658317870938,
+      "learning_rate": 1.2459767948641626e-06,
+      "loss": 0.0835,
+      "step": 160
+    },
+    {
+      "epoch": 0.03659090909090909,
+      "grad_norm": 3.1707408130065264,
+      "learning_rate": 1.2459260845928438e-06,
+      "loss": 0.1229,
+      "step": 161
+    },
+    {
+      "epoch": 0.03681818181818182,
+      "grad_norm": 3.0335907524673154,
+      "learning_rate": 1.245875057777385e-06,
+      "loss": 0.1188,
+      "step": 162
+    },
+    {
+      "epoch": 0.03704545454545455,
+      "grad_norm": 2.3200226209196955,
+      "learning_rate": 1.245823714443799e-06,
+      "loss": 0.064,
+      "step": 163
+    },
+    {
+      "epoch": 0.03727272727272727,
+      "grad_norm": 1.9954265011198742,
+      "learning_rate": 1.24577205461826e-06,
+      "loss": 0.0638,
+      "step": 164
+    },
+    {
+      "epoch": 0.0375,
+      "grad_norm": 2.236806781948131,
+      "learning_rate": 1.2457200783271044e-06,
+      "loss": 0.0863,
+      "step": 165
+    },
+    {
+      "epoch": 0.03772727272727273,
+      "grad_norm": 3.0742733194769123,
+      "learning_rate": 1.245667785596829e-06,
+      "loss": 0.2,
+      "step": 166
+    },
+    {
+      "epoch": 0.037954545454545456,
+      "grad_norm": 2.2520226325508643,
+      "learning_rate": 1.2456151764540924e-06,
+      "loss": 0.0809,
+      "step": 167
+    },
+    {
+      "epoch": 0.038181818181818185,
+      "grad_norm": 2.15663011362523,
+      "learning_rate": 1.2455622509257147e-06,
+      "loss": 0.0831,
+      "step": 168
+    },
+    {
+      "epoch": 0.03840909090909091,
+      "grad_norm": 3.7997476231270535,
+      "learning_rate": 1.2455090090386765e-06,
+      "loss": 0.1603,
+      "step": 169
+    },
+    {
+      "epoch": 0.038636363636363635,
+      "grad_norm": 3.9530380374838114,
+      "learning_rate": 1.2454554508201205e-06,
+      "loss": 0.075,
+      "step": 170
+    },
+    {
+      "epoch": 0.038863636363636364,
+      "grad_norm": 2.4758478414591374,
+      "learning_rate": 1.2454015762973505e-06,
+      "loss": 0.1126,
+      "step": 171
+    },
+    {
+      "epoch": 0.03909090909090909,
+      "grad_norm": 2.9280923203486258,
+      "learning_rate": 1.2453473854978307e-06,
+      "loss": 0.1125,
+      "step": 172
+    },
+    {
+      "epoch": 0.03931818181818182,
+      "grad_norm": 1.857509727053121,
+      "learning_rate": 1.2452928784491877e-06,
+      "loss": 0.1214,
+      "step": 173
+    },
+    {
+      "epoch": 0.03954545454545454,
+      "grad_norm": 1.7565974252145562,
+      "learning_rate": 1.245238055179209e-06,
+      "loss": 0.0941,
+      "step": 174
+    },
+    {
+      "epoch": 0.03977272727272727,
+      "grad_norm": 3.214045178112143,
+      "learning_rate": 1.245182915715843e-06,
+      "loss": 0.1067,
+      "step": 175
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.1459596309788425,
+      "learning_rate": 1.2451274600871991e-06,
+      "loss": 0.0903,
+      "step": 176
+    },
+    {
+      "epoch": 0.04022727272727273,
+      "grad_norm": 2.717741153219287,
+      "learning_rate": 1.245071688321549e-06,
+      "loss": 0.0933,
+      "step": 177
+    },
+    {
+      "epoch": 0.04045454545454545,
+      "grad_norm": 2.433999370162766,
+      "learning_rate": 1.2450156004473238e-06,
+      "loss": 0.1544,
+      "step": 178
+    },
+    {
+      "epoch": 0.04068181818181818,
+      "grad_norm": 1.522448767303929,
+      "learning_rate": 1.2449591964931173e-06,
+      "loss": 0.1366,
+      "step": 179
+    },
+    {
+      "epoch": 0.04090909090909091,
+      "grad_norm": 3.3266491447077904,
+      "learning_rate": 1.2449024764876841e-06,
+      "loss": 0.1176,
+      "step": 180
+    },
+    {
+      "epoch": 0.04113636363636364,
+      "grad_norm": 4.043282241976147,
+      "learning_rate": 1.2448454404599393e-06,
+      "loss": 0.2645,
+      "step": 181
+    },
+    {
+      "epoch": 0.041363636363636366,
+      "grad_norm": 3.4155863239617363,
+      "learning_rate": 1.2447880884389597e-06,
+      "loss": 0.0803,
+      "step": 182
+    },
+    {
+      "epoch": 0.04159090909090909,
+      "grad_norm": 2.9942111557293773,
+      "learning_rate": 1.2447304204539827e-06,
+      "loss": 0.0674,
+      "step": 183
+    },
+    {
+      "epoch": 0.04181818181818182,
+      "grad_norm": 2.418264089227234,
+      "learning_rate": 1.2446724365344076e-06,
+      "loss": 0.0624,
+      "step": 184
+    },
+    {
+      "epoch": 0.042045454545454546,
+      "grad_norm": 3.0587436770158467,
+      "learning_rate": 1.2446141367097936e-06,
+      "loss": 0.0967,
+      "step": 185
+    },
+    {
+      "epoch": 0.042272727272727274,
+      "grad_norm": 4.4237769530614175,
+      "learning_rate": 1.244555521009862e-06,
+      "loss": 0.1264,
+      "step": 186
+    },
+    {
+      "epoch": 0.0425,
+      "grad_norm": 3.1564699048311935,
+      "learning_rate": 1.2444965894644946e-06,
+      "loss": 0.0967,
+      "step": 187
+    },
+    {
+      "epoch": 0.042727272727272725,
+      "grad_norm": 6.667224811522091,
+      "learning_rate": 1.2444373421037345e-06,
+      "loss": 0.1933,
+      "step": 188
+    },
+    {
+      "epoch": 0.042954545454545454,
+      "grad_norm": 3.4612803721893286,
+      "learning_rate": 1.2443777789577852e-06,
+      "loss": 0.1177,
+      "step": 189
+    },
+    {
+      "epoch": 0.04318181818181818,
+      "grad_norm": 3.0645036281205753,
+      "learning_rate": 1.244317900057012e-06,
+      "loss": 0.0872,
+      "step": 190
+    },
+    {
+      "epoch": 0.04340909090909091,
+      "grad_norm": 2.7678849500637432,
+      "learning_rate": 1.2442577054319405e-06,
+      "loss": 0.1107,
+      "step": 191
+    },
+    {
+      "epoch": 0.04363636363636364,
+      "grad_norm": 2.429853550112927,
+      "learning_rate": 1.2441971951132578e-06,
+      "loss": 0.0728,
+      "step": 192
+    },
+    {
+      "epoch": 0.04386363636363636,
+      "grad_norm": 3.36528155987506,
+      "learning_rate": 1.2441363691318114e-06,
+      "loss": 0.1021,
+      "step": 193
+    },
+    {
+      "epoch": 0.04409090909090909,
+      "grad_norm": 2.92737772200131,
+      "learning_rate": 1.2440752275186102e-06,
+      "loss": 0.0724,
+      "step": 194
+    },
+    {
+      "epoch": 0.04431818181818182,
+      "grad_norm": 1.680881694384273,
+      "learning_rate": 1.244013770304824e-06,
+      "loss": 0.0557,
+      "step": 195
+    },
+    {
+      "epoch": 0.04454545454545455,
+      "grad_norm": 1.3839129700771302,
+      "learning_rate": 1.2439519975217828e-06,
+      "loss": 0.0334,
+      "step": 196
+    },
+    {
+      "epoch": 0.04477272727272727,
+      "grad_norm": 3.580271334258035,
+      "learning_rate": 1.2438899092009783e-06,
+      "loss": 0.0899,
+      "step": 197
+    },
+    {
+      "epoch": 0.045,
+      "grad_norm": 3.6568327278776147,
+      "learning_rate": 1.2438275053740624e-06,
+      "loss": 0.1196,
+      "step": 198
+    },
+    {
+      "epoch": 0.04522727272727273,
+      "grad_norm": 2.5412787688165275,
+      "learning_rate": 1.2437647860728487e-06,
+      "loss": 0.0895,
+      "step": 199
+    },
+    {
+      "epoch": 0.045454545454545456,
+      "grad_norm": 3.3934172415051176,
+      "learning_rate": 1.2437017513293107e-06,
+      "loss": 0.1361,
+      "step": 200
+    },
+    {
+      "epoch": 0.045681818181818185,
+      "grad_norm": 2.555977878825042,
+      "learning_rate": 1.243638401175583e-06,
+      "loss": 0.1095,
+      "step": 201
+    },
+    {
+      "epoch": 0.045909090909090906,
+      "grad_norm": 3.716546846964998,
+      "learning_rate": 1.2435747356439614e-06,
+      "loss": 0.2058,
+      "step": 202
+    },
+    {
+      "epoch": 0.046136363636363635,
+      "grad_norm": 2.4595525452664946,
+      "learning_rate": 1.2435107547669022e-06,
+      "loss": 0.0859,
+      "step": 203
+    },
+    {
+      "epoch": 0.046363636363636364,
+      "grad_norm": 2.6578162895974224,
+      "learning_rate": 1.2434464585770226e-06,
+      "loss": 0.0536,
+      "step": 204
+    },
+    {
+      "epoch": 0.04659090909090909,
+      "grad_norm": 3.897584603854884,
+      "learning_rate": 1.2433818471070998e-06,
+      "loss": 0.1557,
+      "step": 205
+    },
+    {
+      "epoch": 0.04681818181818182,
+      "grad_norm": 6.822349149466114,
+      "learning_rate": 1.2433169203900726e-06,
+      "loss": 0.266,
+      "step": 206
+    },
+    {
+      "epoch": 0.04704545454545454,
+      "grad_norm": 3.4167912842651313,
+      "learning_rate": 1.2432516784590405e-06,
+      "loss": 0.0961,
+      "step": 207
+    },
+    {
+      "epoch": 0.04727272727272727,
+      "grad_norm": 3.4982383483847217,
+      "learning_rate": 1.243186121347263e-06,
+      "loss": 0.1032,
+      "step": 208
+    },
+    {
+      "epoch": 0.0475,
+      "grad_norm": 2.9308534972205136,
+      "learning_rate": 1.243120249088161e-06,
+      "loss": 0.0921,
+      "step": 209
+    },
+    {
+      "epoch": 0.04772727272727273,
+      "grad_norm": 2.1650949857926864,
+      "learning_rate": 1.2430540617153156e-06,
+      "loss": 0.07,
+      "step": 210
+    },
+    {
+      "epoch": 0.04795454545454545,
+      "grad_norm": 1.877123728582618,
+      "learning_rate": 1.2429875592624685e-06,
+      "loss": 0.1097,
+      "step": 211
+    },
+    {
+      "epoch": 0.04818181818181818,
+      "grad_norm": 2.8812892658842917,
+      "learning_rate": 1.2429207417635226e-06,
+      "loss": 0.0954,
+      "step": 212
+    },
+    {
+      "epoch": 0.04840909090909091,
+      "grad_norm": 1.9447089447685935,
+      "learning_rate": 1.242853609252541e-06,
+      "loss": 0.0868,
+      "step": 213
+    },
+    {
+      "epoch": 0.04863636363636364,
+      "grad_norm": 4.572871957317649,
+      "learning_rate": 1.2427861617637472e-06,
+      "loss": 0.1737,
+      "step": 214
+    },
+    {
+      "epoch": 0.048863636363636366,
+      "grad_norm": 2.6223510240500882,
+      "learning_rate": 1.2427183993315256e-06,
+      "loss": 0.0562,
+      "step": 215
+    },
+    {
+      "epoch": 0.04909090909090909,
+      "grad_norm": 3.0747367957710385,
+      "learning_rate": 1.2426503219904213e-06,
+      "loss": 0.1262,
+      "step": 216
+    },
+    {
+      "epoch": 0.04931818181818182,
+      "grad_norm": 2.390349193340806,
+      "learning_rate": 1.242581929775139e-06,
+      "loss": 0.0824,
+      "step": 217
+    },
+    {
+      "epoch": 0.049545454545454545,
+      "grad_norm": 2.547736820961635,
+      "learning_rate": 1.2425132227205456e-06,
+      "loss": 0.0705,
+      "step": 218
+    },
+    {
+      "epoch": 0.049772727272727274,
+      "grad_norm": 1.012061665511356,
+      "learning_rate": 1.2424442008616667e-06,
+      "loss": 0.0371,
+      "step": 219
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7563787988146093,
+      "learning_rate": 1.2423748642336894e-06,
+      "loss": 0.0995,
+      "step": 220
+    },
+    {
+      "epoch": 0.050227272727272725,
+      "grad_norm": 2.260009759505115,
+      "learning_rate": 1.2423052128719611e-06,
+      "loss": 0.0758,
+      "step": 221
+    },
+    {
+      "epoch": 0.05045454545454545,
+      "grad_norm": 1.3375603854760587,
+      "learning_rate": 1.24223524681199e-06,
+      "loss": 0.0628,
+      "step": 222
+    },
+    {
+      "epoch": 0.05068181818181818,
+      "grad_norm": 2.6771117922046397,
+      "learning_rate": 1.2421649660894438e-06,
+      "loss": 0.1246,
+      "step": 223
+    },
+    {
+      "epoch": 0.05090909090909091,
+      "grad_norm": 3.7619917996055863,
+      "learning_rate": 1.2420943707401514e-06,
+      "loss": 0.0976,
+      "step": 224
+    },
+    {
+      "epoch": 0.05113636363636364,
+      "grad_norm": 3.0848514275552144,
+      "learning_rate": 1.2420234608001017e-06,
+      "loss": 0.1937,
+      "step": 225
+    },
+    {
+      "epoch": 0.05136363636363636,
+      "grad_norm": 2.4374680912316053,
+      "learning_rate": 1.2419522363054446e-06,
+      "loss": 0.0848,
+      "step": 226
+    },
+    {
+      "epoch": 0.05159090909090909,
+      "grad_norm": 3.7191848287996025,
+      "learning_rate": 1.2418806972924893e-06,
+      "loss": 0.0953,
+      "step": 227
+    },
+    {
+      "epoch": 0.05181818181818182,
+      "grad_norm": 2.66335814077733,
+      "learning_rate": 1.2418088437977063e-06,
+      "loss": 0.1009,
+      "step": 228
+    },
+    {
+      "epoch": 0.05204545454545455,
+      "grad_norm": 2.412809406877248,
+      "learning_rate": 1.241736675857726e-06,
+      "loss": 0.0675,
+      "step": 229
+    },
+    {
+      "epoch": 0.05227272727272727,
+      "grad_norm": 4.65044604462548,
+      "learning_rate": 1.241664193509339e-06,
+      "loss": 0.096,
+      "step": 230
+    },
+    {
+      "epoch": 0.0525,
+      "grad_norm": 5.09363480299026,
+      "learning_rate": 1.2415913967894966e-06,
+      "loss": 0.2421,
+      "step": 231
+    },
+    {
+      "epoch": 0.05272727272727273,
+      "grad_norm": 4.056815516837104,
+      "learning_rate": 1.2415182857353098e-06,
+      "loss": 0.1771,
+      "step": 232
+    },
+    {
+      "epoch": 0.052954545454545456,
+      "grad_norm": 2.4409832544314236,
+      "learning_rate": 1.2414448603840504e-06,
+      "loss": 0.1381,
+      "step": 233
+    },
+    {
+      "epoch": 0.053181818181818184,
+      "grad_norm": 3.30516224546615,
+      "learning_rate": 1.24137112077315e-06,
+      "loss": 0.2239,
+      "step": 234
+    },
+    {
+      "epoch": 0.053409090909090906,
+      "grad_norm": 4.750181242796731,
+      "learning_rate": 1.2412970669402005e-06,
+      "loss": 0.3172,
+      "step": 235
+    },
+    {
+      "epoch": 0.053636363636363635,
+      "grad_norm": 3.3363001291418883,
+      "learning_rate": 1.2412226989229542e-06,
+      "loss": 0.1871,
+      "step": 236
+    },
+    {
+      "epoch": 0.053863636363636364,
+      "grad_norm": 4.902123827255867,
+      "learning_rate": 1.2411480167593237e-06,
+      "loss": 0.1167,
+      "step": 237
+    },
+    {
+      "epoch": 0.05409090909090909,
+      "grad_norm": 2.140890406636027,
+      "learning_rate": 1.241073020487381e-06,
+      "loss": 0.0991,
+      "step": 238
+    },
+    {
+      "epoch": 0.05431818181818182,
+      "grad_norm": 1.460656936347126,
+      "learning_rate": 1.2409977101453591e-06,
+      "loss": 0.0585,
+      "step": 239
+    },
+    {
+      "epoch": 0.05454545454545454,
+      "grad_norm": 3.870074973045016,
+      "learning_rate": 1.2409220857716506e-06,
+      "loss": 0.1517,
+      "step": 240
+    },
+    {
+      "epoch": 0.05477272727272727,
+      "grad_norm": 3.159943045639632,
+      "learning_rate": 1.2408461474048083e-06,
+      "loss": 0.1196,
+      "step": 241
+    },
+    {
+      "epoch": 0.055,
+      "grad_norm": 4.913069714809017,
+      "learning_rate": 1.240769895083545e-06,
+      "loss": 0.1065,
+      "step": 242
+    },
+    {
+      "epoch": 0.05522727272727273,
+      "grad_norm": 2.699614594241355,
+      "learning_rate": 1.2406933288467337e-06,
+      "loss": 0.163,
+      "step": 243
+    },
+    {
+      "epoch": 0.05545454545454546,
+      "grad_norm": 2.587714516355004,
+      "learning_rate": 1.2406164487334077e-06,
+      "loss": 0.1627,
+      "step": 244
+    },
+    {
+      "epoch": 0.05568181818181818,
+      "grad_norm": 3.600462258550555,
+      "learning_rate": 1.2405392547827594e-06,
+      "loss": 0.1118,
+      "step": 245
+    },
+    {
+      "epoch": 0.05590909090909091,
+      "grad_norm": 2.1014376490107316,
+      "learning_rate": 1.2404617470341423e-06,
+      "loss": 0.063,
+      "step": 246
+    },
+    {
+      "epoch": 0.05613636363636364,
+      "grad_norm": 3.921834652165585,
+      "learning_rate": 1.2403839255270693e-06,
+      "loss": 0.1024,
+      "step": 247
+    },
+    {
+      "epoch": 0.056363636363636366,
+      "grad_norm": 2.633424805673825,
+      "learning_rate": 1.2403057903012128e-06,
+      "loss": 0.1147,
+      "step": 248
+    },
+    {
+      "epoch": 0.05659090909090909,
+      "grad_norm": 2.774498349726601,
+      "learning_rate": 1.240227341396406e-06,
+      "loss": 0.0734,
+      "step": 249
+    },
+    {
+      "epoch": 0.056818181818181816,
+      "grad_norm": 3.1093286010353913,
+      "learning_rate": 1.2401485788526418e-06,
+      "loss": 0.2027,
+      "step": 250
+    },
+    {
+      "epoch": 0.057045454545454545,
+      "grad_norm": 2.6714784300800973,
+      "learning_rate": 1.2400695027100725e-06,
+      "loss": 0.1348,
+      "step": 251
+    },
+    {
+      "epoch": 0.057272727272727274,
+      "grad_norm": 2.841094984583613,
+      "learning_rate": 1.2399901130090112e-06,
+      "loss": 0.1093,
+      "step": 252
+    },
+    {
+      "epoch": 0.0575,
+      "grad_norm": 3.9926828089880555,
+      "learning_rate": 1.2399104097899295e-06,
+      "loss": 0.1566,
+      "step": 253
+    },
+    {
+      "epoch": 0.057727272727272724,
+      "grad_norm": 2.0426673539860234,
+      "learning_rate": 1.2398303930934601e-06,
+      "loss": 0.0719,
+      "step": 254
+    },
+    {
+      "epoch": 0.05795454545454545,
+      "grad_norm": 3.4022810910147174,
+      "learning_rate": 1.2397500629603948e-06,
+      "loss": 0.1013,
+      "step": 255
+    },
+    {
+      "epoch": 0.05818181818181818,
+      "grad_norm": 3.079425326642514,
+      "learning_rate": 1.2396694194316851e-06,
+      "loss": 0.0766,
+      "step": 256
+    },
+    {
+      "epoch": 0.05840909090909091,
+      "grad_norm": 2.267428155252032,
+      "learning_rate": 1.2395884625484433e-06,
+      "loss": 0.1396,
+      "step": 257
+    },
+    {
+      "epoch": 0.05863636363636364,
+      "grad_norm": 4.2457069865097665,
+      "learning_rate": 1.2395071923519403e-06,
+      "loss": 0.1329,
+      "step": 258
+    },
+    {
+      "epoch": 0.05886363636363636,
+      "grad_norm": 4.326598150136959,
+      "learning_rate": 1.2394256088836069e-06,
+      "loss": 0.1462,
+      "step": 259
+    },
+    {
+      "epoch": 0.05909090909090909,
+      "grad_norm": 1.8812568816427564,
+      "learning_rate": 1.2393437121850342e-06,
+      "loss": 0.1002,
+      "step": 260
+    },
+    {
+      "epoch": 0.05931818181818182,
+      "grad_norm": 3.046227516360263,
+      "learning_rate": 1.2392615022979723e-06,
+      "loss": 0.108,
+      "step": 261
+    },
+    {
+      "epoch": 0.05954545454545455,
+      "grad_norm": 3.241942095267392,
+      "learning_rate": 1.2391789792643317e-06,
+      "loss": 0.103,
+      "step": 262
+    },
+    {
+      "epoch": 0.059772727272727276,
+      "grad_norm": 5.181675389134145,
+      "learning_rate": 1.2390961431261814e-06,
+      "loss": 0.2546,
+      "step": 263
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4263236911686374,
+      "learning_rate": 1.2390129939257515e-06,
+      "loss": 0.1358,
+      "step": 264
+    },
+    {
+      "epoch": 0.060227272727272727,
+      "grad_norm": 3.8158275451730352,
+      "learning_rate": 1.2389295317054306e-06,
+      "loss": 0.1787,
+      "step": 265
+    },
+    {
+      "epoch": 0.060454545454545455,
+      "grad_norm": 3.4937371000310717,
+      "learning_rate": 1.238845756507767e-06,
+      "loss": 0.1129,
+      "step": 266
+    },
+    {
+      "epoch": 0.060681818181818184,
+      "grad_norm": 5.087808693070465,
+      "learning_rate": 1.2387616683754691e-06,
+      "loss": 0.1218,
+      "step": 267
+    },
+    {
+      "epoch": 0.060909090909090906,
+      "grad_norm": 3.3657773309614973,
+      "learning_rate": 1.2386772673514044e-06,
+      "loss": 0.0644,
+      "step": 268
+    },
+    {
+      "epoch": 0.061136363636363635,
+      "grad_norm": 3.765264072041106,
+      "learning_rate": 1.2385925534786e-06,
+      "loss": 0.1217,
+      "step": 269
+    },
+    {
+      "epoch": 0.06136363636363636,
+      "grad_norm": 1.986880990295197,
+      "learning_rate": 1.2385075268002423e-06,
+      "loss": 0.0744,
+      "step": 270
+    },
+    {
+      "epoch": 0.06159090909090909,
+      "grad_norm": 1.654849985156512,
+      "learning_rate": 1.2384221873596775e-06,
+      "loss": 0.0575,
+      "step": 271
+    },
+    {
+      "epoch": 0.06181818181818182,
+      "grad_norm": 3.6900447222147807,
+      "learning_rate": 1.2383365352004111e-06,
+      "loss": 0.1439,
+      "step": 272
+    },
+    {
+      "epoch": 0.06204545454545454,
+      "grad_norm": 2.4405793751846394,
+      "learning_rate": 1.238250570366108e-06,
+      "loss": 0.1299,
+      "step": 273
+    },
+    {
+      "epoch": 0.06227272727272727,
+      "grad_norm": 2.6310308209955564,
+      "learning_rate": 1.2381642929005927e-06,
+      "loss": 0.0983,
+      "step": 274
+    },
+    {
+      "epoch": 0.0625,
+      "grad_norm": 4.5925254895336725,
+      "learning_rate": 1.238077702847849e-06,
+      "loss": 0.1707,
+      "step": 275
+    },
+    {
+      "epoch": 0.06272727272727273,
+      "grad_norm": 2.9535456356683905,
+      "learning_rate": 1.2379908002520191e-06,
+      "loss": 0.1576,
+      "step": 276
+    },
+    {
+      "epoch": 0.06295454545454546,
+      "grad_norm": 3.785953958589258,
+      "learning_rate": 1.2379035851574063e-06,
+      "loss": 0.1612,
+      "step": 277
