diff --git "a/epoch_4/checkpoint-2200/trainer_state.json" "b/epoch_4/checkpoint-2200/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/epoch_4/checkpoint-2200/trainer_state.json"
@@ -0,0 +1,15434 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.5,
+  "eval_steps": 500,
+  "global_step": 2200,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00022727272727272727,
+      "grad_norm": 3.9157455186995223,
+      "learning_rate": 1.25e-06,
+      "loss": 0.2159,
+      "step": 1
+    },
+    {
+      "epoch": 0.00045454545454545455,
+      "grad_norm": 4.056530013070645,
+      "learning_rate": 1.249999840689502e-06,
+      "loss": 0.1041,
+      "step": 2
+    },
+    {
+      "epoch": 0.0006818181818181819,
+      "grad_norm": 2.571759518347236,
+      "learning_rate": 1.2499993627580887e-06,
+      "loss": 0.1435,
+      "step": 3
+    },
+    {
+      "epoch": 0.0009090909090909091,
+      "grad_norm": 1.939416625317773,
+      "learning_rate": 1.2499985662060041e-06,
+      "loss": 0.1071,
+      "step": 4
+    },
+    {
+      "epoch": 0.0011363636363636363,
+      "grad_norm": 2.0127894256167225,
+      "learning_rate": 1.249997451033654e-06,
+      "loss": 0.1227,
+      "step": 5
+    },
+    {
+      "epoch": 0.0013636363636363637,
+      "grad_norm": 3.3801466412554784,
+      "learning_rate": 1.249996017241607e-06,
+      "loss": 0.1622,
+      "step": 6
+    },
+    {
+      "epoch": 0.001590909090909091,
+      "grad_norm": 2.3630024979059554,
+      "learning_rate": 1.249994264830594e-06,
+      "loss": 0.1238,
+      "step": 7
+    },
+    {
+      "epoch": 0.0018181818181818182,
+      "grad_norm": 1.786049102461721,
+      "learning_rate": 1.2499921938015086e-06,
+      "loss": 0.0929,
+      "step": 8
+    },
+    {
+      "epoch": 0.0020454545454545456,
+      "grad_norm": 1.6006657864811813,
+      "learning_rate": 1.2499898041554066e-06,
+      "loss": 0.1246,
+      "step": 9
+    },
+    {
+      "epoch": 0.0022727272727272726,
+      "grad_norm": 2.8164510047912423,
+      "learning_rate": 1.2499870958935056e-06,
+      "loss": 0.1861,
+      "step": 10
+    },
+    {
+      "epoch": 0.0025,
+      "grad_norm": 2.721054285611732,
+      "learning_rate": 1.2499840690171872e-06,
+      "loss": 0.1713,
+      "step": 11
+    },
+    {
+      "epoch": 0.0027272727272727275,
+      "grad_norm": 4.294434358023365,
+      "learning_rate": 1.2499807235279937e-06,
+      "loss": 0.112,
+      "step": 12
+    },
+    {
+      "epoch": 0.0029545454545454545,
+      "grad_norm": 1.2091397483957185,
+      "learning_rate": 1.249977059427631e-06,
+      "loss": 0.1515,
+      "step": 13
+    },
+    {
+      "epoch": 0.003181818181818182,
+      "grad_norm": 2.100263003714764,
+      "learning_rate": 1.2499730767179668e-06,
+      "loss": 0.1808,
+      "step": 14
+    },
+    {
+      "epoch": 0.003409090909090909,
+      "grad_norm": 1.8513505934803485,
+      "learning_rate": 1.2499687754010318e-06,
+      "loss": 0.1196,
+      "step": 15
+    },
+    {
+      "epoch": 0.0036363636363636364,
+      "grad_norm": 2.1700625995281495,
+      "learning_rate": 1.2499641554790185e-06,
+      "loss": 0.1467,
+      "step": 16
+    },
+    {
+      "epoch": 0.003863636363636364,
+      "grad_norm": 1.933417222933536,
+      "learning_rate": 1.2499592169542823e-06,
+      "loss": 0.1103,
+      "step": 17
+    },
+    {
+      "epoch": 0.004090909090909091,
+      "grad_norm": 2.5219504331962312,
+      "learning_rate": 1.2499539598293406e-06,
+      "loss": 0.0998,
+      "step": 18
+    },
+    {
+      "epoch": 0.004318181818181818,
+      "grad_norm": 2.100693152573381,
+      "learning_rate": 1.2499483841068736e-06,
+      "loss": 0.1328,
+      "step": 19
+    },
+    {
+      "epoch": 0.004545454545454545,
+      "grad_norm": 2.383494969613959,
+      "learning_rate": 1.2499424897897237e-06,
+      "loss": 0.089,
+      "step": 20
+    },
+    {
+      "epoch": 0.004772727272727273,
+      "grad_norm": 4.1129313222088255,
+      "learning_rate": 1.2499362768808958e-06,
+      "loss": 0.1564,
+      "step": 21
+    },
+    {
+      "epoch": 0.005,
+      "grad_norm": 2.581452642800533,
+      "learning_rate": 1.2499297453835574e-06,
+      "loss": 0.2821,
+      "step": 22
+    },
+    {
+      "epoch": 0.005227272727272727,
+      "grad_norm": 1.6203149951434954,
+      "learning_rate": 1.2499228953010379e-06,
+      "loss": 0.0721,
+      "step": 23
+    },
+    {
+      "epoch": 0.005454545454545455,
+      "grad_norm": 2.936253710956124,
+      "learning_rate": 1.2499157266368298e-06,
+      "loss": 0.088,
+      "step": 24
+    },
+    {
+      "epoch": 0.005681818181818182,
+      "grad_norm": 1.405438760059647,
+      "learning_rate": 1.2499082393945871e-06,
+      "loss": 0.1457,
+      "step": 25
+    },
+    {
+      "epoch": 0.005909090909090909,
+      "grad_norm": 3.3458582482811523,
+      "learning_rate": 1.2499004335781272e-06,
+      "loss": 0.1977,
+      "step": 26
+    },
+    {
+      "epoch": 0.006136363636363636,
+      "grad_norm": 2.2907238368292018,
+      "learning_rate": 1.2498923091914293e-06,
+      "loss": 0.1024,
+      "step": 27
+    },
+    {
+      "epoch": 0.006363636363636364,
+      "grad_norm": 2.9548897248006263,
+      "learning_rate": 1.249883866238635e-06,
+      "loss": 0.1492,
+      "step": 28
+    },
+    {
+      "epoch": 0.006590909090909091,
+      "grad_norm": 2.6467075680221814,
+      "learning_rate": 1.2498751047240488e-06,
+      "loss": 0.1799,
+      "step": 29
+    },
+    {
+      "epoch": 0.006818181818181818,
+      "grad_norm": 1.0307593556256838,
+      "learning_rate": 1.2498660246521371e-06,
+      "loss": 0.0661,
+      "step": 30
+    },
+    {
+      "epoch": 0.007045454545454546,
+      "grad_norm": 2.567080979155604,
+      "learning_rate": 1.2498566260275289e-06,
+      "loss": 0.1793,
+      "step": 31
+    },
+    {
+      "epoch": 0.007272727272727273,
+      "grad_norm": 1.5121837310812607,
+      "learning_rate": 1.2498469088550156e-06,
+      "loss": 0.1373,
+      "step": 32
+    },
+    {
+      "epoch": 0.0075,
+      "grad_norm": 2.5935256740890567,
+      "learning_rate": 1.2498368731395507e-06,
+      "loss": 0.1909,
+      "step": 33
+    },
+    {
+      "epoch": 0.007727272727272728,
+      "grad_norm": 2.7160799335891967,
+      "learning_rate": 1.2498265188862505e-06,
+      "loss": 0.2161,
+      "step": 34
+    },
+    {
+      "epoch": 0.007954545454545454,
+      "grad_norm": 2.8070038043050527,
+      "learning_rate": 1.2498158461003935e-06,
+      "loss": 0.2082,
+      "step": 35
+    },
+    {
+      "epoch": 0.008181818181818182,
+      "grad_norm": 1.8771195232464262,
+      "learning_rate": 1.2498048547874208e-06,
+      "loss": 0.0734,
+      "step": 36
+    },
+    {
+      "epoch": 0.00840909090909091,
+      "grad_norm": 1.721525400769201,
+      "learning_rate": 1.2497935449529355e-06,
+      "loss": 0.134,
+      "step": 37
+    },
+    {
+      "epoch": 0.008636363636363636,
+      "grad_norm": 2.777096381856707,
+      "learning_rate": 1.2497819166027035e-06,
+      "loss": 0.0964,
+      "step": 38
+    },
+    {
+      "epoch": 0.008863636363636363,
+      "grad_norm": 3.095733815125014,
+      "learning_rate": 1.2497699697426523e-06,
+      "loss": 0.1105,
+      "step": 39
+    },
+    {
+      "epoch": 0.00909090909090909,
+      "grad_norm": 3.1689678044240708,
+      "learning_rate": 1.2497577043788732e-06,
+      "loss": 0.1316,
+      "step": 40
+    },
+    {
+      "epoch": 0.009318181818181817,
+      "grad_norm": 2.1706368102008065,
+      "learning_rate": 1.2497451205176183e-06,
+      "loss": 0.1313,
+      "step": 41
+    },
+    {
+      "epoch": 0.009545454545454546,
+      "grad_norm": 3.3798671750054887,
+      "learning_rate": 1.2497322181653032e-06,
+      "loss": 0.141,
+      "step": 42
+    },
+    {
+      "epoch": 0.009772727272727273,
+      "grad_norm": 1.5726849149448088,
+      "learning_rate": 1.249718997328505e-06,
+      "loss": 0.1325,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.090733945171677,
+      "learning_rate": 1.2497054580139642e-06,
+      "loss": 0.1464,
+      "step": 44
+    },
+    {
+      "epoch": 0.010227272727272727,
+      "grad_norm": 4.125202381212123,
+      "learning_rate": 1.2496916002285823e-06,
+      "loss": 0.1831,
+      "step": 45
+    },
+    {
+      "epoch": 0.010454545454545454,
+      "grad_norm": 3.293554767292981,
+      "learning_rate": 1.2496774239794246e-06,
+      "loss": 0.1799,
+      "step": 46
+    },
+    {
+      "epoch": 0.010681818181818181,
+      "grad_norm": 2.136770899743358,
+      "learning_rate": 1.2496629292737176e-06,
+      "loss": 0.1808,
+      "step": 47
+    },
+    {
+      "epoch": 0.01090909090909091,
+      "grad_norm": 3.3843737419833047,
+      "learning_rate": 1.249648116118851e-06,
+      "loss": 0.1068,
+      "step": 48
+    },
+    {
+      "epoch": 0.011136363636363637,
+      "grad_norm": 1.704683887030336,
+      "learning_rate": 1.2496329845223759e-06,
+      "loss": 0.1798,
+      "step": 49
+    },
+    {
+      "epoch": 0.011363636363636364,
+      "grad_norm": 2.163557430051891,
+      "learning_rate": 1.2496175344920069e-06,
+      "loss": 0.0884,
+      "step": 50
+    },
+    {
+      "epoch": 0.011590909090909091,
+      "grad_norm": 2.686021333516474,
+      "learning_rate": 1.24960176603562e-06,
+      "loss": 0.0963,
+      "step": 51
+    },
+    {
+      "epoch": 0.011818181818181818,
+      "grad_norm": 2.1107227199979204,
+      "learning_rate": 1.2495856791612538e-06,
+      "loss": 0.1368,
+      "step": 52
+    },
+    {
+      "epoch": 0.012045454545454545,
+      "grad_norm": 1.9557221221815864,
+      "learning_rate": 1.2495692738771095e-06,
+      "loss": 0.159,
+      "step": 53
+    },
+    {
+      "epoch": 0.012272727272727272,
+      "grad_norm": 1.4619920571066773,
+      "learning_rate": 1.2495525501915503e-06,
+      "loss": 0.0981,
+      "step": 54
+    },
+    {
+      "epoch": 0.0125,
+      "grad_norm": 3.013156279876758,
+      "learning_rate": 1.2495355081131017e-06,
+      "loss": 0.1711,
+      "step": 55
+    },
+    {
+      "epoch": 0.012727272727272728,
+      "grad_norm": 2.0204802923301632,
+      "learning_rate": 1.249518147650452e-06,
+      "loss": 0.1185,
+      "step": 56
+    },
+    {
+      "epoch": 0.012954545454545455,
+      "grad_norm": 1.6746882725712227,
+      "learning_rate": 1.249500468812451e-06,
+      "loss": 0.1413,
+      "step": 57
+    },
+    {
+      "epoch": 0.013181818181818182,
+      "grad_norm": 3.006788844248048,
+      "learning_rate": 1.2494824716081117e-06,
+      "loss": 0.1457,
+      "step": 58
+    },
+    {
+      "epoch": 0.013409090909090909,
+      "grad_norm": 1.8878479948946736,
+      "learning_rate": 1.2494641560466087e-06,
+      "loss": 0.1331,
+      "step": 59
+    },
+    {
+      "epoch": 0.013636363636363636,
+      "grad_norm": 4.645396009956058,
+      "learning_rate": 1.249445522137279e-06,
+      "loss": 0.2544,
+      "step": 60
+    },
+    {
+      "epoch": 0.013863636363636364,
+      "grad_norm": 1.670983120496584,
+      "learning_rate": 1.2494265698896224e-06,
+      "loss": 0.1171,
+      "step": 61
+    },
+    {
+      "epoch": 0.014090909090909091,
+      "grad_norm": 4.506709575048941,
+      "learning_rate": 1.2494072993133005e-06,
+      "loss": 0.1419,
+      "step": 62
+    },
+    {
+      "epoch": 0.014318181818181818,
+      "grad_norm": 1.6932802140082395,
+      "learning_rate": 1.2493877104181373e-06,
+      "loss": 0.1509,
+      "step": 63
+    },
+    {
+      "epoch": 0.014545454545454545,
+      "grad_norm": 2.805248484903931,
+      "learning_rate": 1.249367803214119e-06,
+      "loss": 0.1318,
+      "step": 64
+    },
+    {
+      "epoch": 0.014772727272727272,
+      "grad_norm": 3.611629026022409,
+      "learning_rate": 1.2493475777113945e-06,
+      "loss": 0.117,
+      "step": 65
+    },
+    {
+      "epoch": 0.015,
+      "grad_norm": 2.7391655355979787,
+      "learning_rate": 1.2493270339202742e-06,
+      "loss": 0.2509,
+      "step": 66
+    },
+    {
+      "epoch": 0.015227272727272726,
+      "grad_norm": 3.972791809939906,
+      "learning_rate": 1.2493061718512314e-06,
+      "loss": 0.1238,
+      "step": 67
+    },
+    {
+      "epoch": 0.015454545454545455,
+      "grad_norm": 4.05811014342497,
+      "learning_rate": 1.2492849915149013e-06,
+      "loss": 0.2206,
+      "step": 68
+    },
+    {
+      "epoch": 0.015681818181818182,
+      "grad_norm": 2.5601455016912564,
+      "learning_rate": 1.2492634929220817e-06,
+      "loss": 0.1139,
+      "step": 69
+    },
+    {
+      "epoch": 0.015909090909090907,
+      "grad_norm": 4.566104987253978,
+      "learning_rate": 1.2492416760837326e-06,
+      "loss": 0.1474,
+      "step": 70
+    },
+    {
+      "epoch": 0.016136363636363636,
+      "grad_norm": 2.7944209312940274,
+      "learning_rate": 1.2492195410109757e-06,
+      "loss": 0.2927,
+      "step": 71
+    },
+    {
+      "epoch": 0.016363636363636365,
+      "grad_norm": 1.4585595300779401,
+      "learning_rate": 1.2491970877150955e-06,
+      "loss": 0.1208,
+      "step": 72
+    },
+    {
+      "epoch": 0.01659090909090909,
+      "grad_norm": 2.913388303917948,
+      "learning_rate": 1.2491743162075384e-06,
+      "loss": 0.1034,
+      "step": 73
+    },
+    {
+      "epoch": 0.01681818181818182,
+      "grad_norm": 1.937214359878327,
+      "learning_rate": 1.2491512264999135e-06,
+      "loss": 0.1447,
+      "step": 74
+    },
+    {
+      "epoch": 0.017045454545454544,
+      "grad_norm": 3.212681789562621,
+      "learning_rate": 1.2491278186039916e-06,
+      "loss": 0.2092,
+      "step": 75
+    },
+    {
+      "epoch": 0.017272727272727273,
+      "grad_norm": 3.451344117815038,
+      "learning_rate": 1.2491040925317057e-06,
+      "loss": 0.1427,
+      "step": 76
+    },
+    {
+      "epoch": 0.0175,
+      "grad_norm": 2.94301458375521,
+      "learning_rate": 1.2490800482951515e-06,
+      "loss": 0.1916,
+      "step": 77
+    },
+    {
+      "epoch": 0.017727272727272727,
+      "grad_norm": 2.018899579428167,
+      "learning_rate": 1.2490556859065865e-06,
+      "loss": 0.1215,
+      "step": 78
+    },
+    {
+      "epoch": 0.017954545454545456,
+      "grad_norm": 0.8564588761854233,
+      "learning_rate": 1.2490310053784301e-06,
+      "loss": 0.0617,
+      "step": 79
+    },
+    {
+      "epoch": 0.01818181818181818,
+      "grad_norm": 2.5229341409637356,
+      "learning_rate": 1.249006006723265e-06,
+      "loss": 0.0732,
+      "step": 80
+    },
+    {
+      "epoch": 0.01840909090909091,
+      "grad_norm": 2.3488339348166236,
+      "learning_rate": 1.2489806899538349e-06,
+      "loss": 0.1183,
+      "step": 81
+    },
+    {
+      "epoch": 0.018636363636363635,
+      "grad_norm": 1.930335683649572,
+      "learning_rate": 1.2489550550830462e-06,
+      "loss": 0.129,
+      "step": 82
+    },
+    {
+      "epoch": 0.018863636363636364,
+      "grad_norm": 2.4699780724844356,
+      "learning_rate": 1.2489291021239674e-06,
+      "loss": 0.2026,
+      "step": 83
+    },
+    {
+      "epoch": 0.019090909090909092,
+      "grad_norm": 3.2835345764107355,
+      "learning_rate": 1.2489028310898293e-06,
+      "loss": 0.1154,
+      "step": 84
+    },
+    {
+      "epoch": 0.019318181818181818,
+      "grad_norm": 2.636406250732528,
+      "learning_rate": 1.2488762419940244e-06,
+      "loss": 0.2214,
+      "step": 85
+    },
+    {
+      "epoch": 0.019545454545454546,
+      "grad_norm": 3.200678859440104,
+      "learning_rate": 1.248849334850108e-06,
+      "loss": 0.1666,
+      "step": 86
+    },
+    {
+      "epoch": 0.01977272727272727,
+      "grad_norm": 1.5632662895896403,
+      "learning_rate": 1.2488221096717967e-06,
+      "loss": 0.1481,
+      "step": 87
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.811124970706698,
+      "learning_rate": 1.2487945664729703e-06,
+      "loss": 0.146,
+      "step": 88
+    },
+    {
+      "epoch": 0.020227272727272726,
+      "grad_norm": 2.867389422565866,
+      "learning_rate": 1.2487667052676699e-06,
+      "loss": 0.2211,
+      "step": 89
+    },
+    {
+      "epoch": 0.020454545454545454,
+      "grad_norm": 3.436638598057338,
+      "learning_rate": 1.2487385260700987e-06,
+      "loss": 0.2435,
+      "step": 90
+    },
+    {
+      "epoch": 0.020681818181818183,
+      "grad_norm": 1.7496904686727894,
+      "learning_rate": 1.2487100288946228e-06,
+      "loss": 0.1491,
+      "step": 91
+    },
+    {
+      "epoch": 0.02090909090909091,
+      "grad_norm": 3.455590706887831,
+      "learning_rate": 1.2486812137557693e-06,
+      "loss": 0.1419,
+      "step": 92
+    },
+    {
+      "epoch": 0.021136363636363637,
+      "grad_norm": 3.0197426727296564,
+      "learning_rate": 1.2486520806682283e-06,
+      "loss": 0.135,
+      "step": 93
+    },
+    {
+      "epoch": 0.021363636363636362,
+      "grad_norm": 2.2191387852602693,
+      "learning_rate": 1.248622629646852e-06,
+      "loss": 0.1773,
+      "step": 94
+    },
+    {
+      "epoch": 0.02159090909090909,
+      "grad_norm": 3.585545029458258,
+      "learning_rate": 1.2485928607066537e-06,
+      "loss": 0.175,
+      "step": 95
+    },
+    {
+      "epoch": 0.02181818181818182,
+      "grad_norm": 2.8209369985210873,
+      "learning_rate": 1.2485627738628097e-06,
+      "loss": 0.199,
+      "step": 96
+    },
+    {
+      "epoch": 0.022045454545454545,
+      "grad_norm": 2.4960958244989087,
+      "learning_rate": 1.248532369130658e-06,
+      "loss": 0.1129,
+      "step": 97
+    },
+    {
+      "epoch": 0.022272727272727274,
+      "grad_norm": 3.5924121138241976,
+      "learning_rate": 1.2485016465256987e-06,
+      "loss": 0.1242,
+      "step": 98
+    },
+    {
+      "epoch": 0.0225,
+      "grad_norm": 3.624954418252119,
+      "learning_rate": 1.2484706060635945e-06,
+      "loss": 0.1203,
+      "step": 99
+    },
+    {
+      "epoch": 0.022727272727272728,
+      "grad_norm": 2.5959478530072695,
+      "learning_rate": 1.248439247760169e-06,
+      "loss": 0.1068,
+      "step": 100
+    },
+    {
+      "epoch": 0.022954545454545453,
+      "grad_norm": 3.790331698333609,
+      "learning_rate": 1.2484075716314085e-06,
+      "loss": 0.1211,
+      "step": 101
+    },
+    {
+      "epoch": 0.023181818181818182,
+      "grad_norm": 2.7512214904458894,
+      "learning_rate": 1.2483755776934616e-06,
+      "loss": 0.2594,
+      "step": 102
+    },
+    {
+      "epoch": 0.02340909090909091,
+      "grad_norm": 2.162183785160701,
+      "learning_rate": 1.2483432659626384e-06,
+      "loss": 0.1215,
+      "step": 103
+    },
+    {
+      "epoch": 0.023636363636363636,
+      "grad_norm": 1.6358182135482229,
+      "learning_rate": 1.2483106364554115e-06,
+      "loss": 0.1365,
+      "step": 104
+    },
+    {
+      "epoch": 0.023863636363636365,
+      "grad_norm": 4.3529178714973416,
+      "learning_rate": 1.248277689188415e-06,
+      "loss": 0.1665,
+      "step": 105
+    },
+    {
+      "epoch": 0.02409090909090909,
+      "grad_norm": 2.5021206133167704,
+      "learning_rate": 1.248244424178445e-06,
+      "loss": 0.2466,
+      "step": 106
+    },
+    {
+      "epoch": 0.02431818181818182,
+      "grad_norm": 4.739752270142774,
+      "learning_rate": 1.2482108414424602e-06,
+      "loss": 0.1462,
+      "step": 107
+    },
+    {
+      "epoch": 0.024545454545454544,
+      "grad_norm": 2.4811982413789955,
+      "learning_rate": 1.2481769409975805e-06,
+      "loss": 0.1429,
+      "step": 108
+    },
+    {
+      "epoch": 0.024772727272727273,
+      "grad_norm": 3.042622949636853,
+      "learning_rate": 1.2481427228610881e-06,
+      "loss": 0.1148,
+      "step": 109
+    },
+    {
+      "epoch": 0.025,
+      "grad_norm": 2.3509465897611808,
+      "learning_rate": 1.2481081870504278e-06,
+      "loss": 0.1108,
+      "step": 110
+    },
+    {
+      "epoch": 0.025227272727272727,
+      "grad_norm": 3.0684347353646135,
+      "learning_rate": 1.2480733335832052e-06,
+      "loss": 0.0931,
+      "step": 111
+    },
+    {
+      "epoch": 0.025454545454545455,
+      "grad_norm": 2.7667495796907384,
+      "learning_rate": 1.2480381624771882e-06,
+      "loss": 0.0979,
+      "step": 112
+    },
+    {
+      "epoch": 0.02568181818181818,
+      "grad_norm": 2.1741006971341426,
+      "learning_rate": 1.2480026737503073e-06,
+      "loss": 0.1443,
+      "step": 113
+    },
+    {
+      "epoch": 0.02590909090909091,
+      "grad_norm": 3.230826024202444,
+      "learning_rate": 1.2479668674206543e-06,
+      "loss": 0.1681,
+      "step": 114
+    },
+    {
+      "epoch": 0.026136363636363635,
+      "grad_norm": 2.4553243229144828,
+      "learning_rate": 1.2479307435064827e-06,
+      "loss": 0.1009,
+      "step": 115
+    },
+    {
+      "epoch": 0.026363636363636363,
+      "grad_norm": 1.545051619576096,
+      "learning_rate": 1.2478943020262087e-06,
+      "loss": 0.1371,
+      "step": 116
+    },
+    {
+      "epoch": 0.026590909090909092,
+      "grad_norm": 1.9911957255965105,
+      "learning_rate": 1.2478575429984097e-06,
+      "loss": 0.1582,
+      "step": 117
+    },
+    {
+      "epoch": 0.026818181818181817,
+      "grad_norm": 2.3361818882518413,
+      "learning_rate": 1.2478204664418254e-06,
+      "loss": 0.1252,
+      "step": 118
+    },
+    {
+      "epoch": 0.027045454545454546,
+      "grad_norm": 3.9567869664597985,
+      "learning_rate": 1.2477830723753567e-06,
+      "loss": 0.186,
+      "step": 119
+    },
+    {
+      "epoch": 0.02727272727272727,
+      "grad_norm": 3.1037912693923926,
+      "learning_rate": 1.2477453608180673e-06,
+      "loss": 0.1061,
+      "step": 120
+    },
+    {
+      "epoch": 0.0275,
+      "grad_norm": 3.825225348417575,
+      "learning_rate": 1.2477073317891822e-06,
+      "loss": 0.2861,
+      "step": 121
+    },
+    {
+      "epoch": 0.02772727272727273,
+      "grad_norm": 3.769674570000181,
+      "learning_rate": 1.2476689853080883e-06,
+      "loss": 0.1018,
+      "step": 122
+    },
+    {
+      "epoch": 0.027954545454545454,
+      "grad_norm": 2.2379422155174105,
+      "learning_rate": 1.2476303213943346e-06,
+      "loss": 0.1172,
+      "step": 123
+    },
+    {
+      "epoch": 0.028181818181818183,
+      "grad_norm": 2.1579273672861845,
+      "learning_rate": 1.2475913400676314e-06,
+      "loss": 0.0899,
+      "step": 124
+    },
+    {
+      "epoch": 0.028409090909090908,
+      "grad_norm": 3.3836873275354757,
+      "learning_rate": 1.2475520413478516e-06,
+      "loss": 0.2304,
+      "step": 125
+    },
+    {
+      "epoch": 0.028636363636363637,
+      "grad_norm": 3.0871670960753193,
+      "learning_rate": 1.247512425255029e-06,
+      "loss": 0.0788,
+      "step": 126
+    },
+    {
+      "epoch": 0.028863636363636362,
+      "grad_norm": 2.1566139006075824,
+      "learning_rate": 1.2474724918093594e-06,
+      "loss": 0.1124,
+      "step": 127
+    },
+    {
+      "epoch": 0.02909090909090909,
+      "grad_norm": 2.28821595550464,
+      "learning_rate": 1.2474322410312012e-06,
+      "loss": 0.1188,
+      "step": 128
+    },
+    {
+      "epoch": 0.02931818181818182,
+      "grad_norm": 2.081144028694687,
+      "learning_rate": 1.247391672941074e-06,
+      "loss": 0.0834,
+      "step": 129
+    },
+    {
+      "epoch": 0.029545454545454545,
+      "grad_norm": 2.832582122728977,
+      "learning_rate": 1.2473507875596586e-06,
+      "loss": 0.2121,
+      "step": 130
+    },
+    {
+      "epoch": 0.029772727272727274,
+      "grad_norm": 2.8879294827733584,
+      "learning_rate": 1.2473095849077984e-06,
+      "loss": 0.1334,
+      "step": 131
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2378491169475914,
+      "learning_rate": 1.2472680650064984e-06,
+      "loss": 0.1767,
+      "step": 132
+    },
+    {
+      "epoch": 0.030227272727272728,
+      "grad_norm": 2.8864718981704147,
+      "learning_rate": 1.247226227876925e-06,
+      "loss": 0.1735,
+      "step": 133
+    },
+    {
+      "epoch": 0.030454545454545453,
+      "grad_norm": 1.0769027456123426,
+      "learning_rate": 1.2471840735404066e-06,
+      "loss": 0.0518,
+      "step": 134
+    },
+    {
+      "epoch": 0.03068181818181818,
+      "grad_norm": 4.72244866526862,
+      "learning_rate": 1.2471416020184332e-06,
+      "loss": 0.1506,
+      "step": 135
+    },
+    {
+      "epoch": 0.03090909090909091,
+      "grad_norm": 4.412997971829164,
+      "learning_rate": 1.2470988133326564e-06,
+      "loss": 0.1673,
+      "step": 136
+    },
+    {
+      "epoch": 0.031136363636363636,
+      "grad_norm": 1.7473364256637949,
+      "learning_rate": 1.2470557075048897e-06,
+      "loss": 0.1224,
+      "step": 137
+    },
+    {
+      "epoch": 0.031363636363636364,
+      "grad_norm": 1.6208448524360295,
+      "learning_rate": 1.2470122845571081e-06,
+      "loss": 0.085,
+      "step": 138
+    },
+    {
+      "epoch": 0.03159090909090909,
+      "grad_norm": 2.3967275442111045,
+      "learning_rate": 1.2469685445114486e-06,
+      "loss": 0.1053,
+      "step": 139
+    },
+    {
+      "epoch": 0.031818181818181815,
+      "grad_norm": 2.2292655956739633,
+      "learning_rate": 1.2469244873902089e-06,
+      "loss": 0.1604,
+      "step": 140
+    },
+    {
+      "epoch": 0.032045454545454544,
+      "grad_norm": 1.7396484319567247,
+      "learning_rate": 1.2468801132158499e-06,
+      "loss": 0.1028,
+      "step": 141
+    },
+    {
+      "epoch": 0.03227272727272727,
+      "grad_norm": 3.5928251472292567,
+      "learning_rate": 1.2468354220109926e-06,
+      "loss": 0.1499,
+      "step": 142
+    },
+    {
+      "epoch": 0.0325,
+      "grad_norm": 1.67539064124059,
+      "learning_rate": 1.2467904137984208e-06,
+      "loss": 0.1175,
+      "step": 143
+    },
+    {
+      "epoch": 0.03272727272727273,
+      "grad_norm": 4.4693738723985375,
+      "learning_rate": 1.246745088601079e-06,
+      "loss": 0.1408,
+      "step": 144
+    },
+    {
+      "epoch": 0.03295454545454545,
+      "grad_norm": 2.473273370959692,
+      "learning_rate": 1.246699446442074e-06,
+      "loss": 0.2093,
+      "step": 145
+    },
+    {
+      "epoch": 0.03318181818181818,
+      "grad_norm": 3.2727061306974328,
+      "learning_rate": 1.2466534873446738e-06,
+      "loss": 0.2034,
+      "step": 146
+    },
+    {
+      "epoch": 0.03340909090909091,
+      "grad_norm": 1.325912807404117,
+      "learning_rate": 1.246607211332308e-06,
+      "loss": 0.0926,
+      "step": 147
+    },
+    {
+      "epoch": 0.03363636363636364,
+      "grad_norm": 3.4545736020672733,
+      "learning_rate": 1.2465606184285679e-06,
+      "loss": 0.1085,
+      "step": 148
+    },
+    {
+      "epoch": 0.03386363636363637,
+      "grad_norm": 2.3093751125992634,
+      "learning_rate": 1.2465137086572057e-06,
+      "loss": 0.1136,
+      "step": 149
+    },
+    {
+      "epoch": 0.03409090909090909,
+      "grad_norm": 2.1484956138294833,
+      "learning_rate": 1.2464664820421365e-06,
+      "loss": 0.2293,
+      "step": 150
+    },
+    {
+      "epoch": 0.03431818181818182,
+      "grad_norm": 1.8332160203827814,
+      "learning_rate": 1.246418938607436e-06,
+      "loss": 0.1672,
+      "step": 151
+    },
+    {
+      "epoch": 0.034545454545454546,
+      "grad_norm": 3.233420719630291,
+      "learning_rate": 1.246371078377341e-06,
+      "loss": 0.1204,
+      "step": 152
+    },
+    {
+      "epoch": 0.034772727272727275,
+      "grad_norm": 3.2450989492554276,
+      "learning_rate": 1.246322901376251e-06,
+      "loss": 0.2279,
+      "step": 153
+    },
+    {
+      "epoch": 0.035,
+      "grad_norm": 2.806743931345361,
+      "learning_rate": 1.2462744076287257e-06,
+      "loss": 0.1163,
+      "step": 154
+    },
+    {
+      "epoch": 0.035227272727272725,
+      "grad_norm": 3.460987122171745,
+      "learning_rate": 1.2462255971594874e-06,
+      "loss": 0.1229,
+      "step": 155
+    },
+    {
+      "epoch": 0.035454545454545454,
+      "grad_norm": 2.7810369881112327,
+      "learning_rate": 1.2461764699934192e-06,
+      "loss": 0.0968,
+      "step": 156
+    },
+    {
+      "epoch": 0.03568181818181818,
+      "grad_norm": 4.181813889441663,
+      "learning_rate": 1.2461270261555657e-06,
+      "loss": 0.1739,
+      "step": 157
+    },
+    {
+      "epoch": 0.03590909090909091,
+      "grad_norm": 6.006197431180828,
+      "learning_rate": 1.246077265671133e-06,
+      "loss": 0.2387,
+      "step": 158
+    },
+    {
+      "epoch": 0.03613636363636363,
+      "grad_norm": 2.7649421613921317,
+      "learning_rate": 1.2460271885654891e-06,
+      "loss": 0.119,
+      "step": 159
+    },
+    {
+      "epoch": 0.03636363636363636,
+      "grad_norm": 2.262654241194518,
+      "learning_rate": 1.2459767948641626e-06,
+      "loss": 0.1228,
+      "step": 160
+    },
+    {
+      "epoch": 0.03659090909090909,
+      "grad_norm": 3.2896706683442263,
+      "learning_rate": 1.2459260845928438e-06,
+      "loss": 0.177,
+      "step": 161
+    },
+    {
+      "epoch": 0.03681818181818182,
+      "grad_norm": 2.718185942978294,
+      "learning_rate": 1.245875057777385e-06,
+      "loss": 0.1509,
+      "step": 162
+    },
+    {
+      "epoch": 0.03704545454545455,
+      "grad_norm": 3.2744518424095026,
+      "learning_rate": 1.245823714443799e-06,
+      "loss": 0.1092,
+      "step": 163
+    },
+    {
+      "epoch": 0.03727272727272727,
+      "grad_norm": 3.524657730814499,
+      "learning_rate": 1.24577205461826e-06,
+      "loss": 0.1418,
+      "step": 164
+    },
+    {
+      "epoch": 0.0375,
+      "grad_norm": 2.5373129753575254,
+      "learning_rate": 1.2457200783271044e-06,
+      "loss": 0.1314,
+      "step": 165
+    },
+    {
+      "epoch": 0.03772727272727273,
+      "grad_norm": 2.754146612878973,
+      "learning_rate": 1.245667785596829e-06,
+      "loss": 0.2591,
+      "step": 166
+    },
+    {
+      "epoch": 0.037954545454545456,
+      "grad_norm": 2.170142434117645,
+      "learning_rate": 1.2456151764540924e-06,
+      "loss": 0.1016,
+      "step": 167
+    },
+    {
+      "epoch": 0.038181818181818185,
+      "grad_norm": 2.355791270635665,
+      "learning_rate": 1.2455622509257147e-06,
+      "loss": 0.1111,
+      "step": 168
+    },
+    {
+      "epoch": 0.03840909090909091,
+      "grad_norm": 3.501859615924086,
+      "learning_rate": 1.2455090090386765e-06,
+      "loss": 0.1886,
+      "step": 169
+    },
+    {
+      "epoch": 0.038636363636363635,
+      "grad_norm": 3.28713346787876,
+      "learning_rate": 1.2454554508201205e-06,
+      "loss": 0.0801,
+      "step": 170
+    },
+    {
+      "epoch": 0.038863636363636364,
+      "grad_norm": 2.7885107606417106,
+      "learning_rate": 1.2454015762973505e-06,
+      "loss": 0.149,
+      "step": 171
+    },
+    {
+      "epoch": 0.03909090909090909,
+      "grad_norm": 2.818356763629871,
+      "learning_rate": 1.2453473854978307e-06,
+      "loss": 0.1425,
+      "step": 172
+    },
+    {
+      "epoch": 0.03931818181818182,
+      "grad_norm": 2.481505052276004,
+      "learning_rate": 1.2452928784491877e-06,
+      "loss": 0.1993,
+      "step": 173
+    },
+    {
+      "epoch": 0.03954545454545454,
+      "grad_norm": 1.9465593142172546,
+      "learning_rate": 1.245238055179209e-06,
+      "loss": 0.1482,
+      "step": 174
+    },
+    {
+      "epoch": 0.03977272727272727,
+      "grad_norm": 2.7643260416657025,
+      "learning_rate": 1.245182915715843e-06,
+      "loss": 0.1167,
+      "step": 175
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.356853645822642,
+      "learning_rate": 1.2451274600871991e-06,
+      "loss": 0.0978,
+      "step": 176
+    },
+    {
+      "epoch": 0.04022727272727273,
+      "grad_norm": 2.5862196503150034,
+      "learning_rate": 1.245071688321549e-06,
+      "loss": 0.1278,
+      "step": 177
+    },
+    {
+      "epoch": 0.04045454545454545,
+      "grad_norm": 2.424784144507825,
+      "learning_rate": 1.2450156004473238e-06,
+      "loss": 0.1949,
+      "step": 178
+    },
+    {
+      "epoch": 0.04068181818181818,
+      "grad_norm": 1.5167480770562158,
+      "learning_rate": 1.2449591964931173e-06,
+      "loss": 0.2136,
+      "step": 179
+    },
+    {
+      "epoch": 0.04090909090909091,
+      "grad_norm": 3.504049407316891,
+      "learning_rate": 1.2449024764876841e-06,
+      "loss": 0.1583,
+      "step": 180
+    },
+    {
+      "epoch": 0.04113636363636364,
+      "grad_norm": 2.7876064085836743,
+      "learning_rate": 1.2448454404599393e-06,
+      "loss": 0.2616,
+      "step": 181
+    },
+    {
+      "epoch": 0.041363636363636366,
+      "grad_norm": 4.188049531676353,
+      "learning_rate": 1.2447880884389597e-06,
+      "loss": 0.1237,
+      "step": 182
+    },
+    {
+      "epoch": 0.04159090909090909,
+      "grad_norm": 3.269774129018881,
+      "learning_rate": 1.2447304204539827e-06,
+      "loss": 0.0875,
+      "step": 183
+    },
+    {
+      "epoch": 0.04181818181818182,
+      "grad_norm": 2.5265314094507185,
+      "learning_rate": 1.2446724365344076e-06,
+      "loss": 0.0904,
+      "step": 184
+    },
+    {
+      "epoch": 0.042045454545454546,
+      "grad_norm": 3.0396457657195723,
+      "learning_rate": 1.2446141367097936e-06,
+      "loss": 0.1298,
+      "step": 185
+    },
+    {
+      "epoch": 0.042272727272727274,
+      "grad_norm": 3.191958324561842,
+      "learning_rate": 1.244555521009862e-06,
+      "loss": 0.1459,
+      "step": 186
+    },
+    {
+      "epoch": 0.0425,
+      "grad_norm": 3.7887528315532784,
+      "learning_rate": 1.2444965894644946e-06,
+      "loss": 0.1569,
+      "step": 187
+    },
+    {
+      "epoch": 0.042727272727272725,
+      "grad_norm": 5.45023731492135,
+      "learning_rate": 1.2444373421037345e-06,
+      "loss": 0.2733,
+      "step": 188
+    },
+    {
+      "epoch": 0.042954545454545454,
+      "grad_norm": 2.754817242801967,
+      "learning_rate": 1.2443777789577852e-06,
+      "loss": 0.1175,
+      "step": 189
+    },
+    {
+      "epoch": 0.04318181818181818,
+      "grad_norm": 3.0548377085093716,
+      "learning_rate": 1.244317900057012e-06,
+      "loss": 0.1216,
+      "step": 190
+    },
+    {
+      "epoch": 0.04340909090909091,
+      "grad_norm": 2.639386735893913,
+      "learning_rate": 1.2442577054319405e-06,
+      "loss": 0.142,
+      "step": 191
+    },
+    {
+      "epoch": 0.04363636363636364,
+      "grad_norm": 3.466380370035106,
+      "learning_rate": 1.2441971951132578e-06,
+      "loss": 0.1237,
+      "step": 192
+    },
+    {
+      "epoch": 0.04386363636363636,
+      "grad_norm": 3.68544539309394,
+      "learning_rate": 1.2441363691318114e-06,
+      "loss": 0.1296,
+      "step": 193
+    },
+    {
+      "epoch": 0.04409090909090909,
+      "grad_norm": 2.8824045011396304,
+      "learning_rate": 1.2440752275186102e-06,
+      "loss": 0.1077,
+      "step": 194
+    },
+    {
+      "epoch": 0.04431818181818182,
+      "grad_norm": 1.8573180103946834,
+      "learning_rate": 1.244013770304824e-06,
+      "loss": 0.1189,
+      "step": 195
+    },
+    {
+      "epoch": 0.04454545454545455,
+      "grad_norm": 1.8624525681425799,
+      "learning_rate": 1.2439519975217828e-06,
+      "loss": 0.0655,
+      "step": 196
+    },
+    {
+      "epoch": 0.04477272727272727,
+      "grad_norm": 4.8037178991988325,
+      "learning_rate": 1.2438899092009783e-06,
+      "loss": 0.1718,
+      "step": 197
+    },
+    {
+      "epoch": 0.045,
+      "grad_norm": 3.869042621456544,
+      "learning_rate": 1.2438275053740624e-06,
+      "loss": 0.1697,
+      "step": 198
+    },
+    {
+      "epoch": 0.04522727272727273,
+      "grad_norm": 2.9878562689337422,
+      "learning_rate": 1.2437647860728487e-06,
+      "loss": 0.1229,
+      "step": 199
+    },
+    {
+      "epoch": 0.045454545454545456,
+      "grad_norm": 2.646364994650606,
+      "learning_rate": 1.2437017513293107e-06,
+      "loss": 0.1796,
+      "step": 200
+    },
+    {
+      "epoch": 0.045681818181818185,
+      "grad_norm": 2.283599361625718,
+      "learning_rate": 1.243638401175583e-06,
+      "loss": 0.1142,
+      "step": 201
+    },
+    {
+      "epoch": 0.045909090909090906,
+      "grad_norm": 2.0141963966542455,
+      "learning_rate": 1.2435747356439614e-06,
+      "loss": 0.1687,
+      "step": 202
+    },
+    {
+      "epoch": 0.046136363636363635,
+      "grad_norm": 2.88185819117883,
+      "learning_rate": 1.2435107547669022e-06,