+    },
+    {
+      "epoch": 0.06318181818181819,
+      "grad_norm": 2.7382228340177357,
+      "learning_rate": 1.237816057608472e-06,
+      "loss": 0.1532,
+      "step": 278
+    },
+    {
+      "epoch": 0.06340909090909091,
+      "grad_norm": 2.7532145912398005,
+      "learning_rate": 1.2377282176498371e-06,
+      "loss": 0.0581,
+      "step": 279
+    },
+    {
+      "epoch": 0.06363636363636363,
+      "grad_norm": 4.383515501701478,
+      "learning_rate": 1.2376400653262817e-06,
+      "loss": 0.1983,
+      "step": 280
+    },
+    {
+      "epoch": 0.06386363636363636,
+      "grad_norm": 2.8425813883365993,
+      "learning_rate": 1.237551600682746e-06,
+      "loss": 0.1144,
+      "step": 281
+    },
+    {
+      "epoch": 0.06409090909090909,
+      "grad_norm": 2.350939161092161,
+      "learning_rate": 1.237462823764328e-06,
+      "loss": 0.0955,
+      "step": 282
+    },
+    {
+      "epoch": 0.06431818181818182,
+      "grad_norm": 1.996724867356844,
+      "learning_rate": 1.2373737346162857e-06,
+      "loss": 0.0639,
+      "step": 283
+    },
+    {
+      "epoch": 0.06454545454545454,
+      "grad_norm": 1.9085859990147678,
+      "learning_rate": 1.2372843332840364e-06,
+      "loss": 0.0502,
+      "step": 284
+    },
+    {
+      "epoch": 0.06477272727272727,
+      "grad_norm": 3.51832542966258,
+      "learning_rate": 1.2371946198131563e-06,
+      "loss": 0.0803,
+      "step": 285
+    },
+    {
+      "epoch": 0.065,
+      "grad_norm": 1.7601967545837591,
+      "learning_rate": 1.2371045942493804e-06,
+      "loss": 0.0946,
+      "step": 286
+    },
+    {
+      "epoch": 0.06522727272727273,
+      "grad_norm": 2.2038956569787316,
+      "learning_rate": 1.2370142566386038e-06,
+      "loss": 0.1052,
+      "step": 287
+    },
+    {
+      "epoch": 0.06545454545454546,
+      "grad_norm": 3.1887431312405856,
+      "learning_rate": 1.2369236070268795e-06,
+      "loss": 0.0991,
+      "step": 288
+    },
+    {
+      "epoch": 0.06568181818181819,
+      "grad_norm": 1.4097208525041756,
+      "learning_rate": 1.2368326454604201e-06,
+      "loss": 0.0828,
+      "step": 289
+    },
+    {
+      "epoch": 0.0659090909090909,
+      "grad_norm": 2.536442931661013,
+      "learning_rate": 1.2367413719855976e-06,
+      "loss": 0.0763,
+      "step": 290
+    },
+    {
+      "epoch": 0.06613636363636363,
+      "grad_norm": 1.7252748106496696,
+      "learning_rate": 1.2366497866489423e-06,
+      "loss": 0.0558,
+      "step": 291
+    },
+    {
+      "epoch": 0.06636363636363636,
+      "grad_norm": 2.3801214982066408,
+      "learning_rate": 1.236557889497144e-06,
+      "loss": 0.0983,
+      "step": 292
+    },
+    {
+      "epoch": 0.06659090909090909,
+      "grad_norm": 2.5914386333247177,
+      "learning_rate": 1.236465680577051e-06,
+      "loss": 0.091,
+      "step": 293
+    },
+    {
+      "epoch": 0.06681818181818182,
+      "grad_norm": 2.1788870392905118,
+      "learning_rate": 1.2363731599356712e-06,
+      "loss": 0.1222,
+      "step": 294
+    },
+    {
+      "epoch": 0.06704545454545455,
+      "grad_norm": 3.3445059388136555,
+      "learning_rate": 1.2362803276201709e-06,
+      "loss": 0.1815,
+      "step": 295
+    },
+    {
+      "epoch": 0.06727272727272728,
+      "grad_norm": 3.6062837096175286,
+      "learning_rate": 1.2361871836778755e-06,
+      "loss": 0.0819,
+      "step": 296
+    },
+    {
+      "epoch": 0.0675,
+      "grad_norm": 3.1353860677297267,
+      "learning_rate": 1.236093728156269e-06,
+      "loss": 0.0996,
+      "step": 297
+    },
+    {
+      "epoch": 0.06772727272727273,
+      "grad_norm": 3.795202115844352,
+      "learning_rate": 1.2359999611029944e-06,
+      "loss": 0.1258,
+      "step": 298
+    },
+    {
+      "epoch": 0.06795454545454545,
+      "grad_norm": 4.4425120996211644,
+      "learning_rate": 1.2359058825658534e-06,
+      "loss": 0.1118,
+      "step": 299
+    },
+    {
+      "epoch": 0.06818181818181818,
+      "grad_norm": 3.060625425498013,
+      "learning_rate": 1.2358114925928073e-06,
+      "loss": 0.085,
+      "step": 300
+    },
+    {
+      "epoch": 0.0684090909090909,
+      "grad_norm": 2.541572581911574,
+      "learning_rate": 1.2357167912319747e-06,
+      "loss": 0.0788,
+      "step": 301
+    },
+    {
+      "epoch": 0.06863636363636363,
+      "grad_norm": 3.611123902709692,
+      "learning_rate": 1.2356217785316344e-06,
+      "loss": 0.0738,
+      "step": 302
+    },
+    {
+      "epoch": 0.06886363636363636,
+      "grad_norm": 2.57826970052294,
+      "learning_rate": 1.235526454540223e-06,
+      "loss": 0.1558,
+      "step": 303
+    },
+    {
+      "epoch": 0.06909090909090909,
+      "grad_norm": 1.405756388923129,
+      "learning_rate": 1.2354308193063358e-06,
+      "loss": 0.0558,
+      "step": 304
+    },
+    {
+      "epoch": 0.06931818181818182,
+      "grad_norm": 4.892966445443843,
+      "learning_rate": 1.2353348728787274e-06,
+      "loss": 0.1939,
+      "step": 305
+    },
+    {
+      "epoch": 0.06954545454545455,
+      "grad_norm": 3.405266279984527,
+      "learning_rate": 1.2352386153063107e-06,
+      "loss": 0.0782,
+      "step": 306
+    },
+    {
+      "epoch": 0.06977272727272728,
+      "grad_norm": 4.65382172183155,
+      "learning_rate": 1.2351420466381566e-06,
+      "loss": 0.1097,
+      "step": 307
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6483561501934108,
+      "learning_rate": 1.235045166923496e-06,
+      "loss": 0.0952,
+      "step": 308
+    },
+    {
+      "epoch": 0.07022727272727272,
+      "grad_norm": 2.492520472658006,
+      "learning_rate": 1.2349479762117171e-06,
+      "loss": 0.1081,
+      "step": 309
+    },
+    {
+      "epoch": 0.07045454545454545,
+      "grad_norm": 4.661738484788085,
+      "learning_rate": 1.2348504745523673e-06,
+      "loss": 0.2239,
+      "step": 310
+    },
+    {
+      "epoch": 0.07068181818181818,
+      "grad_norm": 2.836454965603772,
+      "learning_rate": 1.2347526619951523e-06,
+      "loss": 0.1213,
+      "step": 311
+    },
+    {
+      "epoch": 0.07090909090909091,
+      "grad_norm": 4.58311528505282,
+      "learning_rate": 1.2346545385899358e-06,
+      "loss": 0.1464,
+      "step": 312
+    },
+    {
+      "epoch": 0.07113636363636364,
+      "grad_norm": 3.5048453466304923,
+      "learning_rate": 1.2345561043867413e-06,
+      "loss": 0.1603,
+      "step": 313
+    },
+    {
+      "epoch": 0.07136363636363637,
+      "grad_norm": 3.561473338236458,
+      "learning_rate": 1.2344573594357493e-06,
+      "loss": 0.0841,
+      "step": 314
+    },
+    {
+      "epoch": 0.0715909090909091,
+      "grad_norm": 7.099554539977735,
+      "learning_rate": 1.2343583037872998e-06,
+      "loss": 0.1376,
+      "step": 315
+    },
+    {
+      "epoch": 0.07181818181818182,
+      "grad_norm": 2.0454245684933086,
+      "learning_rate": 1.2342589374918905e-06,
+      "loss": 0.1454,
+      "step": 316
+    },
+    {
+      "epoch": 0.07204545454545455,
+      "grad_norm": 2.690682928495695,
+      "learning_rate": 1.2341592606001777e-06,
+      "loss": 0.151,
+      "step": 317
+    },
+    {
+      "epoch": 0.07227272727272727,
+      "grad_norm": 1.9473548981882711,
+      "learning_rate": 1.2340592731629758e-06,
+      "loss": 0.0663,
+      "step": 318
+    },
+    {
+      "epoch": 0.0725,
+      "grad_norm": 1.9817106310961383,
+      "learning_rate": 1.2339589752312581e-06,
+      "loss": 0.0755,
+      "step": 319
+    },
+    {
+      "epoch": 0.07272727272727272,
+      "grad_norm": 3.5702400423961294,
+      "learning_rate": 1.233858366856156e-06,
+      "loss": 0.1301,
+      "step": 320
+    },
+    {
+      "epoch": 0.07295454545454545,
+      "grad_norm": 2.8060422993346554,
+      "learning_rate": 1.2337574480889585e-06,
+      "loss": 0.1066,
+      "step": 321
+    },
+    {
+      "epoch": 0.07318181818181818,
+      "grad_norm": 3.309057271529764,
+      "learning_rate": 1.2336562189811138e-06,
+      "loss": 0.0561,
+      "step": 322
+    },
+    {
+      "epoch": 0.07340909090909091,
+      "grad_norm": 4.600936564276701,
+      "learning_rate": 1.2335546795842276e-06,
+      "loss": 0.1313,
+      "step": 323
+    },
+    {
+      "epoch": 0.07363636363636364,
+      "grad_norm": 5.1141235062756305,
+      "learning_rate": 1.233452829950064e-06,
+      "loss": 0.1269,
+      "step": 324
+    },
+    {
+      "epoch": 0.07386363636363637,
+      "grad_norm": 2.677110953995895,
+      "learning_rate": 1.2333506701305453e-06,
+      "loss": 0.0729,
+      "step": 325
+    },
+    {
+      "epoch": 0.0740909090909091,
+      "grad_norm": 2.7724924929723316,
+      "learning_rate": 1.2332482001777522e-06,
+      "loss": 0.1107,
+      "step": 326
+    },
+    {
+      "epoch": 0.07431818181818182,
+      "grad_norm": 5.232846583422687,
+      "learning_rate": 1.233145420143923e-06,
+      "loss": 0.1248,
+      "step": 327
+    },
+    {
+      "epoch": 0.07454545454545454,
+      "grad_norm": 2.7835450341603054,
+      "learning_rate": 1.2330423300814542e-06,
+      "loss": 0.118,
+      "step": 328
+    },
+    {
+      "epoch": 0.07477272727272727,
+      "grad_norm": 3.274121466447511,
+      "learning_rate": 1.2329389300429008e-06,
+      "loss": 0.1509,
+      "step": 329
+    },
+    {
+      "epoch": 0.075,
+      "grad_norm": 2.8890066579436304,
+      "learning_rate": 1.232835220080975e-06,
+      "loss": 0.1061,
+      "step": 330
+    },
+    {
+      "epoch": 0.07522727272727273,
+      "grad_norm": 1.3461200298680585,
+      "learning_rate": 1.232731200248548e-06,
+      "loss": 0.0633,
+      "step": 331
+    },
+    {
+      "epoch": 0.07545454545454545,
+      "grad_norm": 5.7546563462800595,
+      "learning_rate": 1.232626870598648e-06,
+      "loss": 0.1391,
+      "step": 332
+    },
+    {
+      "epoch": 0.07568181818181818,
+      "grad_norm": 3.0275877351098917,
+      "learning_rate": 1.2325222311844617e-06,
+      "loss": 0.1363,
+      "step": 333
+    },
+    {
+      "epoch": 0.07590909090909091,
+      "grad_norm": 4.366747450337461,
+      "learning_rate": 1.2324172820593339e-06,
+      "loss": 0.1209,
+      "step": 334
+    },
+    {
+      "epoch": 0.07613636363636364,
+      "grad_norm": 3.7087005681939518,
+      "learning_rate": 1.2323120232767667e-06,
+      "loss": 0.1229,
+      "step": 335
+    },
+    {
+      "epoch": 0.07636363636363637,
+      "grad_norm": 3.775294884503768,
+      "learning_rate": 1.2322064548904202e-06,
+      "loss": 0.1068,
+      "step": 336
+    },
+    {
+      "epoch": 0.07659090909090908,
+      "grad_norm": 3.7584262607116914,
+      "learning_rate": 1.232100576954113e-06,
+      "loss": 0.0952,
+      "step": 337
+    },
+    {
+      "epoch": 0.07681818181818181,
+      "grad_norm": 1.7891425379053958,
+      "learning_rate": 1.2319943895218205e-06,
+      "loss": 0.0993,
+      "step": 338
+    },
+    {
+      "epoch": 0.07704545454545454,
+      "grad_norm": 1.827195549561059,
+      "learning_rate": 1.2318878926476765e-06,
+      "loss": 0.1345,
+      "step": 339
+    },
+    {
+      "epoch": 0.07727272727272727,
+      "grad_norm": 2.0304662719547046,
+      "learning_rate": 1.2317810863859728e-06,
+      "loss": 0.0881,
+      "step": 340
+    },
+    {
+      "epoch": 0.0775,
+      "grad_norm": 2.603856208796641,
+      "learning_rate": 1.231673970791158e-06,
+      "loss": 0.0989,
+      "step": 341
+    },
+    {
+      "epoch": 0.07772727272727273,
+      "grad_norm": 2.753960531619496,
+      "learning_rate": 1.231566545917839e-06,
+      "loss": 0.1393,
+      "step": 342
+    },
+    {
+      "epoch": 0.07795454545454546,
+      "grad_norm": 3.670105337059318,
+      "learning_rate": 1.2314588118207808e-06,
+      "loss": 0.1411,
+      "step": 343
+    },
+    {
+      "epoch": 0.07818181818181819,
+      "grad_norm": 3.0453426153803855,
+      "learning_rate": 1.2313507685549054e-06,
+      "loss": 0.0669,
+      "step": 344
+    },
+    {
+      "epoch": 0.07840909090909091,
+      "grad_norm": 3.64975550268497,
+      "learning_rate": 1.231242416175292e-06,
+      "loss": 0.0967,
+      "step": 345
+    },
+    {
+      "epoch": 0.07863636363636364,
+      "grad_norm": 3.7460249859795924,
+      "learning_rate": 1.2311337547371785e-06,
+      "loss": 0.1316,
+      "step": 346
+    },
+    {
+      "epoch": 0.07886363636363636,
+      "grad_norm": 2.5110205934468532,
+      "learning_rate": 1.2310247842959597e-06,
+      "loss": 0.1347,
+      "step": 347
+    },
+    {
+      "epoch": 0.07909090909090909,
+      "grad_norm": 2.578783845598259,
+      "learning_rate": 1.230915504907188e-06,
+      "loss": 0.0859,
+      "step": 348
+    },
+    {
+      "epoch": 0.07931818181818182,
+      "grad_norm": 2.904523661612725,
+      "learning_rate": 1.2308059166265734e-06,
+      "loss": 0.1034,
+      "step": 349
+    },
+    {
+      "epoch": 0.07954545454545454,
+      "grad_norm": 3.8463393430603903,
+      "learning_rate": 1.2306960195099833e-06,
+      "loss": 0.0773,
+      "step": 350
+    },
+    {
+      "epoch": 0.07977272727272727,
+      "grad_norm": 3.4874678834509516,
+      "learning_rate": 1.2305858136134422e-06,
+      "loss": 0.1871,
+      "step": 351
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.3233687777324112,
+      "learning_rate": 1.2304752989931327e-06,
+      "loss": 0.1294,
+      "step": 352
+    },
+    {
+      "epoch": 0.08022727272727273,
+      "grad_norm": 2.5143806178353683,
+      "learning_rate": 1.2303644757053945e-06,
+      "loss": 0.1002,
+      "step": 353
+    },
+    {
+      "epoch": 0.08045454545454546,
+      "grad_norm": 2.5100887020522977,
+      "learning_rate": 1.2302533438067247e-06,
+      "loss": 0.1097,
+      "step": 354
+    },
+    {
+      "epoch": 0.08068181818181819,
+      "grad_norm": 2.6722235964082803,
+      "learning_rate": 1.230141903353777e-06,
+      "loss": 0.0936,
+      "step": 355
+    },
+    {
+      "epoch": 0.0809090909090909,
+      "grad_norm": 1.7718009120749822,
+      "learning_rate": 1.2300301544033636e-06,
+      "loss": 0.0527,
+      "step": 356
+    },
+    {
+      "epoch": 0.08113636363636363,
+      "grad_norm": 1.915075876701033,
+      "learning_rate": 1.2299180970124533e-06,
+      "loss": 0.0978,
+      "step": 357
+    },
+    {
+      "epoch": 0.08136363636363636,
+      "grad_norm": 2.1280843016236988,
+      "learning_rate": 1.2298057312381723e-06,
+      "loss": 0.0686,
+      "step": 358
+    },
+    {
+      "epoch": 0.08159090909090909,
+      "grad_norm": 4.459716743720089,
+      "learning_rate": 1.2296930571378035e-06,
+      "loss": 0.1183,
+      "step": 359
+    },
+    {
+      "epoch": 0.08181818181818182,
+      "grad_norm": 2.9168537440591225,
+      "learning_rate": 1.229580074768788e-06,
+      "loss": 0.1402,
+      "step": 360
+    },
+    {
+      "epoch": 0.08204545454545455,
+      "grad_norm": 2.38220639956826,
+      "learning_rate": 1.2294667841887234e-06,
+      "loss": 0.0956,
+      "step": 361
+    },
+    {
+      "epoch": 0.08227272727272728,
+      "grad_norm": 3.3941245990914704,
+      "learning_rate": 1.2293531854553642e-06,
+      "loss": 0.069,
+      "step": 362
+    },
+    {
+      "epoch": 0.0825,
+      "grad_norm": 2.6982014435057007,
+      "learning_rate": 1.2292392786266225e-06,
+      "loss": 0.1295,
+      "step": 363
+    },
+    {
+      "epoch": 0.08272727272727273,
+      "grad_norm": 2.5429887774673605,
+      "learning_rate": 1.2291250637605672e-06,
+      "loss": 0.0649,
+      "step": 364
+    },
+    {
+      "epoch": 0.08295454545454546,
+      "grad_norm": 4.541719412333879,
+      "learning_rate": 1.2290105409154244e-06,
+      "loss": 0.076,
+      "step": 365
+    },
+    {
+      "epoch": 0.08318181818181818,
+      "grad_norm": 3.46953041168632,
+      "learning_rate": 1.2288957101495772e-06,
+      "loss": 0.0955,
+      "step": 366
+    },
+    {
+      "epoch": 0.0834090909090909,
+      "grad_norm": 2.833747952271272,
+      "learning_rate": 1.2287805715215651e-06,
+      "loss": 0.0979,
+      "step": 367
+    },
+    {
+      "epoch": 0.08363636363636363,
+      "grad_norm": 2.4688973245010732,
+      "learning_rate": 1.2286651250900858e-06,
+      "loss": 0.0702,
+      "step": 368
+    },
+    {
+      "epoch": 0.08386363636363636,
+      "grad_norm": 3.6174807701051046,
+      "learning_rate": 1.2285493709139925e-06,
+      "loss": 0.1192,
+      "step": 369
+    },
+    {
+      "epoch": 0.08409090909090909,
+      "grad_norm": 3.1746195503233454,
+      "learning_rate": 1.2284333090522962e-06,
+      "loss": 0.0805,
+      "step": 370
+    },
+    {
+      "epoch": 0.08431818181818182,
+      "grad_norm": 3.6365274354450974,
+      "learning_rate": 1.2283169395641647e-06,
+      "loss": 0.0965,
+      "step": 371
+    },
+    {
+      "epoch": 0.08454545454545455,
+      "grad_norm": 3.1701725473834976,
+      "learning_rate": 1.228200262508922e-06,
+      "loss": 0.1219,
+      "step": 372
+    },
+    {
+      "epoch": 0.08477272727272728,
+      "grad_norm": 6.611704569027101,
+      "learning_rate": 1.2280832779460494e-06,
+      "loss": 0.1599,
+      "step": 373
+    },
+    {
+      "epoch": 0.085,
+      "grad_norm": 2.3590785477124863,
+      "learning_rate": 1.2279659859351853e-06,
+      "loss": 0.1025,
+      "step": 374
+    },
+    {
+      "epoch": 0.08522727272727272,
+      "grad_norm": 4.919619134503741,
+      "learning_rate": 1.2278483865361239e-06,
+      "loss": 0.1255,
+      "step": 375
+    },
+    {
+      "epoch": 0.08545454545454545,
+      "grad_norm": 1.6390680275869847,
+      "learning_rate": 1.227730479808817e-06,
+      "loss": 0.037,
+      "step": 376
+    },
+    {
+      "epoch": 0.08568181818181818,
+      "grad_norm": 3.7054155554055317,
+      "learning_rate": 1.2276122658133723e-06,
+      "loss": 0.0795,
+      "step": 377
+    },
+    {
+      "epoch": 0.08590909090909091,
+      "grad_norm": 3.3710877295891146,
+      "learning_rate": 1.2274937446100548e-06,
+      "loss": 0.1017,
+      "step": 378
+    },
+    {
+      "epoch": 0.08613636363636364,
+      "grad_norm": 4.049749549786513,
+      "learning_rate": 1.227374916259286e-06,
+      "loss": 0.0977,
+      "step": 379
+    },
+    {
+      "epoch": 0.08636363636363636,
+      "grad_norm": 2.8237932036299034,
+      "learning_rate": 1.2272557808216433e-06,
+      "loss": 0.082,
+      "step": 380
+    },
+    {
+      "epoch": 0.0865909090909091,
+      "grad_norm": 6.472982642514373,
+      "learning_rate": 1.2271363383578619e-06,
+      "loss": 0.1111,
+      "step": 381
+    },
+    {
+      "epoch": 0.08681818181818182,
+      "grad_norm": 2.872276084205482,
+      "learning_rate": 1.2270165889288325e-06,
+      "loss": 0.1416,
+      "step": 382
+    },
+    {
+      "epoch": 0.08704545454545455,
+      "grad_norm": 3.1121878136277763,
+      "learning_rate": 1.2268965325956022e-06,
+      "loss": 0.0709,
+      "step": 383
+    },
+    {
+      "epoch": 0.08727272727272728,
+      "grad_norm": 4.289430038043764,
+      "learning_rate": 1.2267761694193752e-06,
+      "loss": 0.1046,
+      "step": 384
+    },
+    {
+      "epoch": 0.0875,
+      "grad_norm": 3.7226223021658016,
+      "learning_rate": 1.2266554994615121e-06,
+      "loss": 0.0881,
+      "step": 385
+    },
+    {
+      "epoch": 0.08772727272727272,
+      "grad_norm": 2.432238318963267,
+      "learning_rate": 1.2265345227835295e-06,
+      "loss": 0.1079,
+      "step": 386
+    },
+    {
+      "epoch": 0.08795454545454545,
+      "grad_norm": 1.7074767331576395,
+      "learning_rate": 1.2264132394471007e-06,
+      "loss": 0.0657,
+      "step": 387
+    },
+    {
+      "epoch": 0.08818181818181818,
+      "grad_norm": 1.9745542140283514,
+      "learning_rate": 1.226291649514055e-06,
+      "loss": 0.0548,
+      "step": 388
+    },
+    {
+      "epoch": 0.08840909090909091,
+      "grad_norm": 3.486509875354997,
+      "learning_rate": 1.226169753046378e-06,
+      "loss": 0.1422,
+      "step": 389
+    },
+    {
+      "epoch": 0.08863636363636364,
+      "grad_norm": 2.105720275426089,
+      "learning_rate": 1.2260475501062121e-06,
+      "loss": 0.0753,
+      "step": 390
+    },
+    {
+      "epoch": 0.08886363636363637,
+      "grad_norm": 1.6908207330851532,
+      "learning_rate": 1.2259250407558553e-06,
+      "loss": 0.0423,
+      "step": 391
+    },
+    {
+      "epoch": 0.0890909090909091,
+      "grad_norm": 1.9536413900829848,
+      "learning_rate": 1.2258022250577622e-06,
+      "loss": 0.0651,
+      "step": 392
+    },
+    {
+      "epoch": 0.08931818181818182,
+      "grad_norm": 2.107555535166247,
+      "learning_rate": 1.2256791030745434e-06,
+      "loss": 0.0767,
+      "step": 393
+    },
+    {
+      "epoch": 0.08954545454545454,
+      "grad_norm": 3.2413360732050918,
+      "learning_rate": 1.225555674868966e-06,
+      "loss": 0.1408,
+      "step": 394
+    },
+    {
+      "epoch": 0.08977272727272727,
+      "grad_norm": 5.098748129148887,
+      "learning_rate": 1.2254319405039524e-06,
+      "loss": 0.148,
+      "step": 395
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.9137716556988416,
+      "learning_rate": 1.2253079000425818e-06,
+      "loss": 0.0858,
+      "step": 396
+    },
+    {
+      "epoch": 0.09022727272727273,
+      "grad_norm": 2.306099527359283,
+      "learning_rate": 1.2251835535480895e-06,
+      "loss": 0.1314,
+      "step": 397
+    },
+    {
+      "epoch": 0.09045454545454545,
+      "grad_norm": 3.233107573064104,
+      "learning_rate": 1.2250589010838662e-06,
+      "loss": 0.1313,
+      "step": 398
+    },
+    {
+      "epoch": 0.09068181818181818,
+      "grad_norm": 3.6199998689552584,
+      "learning_rate": 1.224933942713459e-06,
+      "loss": 0.1645,
+      "step": 399
+    },
+    {
+      "epoch": 0.09090909090909091,
+      "grad_norm": 2.8815298806137113,
+      "learning_rate": 1.2248086785005709e-06,
+      "loss": 0.0814,
+      "step": 400
+    },
+    {
+      "epoch": 0.09113636363636364,
+      "grad_norm": 2.12725818411472,
+      "learning_rate": 1.2246831085090611e-06,
+      "loss": 0.0677,
+      "step": 401
+    },
+    {
+      "epoch": 0.09136363636363637,
+      "grad_norm": 2.068540185403787,
+      "learning_rate": 1.2245572328029438e-06,
+      "loss": 0.0766,
+      "step": 402
+    },
+    {
+      "epoch": 0.0915909090909091,
+      "grad_norm": 3.3787139639022477,
+      "learning_rate": 1.22443105144639e-06,
+      "loss": 0.1528,
+      "step": 403
+    },
+    {
+      "epoch": 0.09181818181818181,
+      "grad_norm": 3.6734313614551706,
+      "learning_rate": 1.224304564503726e-06,
+      "loss": 0.2006,
+      "step": 404
+    },
+    {
+      "epoch": 0.09204545454545454,
+      "grad_norm": 2.3773106183493704,
+      "learning_rate": 1.224177772039434e-06,
+      "loss": 0.0819,
+      "step": 405
+    },
+    {
+      "epoch": 0.09227272727272727,
+      "grad_norm": 4.479594338984166,
+      "learning_rate": 1.2240506741181523e-06,
+      "loss": 0.1965,
+      "step": 406
+    },
+    {
+      "epoch": 0.0925,
+      "grad_norm": 4.599512341378346,
+      "learning_rate": 1.2239232708046745e-06,
+      "loss": 0.117,
+      "step": 407
+    },
+    {
+      "epoch": 0.09272727272727273,
+      "grad_norm": 4.65201743851853,
+      "learning_rate": 1.2237955621639496e-06,
+      "loss": 0.2369,
+      "step": 408
+    },
+    {
+      "epoch": 0.09295454545454546,
+      "grad_norm": 2.7824021376543677,
+      "learning_rate": 1.223667548261083e-06,
+      "loss": 0.0941,
+      "step": 409
+    },
+    {
+      "epoch": 0.09318181818181819,
+      "grad_norm": 4.289778277836561,
+      "learning_rate": 1.2235392291613353e-06,
+      "loss": 0.1326,
+      "step": 410
+    },
+    {
+      "epoch": 0.09340909090909091,
+      "grad_norm": 3.9360665164441055,
+      "learning_rate": 1.2234106049301228e-06,
+      "loss": 0.1095,
+      "step": 411
+    },
+    {
+      "epoch": 0.09363636363636364,
+      "grad_norm": 4.915602973483901,
+      "learning_rate": 1.2232816756330173e-06,
+      "loss": 0.1522,
+      "step": 412
+    },
+    {
+      "epoch": 0.09386363636363636,
+      "grad_norm": 3.440478039172773,
+      "learning_rate": 1.223152441335746e-06,
+      "loss": 0.1086,
+      "step": 413
+    },
+    {
+      "epoch": 0.09409090909090909,
+      "grad_norm": 3.163613020768659,
+      "learning_rate": 1.223022902104192e-06,
+      "loss": 0.1091,
+      "step": 414
+    },
+    {
+      "epoch": 0.09431818181818181,
+      "grad_norm": 2.7771510795806793,
+      "learning_rate": 1.2228930580043931e-06,
+      "loss": 0.0769,
+      "step": 415
+    },
+    {
+      "epoch": 0.09454545454545454,
+      "grad_norm": 4.646937062796584,
+      "learning_rate": 1.2227629091025437e-06,
+      "loss": 0.1422,
+      "step": 416
+    },
+    {
+      "epoch": 0.09477272727272727,
+      "grad_norm": 4.410836996121542,
+      "learning_rate": 1.2226324554649921e-06,
+      "loss": 0.1196,
+      "step": 417
+    },
+    {
+      "epoch": 0.095,
+      "grad_norm": 2.5231288313876727,
+      "learning_rate": 1.222501697158243e-06,
+      "loss": 0.0799,
+      "step": 418
+    },
+    {
+      "epoch": 0.09522727272727273,
+      "grad_norm": 3.360112046907082,
+      "learning_rate": 1.2223706342489565e-06,
+      "loss": 0.0994,
+      "step": 419
+    },
+    {
+      "epoch": 0.09545454545454546,
+      "grad_norm": 4.946888659745919,
+      "learning_rate": 1.222239266803947e-06,
+      "loss": 0.1911,
+      "step": 420
+    },
+    {
+      "epoch": 0.09568181818181819,
+      "grad_norm": 1.4239021451720755,
+      "learning_rate": 1.2221075948901856e-06,
+      "loss": 0.0615,
+      "step": 421
+    },
+    {
+      "epoch": 0.0959090909090909,
+      "grad_norm": 2.7296283333268834,
+      "learning_rate": 1.221975618574797e-06,
+      "loss": 0.1013,
+      "step": 422
+    },
+    {
+      "epoch": 0.09613636363636363,
+      "grad_norm": 4.974673641627054,
+      "learning_rate": 1.2218433379250623e-06,
+      "loss": 0.129,
+      "step": 423
+    },
+    {
+      "epoch": 0.09636363636363636,
+      "grad_norm": 2.6052782154944776,
+      "learning_rate": 1.2217107530084174e-06,
+      "loss": 0.1174,
+      "step": 424
+    },
+    {
+      "epoch": 0.09659090909090909,
+      "grad_norm": 2.4871443771102544,
+      "learning_rate": 1.2215778638924527e-06,
+      "loss": 0.0799,
+      "step": 425
+    },
+    {
+      "epoch": 0.09681818181818182,
+      "grad_norm": 3.86113905482254,
+      "learning_rate": 1.221444670644915e-06,
+      "loss": 0.0932,
+      "step": 426
+    },
+    {
+      "epoch": 0.09704545454545455,
+      "grad_norm": 4.279554503418284,
+      "learning_rate": 1.221311173333705e-06,
+      "loss": 0.101,
+      "step": 427
+    },
+    {
+      "epoch": 0.09727272727272727,
+      "grad_norm": 1.7158239249526541,
+      "learning_rate": 1.2211773720268784e-06,
+      "loss": 0.0557,
+      "step": 428
+    },
+    {
+      "epoch": 0.0975,
+      "grad_norm": 1.9719435011250115,
+      "learning_rate": 1.2210432667926467e-06,
+      "loss": 0.0636,
+      "step": 429
+    },
+    {
+      "epoch": 0.09772727272727273,
+      "grad_norm": 2.800044602484537,
+      "learning_rate": 1.2209088576993757e-06,
+      "loss": 0.1438,
+      "step": 430
+    },
+    {
+      "epoch": 0.09795454545454546,
+      "grad_norm": 2.2459968059258495,
+      "learning_rate": 1.2207741448155867e-06,
+      "loss": 0.0585,
+      "step": 431
+    },
+    {
+      "epoch": 0.09818181818181818,
+      "grad_norm": 2.737923788075119,
+      "learning_rate": 1.220639128209955e-06,
+      "loss": 0.1149,
+      "step": 432
+    },
+    {
+      "epoch": 0.0984090909090909,
+      "grad_norm": 2.160504662100839,
+      "learning_rate": 1.2205038079513113e-06,
+      "loss": 0.0957,
+      "step": 433
+    },
+    {
+      "epoch": 0.09863636363636363,
+      "grad_norm": 2.3449320078814058,
+      "learning_rate": 1.2203681841086409e-06,
+      "loss": 0.0677,
+      "step": 434
+    },
+    {
+      "epoch": 0.09886363636363636,
+      "grad_norm": 1.7561253489930528,
+      "learning_rate": 1.2202322567510843e-06,
+      "loss": 0.0613,
+      "step": 435
+    },
+    {
+      "epoch": 0.09909090909090909,
+      "grad_norm": 4.3714833019328845,
+      "learning_rate": 1.2200960259479362e-06,
+      "loss": 0.1448,
+      "step": 436
+    },
+    {
+      "epoch": 0.09931818181818182,
+      "grad_norm": 5.241134950935839,
+      "learning_rate": 1.219959491768646e-06,
+      "loss": 0.1817,
+      "step": 437
+    },
+    {
+      "epoch": 0.09954545454545455,
+      "grad_norm": 4.511533086825861,
+      "learning_rate": 1.2198226542828183e-06,
+      "loss": 0.1153,
+      "step": 438
+    },
+    {
+      "epoch": 0.09977272727272728,
+      "grad_norm": 1.9872464735723243,
+      "learning_rate": 1.219685513560212e-06,
+      "loss": 0.067,
+      "step": 439
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.966477682233784,
+      "learning_rate": 1.2195480696707401e-06,
+      "loss": 0.0565,
+      "step": 440
+    },
+    {
+      "epoch": 0.10022727272727272,
+      "grad_norm": 4.920514774069939,
+      "learning_rate": 1.2194103226844711e-06,
+      "loss": 0.2292,
+      "step": 441
+    },
+    {
+      "epoch": 0.10045454545454545,
+      "grad_norm": 3.3871390478027137,
+      "learning_rate": 1.2192722726716272e-06,
+      "loss": 0.1924,
+      "step": 442
+    },
+    {
+      "epoch": 0.10068181818181818,
+      "grad_norm": 2.0351508051987826,
+      "learning_rate": 1.2191339197025857e-06,
+      "loss": 0.0862,
+      "step": 443
+    },
+    {
+      "epoch": 0.1009090909090909,
+      "grad_norm": 2.88571125465246,
+      "learning_rate": 1.2189952638478778e-06,
+      "loss": 0.1076,
+      "step": 444
+    },
+    {
+      "epoch": 0.10113636363636364,
+      "grad_norm": 3.9220428041413,
+      "learning_rate": 1.2188563051781894e-06,
+      "loss": 0.0921,
+      "step": 445
+    },
+    {
+      "epoch": 0.10136363636363636,
+      "grad_norm": 2.308630021933964,
+      "learning_rate": 1.2187170437643608e-06,
+      "loss": 0.0786,
+      "step": 446
+    },
+    {
+      "epoch": 0.10159090909090909,
+      "grad_norm": 1.7021008056728288,
+      "learning_rate": 1.2185774796773864e-06,
+      "loss": 0.0489,
+      "step": 447
+    },
+    {
+      "epoch": 0.10181818181818182,
+      "grad_norm": 2.8366182130249156,
+      "learning_rate": 1.2184376129884156e-06,
+      "loss": 0.0711,
+      "step": 448
+    },
+    {
+      "epoch": 0.10204545454545455,
+      "grad_norm": 2.494654489597956,
+      "learning_rate": 1.2182974437687512e-06,
+      "loss": 0.077,
+      "step": 449
+    },
+    {
+      "epoch": 0.10227272727272728,
+      "grad_norm": 1.8789971456688084,
+      "learning_rate": 1.2181569720898503e-06,
+      "loss": 0.0603,
+      "step": 450
+    },
+    {
+      "epoch": 0.1025,
+      "grad_norm": 7.572598925909453,
+      "learning_rate": 1.2180161980233249e-06,
+      "loss": 0.2909,
+      "step": 451
+    },
+    {
+      "epoch": 0.10272727272727272,
+      "grad_norm": 3.251456224316626,
+      "learning_rate": 1.2178751216409404e-06,
+      "loss": 0.1156,
+      "step": 452
+    },
+    {
+      "epoch": 0.10295454545454545,
+      "grad_norm": 2.8783591202595775,
+      "learning_rate": 1.217733743014617e-06,
+      "loss": 0.0812,
+      "step": 453
+    },
+    {
+      "epoch": 0.10318181818181818,
+      "grad_norm": 1.779008544511726,
+      "learning_rate": 1.2175920622164284e-06,
+      "loss": 0.0822,
+      "step": 454
+    },
+    {
+      "epoch": 0.10340909090909091,
+      "grad_norm": 3.151791231016802,
+      "learning_rate": 1.2174500793186024e-06,
+      "loss": 0.1666,
+      "step": 455
+    },
+    {
+      "epoch": 0.10363636363636364,
+      "grad_norm": 6.270849489622784,
+      "learning_rate": 1.2173077943935212e-06,
+      "loss": 0.2308,
+      "step": 456
+    },
+    {
+      "epoch": 0.10386363636363637,
+      "grad_norm": 3.9063075477089595,
+      "learning_rate": 1.2171652075137209e-06,
+      "loss": 0.0935,
+      "step": 457
+    },
+    {
+      "epoch": 0.1040909090909091,
+      "grad_norm": 3.1178755438985934,
+      "learning_rate": 1.2170223187518908e-06,
+      "loss": 0.1215,
+      "step": 458
+    },
+    {
+      "epoch": 0.10431818181818182,
+      "grad_norm": 3.108509861096335,
+      "learning_rate": 1.216879128180875e-06,
+      "loss": 0.0726,
+      "step": 459
+    },
+    {
+      "epoch": 0.10454545454545454,
+      "grad_norm": 2.471974295276308,
+      "learning_rate": 1.2167356358736714e-06,
+      "loss": 0.0618,
+      "step": 460
+    },
+    {
+      "epoch": 0.10477272727272727,
+      "grad_norm": 2.0732406776270715,
+      "learning_rate": 1.2165918419034312e-06,
+      "loss": 0.0728,
+      "step": 461
+    },
+    {
+      "epoch": 0.105,
+      "grad_norm": 3.043747201893837,
+      "learning_rate": 1.2164477463434599e-06,
+      "loss": 0.1112,
+      "step": 462
+    },
+    {
+      "epoch": 0.10522727272727272,
+      "grad_norm": 3.2055267418794124,
+      "learning_rate": 1.216303349267216e-06,
+      "loss": 0.0791,
+      "step": 463
+    },
+    {
+      "epoch": 0.10545454545454545,
+      "grad_norm": 3.383368019795033,
+      "learning_rate": 1.2161586507483126e-06,
+      "loss": 0.1577,
+      "step": 464
+    },
+    {
+      "epoch": 0.10568181818181818,
+      "grad_norm": 3.9789816443070873,
+      "learning_rate": 1.2160136508605156e-06,
+      "loss": 0.0989,
+      "step": 465
+    },
+    {
+      "epoch": 0.10590909090909091,
+      "grad_norm": 2.500431701045704,
+      "learning_rate": 1.2158683496777457e-06,
+      "loss": 0.0839,
+      "step": 466
+    },
+    {
+      "epoch": 0.10613636363636364,
+      "grad_norm": 2.657039985296244,
+      "learning_rate": 1.215722747274076e-06,
+      "loss": 0.0796,
+      "step": 467
+    },
+    {
+      "epoch": 0.10636363636363637,
+      "grad_norm": 2.4937388081922007,
+      "learning_rate": 1.2155768437237342e-06,
+      "loss": 0.1086,
+      "step": 468
+    },
+    {
+      "epoch": 0.1065909090909091,
+      "grad_norm": 2.5615750491399636,
+      "learning_rate": 1.2154306391011003e-06,
+      "loss": 0.1228,
+      "step": 469
+    },
+    {
+      "epoch": 0.10681818181818181,
+      "grad_norm": 2.1009193972019267,
+      "learning_rate": 1.215284133480709e-06,
+      "loss": 0.0656,
+      "step": 470
+    },
+    {
+      "epoch": 0.10704545454545454,
+      "grad_norm": 4.025654861290522,
+      "learning_rate": 1.2151373269372476e-06,
+      "loss": 0.1972,
+      "step": 471
+    },
+    {
+      "epoch": 0.10727272727272727,
+      "grad_norm": 1.7091636419981506,
+      "learning_rate": 1.2149902195455574e-06,
+      "loss": 0.0759,
+      "step": 472
+    },
+    {
+      "epoch": 0.1075,
+      "grad_norm": 2.4698413978210603,
+      "learning_rate": 1.2148428113806326e-06,
+      "loss": 0.1257,
+      "step": 473
+    },
+    {
+      "epoch": 0.10772727272727273,
+      "grad_norm": 2.882577018820717,
+      "learning_rate": 1.214695102517621e-06,
+      "loss": 0.1052,
+      "step": 474
+    },
+    {
+      "epoch": 0.10795454545454546,
+      "grad_norm": 5.414616020041416,
+      "learning_rate": 1.214547093031824e-06,
+      "loss": 0.1681,
+      "step": 475
+    },
+    {
+      "epoch": 0.10818181818181818,
+      "grad_norm": 3.2634120664356967,
+      "learning_rate": 1.2143987829986953e-06,
+      "loss": 0.1293,
+      "step": 476
+    },
+    {
+      "epoch": 0.10840909090909091,
+      "grad_norm": 2.2416444836113842,
+      "learning_rate": 1.2142501724938425e-06,
+      "loss": 0.1309,
+      "step": 477
+    },
+    {
+      "epoch": 0.10863636363636364,
+      "grad_norm": 1.466887546469956,
+      "learning_rate": 1.2141012615930266e-06,
+      "loss": 0.0467,
+      "step": 478
+    },
+    {
+      "epoch": 0.10886363636363636,
+      "grad_norm": 3.2563521071124,
+      "learning_rate": 1.2139520503721614e-06,
+      "loss": 0.2101,
+      "step": 479
+    },
+    {
+      "epoch": 0.10909090909090909,
+      "grad_norm": 3.873460990896338,
+      "learning_rate": 1.2138025389073133e-06,
+      "loss": 0.1078,
+      "step": 480
+    },
+    {
+      "epoch": 0.10931818181818181,
+      "grad_norm": 3.8669089729214634,
+      "learning_rate": 1.2136527272747032e-06,
+      "loss": 0.0754,
+      "step": 481
+    },
+    {
+      "epoch": 0.10954545454545454,
+      "grad_norm": 6.2375879337062035,
+      "learning_rate": 1.2135026155507033e-06,
+      "loss": 0.1833,
+      "step": 482
+    },
+    {
+      "epoch": 0.10977272727272727,
+      "grad_norm": 3.3459487445038585,
+      "learning_rate": 1.2133522038118398e-06,
+      "loss": 0.0923,
+      "step": 483
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.466574087179955,
+      "learning_rate": 1.2132014921347917e-06,
+      "loss": 0.1096,
+      "step": 484
+    },
+    {
+      "epoch": 0.11022727272727273,
+      "grad_norm": 2.167393131020655,
+      "learning_rate": 1.213050480596391e-06,
+      "loss": 0.1058,
+      "step": 485
+    },
+    {
+      "epoch": 0.11045454545454546,
+      "grad_norm": 2.8914726679452816,
+      "learning_rate": 1.2128991692736223e-06,
+      "loss": 0.0797,
+      "step": 486
+    },
+    {
+      "epoch": 0.11068181818181819,
+      "grad_norm": 2.1755687565891066,
+      "learning_rate": 1.2127475582436232e-06,
+      "loss": 0.079,
+      "step": 487
+    },
+    {
+      "epoch": 0.11090909090909092,
+      "grad_norm": 3.492802551825463,
+      "learning_rate": 1.2125956475836837e-06,
+      "loss": 0.138,
+      "step": 488
+    },
+    {
+      "epoch": 0.11113636363636363,
+      "grad_norm": 1.7902508501797942,
+      "learning_rate": 1.2124434373712473e-06,
+      "loss": 0.0908,
+      "step": 489
+    },
+    {
+      "epoch": 0.11136363636363636,
+      "grad_norm": 4.034809847891115,
+      "learning_rate": 1.2122909276839095e-06,
+      "loss": 0.1272,
+      "step": 490
+    },
+    {
+      "epoch": 0.11159090909090909,
+      "grad_norm": 3.1685367135864575,
+      "learning_rate": 1.2121381185994192e-06,
+      "loss": 0.0878,
+      "step": 491
+    },
+    {
+      "epoch": 0.11181818181818182,
+      "grad_norm": 1.589573409739006,
+      "learning_rate": 1.211985010195677e-06,
+      "loss": 0.0479,
+      "step": 492
+    },
+    {
+      "epoch": 0.11204545454545455,
+      "grad_norm": 2.504739016351825,
+      "learning_rate": 1.2118316025507369e-06,
+      "loss": 0.0802,
+      "step": 493
+    },
+    {
+      "epoch": 0.11227272727272727,
+      "grad_norm": 1.9540195021173163,
+      "learning_rate": 1.2116778957428046e-06,
+      "loss": 0.1181,
+      "step": 494
+    },
+    {
+      "epoch": 0.1125,
+      "grad_norm": 3.3334987373499483,
+      "learning_rate": 1.2115238898502395e-06,
+      "loss": 0.0742,
+      "step": 495
+    },
+    {
+      "epoch": 0.11272727272727273,
+      "grad_norm": 3.025530391416926,
+      "learning_rate": 1.2113695849515527e-06,
+      "loss": 0.192,
+      "step": 496
+    },
+    {
+      "epoch": 0.11295454545454546,
+      "grad_norm": 6.741411762616305,
+      "learning_rate": 1.2112149811254076e-06,
+      "loss": 0.1643,
+      "step": 497
+    },
+    {
+      "epoch": 0.11318181818181818,
+      "grad_norm": 2.9075771835692916,
+      "learning_rate": 1.2110600784506202e-06,
+      "loss": 0.0944,
+      "step": 498
+    },
+    {
+      "epoch": 0.1134090909090909,
+      "grad_norm": 2.3717464191846527,
+      "learning_rate": 1.2109048770061593e-06,
+      "loss": 0.0694,
+      "step": 499
+    },
+    {
+      "epoch": 0.11363636363636363,
+      "grad_norm": 2.831854491797305,
+      "learning_rate": 1.2107493768711453e-06,
+      "loss": 0.0985,
+      "step": 500
+    },
+    {
+      "epoch": 0.11386363636363636,
+      "grad_norm": 1.9860156119638017,
+      "learning_rate": 1.2105935781248513e-06,
+      "loss": 0.0867,
+      "step": 501
+    },
+    {
+      "epoch": 0.11409090909090909,
+      "grad_norm": 5.228951709825794,
+      "learning_rate": 1.2104374808467023e-06,
+      "loss": 0.2457,
+      "step": 502
+    },
+    {
+      "epoch": 0.11431818181818182,
+      "grad_norm": 3.510768700023396,
+      "learning_rate": 1.2102810851162758e-06,
+      "loss": 0.1022,
+      "step": 503
+    },
+    {
+      "epoch": 0.11454545454545455,
+      "grad_norm": 3.5140110574002477,
+      "learning_rate": 1.2101243910133013e-06,
+      "loss": 0.0812,
+      "step": 504
+    },
+    {
+      "epoch": 0.11477272727272728,
+      "grad_norm": 2.554267446591213,
+      "learning_rate": 1.2099673986176604e-06,
+      "loss": 0.136,
+      "step": 505
+    },
+    {
+      "epoch": 0.115,
+      "grad_norm": 2.518498587543057,
+      "learning_rate": 1.2098101080093873e-06,
+      "loss": 0.1217,
+      "step": 506
+    },
+    {
+      "epoch": 0.11522727272727273,
+      "grad_norm": 3.8573725853503675,
+      "learning_rate": 1.2096525192686673e-06,
+      "loss": 0.2324,
+      "step": 507
+    },
+    {
+      "epoch": 0.11545454545454545,
+      "grad_norm": 5.743189687413489,
+      "learning_rate": 1.209494632475838e-06,
+      "loss": 0.1057,
+      "step": 508
+    },
+    {
+      "epoch": 0.11568181818181818,
+      "grad_norm": 2.959400513674917,
+      "learning_rate": 1.2093364477113893e-06,
+      "loss": 0.0781,
+      "step": 509
+    },
+    {
+      "epoch": 0.1159090909090909,
+      "grad_norm": 4.277580657761236,
+      "learning_rate": 1.2091779650559628e-06,
+      "loss": 0.0977,
+      "step": 510
+    },
+    {
+      "epoch": 0.11613636363636363,
+      "grad_norm": 2.517412718608715,
+      "learning_rate": 1.209019184590352e-06,
+      "loss": 0.0927,
+      "step": 511
+    },
+    {
+      "epoch": 0.11636363636363636,
+      "grad_norm": 4.4850972401394715,
+      "learning_rate": 1.2088601063955018e-06,
+      "loss": 0.1107,
+      "step": 512
+    },
+    {
+      "epoch": 0.11659090909090909,
+      "grad_norm": 4.277166175095006,