+      "loss": 0.1294,
+      "step": 203
+    },
+    {
+      "epoch": 0.046363636363636364,
+      "grad_norm": 2.8723417881867848,
+      "learning_rate": 1.2434464585770226e-06,
+      "loss": 0.0823,
+      "step": 204
+    },
+    {
+      "epoch": 0.04659090909090909,
+      "grad_norm": 2.106970976966765,
+      "learning_rate": 1.2433818471070998e-06,
+      "loss": 0.1554,
+      "step": 205
+    },
+    {
+      "epoch": 0.04681818181818182,
+      "grad_norm": 3.16499177363157,
+      "learning_rate": 1.2433169203900726e-06,
+      "loss": 0.1807,
+      "step": 206
+    },
+    {
+      "epoch": 0.04704545454545454,
+      "grad_norm": 3.1600104847226778,
+      "learning_rate": 1.2432516784590405e-06,
+      "loss": 0.1191,
+      "step": 207
+    },
+    {
+      "epoch": 0.04727272727272727,
+      "grad_norm": 3.3124875443634525,
+      "learning_rate": 1.243186121347263e-06,
+      "loss": 0.1236,
+      "step": 208
+    },
+    {
+      "epoch": 0.0475,
+      "grad_norm": 1.864014171448486,
+      "learning_rate": 1.243120249088161e-06,
+      "loss": 0.088,
+      "step": 209
+    },
+    {
+      "epoch": 0.04772727272727273,
+      "grad_norm": 2.610280431852156,
+      "learning_rate": 1.2430540617153156e-06,
+      "loss": 0.1273,
+      "step": 210
+    },
+    {
+      "epoch": 0.04795454545454545,
+      "grad_norm": 1.7258742021312505,
+      "learning_rate": 1.2429875592624685e-06,
+      "loss": 0.1451,
+      "step": 211
+    },
+    {
+      "epoch": 0.04818181818181818,
+      "grad_norm": 0.6993633257446737,
+      "learning_rate": 1.2429207417635226e-06,
+      "loss": 0.0404,
+      "step": 212
+    },
+    {
+      "epoch": 0.04840909090909091,
+      "grad_norm": 1.9043416437019283,
+      "learning_rate": 1.242853609252541e-06,
+      "loss": 0.1478,
+      "step": 213
+    },
+    {
+      "epoch": 0.04863636363636364,
+      "grad_norm": 1.9573910066768203,
+      "learning_rate": 1.2427861617637472e-06,
+      "loss": 0.1265,
+      "step": 214
+    },
+    {
+      "epoch": 0.048863636363636366,
+      "grad_norm": 4.156755581065251,
+      "learning_rate": 1.2427183993315256e-06,
+      "loss": 0.0986,
+      "step": 215
+    },
+    {
+      "epoch": 0.04909090909090909,
+      "grad_norm": 1.9380150097538975,
+      "learning_rate": 1.2426503219904213e-06,
+      "loss": 0.1023,
+      "step": 216
+    },
+    {
+      "epoch": 0.04931818181818182,
+      "grad_norm": 2.7387162927934257,
+      "learning_rate": 1.242581929775139e-06,
+      "loss": 0.1338,
+      "step": 217
+    },
+    {
+      "epoch": 0.049545454545454545,
+      "grad_norm": 5.64193223190077,
+      "learning_rate": 1.2425132227205456e-06,
+      "loss": 0.1819,
+      "step": 218
+    },
+    {
+      "epoch": 0.049772727272727274,
+      "grad_norm": 1.1008157177143347,
+      "learning_rate": 1.2424442008616667e-06,
+      "loss": 0.0443,
+      "step": 219
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5439436074399002,
+      "learning_rate": 1.2423748642336894e-06,
+      "loss": 0.1346,
+      "step": 220
+    },
+    {
+      "epoch": 0.050227272727272725,
+      "grad_norm": 1.7562341337137453,
+      "learning_rate": 1.2423052128719611e-06,
+      "loss": 0.0722,
+      "step": 221
+    },
+    {
+      "epoch": 0.05045454545454545,
+      "grad_norm": 1.3123969148516863,
+      "learning_rate": 1.24223524681199e-06,
+      "loss": 0.0955,
+      "step": 222
+    },
+    {
+      "epoch": 0.05068181818181818,
+      "grad_norm": 2.051906705201393,
+      "learning_rate": 1.2421649660894438e-06,
+      "loss": 0.1113,
+      "step": 223
+    },
+    {
+      "epoch": 0.05090909090909091,
+      "grad_norm": 3.2870157705830825,
+      "learning_rate": 1.2420943707401514e-06,
+      "loss": 0.1128,
+      "step": 224
+    },
+    {
+      "epoch": 0.05113636363636364,
+      "grad_norm": 2.9949884334936985,
+      "learning_rate": 1.2420234608001017e-06,
+      "loss": 0.264,
+      "step": 225
+    },
+    {
+      "epoch": 0.05136363636363636,
+      "grad_norm": 2.4157050048551376,
+      "learning_rate": 1.2419522363054446e-06,
+      "loss": 0.1153,
+      "step": 226
+    },
+    {
+      "epoch": 0.05159090909090909,
+      "grad_norm": 2.7584081065097177,
+      "learning_rate": 1.2418806972924893e-06,
+      "loss": 0.0828,
+      "step": 227
+    },
+    {
+      "epoch": 0.05181818181818182,
+      "grad_norm": 3.3021884255619685,
+      "learning_rate": 1.2418088437977063e-06,
+      "loss": 0.1351,
+      "step": 228
+    },
+    {
+      "epoch": 0.05204545454545455,
+      "grad_norm": 1.865539004075356,
+      "learning_rate": 1.241736675857726e-06,
+      "loss": 0.0783,
+      "step": 229
+    },
+    {
+      "epoch": 0.05227272727272727,
+      "grad_norm": 3.7000179659279104,
+      "learning_rate": 1.241664193509339e-06,
+      "loss": 0.1048,
+      "step": 230
+    },
+    {
+      "epoch": 0.0525,
+      "grad_norm": 3.204468627639606,
+      "learning_rate": 1.2415913967894966e-06,
+      "loss": 0.2544,
+      "step": 231
+    },
+    {
+      "epoch": 0.05272727272727273,
+      "grad_norm": 4.852521100892061,
+      "learning_rate": 1.2415182857353098e-06,
+      "loss": 0.2788,
+      "step": 232
+    },
+    {
+      "epoch": 0.052954545454545456,
+      "grad_norm": 2.674894409516308,
+      "learning_rate": 1.2414448603840504e-06,
+      "loss": 0.2185,
+      "step": 233
+    },
+    {
+      "epoch": 0.053181818181818184,
+      "grad_norm": 3.03364506628968,
+      "learning_rate": 1.24137112077315e-06,
+      "loss": 0.2505,
+      "step": 234
+    },
+    {
+      "epoch": 0.053409090909090906,
+      "grad_norm": 3.630449093327049,
+      "learning_rate": 1.2412970669402005e-06,
+      "loss": 0.3191,
+      "step": 235
+    },
+    {
+      "epoch": 0.053636363636363635,
+      "grad_norm": 2.588560263404003,
+      "learning_rate": 1.2412226989229542e-06,
+      "loss": 0.2163,
+      "step": 236
+    },
+    {
+      "epoch": 0.053863636363636364,
+      "grad_norm": 3.9769129237150245,
+      "learning_rate": 1.2411480167593237e-06,
+      "loss": 0.1077,
+      "step": 237
+    },
+    {
+      "epoch": 0.05409090909090909,
+      "grad_norm": 2.520145916247412,
+      "learning_rate": 1.241073020487381e-06,
+      "loss": 0.13,
+      "step": 238
+    },
+    {
+      "epoch": 0.05431818181818182,
+      "grad_norm": 1.4410843509887616,
+      "learning_rate": 1.2409977101453591e-06,
+      "loss": 0.1064,
+      "step": 239
+    },
+    {
+      "epoch": 0.05454545454545454,
+      "grad_norm": 2.8419658024097467,
+      "learning_rate": 1.2409220857716506e-06,
+      "loss": 0.1426,
+      "step": 240
+    },
+    {
+      "epoch": 0.05477272727272727,
+      "grad_norm": 2.312156363289563,
+      "learning_rate": 1.2408461474048083e-06,
+      "loss": 0.0995,
+      "step": 241
+    },
+    {
+      "epoch": 0.055,
+      "grad_norm": 5.080611895503526,
+      "learning_rate": 1.240769895083545e-06,
+      "loss": 0.1296,
+      "step": 242
+    },
+    {
+      "epoch": 0.05522727272727273,
+      "grad_norm": 2.77091289283022,
+      "learning_rate": 1.2406933288467337e-06,
+      "loss": 0.2343,
+      "step": 243
+    },
+    {
+      "epoch": 0.05545454545454546,
+      "grad_norm": 2.7566547821574683,
+      "learning_rate": 1.2406164487334077e-06,
+      "loss": 0.249,
+      "step": 244
+    },
+    {
+      "epoch": 0.05568181818181818,
+      "grad_norm": 3.6482010847990827,
+      "learning_rate": 1.2405392547827594e-06,
+      "loss": 0.1609,
+      "step": 245
+    },
+    {
+      "epoch": 0.05590909090909091,
+      "grad_norm": 1.4650059168862144,
+      "learning_rate": 1.2404617470341423e-06,
+      "loss": 0.0532,
+      "step": 246
+    },
+    {
+      "epoch": 0.05613636363636364,
+      "grad_norm": 3.0146435270136416,
+      "learning_rate": 1.2403839255270693e-06,
+      "loss": 0.1063,
+      "step": 247
+    },
+    {
+      "epoch": 0.056363636363636366,
+      "grad_norm": 1.866016419427917,
+      "learning_rate": 1.2403057903012128e-06,
+      "loss": 0.1526,
+      "step": 248
+    },
+    {
+      "epoch": 0.05659090909090909,
+      "grad_norm": 2.654715786547076,
+      "learning_rate": 1.240227341396406e-06,
+      "loss": 0.1014,
+      "step": 249
+    },
+    {
+      "epoch": 0.056818181818181816,
+      "grad_norm": 1.0322983298092598,
+      "learning_rate": 1.2401485788526418e-06,
+      "loss": 0.1125,
+      "step": 250
+    },
+    {
+      "epoch": 0.057045454545454545,
+      "grad_norm": 3.2986213867502063,
+      "learning_rate": 1.2400695027100725e-06,
+      "loss": 0.2034,
+      "step": 251
+    },
+    {
+      "epoch": 0.057272727272727274,
+      "grad_norm": 3.5416559384400346,
+      "learning_rate": 1.2399901130090112e-06,
+      "loss": 0.1696,
+      "step": 252
+    },
+    {
+      "epoch": 0.0575,
+      "grad_norm": 1.4300547107197759,
+      "learning_rate": 1.2399104097899295e-06,
+      "loss": 0.1007,
+      "step": 253
+    },
+    {
+      "epoch": 0.057727272727272724,
+      "grad_norm": 1.429434314431561,
+      "learning_rate": 1.2398303930934601e-06,
+      "loss": 0.0757,
+      "step": 254
+    },
+    {
+      "epoch": 0.05795454545454545,
+      "grad_norm": 2.4758928845338746,
+      "learning_rate": 1.2397500629603948e-06,
+      "loss": 0.1081,
+      "step": 255
+    },
+    {
+      "epoch": 0.05818181818181818,
+      "grad_norm": 2.7980189995970743,
+      "learning_rate": 1.2396694194316851e-06,
+      "loss": 0.0887,
+      "step": 256
+    },
+    {
+      "epoch": 0.05840909090909091,
+      "grad_norm": 3.000697492517287,
+      "learning_rate": 1.2395884625484433e-06,
+      "loss": 0.2617,
+      "step": 257
+    },
+    {
+      "epoch": 0.05863636363636364,
+      "grad_norm": 4.185118660498254,
+      "learning_rate": 1.2395071923519403e-06,
+      "loss": 0.1726,
+      "step": 258
+    },
+    {
+      "epoch": 0.05886363636363636,
+      "grad_norm": 1.9048535451694386,
+      "learning_rate": 1.2394256088836069e-06,
+      "loss": 0.1084,
+      "step": 259
+    },
+    {
+      "epoch": 0.05909090909090909,
+      "grad_norm": 2.0801655730274153,
+      "learning_rate": 1.2393437121850342e-06,
+      "loss": 0.1551,
+      "step": 260
+    },
+    {
+      "epoch": 0.05931818181818182,
+      "grad_norm": 3.217851949386645,
+      "learning_rate": 1.2392615022979723e-06,
+      "loss": 0.1788,
+      "step": 261
+    },
+    {
+      "epoch": 0.05954545454545455,
+      "grad_norm": 3.279940241351343,
+      "learning_rate": 1.2391789792643317e-06,
+      "loss": 0.1331,
+      "step": 262
+    },
+    {
+      "epoch": 0.059772727272727276,
+      "grad_norm": 2.5153381270102573,
+      "learning_rate": 1.2390961431261814e-06,
+      "loss": 0.2077,
+      "step": 263
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.3072227612524547,
+      "learning_rate": 1.2390129939257515e-06,
+      "loss": 0.1651,
+      "step": 264
+    },
+    {
+      "epoch": 0.060227272727272727,
+      "grad_norm": 4.196404705383203,
+      "learning_rate": 1.2389295317054306e-06,
+      "loss": 0.2765,
+      "step": 265
+    },
+    {
+      "epoch": 0.060454545454545455,
+      "grad_norm": 2.809482163800136,
+      "learning_rate": 1.238845756507767e-06,
+      "loss": 0.1187,
+      "step": 266
+    },
+    {
+      "epoch": 0.060681818181818184,
+      "grad_norm": 3.1723897177592018,
+      "learning_rate": 1.2387616683754691e-06,
+      "loss": 0.1293,
+      "step": 267
+    },
+    {
+      "epoch": 0.060909090909090906,
+      "grad_norm": 2.272359295307085,
+      "learning_rate": 1.2386772673514044e-06,
+      "loss": 0.0629,
+      "step": 268
+    },
+    {
+      "epoch": 0.061136363636363635,
+      "grad_norm": 3.5320897525199983,
+      "learning_rate": 1.2385925534786e-06,
+      "loss": 0.1277,
+      "step": 269
+    },
+    {
+      "epoch": 0.06136363636363636,
+      "grad_norm": 2.742439382587011,
+      "learning_rate": 1.2385075268002423e-06,
+      "loss": 0.1185,
+      "step": 270
+    },
+    {
+      "epoch": 0.06159090909090909,
+      "grad_norm": 1.8201996571872956,
+      "learning_rate": 1.2384221873596775e-06,
+      "loss": 0.0689,
+      "step": 271
+    },
+    {
+      "epoch": 0.06181818181818182,
+      "grad_norm": 2.882446419696092,
+      "learning_rate": 1.2383365352004111e-06,
+      "loss": 0.1574,
+      "step": 272
+    },
+    {
+      "epoch": 0.06204545454545454,
+      "grad_norm": 2.070694394592025,
+      "learning_rate": 1.238250570366108e-06,
+      "loss": 0.1796,
+      "step": 273
+    },
+    {
+      "epoch": 0.06227272727272727,
+      "grad_norm": 2.1441299062815786,
+      "learning_rate": 1.2381642929005927e-06,
+      "loss": 0.0977,
+      "step": 274
+    },
+    {
+      "epoch": 0.0625,
+      "grad_norm": 5.18202165704,
+      "learning_rate": 1.238077702847849e-06,
+      "loss": 0.2323,
+      "step": 275
+    },
+    {
+      "epoch": 0.06272727272727273,
+      "grad_norm": 2.991722368826215,
+      "learning_rate": 1.2379908002520191e-06,
+      "loss": 0.2371,
+      "step": 276
+    },
+    {
+      "epoch": 0.06295454545454546,
+      "grad_norm": 1.6394364483767132,
+      "learning_rate": 1.2379035851574063e-06,
+      "loss": 0.0825,
+      "step": 277
+    },
+    {
+      "epoch": 0.06318181818181819,
+      "grad_norm": 1.5635113644989898,
+      "learning_rate": 1.237816057608472e-06,
+      "loss": 0.1449,
+      "step": 278
+    },
+    {
+      "epoch": 0.06340909090909091,
+      "grad_norm": 2.926323946209762,
+      "learning_rate": 1.2377282176498371e-06,
+      "loss": 0.0842,
+      "step": 279
+    },
+    {
+      "epoch": 0.06363636363636363,
+      "grad_norm": 3.2988249649164745,
+      "learning_rate": 1.2376400653262817e-06,
+      "loss": 0.2047,
+      "step": 280
+    },
+    {
+      "epoch": 0.06386363636363636,
+      "grad_norm": 1.8062602999052046,
+      "learning_rate": 1.237551600682746e-06,
+      "loss": 0.103,
+      "step": 281
+    },
+    {
+      "epoch": 0.06409090909090909,
+      "grad_norm": 3.877639222621489,
+      "learning_rate": 1.237462823764328e-06,
+      "loss": 0.2029,
+      "step": 282
+    },
+    {
+      "epoch": 0.06431818181818182,
+      "grad_norm": 2.3115690671067717,
+      "learning_rate": 1.2373737346162857e-06,
+      "loss": 0.0897,
+      "step": 283
+    },
+    {
+      "epoch": 0.06454545454545454,
+      "grad_norm": 2.165961154728473,
+      "learning_rate": 1.2372843332840364e-06,
+      "loss": 0.066,
+      "step": 284
+    },
+    {
+      "epoch": 0.06477272727272727,
+      "grad_norm": 2.927832101133683,
+      "learning_rate": 1.2371946198131563e-06,
+      "loss": 0.0819,
+      "step": 285
+    },
+    {
+      "epoch": 0.065,
+      "grad_norm": 1.7112451649261085,
+      "learning_rate": 1.2371045942493804e-06,
+      "loss": 0.1292,
+      "step": 286
+    },
+    {
+      "epoch": 0.06522727272727273,
+      "grad_norm": 2.5670797557684115,
+      "learning_rate": 1.2370142566386038e-06,
+      "loss": 0.1504,
+      "step": 287
+    },
+    {
+      "epoch": 0.06545454545454546,
+      "grad_norm": 2.8467909901989437,
+      "learning_rate": 1.2369236070268795e-06,
+      "loss": 0.09,
+      "step": 288
+    },
+    {
+      "epoch": 0.06568181818181819,
+      "grad_norm": 3.216612262377372,
+      "learning_rate": 1.2368326454604201e-06,
+      "loss": 0.2207,
+      "step": 289
+    },
+    {
+      "epoch": 0.0659090909090909,
+      "grad_norm": 2.0416512948368952,
+      "learning_rate": 1.2367413719855976e-06,
+      "loss": 0.0783,
+      "step": 290
+    },
+    {
+      "epoch": 0.06613636363636363,
+      "grad_norm": 2.2414828028989975,
+      "learning_rate": 1.2366497866489423e-06,
+      "loss": 0.0826,
+      "step": 291
+    },
+    {
+      "epoch": 0.06636363636363636,
+      "grad_norm": 1.827322332252562,
+      "learning_rate": 1.236557889497144e-06,
+      "loss": 0.0883,
+      "step": 292
+    },
+    {
+      "epoch": 0.06659090909090909,
+      "grad_norm": 3.4443353783113224,
+      "learning_rate": 1.236465680577051e-06,
+      "loss": 0.1443,
+      "step": 293
+    },
+    {
+      "epoch": 0.06681818181818182,
+      "grad_norm": 2.476104021369502,
+      "learning_rate": 1.2363731599356712e-06,
+      "loss": 0.1739,
+      "step": 294
+    },
+    {
+      "epoch": 0.06704545454545455,
+      "grad_norm": 2.0423040201770273,
+      "learning_rate": 1.2362803276201709e-06,
+      "loss": 0.1899,
+      "step": 295
+    },
+    {
+      "epoch": 0.06727272727272728,
+      "grad_norm": 3.8198397151574905,
+      "learning_rate": 1.2361871836778755e-06,
+      "loss": 0.1068,
+      "step": 296
+    },
+    {
+      "epoch": 0.0675,
+      "grad_norm": 2.9893679696642153,
+      "learning_rate": 1.236093728156269e-06,
+      "loss": 0.1486,
+      "step": 297
+    },
+    {
+      "epoch": 0.06772727272727273,
+      "grad_norm": 4.085951333590418,
+      "learning_rate": 1.2359999611029944e-06,
+      "loss": 0.1827,
+      "step": 298
+    },
+    {
+      "epoch": 0.06795454545454545,
+      "grad_norm": 3.658834732115434,
+      "learning_rate": 1.2359058825658534e-06,
+      "loss": 0.1052,
+      "step": 299
+    },
+    {
+      "epoch": 0.06818181818181818,
+      "grad_norm": 2.626169684349872,
+      "learning_rate": 1.2358114925928073e-06,
+      "loss": 0.1267,
+      "step": 300
+    },
+    {
+      "epoch": 0.0684090909090909,
+      "grad_norm": 2.8805508572121306,
+      "learning_rate": 1.2357167912319747e-06,
+      "loss": 0.1357,
+      "step": 301
+    },
+    {
+      "epoch": 0.06863636363636363,
+      "grad_norm": 4.34860492379719,
+      "learning_rate": 1.2356217785316344e-06,
+      "loss": 0.1184,
+      "step": 302
+    },
+    {
+      "epoch": 0.06886363636363636,
+      "grad_norm": 2.501758847846472,
+      "learning_rate": 1.235526454540223e-06,
+      "loss": 0.2022,
+      "step": 303
+    },
+    {
+      "epoch": 0.06909090909090909,
+      "grad_norm": 1.3128613725879015,
+      "learning_rate": 1.2354308193063358e-06,
+      "loss": 0.0892,
+      "step": 304
+    },
+    {
+      "epoch": 0.06931818181818182,
+      "grad_norm": 4.147832425558623,
+      "learning_rate": 1.2353348728787274e-06,
+      "loss": 0.1951,
+      "step": 305
+    },
+    {
+      "epoch": 0.06954545454545455,
+      "grad_norm": 2.9686841516574964,
+      "learning_rate": 1.2352386153063107e-06,
+      "loss": 0.0955,
+      "step": 306
+    },
+    {
+      "epoch": 0.06977272727272728,
+      "grad_norm": 5.280765555736939,
+      "learning_rate": 1.2351420466381566e-06,
+      "loss": 0.1596,
+      "step": 307
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2413992389068813,
+      "learning_rate": 1.235045166923496e-06,
+      "loss": 0.1776,
+      "step": 308
+    },
+    {
+      "epoch": 0.07022727272727272,
+      "grad_norm": 1.757628540101853,
+      "learning_rate": 1.2349479762117171e-06,
+      "loss": 0.1276,
+      "step": 309
+    },
+    {
+      "epoch": 0.07045454545454545,
+      "grad_norm": 2.8669701290580307,
+      "learning_rate": 1.2348504745523673e-06,
+      "loss": 0.1815,
+      "step": 310
+    },
+    {
+      "epoch": 0.07068181818181818,
+      "grad_norm": 3.1779430430829425,
+      "learning_rate": 1.2347526619951523e-06,
+      "loss": 0.1936,
+      "step": 311
+    },
+    {
+      "epoch": 0.07090909090909091,
+      "grad_norm": 4.795727939021179,
+      "learning_rate": 1.2346545385899358e-06,
+      "loss": 0.2285,
+      "step": 312
+    },
+    {
+      "epoch": 0.07113636363636364,
+      "grad_norm": 3.352302580258064,
+      "learning_rate": 1.2345561043867413e-06,
+      "loss": 0.2239,
+      "step": 313
+    },
+    {
+      "epoch": 0.07136363636363637,
+      "grad_norm": 2.633011525783662,
+      "learning_rate": 1.2344573594357493e-06,
+      "loss": 0.0913,
+      "step": 314
+    },
+    {
+      "epoch": 0.0715909090909091,
+      "grad_norm": 2.227503118186183,
+      "learning_rate": 1.2343583037872998e-06,
+      "loss": 0.1207,
+      "step": 315
+    },
+    {
+      "epoch": 0.07181818181818182,
+      "grad_norm": 2.0062788668276696,
+      "learning_rate": 1.2342589374918905e-06,
+      "loss": 0.1614,
+      "step": 316
+    },
+    {
+      "epoch": 0.07204545454545455,
+      "grad_norm": 2.0924475520713264,
+      "learning_rate": 1.2341592606001777e-06,
+      "loss": 0.1747,
+      "step": 317
+    },
+    {
+      "epoch": 0.07227272727272727,
+      "grad_norm": 2.5495818852244603,
+      "learning_rate": 1.2340592731629758e-06,
+      "loss": 0.0934,
+      "step": 318
+    },
+    {
+      "epoch": 0.0725,
+      "grad_norm": 1.761711317107103,
+      "learning_rate": 1.2339589752312581e-06,
+      "loss": 0.0936,
+      "step": 319
+    },
+    {
+      "epoch": 0.07272727272727272,
+      "grad_norm": 3.961796072791857,
+      "learning_rate": 1.233858366856156e-06,
+      "loss": 0.1754,
+      "step": 320
+    },
+    {
+      "epoch": 0.07295454545454545,
+      "grad_norm": 3.145423688488847,
+      "learning_rate": 1.2337574480889585e-06,
+      "loss": 0.1537,
+      "step": 321
+    },
+    {
+      "epoch": 0.07318181818181818,
+      "grad_norm": 2.101484613453523,
+      "learning_rate": 1.2336562189811138e-06,
+      "loss": 0.0911,
+      "step": 322
+    },
+    {
+      "epoch": 0.07340909090909091,
+      "grad_norm": 2.46447192761943,
+      "learning_rate": 1.2335546795842276e-06,
+      "loss": 0.1036,
+      "step": 323
+    },
+    {
+      "epoch": 0.07363636363636364,
+      "grad_norm": 3.534983556887946,
+      "learning_rate": 1.233452829950064e-06,
+      "loss": 0.1334,
+      "step": 324
+    },
+    {
+      "epoch": 0.07386363636363637,
+      "grad_norm": 2.7817621814022355,
+      "learning_rate": 1.2333506701305453e-06,
+      "loss": 0.0954,
+      "step": 325
+    },
+    {
+      "epoch": 0.0740909090909091,
+      "grad_norm": 2.111869962081062,
+      "learning_rate": 1.2332482001777522e-06,
+      "loss": 0.1249,
+      "step": 326
+    },
+    {
+      "epoch": 0.07431818181818182,
+      "grad_norm": 3.424566110103121,
+      "learning_rate": 1.233145420143923e-06,
+      "loss": 0.1586,
+      "step": 327
+    },
+    {
+      "epoch": 0.07454545454545454,
+      "grad_norm": 2.719972547313337,
+      "learning_rate": 1.2330423300814542e-06,
+      "loss": 0.1655,
+      "step": 328
+    },
+    {
+      "epoch": 0.07477272727272727,
+      "grad_norm": 2.677764091353117,
+      "learning_rate": 1.2329389300429008e-06,
+      "loss": 0.1138,
+      "step": 329
+    },
+    {
+      "epoch": 0.075,
+      "grad_norm": 1.5576583036385727,
+      "learning_rate": 1.232835220080975e-06,
+      "loss": 0.0811,
+      "step": 330
+    },
+    {
+      "epoch": 0.07522727272727273,
+      "grad_norm": 3.7983085662803244,
+      "learning_rate": 1.232731200248548e-06,
+      "loss": 0.2137,
+      "step": 331
+    },
+    {
+      "epoch": 0.07545454545454545,
+      "grad_norm": 3.9004837425382006,
+      "learning_rate": 1.232626870598648e-06,
+      "loss": 0.1261,
+      "step": 332
+    },
+    {
+      "epoch": 0.07568181818181818,
+      "grad_norm": 2.593173187104484,
+      "learning_rate": 1.2325222311844617e-06,
+      "loss": 0.2032,
+      "step": 333
+    },
+    {
+      "epoch": 0.07590909090909091,
+      "grad_norm": 3.8083336918082145,
+      "learning_rate": 1.2324172820593339e-06,
+      "loss": 0.1305,
+      "step": 334
+    },
+    {
+      "epoch": 0.07613636363636364,
+      "grad_norm": 2.988832201004671,
+      "learning_rate": 1.2323120232767667e-06,
+      "loss": 0.1195,
+      "step": 335
+    },
+    {
+      "epoch": 0.07636363636363637,
+      "grad_norm": 2.8096160685127645,
+      "learning_rate": 1.2322064548904202e-06,
+      "loss": 0.0982,
+      "step": 336
+    },
+    {
+      "epoch": 0.07659090909090908,
+      "grad_norm": 3.139227749011756,
+      "learning_rate": 1.232100576954113e-06,
+      "loss": 0.1046,
+      "step": 337
+    },
+    {
+      "epoch": 0.07681818181818181,
+      "grad_norm": 1.6844911358397119,
+      "learning_rate": 1.2319943895218205e-06,
+      "loss": 0.1169,
+      "step": 338
+    },
+    {
+      "epoch": 0.07704545454545454,
+      "grad_norm": 1.7417897204536763,
+      "learning_rate": 1.2318878926476765e-06,
+      "loss": 0.1869,
+      "step": 339
+    },
+    {
+      "epoch": 0.07727272727272727,
+      "grad_norm": 2.421522727702754,
+      "learning_rate": 1.2317810863859728e-06,
+      "loss": 0.1079,
+      "step": 340
+    },
+    {
+      "epoch": 0.0775,
+      "grad_norm": 2.6726505222229746,
+      "learning_rate": 1.231673970791158e-06,
+      "loss": 0.1258,
+      "step": 341
+    },
+    {
+      "epoch": 0.07772727272727273,
+      "grad_norm": 3.051584557057126,
+      "learning_rate": 1.231566545917839e-06,
+      "loss": 0.1783,
+      "step": 342
+    },
+    {
+      "epoch": 0.07795454545454546,
+      "grad_norm": 4.604500510357601,
+      "learning_rate": 1.2314588118207808e-06,
+      "loss": 0.1911,
+      "step": 343
+    },
+    {
+      "epoch": 0.07818181818181819,
+      "grad_norm": 3.206923994962591,
+      "learning_rate": 1.2313507685549054e-06,
+      "loss": 0.0929,
+      "step": 344
+    },
+    {
+      "epoch": 0.07840909090909091,
+      "grad_norm": 2.5514106614836973,
+      "learning_rate": 1.231242416175292e-06,
+      "loss": 0.0942,
+      "step": 345
+    },
+    {
+      "epoch": 0.07863636363636364,
+      "grad_norm": 3.314279988350679,
+      "learning_rate": 1.2311337547371785e-06,
+      "loss": 0.1646,
+      "step": 346
+    },
+    {
+      "epoch": 0.07886363636363636,
+      "grad_norm": 3.330173754439046,
+      "learning_rate": 1.2310247842959597e-06,
+      "loss": 0.2494,
+      "step": 347
+    },
+    {
+      "epoch": 0.07909090909090909,
+      "grad_norm": 2.2320164202385793,
+      "learning_rate": 1.230915504907188e-06,
+      "loss": 0.0925,
+      "step": 348
+    },
+    {
+      "epoch": 0.07931818181818182,
+      "grad_norm": 5.493228887892353,
+      "learning_rate": 1.2308059166265734e-06,
+      "loss": 0.2144,
+      "step": 349
+    },
+    {
+      "epoch": 0.07954545454545454,
+      "grad_norm": 1.6132105727497863,
+      "learning_rate": 1.2306960195099833e-06,
+      "loss": 0.0767,
+      "step": 350
+    },
+    {
+      "epoch": 0.07977272727272727,
+      "grad_norm": 3.186030629569304,
+      "learning_rate": 1.2305858136134422e-06,
+      "loss": 0.1736,
+      "step": 351
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9746632216828104,
+      "learning_rate": 1.2304752989931327e-06,
+      "loss": 0.1414,
+      "step": 352
+    },
+    {
+      "epoch": 0.08022727272727273,
+      "grad_norm": 2.7167161743987926,
+      "learning_rate": 1.2303644757053945e-06,
+      "loss": 0.1385,
+      "step": 353
+    },
+    {
+      "epoch": 0.08045454545454546,
+      "grad_norm": 1.8988997342639247,
+      "learning_rate": 1.2302533438067247e-06,
+      "loss": 0.1431,
+      "step": 354
+    },
+    {
+      "epoch": 0.08068181818181819,
+      "grad_norm": 3.1839309995414604,
+      "learning_rate": 1.230141903353777e-06,
+      "loss": 0.1253,
+      "step": 355
+    },
+    {
+      "epoch": 0.0809090909090909,
+      "grad_norm": 1.7780877183044514,
+      "learning_rate": 1.2300301544033636e-06,
+      "loss": 0.0929,
+      "step": 356
+    },
+    {
+      "epoch": 0.08113636363636363,
+      "grad_norm": 2.1204212242316385,
+      "learning_rate": 1.2299180970124533e-06,
+      "loss": 0.1398,
+      "step": 357
+    },
+    {
+      "epoch": 0.08136363636363636,
+      "grad_norm": 1.8549587886559258,
+      "learning_rate": 1.2298057312381723e-06,
+      "loss": 0.1058,
+      "step": 358
+    },
+    {
+      "epoch": 0.08159090909090909,
+      "grad_norm": 4.913124012193084,
+      "learning_rate": 1.2296930571378035e-06,
+      "loss": 0.1564,
+      "step": 359
+    },
+    {
+      "epoch": 0.08181818181818182,
+      "grad_norm": 1.316925981658509,
+      "learning_rate": 1.229580074768788e-06,
+      "loss": 0.0993,
+      "step": 360
+    },
+    {
+      "epoch": 0.08204545454545455,
+      "grad_norm": 2.828696579317692,
+      "learning_rate": 1.2294667841887234e-06,
+      "loss": 0.1417,
+      "step": 361
+    },
+    {
+      "epoch": 0.08227272727272728,
+      "grad_norm": 3.5307662503778494,
+      "learning_rate": 1.2293531854553642e-06,
+      "loss": 0.1003,
+      "step": 362
+    },
+    {
+      "epoch": 0.0825,
+      "grad_norm": 1.7003063487557772,
+      "learning_rate": 1.2292392786266225e-06,
+      "loss": 0.1281,
+      "step": 363
+    },
+    {
+      "epoch": 0.08272727272727273,
+      "grad_norm": 3.173107868105214,
+      "learning_rate": 1.2291250637605672e-06,
+      "loss": 0.0973,
+      "step": 364
+    },
+    {
+      "epoch": 0.08295454545454546,
+      "grad_norm": 3.328478247063708,
+      "learning_rate": 1.2290105409154244e-06,
+      "loss": 0.0903,
+      "step": 365
+    },
+    {
+      "epoch": 0.08318181818181818,
+      "grad_norm": 3.1923943860860162,
+      "learning_rate": 1.2288957101495772e-06,
+      "loss": 0.099,
+      "step": 366
+    },
+    {
+      "epoch": 0.0834090909090909,
+      "grad_norm": 3.060692470506418,
+      "learning_rate": 1.2287805715215651e-06,
+      "loss": 0.1039,
+      "step": 367
+    },
+    {
+      "epoch": 0.08363636363636363,
+      "grad_norm": 3.543698132513649,
+      "learning_rate": 1.2286651250900858e-06,
+      "loss": 0.1132,
+      "step": 368
+    },
+    {
+      "epoch": 0.08386363636363636,
+      "grad_norm": 3.7186773886717934,
+      "learning_rate": 1.2285493709139925e-06,
+      "loss": 0.1675,
+      "step": 369
+    },
+    {
+      "epoch": 0.08409090909090909,
+      "grad_norm": 2.7351099292210166,
+      "learning_rate": 1.2284333090522962e-06,
+      "loss": 0.0945,
+      "step": 370
+    },
+    {
+      "epoch": 0.08431818181818182,
+      "grad_norm": 3.950774147487926,
+      "learning_rate": 1.2283169395641647e-06,
+      "loss": 0.1142,
+      "step": 371
+    },
+    {
+      "epoch": 0.08454545454545455,
+      "grad_norm": 1.9837415596917587,
+      "learning_rate": 1.228200262508922e-06,
+      "loss": 0.1247,
+      "step": 372
+    },
+    {
+      "epoch": 0.08477272727272728,
+      "grad_norm": 4.24524958508508,
+      "learning_rate": 1.2280832779460494e-06,
+      "loss": 0.1837,
+      "step": 373
+    },
+    {
+      "epoch": 0.085,
+      "grad_norm": 1.4695248810613124,
+      "learning_rate": 1.2279659859351853e-06,
+      "loss": 0.1056,
+      "step": 374
+    },
+    {
+      "epoch": 0.08522727272727272,
+      "grad_norm": 3.9479862024848464,
+      "learning_rate": 1.2278483865361239e-06,
+      "loss": 0.1293,
+      "step": 375
+    },
+    {
+      "epoch": 0.08545454545454545,
+      "grad_norm": 1.5125555431997273,
+      "learning_rate": 1.227730479808817e-06,
+      "loss": 0.0395,
+      "step": 376
+    },
+    {
+      "epoch": 0.08568181818181818,
+      "grad_norm": 3.5278832573637517,
+      "learning_rate": 1.2276122658133723e-06,
+      "loss": 0.0972,
+      "step": 377
+    },
+    {
+      "epoch": 0.08590909090909091,
+      "grad_norm": 2.792201995095852,
+      "learning_rate": 1.2274937446100548e-06,
+      "loss": 0.117,
+      "step": 378
+    },
+    {
+      "epoch": 0.08613636363636364,
+      "grad_norm": 4.206107256448596,
+      "learning_rate": 1.227374916259286e-06,
+      "loss": 0.1254,
+      "step": 379
+    },
+    {
+      "epoch": 0.08636363636363636,
+      "grad_norm": 2.7883818456718203,
+      "learning_rate": 1.2272557808216433e-06,
+      "loss": 0.109,
+      "step": 380
+    },
+    {
+      "epoch": 0.0865909090909091,
+      "grad_norm": 3.1722449636123407,
+      "learning_rate": 1.2271363383578619e-06,
+      "loss": 0.0945,
+      "step": 381
+    },
+    {
+      "epoch": 0.08681818181818182,
+      "grad_norm": 2.0183961894181888,
+      "learning_rate": 1.2270165889288325e-06,
+      "loss": 0.1203,
+      "step": 382
+    },
+    {
+      "epoch": 0.08704545454545455,
+      "grad_norm": 2.8124095180254196,
+      "learning_rate": 1.2268965325956022e-06,
+      "loss": 0.0792,
+      "step": 383
+    },
+    {
+      "epoch": 0.08727272727272728,
+      "grad_norm": 3.568789457511722,
+      "learning_rate": 1.2267761694193752e-06,
+      "loss": 0.1344,
+      "step": 384
+    },
+    {
+      "epoch": 0.0875,
+      "grad_norm": 5.124744434980692,
+      "learning_rate": 1.2266554994615121e-06,
+      "loss": 0.1313,
+      "step": 385
+    },
+    {
+      "epoch": 0.08772727272727272,
+      "grad_norm": 2.5983615591028246,
+      "learning_rate": 1.2265345227835295e-06,
+      "loss": 0.1633,
+      "step": 386
+    },
+    {
+      "epoch": 0.08795454545454545,
+      "grad_norm": 1.784365549960095,
+      "learning_rate": 1.2264132394471007e-06,
+      "loss": 0.0968,
+      "step": 387
+    },
+    {
+      "epoch": 0.08818181818181818,
+      "grad_norm": 2.7499936827992904,
+      "learning_rate": 1.226291649514055e-06,
+      "loss": 0.0849,
+      "step": 388
+    },
+    {
+      "epoch": 0.08840909090909091,
+      "grad_norm": 3.2688251516969173,
+      "learning_rate": 1.226169753046378e-06,
+      "loss": 0.1984,
+      "step": 389
+    },
+    {
+      "epoch": 0.08863636363636364,
+      "grad_norm": 1.8986416717079753,
+      "learning_rate": 1.2260475501062121e-06,
+      "loss": 0.0834,
+      "step": 390
+    },
+    {
+      "epoch": 0.08886363636363637,
+      "grad_norm": 2.243416137456035,
+      "learning_rate": 1.2259250407558553e-06,
+      "loss": 0.1042,
+      "step": 391
+    },
+    {
+      "epoch": 0.0890909090909091,
+      "grad_norm": 2.5084698733482167,
+      "learning_rate": 1.2258022250577622e-06,
+      "loss": 0.0956,
+      "step": 392
+    },
+    {
+      "epoch": 0.08931818181818182,
+      "grad_norm": 2.6992596609034516,
+      "learning_rate": 1.2256791030745434e-06,
+      "loss": 0.1258,
+      "step": 393
+    },
+    {
+      "epoch": 0.08954545454545454,
+      "grad_norm": 2.7173423689572096,
+      "learning_rate": 1.225555674868966e-06,
+      "loss": 0.1273,
+      "step": 394
+    },
+    {
+      "epoch": 0.08977272727272727,
+      "grad_norm": 3.937048903571977,
+      "learning_rate": 1.2254319405039524e-06,
+      "loss": 0.1728,
+      "step": 395
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7423951972662466,
+      "learning_rate": 1.2253079000425818e-06,
+      "loss": 0.1106,
+      "step": 396
+    },
+    {
+      "epoch": 0.09022727272727273,
+      "grad_norm": 1.6532407639933144,
+      "learning_rate": 1.2251835535480895e-06,
+      "loss": 0.1434,
+      "step": 397
+    },
+    {
+      "epoch": 0.09045454545454545,
+      "grad_norm": 2.9213544370927202,
+      "learning_rate": 1.2250589010838662e-06,
+      "loss": 0.1482,
+      "step": 398
+    },
+    {
+      "epoch": 0.09068181818181818,
+      "grad_norm": 2.033761133619529,
+      "learning_rate": 1.224933942713459e-06,
+      "loss": 0.1365,
+      "step": 399
+    },
+    {
+      "epoch": 0.09090909090909091,
+      "grad_norm": 3.8716344931091204,
+      "learning_rate": 1.2248086785005709e-06,
+      "loss": 0.1209,
+      "step": 400
+    },
+    {
+      "epoch": 0.09113636363636364,
+      "grad_norm": 2.473253846100544,
+      "learning_rate": 1.2246831085090611e-06,
+      "loss": 0.1569,
+      "step": 401
+    },
+    {
+      "epoch": 0.09136363636363637,
+      "grad_norm": 2.110843601211002,
+      "learning_rate": 1.2245572328029438e-06,
+      "loss": 0.0977,
+      "step": 402
+    },
+    {
+      "epoch": 0.0915909090909091,
+      "grad_norm": 3.4129633911409027,
+      "learning_rate": 1.22443105144639e-06,
+      "loss": 0.2042,
+      "step": 403
+    },
+    {
+      "epoch": 0.09181818181818181,
+      "grad_norm": 2.861079755271948,
+      "learning_rate": 1.224304564503726e-06,
+      "loss": 0.2059,
+      "step": 404
+    },
+    {
+      "epoch": 0.09204545454545454,
+      "grad_norm": 3.2130254765654485,
+      "learning_rate": 1.224177772039434e-06,
+      "loss": 0.1349,
+      "step": 405
+    },
+    {
+      "epoch": 0.09227272727272727,
+      "grad_norm": 2.5626113263600816,
+      "learning_rate": 1.2240506741181523e-06,
+      "loss": 0.1633,
+      "step": 406
+    },
+    {
+      "epoch": 0.0925,
+      "grad_norm": 3.327617410472412,
+      "learning_rate": 1.2239232708046745e-06,
+      "loss": 0.099,
+      "step": 407
+    },
+    {
+      "epoch": 0.09272727272727273,