+      "learning_rate": 1.20870073055251e-06,
+      "loss": 0.1196,
+      "step": 513
+    },
+    {
+      "epoch": 0.11681818181818182,
+      "grad_norm": 3.3188059291020373,
+      "learning_rate": 1.2085410571426244e-06,
+      "loss": 0.1557,
+      "step": 514
+    },
+    {
+      "epoch": 0.11704545454545455,
+      "grad_norm": 2.835621253981015,
+      "learning_rate": 1.208381086247246e-06,
+      "loss": 0.1512,
+      "step": 515
+    },
+    {
+      "epoch": 0.11727272727272728,
+      "grad_norm": 2.5522170088661316,
+      "learning_rate": 1.2082208179479272e-06,
+      "loss": 0.1309,
+      "step": 516
+    },
+    {
+      "epoch": 0.1175,
+      "grad_norm": 5.645393854993232,
+      "learning_rate": 1.2080602523263715e-06,
+      "loss": 0.1793,
+      "step": 517
+    },
+    {
+      "epoch": 0.11772727272727272,
+      "grad_norm": 2.902480831513025,
+      "learning_rate": 1.207899389464434e-06,
+      "loss": 0.0973,
+      "step": 518
+    },
+    {
+      "epoch": 0.11795454545454545,
+      "grad_norm": 3.0474449053407677,
+      "learning_rate": 1.2077382294441218e-06,
+      "loss": 0.0922,
+      "step": 519
+    },
+    {
+      "epoch": 0.11818181818181818,
+      "grad_norm": 2.1972480331566917,
+      "learning_rate": 1.2075767723475932e-06,
+      "loss": 0.0479,
+      "step": 520
+    },
+    {
+      "epoch": 0.11840909090909091,
+      "grad_norm": 4.718271313739516,
+      "learning_rate": 1.2074150182571579e-06,
+      "loss": 0.1399,
+      "step": 521
+    },
+    {
+      "epoch": 0.11863636363636364,
+      "grad_norm": 3.1548176286505325,
+      "learning_rate": 1.2072529672552771e-06,
+      "loss": 0.1138,
+      "step": 522
+    },
+    {
+      "epoch": 0.11886363636363637,
+      "grad_norm": 3.5207935721656973,
+      "learning_rate": 1.2070906194245634e-06,
+      "loss": 0.1088,
+      "step": 523
+    },
+    {
+      "epoch": 0.1190909090909091,
+      "grad_norm": 3.0581878473011614,
+      "learning_rate": 1.2069279748477812e-06,
+      "loss": 0.0786,
+      "step": 524
+    },
+    {
+      "epoch": 0.11931818181818182,
+      "grad_norm": 2.5403240998391845,
+      "learning_rate": 1.206765033607845e-06,
+      "loss": 0.1118,
+      "step": 525
+    },
+    {
+      "epoch": 0.11954545454545455,
+      "grad_norm": 2.581209186135896,
+      "learning_rate": 1.2066017957878212e-06,
+      "loss": 0.1475,
+      "step": 526
+    },
+    {
+      "epoch": 0.11977272727272727,
+      "grad_norm": 2.7978694227102583,
+      "learning_rate": 1.2064382614709276e-06,
+      "loss": 0.1583,
+      "step": 527
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.230977957726528,
+      "learning_rate": 1.206274430740533e-06,
+      "loss": 0.1672,
+      "step": 528
+    },
+    {
+      "epoch": 0.12022727272727272,
+      "grad_norm": 4.19842301957619,
+      "learning_rate": 1.2061103036801573e-06,
+      "loss": 0.1482,
+      "step": 529
+    },
+    {
+      "epoch": 0.12045454545454545,
+      "grad_norm": 5.770691163267776,
+      "learning_rate": 1.2059458803734712e-06,
+      "loss": 0.1298,
+      "step": 530
+    },
+    {
+      "epoch": 0.12068181818181818,
+      "grad_norm": 3.972768667356933,
+      "learning_rate": 1.2057811609042968e-06,
+      "loss": 0.0998,
+      "step": 531
+    },
+    {
+      "epoch": 0.12090909090909091,
+      "grad_norm": 5.080522193929101,
+      "learning_rate": 1.205616145356607e-06,
+      "loss": 0.1786,
+      "step": 532
+    },
+    {
+      "epoch": 0.12113636363636364,
+      "grad_norm": 3.6416215519669968,
+      "learning_rate": 1.2054508338145257e-06,
+      "loss": 0.1879,
+      "step": 533
+    },
+    {
+      "epoch": 0.12136363636363637,
+      "grad_norm": 3.3554603219618446,
+      "learning_rate": 1.2052852263623274e-06,
+      "loss": 0.1129,
+      "step": 534
+    },
+    {
+      "epoch": 0.1215909090909091,
+      "grad_norm": 2.9806298846905546,
+      "learning_rate": 1.2051193230844382e-06,
+      "loss": 0.0715,
+      "step": 535
+    },
+    {
+      "epoch": 0.12181818181818181,
+      "grad_norm": 3.233094123117659,
+      "learning_rate": 1.2049531240654343e-06,
+      "loss": 0.1178,
+      "step": 536
+    },
+    {
+      "epoch": 0.12204545454545454,
+      "grad_norm": 3.708896515288204,
+      "learning_rate": 1.2047866293900428e-06,
+      "loss": 0.1783,
+      "step": 537
+    },
+    {
+      "epoch": 0.12227272727272727,
+      "grad_norm": 2.6602901122102716,
+      "learning_rate": 1.2046198391431415e-06,
+      "loss": 0.1418,
+      "step": 538
+    },
+    {
+      "epoch": 0.1225,
+      "grad_norm": 4.669634772140572,
+      "learning_rate": 1.2044527534097595e-06,
+      "loss": 0.173,
+      "step": 539
+    },
+    {
+      "epoch": 0.12272727272727273,
+      "grad_norm": 1.5028169649693566,
+      "learning_rate": 1.2042853722750756e-06,
+      "loss": 0.039,
+      "step": 540
+    },
+    {
+      "epoch": 0.12295454545454546,
+      "grad_norm": 2.904497335692425,
+      "learning_rate": 1.2041176958244197e-06,
+      "loss": 0.0888,
+      "step": 541
+    },
+    {
+      "epoch": 0.12318181818181818,
+      "grad_norm": 2.6451929963797713,
+      "learning_rate": 1.2039497241432724e-06,
+      "loss": 0.0682,
+      "step": 542
+    },
+    {
+      "epoch": 0.12340909090909091,
+      "grad_norm": 3.7170549013934435,
+      "learning_rate": 1.2037814573172642e-06,
+      "loss": 0.1411,
+      "step": 543
+    },
+    {
+      "epoch": 0.12363636363636364,
+      "grad_norm": 2.474660313130358,
+      "learning_rate": 1.2036128954321768e-06,
+      "loss": 0.0947,
+      "step": 544
+    },
+    {
+      "epoch": 0.12386363636363637,
+      "grad_norm": 3.1889690497542715,
+      "learning_rate": 1.2034440385739418e-06,
+      "loss": 0.0817,
+      "step": 545
+    },
+    {
+      "epoch": 0.12409090909090909,
+      "grad_norm": 4.537820626453625,
+      "learning_rate": 1.2032748868286415e-06,
+      "loss": 0.2305,
+      "step": 546
+    },
+    {
+      "epoch": 0.12431818181818181,
+      "grad_norm": 3.4406941290200383,
+      "learning_rate": 1.2031054402825082e-06,
+      "loss": 0.0741,
+      "step": 547
+    },
+    {
+      "epoch": 0.12454545454545454,
+      "grad_norm": 2.195149120400146,
+      "learning_rate": 1.2029356990219248e-06,
+      "loss": 0.1076,
+      "step": 548
+    },
+    {
+      "epoch": 0.12477272727272727,
+      "grad_norm": 2.060858796740431,
+      "learning_rate": 1.2027656631334242e-06,
+      "loss": 0.1402,
+      "step": 549
+    },
+    {
+      "epoch": 0.125,
+      "grad_norm": 2.8721644779044535,
+      "learning_rate": 1.2025953327036897e-06,
+      "loss": 0.1092,
+      "step": 550
+    },
+    {
+      "epoch": 0.12522727272727271,
+      "grad_norm": 3.367360775893672,
+      "learning_rate": 1.2024247078195542e-06,
+      "loss": 0.1058,
+      "step": 551
+    },
+    {
+      "epoch": 0.12545454545454546,
+      "grad_norm": 7.355200822564602,
+      "learning_rate": 1.202253788568002e-06,
+      "loss": 0.1409,
+      "step": 552
+    },
+    {
+      "epoch": 0.12568181818181817,
+      "grad_norm": 3.223527339373556,
+      "learning_rate": 1.202082575036166e-06,
+      "loss": 0.1908,
+      "step": 553
+    },
+    {
+      "epoch": 0.12590909090909091,
+      "grad_norm": 2.5935790347418846,
+      "learning_rate": 1.2019110673113302e-06,
+      "loss": 0.126,
+      "step": 554
+    },
+    {
+      "epoch": 0.12613636363636363,
+      "grad_norm": 2.6788096848560268,
+      "learning_rate": 1.2017392654809278e-06,
+      "loss": 0.0734,
+      "step": 555
+    },
+    {
+      "epoch": 0.12636363636363637,
+      "grad_norm": 2.3283744556449935,
+      "learning_rate": 1.2015671696325423e-06,
+      "loss": 0.1331,
+      "step": 556
+    },
+    {
+      "epoch": 0.1265909090909091,
+      "grad_norm": 5.27073941861621,
+      "learning_rate": 1.2013947798539073e-06,
+      "loss": 0.0878,
+      "step": 557
+    },
+    {
+      "epoch": 0.12681818181818183,
+      "grad_norm": 2.606289849041831,
+      "learning_rate": 1.2012220962329058e-06,
+      "loss": 0.0542,
+      "step": 558
+    },
+    {
+      "epoch": 0.12704545454545454,
+      "grad_norm": 6.526314652415761,
+      "learning_rate": 1.201049118857571e-06,
+      "loss": 0.233,
+      "step": 559
+    },
+    {
+      "epoch": 0.12727272727272726,
+      "grad_norm": 2.2175172650303594,
+      "learning_rate": 1.2008758478160853e-06,
+      "loss": 0.0774,
+      "step": 560
+    },
+    {
+      "epoch": 0.1275,
+      "grad_norm": 3.6763556159095567,
+      "learning_rate": 1.2007022831967813e-06,
+      "loss": 0.1373,
+      "step": 561
+    },
+    {
+      "epoch": 0.12772727272727272,
+      "grad_norm": 2.8295415297405615,
+      "learning_rate": 1.2005284250881417e-06,
+      "loss": 0.1133,
+      "step": 562
+    },
+    {
+      "epoch": 0.12795454545454546,
+      "grad_norm": 2.974517601517034,
+      "learning_rate": 1.2003542735787973e-06,
+      "loss": 0.1333,
+      "step": 563
+    },
+    {
+      "epoch": 0.12818181818181817,
+      "grad_norm": 2.715704825781891,
+      "learning_rate": 1.20017982875753e-06,
+      "loss": 0.0901,
+      "step": 564
+    },
+    {
+      "epoch": 0.12840909090909092,
+      "grad_norm": 1.9739259407297394,
+      "learning_rate": 1.2000050907132705e-06,
+      "loss": 0.067,
+      "step": 565
+    },
+    {
+      "epoch": 0.12863636363636363,
+      "grad_norm": 2.906735651308808,
+      "learning_rate": 1.1998300595350993e-06,
+      "loss": 0.1435,
+      "step": 566
+    },
+    {
+      "epoch": 0.12886363636363637,
+      "grad_norm": 3.1949368752448484,
+      "learning_rate": 1.1996547353122461e-06,
+      "loss": 0.0796,
+      "step": 567
+    },
+    {
+      "epoch": 0.1290909090909091,
+      "grad_norm": 3.8738935296067205,
+      "learning_rate": 1.1994791181340897e-06,
+      "loss": 0.14,
+      "step": 568
+    },
+    {
+      "epoch": 0.1293181818181818,
+      "grad_norm": 2.9128094894337924,
+      "learning_rate": 1.1993032080901593e-06,
+      "loss": 0.1087,
+      "step": 569
+    },
+    {
+      "epoch": 0.12954545454545455,
+      "grad_norm": 2.1283511235172416,
+      "learning_rate": 1.1991270052701323e-06,
+      "loss": 0.0727,
+      "step": 570
+    },
+    {
+      "epoch": 0.12977272727272726,
+      "grad_norm": 2.95662718032164,
+      "learning_rate": 1.1989505097638357e-06,
+      "loss": 0.1825,
+      "step": 571
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0152576822935546,
+      "learning_rate": 1.198773721661246e-06,
+      "loss": 0.1685,
+      "step": 572
+    },
+    {
+      "epoch": 0.13022727272727272,
+      "grad_norm": 3.716459362719905,
+      "learning_rate": 1.1985966410524883e-06,
+      "loss": 0.1062,
+      "step": 573
+    },
+    {
+      "epoch": 0.13045454545454546,
+      "grad_norm": 2.3425653658909917,
+      "learning_rate": 1.1984192680278376e-06,
+      "loss": 0.0576,
+      "step": 574
+    },
+    {
+      "epoch": 0.13068181818181818,
+      "grad_norm": 3.298436531129224,
+      "learning_rate": 1.1982416026777172e-06,
+      "loss": 0.1288,
+      "step": 575
+    },
+    {
+      "epoch": 0.13090909090909092,
+      "grad_norm": 2.6354717815457924,
+      "learning_rate": 1.1980636450926999e-06,
+      "loss": 0.0695,
+      "step": 576
+    },
+    {
+      "epoch": 0.13113636363636363,
+      "grad_norm": 1.4220237634639286,
+      "learning_rate": 1.1978853953635074e-06,
+      "loss": 0.0528,
+      "step": 577
+    },
+    {
+      "epoch": 0.13136363636363638,
+      "grad_norm": 2.468609110409853,
+      "learning_rate": 1.1977068535810101e-06,
+      "loss": 0.104,
+      "step": 578
+    },
+    {
+      "epoch": 0.1315909090909091,
+      "grad_norm": 4.366562461588067,
+      "learning_rate": 1.1975280198362276e-06,
+      "loss": 0.1798,
+      "step": 579
+    },
+    {
+      "epoch": 0.1318181818181818,
+      "grad_norm": 3.202193742998132,
+      "learning_rate": 1.1973488942203282e-06,
+      "loss": 0.0977,
+      "step": 580
+    },
+    {
+      "epoch": 0.13204545454545455,
+      "grad_norm": 2.5513005244661504,
+      "learning_rate": 1.197169476824629e-06,
+      "loss": 0.0861,
+      "step": 581
+    },
+    {
+      "epoch": 0.13227272727272726,
+      "grad_norm": 3.5563258599113,
+      "learning_rate": 1.1969897677405956e-06,
+      "loss": 0.1155,
+      "step": 582
+    },
+    {
+      "epoch": 0.1325,
+      "grad_norm": 2.768239718969493,
+      "learning_rate": 1.1968097670598428e-06,
+      "loss": 0.1535,
+      "step": 583
+    },
+    {
+      "epoch": 0.13272727272727272,
+      "grad_norm": 4.5655239624813495,
+      "learning_rate": 1.1966294748741336e-06,
+      "loss": 0.0901,
+      "step": 584
+    },
+    {
+      "epoch": 0.13295454545454546,
+      "grad_norm": 2.5958170708040393,
+      "learning_rate": 1.19644889127538e-06,
+      "loss": 0.078,
+      "step": 585
+    },
+    {
+      "epoch": 0.13318181818181818,
+      "grad_norm": 3.1883648289427304,
+      "learning_rate": 1.1962680163556424e-06,
+      "loss": 0.0712,
+      "step": 586
+    },
+    {
+      "epoch": 0.13340909090909092,
+      "grad_norm": 2.1320009706723733,
+      "learning_rate": 1.1960868502071294e-06,
+      "loss": 0.1475,
+      "step": 587
+    },
+    {
+      "epoch": 0.13363636363636364,
+      "grad_norm": 1.9940879792713833,
+      "learning_rate": 1.1959053929221984e-06,
+      "loss": 0.0981,
+      "step": 588
+    },
+    {
+      "epoch": 0.13386363636363635,
+      "grad_norm": 3.4463318823874394,
+      "learning_rate": 1.1957236445933553e-06,
+      "loss": 0.1088,
+      "step": 589
+    },
+    {
+      "epoch": 0.1340909090909091,
+      "grad_norm": 1.8900755034083492,
+      "learning_rate": 1.1955416053132542e-06,
+      "loss": 0.0494,
+      "step": 590
+    },
+    {
+      "epoch": 0.1343181818181818,
+      "grad_norm": 3.9000513472596174,
+      "learning_rate": 1.1953592751746976e-06,
+      "loss": 0.1265,
+      "step": 591
+    },
+    {
+      "epoch": 0.13454545454545455,
+      "grad_norm": 4.291095102811638,
+      "learning_rate": 1.1951766542706362e-06,
+      "loss": 0.1225,
+      "step": 592
+    },
+    {
+      "epoch": 0.13477272727272727,
+      "grad_norm": 2.6405027569368666,
+      "learning_rate": 1.1949937426941689e-06,
+      "loss": 0.0992,
+      "step": 593
+    },
+    {
+      "epoch": 0.135,
+      "grad_norm": 2.6465513102605644,
+      "learning_rate": 1.1948105405385428e-06,
+      "loss": 0.1436,
+      "step": 594
+    },
+    {
+      "epoch": 0.13522727272727272,
+      "grad_norm": 2.259803623482083,
+      "learning_rate": 1.1946270478971533e-06,
+      "loss": 0.1057,
+      "step": 595
+    },
+    {
+      "epoch": 0.13545454545454547,
+      "grad_norm": 3.068275739482322,
+      "learning_rate": 1.194443264863544e-06,
+      "loss": 0.1208,
+      "step": 596
+    },
+    {
+      "epoch": 0.13568181818181818,
+      "grad_norm": 3.3234464117570095,
+      "learning_rate": 1.1942591915314058e-06,
+      "loss": 0.1977,
+      "step": 597
+    },
+    {
+      "epoch": 0.1359090909090909,
+      "grad_norm": 4.1562159510577,
+      "learning_rate": 1.1940748279945784e-06,
+      "loss": 0.128,
+      "step": 598
+    },
+    {
+      "epoch": 0.13613636363636364,
+      "grad_norm": 1.4529094160225136,
+      "learning_rate": 1.1938901743470494e-06,
+      "loss": 0.0663,
+      "step": 599
+    },
+    {
+      "epoch": 0.13636363636363635,
+      "grad_norm": 3.815547251224526,
+      "learning_rate": 1.1937052306829534e-06,
+      "loss": 0.0906,
+      "step": 600
+    },
+    {
+      "epoch": 0.1365909090909091,
+      "grad_norm": 2.129072993389739,
+      "learning_rate": 1.1935199970965741e-06,
+      "loss": 0.0693,
+      "step": 601
+    },
+    {
+      "epoch": 0.1368181818181818,
+      "grad_norm": 4.248435434354939,
+      "learning_rate": 1.193334473682342e-06,
+      "loss": 0.1476,
+      "step": 602
+    },
+    {
+      "epoch": 0.13704545454545455,
+      "grad_norm": 2.5575392272255626,
+      "learning_rate": 1.193148660534836e-06,
+      "loss": 0.1503,
+      "step": 603
+    },
+    {
+      "epoch": 0.13727272727272727,
+      "grad_norm": 3.010421683808598,
+      "learning_rate": 1.1929625577487825e-06,
+      "loss": 0.1423,
+      "step": 604
+    },
+    {
+      "epoch": 0.1375,
+      "grad_norm": 2.944071062469372,
+      "learning_rate": 1.1927761654190552e-06,
+      "loss": 0.0726,
+      "step": 605
+    },
+    {
+      "epoch": 0.13772727272727273,
+      "grad_norm": 3.832747957363785,
+      "learning_rate": 1.192589483640676e-06,
+      "loss": 0.1073,
+      "step": 606
+    },
+    {
+      "epoch": 0.13795454545454544,
+      "grad_norm": 3.856733174427242,
+      "learning_rate": 1.1924025125088138e-06,
+      "loss": 0.1112,
+      "step": 607
+    },
+    {
+      "epoch": 0.13818181818181818,
+      "grad_norm": 2.751887610692139,
+      "learning_rate": 1.1922152521187854e-06,
+      "loss": 0.1334,
+      "step": 608
+    },
+    {
+      "epoch": 0.1384090909090909,
+      "grad_norm": 5.0873022883095285,
+      "learning_rate": 1.1920277025660553e-06,
+      "loss": 0.1435,
+      "step": 609
+    },
+    {
+      "epoch": 0.13863636363636364,
+      "grad_norm": 4.8249303136466475,
+      "learning_rate": 1.1918398639462345e-06,
+      "loss": 0.1195,
+      "step": 610
+    },
+    {
+      "epoch": 0.13886363636363636,
+      "grad_norm": 4.874472213239946,
+      "learning_rate": 1.1916517363550821e-06,
+      "loss": 0.1632,
+      "step": 611
+    },
+    {
+      "epoch": 0.1390909090909091,
+      "grad_norm": 2.481069373476785,
+      "learning_rate": 1.1914633198885047e-06,
+      "loss": 0.1337,
+      "step": 612
+    },
+    {
+      "epoch": 0.1393181818181818,
+      "grad_norm": 2.556171296030753,
+      "learning_rate": 1.1912746146425555e-06,
+      "loss": 0.0844,
+      "step": 613
+    },
+    {
+      "epoch": 0.13954545454545456,
+      "grad_norm": 3.73896252712052,
+      "learning_rate": 1.1910856207134352e-06,
+      "loss": 0.1516,
+      "step": 614
+    },
+    {
+      "epoch": 0.13977272727272727,
+      "grad_norm": 6.2522296867232505,
+      "learning_rate": 1.1908963381974916e-06,
+      "loss": 0.1254,
+      "step": 615
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.687889077036656,
+      "learning_rate": 1.19070676719122e-06,
+      "loss": 0.2058,
+      "step": 616
+    },
+    {
+      "epoch": 0.14022727272727273,
+      "grad_norm": 3.0307348123486886,
+      "learning_rate": 1.1905169077912623e-06,
+      "loss": 0.1165,
+      "step": 617
+    },
+    {
+      "epoch": 0.14045454545454544,
+      "grad_norm": 2.4821093454838543,
+      "learning_rate": 1.1903267600944077e-06,
+      "loss": 0.0626,
+      "step": 618
+    },
+    {
+      "epoch": 0.14068181818181819,
+      "grad_norm": 4.121754502209153,
+      "learning_rate": 1.1901363241975921e-06,
+      "loss": 0.0986,
+      "step": 619
+    },
+    {
+      "epoch": 0.1409090909090909,
+      "grad_norm": 3.071974785711444,
+      "learning_rate": 1.1899456001978987e-06,
+      "loss": 0.1503,
+      "step": 620
+    },
+    {
+      "epoch": 0.14113636363636364,
+      "grad_norm": 3.947611533293796,
+      "learning_rate": 1.1897545881925573e-06,
+      "loss": 0.1389,
+      "step": 621
+    },
+    {
+      "epoch": 0.14136363636363636,
+      "grad_norm": 2.076988193267275,
+      "learning_rate": 1.1895632882789447e-06,
+      "loss": 0.0882,
+      "step": 622
+    },
+    {
+      "epoch": 0.1415909090909091,
+      "grad_norm": 2.136166779141532,
+      "learning_rate": 1.1893717005545843e-06,
+      "loss": 0.0876,
+      "step": 623
+    },
+    {
+      "epoch": 0.14181818181818182,
+      "grad_norm": 5.3865638412573835,
+      "learning_rate": 1.189179825117146e-06,
+      "loss": 0.1935,
+      "step": 624
+    },
+    {
+      "epoch": 0.14204545454545456,
+      "grad_norm": 3.5245336234055045,
+      "learning_rate": 1.1889876620644472e-06,
+      "loss": 0.0805,
+      "step": 625
+    },
+    {
+      "epoch": 0.14227272727272727,
+      "grad_norm": 4.160973452396278,
+      "learning_rate": 1.1887952114944509e-06,
+      "loss": 0.0884,
+      "step": 626
+    },
+    {
+      "epoch": 0.1425,
+      "grad_norm": 1.4391339653094455,
+      "learning_rate": 1.1886024735052676e-06,
+      "loss": 0.076,
+      "step": 627
+    },
+    {
+      "epoch": 0.14272727272727273,
+      "grad_norm": 2.868684131769496,
+      "learning_rate": 1.1884094481951535e-06,
+      "loss": 0.1337,
+      "step": 628
+    },
+    {
+      "epoch": 0.14295454545454545,
+      "grad_norm": 6.437272841990759,
+      "learning_rate": 1.1882161356625122e-06,
+      "loss": 0.1887,
+      "step": 629
+    },
+    {
+      "epoch": 0.1431818181818182,
+      "grad_norm": 2.6513885033687927,
+      "learning_rate": 1.1880225360058925e-06,
+      "loss": 0.1202,
+      "step": 630
+    },
+    {
+      "epoch": 0.1434090909090909,
+      "grad_norm": 3.5666999999288826,
+      "learning_rate": 1.1878286493239907e-06,
+      "loss": 0.1429,
+      "step": 631
+    },
+    {
+      "epoch": 0.14363636363636365,
+      "grad_norm": 1.919266408920944,
+      "learning_rate": 1.187634475715649e-06,
+      "loss": 0.07,
+      "step": 632
+    },
+    {
+      "epoch": 0.14386363636363636,
+      "grad_norm": 2.570698788827823,
+      "learning_rate": 1.1874400152798557e-06,
+      "loss": 0.0937,
+      "step": 633
+    },
+    {
+      "epoch": 0.1440909090909091,
+      "grad_norm": 2.6570945083915496,
+      "learning_rate": 1.1872452681157453e-06,
+      "loss": 0.0692,
+      "step": 634
+    },
+    {
+      "epoch": 0.14431818181818182,
+      "grad_norm": 2.464540582208454,
+      "learning_rate": 1.1870502343225992e-06,
+      "loss": 0.0699,
+      "step": 635
+    },
+    {
+      "epoch": 0.14454545454545453,
+      "grad_norm": 2.2956454633742513,
+      "learning_rate": 1.186854913999844e-06,
+      "loss": 0.1185,
+      "step": 636
+    },
+    {
+      "epoch": 0.14477272727272728,
+      "grad_norm": 3.2561941076718606,
+      "learning_rate": 1.1866593072470527e-06,
+      "loss": 0.1012,
+      "step": 637
+    },
+    {
+      "epoch": 0.145,
+      "grad_norm": 2.77513438199878,
+      "learning_rate": 1.1864634141639448e-06,
+      "loss": 0.0852,
+      "step": 638
+    },
+    {
+      "epoch": 0.14522727272727273,
+      "grad_norm": 3.332667452259652,
+      "learning_rate": 1.1862672348503848e-06,
+      "loss": 0.1335,
+      "step": 639
+    },
+    {
+      "epoch": 0.14545454545454545,
+      "grad_norm": 1.4736108733640751,
+      "learning_rate": 1.186070769406384e-06,
+      "loss": 0.0483,
+      "step": 640
+    },
+    {
+      "epoch": 0.1456818181818182,
+      "grad_norm": 4.5338067314683235,
+      "learning_rate": 1.185874017932099e-06,
+      "loss": 0.1104,
+      "step": 641
+    },
+    {
+      "epoch": 0.1459090909090909,
+      "grad_norm": 2.581622979177194,
+      "learning_rate": 1.1856769805278327e-06,
+      "loss": 0.0758,
+      "step": 642
+    },
+    {
+      "epoch": 0.14613636363636365,
+      "grad_norm": 4.8451302603373225,
+      "learning_rate": 1.1854796572940332e-06,
+      "loss": 0.255,
+      "step": 643
+    },
+    {
+      "epoch": 0.14636363636363636,
+      "grad_norm": 3.4060902404925764,
+      "learning_rate": 1.1852820483312951e-06,
+      "loss": 0.165,
+      "step": 644
+    },
+    {
+      "epoch": 0.14659090909090908,
+      "grad_norm": 3.7697134812372486,
+      "learning_rate": 1.1850841537403577e-06,
+      "loss": 0.0841,
+      "step": 645
+    },
+    {
+      "epoch": 0.14681818181818182,
+      "grad_norm": 4.04555361809533,
+      "learning_rate": 1.1848859736221062e-06,
+      "loss": 0.1852,
+      "step": 646
+    },
+    {
+      "epoch": 0.14704545454545453,
+      "grad_norm": 5.292624605883491,
+      "learning_rate": 1.1846875080775724e-06,
+      "loss": 0.1396,
+      "step": 647
+    },
+    {
+      "epoch": 0.14727272727272728,
+      "grad_norm": 2.9010815097955596,
+      "learning_rate": 1.1844887572079322e-06,
+      "loss": 0.0809,
+      "step": 648
+    },
+    {
+      "epoch": 0.1475,
+      "grad_norm": 3.0154321976547265,
+      "learning_rate": 1.1842897211145075e-06,
+      "loss": 0.1082,
+      "step": 649
+    },
+    {
+      "epoch": 0.14772727272727273,
+      "grad_norm": 3.603164933951076,
+      "learning_rate": 1.1840903998987657e-06,
+      "loss": 0.0904,
+      "step": 650
+    },
+    {
+      "epoch": 0.14795454545454545,
+      "grad_norm": 2.611926077443849,
+      "learning_rate": 1.1838907936623196e-06,
+      "loss": 0.0923,
+      "step": 651
+    },
+    {
+      "epoch": 0.1481818181818182,
+      "grad_norm": 3.5045755536602647,
+      "learning_rate": 1.183690902506927e-06,
+      "loss": 0.1367,
+      "step": 652
+    },
+    {
+      "epoch": 0.1484090909090909,
+      "grad_norm": 2.079001413436961,
+      "learning_rate": 1.1834907265344913e-06,
+      "loss": 0.1189,
+      "step": 653
+    },
+    {
+      "epoch": 0.14863636363636365,
+      "grad_norm": 1.927037422831208,
+      "learning_rate": 1.1832902658470608e-06,
+      "loss": 0.115,
+      "step": 654
+    },
+    {
+      "epoch": 0.14886363636363636,
+      "grad_norm": 1.97236561610899,
+      "learning_rate": 1.1830895205468293e-06,
+      "loss": 0.06,
+      "step": 655
+    },
+    {
+      "epoch": 0.14909090909090908,
+      "grad_norm": 2.8828111328774493,
+      "learning_rate": 1.182888490736135e-06,
+      "loss": 0.0996,
+      "step": 656
+    },
+    {
+      "epoch": 0.14931818181818182,
+      "grad_norm": 4.106350195454931,
+      "learning_rate": 1.1826871765174622e-06,
+      "loss": 0.1652,
+      "step": 657
+    },
+    {
+      "epoch": 0.14954545454545454,
+      "grad_norm": 4.5075997872634925,
+      "learning_rate": 1.1824855779934392e-06,
+      "loss": 0.1068,
+      "step": 658
+    },
+    {
+      "epoch": 0.14977272727272728,
+      "grad_norm": 2.7307866883608933,
+      "learning_rate": 1.1822836952668397e-06,
+      "loss": 0.129,
+      "step": 659
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8969430815297914,
+      "learning_rate": 1.182081528440582e-06,
+      "loss": 0.0927,
+      "step": 660
+    },
+    {
+      "epoch": 0.15022727272727274,
+      "grad_norm": 2.222251072329309,
+      "learning_rate": 1.18187907761773e-06,
+      "loss": 0.0806,
+      "step": 661
+    },
+    {
+      "epoch": 0.15045454545454545,
+      "grad_norm": 2.855738831847418,
+      "learning_rate": 1.1816763429014917e-06,
+      "loss": 0.1133,
+      "step": 662
+    },
+    {
+      "epoch": 0.1506818181818182,
+      "grad_norm": 4.137168631966571,
+      "learning_rate": 1.1814733243952193e-06,
+      "loss": 0.1729,
+      "step": 663
+    },
+    {
+      "epoch": 0.1509090909090909,
+      "grad_norm": 5.250074138141893,
+      "learning_rate": 1.1812700222024111e-06,
+      "loss": 0.2671,
+      "step": 664
+    },
+    {
+      "epoch": 0.15113636363636362,
+      "grad_norm": 4.095009204321318,
+      "learning_rate": 1.1810664364267092e-06,
+      "loss": 0.0975,
+      "step": 665
+    },
+    {
+      "epoch": 0.15136363636363637,
+      "grad_norm": 1.9872727288340575,
+      "learning_rate": 1.1808625671718999e-06,
+      "loss": 0.0559,
+      "step": 666
+    },
+    {
+      "epoch": 0.15159090909090908,
+      "grad_norm": 2.9314970455180593,
+      "learning_rate": 1.1806584145419144e-06,
+      "loss": 0.1494,
+      "step": 667
+    },
+    {
+      "epoch": 0.15181818181818182,
+      "grad_norm": 3.22472741979199,
+      "learning_rate": 1.1804539786408292e-06,
+      "loss": 0.1201,
+      "step": 668
+    },
+    {
+      "epoch": 0.15204545454545454,
+      "grad_norm": 2.8590813055772557,
+      "learning_rate": 1.1802492595728634e-06,
+      "loss": 0.0895,
+      "step": 669
+    },
+    {
+      "epoch": 0.15227272727272728,
+      "grad_norm": 2.4103185687009576,
+      "learning_rate": 1.1800442574423823e-06,
+      "loss": 0.0843,
+      "step": 670
+    },
+    {
+      "epoch": 0.1525,
+      "grad_norm": 3.191758442329034,
+      "learning_rate": 1.1798389723538942e-06,
+      "loss": 0.0784,
+      "step": 671
+    },
+    {
+      "epoch": 0.15272727272727274,
+      "grad_norm": 3.8816280749453114,
+      "learning_rate": 1.1796334044120522e-06,
+      "loss": 0.153,
+      "step": 672
+    },
+    {
+      "epoch": 0.15295454545454545,
+      "grad_norm": 2.7772880110025238,
+      "learning_rate": 1.1794275537216534e-06,
+      "loss": 0.1308,
+      "step": 673
+    },
+    {
+      "epoch": 0.15318181818181817,
+      "grad_norm": 1.7311480217251134,
+      "learning_rate": 1.1792214203876396e-06,
+      "loss": 0.0612,
+      "step": 674
+    },
+    {
+      "epoch": 0.1534090909090909,
+      "grad_norm": 1.8326204714498646,
+      "learning_rate": 1.1790150045150958e-06,
+      "loss": 0.0819,
+      "step": 675
+    },
+    {
+      "epoch": 0.15363636363636363,
+      "grad_norm": 3.7804068167345344,
+      "learning_rate": 1.1788083062092518e-06,
+      "loss": 0.0889,
+      "step": 676
+    },
+    {
+      "epoch": 0.15386363636363637,
+      "grad_norm": 2.429048128257739,
+      "learning_rate": 1.1786013255754808e-06,
+      "loss": 0.1183,
+      "step": 677
+    },
+    {
+      "epoch": 0.15409090909090908,
+      "grad_norm": 3.6912740206894132,
+      "learning_rate": 1.1783940627193002e-06,
+      "loss": 0.1342,
+      "step": 678
+    },
+    {
+      "epoch": 0.15431818181818183,
+      "grad_norm": 2.315919697767897,
+      "learning_rate": 1.1781865177463717e-06,
+      "loss": 0.0948,
+      "step": 679
+    },
+    {
+      "epoch": 0.15454545454545454,
+      "grad_norm": 4.732313717049891,
+      "learning_rate": 1.1779786907625e-06,
+      "loss": 0.2081,
+      "step": 680
+    },
+    {
+      "epoch": 0.15477272727272728,
+      "grad_norm": 3.0643026784385503,
+      "learning_rate": 1.177770581873634e-06,
+      "loss": 0.1308,
+      "step": 681
+    },
+    {
+      "epoch": 0.155,
+      "grad_norm": 3.596674366098283,
+      "learning_rate": 1.1775621911858665e-06,
+      "loss": 0.1835,
+      "step": 682
+    },
+    {
+      "epoch": 0.1552272727272727,
+      "grad_norm": 2.7629336340919957,
+      "learning_rate": 1.1773535188054336e-06,
+      "loss": 0.0822,
+      "step": 683
+    },
+    {
+      "epoch": 0.15545454545454546,
+      "grad_norm": 3.3873361914536924,
+      "learning_rate": 1.177144564838715e-06,
+      "loss": 0.1179,
+      "step": 684
+    },
+    {
+      "epoch": 0.15568181818181817,
+      "grad_norm": 2.0404782084317112,
+      "learning_rate": 1.1769353293922341e-06,
+      "loss": 0.1189,
+      "step": 685
+    },
+    {
+      "epoch": 0.1559090909090909,
+      "grad_norm": 2.833626094610276,
+      "learning_rate": 1.1767258125726584e-06,
+      "loss": 0.0793,
+      "step": 686
+    },
+    {
+      "epoch": 0.15613636363636363,
+      "grad_norm": 3.78825191552253,
+      "learning_rate": 1.1765160144867972e-06,
+      "loss": 0.1063,
+      "step": 687
+    },
+    {
+      "epoch": 0.15636363636363637,
+      "grad_norm": 2.6397145292980664,
+      "learning_rate": 1.176305935241605e-06,
+      "loss": 0.0804,
+      "step": 688
+    },
+    {
+      "epoch": 0.1565909090909091,
+      "grad_norm": 3.8775177404721846,
+      "learning_rate": 1.1760955749441786e-06,
+      "loss": 0.0998,
+      "step": 689
+    },
+    {
+      "epoch": 0.15681818181818183,
+      "grad_norm": 1.7480890862718244,
+      "learning_rate": 1.1758849337017587e-06,
+      "loss": 0.0372,
+      "step": 690
+    },
+    {
+      "epoch": 0.15704545454545454,
+      "grad_norm": 4.230959165744064,
+      "learning_rate": 1.175674011621728e-06,
+      "loss": 0.2064,
+      "step": 691
+    },
+    {
+      "epoch": 0.1572727272727273,
+      "grad_norm": 3.6828290600360227,
+      "learning_rate": 1.1754628088116138e-06,
+      "loss": 0.086,
+      "step": 692
+    },
+    {
+      "epoch": 0.1575,
+      "grad_norm": 2.425606630823805,
+      "learning_rate": 1.1752513253790861e-06,
+      "loss": 0.0977,
+      "step": 693
+    },
+    {
+      "epoch": 0.15772727272727272,
+      "grad_norm": 4.4212649317439485,
+      "learning_rate": 1.1750395614319576e-06,
+      "loss": 0.1031,
+      "step": 694
+    },
+    {
+      "epoch": 0.15795454545454546,
+      "grad_norm": 4.531721414793234,
+      "learning_rate": 1.174827517078184e-06,
+      "loss": 0.1318,
+      "step": 695
+    },
+    {
+      "epoch": 0.15818181818181817,
+      "grad_norm": 1.2425254927354163,
+      "learning_rate": 1.1746151924258644e-06,
+      "loss": 0.046,
+      "step": 696
+    },
+    {
+      "epoch": 0.15840909090909092,
+      "grad_norm": 4.317477038748653,
+      "learning_rate": 1.1744025875832405e-06,
+      "loss": 0.1605,
+      "step": 697
+    },
+    {
+      "epoch": 0.15863636363636363,
+      "grad_norm": 2.2894514497196545,
+      "learning_rate": 1.174189702658697e-06,
+      "loss": 0.0667,
+      "step": 698
+    },
+    {
+      "epoch": 0.15886363636363637,
+      "grad_norm": 3.1747794315777975,
+      "learning_rate": 1.173976537760761e-06,
+      "loss": 0.178,
+      "step": 699
+    },
+    {
+      "epoch": 0.1590909090909091,
+      "grad_norm": 3.6397940252033054,
+      "learning_rate": 1.1737630929981026e-06,
+      "loss": 0.1234,
+      "step": 700
+    },
+    {
+      "epoch": 0.15931818181818183,
+      "grad_norm": 3.600394223025497,
+      "learning_rate": 1.1735493684795348e-06,
+      "loss": 0.1283,
+      "step": 701
+    },
+    {
+      "epoch": 0.15954545454545455,
+      "grad_norm": 2.1914776151398097,
+      "learning_rate": 1.173335364314013e-06,
+      "loss": 0.0861,
+      "step": 702
+    },
+    {
+      "epoch": 0.15977272727272726,
+      "grad_norm": 4.23410141845482,
+      "learning_rate": 1.173121080610635e-06,
+      "loss": 0.1418,
+      "step": 703
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8336906213405406,
+      "learning_rate": 1.1729065174786414e-06,
+      "loss": 0.0814,
+      "step": 704
+    },
+    {
+      "epoch": 0.16022727272727272,
+      "grad_norm": 1.324797775805725,
+      "learning_rate": 1.1726916750274148e-06,
+      "loss": 0.073,
+      "step": 705
+    },
+    {
+      "epoch": 0.16045454545454546,
+      "grad_norm": 1.1113270460751639,
+      "learning_rate": 1.1724765533664808e-06,
+      "loss": 0.0325,
+      "step": 706
+    },
+    {
+      "epoch": 0.16068181818181818,
+      "grad_norm": 3.329670829994521,
+      "learning_rate": 1.1722611526055073e-06,
+      "loss": 0.0846,
+      "step": 707
+    },
+    {
+      "epoch": 0.16090909090909092,
+      "grad_norm": 4.423260181219768,
+      "learning_rate": 1.1720454728543034e-06,
+      "loss": 0.0811,
+      "step": 708
+    },
+    {
+      "epoch": 0.16113636363636363,
+      "grad_norm": 3.861828831504347,
+      "learning_rate": 1.171829514222822e-06,
+      "loss": 0.1481,
+      "step": 709
+    },
+    {
+      "epoch": 0.16136363636363638,
+      "grad_norm": 3.63402240520416,
+      "learning_rate": 1.1716132768211572e-06,
+      "loss": 0.1154,
+      "step": 710
+    },
+    {
+      "epoch": 0.1615909090909091,
+      "grad_norm": 2.37915520220192,
+      "learning_rate": 1.1713967607595455e-06,
+      "loss": 0.0578,
+      "step": 711
+    },
+    {
+      "epoch": 0.1618181818181818,
+      "grad_norm": 4.41815181178646,
+      "learning_rate": 1.1711799661483653e-06,
+      "loss": 0.1123,
+      "step": 712
+    },
+    {
+      "epoch": 0.16204545454545455,
+      "grad_norm": 4.986293158191178,
+      "learning_rate": 1.170962893098137e-06,
+      "loss": 0.1236,
+      "step": 713
+    },
+    {
+      "epoch": 0.16227272727272726,
+      "grad_norm": 2.908333267858207,
+      "learning_rate": 1.1707455417195231e-06,
+      "loss": 0.1289,
+      "step": 714
+    },
+    {
+      "epoch": 0.1625,
+      "grad_norm": 3.5560662000352004,
+      "learning_rate": 1.170527912123328e-06,
+      "loss": 0.0964,
+      "step": 715
+    },
+    {
+      "epoch": 0.16272727272727272,
+      "grad_norm": 4.497015016872193,
+      "learning_rate": 1.1703100044204984e-06,
+      "loss": 0.1092,
+      "step": 716
+    },
+    {
+      "epoch": 0.16295454545454546,
+      "grad_norm": 2.8435936401692947,
+      "learning_rate": 1.1700918187221214e-06,
+      "loss": 0.1399,
+      "step": 717
+    },
+    {
+      "epoch": 0.16318181818181818,
+      "grad_norm": 3.4768244142021674,
+      "learning_rate": 1.169873355139427e-06,
+      "loss": 0.1007,
+      "step": 718
+    },
+    {
+      "epoch": 0.16340909090909092,
+      "grad_norm": 4.156552840769318,
+      "learning_rate": 1.1696546137837865e-06,
+      "loss": 0.1812,
+      "step": 719
+    },
+    {
+      "epoch": 0.16363636363636364,
+      "grad_norm": 3.6314627894350417,
+      "learning_rate": 1.169435594766713e-06,
+      "loss": 0.111,
+      "step": 720
+    },
+    {
+      "epoch": 0.16386363636363635,
+      "grad_norm": 2.248908961353641,
+      "learning_rate": 1.1692162981998608e-06,
+      "loss": 0.155,
+      "step": 721
+    },
+    {
+      "epoch": 0.1640909090909091,
+      "grad_norm": 3.5815672986259783,
+      "learning_rate": 1.1689967241950263e-06,
+      "loss": 0.105,
+      "step": 722
+    },
+    {
+      "epoch": 0.1643181818181818,
+      "grad_norm": 1.8662592496151242,
+      "learning_rate": 1.168776872864146e-06,
+      "loss": 0.0845,
+      "step": 723
+    },
+    {
+      "epoch": 0.16454545454545455,
+      "grad_norm": 3.5355589094258724,
+      "learning_rate": 1.1685567443192996e-06,
+      "loss": 0.1147,
+      "step": 724
+    },
+    {
+      "epoch": 0.16477272727272727,
+      "grad_norm": 2.7247318067219592,
+      "learning_rate": 1.168336338672707e-06,
+      "loss": 0.1253,
+      "step": 725
+    },
+    {
+      "epoch": 0.165,
+      "grad_norm": 3.5270674462195535,
+      "learning_rate": 1.1681156560367296e-06,
+      "loss": 0.1317,
+      "step": 726
+    },
+    {
+      "epoch": 0.16522727272727272,
+      "grad_norm": 3.8362477631577714,
+      "learning_rate": 1.1678946965238697e-06,
+      "loss": 0.2071,
+      "step": 727
+    },
+    {
+      "epoch": 0.16545454545454547,
+      "grad_norm": 2.445468298940002,
+      "learning_rate": 1.1676734602467713e-06,
+      "loss": 0.0797,
+      "step": 728
+    },
+    {
+      "epoch": 0.16568181818181818,
+      "grad_norm": 2.663027101554758,
+      "learning_rate": 1.1674519473182192e-06,
+      "loss": 0.0547,
+      "step": 729
+    },
+    {
+      "epoch": 0.16590909090909092,
+      "grad_norm": 3.9377492640314444,
+      "learning_rate": 1.1672301578511392e-06,
+      "loss": 0.1422,
+      "step": 730
+    },
+    {
+      "epoch": 0.16613636363636364,
+      "grad_norm": 3.1811680461685614,
+      "learning_rate": 1.167008091958598e-06,
+      "loss": 0.0796,
+      "step": 731
+    },
+    {
+      "epoch": 0.16636363636363635,
+      "grad_norm": 2.375864838080215,
+      "learning_rate": 1.1667857497538037e-06,
+      "loss": 0.0606,
+      "step": 732
+    },
+    {
+      "epoch": 0.1665909090909091,
+      "grad_norm": 2.5147347072402586,
+      "learning_rate": 1.166563131350105e-06,
+      "loss": 0.1209,
+      "step": 733
+    },
+    {
+      "epoch": 0.1668181818181818,
+      "grad_norm": 3.0045790328544033,
+      "learning_rate": 1.166340236860991e-06,
+      "loss": 0.2008,
+      "step": 734
+    },
+    {
+      "epoch": 0.16704545454545455,
+      "grad_norm": 2.217801589556193,
+      "learning_rate": 1.166117066400092e-06,
+      "loss": 0.1056,
+      "step": 735
+    },
+    {
+      "epoch": 0.16727272727272727,
+      "grad_norm": 5.587745484163016,
+      "learning_rate": 1.1658936200811789e-06,
+      "loss": 0.1688,
+      "step": 736
+    },
+    {
+      "epoch": 0.1675,
+      "grad_norm": 2.250124039431649,
+      "learning_rate": 1.1656698980181633e-06,
+      "loss": 0.0974,
+      "step": 737
+    },
+    {
+      "epoch": 0.16772727272727272,
+      "grad_norm": 3.80028598184578,
+      "learning_rate": 1.1654459003250971e-06,
+      "loss": 0.0992,
+      "step": 738
+    },
+    {
+      "epoch": 0.16795454545454547,
+      "grad_norm": 1.439158858739717,
+      "learning_rate": 1.1652216271161728e-06,
+      "loss": 0.0319,
+      "step": 739
+    },
+    {
+      "epoch": 0.16818181818181818,
+      "grad_norm": 2.2130946895334658,
+      "learning_rate": 1.1649970785057238e-06,
+      "loss": 0.1201,
+      "step": 740
+    },
+    {
+      "epoch": 0.1684090909090909,
+      "grad_norm": 2.7739842772717784,
+      "learning_rate": 1.1647722546082232e-06,
+      "loss": 0.0701,
+      "step": 741
+    },
+    {
+      "epoch": 0.16863636363636364,
+      "grad_norm": 2.609228118435818,
+      "learning_rate": 1.164547155538285e-06,
+      "loss": 0.0932,
+      "step": 742
+    },
+    {
+      "epoch": 0.16886363636363635,
+      "grad_norm": 3.703820173450133,
+      "learning_rate": 1.1643217814106633e-06,
+      "loss": 0.0899,
+      "step": 743
+    },
+    {
+      "epoch": 0.1690909090909091,
+      "grad_norm": 3.1694303813113804,
+      "learning_rate": 1.1640961323402522e-06,
+      "loss": 0.0868,
+      "step": 744
+    },
+    {
+      "epoch": 0.1693181818181818,
+      "grad_norm": 2.8568845461742995,
+      "learning_rate": 1.163870208442086e-06,
+      "loss": 0.1187,
+      "step": 745
+    },
+    {
+      "epoch": 0.16954545454545455,
+      "grad_norm": 3.797729399162994,
+      "learning_rate": 1.1636440098313398e-06,
+      "loss": 0.1722,
+      "step": 746
+    },
+    {
+      "epoch": 0.16977272727272727,
+      "grad_norm": 2.8670068838207574,
+      "learning_rate": 1.1634175366233278e-06,
+      "loss": 0.1619,
+      "step": 747
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.087548193361291,
+      "learning_rate": 1.1631907889335046e-06,
+      "loss": 0.0692,
+      "step": 748
+    },
+    {