+      "grad_norm": 3.375302501348502,
+      "learning_rate": 1.2237955621639496e-06,
+      "loss": 0.1827,
+      "step": 408
+    },
+    {
+      "epoch": 0.09295454545454546,
+      "grad_norm": 3.2906592289295977,
+      "learning_rate": 1.223667548261083e-06,
+      "loss": 0.1353,
+      "step": 409
+    },
+    {
+      "epoch": 0.09318181818181819,
+      "grad_norm": 3.138608940337265,
+      "learning_rate": 1.2235392291613353e-06,
+      "loss": 0.1383,
+      "step": 410
+    },
+    {
+      "epoch": 0.09340909090909091,
+      "grad_norm": 3.4729487641130645,
+      "learning_rate": 1.2234106049301228e-06,
+      "loss": 0.118,
+      "step": 411
+    },
+    {
+      "epoch": 0.09363636363636364,
+      "grad_norm": 3.6371170106616537,
+      "learning_rate": 1.2232816756330173e-06,
+      "loss": 0.148,
+      "step": 412
+    },
+    {
+      "epoch": 0.09386363636363636,
+      "grad_norm": 2.1816440673879964,
+      "learning_rate": 1.223152441335746e-06,
+      "loss": 0.1,
+      "step": 413
+    },
+    {
+      "epoch": 0.09409090909090909,
+      "grad_norm": 4.083313885951214,
+      "learning_rate": 1.223022902104192e-06,
+      "loss": 0.1722,
+      "step": 414
+    },
+    {
+      "epoch": 0.09431818181818181,
+      "grad_norm": 2.2622193186301818,
+      "learning_rate": 1.2228930580043931e-06,
+      "loss": 0.0894,
+      "step": 415
+    },
+    {
+      "epoch": 0.09454545454545454,
+      "grad_norm": 3.310400756032989,
+      "learning_rate": 1.2227629091025437e-06,
+      "loss": 0.138,
+      "step": 416
+    },
+    {
+      "epoch": 0.09477272727272727,
+      "grad_norm": 6.625132888914451,
+      "learning_rate": 1.2226324554649921e-06,
+      "loss": 0.1767,
+      "step": 417
+    },
+    {
+      "epoch": 0.095,
+      "grad_norm": 2.3210375835096895,
+      "learning_rate": 1.222501697158243e-06,
+      "loss": 0.1191,
+      "step": 418
+    },
+    {
+      "epoch": 0.09522727272727273,
+      "grad_norm": 4.773934892610083,
+      "learning_rate": 1.2223706342489565e-06,
+      "loss": 0.1477,
+      "step": 419
+    },
+    {
+      "epoch": 0.09545454545454546,
+      "grad_norm": 4.685103560062665,
+      "learning_rate": 1.222239266803947e-06,
+      "loss": 0.2359,
+      "step": 420
+    },
+    {
+      "epoch": 0.09568181818181819,
+      "grad_norm": 1.4397459693065997,
+      "learning_rate": 1.2221075948901856e-06,
+      "loss": 0.0914,
+      "step": 421
+    },
+    {
+      "epoch": 0.0959090909090909,
+      "grad_norm": 2.922514085572872,
+      "learning_rate": 1.221975618574797e-06,
+      "loss": 0.1364,
+      "step": 422
+    },
+    {
+      "epoch": 0.09613636363636363,
+      "grad_norm": 3.4867490303525948,
+      "learning_rate": 1.2218433379250623e-06,
+      "loss": 0.1368,
+      "step": 423
+    },
+    {
+      "epoch": 0.09636363636363636,
+      "grad_norm": 2.5829654890490743,
+      "learning_rate": 1.2217107530084174e-06,
+      "loss": 0.1386,
+      "step": 424
+    },
+    {
+      "epoch": 0.09659090909090909,
+      "grad_norm": 2.842358737121686,
+      "learning_rate": 1.2215778638924527e-06,
+      "loss": 0.1055,
+      "step": 425
+    },
+    {
+      "epoch": 0.09681818181818182,
+      "grad_norm": 2.7938605213363155,
+      "learning_rate": 1.221444670644915e-06,
+      "loss": 0.0815,
+      "step": 426
+    },
+    {
+      "epoch": 0.09704545454545455,
+      "grad_norm": 3.7962895538956802,
+      "learning_rate": 1.221311173333705e-06,
+      "loss": 0.0981,
+      "step": 427
+    },
+    {
+      "epoch": 0.09727272727272727,
+      "grad_norm": 2.4987425160238925,
+      "learning_rate": 1.2211773720268784e-06,
+      "loss": 0.1013,
+      "step": 428
+    },
+    {
+      "epoch": 0.0975,
+      "grad_norm": 3.0276172616365145,
+      "learning_rate": 1.2210432667926467e-06,
+      "loss": 0.1281,
+      "step": 429
+    },
+    {
+      "epoch": 0.09772727272727273,
+      "grad_norm": 3.1697582811970086,
+      "learning_rate": 1.2209088576993757e-06,
+      "loss": 0.22,
+      "step": 430
+    },
+    {
+      "epoch": 0.09795454545454546,
+      "grad_norm": 2.5248741315941308,
+      "learning_rate": 1.2207741448155867e-06,
+      "loss": 0.0733,
+      "step": 431
+    },
+    {
+      "epoch": 0.09818181818181818,
+      "grad_norm": 3.9524719886765083,
+      "learning_rate": 1.220639128209955e-06,
+      "loss": 0.195,
+      "step": 432
+    },
+    {
+      "epoch": 0.0984090909090909,
+      "grad_norm": 2.209319917992543,
+      "learning_rate": 1.2205038079513113e-06,
+      "loss": 0.1642,
+      "step": 433
+    },
+    {
+      "epoch": 0.09863636363636363,
+      "grad_norm": 2.7857191032778577,
+      "learning_rate": 1.2203681841086409e-06,
+      "loss": 0.0949,
+      "step": 434
+    },
+    {
+      "epoch": 0.09886363636363636,
+      "grad_norm": 2.2805049414735854,
+      "learning_rate": 1.2202322567510843e-06,
+      "loss": 0.0982,
+      "step": 435
+    },
+    {
+      "epoch": 0.09909090909090909,
+      "grad_norm": 4.943265095811902,
+      "learning_rate": 1.2200960259479362e-06,
+      "loss": 0.2074,
+      "step": 436
+    },
+    {
+      "epoch": 0.09931818181818182,
+      "grad_norm": 3.2366469468902115,
+      "learning_rate": 1.219959491768646e-06,
+      "loss": 0.1941,
+      "step": 437
+    },
+    {
+      "epoch": 0.09954545454545455,
+      "grad_norm": 5.657854198189881,
+      "learning_rate": 1.2198226542828183e-06,
+      "loss": 0.1804,
+      "step": 438
+    },
+    {
+      "epoch": 0.09977272727272728,
+      "grad_norm": 1.95064659993853,
+      "learning_rate": 1.219685513560212e-06,
+      "loss": 0.115,
+      "step": 439
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9112726645217122,
+      "learning_rate": 1.2195480696707401e-06,
+      "loss": 0.0618,
+      "step": 440
+    },
+    {
+      "epoch": 0.10022727272727272,
+      "grad_norm": 3.2910770175405526,
+      "learning_rate": 1.2194103226844711e-06,
+      "loss": 0.2117,
+      "step": 441
+    },
+    {
+      "epoch": 0.10045454545454545,
+      "grad_norm": 2.9398765527020716,
+      "learning_rate": 1.2192722726716272e-06,
+      "loss": 0.2561,
+      "step": 442
+    },
+    {
+      "epoch": 0.10068181818181818,
+      "grad_norm": 2.840216818916321,
+      "learning_rate": 1.2191339197025857e-06,
+      "loss": 0.1335,
+      "step": 443
+    },
+    {
+      "epoch": 0.1009090909090909,
+      "grad_norm": 2.067390795349421,
+      "learning_rate": 1.2189952638478778e-06,
+      "loss": 0.1284,
+      "step": 444
+    },
+    {
+      "epoch": 0.10113636363636364,
+      "grad_norm": 4.605486897461516,
+      "learning_rate": 1.2188563051781894e-06,
+      "loss": 0.1291,
+      "step": 445
+    },
+    {
+      "epoch": 0.10136363636363636,
+      "grad_norm": 2.6541650007587445,
+      "learning_rate": 1.2187170437643608e-06,
+      "loss": 0.1127,
+      "step": 446
+    },
+    {
+      "epoch": 0.10159090909090909,
+      "grad_norm": 1.9541237136717886,
+      "learning_rate": 1.2185774796773864e-06,
+      "loss": 0.0898,
+      "step": 447
+    },
+    {
+      "epoch": 0.10181818181818182,
+      "grad_norm": 2.6505058569437825,
+      "learning_rate": 1.2184376129884156e-06,
+      "loss": 0.0915,
+      "step": 448
+    },
+    {
+      "epoch": 0.10204545454545455,
+      "grad_norm": 2.551185533125964,
+      "learning_rate": 1.2182974437687512e-06,
+      "loss": 0.1059,
+      "step": 449
+    },
+    {
+      "epoch": 0.10227272727272728,
+      "grad_norm": 2.190339982460119,
+      "learning_rate": 1.2181569720898503e-06,
+      "loss": 0.0842,
+      "step": 450
+    },
+    {
+      "epoch": 0.1025,
+      "grad_norm": 3.8443821400026104,
+      "learning_rate": 1.2180161980233249e-06,
+      "loss": 0.2465,
+      "step": 451
+    },
+    {
+      "epoch": 0.10272727272727272,
+      "grad_norm": 4.104637593999546,
+      "learning_rate": 1.2178751216409404e-06,
+      "loss": 0.1724,
+      "step": 452
+    },
+    {
+      "epoch": 0.10295454545454545,
+      "grad_norm": 3.2070647304648228,
+      "learning_rate": 1.217733743014617e-06,
+      "loss": 0.1452,
+      "step": 453
+    },
+    {
+      "epoch": 0.10318181818181818,
+      "grad_norm": 3.574129525787763,
+      "learning_rate": 1.2175920622164284e-06,
+      "loss": 0.1631,
+      "step": 454
+    },
+    {
+      "epoch": 0.10340909090909091,
+      "grad_norm": 1.91434921246329,
+      "learning_rate": 1.2174500793186024e-06,
+      "loss": 0.1585,
+      "step": 455
+    },
+    {
+      "epoch": 0.10363636363636364,
+      "grad_norm": 1.845202494070972,
+      "learning_rate": 1.2173077943935212e-06,
+      "loss": 0.1533,
+      "step": 456
+    },
+    {
+      "epoch": 0.10386363636363637,
+      "grad_norm": 6.601021962309287,
+      "learning_rate": 1.2171652075137209e-06,
+      "loss": 0.1769,
+      "step": 457
+    },
+    {
+      "epoch": 0.1040909090909091,
+      "grad_norm": 2.2477135348216506,
+      "learning_rate": 1.2170223187518908e-06,
+      "loss": 0.1363,
+      "step": 458
+    },
+    {
+      "epoch": 0.10431818181818182,
+      "grad_norm": 4.252331158195109,
+      "learning_rate": 1.216879128180875e-06,
+      "loss": 0.1201,
+      "step": 459
+    },
+    {
+      "epoch": 0.10454545454545454,
+      "grad_norm": 2.36006859742249,
+      "learning_rate": 1.2167356358736714e-06,
+      "loss": 0.0812,
+      "step": 460
+    },
+    {
+      "epoch": 0.10477272727272727,
+      "grad_norm": 1.8118387079859375,
+      "learning_rate": 1.2165918419034312e-06,
+      "loss": 0.0814,
+      "step": 461
+    },
+    {
+      "epoch": 0.105,
+      "grad_norm": 3.221545845418636,
+      "learning_rate": 1.2164477463434599e-06,
+      "loss": 0.1185,
+      "step": 462
+    },
+    {
+      "epoch": 0.10522727272727272,
+      "grad_norm": 3.18426474899419,
+      "learning_rate": 1.216303349267216e-06,
+      "loss": 0.1217,
+      "step": 463
+    },
+    {
+      "epoch": 0.10545454545454545,
+      "grad_norm": 1.956310191265291,
+      "learning_rate": 1.2161586507483126e-06,
+      "loss": 0.1436,
+      "step": 464
+    },
+    {
+      "epoch": 0.10568181818181818,
+      "grad_norm": 3.701264348511931,
+      "learning_rate": 1.2160136508605156e-06,
+      "loss": 0.1244,
+      "step": 465
+    },
+    {
+      "epoch": 0.10590909090909091,
+      "grad_norm": 3.039201225257589,
+      "learning_rate": 1.2158683496777457e-06,
+      "loss": 0.1192,
+      "step": 466
+    },
+    {
+      "epoch": 0.10613636363636364,
+      "grad_norm": 3.468836655817209,
+      "learning_rate": 1.215722747274076e-06,
+      "loss": 0.1405,
+      "step": 467
+    },
+    {
+      "epoch": 0.10636363636363637,
+      "grad_norm": 3.4210104798821335,
+      "learning_rate": 1.2155768437237342e-06,
+      "loss": 0.1984,
+      "step": 468
+    },
+    {
+      "epoch": 0.1065909090909091,
+      "grad_norm": 2.921845492744074,
+      "learning_rate": 1.2154306391011003e-06,
+      "loss": 0.1612,
+      "step": 469
+    },
+    {
+      "epoch": 0.10681818181818181,
+      "grad_norm": 3.7934850171718204,
+      "learning_rate": 1.215284133480709e-06,
+      "loss": 0.1426,
+      "step": 470
+    },
+    {
+      "epoch": 0.10704545454545454,
+      "grad_norm": 3.687141028611211,
+      "learning_rate": 1.2151373269372476e-06,
+      "loss": 0.2455,
+      "step": 471
+    },
+    {
+      "epoch": 0.10727272727272727,
+      "grad_norm": 1.573995179180965,
+      "learning_rate": 1.2149902195455574e-06,
+      "loss": 0.1049,
+      "step": 472
+    },
+    {
+      "epoch": 0.1075,
+      "grad_norm": 2.1216009281028607,
+      "learning_rate": 1.2148428113806326e-06,
+      "loss": 0.1743,
+      "step": 473
+    },
+    {
+      "epoch": 0.10772727272727273,
+      "grad_norm": 2.44223721026973,
+      "learning_rate": 1.214695102517621e-06,
+      "loss": 0.1354,
+      "step": 474
+    },
+    {
+      "epoch": 0.10795454545454546,
+      "grad_norm": 3.3874508247547532,
+      "learning_rate": 1.214547093031824e-06,
+      "loss": 0.1549,
+      "step": 475
+    },
+    {
+      "epoch": 0.10818181818181818,
+      "grad_norm": 1.9314908859928415,
+      "learning_rate": 1.2143987829986953e-06,
+      "loss": 0.1125,
+      "step": 476
+    },
+    {
+      "epoch": 0.10840909090909091,
+      "grad_norm": 1.894094887161704,
+      "learning_rate": 1.2142501724938425e-06,
+      "loss": 0.1676,
+      "step": 477
+    },
+    {
+      "epoch": 0.10863636363636364,
+      "grad_norm": 1.8828162228030776,
+      "learning_rate": 1.2141012615930266e-06,
+      "loss": 0.0693,
+      "step": 478
+    },
+    {
+      "epoch": 0.10886363636363636,
+      "grad_norm": 3.449195922612347,
+      "learning_rate": 1.2139520503721614e-06,
+      "loss": 0.2871,
+      "step": 479
+    },
+    {
+      "epoch": 0.10909090909090909,
+      "grad_norm": 3.036524839362684,
+      "learning_rate": 1.2138025389073133e-06,
+      "loss": 0.115,
+      "step": 480
+    },
+    {
+      "epoch": 0.10931818181818181,
+      "grad_norm": 4.576687312747585,
+      "learning_rate": 1.2136527272747032e-06,
+      "loss": 0.1254,
+      "step": 481
+    },
+    {
+      "epoch": 0.10954545454545454,
+      "grad_norm": 3.9659075851539485,
+      "learning_rate": 1.2135026155507033e-06,
+      "loss": 0.1764,
+      "step": 482
+    },
+    {
+      "epoch": 0.10977272727272727,
+      "grad_norm": 4.305127998824357,
+      "learning_rate": 1.2133522038118398e-06,
+      "loss": 0.1551,
+      "step": 483
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.940849149203269,
+      "learning_rate": 1.2132014921347917e-06,
+      "loss": 0.1031,
+      "step": 484
+    },
+    {
+      "epoch": 0.11022727272727273,
+      "grad_norm": 1.7787889363481644,
+      "learning_rate": 1.213050480596391e-06,
+      "loss": 0.1148,
+      "step": 485
+    },
+    {
+      "epoch": 0.11045454545454546,
+      "grad_norm": 2.4746077117459575,
+      "learning_rate": 1.2128991692736223e-06,
+      "loss": 0.0867,
+      "step": 486
+    },
+    {
+      "epoch": 0.11068181818181819,
+      "grad_norm": 3.578459985211981,
+      "learning_rate": 1.2127475582436232e-06,
+      "loss": 0.1229,
+      "step": 487
+    },
+    {
+      "epoch": 0.11090909090909092,
+      "grad_norm": 2.8987136912282545,
+      "learning_rate": 1.2125956475836837e-06,
+      "loss": 0.1391,
+      "step": 488
+    },
+    {
+      "epoch": 0.11113636363636363,
+      "grad_norm": 2.179257466214055,
+      "learning_rate": 1.2124434373712473e-06,
+      "loss": 0.1506,
+      "step": 489
+    },
+    {
+      "epoch": 0.11136363636363636,
+      "grad_norm": 3.2862931476506234,
+      "learning_rate": 1.2122909276839095e-06,
+      "loss": 0.1443,
+      "step": 490
+    },
+    {
+      "epoch": 0.11159090909090909,
+      "grad_norm": 2.980983063048763,
+      "learning_rate": 1.2121381185994192e-06,
+      "loss": 0.1271,
+      "step": 491
+    },
+    {
+      "epoch": 0.11181818181818182,
+      "grad_norm": 1.845758134101968,
+      "learning_rate": 1.211985010195677e-06,
+      "loss": 0.0655,
+      "step": 492
+    },
+    {
+      "epoch": 0.11204545454545455,
+      "grad_norm": 1.6928715864406667,
+      "learning_rate": 1.2118316025507369e-06,
+      "loss": 0.098,
+      "step": 493
+    },
+    {
+      "epoch": 0.11227272727272727,
+      "grad_norm": 1.7112690310920473,
+      "learning_rate": 1.2116778957428046e-06,
+      "loss": 0.1522,
+      "step": 494
+    },
+    {
+      "epoch": 0.1125,
+      "grad_norm": 4.292575140097939,
+      "learning_rate": 1.2115238898502395e-06,
+      "loss": 0.1074,
+      "step": 495
+    },
+    {
+      "epoch": 0.11272727272727273,
+      "grad_norm": 1.9950083402002665,
+      "learning_rate": 1.2113695849515527e-06,
+      "loss": 0.1768,
+      "step": 496
+    },
+    {
+      "epoch": 0.11295454545454546,
+      "grad_norm": 4.537042003744556,
+      "learning_rate": 1.2112149811254076e-06,
+      "loss": 0.1832,
+      "step": 497
+    },
+    {
+      "epoch": 0.11318181818181818,
+      "grad_norm": 2.9095376143874376,
+      "learning_rate": 1.2110600784506202e-06,
+      "loss": 0.1235,
+      "step": 498
+    },
+    {
+      "epoch": 0.1134090909090909,
+      "grad_norm": 3.87484602234599,
+      "learning_rate": 1.2109048770061593e-06,
+      "loss": 0.1387,
+      "step": 499
+    },
+    {
+      "epoch": 0.11363636363636363,
+      "grad_norm": 2.5614459802034313,
+      "learning_rate": 1.2107493768711453e-06,
+      "loss": 0.117,
+      "step": 500
+    },
+    {
+      "epoch": 0.11386363636363636,
+      "grad_norm": 2.3068120588944536,
+      "learning_rate": 1.2105935781248513e-06,
+      "loss": 0.1167,
+      "step": 501
+    },
+    {
+      "epoch": 0.11409090909090909,
+      "grad_norm": 3.240528277388154,
+      "learning_rate": 1.2104374808467023e-06,
+      "loss": 0.1783,
+      "step": 502
+    },
+    {
+      "epoch": 0.11431818181818182,
+      "grad_norm": 3.664831460796498,
+      "learning_rate": 1.2102810851162758e-06,
+      "loss": 0.1485,
+      "step": 503
+    },
+    {
+      "epoch": 0.11454545454545455,
+      "grad_norm": 2.7902201224073395,
+      "learning_rate": 1.2101243910133013e-06,
+      "loss": 0.0833,
+      "step": 504
+    },
+    {
+      "epoch": 0.11477272727272728,
+      "grad_norm": 1.7280747084460668,
+      "learning_rate": 1.2099673986176604e-06,
+      "loss": 0.1177,
+      "step": 505
+    },
+    {
+      "epoch": 0.115,
+      "grad_norm": 2.8690784004733616,
+      "learning_rate": 1.2098101080093873e-06,
+      "loss": 0.2038,
+      "step": 506
+    },
+    {
+      "epoch": 0.11522727272727273,
+      "grad_norm": 2.905852633008701,
+      "learning_rate": 1.2096525192686673e-06,
+      "loss": 0.2562,
+      "step": 507
+    },
+    {
+      "epoch": 0.11545454545454545,
+      "grad_norm": 3.226408238742276,
+      "learning_rate": 1.209494632475838e-06,
+      "loss": 0.0997,
+      "step": 508
+    },
+    {
+      "epoch": 0.11568181818181818,
+      "grad_norm": 2.6835813911577544,
+      "learning_rate": 1.2093364477113893e-06,
+      "loss": 0.0953,
+      "step": 509
+    },
+    {
+      "epoch": 0.1159090909090909,
+      "grad_norm": 2.1810994785912254,
+      "learning_rate": 1.2091779650559628e-06,
+      "loss": 0.0903,
+      "step": 510
+    },
+    {
+      "epoch": 0.11613636363636363,
+      "grad_norm": 2.52387505113613,
+      "learning_rate": 1.209019184590352e-06,
+      "loss": 0.1011,
+      "step": 511
+    },
+    {
+      "epoch": 0.11636363636363636,
+      "grad_norm": 4.545750767670771,
+      "learning_rate": 1.2088601063955018e-06,
+      "loss": 0.1517,
+      "step": 512
+    },
+    {
+      "epoch": 0.11659090909090909,
+      "grad_norm": 3.636685087233929,
+      "learning_rate": 1.20870073055251e-06,
+      "loss": 0.1349,
+      "step": 513
+    },
+    {
+      "epoch": 0.11681818181818182,
+      "grad_norm": 3.374342233183646,
+      "learning_rate": 1.2085410571426244e-06,
+      "loss": 0.2035,
+      "step": 514
+    },
+    {
+      "epoch": 0.11704545454545455,
+      "grad_norm": 1.1611250524096484,
+      "learning_rate": 1.208381086247246e-06,
+      "loss": 0.0971,
+      "step": 515
+    },
+    {
+      "epoch": 0.11727272727272728,
+      "grad_norm": 1.7114008100061284,
+      "learning_rate": 1.2082208179479272e-06,
+      "loss": 0.1239,
+      "step": 516
+    },
+    {
+      "epoch": 0.1175,
+      "grad_norm": 4.367587638775703,
+      "learning_rate": 1.2080602523263715e-06,
+      "loss": 0.1794,
+      "step": 517
+    },
+    {
+      "epoch": 0.11772727272727272,
+      "grad_norm": 3.1502195903412744,
+      "learning_rate": 1.207899389464434e-06,
+      "loss": 0.1363,
+      "step": 518
+    },
+    {
+      "epoch": 0.11795454545454545,
+      "grad_norm": 3.3036590083913464,
+      "learning_rate": 1.2077382294441218e-06,
+      "loss": 0.1345,
+      "step": 519
+    },
+    {
+      "epoch": 0.11818181818181818,
+      "grad_norm": 2.9220942134431107,
+      "learning_rate": 1.2075767723475932e-06,
+      "loss": 0.0985,
+      "step": 520
+    },
+    {
+      "epoch": 0.11840909090909091,
+      "grad_norm": 4.032682593359572,
+      "learning_rate": 1.2074150182571579e-06,
+      "loss": 0.1798,
+      "step": 521
+    },
+    {
+      "epoch": 0.11863636363636364,
+      "grad_norm": 3.5016932539053665,
+      "learning_rate": 1.2072529672552771e-06,
+      "loss": 0.1482,
+      "step": 522
+    },
+    {
+      "epoch": 0.11886363636363637,
+      "grad_norm": 4.3105714404331135,
+      "learning_rate": 1.2070906194245634e-06,
+      "loss": 0.145,
+      "step": 523
+    },
+    {
+      "epoch": 0.1190909090909091,
+      "grad_norm": 2.556719172503092,
+      "learning_rate": 1.2069279748477812e-06,
+      "loss": 0.082,
+      "step": 524
+    },
+    {
+      "epoch": 0.11931818181818182,
+      "grad_norm": 2.013252433733995,
+      "learning_rate": 1.206765033607845e-06,
+      "loss": 0.1008,
+      "step": 525
+    },
+    {
+      "epoch": 0.11954545454545455,
+      "grad_norm": 2.629378975428092,
+      "learning_rate": 1.2066017957878212e-06,
+      "loss": 0.2058,
+      "step": 526
+    },
+    {
+      "epoch": 0.11977272727272727,
+      "grad_norm": 2.4516406956651875,
+      "learning_rate": 1.2064382614709276e-06,
+      "loss": 0.2118,
+      "step": 527
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.419768494059278,
+      "learning_rate": 1.206274430740533e-06,
+      "loss": 0.192,
+      "step": 528
+    },
+    {
+      "epoch": 0.12022727272727272,
+      "grad_norm": 4.32941899528697,
+      "learning_rate": 1.2061103036801573e-06,
+      "loss": 0.1827,
+      "step": 529
+    },
+    {
+      "epoch": 0.12045454545454545,
+      "grad_norm": 3.830374434250268,
+      "learning_rate": 1.2059458803734712e-06,
+      "loss": 0.1182,
+      "step": 530
+    },
+    {
+      "epoch": 0.12068181818181818,
+      "grad_norm": 2.2691564410476843,
+      "learning_rate": 1.2057811609042968e-06,
+      "loss": 0.1459,
+      "step": 531
+    },
+    {
+      "epoch": 0.12090909090909091,
+      "grad_norm": 3.8384096097139437,
+      "learning_rate": 1.205616145356607e-06,
+      "loss": 0.2172,
+      "step": 532
+    },
+    {
+      "epoch": 0.12113636363636364,
+      "grad_norm": 1.7526183548475944,
+      "learning_rate": 1.2054508338145257e-06,
+      "loss": 0.1213,
+      "step": 533
+    },
+    {
+      "epoch": 0.12136363636363637,
+      "grad_norm": 4.898460359781565,
+      "learning_rate": 1.2052852263623274e-06,
+      "loss": 0.1941,
+      "step": 534
+    },
+    {
+      "epoch": 0.1215909090909091,
+      "grad_norm": 2.697277694409338,
+      "learning_rate": 1.2051193230844382e-06,
+      "loss": 0.0909,
+      "step": 535
+    },
+    {
+      "epoch": 0.12181818181818181,
+      "grad_norm": 2.399772219862413,
+      "learning_rate": 1.2049531240654343e-06,
+      "loss": 0.1044,
+      "step": 536
+    },
+    {
+      "epoch": 0.12204545454545454,
+      "grad_norm": 2.833273520575603,
+      "learning_rate": 1.2047866293900428e-06,
+      "loss": 0.1911,
+      "step": 537
+    },
+    {
+      "epoch": 0.12227272727272727,
+      "grad_norm": 2.4279726356324476,
+      "learning_rate": 1.2046198391431415e-06,
+      "loss": 0.1954,
+      "step": 538
+    },
+    {
+      "epoch": 0.1225,
+      "grad_norm": 3.7944571617253064,
+      "learning_rate": 1.2044527534097595e-06,
+      "loss": 0.1605,
+      "step": 539
+    },
+    {
+      "epoch": 0.12272727272727273,
+      "grad_norm": 1.5480837031717636,
+      "learning_rate": 1.2042853722750756e-06,
+      "loss": 0.0532,
+      "step": 540
+    },
+    {
+      "epoch": 0.12295454545454546,
+      "grad_norm": 3.193230059256664,
+      "learning_rate": 1.2041176958244197e-06,
+      "loss": 0.0845,
+      "step": 541
+    },
+    {
+      "epoch": 0.12318181818181818,
+      "grad_norm": 3.4892462151970696,
+      "learning_rate": 1.2039497241432724e-06,
+      "loss": 0.0904,
+      "step": 542
+    },
+    {
+      "epoch": 0.12340909090909091,
+      "grad_norm": 1.9330022602538879,
+      "learning_rate": 1.2037814573172642e-06,
+      "loss": 0.1256,
+      "step": 543
+    },
+    {
+      "epoch": 0.12363636363636364,
+      "grad_norm": 2.6872356325795965,
+      "learning_rate": 1.2036128954321768e-06,
+      "loss": 0.1338,
+      "step": 544
+    },
+    {
+      "epoch": 0.12386363636363637,
+      "grad_norm": 5.244441136881968,
+      "learning_rate": 1.2034440385739418e-06,
+      "loss": 0.1483,
+      "step": 545
+    },
+    {
+      "epoch": 0.12409090909090909,
+      "grad_norm": 3.948064299821336,
+      "learning_rate": 1.2032748868286415e-06,
+      "loss": 0.2428,
+      "step": 546
+    },
+    {
+      "epoch": 0.12431818181818181,
+      "grad_norm": 4.727474349290802,
+      "learning_rate": 1.2031054402825082e-06,
+      "loss": 0.1278,
+      "step": 547
+    },
+    {
+      "epoch": 0.12454545454545454,
+      "grad_norm": 2.150453758225874,
+      "learning_rate": 1.2029356990219248e-06,
+      "loss": 0.1251,
+      "step": 548
+    },
+    {
+      "epoch": 0.12477272727272727,
+      "grad_norm": 2.8964909888861565,
+      "learning_rate": 1.2027656631334242e-06,
+      "loss": 0.2567,
+      "step": 549
+    },
+    {
+      "epoch": 0.125,
+      "grad_norm": 4.308257763115935,
+      "learning_rate": 1.2025953327036897e-06,
+      "loss": 0.1347,
+      "step": 550
+    },
+    {
+      "epoch": 0.12522727272727271,
+      "grad_norm": 3.6043653035760483,
+      "learning_rate": 1.2024247078195542e-06,
+      "loss": 0.1502,
+      "step": 551
+    },
+    {
+      "epoch": 0.12545454545454546,
+      "grad_norm": 2.6503674983649477,
+      "learning_rate": 1.202253788568002e-06,
+      "loss": 0.118,
+      "step": 552
+    },
+    {
+      "epoch": 0.12568181818181817,
+      "grad_norm": 2.2951346960486303,
+      "learning_rate": 1.202082575036166e-06,
+      "loss": 0.2004,
+      "step": 553
+    },
+    {
+      "epoch": 0.12590909090909091,
+      "grad_norm": 3.392632965153945,
+      "learning_rate": 1.2019110673113302e-06,
+      "loss": 0.1987,
+      "step": 554
+    },
+    {
+      "epoch": 0.12613636363636363,
+      "grad_norm": 2.208017855549523,
+      "learning_rate": 1.2017392654809278e-06,
+      "loss": 0.0793,
+      "step": 555
+    },
+    {
+      "epoch": 0.12636363636363637,
+      "grad_norm": 2.117808437476879,
+      "learning_rate": 1.2015671696325423e-06,
+      "loss": 0.1584,
+      "step": 556
+    },
+    {
+      "epoch": 0.1265909090909091,
+      "grad_norm": 4.717970528128702,
+      "learning_rate": 1.2013947798539073e-06,
+      "loss": 0.0956,
+      "step": 557
+    },
+    {
+      "epoch": 0.12681818181818183,
+      "grad_norm": 2.098387420997237,
+      "learning_rate": 1.2012220962329058e-06,
+      "loss": 0.063,
+      "step": 558
+    },
+    {
+      "epoch": 0.12704545454545454,
+      "grad_norm": 2.9813153253764257,
+      "learning_rate": 1.201049118857571e-06,
+      "loss": 0.1427,
+      "step": 559
+    },
+    {
+      "epoch": 0.12727272727272726,
+      "grad_norm": 2.7694431312704273,
+      "learning_rate": 1.2008758478160853e-06,
+      "loss": 0.1427,
+      "step": 560
+    },
+    {
+      "epoch": 0.1275,
+      "grad_norm": 1.9100471451430532,
+      "learning_rate": 1.2007022831967813e-06,
+      "loss": 0.1233,
+      "step": 561
+    },
+    {
+      "epoch": 0.12772727272727272,
+      "grad_norm": 2.5838575307479545,
+      "learning_rate": 1.2005284250881417e-06,
+      "loss": 0.1157,
+      "step": 562
+    },
+    {
+      "epoch": 0.12795454545454546,
+      "grad_norm": 3.4928689598047806,
+      "learning_rate": 1.2003542735787973e-06,
+      "loss": 0.1972,
+      "step": 563
+    },
+    {
+      "epoch": 0.12818181818181817,
+      "grad_norm": 1.7036420975608682,
+      "learning_rate": 1.20017982875753e-06,
+      "loss": 0.0845,
+      "step": 564
+    },
+    {
+      "epoch": 0.12840909090909092,
+      "grad_norm": 2.908004499311132,
+      "learning_rate": 1.2000050907132705e-06,
+      "loss": 0.1354,
+      "step": 565
+    },
+    {
+      "epoch": 0.12863636363636363,
+      "grad_norm": 2.345000670043479,
+      "learning_rate": 1.1998300595350993e-06,
+      "loss": 0.1649,
+      "step": 566
+    },
+    {
+      "epoch": 0.12886363636363637,
+      "grad_norm": 3.6640666262617376,
+      "learning_rate": 1.1996547353122461e-06,
+      "loss": 0.1338,
+      "step": 567
+    },
+    {
+      "epoch": 0.1290909090909091,
+      "grad_norm": 4.661088281313586,
+      "learning_rate": 1.1994791181340897e-06,
+      "loss": 0.1933,
+      "step": 568
+    },
+    {
+      "epoch": 0.1293181818181818,
+      "grad_norm": 3.2870774617872494,
+      "learning_rate": 1.1993032080901593e-06,
+      "loss": 0.1681,
+      "step": 569
+    },
+    {
+      "epoch": 0.12954545454545455,
+      "grad_norm": 1.931721329526802,
+      "learning_rate": 1.1991270052701323e-06,
+      "loss": 0.0983,
+      "step": 570
+    },
+    {
+      "epoch": 0.12977272727272726,
+      "grad_norm": 2.7930699910832484,
+      "learning_rate": 1.1989505097638357e-06,
+      "loss": 0.2281,
+      "step": 571
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7123985458682283,
+      "learning_rate": 1.198773721661246e-06,
+      "loss": 0.2385,
+      "step": 572
+    },
+    {
+      "epoch": 0.13022727272727272,
+      "grad_norm": 3.7076940911618155,
+      "learning_rate": 1.1985966410524883e-06,
+      "loss": 0.1452,
+      "step": 573
+    },
+    {
+      "epoch": 0.13045454545454546,
+      "grad_norm": 3.903896826424082,
+      "learning_rate": 1.1984192680278376e-06,
+      "loss": 0.1112,
+      "step": 574
+    },
+    {
+      "epoch": 0.13068181818181818,
+      "grad_norm": 3.6089194458098697,
+      "learning_rate": 1.1982416026777172e-06,
+      "loss": 0.178,
+      "step": 575
+    },
+    {
+      "epoch": 0.13090909090909092,
+      "grad_norm": 2.7504623360782356,
+      "learning_rate": 1.1980636450926999e-06,
+      "loss": 0.0954,
+      "step": 576
+    },
+    {
+      "epoch": 0.13113636363636363,
+      "grad_norm": 1.4654231028042315,
+      "learning_rate": 1.1978853953635074e-06,
+      "loss": 0.0699,
+      "step": 577
+    },
+    {
+      "epoch": 0.13136363636363638,
+      "grad_norm": 2.894690042007633,
+      "learning_rate": 1.1977068535810101e-06,
+      "loss": 0.1353,
+      "step": 578
+    },
+    {
+      "epoch": 0.1315909090909091,
+      "grad_norm": 3.3088626087814483,
+      "learning_rate": 1.1975280198362276e-06,
+      "loss": 0.1746,
+      "step": 579
+    },
+    {
+      "epoch": 0.1318181818181818,
+      "grad_norm": 4.162288648384007,
+      "learning_rate": 1.1973488942203282e-06,
+      "loss": 0.161,
+      "step": 580
+    },
+    {
+      "epoch": 0.13204545454545455,
+      "grad_norm": 2.2242375148844333,
+      "learning_rate": 1.197169476824629e-06,
+      "loss": 0.0874,
+      "step": 581
+    },
+    {
+      "epoch": 0.13227272727272726,
+      "grad_norm": 2.1105600345869795,
+      "learning_rate": 1.1969897677405956e-06,
+      "loss": 0.1012,
+      "step": 582
+    },
+    {
+      "epoch": 0.1325,
+      "grad_norm": 2.304470221079024,
+      "learning_rate": 1.1968097670598428e-06,
+      "loss": 0.2286,
+      "step": 583
+    },
+    {
+      "epoch": 0.13272727272727272,
+      "grad_norm": 2.3241016386932687,
+      "learning_rate": 1.1966294748741336e-06,
+      "loss": 0.0732,
+      "step": 584
+    },
+    {
+      "epoch": 0.13295454545454546,
+      "grad_norm": 2.4283164633258894,
+      "learning_rate": 1.19644889127538e-06,
+      "loss": 0.0972,
+      "step": 585
+    },
+    {
+      "epoch": 0.13318181818181818,
+      "grad_norm": 3.1848753764268443,
+      "learning_rate": 1.1962680163556424e-06,
+      "loss": 0.0974,
+      "step": 586
+    },
+    {
+      "epoch": 0.13340909090909092,
+      "grad_norm": 1.897853082815604,
+      "learning_rate": 1.1960868502071294e-06,
+      "loss": 0.1577,
+      "step": 587
+    },
+    {
+      "epoch": 0.13363636363636364,
+      "grad_norm": 1.662424269952875,
+      "learning_rate": 1.1959053929221984e-06,
+      "loss": 0.0891,
+      "step": 588
+    },
+    {
+      "epoch": 0.13386363636363635,
+      "grad_norm": 3.603206821148784,
+      "learning_rate": 1.1957236445933553e-06,
+      "loss": 0.1522,
+      "step": 589
+    },
+    {
+      "epoch": 0.1340909090909091,
+      "grad_norm": 3.0261554129810158,
+      "learning_rate": 1.1955416053132542e-06,
+      "loss": 0.0821,
+      "step": 590
+    },
+    {
+      "epoch": 0.1343181818181818,
+      "grad_norm": 3.2318951709217796,
+      "learning_rate": 1.1953592751746976e-06,
+      "loss": 0.1449,
+      "step": 591
+    },
+    {
+      "epoch": 0.13454545454545455,
+      "grad_norm": 2.6462190397470127,
+      "learning_rate": 1.1951766542706362e-06,
+      "loss": 0.1048,
+      "step": 592
+    },
+    {
+      "epoch": 0.13477272727272727,
+      "grad_norm": 2.3706296442605375,
+      "learning_rate": 1.1949937426941689e-06,
+      "loss": 0.1363,
+      "step": 593
+    },
+    {
+      "epoch": 0.135,
+      "grad_norm": 2.8040664839299887,
+      "learning_rate": 1.1948105405385428e-06,
+      "loss": 0.1984,
+      "step": 594
+    },
+    {
+      "epoch": 0.13522727272727272,
+      "grad_norm": 3.962385537043296,
+      "learning_rate": 1.1946270478971533e-06,
+      "loss": 0.1681,
+      "step": 595
+    },
+    {
+      "epoch": 0.13545454545454547,
+      "grad_norm": 4.350715264773958,
+      "learning_rate": 1.194443264863544e-06,
+      "loss": 0.2082,
+      "step": 596
+    },
+    {
+      "epoch": 0.13568181818181818,
+      "grad_norm": 3.0064974941931677,
+      "learning_rate": 1.1942591915314058e-06,
+      "loss": 0.2334,
+      "step": 597
+    },
+    {
+      "epoch": 0.1359090909090909,
+      "grad_norm": 3.4478704951100374,
+      "learning_rate": 1.1940748279945784e-06,
+      "loss": 0.1266,
+      "step": 598
+    },
+    {
+      "epoch": 0.13613636363636364,
+      "grad_norm": 1.5802310892883296,
+      "learning_rate": 1.1938901743470494e-06,
+      "loss": 0.0976,
+      "step": 599
+    },
+    {
+      "epoch": 0.13636363636363635,
+      "grad_norm": 5.469476885440198,
+      "learning_rate": 1.1937052306829534e-06,
+      "loss": 0.1473,
+      "step": 600
+    },
+    {
+      "epoch": 0.1365909090909091,
+      "grad_norm": 2.0599309325992086,
+      "learning_rate": 1.1935199970965741e-06,
+      "loss": 0.0812,
+      "step": 601
+    },
+    {
+      "epoch": 0.1368181818181818,
+      "grad_norm": 3.7876336857485993,
+      "learning_rate": 1.193334473682342e-06,
+      "loss": 0.1856,
+      "step": 602
+    },
+    {
+      "epoch": 0.13704545454545455,
+      "grad_norm": 2.8538898475669727,
+      "learning_rate": 1.193148660534836e-06,
+      "loss": 0.23,
+      "step": 603
+    },
+    {
+      "epoch": 0.13727272727272727,
+      "grad_norm": 2.243760180086315,
+      "learning_rate": 1.1929625577487825e-06,
+      "loss": 0.1202,
+      "step": 604
+    },
+    {
+      "epoch": 0.1375,
+      "grad_norm": 5.164509281966962,
+      "learning_rate": 1.1927761654190552e-06,
+      "loss": 0.1974,
+      "step": 605
+    },
+    {
+      "epoch": 0.13772727272727273,
+      "grad_norm": 4.26123569595119,
+      "learning_rate": 1.192589483640676e-06,
+      "loss": 0.1534,
+      "step": 606
+    },
+    {
+      "epoch": 0.13795454545454544,
+      "grad_norm": 3.654511683439206,
+      "learning_rate": 1.1924025125088138e-06,
+      "loss": 0.1269,
+      "step": 607
+    },
+    {
+      "epoch": 0.13818181818181818,
+      "grad_norm": 2.88607378027308,
+      "learning_rate": 1.1922152521187854e-06,
+      "loss": 0.185,
+      "step": 608
+    },
+    {
+      "epoch": 0.1384090909090909,
+      "grad_norm": 2.142519815976745,
+      "learning_rate": 1.1920277025660553e-06,
+      "loss": 0.1071,
+      "step": 609
+    },
+    {
+      "epoch": 0.13863636363636364,
+      "grad_norm": 4.234853887801776,
+      "learning_rate": 1.1918398639462345e-06,
+      "loss": 0.1407,
+      "step": 610
+    },
+    {
+      "epoch": 0.13886363636363636,
+      "grad_norm": 4.847175124644156,
+      "learning_rate": 1.1916517363550821e-06,
+      "loss": 0.2459,
+      "step": 611
+    },
+    {
+      "epoch": 0.1390909090909091,
+      "grad_norm": 2.6810702667911186,
+      "learning_rate": 1.1914633198885047e-06,
+      "loss": 0.2013,
+      "step": 612
+    },
+    {