+      "epoch": 0.17022727272727273,
+      "grad_norm": 4.414850992047754,
+      "learning_rate": 1.162963766877465e-06,
+      "loss": 0.0881,
+      "step": 749
+    },
+    {
+      "epoch": 0.17045454545454544,
+      "grad_norm": 2.411365953758672,
+      "learning_rate": 1.1627364705709428e-06,
+      "loss": 0.1214,
+      "step": 750
+    },
+    {
+      "epoch": 0.17068181818181818,
+      "grad_norm": 3.2669502739528653,
+      "learning_rate": 1.1625089001298129e-06,
+      "loss": 0.1274,
+      "step": 751
+    },
+    {
+      "epoch": 0.1709090909090909,
+      "grad_norm": 5.155420414197394,
+      "learning_rate": 1.1622810556700889e-06,
+      "loss": 0.2368,
+      "step": 752
+    },
+    {
+      "epoch": 0.17113636363636364,
+      "grad_norm": 2.2393607761538825,
+      "learning_rate": 1.1620529373079246e-06,
+      "loss": 0.0885,
+      "step": 753
+    },
+    {
+      "epoch": 0.17136363636363636,
+      "grad_norm": 3.6007275076022167,
+      "learning_rate": 1.1618245451596128e-06,
+      "loss": 0.084,
+      "step": 754
+    },
+    {
+      "epoch": 0.1715909090909091,
+      "grad_norm": 2.6165665043670323,
+      "learning_rate": 1.161595879341587e-06,
+      "loss": 0.1014,
+      "step": 755
+    },
+    {
+      "epoch": 0.17181818181818181,
+      "grad_norm": 3.6672242287695846,
+      "learning_rate": 1.1613669399704192e-06,
+      "loss": 0.1334,
+      "step": 756
+    },
+    {
+      "epoch": 0.17204545454545456,
+      "grad_norm": 3.8755698559282745,
+      "learning_rate": 1.1611377271628213e-06,
+      "loss": 0.1531,
+      "step": 757
+    },
+    {
+      "epoch": 0.17227272727272727,
+      "grad_norm": 3.4636367504088352,
+      "learning_rate": 1.1609082410356447e-06,
+      "loss": 0.0729,
+      "step": 758
+    },
+    {
+      "epoch": 0.1725,
+      "grad_norm": 4.159111572875205,
+      "learning_rate": 1.1606784817058797e-06,
+      "loss": 0.1439,
+      "step": 759
+    },
+    {
+      "epoch": 0.17272727272727273,
+      "grad_norm": 6.150329364181485,
+      "learning_rate": 1.1604484492906562e-06,
+      "loss": 0.2479,
+      "step": 760
+    },
+    {
+      "epoch": 0.17295454545454544,
+      "grad_norm": 2.840827652346105,
+      "learning_rate": 1.1602181439072432e-06,
+      "loss": 0.0529,
+      "step": 761
+    },
+    {
+      "epoch": 0.1731818181818182,
+      "grad_norm": 3.4976650338811326,
+      "learning_rate": 1.1599875656730492e-06,
+      "loss": 0.1441,
+      "step": 762
+    },
+    {
+      "epoch": 0.1734090909090909,
+      "grad_norm": 3.644483649886925,
+      "learning_rate": 1.159756714705621e-06,
+      "loss": 0.1627,
+      "step": 763
+    },
+    {
+      "epoch": 0.17363636363636364,
+      "grad_norm": 4.6658927901270575,
+      "learning_rate": 1.1595255911226456e-06,
+      "loss": 0.1154,
+      "step": 764
+    },
+    {
+      "epoch": 0.17386363636363636,
+      "grad_norm": 3.3625632086878956,
+      "learning_rate": 1.1592941950419475e-06,
+      "loss": 0.1272,
+      "step": 765
+    },
+    {
+      "epoch": 0.1740909090909091,
+      "grad_norm": 3.155381762903906,
+      "learning_rate": 1.1590625265814918e-06,
+      "loss": 0.1759,
+      "step": 766
+    },
+    {
+      "epoch": 0.17431818181818182,
+      "grad_norm": 2.4035789035627073,
+      "learning_rate": 1.1588305858593811e-06,
+      "loss": 0.083,
+      "step": 767
+    },
+    {
+      "epoch": 0.17454545454545456,
+      "grad_norm": 3.8800065252955807,
+      "learning_rate": 1.1585983729938575e-06,
+      "loss": 0.098,
+      "step": 768
+    },
+    {
+      "epoch": 0.17477272727272727,
+      "grad_norm": 3.2133140224754166,
+      "learning_rate": 1.1583658881033013e-06,
+      "loss": 0.1116,
+      "step": 769
+    },
+    {
+      "epoch": 0.175,
+      "grad_norm": 5.6308258883857665,
+      "learning_rate": 1.158133131306232e-06,
+      "loss": 0.1108,
+      "step": 770
+    },
+    {
+      "epoch": 0.17522727272727273,
+      "grad_norm": 3.34568115493396,
+      "learning_rate": 1.1579001027213078e-06,
+      "loss": 0.0828,
+      "step": 771
+    },
+    {
+      "epoch": 0.17545454545454545,
+      "grad_norm": 2.6258599364639466,
+      "learning_rate": 1.1576668024673248e-06,
+      "loss": 0.0705,
+      "step": 772
+    },
+    {
+      "epoch": 0.1756818181818182,
+      "grad_norm": 3.1555688975176004,
+      "learning_rate": 1.157433230663218e-06,
+      "loss": 0.0651,
+      "step": 773
+    },
+    {
+      "epoch": 0.1759090909090909,
+      "grad_norm": 3.611867108975687,
+      "learning_rate": 1.1571993874280611e-06,
+      "loss": 0.1436,
+      "step": 774
+    },
+    {
+      "epoch": 0.17613636363636365,
+      "grad_norm": 2.408343539432884,
+      "learning_rate": 1.1569652728810658e-06,
+      "loss": 0.0916,
+      "step": 775
+    },
+    {
+      "epoch": 0.17636363636363636,
+      "grad_norm": 4.976882099384663,
+      "learning_rate": 1.156730887141582e-06,
+      "loss": 0.1282,
+      "step": 776
+    },
+    {
+      "epoch": 0.1765909090909091,
+      "grad_norm": 5.294480823644898,
+      "learning_rate": 1.156496230329098e-06,
+      "loss": 0.2772,
+      "step": 777
+    },
+    {
+      "epoch": 0.17681818181818182,
+      "grad_norm": 4.742007157127061,
+      "learning_rate": 1.1562613025632406e-06,
+      "loss": 0.1196,
+      "step": 778
+    },
+    {
+      "epoch": 0.17704545454545453,
+      "grad_norm": 2.07528997472403,
+      "learning_rate": 1.1560261039637744e-06,
+      "loss": 0.0664,
+      "step": 779
+    },
+    {
+      "epoch": 0.17727272727272728,
+      "grad_norm": 2.4312830817964133,
+      "learning_rate": 1.1557906346506021e-06,
+      "loss": 0.0795,
+      "step": 780
+    },
+    {
+      "epoch": 0.1775,
+      "grad_norm": 3.2831149389763197,
+      "learning_rate": 1.1555548947437642e-06,
+      "loss": 0.1186,
+      "step": 781
+    },
+    {
+      "epoch": 0.17772727272727273,
+      "grad_norm": 4.073389939948538,
+      "learning_rate": 1.1553188843634399e-06,
+      "loss": 0.1068,
+      "step": 782
+    },
+    {
+      "epoch": 0.17795454545454545,
+      "grad_norm": 3.431137122030362,
+      "learning_rate": 1.1550826036299455e-06,
+      "loss": 0.2235,
+      "step": 783
+    },
+    {
+      "epoch": 0.1781818181818182,
+      "grad_norm": 2.1508864223211717,
+      "learning_rate": 1.1548460526637354e-06,
+      "loss": 0.0728,
+      "step": 784
+    },
+    {
+      "epoch": 0.1784090909090909,
+      "grad_norm": 2.542600883021502,
+      "learning_rate": 1.1546092315854017e-06,
+      "loss": 0.0713,
+      "step": 785
+    },
+    {
+      "epoch": 0.17863636363636365,
+      "grad_norm": 2.4077832811127644,
+      "learning_rate": 1.1543721405156744e-06,
+      "loss": 0.1331,
+      "step": 786
+    },
+    {
+      "epoch": 0.17886363636363636,
+      "grad_norm": 2.7882544425249454,
+      "learning_rate": 1.154134779575421e-06,
+      "loss": 0.1374,
+      "step": 787
+    },
+    {
+      "epoch": 0.17909090909090908,
+      "grad_norm": 4.0339750921784665,
+      "learning_rate": 1.1538971488856465e-06,
+      "loss": 0.1363,
+      "step": 788
+    },
+    {
+      "epoch": 0.17931818181818182,
+      "grad_norm": 1.7731148396393088,
+      "learning_rate": 1.153659248567494e-06,
+      "loss": 0.0533,
+      "step": 789
+    },
+    {
+      "epoch": 0.17954545454545454,
+      "grad_norm": 3.745684753145709,
+      "learning_rate": 1.1534210787422425e-06,
+      "loss": 0.1091,
+      "step": 790
+    },
+    {
+      "epoch": 0.17977272727272728,
+      "grad_norm": 2.9989434091372162,
+      "learning_rate": 1.1531826395313104e-06,
+      "loss": 0.0749,
+      "step": 791
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.215485933310487,
+      "learning_rate": 1.152943931056252e-06,
+      "loss": 0.2935,
+      "step": 792
+    },
+    {
+      "epoch": 0.18022727272727274,
+      "grad_norm": 2.5129589816791875,
+      "learning_rate": 1.1527049534387595e-06,
+      "loss": 0.1564,
+      "step": 793
+    },
+    {
+      "epoch": 0.18045454545454545,
+      "grad_norm": 3.506011025397826,
+      "learning_rate": 1.1524657068006622e-06,
+      "loss": 0.0851,
+      "step": 794
+    },
+    {
+      "epoch": 0.1806818181818182,
+      "grad_norm": 1.7827392041353078,
+      "learning_rate": 1.1522261912639266e-06,
+      "loss": 0.0586,
+      "step": 795
+    },
+    {
+      "epoch": 0.1809090909090909,
+      "grad_norm": 2.6359555068261318,
+      "learning_rate": 1.151986406950656e-06,
+      "loss": 0.0985,
+      "step": 796
+    },
+    {
+      "epoch": 0.18113636363636362,
+      "grad_norm": 3.131388496467777,
+      "learning_rate": 1.1517463539830908e-06,
+      "loss": 0.1576,
+      "step": 797
+    },
+    {
+      "epoch": 0.18136363636363637,
+      "grad_norm": 2.6525424346183093,
+      "learning_rate": 1.1515060324836088e-06,
+      "loss": 0.0994,
+      "step": 798
+    },
+    {
+      "epoch": 0.18159090909090908,
+      "grad_norm": 1.071077855106917,
+      "learning_rate": 1.151265442574724e-06,
+      "loss": 0.066,
+      "step": 799
+    },
+    {
+      "epoch": 0.18181818181818182,
+      "grad_norm": 5.438824680739619,
+      "learning_rate": 1.151024584379088e-06,
+      "loss": 0.1352,
+      "step": 800
+    },
+    {
+      "epoch": 0.18204545454545454,
+      "grad_norm": 2.1428067404342173,
+      "learning_rate": 1.1507834580194883e-06,
+      "loss": 0.0991,
+      "step": 801
+    },
+    {
+      "epoch": 0.18227272727272728,
+      "grad_norm": 2.584642668107492,
+      "learning_rate": 1.15054206361885e-06,
+      "loss": 0.0615,
+      "step": 802
+    },
+    {
+      "epoch": 0.1825,
+      "grad_norm": 3.139534261632662,
+      "learning_rate": 1.1503004013002341e-06,
+      "loss": 0.0801,
+      "step": 803
+    },
+    {
+      "epoch": 0.18272727272727274,
+      "grad_norm": 4.277335163917085,
+      "learning_rate": 1.1500584711868388e-06,
+      "loss": 0.0937,
+      "step": 804
+    },
+    {
+      "epoch": 0.18295454545454545,
+      "grad_norm": 3.3885257254631673,
+      "learning_rate": 1.1498162734019983e-06,
+      "loss": 0.0878,
+      "step": 805
+    },
+    {
+      "epoch": 0.1831818181818182,
+      "grad_norm": 3.299063354174504,
+      "learning_rate": 1.1495738080691834e-06,
+      "loss": 0.1017,
+      "step": 806
+    },
+    {
+      "epoch": 0.1834090909090909,
+      "grad_norm": 1.5696364401183154,
+      "learning_rate": 1.1493310753120017e-06,
+      "loss": 0.067,
+      "step": 807
+    },
+    {
+      "epoch": 0.18363636363636363,
+      "grad_norm": 3.2180129149672325,
+      "learning_rate": 1.1490880752541967e-06,
+      "loss": 0.1305,
+      "step": 808
+    },
+    {
+      "epoch": 0.18386363636363637,
+      "grad_norm": 2.6011996481905824,
+      "learning_rate": 1.148844808019648e-06,
+      "loss": 0.1157,
+      "step": 809
+    },
+    {
+      "epoch": 0.18409090909090908,
+      "grad_norm": 3.4288903997019036,
+      "learning_rate": 1.1486012737323718e-06,
+      "loss": 0.0795,
+      "step": 810
+    },
+    {
+      "epoch": 0.18431818181818183,
+      "grad_norm": 2.681352230471287,
+      "learning_rate": 1.1483574725165207e-06,
+      "loss": 0.1187,
+      "step": 811
+    },
+    {
+      "epoch": 0.18454545454545454,
+      "grad_norm": 1.44422465743348,
+      "learning_rate": 1.1481134044963825e-06,
+      "loss": 0.0535,
+      "step": 812
+    },
+    {
+      "epoch": 0.18477272727272728,
+      "grad_norm": 2.2931329758990957,
+      "learning_rate": 1.1478690697963814e-06,
+      "loss": 0.0903,
+      "step": 813
+    },
+    {
+      "epoch": 0.185,
+      "grad_norm": 2.799907754892058,
+      "learning_rate": 1.1476244685410784e-06,
+      "loss": 0.0669,
+      "step": 814
+    },
+    {
+      "epoch": 0.18522727272727274,
+      "grad_norm": 3.5617296862325074,
+      "learning_rate": 1.147379600855169e-06,
+      "loss": 0.1832,
+      "step": 815
+    },
+    {
+      "epoch": 0.18545454545454546,
+      "grad_norm": 2.4880736924318243,
+      "learning_rate": 1.1471344668634854e-06,
+      "loss": 0.112,
+      "step": 816
+    },
+    {
+      "epoch": 0.18568181818181817,
+      "grad_norm": 2.4444501349390304,
+      "learning_rate": 1.1468890666909954e-06,
+      "loss": 0.1155,
+      "step": 817
+    },
+    {
+      "epoch": 0.1859090909090909,
+      "grad_norm": 3.813853261716941,
+      "learning_rate": 1.1466434004628023e-06,
+      "loss": 0.1429,
+      "step": 818
+    },
+    {
+      "epoch": 0.18613636363636363,
+      "grad_norm": 5.421205135165792,
+      "learning_rate": 1.1463974683041455e-06,
+      "loss": 0.2755,
+      "step": 819
+    },
+    {
+      "epoch": 0.18636363636363637,
+      "grad_norm": 3.609600274274101,
+      "learning_rate": 1.1461512703403992e-06,
+      "loss": 0.1184,
+      "step": 820
+    },
+    {
+      "epoch": 0.18659090909090909,
+      "grad_norm": 1.8615631846224745,
+      "learning_rate": 1.1459048066970736e-06,
+      "loss": 0.0818,
+      "step": 821
+    },
+    {
+      "epoch": 0.18681818181818183,
+      "grad_norm": 2.4289052703794325,
+      "learning_rate": 1.1456580774998146e-06,
+      "loss": 0.0764,
+      "step": 822
+    },
+    {
+      "epoch": 0.18704545454545454,
+      "grad_norm": 2.713828246740441,
+      "learning_rate": 1.1454110828744027e-06,
+      "loss": 0.0912,
+      "step": 823
+    },
+    {
+      "epoch": 0.18727272727272729,
+      "grad_norm": 2.8685763209763637,
+      "learning_rate": 1.1451638229467547e-06,
+      "loss": 0.0738,
+      "step": 824
+    },
+    {
+      "epoch": 0.1875,
+      "grad_norm": 3.3705690705593923,
+      "learning_rate": 1.1449162978429218e-06,
+      "loss": 0.0788,
+      "step": 825
+    },
+    {
+      "epoch": 0.18772727272727271,
+      "grad_norm": 3.254820549961461,
+      "learning_rate": 1.144668507689091e-06,
+      "loss": 0.0681,
+      "step": 826
+    },
+    {
+      "epoch": 0.18795454545454546,
+      "grad_norm": 2.1933541978238735,
+      "learning_rate": 1.1444204526115837e-06,
+      "loss": 0.0646,
+      "step": 827
+    },
+    {
+      "epoch": 0.18818181818181817,
+      "grad_norm": 4.211851612231775,
+      "learning_rate": 1.144172132736857e-06,
+      "loss": 0.1317,
+      "step": 828
+    },
+    {
+      "epoch": 0.18840909090909091,
+      "grad_norm": 3.8238801846582278,
+      "learning_rate": 1.1439235481915028e-06,
+      "loss": 0.0938,
+      "step": 829
+    },
+    {
+      "epoch": 0.18863636363636363,
+      "grad_norm": 8.657468174143784,
+      "learning_rate": 1.1436746991022479e-06,
+      "loss": 0.1045,
+      "step": 830
+    },
+    {
+      "epoch": 0.18886363636363637,
+      "grad_norm": 3.10943093139228,
+      "learning_rate": 1.143425585595954e-06,
+      "loss": 0.1194,
+      "step": 831
+    },
+    {
+      "epoch": 0.1890909090909091,
+      "grad_norm": 2.3018430914505403,
+      "learning_rate": 1.1431762077996174e-06,
+      "loss": 0.1219,
+      "step": 832
+    },
+    {
+      "epoch": 0.18931818181818183,
+      "grad_norm": 4.427509015662018,
+      "learning_rate": 1.1429265658403698e-06,
+      "loss": 0.0977,
+      "step": 833
+    },
+    {
+      "epoch": 0.18954545454545454,
+      "grad_norm": 3.684326637673178,
+      "learning_rate": 1.1426766598454768e-06,
+      "loss": 0.0846,
+      "step": 834
+    },
+    {
+      "epoch": 0.18977272727272726,
+      "grad_norm": 5.163030810979085,
+      "learning_rate": 1.1424264899423383e-06,
+      "loss": 0.1853,
+      "step": 835
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.3060632481428494,
+      "learning_rate": 1.1421760562584901e-06,
+      "loss": 0.1056,
+      "step": 836
+    },
+    {
+      "epoch": 0.19022727272727272,
+      "grad_norm": 3.9361669806006496,
+      "learning_rate": 1.1419253589216012e-06,
+      "loss": 0.2202,
+      "step": 837
+    },
+    {
+      "epoch": 0.19045454545454546,
+      "grad_norm": 2.8331931002753183,
+      "learning_rate": 1.1416743980594758e-06,
+      "loss": 0.0619,
+      "step": 838
+    },
+    {
+      "epoch": 0.19068181818181817,
+      "grad_norm": 4.096429435206798,
+      "learning_rate": 1.1414231738000521e-06,
+      "loss": 0.1873,
+      "step": 839
+    },
+    {
+      "epoch": 0.19090909090909092,
+      "grad_norm": 2.16119928523465,
+      "learning_rate": 1.1411716862714027e-06,
+      "loss": 0.0545,
+      "step": 840
+    },
+    {
+      "epoch": 0.19113636363636363,
+      "grad_norm": 5.550131065652608,
+      "learning_rate": 1.1409199356017339e-06,
+      "loss": 0.1225,
+      "step": 841
+    },
+    {
+      "epoch": 0.19136363636363637,
+      "grad_norm": 2.4101277608947926,
+      "learning_rate": 1.140667921919387e-06,
+      "loss": 0.1466,
+      "step": 842
+    },
+    {
+      "epoch": 0.1915909090909091,
+      "grad_norm": 3.795644548891534,
+      "learning_rate": 1.1404156453528367e-06,
+      "loss": 0.1144,
+      "step": 843
+    },
+    {
+      "epoch": 0.1918181818181818,
+      "grad_norm": 2.8182678499370364,
+      "learning_rate": 1.1401631060306921e-06,
+      "loss": 0.0598,
+      "step": 844
+    },
+    {
+      "epoch": 0.19204545454545455,
+      "grad_norm": 2.6353881625700155,
+      "learning_rate": 1.1399103040816963e-06,
+      "loss": 0.0938,
+      "step": 845
+    },
+    {
+      "epoch": 0.19227272727272726,
+      "grad_norm": 3.6059390754335956,
+      "learning_rate": 1.1396572396347257e-06,
+      "loss": 0.162,
+      "step": 846
+    },
+    {
+      "epoch": 0.1925,
+      "grad_norm": 2.8364878586325517,
+      "learning_rate": 1.1394039128187914e-06,
+      "loss": 0.1038,
+      "step": 847
+    },
+    {
+      "epoch": 0.19272727272727272,
+      "grad_norm": 3.6802409394806546,
+      "learning_rate": 1.1391503237630375e-06,
+      "loss": 0.1348,
+      "step": 848
+    },
+    {
+      "epoch": 0.19295454545454546,
+      "grad_norm": 1.838195768985246,
+      "learning_rate": 1.1388964725967423e-06,
+      "loss": 0.0659,
+      "step": 849
+    },
+    {
+      "epoch": 0.19318181818181818,
+      "grad_norm": 2.1533771310840946,
+      "learning_rate": 1.138642359449317e-06,
+      "loss": 0.0548,
+      "step": 850
+    },
+    {
+      "epoch": 0.19340909090909092,
+      "grad_norm": 2.1550316935790157,
+      "learning_rate": 1.1383879844503073e-06,
+      "loss": 0.0543,
+      "step": 851
+    },
+    {
+      "epoch": 0.19363636363636363,
+      "grad_norm": 3.316083219459905,
+      "learning_rate": 1.1381333477293918e-06,
+      "loss": 0.1225,
+      "step": 852
+    },
+    {
+      "epoch": 0.19386363636363638,
+      "grad_norm": 2.2512138983381855,
+      "learning_rate": 1.137878449416383e-06,
+      "loss": 0.074,
+      "step": 853
+    },
+    {
+      "epoch": 0.1940909090909091,
+      "grad_norm": 1.960701390128428,
+      "learning_rate": 1.137623289641226e-06,
+      "loss": 0.0941,
+      "step": 854
+    },
+    {
+      "epoch": 0.1943181818181818,
+      "grad_norm": 3.3429001180943536,
+      "learning_rate": 1.1373678685339994e-06,
+      "loss": 0.1151,
+      "step": 855
+    },
+    {
+      "epoch": 0.19454545454545455,
+      "grad_norm": 1.4766082260499962,
+      "learning_rate": 1.137112186224916e-06,
+      "loss": 0.0827,
+      "step": 856
+    },
+    {
+      "epoch": 0.19477272727272726,
+      "grad_norm": 5.0914055375258584,
+      "learning_rate": 1.1368562428443205e-06,
+      "loss": 0.1422,
+      "step": 857
+    },
+    {
+      "epoch": 0.195,
+      "grad_norm": 3.6649189418484878,
+      "learning_rate": 1.1366000385226913e-06,
+      "loss": 0.0883,
+      "step": 858
+    },
+    {
+      "epoch": 0.19522727272727272,
+      "grad_norm": 3.0223492961070133,
+      "learning_rate": 1.1363435733906398e-06,
+      "loss": 0.0727,
+      "step": 859
+    },
+    {
+      "epoch": 0.19545454545454546,
+      "grad_norm": 4.775911421467254,
+      "learning_rate": 1.13608684757891e-06,
+      "loss": 0.1403,
+      "step": 860
+    },
+    {
+      "epoch": 0.19568181818181818,
+      "grad_norm": 1.8306464787393781,
+      "learning_rate": 1.1358298612183793e-06,
+      "loss": 0.0519,
+      "step": 861
+    },
+    {
+      "epoch": 0.19590909090909092,
+      "grad_norm": 2.560261151483981,
+      "learning_rate": 1.135572614440058e-06,
+      "loss": 0.1326,
+      "step": 862
+    },
+    {
+      "epoch": 0.19613636363636364,
+      "grad_norm": 4.361080293809697,
+      "learning_rate": 1.1353151073750882e-06,
+      "loss": 0.1122,
+      "step": 863
+    },
+    {
+      "epoch": 0.19636363636363635,
+      "grad_norm": 5.780141713415176,
+      "learning_rate": 1.1350573401547457e-06,
+      "loss": 0.1673,
+      "step": 864
+    },
+    {
+      "epoch": 0.1965909090909091,
+      "grad_norm": 5.839517715657577,
+      "learning_rate": 1.1347993129104386e-06,
+      "loss": 0.1308,
+      "step": 865
+    },
+    {
+      "epoch": 0.1968181818181818,
+      "grad_norm": 5.86249892283722,
+      "learning_rate": 1.1345410257737078e-06,
+      "loss": 0.1348,