+      "epoch": 0.1393181818181818,
+      "grad_norm": 0.9282082803774457,
+      "learning_rate": 1.1912746146425555e-06,
+      "loss": 0.0404,
+      "step": 613
+    },
+    {
+      "epoch": 0.13954545454545456,
+      "grad_norm": 2.7637210079594876,
+      "learning_rate": 1.1910856207134352e-06,
+      "loss": 0.1719,
+      "step": 614
+    },
+    {
+      "epoch": 0.13977272727272727,
+      "grad_norm": 3.4953702693791397,
+      "learning_rate": 1.1908963381974916e-06,
+      "loss": 0.0953,
+      "step": 615
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2979343046971703,
+      "learning_rate": 1.19070676719122e-06,
+      "loss": 0.2989,
+      "step": 616
+    },
+    {
+      "epoch": 0.14022727272727273,
+      "grad_norm": 3.567295341985518,
+      "learning_rate": 1.1905169077912623e-06,
+      "loss": 0.1642,
+      "step": 617
+    },
+    {
+      "epoch": 0.14045454545454544,
+      "grad_norm": 3.003479520144108,
+      "learning_rate": 1.1903267600944077e-06,
+      "loss": 0.0994,
+      "step": 618
+    },
+    {
+      "epoch": 0.14068181818181819,
+      "grad_norm": 3.511446640079229,
+      "learning_rate": 1.1901363241975921e-06,
+      "loss": 0.1204,
+      "step": 619
+    },
+    {
+      "epoch": 0.1409090909090909,
+      "grad_norm": 1.9171796828882752,
+      "learning_rate": 1.1899456001978987e-06,
+      "loss": 0.1284,
+      "step": 620
+    },
+    {
+      "epoch": 0.14113636363636364,
+      "grad_norm": 2.4108724575341403,
+      "learning_rate": 1.1897545881925573e-06,
+      "loss": 0.127,
+      "step": 621
+    },
+    {
+      "epoch": 0.14136363636363636,
+      "grad_norm": 2.0620324329434685,
+      "learning_rate": 1.1895632882789447e-06,
+      "loss": 0.1546,
+      "step": 622
+    },
+    {
+      "epoch": 0.1415909090909091,
+      "grad_norm": 2.3192045591918187,
+      "learning_rate": 1.1893717005545843e-06,
+      "loss": 0.1314,
+      "step": 623
+    },
+    {
+      "epoch": 0.14181818181818182,
+      "grad_norm": 3.389403913289586,
+      "learning_rate": 1.189179825117146e-06,
+      "loss": 0.1796,
+      "step": 624
+    },
+    {
+      "epoch": 0.14204545454545456,
+      "grad_norm": 3.47646641688817,
+      "learning_rate": 1.1889876620644472e-06,
+      "loss": 0.1212,
+      "step": 625
+    },
+    {
+      "epoch": 0.14227272727272727,
+      "grad_norm": 4.970814693649349,
+      "learning_rate": 1.1887952114944509e-06,
+      "loss": 0.1268,
+      "step": 626
+    },
+    {
+      "epoch": 0.1425,
+      "grad_norm": 1.379310731725192,
+      "learning_rate": 1.1886024735052676e-06,
+      "loss": 0.1013,
+      "step": 627
+    },
+    {
+      "epoch": 0.14272727272727273,
+      "grad_norm": 2.89035106180528,
+      "learning_rate": 1.1884094481951535e-06,
+      "loss": 0.1863,
+      "step": 628
+    },
+    {
+      "epoch": 0.14295454545454545,
+      "grad_norm": 4.142207318726914,
+      "learning_rate": 1.1882161356625122e-06,
+      "loss": 0.1725,
+      "step": 629
+    },
+    {
+      "epoch": 0.1431818181818182,
+      "grad_norm": 2.3708635028677083,
+      "learning_rate": 1.1880225360058925e-06,
+      "loss": 0.1473,
+      "step": 630
+    },
+    {
+      "epoch": 0.1434090909090909,
+      "grad_norm": 3.164006357166789,
+      "learning_rate": 1.1878286493239907e-06,
+      "loss": 0.1875,
+      "step": 631
+    },
+    {
+      "epoch": 0.14363636363636365,
+      "grad_norm": 2.012891475566374,
+      "learning_rate": 1.187634475715649e-06,
+      "loss": 0.0871,
+      "step": 632
+    },
+    {
+      "epoch": 0.14386363636363636,
+      "grad_norm": 1.4996893437139978,
+      "learning_rate": 1.1874400152798557e-06,
+      "loss": 0.0818,
+      "step": 633
+    },
+    {
+      "epoch": 0.1440909090909091,
+      "grad_norm": 2.9724805179056135,
+      "learning_rate": 1.1872452681157453e-06,
+      "loss": 0.0876,
+      "step": 634
+    },
+    {
+      "epoch": 0.14431818181818182,
+      "grad_norm": 2.478648639009007,
+      "learning_rate": 1.1870502343225992e-06,
+      "loss": 0.0843,
+      "step": 635
+    },
+    {
+      "epoch": 0.14454545454545453,
+      "grad_norm": 2.3989686437463544,
+      "learning_rate": 1.186854913999844e-06,
+      "loss": 0.1612,
+      "step": 636
+    },
+    {
+      "epoch": 0.14477272727272728,
+      "grad_norm": 2.846010490040111,
+      "learning_rate": 1.1866593072470527e-06,
+      "loss": 0.1189,
+      "step": 637
+    },
+    {
+      "epoch": 0.145,
+      "grad_norm": 2.840094600150523,
+      "learning_rate": 1.1864634141639448e-06,
+      "loss": 0.1087,
+      "step": 638
+    },
+    {
+      "epoch": 0.14522727272727273,
+      "grad_norm": 4.366873590117856,
+      "learning_rate": 1.1862672348503848e-06,
+      "loss": 0.1777,
+      "step": 639
+    },
+    {
+      "epoch": 0.14545454545454545,
+      "grad_norm": 2.9445837118851625,
+      "learning_rate": 1.186070769406384e-06,
+      "loss": 0.1197,
+      "step": 640
+    },
+    {
+      "epoch": 0.1456818181818182,
+      "grad_norm": 4.3467521538212495,
+      "learning_rate": 1.185874017932099e-06,
+      "loss": 0.14,
+      "step": 641
+    },
+    {
+      "epoch": 0.1459090909090909,
+      "grad_norm": 2.7107082632244563,
+      "learning_rate": 1.1856769805278327e-06,
+      "loss": 0.1088,
+      "step": 642
+    },
+    {
+      "epoch": 0.14613636363636365,
+      "grad_norm": 3.0572416947324084,
+      "learning_rate": 1.1854796572940332e-06,
+      "loss": 0.2166,
+      "step": 643
+    },
+    {
+      "epoch": 0.14636363636363636,
+      "grad_norm": 1.7294876653987825,
+      "learning_rate": 1.1852820483312951e-06,
+      "loss": 0.1229,
+      "step": 644
+    },
+    {
+      "epoch": 0.14659090909090908,
+      "grad_norm": 2.2077723153086577,
+      "learning_rate": 1.1850841537403577e-06,
+      "loss": 0.0724,
+      "step": 645
+    },
+    {
+      "epoch": 0.14681818181818182,
+      "grad_norm": 3.2267668323284795,
+      "learning_rate": 1.1848859736221062e-06,
+      "loss": 0.1787,
+      "step": 646
+    },
+    {
+      "epoch": 0.14704545454545453,
+      "grad_norm": 6.311236342788911,
+      "learning_rate": 1.1846875080775724e-06,
+      "loss": 0.1886,
+      "step": 647
+    },
+    {
+      "epoch": 0.14727272727272728,
+      "grad_norm": 2.0051481855087507,
+      "learning_rate": 1.1844887572079322e-06,
+      "loss": 0.0948,
+      "step": 648
+    },
+    {
+      "epoch": 0.1475,
+      "grad_norm": 3.039269472251779,
+      "learning_rate": 1.1842897211145075e-06,
+      "loss": 0.1275,
+      "step": 649
+    },
+    {
+      "epoch": 0.14772727272727273,
+      "grad_norm": 4.005537587249589,
+      "learning_rate": 1.1840903998987657e-06,
+      "loss": 0.1124,
+      "step": 650
+    },
+    {
+      "epoch": 0.14795454545454545,
+      "grad_norm": 1.281619369406636,
+      "learning_rate": 1.1838907936623196e-06,
+      "loss": 0.0666,
+      "step": 651
+    },
+    {
+      "epoch": 0.1481818181818182,
+      "grad_norm": 2.2042880744049156,
+      "learning_rate": 1.183690902506927e-06,
+      "loss": 0.1043,
+      "step": 652
+    },
+    {
+      "epoch": 0.1484090909090909,
+      "grad_norm": 2.31369437669866,
+      "learning_rate": 1.1834907265344913e-06,
+      "loss": 0.1552,
+      "step": 653
+    },
+    {
+      "epoch": 0.14863636363636365,
+      "grad_norm": 2.4636723773691416,
+      "learning_rate": 1.1832902658470608e-06,
+      "loss": 0.162,
+      "step": 654
+    },
+    {
+      "epoch": 0.14886363636363636,
+      "grad_norm": 2.375174731872374,
+      "learning_rate": 1.1830895205468293e-06,
+      "loss": 0.092,
+      "step": 655
+    },
+    {
+      "epoch": 0.14909090909090908,
+      "grad_norm": 2.4162840663153586,
+      "learning_rate": 1.182888490736135e-06,
+      "loss": 0.1092,
+      "step": 656
+    },
+    {
+      "epoch": 0.14931818181818182,
+      "grad_norm": 2.6126138505316114,
+      "learning_rate": 1.1826871765174622e-06,
+      "loss": 0.1421,
+      "step": 657
+    },
+    {
+      "epoch": 0.14954545454545454,
+      "grad_norm": 3.4322474711795823,
+      "learning_rate": 1.1824855779934392e-06,
+      "loss": 0.1097,
+      "step": 658
+    },
+    {
+      "epoch": 0.14977272727272728,
+      "grad_norm": 2.980748123530426,
+      "learning_rate": 1.1822836952668397e-06,
+      "loss": 0.1706,
+      "step": 659
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3075328167430285,
+      "learning_rate": 1.182081528440582e-06,
+      "loss": 0.1672,
+      "step": 660
+    },
+    {
+      "epoch": 0.15022727272727274,
+      "grad_norm": 2.2767094309710307,
+      "learning_rate": 1.18187907761773e-06,
+      "loss": 0.0987,
+      "step": 661
+    },
+    {
+      "epoch": 0.15045454545454545,
+      "grad_norm": 2.6697459385294264,
+      "learning_rate": 1.1816763429014917e-06,
+      "loss": 0.1393,
+      "step": 662
+    },
+    {
+      "epoch": 0.1506818181818182,
+      "grad_norm": 2.7861667529239034,
+      "learning_rate": 1.1814733243952193e-06,
+      "loss": 0.2047,
+      "step": 663
+    },
+    {
+      "epoch": 0.1509090909090909,
+      "grad_norm": 4.791942950897901,
+      "learning_rate": 1.1812700222024111e-06,
+      "loss": 0.3118,
+      "step": 664
+    },
+    {
+      "epoch": 0.15113636363636362,
+      "grad_norm": 3.5181082061061772,
+      "learning_rate": 1.1810664364267092e-06,
+      "loss": 0.1188,
+      "step": 665
+    },
+    {
+      "epoch": 0.15136363636363637,
+      "grad_norm": 2.5117207418715886,
+      "learning_rate": 1.1808625671718999e-06,
+      "loss": 0.0803,
+      "step": 666
+    },
+    {
+      "epoch": 0.15159090909090908,
+      "grad_norm": 3.0096279077338823,
+      "learning_rate": 1.1806584145419144e-06,
+      "loss": 0.2191,
+      "step": 667
+    },
+    {
+      "epoch": 0.15181818181818182,
+      "grad_norm": 3.7087275544111145,
+      "learning_rate": 1.1804539786408292e-06,
+      "loss": 0.1637,
+      "step": 668
+    },
+    {
+      "epoch": 0.15204545454545454,
+      "grad_norm": 3.3191378112368946,
+      "learning_rate": 1.1802492595728634e-06,
+      "loss": 0.1297,
+      "step": 669
+    },
+    {
+      "epoch": 0.15227272727272728,
+      "grad_norm": 2.1698648807784515,
+      "learning_rate": 1.1800442574423823e-06,
+      "loss": 0.091,
+      "step": 670
+    },
+    {
+      "epoch": 0.1525,
+      "grad_norm": 3.431341946362163,
+      "learning_rate": 1.1798389723538942e-06,
+      "loss": 0.0866,
+      "step": 671
+    },
+    {
+      "epoch": 0.15272727272727274,
+      "grad_norm": 4.153382296642683,
+      "learning_rate": 1.1796334044120522e-06,
+      "loss": 0.2152,
+      "step": 672
+    },
+    {
+      "epoch": 0.15295454545454545,
+      "grad_norm": 3.5445712889443572,
+      "learning_rate": 1.1794275537216534e-06,
+      "loss": 0.2061,
+      "step": 673
+    },
+    {
+      "epoch": 0.15318181818181817,
+      "grad_norm": 2.0500779899504833,
+      "learning_rate": 1.1792214203876396e-06,
+      "loss": 0.098,
+      "step": 674
+    },
+    {
+      "epoch": 0.1534090909090909,
+      "grad_norm": 1.8909413828445034,
+      "learning_rate": 1.1790150045150958e-06,
+      "loss": 0.1143,
+      "step": 675
+    },
+    {
+      "epoch": 0.15363636363636363,
+      "grad_norm": 2.9091838294689625,
+      "learning_rate": 1.1788083062092518e-06,
+      "loss": 0.0867,
+      "step": 676
+    },
+    {
+      "epoch": 0.15386363636363637,
+      "grad_norm": 2.245778846358278,
+      "learning_rate": 1.1786013255754808e-06,
+      "loss": 0.1523,
+      "step": 677
+    },
+    {
+      "epoch": 0.15409090909090908,
+      "grad_norm": 4.018810593072499,
+      "learning_rate": 1.1783940627193002e-06,
+      "loss": 0.1594,
+      "step": 678
+    },
+    {
+      "epoch": 0.15431818181818183,
+      "grad_norm": 2.4918503999416375,
+      "learning_rate": 1.1781865177463717e-06,
+      "loss": 0.1434,
+      "step": 679
+    },
+    {
+      "epoch": 0.15454545454545454,
+      "grad_norm": 1.7463432237419554,
+      "learning_rate": 1.1779786907625e-06,
+      "loss": 0.1247,
+      "step": 680
+    },
+    {
+      "epoch": 0.15477272727272728,
+      "grad_norm": 2.0176686430118917,
+      "learning_rate": 1.177770581873634e-06,
+      "loss": 0.1473,
+      "step": 681
+    },
+    {
+      "epoch": 0.155,
+      "grad_norm": 2.747145536579915,
+      "learning_rate": 1.1775621911858665e-06,
+      "loss": 0.2038,
+      "step": 682
+    },
+    {
+      "epoch": 0.1552272727272727,
+      "grad_norm": 3.1510741981191193,
+      "learning_rate": 1.1773535188054336e-06,
+      "loss": 0.1128,
+      "step": 683
+    },
+    {
+      "epoch": 0.15545454545454546,
+      "grad_norm": 4.029277786302326,
+      "learning_rate": 1.177144564838715e-06,
+      "loss": 0.1714,
+      "step": 684
+    },
+    {
+      "epoch": 0.15568181818181817,
+      "grad_norm": 1.8601784285123628,
+      "learning_rate": 1.1769353293922341e-06,
+      "loss": 0.1441,
+      "step": 685
+    },
+    {
+      "epoch": 0.1559090909090909,
+      "grad_norm": 2.9784160155105037,
+      "learning_rate": 1.1767258125726584e-06,
+      "loss": 0.1009,
+      "step": 686
+    },
+    {
+      "epoch": 0.15613636363636363,
+      "grad_norm": 3.5876129635968455,
+      "learning_rate": 1.1765160144867972e-06,
+      "loss": 0.1317,
+      "step": 687
+    },
+    {
+      "epoch": 0.15636363636363637,
+      "grad_norm": 3.2441988651574536,
+      "learning_rate": 1.176305935241605e-06,
+      "loss": 0.1385,
+      "step": 688
+    },
+    {
+      "epoch": 0.1565909090909091,
+      "grad_norm": 4.203294202137399,
+      "learning_rate": 1.1760955749441786e-06,
+      "loss": 0.1737,
+      "step": 689
+    },
+    {
+      "epoch": 0.15681818181818183,
+      "grad_norm": 1.919232559416769,
+      "learning_rate": 1.1758849337017587e-06,
+      "loss": 0.0686,
+      "step": 690
+    },
+    {
+      "epoch": 0.15704545454545454,
+      "grad_norm": 2.3026456663388375,
+      "learning_rate": 1.175674011621728e-06,
+      "loss": 0.1316,
+      "step": 691
+    },
+    {
+      "epoch": 0.1572727272727273,
+      "grad_norm": 3.9443586520938387,
+      "learning_rate": 1.1754628088116138e-06,
+      "loss": 0.1154,
+      "step": 692
+    },
+    {
+      "epoch": 0.1575,
+      "grad_norm": 2.7876319636069824,
+      "learning_rate": 1.1752513253790861e-06,
+      "loss": 0.1275,
+      "step": 693
+    },
+    {
+      "epoch": 0.15772727272727272,
+      "grad_norm": 4.42495785538581,
+      "learning_rate": 1.1750395614319576e-06,
+      "loss": 0.1375,
+      "step": 694
+    },
+    {
+      "epoch": 0.15795454545454546,
+      "grad_norm": 4.412921738694996,
+      "learning_rate": 1.174827517078184e-06,
+      "loss": 0.185,
+      "step": 695
+    },
+    {
+      "epoch": 0.15818181818181817,
+      "grad_norm": 1.3120321617270885,
+      "learning_rate": 1.1746151924258644e-06,
+      "loss": 0.0703,
+      "step": 696
+    },
+    {
+      "epoch": 0.15840909090909092,
+      "grad_norm": 4.635825956597518,
+      "learning_rate": 1.1744025875832405e-06,
+      "loss": 0.2247,
+      "step": 697
+    },
+    {
+      "epoch": 0.15863636363636363,
+      "grad_norm": 2.7241427213406566,
+      "learning_rate": 1.174189702658697e-06,
+      "loss": 0.0974,
+      "step": 698
+    },
+    {
+      "epoch": 0.15886363636363637,
+      "grad_norm": 4.3607626592669515,
+      "learning_rate": 1.173976537760761e-06,
+      "loss": 0.2763,
+      "step": 699
+    },
+    {
+      "epoch": 0.1590909090909091,
+      "grad_norm": 3.4955896156316646,
+      "learning_rate": 1.1737630929981026e-06,
+      "loss": 0.1791,
+      "step": 700
+    },
+    {
+      "epoch": 0.15931818181818183,
+      "grad_norm": 4.142033011564397,
+      "learning_rate": 1.1735493684795348e-06,
+      "loss": 0.1625,
+      "step": 701
+    },
+    {
+      "epoch": 0.15954545454545455,
+      "grad_norm": 2.4010837798923816,
+      "learning_rate": 1.173335364314013e-06,
+      "loss": 0.1307,
+      "step": 702
+    },
+    {
+      "epoch": 0.15977272727272726,
+      "grad_norm": 2.205409683963526,
+      "learning_rate": 1.173121080610635e-06,
+      "loss": 0.1052,
+      "step": 703
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.077776081514888,
+      "learning_rate": 1.1729065174786414e-06,
+      "loss": 0.079,
+      "step": 704
+    },
+    {
+      "epoch": 0.16022727272727272,
+      "grad_norm": 1.4319932933937107,
+      "learning_rate": 1.1726916750274148e-06,
+      "loss": 0.1054,
+      "step": 705
+    },
+    {
+      "epoch": 0.16045454545454546,
+      "grad_norm": 1.2346306307681347,
+      "learning_rate": 1.1724765533664808e-06,
+      "loss": 0.0623,
+      "step": 706
+    },
+    {
+      "epoch": 0.16068181818181818,
+      "grad_norm": 5.388785271504581,
+      "learning_rate": 1.1722611526055073e-06,
+      "loss": 0.1167,
+      "step": 707
+    },
+    {
+      "epoch": 0.16090909090909092,
+      "grad_norm": 2.314170205547252,
+      "learning_rate": 1.1720454728543034e-06,
+      "loss": 0.0661,
+      "step": 708
+    },
+    {
+      "epoch": 0.16113636363636363,
+      "grad_norm": 2.418063295938299,
+      "learning_rate": 1.171829514222822e-06,
+      "loss": 0.1307,
+      "step": 709
+    },
+    {
+      "epoch": 0.16136363636363638,
+      "grad_norm": 5.959087069962887,
+      "learning_rate": 1.1716132768211572e-06,
+      "loss": 0.2376,
+      "step": 710
+    },
+    {
+      "epoch": 0.1615909090909091,
+      "grad_norm": 2.8307364485663173,
+      "learning_rate": 1.1713967607595455e-06,
+      "loss": 0.0959,
+      "step": 711
+    },
+    {
+      "epoch": 0.1618181818181818,
+      "grad_norm": 6.681362226085489,
+      "learning_rate": 1.1711799661483653e-06,
+      "loss": 0.1507,
+      "step": 712
+    },
+    {
+      "epoch": 0.16204545454545455,
+      "grad_norm": 2.4054067679003017,
+      "learning_rate": 1.170962893098137e-06,
+      "loss": 0.111,
+      "step": 713
+    },
+    {
+      "epoch": 0.16227272727272726,
+      "grad_norm": 2.0632433425639585,
+      "learning_rate": 1.1707455417195231e-06,
+      "loss": 0.1316,
+      "step": 714
+    },
+    {
+      "epoch": 0.1625,
+      "grad_norm": 3.1620604263130843,
+      "learning_rate": 1.170527912123328e-06,
+      "loss": 0.1007,
+      "step": 715
+    },
+    {
+      "epoch": 0.16272727272727272,
+      "grad_norm": 6.379889387763465,
+      "learning_rate": 1.1703100044204984e-06,
+      "loss": 0.1653,
+      "step": 716
+    },
+    {
+      "epoch": 0.16295454545454546,
+      "grad_norm": 3.309482698057941,
+      "learning_rate": 1.1700918187221214e-06,
+      "loss": 0.226,
+      "step": 717
+    },
+    {
+      "epoch": 0.16318181818181818,
+      "grad_norm": 2.748635838382098,
+      "learning_rate": 1.169873355139427e-06,
+      "loss": 0.1165,
+      "step": 718
+    },
+    {
+      "epoch": 0.16340909090909092,
+      "grad_norm": 2.022588504874839,
+      "learning_rate": 1.1696546137837865e-06,
+      "loss": 0.141,
+      "step": 719
+    },
+    {
+      "epoch": 0.16363636363636364,
+      "grad_norm": 3.191057675975919,
+      "learning_rate": 1.169435594766713e-06,
+      "loss": 0.1208,
+      "step": 720
+    },
+    {
+      "epoch": 0.16386363636363635,
+      "grad_norm": 1.2802870857450932,
+      "learning_rate": 1.1692162981998608e-06,
+      "loss": 0.135,
+      "step": 721
+    },
+    {
+      "epoch": 0.1640909090909091,
+      "grad_norm": 3.8391698276745956,
+      "learning_rate": 1.1689967241950263e-06,
+      "loss": 0.1385,
+      "step": 722
+    },
+    {
+      "epoch": 0.1643181818181818,
+      "grad_norm": 1.7292825191602952,
+      "learning_rate": 1.168776872864146e-06,
+      "loss": 0.124,
+      "step": 723
+    },
+    {
+      "epoch": 0.16454545454545455,
+      "grad_norm": 3.2795303208570163,
+      "learning_rate": 1.1685567443192996e-06,
+      "loss": 0.1109,
+      "step": 724
+    },
+    {
+      "epoch": 0.16477272727272727,
+      "grad_norm": 2.6299019638880696,
+      "learning_rate": 1.168336338672707e-06,
+      "loss": 0.1597,
+      "step": 725
+    },
+    {
+      "epoch": 0.165,
+      "grad_norm": 3.351678847020426,
+      "learning_rate": 1.1681156560367296e-06,
+      "loss": 0.1513,
+      "step": 726
+    },
+    {
+      "epoch": 0.16522727272727272,
+      "grad_norm": 2.6050465846386937,
+      "learning_rate": 1.1678946965238697e-06,
+      "loss": 0.1982,
+      "step": 727
+    },
+    {
+      "epoch": 0.16545454545454547,
+      "grad_norm": 1.636700376895789,
+      "learning_rate": 1.1676734602467713e-06,
+      "loss": 0.0736,
+      "step": 728
+    },
+    {
+      "epoch": 0.16568181818181818,
+      "grad_norm": 2.744581614071777,
+      "learning_rate": 1.1674519473182192e-06,
+      "loss": 0.0709,
+      "step": 729
+    },
+    {
+      "epoch": 0.16590909090909092,
+      "grad_norm": 2.6395163493229963,
+      "learning_rate": 1.1672301578511392e-06,
+      "loss": 0.1673,
+      "step": 730
+    },
+    {
+      "epoch": 0.16613636363636364,
+      "grad_norm": 2.6282749168817507,
+      "learning_rate": 1.167008091958598e-06,
+      "loss": 0.0909,
+      "step": 731
+    },
+    {
+      "epoch": 0.16636363636363635,
+      "grad_norm": 2.6151049942306632,
+      "learning_rate": 1.1667857497538037e-06,
+      "loss": 0.0808,
+      "step": 732
+    },
+    {
+      "epoch": 0.1665909090909091,
+      "grad_norm": 3.886652448954216,
+      "learning_rate": 1.166563131350105e-06,
+      "loss": 0.2248,
+      "step": 733
+    },
+    {
+      "epoch": 0.1668181818181818,
+      "grad_norm": 2.3948049185305127,
+      "learning_rate": 1.166340236860991e-06,
+      "loss": 0.2084,
+      "step": 734
+    },
+    {
+      "epoch": 0.16704545454545455,
+      "grad_norm": 2.1731692572173267,
+      "learning_rate": 1.166117066400092e-06,
+      "loss": 0.1566,
+      "step": 735
+    },
+    {
+      "epoch": 0.16727272727272727,
+      "grad_norm": 4.488623179972788,
+      "learning_rate": 1.1658936200811789e-06,
+      "loss": 0.1921,
+      "step": 736
+    },
+    {
+      "epoch": 0.1675,
+      "grad_norm": 2.7339134175043935,
+      "learning_rate": 1.1656698980181633e-06,
+      "loss": 0.1844,
+      "step": 737
+    },
+    {
+      "epoch": 0.16772727272727272,
+      "grad_norm": 3.3242848191560617,
+      "learning_rate": 1.1654459003250971e-06,
+      "loss": 0.1269,
+      "step": 738
+    },
+    {
+      "epoch": 0.16795454545454547,
+      "grad_norm": 1.68341970775111,
+      "learning_rate": 1.1652216271161728e-06,
+      "loss": 0.052,
+      "step": 739
+    },
+    {
+      "epoch": 0.16818181818181818,
+      "grad_norm": 2.7139465981471074,
+      "learning_rate": 1.1649970785057238e-06,
+      "loss": 0.1893,
+      "step": 740
+    },
+    {
+      "epoch": 0.1684090909090909,
+      "grad_norm": 2.593443849301937,
+      "learning_rate": 1.1647722546082232e-06,
+      "loss": 0.0748,
+      "step": 741
+    },
+    {
+      "epoch": 0.16863636363636364,
+      "grad_norm": 3.2702456330675624,
+      "learning_rate": 1.164547155538285e-06,
+      "loss": 0.1426,
+      "step": 742
+    },
+    {
+      "epoch": 0.16886363636363635,
+      "grad_norm": 3.573667296858276,
+      "learning_rate": 1.1643217814106633e-06,
+      "loss": 0.1132,
+      "step": 743
+    },
+    {
+      "epoch": 0.1690909090909091,
+      "grad_norm": 3.7194920480146134,
+      "learning_rate": 1.1640961323402522e-06,
+      "loss": 0.1271,
+      "step": 744
+    },
+    {
+      "epoch": 0.1693181818181818,
+      "grad_norm": 2.950558394662862,
+      "learning_rate": 1.163870208442086e-06,
+      "loss": 0.1399,
+      "step": 745
+    },
+    {
+      "epoch": 0.16954545454545455,
+      "grad_norm": 2.9611810412801685,
+      "learning_rate": 1.1636440098313398e-06,
+      "loss": 0.1567,
+      "step": 746
+    },
+    {
+      "epoch": 0.16977272727272727,
+      "grad_norm": 2.0824725181288923,
+      "learning_rate": 1.1634175366233278e-06,
+      "loss": 0.204,
+      "step": 747
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.364186147315778,
+      "learning_rate": 1.1631907889335046e-06,
+      "loss": 0.127,
+      "step": 748
+    },
+    {
+      "epoch": 0.17022727272727273,
+      "grad_norm": 3.8788218098745575,
+      "learning_rate": 1.162963766877465e-06,
+      "loss": 0.1303,
+      "step": 749
+    },
+    {
+      "epoch": 0.17045454545454544,
+      "grad_norm": 2.2417412756946202,
+      "learning_rate": 1.1627364705709428e-06,
+      "loss": 0.1509,
+      "step": 750
+    },
+    {
+      "epoch": 0.17068181818181818,
+      "grad_norm": 2.842048294156292,
+      "learning_rate": 1.1625089001298129e-06,
+      "loss": 0.1327,
+      "step": 751
+    },
+    {
+      "epoch": 0.1709090909090909,
+      "grad_norm": 4.29305276257053,
+      "learning_rate": 1.1622810556700889e-06,
+      "loss": 0.2562,
+      "step": 752
+    },
+    {
+      "epoch": 0.17113636363636364,
+      "grad_norm": 2.761534576158799,
+      "learning_rate": 1.1620529373079246e-06,
+      "loss": 0.1311,
+      "step": 753
+    },
+    {
+      "epoch": 0.17136363636363636,
+      "grad_norm": 2.5152999927690254,
+      "learning_rate": 1.1618245451596128e-06,
+      "loss": 0.0815,
+      "step": 754
+    },
+    {
+      "epoch": 0.1715909090909091,
+      "grad_norm": 1.8540579522207619,
+      "learning_rate": 1.161595879341587e-06,
+      "loss": 0.0669,
+      "step": 755
+    },
+    {
+      "epoch": 0.17181818181818181,
+      "grad_norm": 3.1281319361472573,
+      "learning_rate": 1.1613669399704192e-06,
+      "loss": 0.1412,
+      "step": 756
+    },
+    {
+      "epoch": 0.17204545454545456,
+      "grad_norm": 2.8443966158376814,
+      "learning_rate": 1.1611377271628213e-06,
+      "loss": 0.152,
+      "step": 757
+    },
+    {
+      "epoch": 0.17227272727272727,
+      "grad_norm": 3.207403833018733,
+      "learning_rate": 1.1609082410356447e-06,
+      "loss": 0.1002,
+      "step": 758
+    },
+    {
+      "epoch": 0.1725,
+      "grad_norm": 3.422950619905195,
+      "learning_rate": 1.1606784817058797e-06,
+      "loss": 0.1737,
+      "step": 759
+    },
+    {
+      "epoch": 0.17272727272727273,
+      "grad_norm": 3.2687819031909506,
+      "learning_rate": 1.1604484492906562e-06,
+      "loss": 0.1867,
+      "step": 760
+    },
+    {
+      "epoch": 0.17295454545454544,
+      "grad_norm": 3.214210831160074,
+      "learning_rate": 1.1602181439072432e-06,
+      "loss": 0.0691,
+      "step": 761
+    },
+    {
+      "epoch": 0.1731818181818182,
+      "grad_norm": 3.8790707494088346,
+      "learning_rate": 1.1599875656730492e-06,
+      "loss": 0.1843,
+      "step": 762
+    },
+    {
+      "epoch": 0.1734090909090909,
+      "grad_norm": 1.687132461957667,
+      "learning_rate": 1.159756714705621e-06,
+      "loss": 0.1179,
+      "step": 763
+    },
+    {
+      "epoch": 0.17363636363636364,
+      "grad_norm": 4.14352268585676,
+      "learning_rate": 1.1595255911226456e-06,
+      "loss": 0.1386,
+      "step": 764
+    },
+    {
+      "epoch": 0.17386363636363636,
+      "grad_norm": 3.714452698007248,
+      "learning_rate": 1.1592941950419475e-06,
+      "loss": 0.1369,
+      "step": 765
+    },
+    {
+      "epoch": 0.1740909090909091,
+      "grad_norm": 1.789381702846916,
+      "learning_rate": 1.1590625265814918e-06,
+      "loss": 0.1635,
+      "step": 766
+    },
+    {
+      "epoch": 0.17431818181818182,
+      "grad_norm": 2.607001331930699,
+      "learning_rate": 1.1588305858593811e-06,
+      "loss": 0.128,
+      "step": 767
+    },
+    {
+      "epoch": 0.17454545454545456,
+      "grad_norm": 3.4989510487110964,
+      "learning_rate": 1.1585983729938575e-06,
+      "loss": 0.1287,
+      "step": 768
+    },
+    {
+      "epoch": 0.17477272727272727,
+      "grad_norm": 2.6836259080834473,
+      "learning_rate": 1.1583658881033013e-06,
+      "loss": 0.1004,
+      "step": 769
+    },
+    {
+      "epoch": 0.175,
+      "grad_norm": 5.482137218229676,
+      "learning_rate": 1.158133131306232e-06,
+      "loss": 0.1289,
+      "step": 770
+    },
+    {
+      "epoch": 0.17522727272727273,
+      "grad_norm": 3.056367376708317,
+      "learning_rate": 1.1579001027213078e-06,
+      "loss": 0.0944,
+      "step": 771
+    },
+    {
+      "epoch": 0.17545454545454545,
+      "grad_norm": 2.6085852855265808,
+      "learning_rate": 1.1576668024673248e-06,
+      "loss": 0.1161,
+      "step": 772
+    },
+    {
+      "epoch": 0.1756818181818182,
+      "grad_norm": 3.3670444379456463,
+      "learning_rate": 1.157433230663218e-06,
+      "loss": 0.0862,
+      "step": 773
+    },
+    {
+      "epoch": 0.1759090909090909,
+      "grad_norm": 3.9124094933737474,
+      "learning_rate": 1.1571993874280611e-06,
+      "loss": 0.1653,
+      "step": 774
+    },
+    {
+      "epoch": 0.17613636363636365,
+      "grad_norm": 2.098372213664325,
+      "learning_rate": 1.1569652728810658e-06,
+      "loss": 0.1056,
+      "step": 775
+    },
+    {
+      "epoch": 0.17636363636363636,
+      "grad_norm": 4.856787230693005,
+      "learning_rate": 1.156730887141582e-06,
+      "loss": 0.1548,
+      "step": 776
+    },
+    {
+      "epoch": 0.1765909090909091,
+      "grad_norm": 3.238756391913626,
+      "learning_rate": 1.156496230329098e-06,
+      "loss": 0.2339,
+      "step": 777
+    },
+    {
+      "epoch": 0.17681818181818182,
+      "grad_norm": 1.595387561954138,
+      "learning_rate": 1.1562613025632406e-06,
+      "loss": 0.0849,
+      "step": 778
+    },
+    {
+      "epoch": 0.17704545454545453,
+      "grad_norm": 2.0727998006207287,
+      "learning_rate": 1.1560261039637744e-06,
+      "loss": 0.0652,
+      "step": 779
+    },
+    {
+      "epoch": 0.17727272727272728,
+      "grad_norm": 2.2064050919258307,
+      "learning_rate": 1.1557906346506021e-06,
+      "loss": 0.0835,
+      "step": 780
+    },
+    {
+      "epoch": 0.1775,
+      "grad_norm": 1.9946221213508033,
+      "learning_rate": 1.1555548947437642e-06,
+      "loss": 0.1185,
+      "step": 781
+    },
+    {
+      "epoch": 0.17772727272727273,
+      "grad_norm": 3.129919046350792,
+      "learning_rate": 1.1553188843634399e-06,
+      "loss": 0.0898,
+      "step": 782
+    },
+    {
+      "epoch": 0.17795454545454545,
+      "grad_norm": 2.886433483282702,
+      "learning_rate": 1.1550826036299455e-06,
+      "loss": 0.2483,
+      "step": 783
+    },
+    {
+      "epoch": 0.1781818181818182,
+      "grad_norm": 2.394012602417068,
+      "learning_rate": 1.1548460526637354e-06,
+      "loss": 0.1536,
+      "step": 784
+    },
+    {
+      "epoch": 0.1784090909090909,
+      "grad_norm": 2.9399967933076816,
+      "learning_rate": 1.1546092315854017e-06,
+      "loss": 0.1504,
+      "step": 785
+    },
+    {
+      "epoch": 0.17863636363636365,
+      "grad_norm": 2.4053756077601784,
+      "learning_rate": 1.1543721405156744e-06,
+      "loss": 0.1932,
+      "step": 786
+    },
+    {
+      "epoch": 0.17886363636363636,
+      "grad_norm": 1.9507459878797975,
+      "learning_rate": 1.154134779575421e-06,
+      "loss": 0.1568,
+      "step": 787
+    },
+    {
+      "epoch": 0.17909090909090908,
+      "grad_norm": 3.8163320945783927,
+      "learning_rate": 1.1538971488856465e-06,
+      "loss": 0.1545,
+      "step": 788
+    },
+    {
+      "epoch": 0.17931818181818182,
+      "grad_norm": 1.6661912582967584,
+      "learning_rate": 1.153659248567494e-06,
+      "loss": 0.0652,
+      "step": 789
+    },
+    {
+      "epoch": 0.17954545454545454,
+      "grad_norm": 2.403519177990125,
+      "learning_rate": 1.1534210787422425e-06,
+      "loss": 0.1127,
+      "step": 790
+    },
+    {
+      "epoch": 0.17977272727272728,
+      "grad_norm": 3.236244380538794,
+      "learning_rate": 1.1531826395313104e-06,
+      "loss": 0.0983,
+      "step": 791
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.122823591783606,
+      "learning_rate": 1.152943931056252e-06,
+      "loss": 0.3468,
+      "step": 792
+    },
+    {
+      "epoch": 0.18022727272727274,
+      "grad_norm": 2.4556581997450757,
+      "learning_rate": 1.1527049534387595e-06,
+      "loss": 0.2062,
+      "step": 793
+    },
+    {
+      "epoch": 0.18045454545454545,
+      "grad_norm": 3.064315627085115,
+      "learning_rate": 1.1524657068006622e-06,
+      "loss": 0.102,
+      "step": 794
+    },
+    {
+      "epoch": 0.1806818181818182,
+      "grad_norm": 2.24227763719453,
+      "learning_rate": 1.1522261912639266e-06,
+      "loss": 0.1154,
+      "step": 795
+    },
+    {
+      "epoch": 0.1809090909090909,
+      "grad_norm": 3.1021337311882817,
+      "learning_rate": 1.151986406950656e-06,
+      "loss": 0.1356,
+      "step": 796
+    },
+    {
+      "epoch": 0.18113636363636362,
+      "grad_norm": 2.309947804508055,
+      "learning_rate": 1.1517463539830908e-06,
+      "loss": 0.1851,
+      "step": 797
+    },
+    {
+      "epoch": 0.18136363636363637,
+      "grad_norm": 2.3461503275742617,
+      "learning_rate": 1.1515060324836088e-06,
+      "loss": 0.1146,
+      "step": 798
+    },
+    {
+      "epoch": 0.18159090909090908,
+      "grad_norm": 0.942908115083293,
+      "learning_rate": 1.151265442574724e-06,
+      "loss": 0.0868,
+      "step": 799
+    },
+    {
+      "epoch": 0.18181818181818182,
+      "grad_norm": 5.253767738350941,
+      "learning_rate": 1.151024584379088e-06,
+      "loss": 0.1662,
+      "step": 800
+    },
+    {
+      "epoch": 0.18204545454545454,
+      "grad_norm": 2.1044537190139914,
+      "learning_rate": 1.1507834580194883e-06,
+      "loss": 0.1585,
+      "step": 801
+    },
+    {
+      "epoch": 0.18227272727272728,
+      "grad_norm": 2.2973847557647202,
+      "learning_rate": 1.15054206361885e-06,
+      "loss": 0.0663,
+      "step": 802
+    },
+    {
+      "epoch": 0.1825,
+      "grad_norm": 2.6834847508092166,
+      "learning_rate": 1.1503004013002341e-06,
+      "loss": 0.103,
+      "step": 803
+    },
+    {
+      "epoch": 0.18272727272727274,
+      "grad_norm": 3.57442874476484,
+      "learning_rate": 1.1500584711868388e-06,
+      "loss": 0.1276,
+      "step": 804
+    },
+    {
+      "epoch": 0.18295454545454545,
+      "grad_norm": 3.086493657377809,
+      "learning_rate": 1.1498162734019983e-06,
+      "loss": 0.0971,
+      "step": 805
+    },
+    {
+      "epoch": 0.1831818181818182,
+      "grad_norm": 4.351195022145964,
+      "learning_rate": 1.1495738080691834e-06,
+      "loss": 0.1499,
+      "step": 806
+    },
+    {
+      "epoch": 0.1834090909090909,
+      "grad_norm": 1.6156089715781095,
+      "learning_rate": 1.1493310753120017e-06,
+      "loss": 0.1203,
+      "step": 807
+    },
+    {
+      "epoch": 0.18363636363636363,
+      "grad_norm": 2.70754299009439,
+      "learning_rate": 1.1490880752541967e-06,
+      "loss": 0.1326,
+      "step": 808
+    },
+    {
+      "epoch": 0.18386363636363637,
+      "grad_norm": 2.5345969833080644,
+      "learning_rate": 1.148844808019648e-06,
+      "loss": 0.1719,
+      "step": 809
+    },
+    {
+      "epoch": 0.18409090909090908,
+      "grad_norm": 3.3577061986890184,
+      "learning_rate": 1.1486012737323718e-06,
+      "loss": 0.118,
+      "step": 810
+    },
+    {
+      "epoch": 0.18431818181818183,
+      "grad_norm": 2.438615273574655,
+      "learning_rate": 1.1483574725165207e-06,
+      "loss": 0.1311,
+      "step": 811
+    },
+    {
+      "epoch": 0.18454545454545454,
+      "grad_norm": 1.1633166691708112,
+      "learning_rate": 1.1481134044963825e-06,
+      "loss": 0.0904,
+      "step": 812
+    },
+    {
+      "epoch": 0.18477272727272728,
+      "grad_norm": 2.3699404360953973,
+      "learning_rate": 1.1478690697963814e-06,
+      "loss": 0.1526,
+      "step": 813
+    },
+    {
+      "epoch": 0.185,
+      "grad_norm": 2.7464445198549066,
+      "learning_rate": 1.1476244685410784e-06,
+      "loss": 0.0781,
+      "step": 814
+    },
+    {
+      "epoch": 0.18522727272727274,
+      "grad_norm": 2.8879685130608843,
+      "learning_rate": 1.147379600855169e-06,
+      "loss": 0.2298,
+      "step": 815
+    },
+    {
+      "epoch": 0.18545454545454546,
+      "grad_norm": 3.064297916245351,
+      "learning_rate": 1.1471344668634854e-06,
+      "loss": 0.1948,
+      "step": 816
+    },
+    {
+      "epoch": 0.18568181818181817,
+      "grad_norm": 2.188691429463574,
+      "learning_rate": 1.1468890666909954e-06,
+      "loss": 0.175,
+      "step": 817
+    },