+      "step": 866
+    },
+    {
+      "epoch": 0.19704545454545455,
+      "grad_norm": 2.4773347048259873,
+      "learning_rate": 1.1342824788762258e-06,
+      "loss": 0.086,
+      "step": 867
+    },
+    {
+      "epoch": 0.19727272727272727,
+      "grad_norm": 2.1062215228256287,
+      "learning_rate": 1.1340236723497985e-06,
+      "loss": 0.0828,
+      "step": 868
+    },
+    {
+      "epoch": 0.1975,
+      "grad_norm": 2.322401525588964,
+      "learning_rate": 1.133764606326364e-06,
+      "loss": 0.0634,
+      "step": 869
+    },
+    {
+      "epoch": 0.19772727272727272,
+      "grad_norm": 3.0304941191038757,
+      "learning_rate": 1.1335052809379921e-06,
+      "loss": 0.0856,
+      "step": 870
+    },
+    {
+      "epoch": 0.19795454545454547,
+      "grad_norm": 2.826869993927377,
+      "learning_rate": 1.1332456963168854e-06,
+      "loss": 0.1027,
+      "step": 871
+    },
+    {
+      "epoch": 0.19818181818181818,
+      "grad_norm": 3.894930692754439,
+      "learning_rate": 1.1329858525953785e-06,
+      "loss": 0.1637,
+      "step": 872
+    },
+    {
+      "epoch": 0.1984090909090909,
+      "grad_norm": 3.554701139789984,
+      "learning_rate": 1.132725749905938e-06,
+      "loss": 0.0967,
+      "step": 873
+    },
+    {
+      "epoch": 0.19863636363636364,
+      "grad_norm": 4.171259007668478,
+      "learning_rate": 1.132465388381163e-06,
+      "loss": 0.1455,
+      "step": 874
+    },
+    {
+      "epoch": 0.19886363636363635,
+      "grad_norm": 3.9600508087933113,
+      "learning_rate": 1.1322047681537834e-06,
+      "loss": 0.0719,
+      "step": 875
+    },
+    {
+      "epoch": 0.1990909090909091,
+      "grad_norm": 5.976129345335904,
+      "learning_rate": 1.131943889356662e-06,
+      "loss": 0.1777,
+      "step": 876
+    },
+    {
+      "epoch": 0.1993181818181818,
+      "grad_norm": 2.275263782788514,
+      "learning_rate": 1.1316827521227935e-06,
+      "loss": 0.0681,
+      "step": 877
+    },
+    {
+      "epoch": 0.19954545454545455,
+      "grad_norm": 3.5336963821913736,
+      "learning_rate": 1.1314213565853036e-06,
+      "loss": 0.0928,
+      "step": 878
+    },
+    {
+      "epoch": 0.19977272727272727,
+      "grad_norm": 3.3253983060704693,
+      "learning_rate": 1.1311597028774503e-06,
+      "loss": 0.1077,
+      "step": 879
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.408373012497816,
+      "learning_rate": 1.1308977911326229e-06,
+      "loss": 0.1048,
+      "step": 880
+    },
+    {
+      "epoch": 0.20022727272727273,
+      "grad_norm": 1.1404142454775106,
+      "learning_rate": 1.1306356214843423e-06,
+      "loss": 0.0502,
+      "step": 881
+    },
+    {
+      "epoch": 0.20045454545454544,
+      "grad_norm": 2.7407103948086107,
+      "learning_rate": 1.1303731940662608e-06,
+      "loss": 0.0768,
+      "step": 882
+    },
+    {
+      "epoch": 0.20068181818181818,
+      "grad_norm": 2.373689445874234,
+      "learning_rate": 1.1301105090121624e-06,
+      "loss": 0.1118,
+      "step": 883
+    },
+    {
+      "epoch": 0.2009090909090909,
+      "grad_norm": 2.523364804875991,
+      "learning_rate": 1.1298475664559622e-06,
+      "loss": 0.0987,
+      "step": 884
+    },
+    {
+      "epoch": 0.20113636363636364,
+      "grad_norm": 2.917457869309103,
+      "learning_rate": 1.1295843665317067e-06,
+      "loss": 0.0812,
+      "step": 885
+    },
+    {
+      "epoch": 0.20136363636363636,
+      "grad_norm": 2.91517100930304,
+      "learning_rate": 1.1293209093735732e-06,
+      "loss": 0.0878,
+      "step": 886
+    },
+    {
+      "epoch": 0.2015909090909091,
+      "grad_norm": 3.307951197493292,
+      "learning_rate": 1.129057195115871e-06,
+      "loss": 0.0955,
+      "step": 887
+    },
+    {
+      "epoch": 0.2018181818181818,
+      "grad_norm": 3.2338834501345723,
+      "learning_rate": 1.1287932238930397e-06,
+      "loss": 0.095,
+      "step": 888
+    },
+    {
+      "epoch": 0.20204545454545456,
+      "grad_norm": 2.0160635713962636,
+      "learning_rate": 1.12852899583965e-06,
+      "loss": 0.0944,
+      "step": 889
+    },
+    {
+      "epoch": 0.20227272727272727,
+      "grad_norm": 2.0417187168564435,
+      "learning_rate": 1.1282645110904036e-06,
+      "loss": 0.0848,
+      "step": 890
+    },
+    {
+      "epoch": 0.2025,
+      "grad_norm": 2.068173970557142,
+      "learning_rate": 1.1279997697801334e-06,
+      "loss": 0.0877,
+      "step": 891
+    },
+    {
+      "epoch": 0.20272727272727273,
+      "grad_norm": 1.5156010004576486,
+      "learning_rate": 1.1277347720438028e-06,
+      "loss": 0.0769,
+      "step": 892
+    },
+    {
+      "epoch": 0.20295454545454544,
+      "grad_norm": 2.080558521114398,
+      "learning_rate": 1.127469518016506e-06,
+      "loss": 0.0711,
+      "step": 893
+    },
+    {
+      "epoch": 0.20318181818181819,
+      "grad_norm": 2.7548962731781175,
+      "learning_rate": 1.1272040078334675e-06,
+      "loss": 0.1318,
+      "step": 894
+    },
+    {
+      "epoch": 0.2034090909090909,
+      "grad_norm": 2.166531975684643,
+      "learning_rate": 1.1269382416300429e-06,
+      "loss": 0.0744,
+      "step": 895
+    },
+    {
+      "epoch": 0.20363636363636364,
+      "grad_norm": 4.859934246879939,
+      "learning_rate": 1.126672219541718e-06,
+      "loss": 0.1879,
+      "step": 896
+    },
+    {
+      "epoch": 0.20386363636363636,
+      "grad_norm": 3.165261085320356,
+      "learning_rate": 1.1264059417041095e-06,
+      "loss": 0.1116,
+      "step": 897
+    },
+    {
+      "epoch": 0.2040909090909091,
+      "grad_norm": 2.581609984811572,
+      "learning_rate": 1.1261394082529638e-06,
+      "loss": 0.0747,
+      "step": 898
+    },
+    {
+      "epoch": 0.20431818181818182,
+      "grad_norm": 7.595407532118096,
+      "learning_rate": 1.1258726193241578e-06,
+      "loss": 0.1331,
+      "step": 899
+    },
+    {
+      "epoch": 0.20454545454545456,
+      "grad_norm": 2.460239281518681,
+      "learning_rate": 1.1256055750536992e-06,
+      "loss": 0.0569,
+      "step": 900
+    },
+    {
+      "epoch": 0.20477272727272727,
+      "grad_norm": 3.9890344248979335,
+      "learning_rate": 1.1253382755777253e-06,
+      "loss": 0.1031,
+      "step": 901
+    },
+    {
+      "epoch": 0.205,
+      "grad_norm": 2.303899054124794,
+      "learning_rate": 1.1250707210325036e-06,
+      "loss": 0.1064,
+      "step": 902
+    },
+    {
+      "epoch": 0.20522727272727273,
+      "grad_norm": 3.7644569912194763,
+      "learning_rate": 1.1248029115544319e-06,
+      "loss": 0.1531,
+      "step": 903
+    },
+    {
+      "epoch": 0.20545454545454545,
+      "grad_norm": 2.361898802494317,
+      "learning_rate": 1.1245348472800372e-06,
+      "loss": 0.1254,
+      "step": 904
+    },
+    {
+      "epoch": 0.2056818181818182,
+      "grad_norm": 3.2077710240271395,
+      "learning_rate": 1.1242665283459775e-06,
+      "loss": 0.1178,
+      "step": 905
+    },
+    {
+      "epoch": 0.2059090909090909,
+      "grad_norm": 4.5728122439497465,
+      "learning_rate": 1.1239979548890398e-06,
+      "loss": 0.1427,
+      "step": 906
+    },
+    {
+      "epoch": 0.20613636363636365,
+      "grad_norm": 2.143135780195684,
+      "learning_rate": 1.1237291270461413e-06,
+      "loss": 0.0906,
+      "step": 907
+    },
+    {
+      "epoch": 0.20636363636363636,
+      "grad_norm": 2.377923485082702,
+      "learning_rate": 1.1234600449543288e-06,
+      "loss": 0.0928,
+      "step": 908
+    },
+    {
+      "epoch": 0.2065909090909091,
+      "grad_norm": 3.95466741020973,
+      "learning_rate": 1.123190708750778e-06,
+      "loss": 0.1844,
+      "step": 909
+    },
+    {
+      "epoch": 0.20681818181818182,
+      "grad_norm": 3.6946488094845544,
+      "learning_rate": 1.1229211185727957e-06,
+      "loss": 0.1023,
+      "step": 910
+    },
+    {
+      "epoch": 0.20704545454545453,
+      "grad_norm": 4.3259911216107705,
+      "learning_rate": 1.1226512745578166e-06,
+      "loss": 0.061,
+      "step": 911
+    },
+    {
+      "epoch": 0.20727272727272728,
+      "grad_norm": 4.401604638829085,
+      "learning_rate": 1.1223811768434057e-06,
+      "loss": 0.1589,
+      "step": 912
+    },
+    {
+      "epoch": 0.2075,
+      "grad_norm": 1.6671112651180322,
+      "learning_rate": 1.122110825567257e-06,
+      "loss": 0.0733,
+      "step": 913
+    },
+    {
+      "epoch": 0.20772727272727273,
+      "grad_norm": 3.6022673624665527,
+      "learning_rate": 1.1218402208671938e-06,
+      "loss": 0.0915,
+      "step": 914
+    },
+    {
+      "epoch": 0.20795454545454545,
+      "grad_norm": 3.7529649858485694,
+      "learning_rate": 1.1215693628811688e-06,
+      "loss": 0.1332,
+      "step": 915
+    },
+    {
+      "epoch": 0.2081818181818182,
+      "grad_norm": 3.4313784464106734,
+      "learning_rate": 1.1212982517472636e-06,
+      "loss": 0.1177,
+      "step": 916
+    },
+    {
+      "epoch": 0.2084090909090909,
+      "grad_norm": 3.204234481204852,
+      "learning_rate": 1.1210268876036888e-06,
+      "loss": 0.071,
+      "step": 917
+    },
+    {
+      "epoch": 0.20863636363636365,
+      "grad_norm": 5.550411836893469,
+      "learning_rate": 1.1207552705887841e-06,
+      "loss": 0.1424,
+      "step": 918
+    },
+    {
+      "epoch": 0.20886363636363636,
+      "grad_norm": 3.796166977580573,
+      "learning_rate": 1.1204834008410184e-06,
+      "loss": 0.1275,
+      "step": 919
+    },
+    {
+      "epoch": 0.20909090909090908,
+      "grad_norm": 4.448753427467119,
+      "learning_rate": 1.1202112784989891e-06,
+      "loss": 0.1215,
+      "step": 920
+    },
+    {
+      "epoch": 0.20931818181818182,
+      "grad_norm": 3.603697902104071,
+      "learning_rate": 1.1199389037014221e-06,
+      "loss": 0.1349,
+      "step": 921
+    },
+    {
+      "epoch": 0.20954545454545453,
+      "grad_norm": 2.6656067554834175,
+      "learning_rate": 1.1196662765871725e-06,
+      "loss": 0.0758,
+      "step": 922
+    },
+    {
+      "epoch": 0.20977272727272728,
+      "grad_norm": 2.522427090552048,
+      "learning_rate": 1.119393397295224e-06,
+      "loss": 0.1295,
+      "step": 923
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.5500355457890977,
+      "learning_rate": 1.1191202659646883e-06,
+      "loss": 0.1163,
+      "step": 924
+    },
+    {
+      "epoch": 0.21022727272727273,
+      "grad_norm": 3.314367017367459,
+      "learning_rate": 1.1188468827348066e-06,
+      "loss": 0.2032,
+      "step": 925
+    },
+    {
+      "epoch": 0.21045454545454545,
+      "grad_norm": 3.1745714374101732,
+      "learning_rate": 1.1185732477449475e-06,
+      "loss": 0.0621,
+      "step": 926
+    },
+    {
+      "epoch": 0.2106818181818182,
+      "grad_norm": 2.6629434499497235,
+      "learning_rate": 1.1182993611346084e-06,
+      "loss": 0.0951,
+      "step": 927
+    },
+    {
+      "epoch": 0.2109090909090909,
+      "grad_norm": 2.109660815852733,
+      "learning_rate": 1.1180252230434151e-06,
+      "loss": 0.096,
+      "step": 928
+    },
+    {
+      "epoch": 0.21113636363636365,
+      "grad_norm": 2.3764709838216462,
+      "learning_rate": 1.1177508336111215e-06,
+      "loss": 0.081,
+      "step": 929
+    },
+    {
+      "epoch": 0.21136363636363636,
+      "grad_norm": 4.423531101074121,
+      "learning_rate": 1.1174761929776094e-06,
+      "loss": 0.1349,
+      "step": 930
+    },
+    {
+      "epoch": 0.21159090909090908,
+      "grad_norm": 6.046686138198189,
+      "learning_rate": 1.1172013012828887e-06,
+      "loss": 0.214,
+      "step": 931
+    },
+    {
+      "epoch": 0.21181818181818182,
+      "grad_norm": 3.1481418952455025,
+      "learning_rate": 1.1169261586670976e-06,
+      "loss": 0.0763,
+      "step": 932
+    },
+    {
+      "epoch": 0.21204545454545454,
+      "grad_norm": 2.1896468943616814,
+      "learning_rate": 1.1166507652705021e-06,
+      "loss": 0.0732,
+      "step": 933
+    },
+    {
+      "epoch": 0.21227272727272728,
+      "grad_norm": 1.612292051343508,
+      "learning_rate": 1.1163751212334962e-06,
+      "loss": 0.0627,
+      "step": 934
+    },
+    {
+      "epoch": 0.2125,
+      "grad_norm": 3.317046111551942,
+      "learning_rate": 1.116099226696601e-06,
+      "loss": 0.086,
+      "step": 935
+    },
+    {
+      "epoch": 0.21272727272727274,
+      "grad_norm": 2.4966447432976633,
+      "learning_rate": 1.1158230818004656e-06,
+      "loss": 0.1291,
+      "step": 936
+    },
+    {
+      "epoch": 0.21295454545454545,
+      "grad_norm": 2.6833283960400474,
+      "learning_rate": 1.1155466866858677e-06,
+      "loss": 0.0992,
+      "step": 937
+    },
+    {
+      "epoch": 0.2131818181818182,
+      "grad_norm": 2.679797743777146,
+      "learning_rate": 1.1152700414937111e-06,
+      "loss": 0.1006,
+      "step": 938
+    },
+    {
+      "epoch": 0.2134090909090909,
+      "grad_norm": 2.4526717287017776,
+      "learning_rate": 1.114993146365028e-06,
+      "loss": 0.1083,
+      "step": 939
+    },
+    {
+      "epoch": 0.21363636363636362,
+      "grad_norm": 4.997802297801097,
+      "learning_rate": 1.1147160014409779e-06,
+      "loss": 0.148,
+      "step": 940
+    },
+    {
+      "epoch": 0.21386363636363637,
+      "grad_norm": 3.631739963959465,
+      "learning_rate": 1.1144386068628472e-06,
+      "loss": 0.0986,
+      "step": 941
+    },
+    {
+      "epoch": 0.21409090909090908,
+      "grad_norm": 4.673512937040969,
+      "learning_rate": 1.1141609627720501e-06,
+      "loss": 0.122,
+      "step": 942
+    },
+    {
+      "epoch": 0.21431818181818182,
+      "grad_norm": 6.356064996925455,
+      "learning_rate": 1.1138830693101277e-06,
+      "loss": 0.1312,
+      "step": 943
+    },
+    {
+      "epoch": 0.21454545454545454,
+      "grad_norm": 2.247979228235356,
+      "learning_rate": 1.1136049266187481e-06,
+      "loss": 0.0574,
+      "step": 944
+    },
+    {
+      "epoch": 0.21477272727272728,
+      "grad_norm": 2.4398860326798513,
+      "learning_rate": 1.113326534839707e-06,
+      "loss": 0.0656,
+      "step": 945
+    },
+    {
+      "epoch": 0.215,
+      "grad_norm": 2.944609555841561,
+      "learning_rate": 1.1130478941149268e-06,
+      "loss": 0.1414,
+      "step": 946
+    },
+    {
+      "epoch": 0.21522727272727274,
+      "grad_norm": 1.8780586828444517,
+      "learning_rate": 1.1127690045864561e-06,
+      "loss": 0.0642,
+      "step": 947
+    },
+    {
+      "epoch": 0.21545454545454545,
+      "grad_norm": 2.0624939432771487,
+      "learning_rate": 1.1124898663964718e-06,
+      "loss": 0.0558,
+      "step": 948
+    },
+    {
+      "epoch": 0.21568181818181817,
+      "grad_norm": 2.047645031880203,
+      "learning_rate": 1.1122104796872763e-06,
+      "loss": 0.0826,
+      "step": 949
+    },
+    {
+      "epoch": 0.2159090909090909,
+      "grad_norm": 2.4564226331161354,
+      "learning_rate": 1.1119308446012993e-06,
+      "loss": 0.1094,
+      "step": 950
+    },
+    {
+      "epoch": 0.21613636363636363,
+      "grad_norm": 2.9508930203500405,
+      "learning_rate": 1.111650961281097e-06,
+      "loss": 0.0956,
+      "step": 951
+    },
+    {
+      "epoch": 0.21636363636363637,
+      "grad_norm": 2.584735875741647,
+      "learning_rate": 1.111370829869352e-06,
+      "loss": 0.0672,
+      "step": 952
+    },
+    {
+      "epoch": 0.21659090909090908,
+      "grad_norm": 2.3544927742886177,
+      "learning_rate": 1.1110904505088738e-06,
+      "loss": 0.1015,
+      "step": 953
+    },
+    {
+      "epoch": 0.21681818181818183,
+      "grad_norm": 2.0923619070138004,
+      "learning_rate": 1.1108098233425977e-06,
+      "loss": 0.0789,
+      "step": 954
+    },
+    {
+      "epoch": 0.21704545454545454,
+      "grad_norm": 2.827344856451054,
+      "learning_rate": 1.1105289485135855e-06,
+      "loss": 0.1264,
+      "step": 955
+    },
+    {
+      "epoch": 0.21727272727272728,
+      "grad_norm": 2.7625200370114693,
+      "learning_rate": 1.1102478261650258e-06,
+      "loss": 0.1385,
+      "step": 956
+    },
+    {
+      "epoch": 0.2175,
+      "grad_norm": 4.273223319040278,
+      "learning_rate": 1.1099664564402327e-06,
+      "loss": 0.093,
+      "step": 957
+    },
+    {
+      "epoch": 0.2177272727272727,
+      "grad_norm": 3.957015148241719,
+      "learning_rate": 1.1096848394826467e-06,
+      "loss": 0.0937,
+      "step": 958
+    },
+    {
+      "epoch": 0.21795454545454546,
+      "grad_norm": 2.6039014203138406,
+      "learning_rate": 1.1094029754358343e-06,
+      "loss": 0.0893,
+      "step": 959
+    },
+    {
+      "epoch": 0.21818181818181817,
+      "grad_norm": 4.337016858549236,
+      "learning_rate": 1.1091208644434882e-06,
+      "loss": 0.0853,
+      "step": 960
+    },
+    {
+      "epoch": 0.2184090909090909,
+      "grad_norm": 2.7783412449963016,
+      "learning_rate": 1.1088385066494267e-06,
+      "loss": 0.1019,
+      "step": 961
+    },
+    {
+      "epoch": 0.21863636363636363,
+      "grad_norm": 4.567262705167418,
+      "learning_rate": 1.1085559021975937e-06,
+      "loss": 0.162,
+      "step": 962
+    },
+    {
+      "epoch": 0.21886363636363637,
+      "grad_norm": 2.7701678941373875,
+      "learning_rate": 1.1082730512320597e-06,
+      "loss": 0.0924,
+      "step": 963
+    },
+    {
+      "epoch": 0.2190909090909091,
+      "grad_norm": 4.266085656521103,
+      "learning_rate": 1.1079899538970196e-06,
+      "loss": 0.2061,
+      "step": 964
+    },
+    {
+      "epoch": 0.21931818181818183,
+      "grad_norm": 4.47817591540772,
+      "learning_rate": 1.1077066103367954e-06,
+      "loss": 0.1175,
+      "step": 965
+    },
+    {
+      "epoch": 0.21954545454545454,
+      "grad_norm": 3.883728443912027,
+      "learning_rate": 1.1074230206958332e-06,
+      "loss": 0.1186,
+      "step": 966
+    },
+    {
+      "epoch": 0.2197727272727273,
+      "grad_norm": 3.4387024504468067,
+      "learning_rate": 1.1071391851187052e-06,
+      "loss": 0.1316,
+      "step": 967
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.109298772287061,
+      "learning_rate": 1.1068551037501093e-06,
+      "loss": 0.2123,
+      "step": 968
+    },
+    {
+      "epoch": 0.22022727272727272,
+      "grad_norm": 7.9804556432768585,
+      "learning_rate": 1.1065707767348684e-06,
+      "loss": 0.2646,
+      "step": 969
+    },
+    {
+      "epoch": 0.22045454545454546,
+      "grad_norm": 3.2426825766170913,
+      "learning_rate": 1.1062862042179302e-06,
+      "loss": 0.0921,
+      "step": 970
+    },
+    {
+      "epoch": 0.22068181818181817,
+      "grad_norm": 2.654212119535768,
+      "learning_rate": 1.1060013863443683e-06,
+      "loss": 0.0832,
+      "step": 971
+    },
+    {
+      "epoch": 0.22090909090909092,
+      "grad_norm": 2.3292926977665838,
+      "learning_rate": 1.1057163232593808e-06,
+      "loss": 0.0583,
+      "step": 972
+    },
+    {
+      "epoch": 0.22113636363636363,
+      "grad_norm": 7.754932245788428,
+      "learning_rate": 1.1054310151082913e-06,
+      "loss": 0.3362,
+      "step": 973
+    },
+    {
+      "epoch": 0.22136363636363637,
+      "grad_norm": 1.848455410558129,
+      "learning_rate": 1.1051454620365475e-06,
+      "loss": 0.0732,
+      "step": 974
+    },
+    {
+      "epoch": 0.2215909090909091,
+      "grad_norm": 4.175374511241659,
+      "learning_rate": 1.1048596641897233e-06,
+      "loss": 0.1147,
+      "step": 975
+    },
+    {
+      "epoch": 0.22181818181818183,
+      "grad_norm": 2.775054721276452,
+      "learning_rate": 1.104573621713516e-06,
+      "loss": 0.2448,
+      "step": 976
+    },
+    {
+      "epoch": 0.22204545454545455,
+      "grad_norm": 3.0619541038648164,
+      "learning_rate": 1.1042873347537485e-06,
+      "loss": 0.0676,
+      "step": 977
+    },
+    {
+      "epoch": 0.22227272727272726,
+      "grad_norm": 2.3963200464235466,
+      "learning_rate": 1.1040008034563682e-06,
+      "loss": 0.0595,
+      "step": 978
+    },
+    {
+      "epoch": 0.2225,
+      "grad_norm": 2.724124197152781,
+      "learning_rate": 1.1037140279674468e-06,
+      "loss": 0.0919,
+      "step": 979
+    },
+    {
+      "epoch": 0.22272727272727272,
+      "grad_norm": 2.2618149150379936,
+      "learning_rate": 1.1034270084331803e-06,
+      "loss": 0.0934,
+      "step": 980
+    },
+    {
+      "epoch": 0.22295454545454546,
+      "grad_norm": 2.4538484431914247,
+      "learning_rate": 1.1031397449998896e-06,
+      "loss": 0.1762,
+      "step": 981
+    },
+    {
+      "epoch": 0.22318181818181818,
+      "grad_norm": 3.1096558602754,
+      "learning_rate": 1.10285223781402e-06,
+      "loss": 0.0903,
+      "step": 982
+    },
+    {
+      "epoch": 0.22340909090909092,
+      "grad_norm": 2.6710422364710147,
+      "learning_rate": 1.1025644870221405e-06,
+      "loss": 0.1155,
+      "step": 983
+    },
+    {
+      "epoch": 0.22363636363636363,
+      "grad_norm": 4.404314169230577,
+      "learning_rate": 1.1022764927709447e-06,