+    {
+      "epoch": 0.1859090909090909,
+      "grad_norm": 3.601284618296476,
+      "learning_rate": 1.1466434004628023e-06,
+      "loss": 0.1424,
+      "step": 818
+    },
+    {
+      "epoch": 0.18613636363636363,
+      "grad_norm": 4.044923514227203,
+      "learning_rate": 1.1463974683041455e-06,
+      "loss": 0.2224,
+      "step": 819
+    },
+    {
+      "epoch": 0.18636363636363637,
+      "grad_norm": 3.1276602862595073,
+      "learning_rate": 1.1461512703403992e-06,
+      "loss": 0.1508,
+      "step": 820
+    },
+    {
+      "epoch": 0.18659090909090909,
+      "grad_norm": 2.0888530301886923,
+      "learning_rate": 1.1459048066970736e-06,
+      "loss": 0.1487,
+      "step": 821
+    },
+    {
+      "epoch": 0.18681818181818183,
+      "grad_norm": 2.494265349166839,
+      "learning_rate": 1.1456580774998146e-06,
+      "loss": 0.1014,
+      "step": 822
+    },
+    {
+      "epoch": 0.18704545454545454,
+      "grad_norm": 2.1786462475879738,
+      "learning_rate": 1.1454110828744027e-06,
+      "loss": 0.1205,
+      "step": 823
+    },
+    {
+      "epoch": 0.18727272727272729,
+      "grad_norm": 2.720243495612787,
+      "learning_rate": 1.1451638229467547e-06,
+      "loss": 0.0863,
+      "step": 824
+    },
+    {
+      "epoch": 0.1875,
+      "grad_norm": 2.9322880363083734,
+      "learning_rate": 1.1449162978429218e-06,
+      "loss": 0.0924,
+      "step": 825
+    },
+    {
+      "epoch": 0.18772727272727271,
+      "grad_norm": 4.562235969968607,
+      "learning_rate": 1.144668507689091e-06,
+      "loss": 0.09,
+      "step": 826
+    },
+    {
+      "epoch": 0.18795454545454546,
+      "grad_norm": 5.296921231454007,
+      "learning_rate": 1.1444204526115837e-06,
+      "loss": 0.1391,
+      "step": 827
+    },
+    {
+      "epoch": 0.18818181818181817,
+      "grad_norm": 2.097742080295468,
+      "learning_rate": 1.144172132736857e-06,
+      "loss": 0.1163,
+      "step": 828
+    },
+    {
+      "epoch": 0.18840909090909091,
+      "grad_norm": 3.919117935193303,
+      "learning_rate": 1.1439235481915028e-06,
+      "loss": 0.1186,
+      "step": 829
+    },
+    {
+      "epoch": 0.18863636363636363,
+      "grad_norm": 2.880486406032231,
+      "learning_rate": 1.1436746991022479e-06,
+      "loss": 0.1004,
+      "step": 830
+    },
+    {
+      "epoch": 0.18886363636363637,
+      "grad_norm": 3.3546769794061704,
+      "learning_rate": 1.143425585595954e-06,
+      "loss": 0.1239,
+      "step": 831
+    },
+    {
+      "epoch": 0.1890909090909091,
+      "grad_norm": 3.0779119093071983,
+      "learning_rate": 1.1431762077996174e-06,
+      "loss": 0.1695,
+      "step": 832
+    },
+    {
+      "epoch": 0.18931818181818183,
+      "grad_norm": 2.1119083798272213,
+      "learning_rate": 1.1429265658403698e-06,
+      "loss": 0.0827,
+      "step": 833
+    },
+    {
+      "epoch": 0.18954545454545454,
+      "grad_norm": 4.262897532682798,
+      "learning_rate": 1.1426766598454768e-06,
+      "loss": 0.1279,
+      "step": 834
+    },
+    {
+      "epoch": 0.18977272727272726,
+      "grad_norm": 2.7040659996546705,
+      "learning_rate": 1.1424264899423383e-06,
+      "loss": 0.1629,
+      "step": 835
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1278986481807958,
+      "learning_rate": 1.1421760562584901e-06,
+      "loss": 0.1207,
+      "step": 836
+    },
+    {
+      "epoch": 0.19022727272727272,
+      "grad_norm": 2.3069458205276927,
+      "learning_rate": 1.1419253589216012e-06,
+      "loss": 0.1782,
+      "step": 837
+    },
+    {
+      "epoch": 0.19045454545454546,
+      "grad_norm": 3.767012878461527,
+      "learning_rate": 1.1416743980594758e-06,
+      "loss": 0.1153,
+      "step": 838
+    },
+    {
+      "epoch": 0.19068181818181817,
+      "grad_norm": 2.6741229655400875,
+      "learning_rate": 1.1414231738000521e-06,
+      "loss": 0.1789,
+      "step": 839
+    },
+    {
+      "epoch": 0.19090909090909092,
+      "grad_norm": 2.101009507945506,
+      "learning_rate": 1.1411716862714027e-06,
+      "loss": 0.0653,
+      "step": 840
+    },
+    {
+      "epoch": 0.19113636363636363,
+      "grad_norm": 3.689397448197424,
+      "learning_rate": 1.1409199356017339e-06,
+      "loss": 0.1148,
+      "step": 841
+    },
+    {
+      "epoch": 0.19136363636363637,
+      "grad_norm": 2.0420772527481375,
+      "learning_rate": 1.140667921919387e-06,
+      "loss": 0.2031,
+      "step": 842
+    },
+    {
+      "epoch": 0.1915909090909091,
+      "grad_norm": 2.026188196012587,
+      "learning_rate": 1.1404156453528367e-06,
+      "loss": 0.0976,
+      "step": 843
+    },
+    {
+      "epoch": 0.1918181818181818,
+      "grad_norm": 6.012679607100995,
+      "learning_rate": 1.1401631060306921e-06,
+      "loss": 0.1494,
+      "step": 844
+    },
+    {
+      "epoch": 0.19204545454545455,
+      "grad_norm": 2.423274028563013,
+      "learning_rate": 1.1399103040816963e-06,
+      "loss": 0.124,
+      "step": 845
+    },
+    {
+      "epoch": 0.19227272727272726,
+      "grad_norm": 2.794934111794361,
+      "learning_rate": 1.1396572396347257e-06,
+      "loss": 0.2355,
+      "step": 846
+    },
+    {
+      "epoch": 0.1925,
+      "grad_norm": 2.6667405221550733,
+      "learning_rate": 1.1394039128187914e-06,
+      "loss": 0.1304,
+      "step": 847
+    },
+    {
+      "epoch": 0.19272727272727272,
+      "grad_norm": 3.30636603306502,
+      "learning_rate": 1.1391503237630375e-06,
+      "loss": 0.1543,
+      "step": 848
+    },
+    {
+      "epoch": 0.19295454545454546,
+      "grad_norm": 1.8806280285001553,
+      "learning_rate": 1.1388964725967423e-06,
+      "loss": 0.0975,
+      "step": 849
+    },
+    {
+      "epoch": 0.19318181818181818,
+      "grad_norm": 2.3424660602947536,
+      "learning_rate": 1.138642359449317e-06,
+      "loss": 0.0582,
+      "step": 850
+    },
+    {
+      "epoch": 0.19340909090909092,
+      "grad_norm": 2.4441563812127862,
+      "learning_rate": 1.1383879844503073e-06,
+      "loss": 0.0748,
+      "step": 851
+    },
+    {
+      "epoch": 0.19363636363636363,
+      "grad_norm": 4.225068347382088,
+      "learning_rate": 1.1381333477293918e-06,
+      "loss": 0.1792,
+      "step": 852
+    },
+    {
+      "epoch": 0.19386363636363638,
+      "grad_norm": 2.2576982964584924,
+      "learning_rate": 1.137878449416383e-06,
+      "loss": 0.1127,
+      "step": 853
+    },
+    {
+      "epoch": 0.1940909090909091,
+      "grad_norm": 1.7725145986402941,
+      "learning_rate": 1.137623289641226e-06,
+      "loss": 0.1166,
+      "step": 854
+    },
+    {
+      "epoch": 0.1943181818181818,
+      "grad_norm": 3.923996731343817,
+      "learning_rate": 1.1373678685339994e-06,
+      "loss": 0.1696,
+      "step": 855
+    },
+    {
+      "epoch": 0.19454545454545455,
+      "grad_norm": 1.4670578602291842,
+      "learning_rate": 1.137112186224916e-06,
+      "loss": 0.117,
+      "step": 856
+    },
+    {
+      "epoch": 0.19477272727272726,
+      "grad_norm": 5.245237689880103,
+      "learning_rate": 1.1368562428443205e-06,
+      "loss": 0.1649,
+      "step": 857
+    },
+    {
+      "epoch": 0.195,
+      "grad_norm": 6.813292589259737,
+      "learning_rate": 1.1366000385226913e-06,
+      "loss": 0.2304,
+      "step": 858
+    },
+    {
+      "epoch": 0.19522727272727272,
+      "grad_norm": 2.217787446965733,
+      "learning_rate": 1.1363435733906398e-06,
+      "loss": 0.0719,
+      "step": 859
+    },
+    {
+      "epoch": 0.19545454545454546,
+      "grad_norm": 2.397366324280405,
+      "learning_rate": 1.13608684757891e-06,
+      "loss": 0.1213,
+      "step": 860
+    },
+    {
+      "epoch": 0.19568181818181818,
+      "grad_norm": 2.5570553414412687,
+      "learning_rate": 1.1358298612183793e-06,
+      "loss": 0.0863,
+      "step": 861
+    },
+    {
+      "epoch": 0.19590909090909092,
+      "grad_norm": 2.8288238369595455,
+      "learning_rate": 1.135572614440058e-06,
+      "loss": 0.2539,
+      "step": 862
+    },
+    {
+      "epoch": 0.19613636363636364,
+      "grad_norm": 6.300657019943351,
+      "learning_rate": 1.1353151073750882e-06,
+      "loss": 0.2171,
+      "step": 863
+    },
+    {
+      "epoch": 0.19636363636363635,
+      "grad_norm": 3.5311844248331283,
+      "learning_rate": 1.1350573401547457e-06,
+      "loss": 0.1387,
+      "step": 864
+    },
+    {
+      "epoch": 0.1965909090909091,
+      "grad_norm": 5.242792892906364,
+      "learning_rate": 1.1347993129104386e-06,
+      "loss": 0.1572,
+      "step": 865
+    },
+    {
+      "epoch": 0.1968181818181818,
+      "grad_norm": 4.707273140428848,
+      "learning_rate": 1.1345410257737078e-06,
+      "loss": 0.1402,
+      "step": 866
+    },
+    {
+      "epoch": 0.19704545454545455,
+      "grad_norm": 2.5120161670170797,
+      "learning_rate": 1.1342824788762258e-06,
+      "loss": 0.0992,
+      "step": 867
+    },
+    {
+      "epoch": 0.19727272727272727,
+      "grad_norm": 2.4152824803402875,
+      "learning_rate": 1.1340236723497985e-06,
+      "loss": 0.113,
+      "step": 868
+    },
+    {
+      "epoch": 0.1975,
+      "grad_norm": 3.574264532835847,
+      "learning_rate": 1.133764606326364e-06,
+      "loss": 0.0918,
+      "step": 869
+    },
+    {
+      "epoch": 0.19772727272727272,
+      "grad_norm": 1.9916657489331235,
+      "learning_rate": 1.1335052809379921e-06,
+      "loss": 0.0855,
+      "step": 870
+    },
+    {
+      "epoch": 0.19795454545454547,
+      "grad_norm": 2.4090547500582047,
+      "learning_rate": 1.1332456963168854e-06,
+      "loss": 0.1141,
+      "step": 871
+    },
+    {
+      "epoch": 0.19818181818181818,
+      "grad_norm": 3.2552556158403405,
+      "learning_rate": 1.1329858525953785e-06,
+      "loss": 0.1718,
+      "step": 872
+    },
+    {
+      "epoch": 0.1984090909090909,
+      "grad_norm": 2.549162059679049,
+      "learning_rate": 1.132725749905938e-06,
+      "loss": 0.1004,
+      "step": 873
+    },
+    {
+      "epoch": 0.19863636363636364,
+      "grad_norm": 2.744893739793677,
+      "learning_rate": 1.132465388381163e-06,
+      "loss": 0.1362,
+      "step": 874
+    },
+    {
+      "epoch": 0.19886363636363635,
+      "grad_norm": 6.228555512888278,
+      "learning_rate": 1.1322047681537834e-06,
+      "loss": 0.134,
+      "step": 875
+    },
+    {
+      "epoch": 0.1990909090909091,
+      "grad_norm": 3.0396334940780663,
+      "learning_rate": 1.131943889356662e-06,
+      "loss": 0.1455,
+      "step": 876
+    },
+    {
+      "epoch": 0.1993181818181818,
+      "grad_norm": 2.27528457032552,
+      "learning_rate": 1.1316827521227935e-06,
+      "loss": 0.0779,
+      "step": 877
+    },
+    {
+      "epoch": 0.19954545454545455,
+      "grad_norm": 5.399676208752103,
+      "learning_rate": 1.1314213565853036e-06,
+      "loss": 0.1722,
+      "step": 878
+    },
+    {
+      "epoch": 0.19977272727272727,
+      "grad_norm": 4.269701909032954,
+      "learning_rate": 1.1311597028774503e-06,
+      "loss": 0.2082,
+      "step": 879
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.359319367499375,
+      "learning_rate": 1.1308977911326229e-06,
+      "loss": 0.128,
+      "step": 880
+    },
+    {
+      "epoch": 0.20022727272727273,
+      "grad_norm": 1.3252010285703189,
+      "learning_rate": 1.1306356214843423e-06,
+      "loss": 0.0955,
+      "step": 881
+    },
+    {
+      "epoch": 0.20045454545454544,
+      "grad_norm": 3.279146239163216,
+      "learning_rate": 1.1303731940662608e-06,
+      "loss": 0.1206,
+      "step": 882
+    },
+    {
+      "epoch": 0.20068181818181818,
+      "grad_norm": 2.31483626518365,
+      "learning_rate": 1.1301105090121624e-06,
+      "loss": 0.1602,
+      "step": 883
+    },
+    {
+      "epoch": 0.2009090909090909,
+      "grad_norm": 3.3023889936699606,
+      "learning_rate": 1.1298475664559622e-06,
+      "loss": 0.1623,
+      "step": 884
+    },
+    {
+      "epoch": 0.20113636363636364,
+      "grad_norm": 3.114987568643515,
+      "learning_rate": 1.1295843665317067e-06,
+      "loss": 0.1231,
+      "step": 885
+    },
+    {
+      "epoch": 0.20136363636363636,
+      "grad_norm": 3.0708139471057856,
+      "learning_rate": 1.1293209093735732e-06,
+      "loss": 0.108,
+      "step": 886
+    },
+    {
+      "epoch": 0.2015909090909091,
+      "grad_norm": 5.7094438118297886,
+      "learning_rate": 1.129057195115871e-06,
+      "loss": 0.2208,
+      "step": 887
+    },
+    {
+      "epoch": 0.2018181818181818,
+      "grad_norm": 2.6711769318759164,
+      "learning_rate": 1.1287932238930397e-06,
+      "loss": 0.1125,
+      "step": 888
+    },
+    {
+      "epoch": 0.20204545454545456,
+      "grad_norm": 2.0402523276660633,
+      "learning_rate": 1.12852899583965e-06,
+      "loss": 0.1404,
+      "step": 889
+    },
+    {
+      "epoch": 0.20227272727272727,
+      "grad_norm": 1.8261942619085358,
+      "learning_rate": 1.1282645110904036e-06,
+      "loss": 0.1148,
+      "step": 890
+    },
+    {
+      "epoch": 0.2025,
+      "grad_norm": 1.6194431003325316,
+      "learning_rate": 1.1279997697801334e-06,
+      "loss": 0.1125,
+      "step": 891
+    },
+    {
+      "epoch": 0.20272727272727273,
+      "grad_norm": 1.3206650109215734,
+      "learning_rate": 1.1277347720438028e-06,
+      "loss": 0.1133,
+      "step": 892
+    },
+    {
+      "epoch": 0.20295454545454544,
+      "grad_norm": 1.7701566234625217,
+      "learning_rate": 1.127469518016506e-06,
+      "loss": 0.094,
+      "step": 893
+    },
+    {
+      "epoch": 0.20318181818181819,
+      "grad_norm": 2.5929160004023983,
+      "learning_rate": 1.1272040078334675e-06,
+      "loss": 0.1602,
+      "step": 894
+    },
+    {
+      "epoch": 0.2034090909090909,
+      "grad_norm": 2.248541580867578,
+      "learning_rate": 1.1269382416300429e-06,
+      "loss": 0.1006,
+      "step": 895
+    },
+    {
+      "epoch": 0.20363636363636364,
+      "grad_norm": 2.9153956540464256,
+      "learning_rate": 1.126672219541718e-06,
+      "loss": 0.2115,
+      "step": 896
+    },
+    {
+      "epoch": 0.20386363636363636,
+      "grad_norm": 3.65387102331878,
+      "learning_rate": 1.1264059417041095e-06,
+      "loss": 0.1454,
+      "step": 897
+    },
+    {
+      "epoch": 0.2040909090909091,
+      "grad_norm": 1.5094855026951302,
+      "learning_rate": 1.1261394082529638e-06,
+      "loss": 0.0551,
+      "step": 898
+    },
+    {
+      "epoch": 0.20431818181818182,
+      "grad_norm": 3.864726857371483,
+      "learning_rate": 1.1258726193241578e-06,
+      "loss": 0.1419,
+      "step": 899
+    },
+    {
+      "epoch": 0.20454545454545456,
+      "grad_norm": 2.850781671661656,
+      "learning_rate": 1.1256055750536992e-06,
+      "loss": 0.0795,
+      "step": 900
+    },
+    {
+      "epoch": 0.20477272727272727,
+      "grad_norm": 4.739241650980207,
+      "learning_rate": 1.1253382755777253e-06,
+      "loss": 0.1467,
+      "step": 901
+    },
+    {
+      "epoch": 0.205,
+      "grad_norm": 2.4142682855998765,
+      "learning_rate": 1.1250707210325036e-06,
+      "loss": 0.2125,
+      "step": 902
+    },
+    {
+      "epoch": 0.20522727272727273,
+      "grad_norm": 2.437738528653348,
+      "learning_rate": 1.1248029115544319e-06,
+      "loss": 0.1445,
+      "step": 903
+    },
+    {
+      "epoch": 0.20545454545454545,
+      "grad_norm": 2.0506459414562177,
+      "learning_rate": 1.1245348472800372e-06,
+      "loss": 0.1562,
+      "step": 904
+    },
+    {
+      "epoch": 0.2056818181818182,
+      "grad_norm": 4.821329278134585,
+      "learning_rate": 1.1242665283459775e-06,
+      "loss": 0.205,
+      "step": 905
+    },
+    {
+      "epoch": 0.2059090909090909,
+      "grad_norm": 4.947066169721815,
+      "learning_rate": 1.1239979548890398e-06,
+      "loss": 0.1866,
+      "step": 906
+    },
+    {
+      "epoch": 0.20613636363636365,
+      "grad_norm": 2.3156689893828375,
+      "learning_rate": 1.1237291270461413e-06,
+      "loss": 0.1181,
+      "step": 907
+    },
+    {
+      "epoch": 0.20636363636363636,
+      "grad_norm": 1.4177210897394543,
+      "learning_rate": 1.1234600449543288e-06,
+      "loss": 0.1124,
+      "step": 908
+    },
+    {
+      "epoch": 0.2065909090909091,
+      "grad_norm": 2.8389581420918675,
+      "learning_rate": 1.123190708750778e-06,
+      "loss": 0.1638,
+      "step": 909
+    },
+    {
+      "epoch": 0.20681818181818182,
+      "grad_norm": 3.8570171047006285,
+      "learning_rate": 1.1229211185727957e-06,
+      "loss": 0.13,
+      "step": 910
+    },
+    {
+      "epoch": 0.20704545454545453,
+      "grad_norm": 2.1274411895782115,
+      "learning_rate": 1.1226512745578166e-06,
+      "loss": 0.0876,
+      "step": 911
+    },
+    {
+      "epoch": 0.20727272727272728,
+      "grad_norm": 3.0596409134864366,
+      "learning_rate": 1.1223811768434057e-06,
+      "loss": 0.1303,
+      "step": 912
+    },
+    {
+      "epoch": 0.2075,
+      "grad_norm": 1.639377229813019,
+      "learning_rate": 1.122110825567257e-06,
+      "loss": 0.1302,
+      "step": 913
+    },
+    {
+      "epoch": 0.20772727272727273,
+      "grad_norm": 3.1928300415418227,
+      "learning_rate": 1.1218402208671938e-06,
+      "loss": 0.1128,
+      "step": 914
+    },
+    {
+      "epoch": 0.20795454545454545,
+      "grad_norm": 3.434681729591869,
+      "learning_rate": 1.1215693628811688e-06,
+      "loss": 0.1671,
+      "step": 915
+    },
+    {
+      "epoch": 0.2081818181818182,
+      "grad_norm": 2.3033193556128015,
+      "learning_rate": 1.1212982517472636e-06,
+      "loss": 0.0877,
+      "step": 916
+    },
+    {
+      "epoch": 0.2084090909090909,
+      "grad_norm": 2.699665931194784,
+      "learning_rate": 1.1210268876036888e-06,
+      "loss": 0.0827,
+      "step": 917
+    },
+    {
+      "epoch": 0.20863636363636365,
+      "grad_norm": 3.2319152945051206,
+      "learning_rate": 1.1207552705887841e-06,
+      "loss": 0.1262,
+      "step": 918
+    },
+    {
+      "epoch": 0.20886363636363636,
+      "grad_norm": 2.133130122228573,
+      "learning_rate": 1.1204834008410184e-06,
+      "loss": 0.1101,
+      "step": 919
+    },
+    {
+      "epoch": 0.20909090909090908,
+      "grad_norm": 2.774745813434787,
+      "learning_rate": 1.1202112784989891e-06,
+      "loss": 0.1443,
+      "step": 920
+    },
+    {
+      "epoch": 0.20931818181818182,
+      "grad_norm": 3.252816756951679,
+      "learning_rate": 1.1199389037014221e-06,
+      "loss": 0.1607,
+      "step": 921
+    },
+    {
+      "epoch": 0.20954545454545453,
+      "grad_norm": 2.5134713298937865,
+      "learning_rate": 1.1196662765871725e-06,
+      "loss": 0.1029,
+      "step": 922
+    },
+    {
+      "epoch": 0.20977272727272728,
+      "grad_norm": 1.9727785243064362,
+      "learning_rate": 1.119393397295224e-06,
+      "loss": 0.1487,
+      "step": 923
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.08094999061293,
+      "learning_rate": 1.1191202659646883e-06,
+      "loss": 0.1205,
+      "step": 924
+    },
+    {
+      "epoch": 0.21022727272727273,
+      "grad_norm": 3.342636897922391,
+      "learning_rate": 1.1188468827348066e-06,
+      "loss": 0.227,
+      "step": 925
+    },
+    {
+      "epoch": 0.21045454545454545,
+      "grad_norm": 1.5144360248913569,
+      "learning_rate": 1.1185732477449475e-06,
+      "loss": 0.0549,
+      "step": 926
+    },
+    {
+      "epoch": 0.2106818181818182,
+      "grad_norm": 1.6019820738101995,
+      "learning_rate": 1.1182993611346084e-06,
+      "loss": 0.069,
+      "step": 927
+    },
+    {
+      "epoch": 0.2109090909090909,
+      "grad_norm": 1.974698756000631,
+      "learning_rate": 1.1180252230434151e-06,
+      "loss": 0.1271,
+      "step": 928
+    },
+    {
+      "epoch": 0.21113636363636365,
+      "grad_norm": 2.3185819260946667,
+      "learning_rate": 1.1177508336111215e-06,
+      "loss": 0.0966,
+      "step": 929
+    },
+    {
+      "epoch": 0.21136363636363636,
+      "grad_norm": 2.3660247775228154,
+      "learning_rate": 1.1174761929776094e-06,
+      "loss": 0.0909,
+      "step": 930
+    },
+    {
+      "epoch": 0.21159090909090908,
+      "grad_norm": 3.5750836513779767,
+      "learning_rate": 1.1172013012828887e-06,
+      "loss": 0.2178,
+      "step": 931
+    },
+    {
+      "epoch": 0.21181818181818182,
+      "grad_norm": 3.3553411849678136,
+      "learning_rate": 1.1169261586670976e-06,
+      "loss": 0.1,
+      "step": 932
+    },
+    {
+      "epoch": 0.21204545454545454,
+      "grad_norm": 2.461573990026709,
+      "learning_rate": 1.1166507652705021e-06,
+      "loss": 0.1204,
+      "step": 933
+    },
+    {
+      "epoch": 0.21227272727272728,
+      "grad_norm": 2.132948853011121,
+      "learning_rate": 1.1163751212334962e-06,
+      "loss": 0.1346,
+      "step": 934
+    },
+    {
+      "epoch": 0.2125,
+      "grad_norm": 5.0499737173683785,
+      "learning_rate": 1.116099226696601e-06,
+      "loss": 0.1486,
+      "step": 935
+    },
+    {
+      "epoch": 0.21272727272727274,
+      "grad_norm": 1.5105585328399602,
+      "learning_rate": 1.1158230818004656e-06,
+      "loss": 0.1098,
+      "step": 936
+    },
+    {
+      "epoch": 0.21295454545454545,
+      "grad_norm": 3.5358893745787188,
+      "learning_rate": 1.1155466866858677e-06,
+      "loss": 0.1392,
+      "step": 937
+    },
+    {
+      "epoch": 0.2131818181818182,
+      "grad_norm": 3.592774850320713,
+      "learning_rate": 1.1152700414937111e-06,
+      "loss": 0.1668,
+      "step": 938
+    },
+    {
+      "epoch": 0.2134090909090909,
+      "grad_norm": 2.3492875337615517,
+      "learning_rate": 1.114993146365028e-06,
+      "loss": 0.1373,
+      "step": 939
+    },
+    {
+      "epoch": 0.21363636363636362,
+      "grad_norm": 4.68217978312938,
+      "learning_rate": 1.1147160014409779e-06,
+      "loss": 0.187,
+      "step": 940
+    },
+    {
+      "epoch": 0.21386363636363637,
+      "grad_norm": 3.022198519387063,
+      "learning_rate": 1.1144386068628472e-06,
+      "loss": 0.1023,
+      "step": 941
+    },
+    {
+      "epoch": 0.21409090909090908,
+      "grad_norm": 3.38079062337303,
+      "learning_rate": 1.1141609627720501e-06,
+      "loss": 0.1048,
+      "step": 942
+    },
+    {
+      "epoch": 0.21431818181818182,
+      "grad_norm": 3.7223716539133034,
+      "learning_rate": 1.1138830693101277e-06,
+      "loss": 0.1141,
+      "step": 943
+    },
+    {
+      "epoch": 0.21454545454545454,
+      "grad_norm": 3.0069098849793154,
+      "learning_rate": 1.1136049266187481e-06,
+      "loss": 0.108,
+      "step": 944
+    },
+    {
+      "epoch": 0.21477272727272728,
+      "grad_norm": 5.2311974782528985,
+      "learning_rate": 1.113326534839707e-06,
+      "loss": 0.1526,
+      "step": 945
+    },
+    {
+      "epoch": 0.215,
+      "grad_norm": 3.119040125040404,
+      "learning_rate": 1.1130478941149268e-06,
+      "loss": 0.1452,
+      "step": 946
+    },
+    {
+      "epoch": 0.21522727272727274,
+      "grad_norm": 1.4890189795841464,
+      "learning_rate": 1.1127690045864561e-06,
+      "loss": 0.0759,
+      "step": 947
+    },
+    {
+      "epoch": 0.21545454545454545,
+      "grad_norm": 2.2018102674880646,
+      "learning_rate": 1.1124898663964718e-06,
+      "loss": 0.0753,
+      "step": 948
+    },
+    {
+      "epoch": 0.21568181818181817,
+      "grad_norm": 2.760261137195679,
+      "learning_rate": 1.1122104796872763e-06,
+      "loss": 0.1072,
+      "step": 949
+    },
+    {
+      "epoch": 0.2159090909090909,
+      "grad_norm": 3.3409885975170464,
+      "learning_rate": 1.1119308446012993e-06,
+      "loss": 0.1217,
+      "step": 950
+    },
+    {
+      "epoch": 0.21613636363636363,
+      "grad_norm": 2.716854908988171,
+      "learning_rate": 1.111650961281097e-06,
+      "loss": 0.1054,
+      "step": 951
+    },
+    {
+      "epoch": 0.21636363636363637,
+      "grad_norm": 3.493928291069267,
+      "learning_rate": 1.111370829869352e-06,
+      "loss": 0.1476,
+      "step": 952
+    },
+    {
+      "epoch": 0.21659090909090908,
+      "grad_norm": 2.434355429716211,
+      "learning_rate": 1.1110904505088738e-06,
+      "loss": 0.155,
+      "step": 953
+    },
+    {
+      "epoch": 0.21681818181818183,
+      "grad_norm": 2.150523888824578,
+      "learning_rate": 1.1108098233425977e-06,
+      "loss": 0.0834,
+      "step": 954
+    },
+    {
+      "epoch": 0.21704545454545454,
+      "grad_norm": 3.784235712659188,
+      "learning_rate": 1.1105289485135855e-06,
+      "loss": 0.2196,
+      "step": 955
+    },
+    {
+      "epoch": 0.21727272727272728,
+      "grad_norm": 2.031967467476721,
+      "learning_rate": 1.1102478261650258e-06,
+      "loss": 0.1159,
+      "step": 956
+    },
+    {
+      "epoch": 0.2175,
+      "grad_norm": 2.6180727716178893,
+      "learning_rate": 1.1099664564402327e-06,
+      "loss": 0.0697,
+      "step": 957
+    },
+    {
+      "epoch": 0.2177272727272727,
+      "grad_norm": 3.7527071781898464,
+      "learning_rate": 1.1096848394826467e-06,
+      "loss": 0.1202,
+      "step": 958
+    },
+    {
+      "epoch": 0.21795454545454546,
+      "grad_norm": 2.7842181321358255,
+      "learning_rate": 1.1094029754358343e-06,
+      "loss": 0.1357,
+      "step": 959
+    },
+    {
+      "epoch": 0.21818181818181817,
+      "grad_norm": 3.4205439007813663,
+      "learning_rate": 1.1091208644434882e-06,
+      "loss": 0.1041,
+      "step": 960
+    },
+    {
+      "epoch": 0.2184090909090909,
+      "grad_norm": 4.247970764538057,
+      "learning_rate": 1.1088385066494267e-06,
+      "loss": 0.244,
+      "step": 961
+    },
+    {
+      "epoch": 0.21863636363636363,
+      "grad_norm": 2.9884914959346935,
+      "learning_rate": 1.1085559021975937e-06,
+      "loss": 0.1512,
+      "step": 962
+    },
+    {
+      "epoch": 0.21886363636363637,
+      "grad_norm": 2.888953149227291,
+      "learning_rate": 1.1082730512320597e-06,
+      "loss": 0.0791,
+      "step": 963
+    },
+    {
+      "epoch": 0.2190909090909091,
+      "grad_norm": 2.6754049899904646,
+      "learning_rate": 1.1079899538970196e-06,
+      "loss": 0.2173,
+      "step": 964
+    },
+    {
+      "epoch": 0.21931818181818183,
+      "grad_norm": 4.135399295048608,
+      "learning_rate": 1.1077066103367954e-06,
+      "loss": 0.1597,
+      "step": 965
+    },
+    {
+      "epoch": 0.21954545454545454,
+      "grad_norm": 3.030186668045807,
+      "learning_rate": 1.1074230206958332e-06,
+      "loss": 0.1223,
+      "step": 966
+    },
+    {
+      "epoch": 0.2197727272727273,
+      "grad_norm": 4.654726687612204,
+      "learning_rate": 1.1071391851187052e-06,
+      "loss": 0.2229,
+      "step": 967
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.269763035774091,
+      "learning_rate": 1.1068551037501093e-06,
+      "loss": 0.2342,
+      "step": 968
+    },
+    {
+      "epoch": 0.22022727272727272,
+      "grad_norm": 2.3883820110800347,
+      "learning_rate": 1.1065707767348684e-06,
+      "loss": 0.2017,
+      "step": 969
+    },
+    {
+      "epoch": 0.22045454545454546,
+      "grad_norm": 5.324362128340279,
+      "learning_rate": 1.1062862042179302e-06,
+      "loss": 0.1029,
+      "step": 970
+    },
+    {
+      "epoch": 0.22068181818181817,
+      "grad_norm": 3.698309881204802,
+      "learning_rate": 1.1060013863443683e-06,
+      "loss": 0.1437,
+      "step": 971
+    },
+    {
+      "epoch": 0.22090909090909092,
+      "grad_norm": 2.8806368142725765,
+      "learning_rate": 1.1057163232593808e-06,
+      "loss": 0.1043,
+      "step": 972
+    },
+    {
+      "epoch": 0.22113636363636363,
+      "grad_norm": 2.503596957560977,
+      "learning_rate": 1.1054310151082913e-06,
+      "loss": 0.2415,
+      "step": 973
+    },
+    {
+      "epoch": 0.22136363636363637,
+      "grad_norm": 1.9210715547743957,
+      "learning_rate": 1.1051454620365475e-06,
+      "loss": 0.1009,
+      "step": 974
+    },
+    {
+      "epoch": 0.2215909090909091,
+      "grad_norm": 2.3220872576228984,
+      "learning_rate": 1.1048596641897233e-06,
+      "loss": 0.1196,
+      "step": 975
+    },
+    {
+      "epoch": 0.22181818181818183,
+      "grad_norm": 2.949017935093454,
+      "learning_rate": 1.104573621713516e-06,
+      "loss": 0.3316,
+      "step": 976
+    },
+    {
+      "epoch": 0.22204545454545455,
+      "grad_norm": 2.79464745258525,
+      "learning_rate": 1.1042873347537485e-06,
+      "loss": 0.083,
+      "step": 977
+    },
+    {
+      "epoch": 0.22227272727272726,
+      "grad_norm": 2.750644243831504,
+      "learning_rate": 1.1040008034563682e-06,
+      "loss": 0.0826,
+      "step": 978
+    },
+    {
+      "epoch": 0.2225,
+      "grad_norm": 2.572646414202906,
+      "learning_rate": 1.1037140279674468e-06,
+      "loss": 0.1138,
+      "step": 979
+    },
+    {
+      "epoch": 0.22272727272727272,
+      "grad_norm": 2.8280978926544553,
+      "learning_rate": 1.1034270084331803e-06,
+      "loss": 0.1364,
+      "step": 980
+    },
+    {
+      "epoch": 0.22295454545454546,
+      "grad_norm": 1.8900358810799298,
+      "learning_rate": 1.1031397449998896e-06,
+      "loss": 0.1814,
+      "step": 981
+    },
+    {
+      "epoch": 0.22318181818181818,
+      "grad_norm": 3.548172063137955,
+      "learning_rate": 1.10285223781402e-06,
+      "loss": 0.1136,
+      "step": 982
+    },
+    {
+      "epoch": 0.22340909090909092,
+      "grad_norm": 2.430967031163591,
+      "learning_rate": 1.1025644870221405e-06,
+      "loss": 0.1367,
+      "step": 983
+    },
+    {
+      "epoch": 0.22363636363636363,
+      "grad_norm": 4.976998296880154,
+      "learning_rate": 1.1022764927709447e-06,
+      "loss": 0.0902,
+      "step": 984
+    },
+    {
+      "epoch": 0.22386363636363638,
+      "grad_norm": 3.4098111165266407,
+      "learning_rate": 1.1019882552072502e-06,
+      "loss": 0.1562,
+      "step": 985
+    },
+    {
+      "epoch": 0.2240909090909091,
+      "grad_norm": 4.776350175294678,
+      "learning_rate": 1.101699774477999e-06,
+      "loss": 0.1678,
+      "step": 986
+    },
+    {
+      "epoch": 0.2243181818181818,
+      "grad_norm": 3.241579437364959,
+      "learning_rate": 1.1014110507302563e-06,
+      "loss": 0.0978,
+      "step": 987
+    },
+    {
+      "epoch": 0.22454545454545455,
+      "grad_norm": 2.1785201003945764,
+      "learning_rate": 1.1011220841112118e-06,
+      "loss": 0.0776,
+      "step": 988
+    },
+    {
+      "epoch": 0.22477272727272726,
+      "grad_norm": 2.8270550637915717,
+      "learning_rate": 1.1008328747681788e-06,
+      "loss": 0.1289,
+      "step": 989
+    },
+    {
+      "epoch": 0.225,
+      "grad_norm": 4.0570559584431445,
+      "learning_rate": 1.1005434228485945e-06,
+      "loss": 0.1378,
+      "step": 990
+    },
+    {
+      "epoch": 0.22522727272727272,
+      "grad_norm": 3.5179356538335242,
+      "learning_rate": 1.1002537285000196e-06,
+      "loss": 0.1251,
+      "step": 991
+    },
+    {
+      "epoch": 0.22545454545454546,
+      "grad_norm": 3.6509511027654824,
+      "learning_rate": 1.099963791870138e-06,
+      "loss": 0.1942,
+      "step": 992
+    },
+    {
+      "epoch": 0.22568181818181818,
+      "grad_norm": 3.05281277328344,
+      "learning_rate": 1.099673613106758e-06,
+      "loss": 0.1753,
+      "step": 993
+    },
+    {
+      "epoch": 0.22590909090909092,
+      "grad_norm": 3.347653752266796,
+      "learning_rate": 1.0993831923578107e-06,
+      "loss": 0.1685,
+      "step": 994
+    },
+    {
+      "epoch": 0.22613636363636364,
+      "grad_norm": 2.1493293376718876,
+      "learning_rate": 1.099092529771351e-06,
+      "loss": 0.1795,
+      "step": 995
+    },
+    {
+      "epoch": 0.22636363636363635,
+      "grad_norm": 3.570772278913998,
+      "learning_rate": 1.0988016254955565e-06,
+      "loss": 0.0947,
+      "step": 996
+    },
+    {
+      "epoch": 0.2265909090909091,
+      "grad_norm": 3.6802200327436787,
+      "learning_rate": 1.0985104796787285e-06,
+      "loss": 0.1257,
+      "step": 997
+    },
+    {
+      "epoch": 0.2268181818181818,
+      "grad_norm": 2.95563082955171,
+      "learning_rate": 1.0982190924692911e-06,
+      "loss": 0.1281,
+      "step": 998
+    },
+    {
+      "epoch": 0.22704545454545455,
+      "grad_norm": 2.5893040273448125,
+      "learning_rate": 1.0979274640157919e-06,
+      "loss": 0.1006,
+      "step": 999
+    },
+    {
+      "epoch": 0.22727272727272727,
+      "grad_norm": 2.5655108021263384,
+      "learning_rate": 1.097635594466901e-06,
+      "loss": 0.0953,
+      "step": 1000
+    },
+    {
+      "epoch": 0.2275,
+      "grad_norm": 2.7240844773411212,
+      "learning_rate": 1.0973434839714116e-06,
+      "loss": 0.1857,
+      "step": 1001
+    },
+    {
+      "epoch": 0.22772727272727272,
+      "grad_norm": 3.5866761555075755,
+      "learning_rate": 1.0970511326782398e-06,
+      "loss": 0.1065,
+      "step": 1002
+    },
+    {
+      "epoch": 0.22795454545454547,
+      "grad_norm": 2.251929251515994,
+      "learning_rate": 1.0967585407364246e-06,
+      "loss": 0.1693,
+      "step": 1003
+    },
+    {
+      "epoch": 0.22818181818181818,
+      "grad_norm": 2.725909396049706,
+      "learning_rate": 1.0964657082951274e-06,
+      "loss": 0.165,
+      "step": 1004
+    },
+    {
+      "epoch": 0.22840909090909092,
+      "grad_norm": 3.105840927643887,
+      "learning_rate": 1.0961726355036324e-06,
+      "loss": 0.1241,
+      "step": 1005
+    },
+    {
+      "epoch": 0.22863636363636364,
+      "grad_norm": 3.537168974510694,
+      "learning_rate": 1.0958793225113459e-06,
+      "loss": 0.2381,
+      "step": 1006
+    },
+    {
+      "epoch": 0.22886363636363635,
+      "grad_norm": 3.140521639553628,
+      "learning_rate": 1.0955857694677971e-06,
+      "loss": 0.1287,
+      "step": 1007
+    },
+    {
+      "epoch": 0.2290909090909091,
+      "grad_norm": 3.42767310942977,
+      "learning_rate": 1.0952919765226378e-06,
+      "loss": 0.155,
+      "step": 1008
+    },
+    {
+      "epoch": 0.2293181818181818,
+      "grad_norm": 0.9349201869899464,
+      "learning_rate": 1.0949979438256415e-06,
+      "loss": 0.0622,
+      "step": 1009
+    },
+    {
+      "epoch": 0.22954545454545455,
+      "grad_norm": 3.0231817541182413,
+      "learning_rate": 1.0947036715267039e-06,
+      "loss": 0.1032,
+      "step": 1010
+    },
+    {
+      "epoch": 0.22977272727272727,
+      "grad_norm": 4.852191624256205,
+      "learning_rate": 1.0944091597758438e-06,
+      "loss": 0.1539,
+      "step": 1011
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7867523869746207,
+      "learning_rate": 1.0941144087232008e-06,
+      "loss": 0.1967,
+      "step": 1012
+    },
+    {
+      "epoch": 0.23022727272727272,
+      "grad_norm": 3.556241020496331,
+      "learning_rate": 1.0938194185190374e-06,
+      "loss": 0.1137,
+      "step": 1013
+    },
+    {
+      "epoch": 0.23045454545454547,
+      "grad_norm": 2.277983573817439,
+      "learning_rate": 1.0935241893137376e-06,
+      "loss": 0.0826,
+      "step": 1014
+    },
+    {
+      "epoch": 0.23068181818181818,
+      "grad_norm": 2.70278221868962,
+      "learning_rate": 1.0932287212578075e-06,
+      "loss": 0.1061,
+      "step": 1015
+    },
+    {
+      "epoch": 0.2309090909090909,
+      "grad_norm": 3.261919882340595,
+      "learning_rate": 1.0929330145018747e-06,
+      "loss": 0.1012,
+      "step": 1016
+    },
+    {
+      "epoch": 0.23113636363636364,
+      "grad_norm": 3.052034847220137,
+      "learning_rate": 1.0926370691966883e-06,
+      "loss": 0.1472,
+      "step": 1017
+    },
+    {
+      "epoch": 0.23136363636363635,
+      "grad_norm": 3.109175963981429,
+      "learning_rate": 1.0923408854931202e-06,
+      "loss": 0.104,
+      "step": 1018
+    },
+    {
+      "epoch": 0.2315909090909091,
+      "grad_norm": 2.0619052744025312,
+      "learning_rate": 1.0920444635421622e-06,
+      "loss": 0.1396,
+      "step": 1019
+    },
+    {
+      "epoch": 0.2318181818181818,
+      "grad_norm": 2.727094544065828,
+      "learning_rate": 1.0917478034949285e-06,
+      "loss": 0.1002,
+      "step": 1020
+    },
+    {
+      "epoch": 0.23204545454545455,
+      "grad_norm": 3.9111476930566513,
+      "learning_rate": 1.0914509055026545e-06,
+      "loss": 0.1443,
+      "step": 1021
+    },
+    {
+      "epoch": 0.23227272727272727,
+      "grad_norm": 3.0191865826386786,
+      "learning_rate": 1.0911537697166967e-06,