+      "loss": 0.0618,
+      "step": 984
+    },
+    {
+      "epoch": 0.22386363636363638,
+      "grad_norm": 3.1080682317581116,
+      "learning_rate": 1.1019882552072502e-06,
+      "loss": 0.1129,
+      "step": 985
+    },
+    {
+      "epoch": 0.2240909090909091,
+      "grad_norm": 3.4982391770158126,
+      "learning_rate": 1.101699774477999e-06,
+      "loss": 0.0943,
+      "step": 986
+    },
+    {
+      "epoch": 0.2243181818181818,
+      "grad_norm": 3.7455969246134586,
+      "learning_rate": 1.1014110507302563e-06,
+      "loss": 0.0829,
+      "step": 987
+    },
+    {
+      "epoch": 0.22454545454545455,
+      "grad_norm": 3.089772830261481,
+      "learning_rate": 1.1011220841112118e-06,
+      "loss": 0.0916,
+      "step": 988
+    },
+    {
+      "epoch": 0.22477272727272726,
+      "grad_norm": 3.685996926681144,
+      "learning_rate": 1.1008328747681788e-06,
+      "loss": 0.125,
+      "step": 989
+    },
+    {
+      "epoch": 0.225,
+      "grad_norm": 5.5029006850618245,
+      "learning_rate": 1.1005434228485945e-06,
+      "loss": 0.1287,
+      "step": 990
+    },
+    {
+      "epoch": 0.22522727272727272,
+      "grad_norm": 2.503227367309242,
+      "learning_rate": 1.1002537285000196e-06,
+      "loss": 0.074,
+      "step": 991
+    },
+    {
+      "epoch": 0.22545454545454546,
+      "grad_norm": 3.663678384190874,
+      "learning_rate": 1.099963791870138e-06,
+      "loss": 0.1679,
+      "step": 992
+    },
+    {
+      "epoch": 0.22568181818181818,
+      "grad_norm": 3.6473105532007906,
+      "learning_rate": 1.099673613106758e-06,
+      "loss": 0.1851,
+      "step": 993
+    },
+    {
+      "epoch": 0.22590909090909092,
+      "grad_norm": 3.376502978870639,
+      "learning_rate": 1.0993831923578107e-06,
+      "loss": 0.1261,
+      "step": 994
+    },
+    {
+      "epoch": 0.22613636363636364,
+      "grad_norm": 2.050111521397145,
+      "learning_rate": 1.099092529771351e-06,
+      "loss": 0.1158,
+      "step": 995
+    },
+    {
+      "epoch": 0.22636363636363635,
+      "grad_norm": 3.524690286226471,
+      "learning_rate": 1.0988016254955565e-06,
+      "loss": 0.079,
+      "step": 996
+    },
+    {
+      "epoch": 0.2265909090909091,
+      "grad_norm": 3.5070904166489933,
+      "learning_rate": 1.0985104796787285e-06,
+      "loss": 0.0929,
+      "step": 997
+    },
+    {
+      "epoch": 0.2268181818181818,
+      "grad_norm": 4.456391490112464,
+      "learning_rate": 1.0982190924692911e-06,
+      "loss": 0.1171,
+      "step": 998
+    },
+    {
+      "epoch": 0.22704545454545455,
+      "grad_norm": 2.9034448771170953,
+      "learning_rate": 1.0979274640157919e-06,
+      "loss": 0.0891,
+      "step": 999
+    },
+    {
+      "epoch": 0.22727272727272727,
+      "grad_norm": 2.317303423206001,
+      "learning_rate": 1.097635594466901e-06,
+      "loss": 0.0746,
+      "step": 1000
+    },
+    {
+      "epoch": 0.2275,
+      "grad_norm": 2.3898207317284226,
+      "learning_rate": 1.0973434839714116e-06,
+      "loss": 0.1337,
+      "step": 1001
+    },
+    {
+      "epoch": 0.22772727272727272,
+      "grad_norm": 4.5542207241038195,
+      "learning_rate": 1.0970511326782398e-06,
+      "loss": 0.1,
+      "step": 1002
+    },
+    {
+      "epoch": 0.22795454545454547,
+      "grad_norm": 2.5962593883287775,
+      "learning_rate": 1.0967585407364246e-06,
+      "loss": 0.1203,
+      "step": 1003
+    },
+    {
+      "epoch": 0.22818181818181818,
+      "grad_norm": 1.4589261347771723,
+      "learning_rate": 1.0964657082951274e-06,
+      "loss": 0.0855,
+      "step": 1004
+    },
+    {
+      "epoch": 0.22840909090909092,
+      "grad_norm": 4.005718643782005,
+      "learning_rate": 1.0961726355036324e-06,
+      "loss": 0.1141,
+      "step": 1005
+    },
+    {
+      "epoch": 0.22863636363636364,
+      "grad_norm": 3.6227520455795013,
+      "learning_rate": 1.0958793225113459e-06,
+      "loss": 0.2075,
+      "step": 1006
+    },
+    {
+      "epoch": 0.22886363636363635,
+      "grad_norm": 3.352441562049467,
+      "learning_rate": 1.0955857694677971e-06,
+      "loss": 0.1172,
+      "step": 1007
+    },
+    {
+      "epoch": 0.2290909090909091,
+      "grad_norm": 3.5771013991275,
+      "learning_rate": 1.0952919765226378e-06,
+      "loss": 0.1096,
+      "step": 1008
+    },
+    {
+      "epoch": 0.2293181818181818,
+      "grad_norm": 2.820476737214537,
+      "learning_rate": 1.0949979438256415e-06,
+      "loss": 0.1502,
+      "step": 1009
+    },
+    {
+      "epoch": 0.22954545454545455,
+      "grad_norm": 3.422637080217908,
+      "learning_rate": 1.0947036715267039e-06,
+      "loss": 0.079,
+      "step": 1010
+    },
+    {
+      "epoch": 0.22977272727272727,
+      "grad_norm": 2.7288425802149234,
+      "learning_rate": 1.0944091597758438e-06,
+      "loss": 0.087,
+      "step": 1011
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.431064862153476,
+      "learning_rate": 1.0941144087232008e-06,
+      "loss": 0.1365,
+      "step": 1012
+    },
+    {
+      "epoch": 0.23022727272727272,
+      "grad_norm": 3.5215545806808373,
+      "learning_rate": 1.0938194185190374e-06,
+      "loss": 0.0754,
+      "step": 1013
+    },
+    {
+      "epoch": 0.23045454545454547,
+      "grad_norm": 2.44728324136204,
+      "learning_rate": 1.0935241893137376e-06,
+      "loss": 0.0632,
+      "step": 1014
+    },
+    {
+      "epoch": 0.23068181818181818,
+      "grad_norm": 5.0499786599015835,
+      "learning_rate": 1.0932287212578075e-06,
+      "loss": 0.1379,
+      "step": 1015
+    },
+    {
+      "epoch": 0.2309090909090909,
+      "grad_norm": 3.5265344835599852,
+      "learning_rate": 1.0929330145018747e-06,
+      "loss": 0.0807,
+      "step": 1016
+    },
+    {
+      "epoch": 0.23113636363636364,
+      "grad_norm": 3.177449426677741,
+      "learning_rate": 1.0926370691966883e-06,
+      "loss": 0.1184,
+      "step": 1017
+    },
+    {
+      "epoch": 0.23136363636363635,
+      "grad_norm": 2.799778091924247,
+      "learning_rate": 1.0923408854931202e-06,
+      "loss": 0.0815,
+      "step": 1018
+    },
+    {
+      "epoch": 0.2315909090909091,
+      "grad_norm": 3.5022357220937517,
+      "learning_rate": 1.0920444635421622e-06,
+      "loss": 0.1677,
+      "step": 1019
+    },
+    {
+      "epoch": 0.2318181818181818,
+      "grad_norm": 2.409623073624271,
+      "learning_rate": 1.0917478034949285e-06,
+      "loss": 0.0712,
+      "step": 1020
+    },
+    {
+      "epoch": 0.23204545454545455,
+      "grad_norm": 2.825568763157153,
+      "learning_rate": 1.0914509055026545e-06,
+      "loss": 0.0888,
+      "step": 1021
+    },
+    {
+      "epoch": 0.23227272727272727,
+      "grad_norm": 2.446935650086231,
+      "learning_rate": 1.0911537697166967e-06,
+      "loss": 0.0814,
+      "step": 1022
+    },
+    {
+      "epoch": 0.2325,
+      "grad_norm": 3.209321781263389,
+      "learning_rate": 1.0908563962885337e-06,
+      "loss": 0.1928,
+      "step": 1023
+    },
+    {
+      "epoch": 0.23272727272727273,
+      "grad_norm": 3.1835675972038335,
+      "learning_rate": 1.0905587853697638e-06,
+      "loss": 0.1046,
+      "step": 1024
+    },
+    {
+      "epoch": 0.23295454545454544,
+      "grad_norm": 3.250106120972125,
+      "learning_rate": 1.0902609371121076e-06,
+      "loss": 0.112,
+      "step": 1025
+    },
+    {
+      "epoch": 0.23318181818181818,
+      "grad_norm": 2.5655596038091613,
+      "learning_rate": 1.089962851667406e-06,
+      "loss": 0.1314,
+      "step": 1026
+    },
+    {
+      "epoch": 0.2334090909090909,
+      "grad_norm": 3.0928004192484386,
+      "learning_rate": 1.089664529187621e-06,
+      "loss": 0.0922,
+      "step": 1027
+    },
+    {
+      "epoch": 0.23363636363636364,
+      "grad_norm": 3.3630779647276983,
+      "learning_rate": 1.0893659698248358e-06,
+      "loss": 0.1259,
+      "step": 1028
+    },
+    {
+      "epoch": 0.23386363636363636,
+      "grad_norm": 3.0639726100068145,
+      "learning_rate": 1.0890671737312538e-06,
+      "loss": 0.1063,
+      "step": 1029
+    },
+    {
+      "epoch": 0.2340909090909091,
+      "grad_norm": 1.928986513851524,
+      "learning_rate": 1.0887681410591994e-06,
+      "loss": 0.0987,
+      "step": 1030
+    },
+    {
+      "epoch": 0.23431818181818181,
+      "grad_norm": 3.3290930983318745,
+      "learning_rate": 1.0884688719611176e-06,
+      "loss": 0.0727,
+      "step": 1031
+    },
+    {
+      "epoch": 0.23454545454545456,
+      "grad_norm": 2.340235596266924,
+      "learning_rate": 1.0881693665895737e-06,
+      "loss": 0.0935,
+      "step": 1032
+    },
+    {
+      "epoch": 0.23477272727272727,
+      "grad_norm": 2.6097010578731443,
+      "learning_rate": 1.0878696250972536e-06,
+      "loss": 0.0742,
+      "step": 1033
+    },
+    {
+      "epoch": 0.235,
+      "grad_norm": 2.361225734558857,
+      "learning_rate": 1.087569647636964e-06,
+      "loss": 0.0858,
+      "step": 1034
+    },
+    {
+      "epoch": 0.23522727272727273,
+      "grad_norm": 2.0453254421623206,
+      "learning_rate": 1.0872694343616312e-06,
+      "loss": 0.066,
+      "step": 1035
+    },
+    {
+      "epoch": 0.23545454545454544,
+      "grad_norm": 5.213027010474639,
+      "learning_rate": 1.0869689854243019e-06,
+      "loss": 0.2446,
+      "step": 1036
+    },
+    {
+      "epoch": 0.2356818181818182,
+      "grad_norm": 1.1104112070804562,
+      "learning_rate": 1.0866683009781432e-06,
+      "loss": 0.068,
+      "step": 1037
+    },
+    {
+      "epoch": 0.2359090909090909,
+      "grad_norm": 4.448468896953293,
+      "learning_rate": 1.0863673811764419e-06,
+      "loss": 0.1695,
+      "step": 1038
+    },
+    {
+      "epoch": 0.23613636363636364,
+      "grad_norm": 1.9053478319797195,
+      "learning_rate": 1.0860662261726054e-06,
+      "loss": 0.0442,
+      "step": 1039
+    },
+    {
+      "epoch": 0.23636363636363636,
+      "grad_norm": 2.3912744153621093,
+      "learning_rate": 1.0857648361201603e-06,
+      "loss": 0.1416,
+      "step": 1040
+    },
+    {
+      "epoch": 0.2365909090909091,
+      "grad_norm": 6.142496178131957,
+      "learning_rate": 1.085463211172753e-06,
+      "loss": 0.1945,
+      "step": 1041
+    },
+    {
+      "epoch": 0.23681818181818182,
+      "grad_norm": 3.4672080243967924,
+      "learning_rate": 1.0851613514841508e-06,
+      "loss": 0.111,
+      "step": 1042
+    },
+    {
+      "epoch": 0.23704545454545456,
+      "grad_norm": 2.639038515311767,
+      "learning_rate": 1.0848592572082391e-06,
+      "loss": 0.0829,
+      "step": 1043
+    },
+    {
+      "epoch": 0.23727272727272727,
+      "grad_norm": 2.2188111340714682,
+      "learning_rate": 1.084556928499024e-06,
+      "loss": 0.0577,
+      "step": 1044
+    },
+    {
+      "epoch": 0.2375,
+      "grad_norm": 3.3735275365083957,
+      "learning_rate": 1.0842543655106305e-06,
+      "loss": 0.1386,
+      "step": 1045
+    },
+    {
+      "epoch": 0.23772727272727273,
+      "grad_norm": 2.8783031695202452,
+      "learning_rate": 1.0839515683973035e-06,
+      "loss": 0.094,
+      "step": 1046
+    },
+    {
+      "epoch": 0.23795454545454545,
+      "grad_norm": 2.4681897745323265,
+      "learning_rate": 1.083648537313407e-06,
+      "loss": 0.0631,
+      "step": 1047
+    },
+    {
+      "epoch": 0.2381818181818182,
+      "grad_norm": 3.5409382848693745,
+      "learning_rate": 1.083345272413424e-06,
+      "loss": 0.1112,
+      "step": 1048
+    },
+    {
+      "epoch": 0.2384090909090909,
+      "grad_norm": 3.431382674528093,
+      "learning_rate": 1.0830417738519575e-06,
+      "loss": 0.168,
+      "step": 1049
+    },
+    {
+      "epoch": 0.23863636363636365,
+      "grad_norm": 2.623520023331604,
+      "learning_rate": 1.0827380417837287e-06,
+      "loss": 0.0954,
+      "step": 1050
+    },
+    {
+      "epoch": 0.23886363636363636,
+      "grad_norm": 2.6900299562450614,
+      "learning_rate": 1.0824340763635785e-06,
+      "loss": 0.0637,
+      "step": 1051
+    },
+    {
+      "epoch": 0.2390909090909091,
+      "grad_norm": 3.3988135284447307,
+      "learning_rate": 1.0821298777464665e-06,
+      "loss": 0.1049,
+      "step": 1052
+    },
+    {
+      "epoch": 0.23931818181818182,
+      "grad_norm": 1.4388463452619982,
+      "learning_rate": 1.081825446087471e-06,
+      "loss": 0.0443,
+      "step": 1053
+    },
+    {
+      "epoch": 0.23954545454545453,
+      "grad_norm": 4.453600232712395,
+      "learning_rate": 1.0815207815417894e-06,
+      "loss": 0.1187,
+      "step": 1054
+    },
+    {
+      "epoch": 0.23977272727272728,
+      "grad_norm": 2.813752236794351,
+      "learning_rate": 1.081215884264738e-06,
+      "loss": 0.0641,
+      "step": 1055
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.5103600005398614,
+      "learning_rate": 1.0809107544117511e-06,
+      "loss": 0.1622,
+      "step": 1056
+    },
+    {
+      "epoch": 0.24022727272727273,
+      "grad_norm": 3.0614316400223434,
+      "learning_rate": 1.0806053921383823e-06,
+      "loss": 0.0805,
+      "step": 1057
+    },
+    {
+      "epoch": 0.24045454545454545,
+      "grad_norm": 4.599454155105708,
+      "learning_rate": 1.0802997976003031e-06,
+      "loss": 0.1113,
+      "step": 1058
+    },
+    {
+      "epoch": 0.2406818181818182,
+      "grad_norm": 2.946255517287608,
+      "learning_rate": 1.0799939709533036e-06,
+      "loss": 0.1171,
+      "step": 1059
+    },
+    {
+      "epoch": 0.2409090909090909,
+      "grad_norm": 2.313539654888859,
+      "learning_rate": 1.0796879123532924e-06,
+      "loss": 0.1487,
+      "step": 1060
+    },
+    {
+      "epoch": 0.24113636363636365,
+      "grad_norm": 2.5859760865279218,
+      "learning_rate": 1.0793816219562963e-06,
+      "loss": 0.0851,
+      "step": 1061
+    },
+    {
+      "epoch": 0.24136363636363636,
+      "grad_norm": 2.730372275838902,
+      "learning_rate": 1.0790750999184598e-06,
+      "loss": 0.1449,
+      "step": 1062
+    },
+    {
+      "epoch": 0.24159090909090908,
+      "grad_norm": 3.1106198042263378,
+      "learning_rate": 1.0787683463960462e-06,
+      "loss": 0.0985,
+      "step": 1063
+    },
+    {
+      "epoch": 0.24181818181818182,
+      "grad_norm": 2.4312389201931373,
+      "learning_rate": 1.0784613615454365e-06,
+      "loss": 0.0903,
+      "step": 1064
+    },
+    {
+      "epoch": 0.24204545454545454,
+      "grad_norm": 2.618279878946786,
+      "learning_rate": 1.0781541455231294e-06,
+      "loss": 0.1378,
+      "step": 1065
+    },
+    {
+      "epoch": 0.24227272727272728,
+      "grad_norm": 6.440838252004829,
+      "learning_rate": 1.077846698485742e-06,
+      "loss": 0.1975,
+      "step": 1066
+    },
+    {
+      "epoch": 0.2425,
+      "grad_norm": 2.8602768490988733,
+      "learning_rate": 1.0775390205900084e-06,
+      "loss": 0.0743,
+      "step": 1067
+    },
+    {
+      "epoch": 0.24272727272727274,
+      "grad_norm": 2.426755601088994,
+      "learning_rate": 1.0772311119927808e-06,
+      "loss": 0.0998,
+      "step": 1068
+    },
+    {
+      "epoch": 0.24295454545454545,
+      "grad_norm": 4.9494508884835,
+      "learning_rate": 1.0769229728510298e-06,
+      "loss": 0.1687,
+      "step": 1069
+    },
+    {
+      "epoch": 0.2431818181818182,
+      "grad_norm": 3.221663570718401,
+      "learning_rate": 1.0766146033218417e-06,
+      "loss": 0.2123,
+      "step": 1070
+    },
+    {
+      "epoch": 0.2434090909090909,
+      "grad_norm": 2.7965004443252273,
+      "learning_rate": 1.076306003562422e-06,
+      "loss": 0.1214,
+      "step": 1071
+    },
+    {
+      "epoch": 0.24363636363636362,
+      "grad_norm": 2.6322221145774916,
+      "learning_rate": 1.0759971737300928e-06,
+      "loss": 0.0976,
+      "step": 1072
+    },
+    {
+      "epoch": 0.24386363636363637,
+      "grad_norm": 2.6699556056360527,
+      "learning_rate": 1.0756881139822934e-06,
+      "loss": 0.0987,
+      "step": 1073
+    },
+    {
+      "epoch": 0.24409090909090908,
+      "grad_norm": 1.7075151884408495,
+      "learning_rate": 1.0753788244765805e-06,
+      "loss": 0.0704,
+      "step": 1074
+    },
+    {
+      "epoch": 0.24431818181818182,
+      "grad_norm": 3.713544106708803,
+      "learning_rate": 1.0750693053706282e-06,
+      "loss": 0.1216,
+      "step": 1075
+    },
+    {
+      "epoch": 0.24454545454545454,
+      "grad_norm": 2.839568732276037,
+      "learning_rate": 1.0747595568222268e-06,
+      "loss": 0.1676,
+      "step": 1076
+    },
+    {
+      "epoch": 0.24477272727272728,
+      "grad_norm": 5.436580193861712,
+      "learning_rate": 1.0744495789892848e-06,
+      "loss": 0.1536,
+      "step": 1077
+    },
+    {
+      "epoch": 0.245,
+      "grad_norm": 2.166952315452957,
+      "learning_rate": 1.0741393720298263e-06,
+      "loss": 0.0918,
+      "step": 1078
+    },
+    {
+      "epoch": 0.24522727272727274,
+      "grad_norm": 2.2368545994346896,
+      "learning_rate": 1.073828936101993e-06,
+      "loss": 0.1229,
+      "step": 1079
+    },
+    {
+      "epoch": 0.24545454545454545,
+      "grad_norm": 2.856030235398001,
+      "learning_rate": 1.0735182713640436e-06,
+      "loss": 0.1058,
+      "step": 1080
+    },
+    {
+      "epoch": 0.2456818181818182,
+      "grad_norm": 3.705020687317192,
+      "learning_rate": 1.0732073779743523e-06,
+      "loss": 0.1313,
+      "step": 1081
+    },
+    {
+      "epoch": 0.2459090909090909,
+      "grad_norm": 3.369565605292207,
+      "learning_rate": 1.0728962560914108e-06,
+      "loss": 0.1278,
+      "step": 1082
+    },
+    {
+      "epoch": 0.24613636363636363,
+      "grad_norm": 3.2753023462687723,
+      "learning_rate": 1.0725849058738274e-06,
+      "loss": 0.0791,
+      "step": 1083
+    },
+    {
+      "epoch": 0.24636363636363637,
+      "grad_norm": 2.679693942002681,
+      "learning_rate": 1.0722733274803261e-06,
+      "loss": 0.1241,
+      "step": 1084
+    },
+    {
+      "epoch": 0.24659090909090908,
+      "grad_norm": 2.3801136174936413,
+      "learning_rate": 1.0719615210697476e-06,
+      "loss": 0.0766,
+      "step": 1085
+    },
+    {
+      "epoch": 0.24681818181818183,
+      "grad_norm": 2.449329195243545,
+      "learning_rate": 1.0716494868010488e-06,
+      "loss": 0.0994,
+      "step": 1086
+    },
+    {
+      "epoch": 0.24704545454545454,
+      "grad_norm": 3.017857292121828,
+      "learning_rate": 1.071337224833303e-06,
+      "loss": 0.092,
+      "step": 1087
+    },
+    {
+      "epoch": 0.24727272727272728,
+      "grad_norm": 4.834907623969626,
+      "learning_rate": 1.0710247353256988e-06,
+      "loss": 0.1069,
+      "step": 1088
+    },
+    {
+      "epoch": 0.2475,
+      "grad_norm": 2.817122076216917,
+      "learning_rate": 1.0707120184375422e-06,
+      "loss": 0.0953,
+      "step": 1089
+    },
+    {
+      "epoch": 0.24772727272727274,
+      "grad_norm": 2.6444140932984186,
+      "learning_rate": 1.0703990743282534e-06,
+      "loss": 0.0778,
+      "step": 1090
+    },
+    {
+      "epoch": 0.24795454545454546,
+      "grad_norm": 3.0131058641087263,
+      "learning_rate": 1.07008590315737e-06,
+      "loss": 0.1316,
+      "step": 1091
+    },
+    {
+      "epoch": 0.24818181818181817,
+      "grad_norm": 2.650454667338594,
+      "learning_rate": 1.069772505084544e-06,
+      "loss": 0.1216,
+      "step": 1092
+    },
+    {
+      "epoch": 0.2484090909090909,
+      "grad_norm": 1.3364406391602255,
+      "learning_rate": 1.0694588802695443e-06,
+      "loss": 0.0565,
+      "step": 1093
+    },
+    {
+      "epoch": 0.24863636363636363,
+      "grad_norm": 3.0543730423049102,
+      "learning_rate": 1.0691450288722545e-06,
+      "loss": 0.1239,
+      "step": 1094
+    },
+    {
+      "epoch": 0.24886363636363637,
+      "grad_norm": 3.187648535671965,
+      "learning_rate": 1.0688309510526742e-06,
+      "loss": 0.0896,
+      "step": 1095
+    },
+    {
+      "epoch": 0.24909090909090909,
+      "grad_norm": 3.1151811182727296,
+      "learning_rate": 1.0685166469709181e-06,
+      "loss": 0.1363,
+      "step": 1096
+    },
+    {
+      "epoch": 0.24931818181818183,
+      "grad_norm": 2.882473555473901,
+      "learning_rate": 1.0682021167872166e-06,
+      "loss": 0.0681,
+      "step": 1097
+    },
+    {
+      "epoch": 0.24954545454545454,
+      "grad_norm": 4.8287176838478105,
+      "learning_rate": 1.0678873606619152e-06,
+      "loss": 0.2607,
+      "step": 1098
+    },
+    {
+      "epoch": 0.24977272727272729,
+      "grad_norm": 2.7638840844991512,
+      "learning_rate": 1.0675723787554743e-06,
+      "loss": 0.0877,
+      "step": 1099
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.966417813914996,
+      "learning_rate": 1.0672571712284697e-06,
+      "loss": 0.1078,
+      "step": 1100
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 4400,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3406754807808.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}