+      "loss": 0.1242,
+      "step": 1022
+    },
+    {
+      "epoch": 0.2325,
+      "grad_norm": 2.5132795558840857,
+      "learning_rate": 1.0908563962885337e-06,
+      "loss": 0.1777,
+      "step": 1023
+    },
+    {
+      "epoch": 0.23272727272727273,
+      "grad_norm": 3.9179399018012795,
+      "learning_rate": 1.0905587853697638e-06,
+      "loss": 0.1475,
+      "step": 1024
+    },
+    {
+      "epoch": 0.23295454545454544,
+      "grad_norm": 3.433367564686768,
+      "learning_rate": 1.0902609371121076e-06,
+      "loss": 0.1414,
+      "step": 1025
+    },
+    {
+      "epoch": 0.23318181818181818,
+      "grad_norm": 3.008375968991309,
+      "learning_rate": 1.089962851667406e-06,
+      "loss": 0.1888,
+      "step": 1026
+    },
+    {
+      "epoch": 0.2334090909090909,
+      "grad_norm": 3.0946776652749994,
+      "learning_rate": 1.089664529187621e-06,
+      "loss": 0.1213,
+      "step": 1027
+    },
+    {
+      "epoch": 0.23363636363636364,
+      "grad_norm": 2.245229847872642,
+      "learning_rate": 1.0893659698248358e-06,
+      "loss": 0.0845,
+      "step": 1028
+    },
+    {
+      "epoch": 0.23386363636363636,
+      "grad_norm": 2.6283980706368437,
+      "learning_rate": 1.0890671737312538e-06,
+      "loss": 0.1256,
+      "step": 1029
+    },
+    {
+      "epoch": 0.2340909090909091,
+      "grad_norm": 1.4134599455391224,
+      "learning_rate": 1.0887681410591994e-06,
+      "loss": 0.12,
+      "step": 1030
+    },
+    {
+      "epoch": 0.23431818181818181,
+      "grad_norm": 2.8877976074820273,
+      "learning_rate": 1.0884688719611176e-06,
+      "loss": 0.0818,
+      "step": 1031
+    },
+    {
+      "epoch": 0.23454545454545456,
+      "grad_norm": 2.637884980076348,
+      "learning_rate": 1.0881693665895737e-06,
+      "loss": 0.1219,
+      "step": 1032
+    },
+    {
+      "epoch": 0.23477272727272727,
+      "grad_norm": 2.692038701866701,
+      "learning_rate": 1.0878696250972536e-06,
+      "loss": 0.1007,
+      "step": 1033
+    },
+    {
+      "epoch": 0.235,
+      "grad_norm": 2.1786346420797047,
+      "learning_rate": 1.087569647636964e-06,
+      "loss": 0.119,
+      "step": 1034
+    },
+    {
+      "epoch": 0.23522727272727273,
+      "grad_norm": 2.068810991727987,
+      "learning_rate": 1.0872694343616312e-06,
+      "loss": 0.0845,
+      "step": 1035
+    },
+    {
+      "epoch": 0.23545454545454544,
+      "grad_norm": 4.919658840647895,
+      "learning_rate": 1.0869689854243019e-06,
+      "loss": 0.3173,
+      "step": 1036
+    },
+    {
+      "epoch": 0.2356818181818182,
+      "grad_norm": 1.3258282185600323,
+      "learning_rate": 1.0866683009781432e-06,
+      "loss": 0.0951,
+      "step": 1037
+    },
+    {
+      "epoch": 0.2359090909090909,
+      "grad_norm": 3.4736535085661804,
+      "learning_rate": 1.0863673811764419e-06,
+      "loss": 0.144,
+      "step": 1038
+    },
+    {
+      "epoch": 0.23613636363636364,
+      "grad_norm": 2.009542419536427,
+      "learning_rate": 1.0860662261726054e-06,
+      "loss": 0.056,
+      "step": 1039
+    },
+    {
+      "epoch": 0.23636363636363636,
+      "grad_norm": 0.8819106693521577,
+      "learning_rate": 1.0857648361201603e-06,
+      "loss": 0.0799,
+      "step": 1040
+    },
+    {
+      "epoch": 0.2365909090909091,
+      "grad_norm": 4.2639265688479275,
+      "learning_rate": 1.085463211172753e-06,
+      "loss": 0.1513,
+      "step": 1041
+    },
+    {
+      "epoch": 0.23681818181818182,
+      "grad_norm": 3.7006214346382342,
+      "learning_rate": 1.0851613514841508e-06,
+      "loss": 0.1675,
+      "step": 1042
+    },
+    {
+      "epoch": 0.23704545454545456,
+      "grad_norm": 2.550279844867115,
+      "learning_rate": 1.0848592572082391e-06,
+      "loss": 0.1076,
+      "step": 1043
+    },
+    {
+      "epoch": 0.23727272727272727,
+      "grad_norm": 2.6152925625499885,
+      "learning_rate": 1.084556928499024e-06,
+      "loss": 0.0601,
+      "step": 1044
+    },
+    {
+      "epoch": 0.2375,
+      "grad_norm": 2.799825533301437,
+      "learning_rate": 1.0842543655106305e-06,
+      "loss": 0.147,
+      "step": 1045
+    },
+    {
+      "epoch": 0.23772727272727273,
+      "grad_norm": 2.8297783412869606,
+      "learning_rate": 1.0839515683973035e-06,
+      "loss": 0.1125,
+      "step": 1046
+    },
+    {
+      "epoch": 0.23795454545454545,
+      "grad_norm": 2.3193405292936737,
+      "learning_rate": 1.083648537313407e-06,
+      "loss": 0.1045,
+      "step": 1047
+    },
+    {
+      "epoch": 0.2381818181818182,
+      "grad_norm": 2.427682106599837,
+      "learning_rate": 1.083345272413424e-06,
+      "loss": 0.1207,
+      "step": 1048
+    },
+    {
+      "epoch": 0.2384090909090909,
+      "grad_norm": 2.3886543968915546,
+      "learning_rate": 1.0830417738519575e-06,
+      "loss": 0.1634,
+      "step": 1049
+    },
+    {
+      "epoch": 0.23863636363636365,
+      "grad_norm": 2.6501786061385406,
+      "learning_rate": 1.0827380417837287e-06,
+      "loss": 0.1359,
+      "step": 1050
+    },
+    {
+      "epoch": 0.23886363636363636,
+      "grad_norm": 3.287063986605485,
+      "learning_rate": 1.0824340763635785e-06,
+      "loss": 0.1068,
+      "step": 1051
+    },
+    {
+      "epoch": 0.2390909090909091,
+      "grad_norm": 2.6012099329720346,
+      "learning_rate": 1.0821298777464665e-06,
+      "loss": 0.1117,
+      "step": 1052
+    },
+    {
+      "epoch": 0.23931818181818182,
+      "grad_norm": 1.6562418136706265,
+      "learning_rate": 1.081825446087471e-06,
+      "loss": 0.0656,
+      "step": 1053
+    },
+    {
+      "epoch": 0.23954545454545453,
+      "grad_norm": 3.1133210593465614,
+      "learning_rate": 1.0815207815417894e-06,
+      "loss": 0.1145,
+      "step": 1054
+    },
+    {
+      "epoch": 0.23977272727272728,
+      "grad_norm": 3.316270914582336,
+      "learning_rate": 1.081215884264738e-06,
+      "loss": 0.093,
+      "step": 1055
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.954934076565024,
+      "learning_rate": 1.0809107544117511e-06,
+      "loss": 0.2215,
+      "step": 1056
+    },
+    {
+      "epoch": 0.24022727272727273,
+      "grad_norm": 2.827985987143346,
+      "learning_rate": 1.0806053921383823e-06,
+      "loss": 0.1091,
+      "step": 1057
+    },
+    {
+      "epoch": 0.24045454545454545,
+      "grad_norm": 3.753631864332377,
+      "learning_rate": 1.0802997976003031e-06,
+      "loss": 0.1347,
+      "step": 1058
+    },
+    {
+      "epoch": 0.2406818181818182,
+      "grad_norm": 2.3210571459620355,
+      "learning_rate": 1.0799939709533036e-06,
+      "loss": 0.1213,
+      "step": 1059
+    },
+    {
+      "epoch": 0.2409090909090909,
+      "grad_norm": 2.9930101419937474,
+      "learning_rate": 1.0796879123532924e-06,
+      "loss": 0.2363,
+      "step": 1060
+    },
+    {
+      "epoch": 0.24113636363636365,
+      "grad_norm": 2.824235208418317,
+      "learning_rate": 1.0793816219562963e-06,
+      "loss": 0.1123,
+      "step": 1061
+    },
+    {
+      "epoch": 0.24136363636363636,
+      "grad_norm": 2.233166466663316,
+      "learning_rate": 1.0790750999184598e-06,
+      "loss": 0.1771,
+      "step": 1062
+    },
+    {
+      "epoch": 0.24159090909090908,
+      "grad_norm": 4.357506158883664,
+      "learning_rate": 1.0787683463960462e-06,
+      "loss": 0.1122,
+      "step": 1063
+    },
+    {
+      "epoch": 0.24181818181818182,
+      "grad_norm": 2.41811670816456,
+      "learning_rate": 1.0784613615454365e-06,
+      "loss": 0.1344,
+      "step": 1064
+    },
+    {
+      "epoch": 0.24204545454545454,
+      "grad_norm": 2.044594682206005,
+      "learning_rate": 1.0781541455231294e-06,
+      "loss": 0.1237,
+      "step": 1065
+    },
+    {
+      "epoch": 0.24227272727272728,
+      "grad_norm": 5.074094808707105,
+      "learning_rate": 1.077846698485742e-06,
+      "loss": 0.1483,
+      "step": 1066
+    },
+    {
+      "epoch": 0.2425,
+      "grad_norm": 3.431611332095325,
+      "learning_rate": 1.0775390205900084e-06,
+      "loss": 0.1214,
+      "step": 1067
+    },
+    {
+      "epoch": 0.24272727272727274,
+      "grad_norm": 1.7274818652585815,
+      "learning_rate": 1.0772311119927808e-06,
+      "loss": 0.092,
+      "step": 1068
+    },
+    {
+      "epoch": 0.24295454545454545,
+      "grad_norm": 2.518179730346009,
+      "learning_rate": 1.0769229728510298e-06,
+      "loss": 0.1301,
+      "step": 1069
+    },
+    {
+      "epoch": 0.2431818181818182,
+      "grad_norm": 2.679448169829352,
+      "learning_rate": 1.0766146033218417e-06,
+      "loss": 0.2318,
+      "step": 1070
+    },
+    {
+      "epoch": 0.2434090909090909,
+      "grad_norm": 3.8561999460181426,
+      "learning_rate": 1.076306003562422e-06,
+      "loss": 0.2135,
+      "step": 1071
+    },
+    {
+      "epoch": 0.24363636363636362,
+      "grad_norm": 2.6388004592956404,
+      "learning_rate": 1.0759971737300928e-06,
+      "loss": 0.1214,
+      "step": 1072
+    },
+    {
+      "epoch": 0.24386363636363637,
+      "grad_norm": 2.7418257135270263,
+      "learning_rate": 1.0756881139822934e-06,
+      "loss": 0.1224,
+      "step": 1073
+    },
+    {
+      "epoch": 0.24409090909090908,
+      "grad_norm": 2.1496782917069317,
+      "learning_rate": 1.0753788244765805e-06,
+      "loss": 0.1232,
+      "step": 1074
+    },
+    {
+      "epoch": 0.24431818181818182,
+      "grad_norm": 2.288012649033509,
+      "learning_rate": 1.0750693053706282e-06,
+      "loss": 0.102,
+      "step": 1075
+    },
+    {
+      "epoch": 0.24454545454545454,
+      "grad_norm": 5.8589423617292935,
+      "learning_rate": 1.0747595568222268e-06,
+      "loss": 0.3118,
+      "step": 1076
+    },
+    {
+      "epoch": 0.24477272727272728,
+      "grad_norm": 4.2972298670743285,
+      "learning_rate": 1.0744495789892848e-06,
+      "loss": 0.1616,
+      "step": 1077
+    },
+    {
+      "epoch": 0.245,
+      "grad_norm": 2.2628107145482947,
+      "learning_rate": 1.0741393720298263e-06,
+      "loss": 0.1184,
+      "step": 1078
+    },
+    {
+      "epoch": 0.24522727272727274,
+      "grad_norm": 3.3502064179889164,
+      "learning_rate": 1.073828936101993e-06,
+      "loss": 0.2154,
+      "step": 1079
+    },
+    {
+      "epoch": 0.24545454545454545,
+      "grad_norm": 3.641620959204165,
+      "learning_rate": 1.0735182713640436e-06,
+      "loss": 0.1539,
+      "step": 1080
+    },
+    {
+      "epoch": 0.2456818181818182,
+      "grad_norm": 2.1141997809327835,
+      "learning_rate": 1.0732073779743523e-06,
+      "loss": 0.1258,
+      "step": 1081
+    },
+    {
+      "epoch": 0.2459090909090909,
+      "grad_norm": 2.4105081014514513,
+      "learning_rate": 1.0728962560914108e-06,
+      "loss": 0.1359,
+      "step": 1082
+    },
+    {
+      "epoch": 0.24613636363636363,
+      "grad_norm": 3.395747201090836,
+      "learning_rate": 1.0725849058738274e-06,
+      "loss": 0.09,
+      "step": 1083
+    },
+    {
+      "epoch": 0.24636363636363637,
+      "grad_norm": 2.928611083959143,
+      "learning_rate": 1.0722733274803261e-06,
+      "loss": 0.1772,
+      "step": 1084
+    },
+    {
+      "epoch": 0.24659090909090908,
+      "grad_norm": 2.2715805940470846,
+      "learning_rate": 1.0719615210697476e-06,
+      "loss": 0.0991,
+      "step": 1085
+    },
+    {
+      "epoch": 0.24681818181818183,
+      "grad_norm": 2.5672432768767193,
+      "learning_rate": 1.0716494868010488e-06,
+      "loss": 0.1924,
+      "step": 1086
+    },
+    {
+      "epoch": 0.24704545454545454,
+      "grad_norm": 2.756203612952947,
+      "learning_rate": 1.071337224833303e-06,
+      "loss": 0.1161,
+      "step": 1087
+    },
+    {
+      "epoch": 0.24727272727272728,
+      "grad_norm": 3.1068530882871004,
+      "learning_rate": 1.0710247353256988e-06,
+      "loss": 0.0948,
+      "step": 1088
+    },
+    {
+      "epoch": 0.2475,
+      "grad_norm": 3.33292913549004,
+      "learning_rate": 1.0707120184375422e-06,
+      "loss": 0.1528,
+      "step": 1089
+    },
+    {
+      "epoch": 0.24772727272727274,
+      "grad_norm": 2.8953930014200804,
+      "learning_rate": 1.0703990743282534e-06,
+      "loss": 0.1091,
+      "step": 1090
+    },
+    {
+      "epoch": 0.24795454545454546,
+      "grad_norm": 2.029463576531459,
+      "learning_rate": 1.07008590315737e-06,
+      "loss": 0.1102,
+      "step": 1091
+    },
+    {
+      "epoch": 0.24818181818181817,
+      "grad_norm": 3.1502540087687327,
+      "learning_rate": 1.069772505084544e-06,
+      "loss": 0.1645,
+      "step": 1092
+    },
+    {
+      "epoch": 0.2484090909090909,
+      "grad_norm": 1.2895494700644135,
+      "learning_rate": 1.0694588802695443e-06,
+      "loss": 0.088,
+      "step": 1093
+    },
+    {
+      "epoch": 0.24863636363636363,
+      "grad_norm": 4.15481462918436,
+      "learning_rate": 1.0691450288722545e-06,
+      "loss": 0.1617,
+      "step": 1094
+    },
+    {
+      "epoch": 0.24886363636363637,
+      "grad_norm": 2.5418302438293576,
+      "learning_rate": 1.0688309510526742e-06,
+      "loss": 0.1033,
+      "step": 1095
+    },
+    {
+      "epoch": 0.24909090909090909,
+      "grad_norm": 2.964820366183925,
+      "learning_rate": 1.0685166469709181e-06,
+      "loss": 0.1795,
+      "step": 1096
+    },
+    {
+      "epoch": 0.24931818181818183,
+      "grad_norm": 2.8390721115209026,
+      "learning_rate": 1.0682021167872166e-06,
+      "loss": 0.0827,
+      "step": 1097
+    },
+    {
+      "epoch": 0.24954545454545454,
+      "grad_norm": 3.6250007396730357,
+      "learning_rate": 1.0678873606619152e-06,
+      "loss": 0.2644,
+      "step": 1098
+    },
+    {
+      "epoch": 0.24977272727272729,
+      "grad_norm": 2.661399686454449,
+      "learning_rate": 1.0675723787554743e-06,
+      "loss": 0.1153,
+      "step": 1099
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.5815989364802325,
+      "learning_rate": 1.0672571712284697e-06,
+      "loss": 0.0693,
+      "step": 1100
+    },
+    {
+      "epoch": 0.25022727272727274,
+      "grad_norm": 1.4929977048751337,
+      "learning_rate": 1.0669417382415923e-06,
+      "loss": 0.1383,
+      "step": 1101
+    },
+    {
+      "epoch": 0.25045454545454543,
+      "grad_norm": 3.8669518807164605,
+      "learning_rate": 1.0666260799556477e-06,
+      "loss": 0.262,
+      "step": 1102
+    },
+    {
+      "epoch": 0.2506818181818182,
+      "grad_norm": 5.2870780991855,
+      "learning_rate": 1.0663101965315568e-06,
+      "loss": 0.3254,
+      "step": 1103
+    },
+    {
+      "epoch": 0.2509090909090909,
+      "grad_norm": 3.1063174113790346,
+      "learning_rate": 1.0659940881303545e-06,
+      "loss": 0.1545,
+      "step": 1104
+    },
+    {
+      "epoch": 0.25113636363636366,
+      "grad_norm": 1.8967737459743916,
+      "learning_rate": 1.065677754913191e-06,
+      "loss": 0.1603,
+      "step": 1105
+    },
+    {
+      "epoch": 0.25136363636363634,
+      "grad_norm": 2.9740829264558633,
+      "learning_rate": 1.0653611970413311e-06,
+      "loss": 0.1413,
+      "step": 1106
+    },
+    {
+      "epoch": 0.2515909090909091,
+      "grad_norm": 1.669824887342398,
+      "learning_rate": 1.0650444146761537e-06,
+      "loss": 0.0916,
+      "step": 1107
+    },
+    {
+      "epoch": 0.25181818181818183,
+      "grad_norm": 2.0455315785068042,
+      "learning_rate": 1.0647274079791529e-06,
+      "loss": 0.1526,
+      "step": 1108
+    },
+    {
+      "epoch": 0.2520454545454546,
+      "grad_norm": 3.386333353720849,
+      "learning_rate": 1.0644101771119361e-06,
+      "loss": 0.2019,
+      "step": 1109
+    },
+    {
+      "epoch": 0.25227272727272726,
+      "grad_norm": 3.3762245220826204,
+      "learning_rate": 1.0640927222362258e-06,
+      "loss": 0.1498,
+      "step": 1110
+    },
+    {
+      "epoch": 0.2525,
+      "grad_norm": 2.8108878122517544,
+      "learning_rate": 1.0637750435138586e-06,
+      "loss": 0.1622,
+      "step": 1111
+    },
+    {
+      "epoch": 0.25272727272727274,
+      "grad_norm": 3.5900331982258717,
+      "learning_rate": 1.063457141106785e-06,
+      "loss": 0.1119,
+      "step": 1112
+    },
+    {
+      "epoch": 0.25295454545454543,
+      "grad_norm": 3.2926556780336793,
+      "learning_rate": 1.0631390151770693e-06,
+      "loss": 0.0893,
+      "step": 1113
+    },
+    {
+      "epoch": 0.2531818181818182,
+      "grad_norm": 4.104980650307259,
+      "learning_rate": 1.0628206658868906e-06,
+      "loss": 0.1167,
+      "step": 1114
+    },
+    {
+      "epoch": 0.2534090909090909,
+      "grad_norm": 2.6506975549311615,
+      "learning_rate": 1.0625020933985411e-06,
+      "loss": 0.1281,
+      "step": 1115
+    },
+    {
+      "epoch": 0.25363636363636366,
+      "grad_norm": 2.8538229891355735,
+      "learning_rate": 1.0621832978744267e-06,
+      "loss": 0.1495,
+      "step": 1116
+    },
+    {
+      "epoch": 0.25386363636363635,
+      "grad_norm": 2.163785180996921,
+      "learning_rate": 1.0618642794770679e-06,
+      "loss": 0.1547,
+      "step": 1117
+    },
+    {
+      "epoch": 0.2540909090909091,
+      "grad_norm": 3.210802171194901,
+      "learning_rate": 1.0615450383690978e-06,
+      "loss": 0.1307,
+      "step": 1118
+    },
+    {
+      "epoch": 0.25431818181818183,
+      "grad_norm": 4.242208233481095,
+      "learning_rate": 1.0612255747132637e-06,
+      "loss": 0.2379,
+      "step": 1119
+    },
+    {
+      "epoch": 0.2545454545454545,
+      "grad_norm": 2.7050650275559405,
+      "learning_rate": 1.0609058886724258e-06,
+      "loss": 0.1195,
+      "step": 1120
+    },
+    {
+      "epoch": 0.25477272727272726,
+      "grad_norm": 2.7250356109494938,
+      "learning_rate": 1.0605859804095587e-06,
+      "loss": 0.1934,
+      "step": 1121
+    },
+    {
+      "epoch": 0.255,
+      "grad_norm": 3.5342436452027264,
+      "learning_rate": 1.0602658500877486e-06,
+      "loss": 0.141,
+      "step": 1122
+    },
+    {
+      "epoch": 0.25522727272727275,
+      "grad_norm": 3.6683262261308953,
+      "learning_rate": 1.0599454978701965e-06,
+      "loss": 0.1802,
+      "step": 1123
+    },
+    {
+      "epoch": 0.25545454545454543,
+      "grad_norm": 2.98625850934089,
+      "learning_rate": 1.0596249239202158e-06,
+      "loss": 0.1306,
+      "step": 1124
+    },
+    {
+      "epoch": 0.2556818181818182,
+      "grad_norm": 2.6822532320193972,
+      "learning_rate": 1.0593041284012333e-06,
+      "loss": 0.1484,
+      "step": 1125
+    },
+    {
+      "epoch": 0.2559090909090909,
+      "grad_norm": 3.2021923715273166,
+      "learning_rate": 1.058983111476788e-06,
+      "loss": 0.1541,
+      "step": 1126
+    },
+    {
+      "epoch": 0.25613636363636366,
+      "grad_norm": 2.33317890243091,
+      "learning_rate": 1.0586618733105327e-06,
+      "loss": 0.1073,
+      "step": 1127
+    },
+    {
+      "epoch": 0.25636363636363635,
+      "grad_norm": 2.9101747724044436,
+      "learning_rate": 1.0583404140662322e-06,
+      "loss": 0.1221,
+      "step": 1128
+    },
+    {
+      "epoch": 0.2565909090909091,
+      "grad_norm": 3.0665207058498067,
+      "learning_rate": 1.0580187339077647e-06,
+      "loss": 0.1355,
+      "step": 1129
+    },
+    {
+      "epoch": 0.25681818181818183,
+      "grad_norm": 2.124997941407588,
+      "learning_rate": 1.0576968329991205e-06,
+      "loss": 0.2059,
+      "step": 1130
+    },
+    {
+      "epoch": 0.2570454545454545,
+      "grad_norm": 3.869005710940759,
+      "learning_rate": 1.0573747115044025e-06,
+      "loss": 0.1363,
+      "step": 1131
+    },
+    {
+      "epoch": 0.25727272727272726,
+      "grad_norm": 3.039484302640726,
+      "learning_rate": 1.0570523695878264e-06,
+      "loss": 0.1033,
+      "step": 1132
+    },
+    {
+      "epoch": 0.2575,
+      "grad_norm": 4.080946601123749,
+      "learning_rate": 1.05672980741372e-06,
+      "loss": 0.1629,
+      "step": 1133
+    },
+    {
+      "epoch": 0.25772727272727275,
+      "grad_norm": 3.633691899243643,
+      "learning_rate": 1.0564070251465233e-06,
+      "loss": 0.14,
+      "step": 1134
+    },
+    {
+      "epoch": 0.25795454545454544,
+      "grad_norm": 3.306099597511346,
+      "learning_rate": 1.0560840229507888e-06,
+      "loss": 0.1779,
+      "step": 1135
+    },
+    {
+      "epoch": 0.2581818181818182,
+      "grad_norm": 1.685281238911056,
+      "learning_rate": 1.055760800991181e-06,
+      "loss": 0.0913,
+      "step": 1136
+    },
+    {
+      "epoch": 0.2584090909090909,
+      "grad_norm": 1.9736004018296096,
+      "learning_rate": 1.0554373594324762e-06,
+      "loss": 0.1313,
+      "step": 1137
+    },
+    {
+      "epoch": 0.2586363636363636,
+      "grad_norm": 3.3359483851359157,
+      "learning_rate": 1.0551136984395627e-06,
+      "loss": 0.1077,
+      "step": 1138
+    },
+    {
+      "epoch": 0.25886363636363635,
+      "grad_norm": 4.2823319457048825,
+      "learning_rate": 1.0547898181774413e-06,
+      "loss": 0.2117,
+      "step": 1139
+    },
+    {
+      "epoch": 0.2590909090909091,
+      "grad_norm": 4.057011258065592,
+      "learning_rate": 1.0544657188112238e-06,
+      "loss": 0.1678,
+      "step": 1140
+    },
+    {
+      "epoch": 0.25931818181818184,
+      "grad_norm": 2.9161454690239763,
+      "learning_rate": 1.0541414005061334e-06,
+      "loss": 0.1531,
+      "step": 1141
+    },
+    {
+      "epoch": 0.2595454545454545,
+      "grad_norm": 2.945798692707358,
+      "learning_rate": 1.0538168634275067e-06,
+      "loss": 0.1023,
+      "step": 1142
+    },
+    {
+      "epoch": 0.25977272727272727,
+      "grad_norm": 2.604940874062389,
+      "learning_rate": 1.0534921077407898e-06,
+      "loss": 0.0816,
+      "step": 1143
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8118400954961054,
+      "learning_rate": 1.0531671336115409e-06,
+      "loss": 0.1078,
+      "step": 1144
+    },
+    {
+      "epoch": 0.26022727272727275,
+      "grad_norm": 3.1100280462960534,
+      "learning_rate": 1.0528419412054306e-06,
+      "loss": 0.2092,
+      "step": 1145
+    },
+    {
+      "epoch": 0.26045454545454544,
+      "grad_norm": 4.269359688990448,
+      "learning_rate": 1.052516530688239e-06,
+      "loss": 0.3091,
+      "step": 1146
+    },
+    {
+      "epoch": 0.2606818181818182,
+      "grad_norm": 2.7309105767041393,
+      "learning_rate": 1.0521909022258589e-06,
+      "loss": 0.1034,
+      "step": 1147
+    },
+    {
+      "epoch": 0.2609090909090909,
+      "grad_norm": 3.8310450019473166,
+      "learning_rate": 1.0518650559842934e-06,
+      "loss": 0.2042,
+      "step": 1148
+    },
+    {
+      "epoch": 0.2611363636363636,
+      "grad_norm": 3.1639987483714087,
+      "learning_rate": 1.0515389921296568e-06,
+      "loss": 0.1928,
+      "step": 1149
+    },
+    {
+      "epoch": 0.26136363636363635,
+      "grad_norm": 1.7882952967245254,
+      "learning_rate": 1.0512127108281741e-06,
+      "loss": 0.0938,
+      "step": 1150
+    },
+    {
+      "epoch": 0.2615909090909091,
+      "grad_norm": 3.105111724025716,
+      "learning_rate": 1.0508862122461822e-06,
+      "loss": 0.135,
+      "step": 1151
+    },
+    {
+      "epoch": 0.26181818181818184,
+      "grad_norm": 2.2471438892190974,
+      "learning_rate": 1.050559496550127e-06,
+      "loss": 0.0848,
+      "step": 1152
+    },
+    {
+      "epoch": 0.2620454545454545,
+      "grad_norm": 2.3991541261216724,
+      "learning_rate": 1.0502325639065669e-06,
+      "loss": 0.1604,
+      "step": 1153
+    },
+    {
+      "epoch": 0.26227272727272727,
+      "grad_norm": 2.7378067048142873,
+      "learning_rate": 1.0499054144821695e-06,
+      "loss": 0.1592,
+      "step": 1154
+    },
+    {
+      "epoch": 0.2625,
+      "grad_norm": 2.642957172409073,
+      "learning_rate": 1.0495780484437138e-06,
+      "loss": 0.1091,
+      "step": 1155
+    },
+    {
+      "epoch": 0.26272727272727275,
+      "grad_norm": 2.7948860319327,
+      "learning_rate": 1.0492504659580888e-06,
+      "loss": 0.1653,
+      "step": 1156
+    },
+    {
+      "epoch": 0.26295454545454544,
+      "grad_norm": 2.63170473383065,
+      "learning_rate": 1.0489226671922938e-06,
+      "loss": 0.0926,
+      "step": 1157
+    },
+    {
+      "epoch": 0.2631818181818182,
+      "grad_norm": 2.745750022622333,
+      "learning_rate": 1.048594652313439e-06,
+      "loss": 0.1736,
+      "step": 1158
+    },
+    {
+      "epoch": 0.2634090909090909,
+      "grad_norm": 4.228016918013105,
+      "learning_rate": 1.0482664214887436e-06,
+      "loss": 0.1263,
+      "step": 1159
+    },
+    {
+      "epoch": 0.2636363636363636,
+      "grad_norm": 4.102973659161791,
+      "learning_rate": 1.047937974885538e-06,
+      "loss": 0.141,
+      "step": 1160
+    },
+    {
+      "epoch": 0.26386363636363636,
+      "grad_norm": 3.2162838014257975,
+      "learning_rate": 1.0476093126712622e-06,
+      "loss": 0.2158,
+      "step": 1161
+    },
+    {
+      "epoch": 0.2640909090909091,
+      "grad_norm": 2.0563238269114983,
+      "learning_rate": 1.0472804350134661e-06,
+      "loss": 0.1596,
+      "step": 1162
+    },
+    {
+      "epoch": 0.26431818181818184,
+      "grad_norm": 4.758175022570386,
+      "learning_rate": 1.0469513420798091e-06,
+      "loss": 0.2108,
+      "step": 1163
+    },
+    {
+      "epoch": 0.26454545454545453,
+      "grad_norm": 2.8218832948849353,
+      "learning_rate": 1.046622034038061e-06,
+      "loss": 0.1011,
+      "step": 1164
+    },
+    {
+      "epoch": 0.26477272727272727,
+      "grad_norm": 2.050382839878734,
+      "learning_rate": 1.046292511056101e-06,
+      "loss": 0.1177,
+      "step": 1165
+    },
+    {
+      "epoch": 0.265,
+      "grad_norm": 3.95352727779858,
+      "learning_rate": 1.0459627733019172e-06,
+      "loss": 0.1216,
+      "step": 1166
+    },
+    {
+      "epoch": 0.2652272727272727,
+      "grad_norm": 3.106023531965552,
+      "learning_rate": 1.0456328209436085e-06,
+      "loss": 0.1169,
+      "step": 1167
+    },
+    {
+      "epoch": 0.26545454545454544,
+      "grad_norm": 2.3971440428833897,
+      "learning_rate": 1.0453026541493822e-06,
+      "loss": 0.0812,
+      "step": 1168
+    },
+    {
+      "epoch": 0.2656818181818182,
+      "grad_norm": 4.445261593318206,
+      "learning_rate": 1.044972273087555e-06,
+      "loss": 0.117,
+      "step": 1169
+    },
+    {
+      "epoch": 0.26590909090909093,
+      "grad_norm": 2.960038633980034,
+      "learning_rate": 1.0446416779265535e-06,
+      "loss": 0.2082,
+      "step": 1170
+    },
+    {
+      "epoch": 0.2661363636363636,
+      "grad_norm": 2.9876245338493104,
+      "learning_rate": 1.0443108688349127e-06,
+      "loss": 0.2329,
+      "step": 1171
+    },
+    {
+      "epoch": 0.26636363636363636,
+      "grad_norm": 2.41843081207234,
+      "learning_rate": 1.0439798459812772e-06,
+      "loss": 0.1334,
+      "step": 1172
+    },
+    {
+      "epoch": 0.2665909090909091,
+      "grad_norm": 2.751724302124061,
+      "learning_rate": 1.0436486095344e-06,
+      "loss": 0.1366,
+      "step": 1173
+    },
+    {
+      "epoch": 0.26681818181818184,
+      "grad_norm": 2.0652934793986852,
+      "learning_rate": 1.0433171596631433e-06,
+      "loss": 0.1237,
+      "step": 1174
+    },
+    {
+      "epoch": 0.26704545454545453,
+      "grad_norm": 3.1205151452365856,
+      "learning_rate": 1.0429854965364786e-06,
+      "loss": 0.1572,
+      "step": 1175
+    },
+    {
+      "epoch": 0.2672727272727273,
+      "grad_norm": 3.5186249635112845,
+      "learning_rate": 1.0426536203234851e-06,
+      "loss": 0.1066,
+      "step": 1176
+    },
+    {
+      "epoch": 0.2675,
+      "grad_norm": 2.0555585829778447,
+      "learning_rate": 1.0423215311933516e-06,
+      "loss": 0.1101,
+      "step": 1177
+    },
+    {
+      "epoch": 0.2677272727272727,
+      "grad_norm": 1.7537823014344742,
+      "learning_rate": 1.0419892293153749e-06,
+      "loss": 0.1203,
+      "step": 1178
+    },
+    {
+      "epoch": 0.26795454545454545,
+      "grad_norm": 3.1153704839717915,
+      "learning_rate": 1.04165671485896e-06,
+      "loss": 0.1373,
+      "step": 1179
+    },
+    {
+      "epoch": 0.2681818181818182,
+      "grad_norm": 3.688256064349858,
+      "learning_rate": 1.0413239879936213e-06,
+      "loss": 0.247,
+      "step": 1180
+    },
+    {
+      "epoch": 0.26840909090909093,
+      "grad_norm": 2.481792581319086,
+      "learning_rate": 1.0409910488889801e-06,
+      "loss": 0.2242,
+      "step": 1181
+    },
+    {
+      "epoch": 0.2686363636363636,
+      "grad_norm": 2.223525336117619,
+      "learning_rate": 1.0406578977147672e-06,
+      "loss": 0.1748,
+      "step": 1182
+    },
+    {
+      "epoch": 0.26886363636363636,
+      "grad_norm": 3.451573283411608,
+      "learning_rate": 1.0403245346408205e-06,
+      "loss": 0.1305,
+      "step": 1183
+    },
+    {
+      "epoch": 0.2690909090909091,
+      "grad_norm": 2.103540664535555,
+      "learning_rate": 1.0399909598370866e-06,
+      "loss": 0.1518,
+      "step": 1184
+    },
+    {
+      "epoch": 0.26931818181818185,
+      "grad_norm": 1.9859115923781316,
+      "learning_rate": 1.03965717347362e-06,
+      "loss": 0.0836,
+      "step": 1185
+    },
+    {
+      "epoch": 0.26954545454545453,
+      "grad_norm": 3.7587187550089802,
+      "learning_rate": 1.0393231757205824e-06,
+      "loss": 0.1568,
+      "step": 1186
+    },
+    {
+      "epoch": 0.2697727272727273,
+      "grad_norm": 1.3465084917328027,
+      "learning_rate": 1.038988966748244e-06,
+      "loss": 0.1551,
+      "step": 1187
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.076625566330226,
+      "learning_rate": 1.0386545467269823e-06,
+      "loss": 0.153,
+      "step": 1188
+    },
+    {
+      "epoch": 0.2702272727272727,
+      "grad_norm": 1.9296709990409082,
+      "learning_rate": 1.0383199158272825e-06,
+      "loss": 0.1482,
+      "step": 1189
+    },
+    {
+      "epoch": 0.27045454545454545,
+      "grad_norm": 2.323340919101609,
+      "learning_rate": 1.0379850742197374e-06,
+      "loss": 0.1098,
+      "step": 1190
+    },
+    {
+      "epoch": 0.2706818181818182,
+      "grad_norm": 2.907648656621964,
+      "learning_rate": 1.037650022075047e-06,
+      "loss": 0.0908,
+      "step": 1191
+    },
+    {
+      "epoch": 0.27090909090909093,
+      "grad_norm": 3.3877322182021166,
+      "learning_rate": 1.0373147595640183e-06,
+      "loss": 0.2003,
+      "step": 1192
+    },
+    {
+      "epoch": 0.2711363636363636,
+      "grad_norm": 2.598731714841657,
+      "learning_rate": 1.0369792868575669e-06,
+      "loss": 0.1446,
+      "step": 1193
+    },
+    {
+      "epoch": 0.27136363636363636,
+      "grad_norm": 3.838783542621206,
+      "learning_rate": 1.0366436041267142e-06,
+      "loss": 0.2235,
+      "step": 1194
+    },
+    {
+      "epoch": 0.2715909090909091,
+      "grad_norm": 2.940516351148344,
+      "learning_rate": 1.0363077115425888e-06,
+      "loss": 0.1258,
+      "step": 1195
+    },
+    {
+      "epoch": 0.2718181818181818,
+      "grad_norm": 3.85466590886345,
+      "learning_rate": 1.0359716092764268e-06,
+      "loss": 0.1137,
+      "step": 1196
+    },
+    {
+      "epoch": 0.27204545454545453,
+      "grad_norm": 3.1651492430860393,
+      "learning_rate": 1.0356352974995713e-06,
+      "loss": 0.1501,
+      "step": 1197
+    },
+    {
+      "epoch": 0.2722727272727273,
+      "grad_norm": 2.987399819434138,
+      "learning_rate": 1.0352987763834716e-06,
+      "loss": 0.0876,
+      "step": 1198
+    },
+    {
+      "epoch": 0.2725,
+      "grad_norm": 1.6008802229478307,
+      "learning_rate": 1.034962046099684e-06,
+      "loss": 0.1055,
+      "step": 1199
+    },
+    {
+      "epoch": 0.2727272727272727,
+      "grad_norm": 1.5002875972984264,
+      "learning_rate": 1.0346251068198712e-06,
+      "loss": 0.1114,
+      "step": 1200
+    },
+    {
+      "epoch": 0.27295454545454545,
+      "grad_norm": 4.431711057005355,
+      "learning_rate": 1.0342879587158033e-06,
+      "loss": 0.1574,
+      "step": 1201
+    },
+    {
+      "epoch": 0.2731818181818182,
+      "grad_norm": 3.5198631105749163,
+      "learning_rate": 1.0339506019593557e-06,
+      "loss": 0.2101,
+      "step": 1202
+    },
+    {
+      "epoch": 0.27340909090909093,
+      "grad_norm": 3.680094009357352,
+      "learning_rate": 1.033613036722511e-06,
+      "loss": 0.105,
+      "step": 1203
+    },
+    {
+      "epoch": 0.2736363636363636,
+      "grad_norm": 2.0446958410876888,
+      "learning_rate": 1.0332752631773578e-06,
+      "loss": 0.0894,
+      "step": 1204
+    },
+    {
+      "epoch": 0.27386363636363636,
+      "grad_norm": 2.283315545493432,
+      "learning_rate": 1.0329372814960906e-06,
+      "loss": 0.1121,
+      "step": 1205
+    },
+    {
+      "epoch": 0.2740909090909091,
+      "grad_norm": 3.78402533889888,
+      "learning_rate": 1.0325990918510107e-06,
+      "loss": 0.2352,
+      "step": 1206
+    },
+    {
+      "epoch": 0.2743181818181818,
+      "grad_norm": 4.163070599397629,
+      "learning_rate": 1.0322606944145248e-06,
+      "loss": 0.2065,
+      "step": 1207
+    },
+    {
+      "epoch": 0.27454545454545454,
+      "grad_norm": 1.5750889801789514,
+      "learning_rate": 1.0319220893591456e-06,
+      "loss": 0.1604,
+      "step": 1208
+    },
+    {
+      "epoch": 0.2747727272727273,
+      "grad_norm": 3.3728390625847124,
+      "learning_rate": 1.031583276857492e-06,
+      "loss": 0.1084,
+      "step": 1209
+    },
+    {
+      "epoch": 0.275,
+      "grad_norm": 5.555765954102343,
+      "learning_rate": 1.0312442570822883e-06,
+      "loss": 0.2012,
+      "step": 1210
+    },
+    {
+      "epoch": 0.2752272727272727,
+      "grad_norm": 2.9195366370030484,
+      "learning_rate": 1.0309050302063647e-06,
+      "loss": 0.1307,
+      "step": 1211
+    },
+    {
+      "epoch": 0.27545454545454545,
+      "grad_norm": 2.5760484424034784,
+      "learning_rate": 1.0305655964026573e-06,
+      "loss": 0.079,
+      "step": 1212
+    },
+    {
+      "epoch": 0.2756818181818182,
+      "grad_norm": 2.5289673344196943,
+      "learning_rate": 1.0302259558442065e-06,
+      "loss": 0.1431,
+      "step": 1213
+    },
+    {
+      "epoch": 0.2759090909090909,
+      "grad_norm": 3.1924698349917926,
+      "learning_rate": 1.0298861087041593e-06,
+      "loss": 0.0798,
+      "step": 1214
+    },
+    {
+      "epoch": 0.2761363636363636,
+      "grad_norm": 3.4249701785253195,
+      "learning_rate": 1.0295460551557676e-06,
+      "loss": 0.16,
+      "step": 1215
+    },
+    {
+      "epoch": 0.27636363636363637,
+      "grad_norm": 2.5645226760525137,
+      "learning_rate": 1.0292057953723883e-06,
+      "loss": 0.1731,
+      "step": 1216
+    },
+    {
+      "epoch": 0.2765909090909091,
+      "grad_norm": 3.1252342884851503,
+      "learning_rate": 1.028865329527484e-06,
+      "loss": 0.1043,
+      "step": 1217
+    },
+    {
+      "epoch": 0.2768181818181818,
+      "grad_norm": 3.4412095980109987,
+      "learning_rate": 1.0285246577946217e-06,
+      "loss": 0.1487,
+      "step": 1218
+    },
+    {
+      "epoch": 0.27704545454545454,
+      "grad_norm": 2.6919917942454465,
+      "learning_rate": 1.0281837803474739e-06,
+      "loss": 0.0999,
+      "step": 1219
+    },
+    {
+      "epoch": 0.2772727272727273,
+      "grad_norm": 4.108027070798246,
+      "learning_rate": 1.0278426973598176e-06,
+      "loss": 0.149,
+      "step": 1220
+    },
+    {
+      "epoch": 0.2775,
+      "grad_norm": 1.387394906852245,
+      "learning_rate": 1.0275014090055346e-06,
+      "loss": 0.0712,
+      "step": 1221
+    },
+    {
+      "epoch": 0.2777272727272727,
+      "grad_norm": 2.2734736426715436,
+      "learning_rate": 1.0271599154586119e-06,
+      "loss": 0.0975,
+      "step": 1222
+    },
+    {
+      "epoch": 0.27795454545454545,
+      "grad_norm": 2.3865628897621582,
+      "learning_rate": 1.0268182168931404e-06,
+      "loss": 0.1913,
+      "step": 1223
+    },
+    {
+      "epoch": 0.2781818181818182,
+      "grad_norm": 3.53032382662459,
+      "learning_rate": 1.026476313483316e-06,
+      "loss": 0.2502,
+      "step": 1224
+    },
+    {
+      "epoch": 0.2784090909090909,
+      "grad_norm": 2.4243747741356776,
+      "learning_rate": 1.0261342054034387e-06,
+      "loss": 0.0919,
+      "step": 1225
+    },
+    {
+      "epoch": 0.2786363636363636,
+      "grad_norm": 3.8445169147986924,
+      "learning_rate": 1.025791892827913e-06,
+      "loss": 0.1613,
+      "step": 1226
+    },
+    {
+      "epoch": 0.27886363636363637,
+      "grad_norm": 3.0851654483422797,
+      "learning_rate": 1.0254493759312482e-06,
+      "loss": 0.1205,
+      "step": 1227
+    },
+    {
+      "epoch": 0.2790909090909091,
+      "grad_norm": 3.557417726497082,
+      "learning_rate": 1.0251066548880564e-06,
+      "loss": 0.1402,
+      "step": 1228
+    },
+    {
+      "epoch": 0.2793181818181818,
+      "grad_norm": 4.3129289418951995,
+      "learning_rate": 1.024763729873055e-06,
+      "loss": 0.1801,
+      "step": 1229
+    },
+    {
+      "epoch": 0.27954545454545454,
+      "grad_norm": 3.002606190973831,
+      "learning_rate": 1.0244206010610653e-06,
+      "loss": 0.1024,
+      "step": 1230
+    },
+    {
+      "epoch": 0.2797727272727273,
+      "grad_norm": 1.9087802869437935,
+      "learning_rate": 1.0240772686270115e-06,
+      "loss": 0.1354,
+      "step": 1231
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.66057466955001,
+      "learning_rate": 1.0237337327459226e-06,
+      "loss": 0.1137,
+      "step": 1232
+    },
+    {
+      "epoch": 0.2802272727272727,
+      "grad_norm": 2.7039890710365104,
+      "learning_rate": 1.0233899935929311e-06,
+      "loss": 0.1516,
+      "step": 1233
+    },
+    {
+      "epoch": 0.28045454545454546,
+      "grad_norm": 1.8530981479095547,
+      "learning_rate": 1.023046051343273e-06,
+      "loss": 0.1025,
+      "step": 1234
+    },
+    {
+      "epoch": 0.2806818181818182,
+      "grad_norm": 3.0837464482962966,
+      "learning_rate": 1.0227019061722877e-06,
+      "loss": 0.0956,
+      "step": 1235
+    },
+    {
+      "epoch": 0.2809090909090909,
+      "grad_norm": 3.523050624249298,
+      "learning_rate": 1.022357558255418e-06,
+      "loss": 0.1033,
+      "step": 1236
+    },
+    {
+      "epoch": 0.28113636363636363,
+      "grad_norm": 2.682805869757046,
+      "learning_rate": 1.022013007768211e-06,
+      "loss": 0.0916,
+      "step": 1237
+    },
+    {
+      "epoch": 0.28136363636363637,
+      "grad_norm": 2.9542045119245817,
+      "learning_rate": 1.0216682548863155e-06,
+      "loss": 0.2049,
+      "step": 1238
+    },
+    {
+      "epoch": 0.2815909090909091,
+      "grad_norm": 2.3488555378107145,
+      "learning_rate": 1.0213232997854848e-06,
+      "loss": 0.1654,
+      "step": 1239
+    },
+    {
+      "epoch": 0.2818181818181818,
+      "grad_norm": 2.08295954252489,
+      "learning_rate": 1.0209781426415747e-06,
+      "loss": 0.0887,
+      "step": 1240
+    },
+    {
+      "epoch": 0.28204545454545454,
+      "grad_norm": 2.062846656277119,
+      "learning_rate": 1.020632783630544e-06,
+      "loss": 0.0974,
+      "step": 1241
+    },
+    {
+      "epoch": 0.2822727272727273,
+      "grad_norm": 5.039976892823724,
+      "learning_rate": 1.0202872229284546e-06,
+      "loss": 0.1787,
+      "step": 1242
+    },
+    {
+      "epoch": 0.2825,
+      "grad_norm": 2.5189655049623694,
+      "learning_rate": 1.0199414607114713e-06,
+      "loss": 0.12,
+      "step": 1243
+    },
+    {
+      "epoch": 0.2827272727272727,
+      "grad_norm": 4.687340951905225,
+      "learning_rate": 1.0195954971558613e-06,
+      "loss": 0.1556,
+      "step": 1244
+    },
+    {
+      "epoch": 0.28295454545454546,
+      "grad_norm": 2.450782299050815,
+      "learning_rate": 1.0192493324379943e-06,
+      "loss": 0.0894,
+      "step": 1245
+    },
+    {
+      "epoch": 0.2831818181818182,
+      "grad_norm": 3.713648046364993,
+      "learning_rate": 1.0189029667343433e-06,
+      "loss": 0.1226,
+      "step": 1246
+    },
+    {
+      "epoch": 0.2834090909090909,
+      "grad_norm": 4.43266038604316,
+      "learning_rate": 1.0185564002214831e-06,
+      "loss": 0.1305,
+      "step": 1247
+    },
+    {
+      "epoch": 0.28363636363636363,
+      "grad_norm": 2.9982772717128303,
+      "learning_rate": 1.0182096330760912e-06,
+      "loss": 0.1635,
+      "step": 1248
+    },
+    {
+      "epoch": 0.2838636363636364,
+      "grad_norm": 2.4645906958688926,
+      "learning_rate": 1.0178626654749473e-06,
+      "loss": 0.1552,
+      "step": 1249
+    },
+    {
+      "epoch": 0.2840909090909091,
+      "grad_norm": 2.091420740436183,
+      "learning_rate": 1.017515497594933e-06,
+      "loss": 0.1156,
+      "step": 1250
+    },
+    {
+      "epoch": 0.2843181818181818,
+      "grad_norm": 2.5109906190954874,
+      "learning_rate": 1.0171681296130326e-06,
+      "loss": 0.1089,
+      "step": 1251
+    },
+    {
+      "epoch": 0.28454545454545455,
+      "grad_norm": 4.2681764233350785,
+      "learning_rate": 1.0168205617063318e-06,
+      "loss": 0.1423,
+      "step": 1252
+    },
+    {
+      "epoch": 0.2847727272727273,
+      "grad_norm": 2.800916787012341,
+      "learning_rate": 1.0164727940520185e-06,
+      "loss": 0.0903,
+      "step": 1253
+    },
+    {
+      "epoch": 0.285,
+      "grad_norm": 2.465433995497064,
+      "learning_rate": 1.016124826827383e-06,
+      "loss": 0.2051,
+      "step": 1254
+    },
+    {
+      "epoch": 0.2852272727272727,
+      "grad_norm": 2.4098120551722073,
+      "learning_rate": 1.015776660209816e-06,
+      "loss": 0.1135,
+      "step": 1255
+    },
+    {
+      "epoch": 0.28545454545454546,
+      "grad_norm": 4.013886366124501,
+      "learning_rate": 1.0154282943768106e-06,
+      "loss": 0.2614,
+      "step": 1256
+    },
+    {
+      "epoch": 0.2856818181818182,
+      "grad_norm": 2.3481488841511786,
+      "learning_rate": 1.0150797295059623e-06,
+      "loss": 0.1028,
+      "step": 1257
+    },
+    {
+      "epoch": 0.2859090909090909,
+      "grad_norm": 3.331567706580509,
+      "learning_rate": 1.0147309657749663e-06,
+      "loss": 0.1065,
+      "step": 1258
+    },
+    {
+      "epoch": 0.28613636363636363,
+      "grad_norm": 2.4782160347277475,
+      "learning_rate": 1.0143820033616206e-06,
+      "loss": 0.1908,
+      "step": 1259
+    },
+    {
+      "epoch": 0.2863636363636364,
+      "grad_norm": 3.477605214964916,
+      "learning_rate": 1.0140328424438242e-06,
+      "loss": 0.1864,
+      "step": 1260
+    },
+    {
+      "epoch": 0.2865909090909091,
+      "grad_norm": 3.2220424602788786,
+      "learning_rate": 1.0136834831995765e-06,
+      "loss": 0.1462,
+      "step": 1261
+    },
+    {
+      "epoch": 0.2868181818181818,
+      "grad_norm": 1.53671143534119,
+      "learning_rate": 1.0133339258069791e-06,
+      "loss": 0.1748,
+      "step": 1262
+    },
+    {
+      "epoch": 0.28704545454545455,
+      "grad_norm": 4.666495366673696,
+      "learning_rate": 1.0129841704442339e-06,
+      "loss": 0.1844,
+      "step": 1263
+    },
+    {
+      "epoch": 0.2872727272727273,
+      "grad_norm": 1.6940068544527738,
+      "learning_rate": 1.012634217289644e-06,
+      "loss": 0.1385,
+      "step": 1264
+    },
+    {
+      "epoch": 0.2875,
+      "grad_norm": 3.3236447827011983,
+      "learning_rate": 1.0122840665216133e-06,
+      "loss": 0.0952,
+      "step": 1265
+    },
+    {
+      "epoch": 0.2877272727272727,
+      "grad_norm": 2.098549258457893,
+      "learning_rate": 1.0119337183186465e-06,
+      "loss": 0.0857,
+      "step": 1266
+    },
+    {
+      "epoch": 0.28795454545454546,
+      "grad_norm": 5.732429867233003,
+      "learning_rate": 1.0115831728593483e-06,
+      "loss": 0.2075,
+      "step": 1267
+    },
+    {
+      "epoch": 0.2881818181818182,
+      "grad_norm": 3.7427665815869897,
+      "learning_rate": 1.0112324303224255e-06,
+      "loss": 0.1936,
+      "step": 1268
+    },
+    {
+      "epoch": 0.2884090909090909,
+      "grad_norm": 3.282403134582927,
+      "learning_rate": 1.0108814908866837e-06,
+      "loss": 0.1241,
+      "step": 1269
+    },
+    {
+      "epoch": 0.28863636363636364,
+      "grad_norm": 4.823329379934923,
+      "learning_rate": 1.0105303547310298e-06,
+      "loss": 0.1809,
+      "step": 1270
+    },
+    {
+      "epoch": 0.2888636363636364,
+      "grad_norm": 1.7984578657352335,
+      "learning_rate": 1.0101790220344707e-06,
+      "loss": 0.1061,
+      "step": 1271
+    },
+    {
+      "epoch": 0.28909090909090907,
+      "grad_norm": 3.2118939401054147,
+      "learning_rate": 1.0098274929761134e-06,
+      "loss": 0.1734,
+      "step": 1272
+    },
+    {
+      "epoch": 0.2893181818181818,
+      "grad_norm": 2.2536301544671775,
+      "learning_rate": 1.0094757677351658e-06,
+      "loss": 0.1326,
+      "step": 1273
+    },
+    {
+      "epoch": 0.28954545454545455,
+      "grad_norm": 2.573188163660672,
+      "learning_rate": 1.0091238464909343e-06,
+      "loss": 0.1694,
+      "step": 1274
+    },
+    {
+      "epoch": 0.2897727272727273,
+      "grad_norm": 3.3644026378445506,
+      "learning_rate": 1.0087717294228266e-06,
+      "loss": 0.0946,
+      "step": 1275
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9139241425017564,
+      "learning_rate": 1.0084194167103499e-06,
+      "loss": 0.1108,
+      "step": 1276
+    },
+    {
+      "epoch": 0.2902272727272727,
+      "grad_norm": 3.6133670589618685,
+      "learning_rate": 1.0080669085331104e-06,
+      "loss": 0.1033,
+      "step": 1277
+    },
+    {
+      "epoch": 0.29045454545454547,
+      "grad_norm": 3.6329629285779053,
+      "learning_rate": 1.007714205070815e-06,
+      "loss": 0.1363,
+      "step": 1278
+    },
+    {
+      "epoch": 0.2906818181818182,
+      "grad_norm": 3.392935769739737,
+      "learning_rate": 1.0073613065032694e-06,
+      "loss": 0.2473,
+      "step": 1279
+    },
+    {
+      "epoch": 0.2909090909090909,
+      "grad_norm": 3.923413090828727,
+      "learning_rate": 1.007008213010379e-06,
+      "loss": 0.0932,
+      "step": 1280
+    },
+    {
+      "epoch": 0.29113636363636364,
+      "grad_norm": 3.7470039607461927,
+      "learning_rate": 1.0066549247721489e-06,
+      "loss": 0.1905,
+      "step": 1281
+    },
+    {
+      "epoch": 0.2913636363636364,
+      "grad_norm": 3.7695425466720898,
+      "learning_rate": 1.006301441968683e-06,
+      "loss": 0.1235,
+      "step": 1282
+    },
+    {
+      "epoch": 0.29159090909090907,
+      "grad_norm": 2.3909767439084373,
+      "learning_rate": 1.0059477647801845e-06,
+      "loss": 0.1548,
+      "step": 1283
+    },
+    {
+      "epoch": 0.2918181818181818,
+      "grad_norm": 2.817665203018277,
+      "learning_rate": 1.0055938933869557e-06,
+      "loss": 0.2095,
+      "step": 1284
+    },
+    {
+      "epoch": 0.29204545454545455,
+      "grad_norm": 1.8481740652455538,
+      "learning_rate": 1.0052398279693984e-06,
+      "loss": 0.0767,
+      "step": 1285
+    },
+    {
+      "epoch": 0.2922727272727273,
+      "grad_norm": 2.91774609501657,
+      "learning_rate": 1.0048855687080122e-06,
+      "loss": 0.1076,
+      "step": 1286
+    },
+    {
+      "epoch": 0.2925,
+      "grad_norm": 3.1224918951402216,
+      "learning_rate": 1.0045311157833968e-06,
+      "loss": 0.106,
+      "step": 1287
+    },
+    {
+      "epoch": 0.2927272727272727,
+      "grad_norm": 2.532662572143345,
+      "learning_rate": 1.0041764693762499e-06,
+      "loss": 0.0867,
+      "step": 1288
+    },
+    {
+      "epoch": 0.29295454545454547,
+      "grad_norm": 2.473018735325476,
+      "learning_rate": 1.0038216296673678e-06,
+      "loss": 0.069,
+      "step": 1289
+    },
+    {
+      "epoch": 0.29318181818181815,
+      "grad_norm": 1.720480848120981,
+      "learning_rate": 1.0034665968376457e-06,
+      "loss": 0.1207,
+      "step": 1290
+    },
+    {
+      "epoch": 0.2934090909090909,
+      "grad_norm": 4.865547909948531,
+      "learning_rate": 1.0031113710680767e-06,
+      "loss": 0.2132,
+      "step": 1291
+    },
+    {
+      "epoch": 0.29363636363636364,
+      "grad_norm": 4.801645852198265,
+      "learning_rate": 1.002755952539753e-06,
+      "loss": 0.1186,
+      "step": 1292
+    },
+    {
+      "epoch": 0.2938636363636364,
+      "grad_norm": 4.49831065716073,
+      "learning_rate": 1.0024003414338645e-06,
+      "loss": 0.2134,
+      "step": 1293
+    },
+    {
+      "epoch": 0.29409090909090907,
+      "grad_norm": 2.1430344725571784,
+      "learning_rate": 1.0020445379316996e-06,
+      "loss": 0.1126,
+      "step": 1294
+    },
+    {
+      "epoch": 0.2943181818181818,
+      "grad_norm": 3.295072725900097,
+      "learning_rate": 1.0016885422146445e-06,
+      "loss": 0.145,
+      "step": 1295
+    },
+    {
+      "epoch": 0.29454545454545455,
+      "grad_norm": 1.7609069564796003,
+      "learning_rate": 1.0013323544641837e-06,
+      "loss": 0.0781,
+      "step": 1296
+    },
+    {
+      "epoch": 0.2947727272727273,
+      "grad_norm": 2.862338694859953,
+      "learning_rate": 1.0009759748618993e-06,
+      "loss": 0.1108,
+      "step": 1297
+    },
+    {
+      "epoch": 0.295,
+      "grad_norm": 2.527223361758931,
+      "learning_rate": 1.0006194035894714e-06,
+      "loss": 0.1381,
+      "step": 1298
+    },
+    {
+      "epoch": 0.2952272727272727,
+      "grad_norm": 2.1851117306578653,
+      "learning_rate": 1.0002626408286777e-06,
+      "loss": 0.0948,
+      "step": 1299
+    },
+    {
+      "epoch": 0.29545454545454547,
+      "grad_norm": 4.281677444849836,
+      "learning_rate": 9.999056867613935e-07,
+      "loss": 0.1088,
+      "step": 1300
+    },
+    {
+      "epoch": 0.29568181818181816,
+      "grad_norm": 3.1580417233368623,
+      "learning_rate": 9.995485415695919e-07,
+      "loss": 0.1927,
+      "step": 1301
+    },
+    {
+      "epoch": 0.2959090909090909,
+      "grad_norm": 2.9327029018666737,
+      "learning_rate": 9.99191205435343e-07,
+      "loss": 0.1042,
+      "step": 1302
+    },
+    {
+      "epoch": 0.29613636363636364,
+      "grad_norm": 3.049178820933939,
+      "learning_rate": 9.988336785408147e-07,
+      "loss": 0.1542,
+      "step": 1303
+    },
+    {
+      "epoch": 0.2963636363636364,
+      "grad_norm": 2.8816323270330213,
+      "learning_rate": 9.984759610682721e-07,
+      "loss": 0.1226,
+      "step": 1304
+    },
+    {
+      "epoch": 0.29659090909090907,
+      "grad_norm": 2.554895421289665,
+      "learning_rate": 9.981180532000768e-07,
+      "loss": 0.0795,
+      "step": 1305
+    },
+    {
+      "epoch": 0.2968181818181818,
+      "grad_norm": 1.49343375739613,
+      "learning_rate": 9.97759955118688e-07,
+      "loss": 0.0889,
+      "step": 1306
+    },
+    {
+      "epoch": 0.29704545454545456,
+      "grad_norm": 3.59957836579208,
+      "learning_rate": 9.97401667006662e-07,
+      "loss": 0.2435,
+      "step": 1307
+    },
+    {
+      "epoch": 0.2972727272727273,
+      "grad_norm": 1.9157583474000284,
+      "learning_rate": 9.970431890466516e-07,
+      "loss": 0.1478,
+      "step": 1308
+    },
+    {
+      "epoch": 0.2975,
+      "grad_norm": 2.1149785784920256,
+      "learning_rate": 9.96684521421407e-07,
+      "loss": 0.2239,
+      "step": 1309
+    },
+    {
+      "epoch": 0.29772727272727273,
+      "grad_norm": 3.0763793995713935,
+      "learning_rate": 9.963256643137743e-07,
+      "loss": 0.1195,
+      "step": 1310
+    },
+    {
+      "epoch": 0.29795454545454547,
+      "grad_norm": 2.52005558517502,
+      "learning_rate": 9.959666179066965e-07,
+      "loss": 0.1601,
+      "step": 1311
+    },
+    {
+      "epoch": 0.29818181818181816,
+      "grad_norm": 3.355373281244958,
+      "learning_rate": 9.95607382383213e-07,
+      "loss": 0.1105,
+      "step": 1312
+    },
+    {
+      "epoch": 0.2984090909090909,
+      "grad_norm": 4.294359481249815,
+      "learning_rate": 9.952479579264605e-07,
+      "loss": 0.1885,
+      "step": 1313
+    },
+    {
+      "epoch": 0.29863636363636364,
+      "grad_norm": 3.3897578733567375,
+      "learning_rate": 9.948883447196706e-07,
+      "loss": 0.149,
+      "step": 1314
+    },
+    {
+      "epoch": 0.2988636363636364,
+      "grad_norm": 3.0137619476597357,
+      "learning_rate": 9.945285429461722e-07,
+      "loss": 0.128,
+      "step": 1315
+    },
+    {
+      "epoch": 0.2990909090909091,
+      "grad_norm": 2.6188845567613437,
+      "learning_rate": 9.941685527893894e-07,
+      "loss": 0.1046,
+      "step": 1316
+    },
+    {
+      "epoch": 0.2993181818181818,
+      "grad_norm": 2.3062392933373155,
+      "learning_rate": 9.938083744328436e-07,
+      "loss": 0.0805,
+      "step": 1317
+    },
+    {
+      "epoch": 0.29954545454545456,
+      "grad_norm": 3.320355437954206,
+      "learning_rate": 9.934480080601507e-07,
+      "loss": 0.1243,
+      "step": 1318
+    },
+    {
+      "epoch": 0.29977272727272725,
+      "grad_norm": 4.16372063509409,
+      "learning_rate": 9.930874538550238e-07,
+      "loss": 0.1183,
+      "step": 1319
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.7525700334337837,
+      "learning_rate": 9.92726712001271e-07,
+      "loss": 0.1002,
+      "step": 1320
+    },
+    {
+      "epoch": 0.30022727272727273,
+      "grad_norm": 2.5668952122333044,
+      "learning_rate": 9.923657826827958e-07,
+      "loss": 0.1767,
+      "step": 1321
+    },
+    {
+      "epoch": 0.3004545454545455,
+      "grad_norm": 2.8034529369153955,
+      "learning_rate": 9.92004666083598e-07,
+      "loss": 0.1763,
+      "step": 1322
+    },
+    {
+      "epoch": 0.30068181818181816,
+      "grad_norm": 2.8518249924641315,
+      "learning_rate": 9.916433623877722e-07,
+      "loss": 0.1166,
+      "step": 1323
+    },
+    {
+      "epoch": 0.3009090909090909,
+      "grad_norm": 2.9408399707911266,
+      "learning_rate": 9.91281871779509e-07,
+      "loss": 0.1346,
+      "step": 1324
+    },
+    {
+      "epoch": 0.30113636363636365,
+      "grad_norm": 3.320909794955794,
+      "learning_rate": 9.909201944430943e-07,
+      "loss": 0.2913,
+      "step": 1325
+    },
+    {
+      "epoch": 0.3013636363636364,
+      "grad_norm": 2.7892952811933,
+      "learning_rate": 9.905583305629082e-07,
+      "loss": 0.1298,
+      "step": 1326
+    },
+    {
+      "epoch": 0.3015909090909091,
+      "grad_norm": 3.3703917647280583,
+      "learning_rate": 9.90196280323427e-07,
+      "loss": 0.1215,
+      "step": 1327
+    },
+    {
+      "epoch": 0.3018181818181818,
+      "grad_norm": 3.0833600148908977,
+      "learning_rate": 9.898340439092216e-07,
+      "loss": 0.2274,
+      "step": 1328
+    },
+    {
+      "epoch": 0.30204545454545456,
+      "grad_norm": 3.0785923108966595,
+      "learning_rate": 9.894716215049575e-07,
+      "loss": 0.2397,
+      "step": 1329
+    },
+    {
+      "epoch": 0.30227272727272725,
+      "grad_norm": 2.7090784663515786,
+      "learning_rate": 9.891090132953957e-07,
+      "loss": 0.0805,
+      "step": 1330
+    },
+    {
+      "epoch": 0.3025,
+      "grad_norm": 2.6505598509516264,
+      "learning_rate": 9.887462194653912e-07,
+      "loss": 0.1778,
+      "step": 1331
+    },
+    {
+      "epoch": 0.30272727272727273,
+      "grad_norm": 3.8033764924339173,
+      "learning_rate": 9.883832401998942e-07,
+      "loss": 0.1181,
+      "step": 1332
+    },
+    {
+      "epoch": 0.3029545454545455,
+      "grad_norm": 1.6234090695032544,
+      "learning_rate": 9.880200756839492e-07,
+      "loss": 0.1157,
+      "step": 1333
+    },
+    {
+      "epoch": 0.30318181818181816,
+      "grad_norm": 3.775448542701932,
+      "learning_rate": 9.87656726102695e-07,
+      "loss": 0.1356,
+      "step": 1334
+    },
+    {
+      "epoch": 0.3034090909090909,
+      "grad_norm": 3.0646944344303297,
+      "learning_rate": 9.87293191641365e-07,
+      "loss": 0.1575,
+      "step": 1335
+    },
+    {
+      "epoch": 0.30363636363636365,
+      "grad_norm": 2.1481025068990323,
+      "learning_rate": 9.869294724852868e-07,
+      "loss": 0.0823,
+      "step": 1336
+    },
+    {
+      "epoch": 0.3038636363636364,
+      "grad_norm": 2.9598143286603182,
+      "learning_rate": 9.86565568819882e-07,
+      "loss": 0.1544,
+      "step": 1337
+    },
+    {
+      "epoch": 0.3040909090909091,
+      "grad_norm": 3.1440494635036167,
+      "learning_rate": 9.86201480830666e-07,
+      "loss": 0.1939,
+      "step": 1338
+    },
+    {
+      "epoch": 0.3043181818181818,
+      "grad_norm": 1.5547080685143568,
+      "learning_rate": 9.858372087032493e-07,
+      "loss": 0.0507,
+      "step": 1339
+    },
+    {
+      "epoch": 0.30454545454545456,
+      "grad_norm": 4.009379201583973,
+      "learning_rate": 9.854727526233351e-07,
+      "loss": 0.1547,
+      "step": 1340
+    },
+    {
+      "epoch": 0.30477272727272725,
+      "grad_norm": 3.50311248570337,
+      "learning_rate": 9.851081127767206e-07,
+      "loss": 0.1335,
+      "step": 1341
+    },
+    {
+      "epoch": 0.305,
+      "grad_norm": 3.3634163687308103,
+      "learning_rate": 9.847432893492972e-07,
+      "loss": 0.1137,
+      "step": 1342
+    },
+    {
+      "epoch": 0.30522727272727274,
+      "grad_norm": 2.3988016734470414,
+      "learning_rate": 9.843782825270494e-07,
+      "loss": 0.1267,
+      "step": 1343
+    },
+    {
+      "epoch": 0.3054545454545455,
+      "grad_norm": 3.807427250371877,
+      "learning_rate": 9.84013092496055e-07,
+      "loss": 0.2799,
+      "step": 1344
+    },
+    {
+      "epoch": 0.30568181818181817,
+      "grad_norm": 2.68148562014571,
+      "learning_rate": 9.83647719442486e-07,
+      "loss": 0.1798,
+      "step": 1345
+    },
+    {
+      "epoch": 0.3059090909090909,
+      "grad_norm": 3.373555002044503,
+      "learning_rate": 9.832821635526072e-07,
+      "loss": 0.1134,
+      "step": 1346
+    },
+    {
+      "epoch": 0.30613636363636365,
+      "grad_norm": 4.467104449034929,
+      "learning_rate": 9.829164250127761e-07,
+      "loss": 0.1743,
+      "step": 1347
+    },
+    {
+      "epoch": 0.30636363636363634,
+      "grad_norm": 3.449777425805995,
+      "learning_rate": 9.82550504009445e-07,
+      "loss": 0.1383,
+      "step": 1348
+    },
+    {
+      "epoch": 0.3065909090909091,
+      "grad_norm": 3.7464255793210866,
+      "learning_rate": 9.82184400729157e-07,
+      "loss": 0.1177,
+      "step": 1349
+    },
+    {
+      "epoch": 0.3068181818181818,
+      "grad_norm": 3.2416932090025887,
+      "learning_rate": 9.818181153585495e-07,
+      "loss": 0.1097,
+      "step": 1350
+    },
+    {
+      "epoch": 0.30704545454545457,
+      "grad_norm": 4.927955199488367,
+      "learning_rate": 9.814516480843526e-07,
+      "loss": 0.183,
+      "step": 1351
+    },
+    {
+      "epoch": 0.30727272727272725,
+      "grad_norm": 4.130835112868002,
+      "learning_rate": 9.810849990933891e-07,
+      "loss": 0.2214,
+      "step": 1352
+    },
+    {
+      "epoch": 0.3075,
+      "grad_norm": 2.092555318254439,
+      "learning_rate": 9.80718168572574e-07,
+      "loss": 0.146,
+      "step": 1353
+    },
+    {
+      "epoch": 0.30772727272727274,
+      "grad_norm": 4.624238202746889,
+      "learning_rate": 9.803511567089151e-07,
+      "loss": 0.1816,
+      "step": 1354
+    },
+    {
+      "epoch": 0.3079545454545455,
+      "grad_norm": 2.2567251729810964,
+      "learning_rate": 9.799839636895129e-07,
+      "loss": 0.1027,
+      "step": 1355
+    },
+    {
+      "epoch": 0.30818181818181817,
+      "grad_norm": 2.6928136034878647,
+      "learning_rate": 9.796165897015602e-07,
+      "loss": 0.1016,
+      "step": 1356
+    },
+    {
+      "epoch": 0.3084090909090909,
+      "grad_norm": 2.897310614288671,
+      "learning_rate": 9.792490349323415e-07,
+      "loss": 0.1257,
+      "step": 1357
+    },
+    {
+      "epoch": 0.30863636363636365,
+      "grad_norm": 1.9111600235307362,
+      "learning_rate": 9.788812995692342e-07,
+      "loss": 0.1723,
+      "step": 1358
+    },
+    {
+      "epoch": 0.30886363636363634,
+      "grad_norm": 2.640702305942462,
+      "learning_rate": 9.78513383799707e-07,
+      "loss": 0.1606,
+      "step": 1359
+    },
+    {
+      "epoch": 0.3090909090909091,
+      "grad_norm": 1.5528942608378375,
+      "learning_rate": 9.781452878113219e-07,
+      "loss": 0.1407,
+      "step": 1360
+    },
+    {
+      "epoch": 0.3093181818181818,
+      "grad_norm": 4.151701265380046,
+      "learning_rate": 9.77777011791731e-07,
+      "loss": 0.1786,
+      "step": 1361
+    },
+    {
+      "epoch": 0.30954545454545457,
+      "grad_norm": 2.727217967327602,
+      "learning_rate": 9.774085559286793e-07,
+      "loss": 0.1273,
+      "step": 1362
+    },
+    {
+      "epoch": 0.30977272727272726,
+      "grad_norm": 2.2017523567817645,
+      "learning_rate": 9.770399204100032e-07,
+      "loss": 0.1055,
+      "step": 1363
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9542195545641395,
+      "learning_rate": 9.766711054236308e-07,
+      "loss": 0.0402,
+      "step": 1364
+    },
+    {
+      "epoch": 0.31022727272727274,
+      "grad_norm": 4.244110531843053,
+      "learning_rate": 9.763021111575816e-07,
+      "loss": 0.2823,
+      "step": 1365
+    },
+    {
+      "epoch": 0.3104545454545454,
+      "grad_norm": 2.539433640984445,
+      "learning_rate": 9.75932937799967e-07,
+      "loss": 0.1898,
+      "step": 1366
+    },
+    {
+      "epoch": 0.31068181818181817,
+      "grad_norm": 3.577627608830666,
+      "learning_rate": 9.755635855389884e-07,
+      "loss": 0.1274,
+      "step": 1367
+    },
+    {
+      "epoch": 0.3109090909090909,
+      "grad_norm": 2.7227366157499007,
+      "learning_rate": 9.751940545629394e-07,
+      "loss": 0.1191,
+      "step": 1368
+    },
+    {
+      "epoch": 0.31113636363636366,
+      "grad_norm": 4.83668966793199,
+      "learning_rate": 9.748243450602049e-07,
+      "loss": 0.2684,
+      "step": 1369
+    },
+    {
+      "epoch": 0.31136363636363634,
+      "grad_norm": 2.2016367001487636,
+      "learning_rate": 9.7445445721926e-07,
+      "loss": 0.0566,
+      "step": 1370
+    },
+    {
+      "epoch": 0.3115909090909091,
+      "grad_norm": 4.041931435217974,
+      "learning_rate": 9.740843912286717e-07,
+      "loss": 0.198,
+      "step": 1371
+    },
+    {
+      "epoch": 0.3118181818181818,
+      "grad_norm": 4.041761166356787,
+      "learning_rate": 9.737141472770967e-07,
+      "loss": 0.1254,
+      "step": 1372
+    },
+    {
+      "epoch": 0.31204545454545457,
+      "grad_norm": 2.6172983211790877,
+      "learning_rate": 9.733437255532836e-07,
+      "loss": 0.1272,
+      "step": 1373
+    },
+    {
+      "epoch": 0.31227272727272726,
+      "grad_norm": 4.778599732328708,
+      "learning_rate": 9.729731262460702e-07,
+      "loss": 0.1833,
+      "step": 1374
+    },
+    {
+      "epoch": 0.3125,
+      "grad_norm": 2.605311362256887,
+      "learning_rate": 9.726023495443866e-07,
+      "loss": 0.1444,
+      "step": 1375
+    },
+    {
+      "epoch": 0.31272727272727274,
+      "grad_norm": 2.505572326729051,
+      "learning_rate": 9.722313956372515e-07,
+      "loss": 0.1421,
+      "step": 1376
+    },
+    {
+      "epoch": 0.31295454545454543,
+      "grad_norm": 2.5737141698605344,
+      "learning_rate": 9.718602647137754e-07,
+      "loss": 0.1108,
+      "step": 1377
+    },
+    {
+      "epoch": 0.3131818181818182,
+      "grad_norm": 2.1710870728382106,
+      "learning_rate": 9.714889569631583e-07,
+      "loss": 0.1004,
+      "step": 1378
+    },
+    {
+      "epoch": 0.3134090909090909,
+      "grad_norm": 2.724019542844689,
+      "learning_rate": 9.711174725746904e-07,
+      "loss": 0.123,
+      "step": 1379
+    },
+    {
+      "epoch": 0.31363636363636366,
+      "grad_norm": 2.889848520767696,
+      "learning_rate": 9.707458117377525e-07,
+      "loss": 0.122,
+      "step": 1380
+    },
+    {
+      "epoch": 0.31386363636363634,
+      "grad_norm": 1.6182007735389823,
+      "learning_rate": 9.703739746418141e-07,
+      "loss": 0.1401,
+      "step": 1381
+    },
+    {
+      "epoch": 0.3140909090909091,
+      "grad_norm": 2.6844934799875473,
+      "learning_rate": 9.700019614764364e-07,
+      "loss": 0.1192,
+      "step": 1382
+    },
+    {
+      "epoch": 0.31431818181818183,
+      "grad_norm": 2.9823777292406084,
+      "learning_rate": 9.696297724312683e-07,
+      "loss": 0.081,
+      "step": 1383
+    },
+    {
+      "epoch": 0.3145454545454546,
+      "grad_norm": 1.868224847936481,
+      "learning_rate": 9.692574076960501e-07,
+      "loss": 0.0906,
+      "step": 1384
+    },
+    {
+      "epoch": 0.31477272727272726,
+      "grad_norm": 2.3709564273523878,
+      "learning_rate": 9.688848674606108e-07,
+      "loss": 0.1839,
+      "step": 1385
+    },
+    {
+      "epoch": 0.315,
+      "grad_norm": 3.3328869318833894,
+      "learning_rate": 9.68512151914869e-07,
+      "loss": 0.1303,
+      "step": 1386
+    },
+    {
+      "epoch": 0.31522727272727274,
+      "grad_norm": 2.8949698241640864,
+      "learning_rate": 9.681392612488326e-07,
+      "loss": 0.1575,
+      "step": 1387
+    },
+    {
+      "epoch": 0.31545454545454543,
+      "grad_norm": 3.830557620990231,
+      "learning_rate": 9.677661956525986e-07,
+      "loss": 0.1174,
+      "step": 1388
+    },
+    {
+      "epoch": 0.3156818181818182,
+      "grad_norm": 3.1598108283016044,
+      "learning_rate": 9.673929553163541e-07,
+      "loss": 0.1415,
+      "step": 1389
+    },
+    {
+      "epoch": 0.3159090909090909,
+      "grad_norm": 3.3887097834736846,
+      "learning_rate": 9.670195404303744e-07,
+      "loss": 0.1059,
+      "step": 1390
+    },
+    {
+      "epoch": 0.31613636363636366,
+      "grad_norm": 2.4437900931992322,
+      "learning_rate": 9.666459511850237e-07,
+      "loss": 0.0908,
+      "step": 1391
+    },
+    {
+      "epoch": 0.31636363636363635,
+      "grad_norm": 2.951827620044325,
+      "learning_rate": 9.662721877707557e-07,
+      "loss": 0.2625,
+      "step": 1392
+    },
+    {
+      "epoch": 0.3165909090909091,
+      "grad_norm": 3.72757338367346,
+      "learning_rate": 9.658982503781127e-07,
+      "loss": 0.1689,
+      "step": 1393
+    },
+    {
+      "epoch": 0.31681818181818183,
+      "grad_norm": 2.2016990687648503,
+      "learning_rate": 9.655241391977251e-07,
+      "loss": 0.105,
+      "step": 1394
+    },
+    {
+      "epoch": 0.3170454545454545,
+      "grad_norm": 2.038430231022133,
+      "learning_rate": 9.65149854420313e-07,
+      "loss": 0.1662,
+      "step": 1395
+    },
+    {
+      "epoch": 0.31727272727272726,
+      "grad_norm": 2.9243283104284736,
+      "learning_rate": 9.64775396236684e-07,
+      "loss": 0.1236,
+      "step": 1396
+    },
+    {
+      "epoch": 0.3175,
+      "grad_norm": 2.218908856475908,
+      "learning_rate": 9.644007648377347e-07,
+      "loss": 0.1174,
+      "step": 1397
+    },
+    {
+      "epoch": 0.31772727272727275,
+      "grad_norm": 3.134174414423278,
+      "learning_rate": 9.640259604144493e-07,
+      "loss": 0.0898,
+      "step": 1398
+    },
+    {
+      "epoch": 0.31795454545454543,
+      "grad_norm": 2.8894377053750695,
+      "learning_rate": 9.636509831579013e-07,
+      "loss": 0.1272,
+      "step": 1399
+    },
+    {
+      "epoch": 0.3181818181818182,
+      "grad_norm": 2.4996744156762594,
+      "learning_rate": 9.632758332592514e-07,
+      "loss": 0.2096,
+      "step": 1400
+    },
+    {
+      "epoch": 0.3184090909090909,
+      "grad_norm": 4.3388961707823475,
+      "learning_rate": 9.629005109097486e-07,
+      "loss": 0.2186,
+      "step": 1401
+    },
+    {
+      "epoch": 0.31863636363636366,
+      "grad_norm": 1.96125394038812,
+      "learning_rate": 9.6252501630073e-07,
+      "loss": 0.0926,
+      "step": 1402
+    },
+    {
+      "epoch": 0.31886363636363635,
+      "grad_norm": 3.131030072338649,
+      "learning_rate": 9.6214934962362e-07,
+      "loss": 0.0761,
+      "step": 1403
+    },
+    {
+      "epoch": 0.3190909090909091,
+      "grad_norm": 3.569075375429256,
+      "learning_rate": 9.617735110699317e-07,
+      "loss": 0.1975,
+      "step": 1404
+    },
+    {
+      "epoch": 0.31931818181818183,
+      "grad_norm": 2.399677405327604,
+      "learning_rate": 9.613975008312646e-07,
+      "loss": 0.1607,
+      "step": 1405
+    },
+    {
+      "epoch": 0.3195454545454545,
+      "grad_norm": 1.638705299602623,
+      "learning_rate": 9.610213190993063e-07,
+      "loss": 0.1083,
+      "step": 1406
+    },
+    {
+      "epoch": 0.31977272727272726,
+      "grad_norm": 2.22793466798009,
+      "learning_rate": 9.606449660658323e-07,
+      "loss": 0.1518,
+      "step": 1407
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9910543845187434,
+      "learning_rate": 9.602684419227047e-07,
+      "loss": 0.1577,
+      "step": 1408
+    },
+    {
+      "epoch": 0.32022727272727275,
+      "grad_norm": 2.8772829999086635,
+      "learning_rate": 9.59891746861873e-07,
+      "loss": 0.1179,
+      "step": 1409
+    },
+    {
+      "epoch": 0.32045454545454544,
+      "grad_norm": 5.543134639937321,
+      "learning_rate": 9.59514881075374e-07,
+      "loss": 0.2971,
+      "step": 1410
+    },
+    {
+      "epoch": 0.3206818181818182,
+      "grad_norm": 3.095033817653101,
+      "learning_rate": 9.591378447553316e-07,
+      "loss": 0.1273,
+      "step": 1411
+    },
+    {
+      "epoch": 0.3209090909090909,
+      "grad_norm": 2.100000046013805,
+      "learning_rate": 9.587606380939565e-07,
+      "loss": 0.1138,
+      "step": 1412
+    },
+    {
+      "epoch": 0.3211363636363636,
+      "grad_norm": 2.9211799656998743,
+      "learning_rate": 9.58383261283546e-07,
+      "loss": 0.2079,
+      "step": 1413
+    },
+    {
+      "epoch": 0.32136363636363635,
+      "grad_norm": 3.2216141279023653,
+      "learning_rate": 9.580057145164844e-07,
+      "loss": 0.0945,
+      "step": 1414
+    },
+    {
+      "epoch": 0.3215909090909091,
+      "grad_norm": 1.4901272427834402,
+      "learning_rate": 9.576279979852429e-07,
+      "loss": 0.1108,
+      "step": 1415
+    },
+    {
+      "epoch": 0.32181818181818184,
+      "grad_norm": 4.134060884019447,
+      "learning_rate": 9.572501118823787e-07,
+      "loss": 0.1124,
+      "step": 1416
+    },
+    {
+      "epoch": 0.3220454545454545,
+      "grad_norm": 3.062684516028034,
+      "learning_rate": 9.56872056400536e-07,
+      "loss": 0.0749,
+      "step": 1417
+    },
+    {
+      "epoch": 0.32227272727272727,
+      "grad_norm": 2.4686038575930933,
+      "learning_rate": 9.564938317324447e-07,
+      "loss": 0.1579,
+      "step": 1418
+    },
+    {
+      "epoch": 0.3225,
+      "grad_norm": 2.0867120007789817,
+      "learning_rate": 9.561154380709218e-07,
+      "loss": 0.0846,
+      "step": 1419
+    },
+    {
+      "epoch": 0.32272727272727275,
+      "grad_norm": 4.022420829051529,
+      "learning_rate": 9.557368756088691e-07,
+      "loss": 0.1651,
+      "step": 1420
+    },
+    {
+      "epoch": 0.32295454545454544,
+      "grad_norm": 2.7111082637467008,
+      "learning_rate": 9.553581445392762e-07,
+      "loss": 0.1426,
+      "step": 1421
+    },
+    {
+      "epoch": 0.3231818181818182,
+      "grad_norm": 1.9133166624011744,
+      "learning_rate": 9.549792450552176e-07,
+      "loss": 0.1711,
+      "step": 1422
+    },
+    {
+      "epoch": 0.3234090909090909,
+      "grad_norm": 3.040299787254641,
+      "learning_rate": 9.546001773498537e-07,
+      "loss": 0.1394,
+      "step": 1423
+    },
+    {
+      "epoch": 0.3236363636363636,
+      "grad_norm": 2.0438246338657824,
+      "learning_rate": 9.542209416164305e-07,
+      "loss": 0.124,
+      "step": 1424
+    },
+    {
+      "epoch": 0.32386363636363635,
+      "grad_norm": 4.363187247655015,
+      "learning_rate": 9.538415380482804e-07,
+      "loss": 0.1292,
+      "step": 1425
+    },
+    {
+      "epoch": 0.3240909090909091,
+      "grad_norm": 2.0486475015332632,
+      "learning_rate": 9.534619668388206e-07,
+      "loss": 0.1455,
+      "step": 1426
+    },
+    {
+      "epoch": 0.32431818181818184,
+      "grad_norm": 2.772477301875394,
+      "learning_rate": 9.530822281815543e-07,
+      "loss": 0.1081,
+      "step": 1427
+    },
+    {
+      "epoch": 0.3245454545454545,
+      "grad_norm": 2.8505339533369516,
+      "learning_rate": 9.527023222700697e-07,
+      "loss": 0.1484,
+      "step": 1428
+    },
+    {
+      "epoch": 0.32477272727272727,
+      "grad_norm": 2.7976466631523267,
+      "learning_rate": 9.523222492980405e-07,
+      "loss": 0.0772,
+      "step": 1429
+    },
+    {
+      "epoch": 0.325,
+      "grad_norm": 1.8732810863228075,
+      "learning_rate": 9.519420094592252e-07,
+      "loss": 0.0812,
+      "step": 1430
+    },
+    {
+      "epoch": 0.32522727272727275,
+      "grad_norm": 3.328701235720672,
+      "learning_rate": 9.51561602947468e-07,
+      "loss": 0.1126,
+      "step": 1431
+    },
+    {
+      "epoch": 0.32545454545454544,
+      "grad_norm": 3.402821501097956,
+      "learning_rate": 9.511810299566976e-07,
+      "loss": 0.0941,
+      "step": 1432
+    },
+    {
+      "epoch": 0.3256818181818182,
+      "grad_norm": 2.5588543184142765,
+      "learning_rate": 9.508002906809276e-07,
+      "loss": 0.1646,
+      "step": 1433
+    },
+    {
+      "epoch": 0.3259090909090909,
+      "grad_norm": 3.4268718528710664,
+      "learning_rate": 9.504193853142564e-07,
+      "loss": 0.1152,
+      "step": 1434
+    },
+    {
+      "epoch": 0.3261363636363636,
+      "grad_norm": 3.7591500139640397,
+      "learning_rate": 9.500383140508671e-07,
+      "loss": 0.1503,
+      "step": 1435
+    },
+    {
+      "epoch": 0.32636363636363636,
+      "grad_norm": 2.995643866861146,
+      "learning_rate": 9.496570770850274e-07,
+      "loss": 0.0807,
+      "step": 1436
+    },
+    {
+      "epoch": 0.3265909090909091,
+      "grad_norm": 2.1464734473404485,
+      "learning_rate": 9.492756746110898e-07,
+      "loss": 0.0802,
+      "step": 1437
+    },
+    {
+      "epoch": 0.32681818181818184,
+      "grad_norm": 2.1157429213643466,
+      "learning_rate": 9.488941068234904e-07,
+      "loss": 0.1599,
+      "step": 1438
+    },
+    {
+      "epoch": 0.32704545454545453,
+      "grad_norm": 2.5434835535575044,
+      "learning_rate": 9.485123739167501e-07,
+      "loss": 0.1255,
+      "step": 1439
+    },
+    {
+      "epoch": 0.32727272727272727,
+      "grad_norm": 2.7006056497282933,
+      "learning_rate": 9.481304760854743e-07,
+      "loss": 0.1316,
+      "step": 1440
+    },
+    {
+      "epoch": 0.3275,
+      "grad_norm": 3.501093584573387,
+      "learning_rate": 9.477484135243513e-07,
+      "loss": 0.1333,
+      "step": 1441
+    },
+    {
+      "epoch": 0.3277272727272727,
+      "grad_norm": 4.119385468097318,
+      "learning_rate": 9.473661864281547e-07,
+      "loss": 0.2449,
+      "step": 1442
+    },
+    {
+      "epoch": 0.32795454545454544,
+      "grad_norm": 4.285628266720916,
+      "learning_rate": 9.469837949917415e-07,
+      "loss": 0.2006,
+      "step": 1443
+    },
+    {
+      "epoch": 0.3281818181818182,
+      "grad_norm": 3.720190097206956,
+      "learning_rate": 9.466012394100519e-07,
+      "loss": 0.1486,
+      "step": 1444
+    },
+    {
+      "epoch": 0.32840909090909093,
+      "grad_norm": 3.277887657683287,
+      "learning_rate": 9.46218519878111e-07,
+      "loss": 0.1757,
+      "step": 1445
+    },
+    {
+      "epoch": 0.3286363636363636,
+      "grad_norm": 4.015487478379219,
+      "learning_rate": 9.458356365910259e-07,
+      "loss": 0.158,
+      "step": 1446
+    },
+    {
+      "epoch": 0.32886363636363636,
+      "grad_norm": 2.8114250060748702,
+      "learning_rate": 9.454525897439888e-07,
+      "loss": 0.1368,
+      "step": 1447
+    },
+    {
+      "epoch": 0.3290909090909091,
+      "grad_norm": 2.4705744484822056,
+      "learning_rate": 9.450693795322742e-07,
+      "loss": 0.0908,
+      "step": 1448
+    },
+    {
+      "epoch": 0.32931818181818184,
+      "grad_norm": 1.7858088238116192,
+      "learning_rate": 9.446860061512402e-07,
+      "loss": 0.0811,
+      "step": 1449
+    },
+    {
+      "epoch": 0.32954545454545453,
+      "grad_norm": 3.0223335453716413,
+      "learning_rate": 9.443024697963279e-07,
+      "loss": 0.0928,
+      "step": 1450
+    },
+    {
+      "epoch": 0.3297727272727273,
+      "grad_norm": 2.6877910287778684,
+      "learning_rate": 9.439187706630625e-07,
+      "loss": 0.0957,
+      "step": 1451
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.9458432288083274,
+      "learning_rate": 9.435349089470502e-07,
+      "loss": 0.2021,
+      "step": 1452
+    },
+    {
+      "epoch": 0.3302272727272727,
+      "grad_norm": 1.714472473655519,
+      "learning_rate": 9.431508848439822e-07,
+      "loss": 0.0656,
+      "step": 1453
+    },
+    {
+      "epoch": 0.33045454545454545,
+      "grad_norm": 2.30655294165752,
+      "learning_rate": 9.427666985496311e-07,
+      "loss": 0.1645,
+      "step": 1454
+    },
+    {
+      "epoch": 0.3306818181818182,
+      "grad_norm": 2.712822371050581,
+      "learning_rate": 9.423823502598526e-07,
+      "loss": 0.1424,
+      "step": 1455
+    },
+    {
+      "epoch": 0.33090909090909093,
+      "grad_norm": 3.6511965031808993,
+      "learning_rate": 9.419978401705849e-07,
+      "loss": 0.1421,
+      "step": 1456
+    },
+    {
+      "epoch": 0.3311363636363636,
+      "grad_norm": 3.052364393776927,
+      "learning_rate": 9.41613168477849e-07,
+      "loss": 0.0936,
+      "step": 1457
+    },
+    {
+      "epoch": 0.33136363636363636,
+      "grad_norm": 5.557105712905154,
+      "learning_rate": 9.412283353777482e-07,
+      "loss": 0.1895,
+      "step": 1458
+    },
+    {
+      "epoch": 0.3315909090909091,
+      "grad_norm": 3.4383855543413104,
+      "learning_rate": 9.408433410664675e-07,
+      "loss": 0.1361,
+      "step": 1459
+    },
+    {
+      "epoch": 0.33181818181818185,
+      "grad_norm": 2.1900595403013225,
+      "learning_rate": 9.40458185740275e-07,
+      "loss": 0.1002,
+      "step": 1460
+    },
+    {
+      "epoch": 0.33204545454545453,
+      "grad_norm": 2.4725264361642774,
+      "learning_rate": 9.4007286959552e-07,
+      "loss": 0.073,
+      "step": 1461
+    },
+    {
+      "epoch": 0.3322727272727273,
+      "grad_norm": 3.177841913289749,
+      "learning_rate": 9.396873928286346e-07,
+      "loss": 0.0821,
+      "step": 1462
+    },
+    {
+      "epoch": 0.3325,
+      "grad_norm": 2.3792984902180865,
+      "learning_rate": 9.393017556361321e-07,
+      "loss": 0.1702,
+      "step": 1463
+    },
+    {
+      "epoch": 0.3327272727272727,
+      "grad_norm": 4.583218606386164,
+      "learning_rate": 9.389159582146079e-07,
+      "loss": 0.1367,
+      "step": 1464
+    },
+    {
+      "epoch": 0.33295454545454545,
+      "grad_norm": 1.8214654804528765,
+      "learning_rate": 9.385300007607392e-07,
+      "loss": 0.0951,
+      "step": 1465
+    },
+    {
+      "epoch": 0.3331818181818182,
+      "grad_norm": 3.1201876047495247,
+      "learning_rate": 9.381438834712846e-07,
+      "loss": 0.1149,
+      "step": 1466
+    },
+    {
+      "epoch": 0.33340909090909093,
+      "grad_norm": 2.790461863572778,
+      "learning_rate": 9.377576065430841e-07,
+      "loss": 0.0937,
+      "step": 1467
+    },
+    {
+      "epoch": 0.3336363636363636,
+      "grad_norm": 2.3025998342941287,
+      "learning_rate": 9.373711701730593e-07,
+      "loss": 0.0983,
+      "step": 1468
+    },
+    {
+      "epoch": 0.33386363636363636,
+      "grad_norm": 3.518138864020965,
+      "learning_rate": 9.369845745582129e-07,
+      "loss": 0.2032,
+      "step": 1469
+    },
+    {
+      "epoch": 0.3340909090909091,
+      "grad_norm": 1.1822571567587599,
+      "learning_rate": 9.365978198956289e-07,
+      "loss": 0.1126,
+      "step": 1470
+    },
+    {
+      "epoch": 0.3343181818181818,
+      "grad_norm": 2.857675228014516,
+      "learning_rate": 9.362109063824725e-07,
+      "loss": 0.1925,
+      "step": 1471
+    },
+    {
+      "epoch": 0.33454545454545453,
+      "grad_norm": 1.1598847148060965,
+      "learning_rate": 9.358238342159898e-07,
+      "loss": 0.0719,
+      "step": 1472
+    },
+    {
+      "epoch": 0.3347727272727273,
+      "grad_norm": 1.8090967317090128,
+      "learning_rate": 9.354366035935071e-07,
+      "loss": 0.1157,
+      "step": 1473
+    },
+    {
+      "epoch": 0.335,
+      "grad_norm": 3.5569747076888705,
+      "learning_rate": 9.350492147124328e-07,
+      "loss": 0.1584,
+      "step": 1474
+    },
+    {
+      "epoch": 0.3352272727272727,
+      "grad_norm": 2.235620576417635,
+      "learning_rate": 9.346616677702548e-07,
+      "loss": 0.2119,
+      "step": 1475
+    },
+    {
+      "epoch": 0.33545454545454545,
+      "grad_norm": 3.7946169140771504,
+      "learning_rate": 9.342739629645423e-07,
+      "loss": 0.1571,
+      "step": 1476
+    },
+    {
+      "epoch": 0.3356818181818182,
+      "grad_norm": 4.337483593789708,
+      "learning_rate": 9.338861004929447e-07,
+      "loss": 0.1494,
+      "step": 1477
+    },
+    {
+      "epoch": 0.33590909090909093,
+      "grad_norm": 2.4710942508140987,
+      "learning_rate": 9.334980805531918e-07,
+      "loss": 0.1035,
+      "step": 1478
+    },
+    {
+      "epoch": 0.3361363636363636,
+      "grad_norm": 2.673817116525999,
+      "learning_rate": 9.331099033430935e-07,
+      "loss": 0.159,
+      "step": 1479
+    },
+    {
+      "epoch": 0.33636363636363636,
+      "grad_norm": 2.0398041906418483,
+      "learning_rate": 9.327215690605403e-07,
+      "loss": 0.1537,
+      "step": 1480
+    },
+    {
+      "epoch": 0.3365909090909091,
+      "grad_norm": 4.870275716192335,
+      "learning_rate": 9.323330779035025e-07,
+      "loss": 0.1186,
+      "step": 1481
+    },
+    {
+      "epoch": 0.3368181818181818,
+      "grad_norm": 4.283644169180116,
+      "learning_rate": 9.319444300700301e-07,
+      "loss": 0.1174,
+      "step": 1482
+    },
+    {
+      "epoch": 0.33704545454545454,
+      "grad_norm": 1.740126833905034,
+      "learning_rate": 9.315556257582536e-07,
+      "loss": 0.0897,
+      "step": 1483
+    },
+    {
+      "epoch": 0.3372727272727273,
+      "grad_norm": 4.792583237555445,
+      "learning_rate": 9.311666651663831e-07,
+      "loss": 0.1634,
+      "step": 1484
+    },
+    {
+      "epoch": 0.3375,
+      "grad_norm": 1.7394416073621668,
+      "learning_rate": 9.307775484927078e-07,
+      "loss": 0.0947,
+      "step": 1485
+    },
+    {
+      "epoch": 0.3377272727272727,
+      "grad_norm": 1.1449260116183348,
+      "learning_rate": 9.30388275935597e-07,
+      "loss": 0.052,
+      "step": 1486
+    },
+    {
+      "epoch": 0.33795454545454545,
+      "grad_norm": 2.96831216805227,
+      "learning_rate": 9.299988476934995e-07,
+      "loss": 0.1977,
+      "step": 1487
+    },
+    {
+      "epoch": 0.3381818181818182,
+      "grad_norm": 2.986383903263666,
+      "learning_rate": 9.29609263964943e-07,
+      "loss": 0.1927,
+      "step": 1488
+    },
+    {
+      "epoch": 0.3384090909090909,
+      "grad_norm": 3.5313837473032796,
+      "learning_rate": 9.292195249485353e-07,
+      "loss": 0.1266,
+      "step": 1489
+    },
+    {
+      "epoch": 0.3386363636363636,
+      "grad_norm": 2.31027374094743,
+      "learning_rate": 9.288296308429626e-07,
+      "loss": 0.1463,
+      "step": 1490
+    },
+    {
+      "epoch": 0.33886363636363637,
+      "grad_norm": 1.4987471476983172,
+      "learning_rate": 9.284395818469902e-07,
+      "loss": 0.0908,
+      "step": 1491
+    },
+    {
+      "epoch": 0.3390909090909091,
+      "grad_norm": 3.14539102951661,
+      "learning_rate": 9.280493781594628e-07,
+      "loss": 0.285,
+      "step": 1492
+    },
+    {
+      "epoch": 0.3393181818181818,
+      "grad_norm": 2.733655294344683,
+      "learning_rate": 9.276590199793037e-07,
+      "loss": 0.1276,
+      "step": 1493
+    },
+    {
+      "epoch": 0.33954545454545454,
+      "grad_norm": 2.2685551433784727,
+      "learning_rate": 9.272685075055152e-07,
+      "loss": 0.1755,
+      "step": 1494
+    },
+    {
+      "epoch": 0.3397727272727273,
+      "grad_norm": 2.212114513225997,
+      "learning_rate": 9.268778409371778e-07,
+      "loss": 0.1475,
+      "step": 1495
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.374881629644495,
+      "learning_rate": 9.264870204734508e-07,
+      "loss": 0.1148,
+      "step": 1496
+    },
+    {
+      "epoch": 0.3402272727272727,
+      "grad_norm": 4.376261509578244,
+      "learning_rate": 9.260960463135721e-07,
+      "loss": 0.1144,
+      "step": 1497
+    },
+    {
+      "epoch": 0.34045454545454545,
+      "grad_norm": 2.5049590146533953,
+      "learning_rate": 9.257049186568579e-07,
+      "loss": 0.1482,
+      "step": 1498
+    },
+    {
+      "epoch": 0.3406818181818182,
+      "grad_norm": 4.784262375089888,
+      "learning_rate": 9.253136377027022e-07,
+      "loss": 0.2217,
+      "step": 1499
+    },
+    {
+      "epoch": 0.3409090909090909,
+      "grad_norm": 2.5496367302833063,
+      "learning_rate": 9.249222036505779e-07,
+      "loss": 0.1636,
+      "step": 1500
+    },
+    {
+      "epoch": 0.3411363636363636,
+      "grad_norm": 1.9422439062342065,
+      "learning_rate": 9.245306167000357e-07,
+      "loss": 0.1184,
+      "step": 1501
+    },
+    {
+      "epoch": 0.34136363636363637,
+      "grad_norm": 2.6903297594498583,
+      "learning_rate": 9.241388770507036e-07,
+      "loss": 0.1186,
+      "step": 1502
+    },
+    {
+      "epoch": 0.3415909090909091,
+      "grad_norm": 3.1839736985918505,
+      "learning_rate": 9.237469849022881e-07,
+      "loss": 0.118,
+      "step": 1503
+    },
+    {
+      "epoch": 0.3418181818181818,
+      "grad_norm": 4.4797500751673525,
+      "learning_rate": 9.233549404545737e-07,
+      "loss": 0.123,
+      "step": 1504
+    },
+    {
+      "epoch": 0.34204545454545454,
+      "grad_norm": 2.46329606153784,
+      "learning_rate": 9.229627439074219e-07,
+      "loss": 0.1031,
+      "step": 1505
+    },
+    {
+      "epoch": 0.3422727272727273,
+      "grad_norm": 2.0409085694548166,
+      "learning_rate": 9.225703954607718e-07,
+      "loss": 0.0718,
+      "step": 1506
+    },
+    {
+      "epoch": 0.3425,
+      "grad_norm": 1.2501642198436365,
+      "learning_rate": 9.221778953146405e-07,
+      "loss": 0.1439,
+      "step": 1507
+    },
+    {
+      "epoch": 0.3427272727272727,
+      "grad_norm": 1.7764551960318216,
+      "learning_rate": 9.217852436691216e-07,
+      "loss": 0.1232,
+      "step": 1508
+    },
+    {
+      "epoch": 0.34295454545454546,
+      "grad_norm": 3.258456374273045,
+      "learning_rate": 9.213924407243868e-07,
+      "loss": 0.1304,
+      "step": 1509
+    },
+    {
+      "epoch": 0.3431818181818182,
+      "grad_norm": 4.30500951412829,
+      "learning_rate": 9.209994866806846e-07,
+      "loss": 0.2003,
+      "step": 1510
+    },
+    {
+      "epoch": 0.3434090909090909,
+      "grad_norm": 1.1164174067708421,
+      "learning_rate": 9.206063817383399e-07,
+      "loss": 0.0856,
+      "step": 1511
+    },
+    {
+      "epoch": 0.34363636363636363,
+      "grad_norm": 1.9924219461789776,
+      "learning_rate": 9.202131260977555e-07,
+      "loss": 0.1307,
+      "step": 1512
+    },
+    {
+      "epoch": 0.34386363636363637,
+      "grad_norm": 3.4846109629634814,
+      "learning_rate": 9.198197199594105e-07,
+      "loss": 0.1453,
+      "step": 1513
+    },
+    {
+      "epoch": 0.3440909090909091,
+      "grad_norm": 2.2971948216997156,
+      "learning_rate": 9.194261635238609e-07,
+      "loss": 0.0944,
+      "step": 1514
+    },
+    {
+      "epoch": 0.3443181818181818,
+      "grad_norm": 3.5835959788319913,
+      "learning_rate": 9.190324569917393e-07,
+      "loss": 0.1739,
+      "step": 1515
+    },
+    {
+      "epoch": 0.34454545454545454,
+      "grad_norm": 1.6266958381685928,
+      "learning_rate": 9.186386005637546e-07,
+      "loss": 0.1015,
+      "step": 1516
+    },
+    {
+      "epoch": 0.3447727272727273,
+      "grad_norm": 3.433566765086783,
+      "learning_rate": 9.182445944406923e-07,
+      "loss": 0.1015,
+      "step": 1517
+    },
+    {
+      "epoch": 0.345,
+      "grad_norm": 3.1943885625419557,
+      "learning_rate": 9.178504388234143e-07,
+      "loss": 0.0911,
+      "step": 1518
+    },
+    {
+      "epoch": 0.3452272727272727,
+      "grad_norm": 1.8623007014809139,
+      "learning_rate": 9.174561339128584e-07,
+      "loss": 0.0947,
+      "step": 1519
+    },
+    {
+      "epoch": 0.34545454545454546,
+      "grad_norm": 2.45881384467966,
+      "learning_rate": 9.170616799100391e-07,
+      "loss": 0.1159,
+      "step": 1520
+    },
+    {
+      "epoch": 0.3456818181818182,
+      "grad_norm": 2.0569232089673686,
+      "learning_rate": 9.166670770160461e-07,
+      "loss": 0.0858,
+      "step": 1521
+    },
+    {
+      "epoch": 0.3459090909090909,
+      "grad_norm": 2.1949255432352652,
+      "learning_rate": 9.162723254320458e-07,
+      "loss": 0.159,
+      "step": 1522
+    },
+    {
+      "epoch": 0.34613636363636363,
+      "grad_norm": 3.2929828147494353,
+      "learning_rate": 9.158774253592798e-07,
+      "loss": 0.1483,
+      "step": 1523
+    },
+    {
+      "epoch": 0.3463636363636364,
+      "grad_norm": 2.424574880131968,
+      "learning_rate": 9.154823769990656e-07,
+      "loss": 0.1151,
+      "step": 1524
+    },
+    {
+      "epoch": 0.3465909090909091,
+      "grad_norm": 2.7448446561359137,
+      "learning_rate": 9.150871805527965e-07,
+      "loss": 0.1173,
+      "step": 1525
+    },
+    {
+      "epoch": 0.3468181818181818,
+      "grad_norm": 1.851380451603946,
+      "learning_rate": 9.146918362219409e-07,
+      "loss": 0.0968,
+      "step": 1526
+    },
+    {
+      "epoch": 0.34704545454545455,
+      "grad_norm": 4.215004507179443,
+      "learning_rate": 9.142963442080432e-07,
+      "loss": 0.1695,
+      "step": 1527
+    },
+    {
+      "epoch": 0.3472727272727273,
+      "grad_norm": 2.4193696443367605,
+      "learning_rate": 9.139007047127222e-07,
+      "loss": 0.1401,
+      "step": 1528
+    },
+    {
+      "epoch": 0.3475,
+      "grad_norm": 5.841359569333322,
+      "learning_rate": 9.135049179376726e-07,
+      "loss": 0.1581,
+      "step": 1529
+    },
+    {
+      "epoch": 0.3477272727272727,
+      "grad_norm": 3.4126577940582563,
+      "learning_rate": 9.131089840846641e-07,
+      "loss": 0.1562,
+      "step": 1530
+    },
+    {
+      "epoch": 0.34795454545454546,
+      "grad_norm": 1.7132066087282587,
+      "learning_rate": 9.12712903355541e-07,
+      "loss": 0.084,
+      "step": 1531
+    },
+    {
+      "epoch": 0.3481818181818182,
+      "grad_norm": 2.299787573940957,
+      "learning_rate": 9.123166759522229e-07,
+      "loss": 0.1801,
+      "step": 1532
+    },
+    {
+      "epoch": 0.3484090909090909,
+      "grad_norm": 3.7782049538053797,
+      "learning_rate": 9.119203020767038e-07,
+      "loss": 0.1525,
+      "step": 1533
+    },
+    {
+      "epoch": 0.34863636363636363,
+      "grad_norm": 3.3508101919451576,
+      "learning_rate": 9.11523781931053e-07,
+      "loss": 0.0845,
+      "step": 1534
+    },
+    {
+      "epoch": 0.3488636363636364,
+      "grad_norm": 4.910189706185989,
+      "learning_rate": 9.111271157174131e-07,
+      "loss": 0.2347,
+      "step": 1535
+    },
+    {
+      "epoch": 0.3490909090909091,
+      "grad_norm": 2.119087389297756,
+      "learning_rate": 9.107303036380028e-07,
+      "loss": 0.1288,
+      "step": 1536
+    },
+    {
+      "epoch": 0.3493181818181818,
+      "grad_norm": 3.423214921912302,
+      "learning_rate": 9.103333458951142e-07,
+      "loss": 0.1523,
+      "step": 1537
+    },
+    {
+      "epoch": 0.34954545454545455,
+      "grad_norm": 2.237359504432335,
+      "learning_rate": 9.099362426911135e-07,
+      "loss": 0.0892,
+      "step": 1538
+    },
+    {
+      "epoch": 0.3497727272727273,
+      "grad_norm": 3.065470642745802,
+      "learning_rate": 9.095389942284416e-07,
+      "loss": 0.0995,
+      "step": 1539
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.965039988464307,
+      "learning_rate": 9.091416007096131e-07,
+      "loss": 0.246,
+      "step": 1540
+    },
+    {
+      "epoch": 0.3502272727272727,
+      "grad_norm": 2.281901664398071,
+      "learning_rate": 9.087440623372168e-07,
+      "loss": 0.1234,
+      "step": 1541
+    },
+    {
+      "epoch": 0.35045454545454546,
+      "grad_norm": 2.200493307552751,
+      "learning_rate": 9.083463793139152e-07,
+      "loss": 0.0834,
+      "step": 1542
+    },
+    {
+      "epoch": 0.3506818181818182,
+      "grad_norm": 2.1836253123403653,
+      "learning_rate": 9.079485518424444e-07,
+      "loss": 0.121,
+      "step": 1543
+    },
+    {
+      "epoch": 0.3509090909090909,
+      "grad_norm": 4.445518310304057,
+      "learning_rate": 9.075505801256145e-07,
+      "loss": 0.1654,
+      "step": 1544
+    },
+    {
+      "epoch": 0.35113636363636364,
+      "grad_norm": 1.431721430248671,
+      "learning_rate": 9.071524643663089e-07,
+      "loss": 0.0864,
+      "step": 1545
+    },
+    {
+      "epoch": 0.3513636363636364,
+      "grad_norm": 2.524189556571719,
+      "learning_rate": 9.067542047674843e-07,
+      "loss": 0.1567,
+      "step": 1546
+    },
+    {
+      "epoch": 0.35159090909090907,
+      "grad_norm": 3.4842046736395837,
+      "learning_rate": 9.063558015321711e-07,
+      "loss": 0.123,
+      "step": 1547
+    },
+    {
+      "epoch": 0.3518181818181818,
+      "grad_norm": 3.8492862602239852,
+      "learning_rate": 9.059572548634725e-07,
+      "loss": 0.111,
+      "step": 1548
+    },
+    {
+      "epoch": 0.35204545454545455,
+      "grad_norm": 2.2566686319645104,
+      "learning_rate": 9.055585649645651e-07,
+      "loss": 0.177,
+      "step": 1549
+    },
+    {
+      "epoch": 0.3522727272727273,
+      "grad_norm": 3.6199751393531376,
+      "learning_rate": 9.051597320386986e-07,
+      "loss": 0.211,
+      "step": 1550
+    },
+    {
+      "epoch": 0.3525,
+      "grad_norm": 3.0199939357681425,
+      "learning_rate": 9.047607562891954e-07,
+      "loss": 0.093,
+      "step": 1551
+    },
+    {
+      "epoch": 0.3527272727272727,
+      "grad_norm": 2.54030040487978,
+      "learning_rate": 9.043616379194508e-07,
+      "loss": 0.2716,
+      "step": 1552
+    },
+    {
+      "epoch": 0.35295454545454547,
+      "grad_norm": 2.4405452643108596,
+      "learning_rate": 9.039623771329326e-07,
+      "loss": 0.1424,
+      "step": 1553
+    },
+    {
+      "epoch": 0.3531818181818182,
+      "grad_norm": 2.8059224078642977,
+      "learning_rate": 9.035629741331817e-07,
+      "loss": 0.1459,
+      "step": 1554
+    },
+    {
+      "epoch": 0.3534090909090909,
+      "grad_norm": 3.3109178670292283,
+      "learning_rate": 9.031634291238109e-07,
+      "loss": 0.1155,
+      "step": 1555
+    },
+    {
+      "epoch": 0.35363636363636364,
+      "grad_norm": 2.642198100932542,
+      "learning_rate": 9.027637423085061e-07,
+      "loss": 0.2138,
+      "step": 1556
+    },
+    {
+      "epoch": 0.3538636363636364,
+      "grad_norm": 3.526511367950689,
+      "learning_rate": 9.023639138910246e-07,
+      "loss": 0.1561,
+      "step": 1557
+    },
+    {
+      "epoch": 0.35409090909090907,
+      "grad_norm": 5.026063946990607,
+      "learning_rate": 9.019639440751964e-07,
+      "loss": 0.2878,
+      "step": 1558
+    },
+    {
+      "epoch": 0.3543181818181818,
+      "grad_norm": 2.585943590801081,
+      "learning_rate": 9.015638330649239e-07,
+      "loss": 0.1229,
+      "step": 1559
+    },
+    {
+      "epoch": 0.35454545454545455,
+      "grad_norm": 3.370290060509222,
+      "learning_rate": 9.011635810641809e-07,
+      "loss": 0.1231,
+      "step": 1560
+    },
+    {
+      "epoch": 0.3547727272727273,
+      "grad_norm": 4.273267872770528,
+      "learning_rate": 9.007631882770133e-07,
+      "loss": 0.1647,
+      "step": 1561
+    },
+    {
+      "epoch": 0.355,
+      "grad_norm": 2.593065590384995,
+      "learning_rate": 9.00362654907539e-07,
+      "loss": 0.0881,
+      "step": 1562
+    },
+    {
+      "epoch": 0.3552272727272727,
+      "grad_norm": 2.154381491470544,
+      "learning_rate": 8.99961981159947e-07,
+      "loss": 0.0663,
+      "step": 1563
+    },
+    {
+      "epoch": 0.35545454545454547,
+      "grad_norm": 2.805709142167706,
+      "learning_rate": 8.995611672384985e-07,
+      "loss": 0.1012,
+      "step": 1564
+    },
+    {
+      "epoch": 0.35568181818181815,
+      "grad_norm": 1.7875072076795673,
+      "learning_rate": 8.991602133475258e-07,
+      "loss": 0.1281,
+      "step": 1565
+    },
+    {
+      "epoch": 0.3559090909090909,
+      "grad_norm": 2.298421060774618,
+      "learning_rate": 8.987591196914325e-07,
+      "loss": 0.0698,
+      "step": 1566
+    },
+    {
+      "epoch": 0.35613636363636364,
+      "grad_norm": 2.0886307204353955,
+      "learning_rate": 8.983578864746935e-07,
+      "loss": 0.1487,
+      "step": 1567
+    },
+    {
+      "epoch": 0.3563636363636364,
+      "grad_norm": 2.5160003038507495,
+      "learning_rate": 8.979565139018554e-07,
+      "loss": 0.1148,
+      "step": 1568
+    },
+    {
+      "epoch": 0.35659090909090907,
+      "grad_norm": 3.195311590033761,
+      "learning_rate": 8.975550021775346e-07,
+      "loss": 0.1001,
+      "step": 1569
+    },
+    {
+      "epoch": 0.3568181818181818,
+      "grad_norm": 4.354843341670446,
+      "learning_rate": 8.9715335150642e-07,
+      "loss": 0.1544,
+      "step": 1570
+    },
+    {
+      "epoch": 0.35704545454545455,
+      "grad_norm": 1.8260772245370762,
+      "learning_rate": 8.967515620932702e-07,
+      "loss": 0.1342,
+      "step": 1571
+    },
+    {
+      "epoch": 0.3572727272727273,
+      "grad_norm": 2.9127008441369284,
+      "learning_rate": 8.963496341429145e-07,
+      "loss": 0.1817,
+      "step": 1572
+    },
+    {
+      "epoch": 0.3575,
+      "grad_norm": 2.9418931441868685,
+      "learning_rate": 8.959475678602539e-07,
+      "loss": 0.2225,
+      "step": 1573
+    },
+    {
+      "epoch": 0.3577272727272727,
+      "grad_norm": 2.237067718160595,
+      "learning_rate": 8.95545363450259e-07,
+      "loss": 0.0974,
+      "step": 1574
+    },
+    {
+      "epoch": 0.35795454545454547,
+      "grad_norm": 2.8364680870309518,
+      "learning_rate": 8.951430211179704e-07,
+      "loss": 0.135,
+      "step": 1575
+    },
+    {
+      "epoch": 0.35818181818181816,
+      "grad_norm": 5.306388249428803,
+      "learning_rate": 8.947405410685004e-07,
+      "loss": 0.1691,
+      "step": 1576
+    },
+    {
+      "epoch": 0.3584090909090909,
+      "grad_norm": 3.7918932127270457,
+      "learning_rate": 8.943379235070307e-07,
+      "loss": 0.1809,
+      "step": 1577
+    },
+    {
+      "epoch": 0.35863636363636364,
+      "grad_norm": 2.6437181223294703,
+      "learning_rate": 8.939351686388127e-07,
+      "loss": 0.1946,
+      "step": 1578
+    },
+    {
+      "epoch": 0.3588636363636364,
+      "grad_norm": 3.1327820206290995,
+      "learning_rate": 8.935322766691684e-07,
+      "loss": 0.1056,
+      "step": 1579
+    },
+    {
+      "epoch": 0.35909090909090907,
+      "grad_norm": 1.7285360244190917,
+      "learning_rate": 8.931292478034899e-07,
+      "loss": 0.1538,
+      "step": 1580
+    },
+    {
+      "epoch": 0.3593181818181818,
+      "grad_norm": 4.067816964904866,
+      "learning_rate": 8.927260822472384e-07,
+      "loss": 0.181,
+      "step": 1581
+    },
+    {
+      "epoch": 0.35954545454545456,
+      "grad_norm": 4.218485290493098,
+      "learning_rate": 8.92322780205945e-07,
+      "loss": 0.1981,
+      "step": 1582
+    },
+    {
+      "epoch": 0.3597727272727273,
+      "grad_norm": 3.0461330680194862,
+      "learning_rate": 8.91919341885211e-07,
+      "loss": 0.1025,
+      "step": 1583
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.260375500472309,
+      "learning_rate": 8.915157674907062e-07,
+      "loss": 0.1377,
+      "step": 1584
+    },
+    {
+      "epoch": 0.36022727272727273,
+      "grad_norm": 2.286311150377566,
+      "learning_rate": 8.911120572281705e-07,
+      "loss": 0.1921,
+      "step": 1585
+    },
+    {
+      "epoch": 0.36045454545454547,
+      "grad_norm": 2.683520296771183,
+      "learning_rate": 8.907082113034126e-07,
+      "loss": 0.266,
+      "step": 1586
+    },
+    {
+      "epoch": 0.36068181818181816,
+      "grad_norm": 2.215908112988123,
+      "learning_rate": 8.903042299223108e-07,
+      "loss": 0.0657,
+      "step": 1587
+    },
+    {
+      "epoch": 0.3609090909090909,
+      "grad_norm": 3.9455292466189222,
+      "learning_rate": 8.89900113290812e-07,
+      "loss": 0.1192,
+      "step": 1588
+    },
+    {
+      "epoch": 0.36113636363636364,
+      "grad_norm": 3.4340361918008946,
+      "learning_rate": 8.894958616149324e-07,
+      "loss": 0.1959,
+      "step": 1589
+    },
+    {
+      "epoch": 0.3613636363636364,
+      "grad_norm": 3.1300866470275364,
+      "learning_rate": 8.890914751007568e-07,
+      "loss": 0.1714,
+      "step": 1590
+    },
+    {
+      "epoch": 0.3615909090909091,
+      "grad_norm": 2.826667075670108,
+      "learning_rate": 8.88686953954439e-07,
+      "loss": 0.1495,
+      "step": 1591
+    },
+    {
+      "epoch": 0.3618181818181818,
+      "grad_norm": 1.788276559668391,
+      "learning_rate": 8.882822983822014e-07,
+      "loss": 0.1341,
+      "step": 1592
+    },
+    {
+      "epoch": 0.36204545454545456,
+      "grad_norm": 3.3355374682276406,
+      "learning_rate": 8.878775085903344e-07,
+      "loss": 0.1812,
+      "step": 1593
+    },
+    {
+      "epoch": 0.36227272727272725,
+      "grad_norm": 3.1844095125648906,
+      "learning_rate": 8.874725847851976e-07,
+      "loss": 0.1838,
+      "step": 1594
+    },
+    {
+      "epoch": 0.3625,
+      "grad_norm": 2.385369951707618,
+      "learning_rate": 8.870675271732184e-07,
+      "loss": 0.1847,
+      "step": 1595
+    },
+    {
+      "epoch": 0.36272727272727273,
+      "grad_norm": 2.485418076760724,
+      "learning_rate": 8.866623359608927e-07,
+      "loss": 0.1493,
+      "step": 1596
+    },
+    {
+      "epoch": 0.3629545454545455,
+      "grad_norm": 3.258146926065343,
+      "learning_rate": 8.862570113547843e-07,
+      "loss": 0.2296,
+      "step": 1597
+    },
+    {
+      "epoch": 0.36318181818181816,
+      "grad_norm": 3.603183806055147,
+      "learning_rate": 8.858515535615249e-07,
+      "loss": 0.2462,
+      "step": 1598
+    },
+    {
+      "epoch": 0.3634090909090909,
+      "grad_norm": 3.560995324618087,
+      "learning_rate": 8.854459627878146e-07,
+      "loss": 0.1638,
+      "step": 1599
+    },
+    {
+      "epoch": 0.36363636363636365,
+      "grad_norm": 2.195892220035855,
+      "learning_rate": 8.85040239240421e-07,
+      "loss": 0.0671,
+      "step": 1600
+    },
+    {
+      "epoch": 0.3638636363636364,
+      "grad_norm": 2.92533121394955,
+      "learning_rate": 8.846343831261792e-07,
+      "loss": 0.1114,
+      "step": 1601
+    },
+    {
+      "epoch": 0.3640909090909091,
+      "grad_norm": 1.9308712336608727,
+      "learning_rate": 8.842283946519918e-07,
+      "loss": 0.1881,
+      "step": 1602
+    },
+    {
+      "epoch": 0.3643181818181818,
+      "grad_norm": 2.170129534153303,
+      "learning_rate": 8.838222740248297e-07,
+      "loss": 0.1074,
+      "step": 1603
+    },
+    {
+      "epoch": 0.36454545454545456,
+      "grad_norm": 2.7674856401218957,
+      "learning_rate": 8.834160214517302e-07,
+      "loss": 0.0908,
+      "step": 1604
+    },
+    {
+      "epoch": 0.36477272727272725,
+      "grad_norm": 3.191857520151726,
+      "learning_rate": 8.830096371397982e-07,
+      "loss": 0.1276,
+      "step": 1605
+    },
+    {
+      "epoch": 0.365,
+      "grad_norm": 1.779275891155299,
+      "learning_rate": 8.826031212962064e-07,
+      "loss": 0.1094,
+      "step": 1606
+    },
+    {
+      "epoch": 0.36522727272727273,
+      "grad_norm": 2.992537208317427,
+      "learning_rate": 8.821964741281932e-07,
+      "loss": 0.092,
+      "step": 1607
+    },
+    {
+      "epoch": 0.3654545454545455,
+      "grad_norm": 5.6188776644396246,
+      "learning_rate": 8.81789695843065e-07,
+      "loss": 0.2093,
+      "step": 1608
+    },
+    {
+      "epoch": 0.36568181818181816,
+      "grad_norm": 2.9385552603664746,
+      "learning_rate": 8.81382786648195e-07,
+      "loss": 0.1854,
+      "step": 1609
+    },
+    {
+      "epoch": 0.3659090909090909,
+      "grad_norm": 3.3347569154817847,
+      "learning_rate": 8.809757467510227e-07,
+      "loss": 0.1999,
+      "step": 1610
+    },
+    {
+      "epoch": 0.36613636363636365,
+      "grad_norm": 4.94557098840462,
+      "learning_rate": 8.805685763590543e-07,
+      "loss": 0.1531,
+      "step": 1611
+    },
+    {
+      "epoch": 0.3663636363636364,
+      "grad_norm": 2.5551537424824193,
+      "learning_rate": 8.80161275679863e-07,
+      "loss": 0.2244,
+      "step": 1612
+    },
+    {
+      "epoch": 0.3665909090909091,
+      "grad_norm": 1.2476047722881767,
+      "learning_rate": 8.797538449210878e-07,
+      "loss": 0.1161,
+      "step": 1613
+    },
+    {
+      "epoch": 0.3668181818181818,
+      "grad_norm": 2.947747940022824,
+      "learning_rate": 8.793462842904344e-07,
+      "loss": 0.1295,
+      "step": 1614
+    },
+    {
+      "epoch": 0.36704545454545456,
+      "grad_norm": 2.506414728718674,
+      "learning_rate": 8.789385939956747e-07,
+      "loss": 0.1248,
+      "step": 1615
+    },
+    {
+      "epoch": 0.36727272727272725,
+      "grad_norm": 3.1407625728749493,
+      "learning_rate": 8.785307742446464e-07,
+      "loss": 0.1037,
+      "step": 1616
+    },
+    {
+      "epoch": 0.3675,
+      "grad_norm": 2.3778699855320387,
+      "learning_rate": 8.781228252452534e-07,
+      "loss": 0.0977,
+      "step": 1617
+    },
+    {
+      "epoch": 0.36772727272727274,
+      "grad_norm": 3.375645408428094,
+      "learning_rate": 8.777147472054657e-07,
+      "loss": 0.1332,
+      "step": 1618
+    },
+    {
+      "epoch": 0.3679545454545455,
+      "grad_norm": 2.193054702243968,
+      "learning_rate": 8.773065403333186e-07,
+      "loss": 0.1563,
+      "step": 1619
+    },
+    {
+      "epoch": 0.36818181818181817,
+      "grad_norm": 2.863132560150171,
+      "learning_rate": 8.768982048369136e-07,
+      "loss": 0.0786,
+      "step": 1620
+    },
+    {
+      "epoch": 0.3684090909090909,
+      "grad_norm": 4.134267353704429,
+      "learning_rate": 8.764897409244176e-07,
+      "loss": 0.1875,
+      "step": 1621
+    },
+    {
+      "epoch": 0.36863636363636365,
+      "grad_norm": 3.839569975064351,
+      "learning_rate": 8.760811488040624e-07,
+      "loss": 0.165,
+      "step": 1622
+    },
+    {
+      "epoch": 0.36886363636363634,
+      "grad_norm": 3.917386598028515,
+      "learning_rate": 8.756724286841458e-07,
+      "loss": 0.1297,
+      "step": 1623
+    },
+    {
+      "epoch": 0.3690909090909091,
+      "grad_norm": 2.8636915223793697,
+      "learning_rate": 8.752635807730314e-07,
+      "loss": 0.1514,
+      "step": 1624
+    },
+    {
+      "epoch": 0.3693181818181818,
+      "grad_norm": 3.8042443633976983,
+      "learning_rate": 8.748546052791462e-07,
+      "loss": 0.1498,
+      "step": 1625
+    },
+    {
+      "epoch": 0.36954545454545457,
+      "grad_norm": 1.7607558212033378,
+      "learning_rate": 8.744455024109838e-07,
+      "loss": 0.0924,
+      "step": 1626
+    },
+    {
+      "epoch": 0.36977272727272725,
+      "grad_norm": 1.9534594732479755,
+      "learning_rate": 8.740362723771024e-07,
+      "loss": 0.1389,
+      "step": 1627
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.130552427723555,
+      "learning_rate": 8.736269153861246e-07,
+      "loss": 0.0697,
+      "step": 1628
+    },
+    {
+      "epoch": 0.37022727272727274,
+      "grad_norm": 1.761947562569739,
+      "learning_rate": 8.732174316467379e-07,
+      "loss": 0.1281,
+      "step": 1629
+    },
+    {
+      "epoch": 0.3704545454545455,
+      "grad_norm": 2.4394805600758667,
+      "learning_rate": 8.728078213676948e-07,
+      "loss": 0.1508,
+      "step": 1630
+    },
+    {
+      "epoch": 0.37068181818181817,
+      "grad_norm": 4.370582455782946,
+      "learning_rate": 8.723980847578116e-07,
+      "loss": 0.1469,
+      "step": 1631
+    },
+    {
+      "epoch": 0.3709090909090909,
+      "grad_norm": 4.0837561552443455,
+      "learning_rate": 8.719882220259697e-07,
+      "loss": 0.1549,
+      "step": 1632
+    },
+    {
+      "epoch": 0.37113636363636365,
+      "grad_norm": 3.558998071605903,
+      "learning_rate": 8.715782333811143e-07,
+      "loss": 0.1167,
+      "step": 1633
+    },
+    {
+      "epoch": 0.37136363636363634,
+      "grad_norm": 6.949943203042626,
+      "learning_rate": 8.711681190322552e-07,
+      "loss": 0.2275,
+      "step": 1634
+    },
+    {
+      "epoch": 0.3715909090909091,
+      "grad_norm": 1.4624796892950473,
+      "learning_rate": 8.707578791884659e-07,
+      "loss": 0.0747,
+      "step": 1635
+    },
+    {
+      "epoch": 0.3718181818181818,
+      "grad_norm": 2.212704573653537,
+      "learning_rate": 8.703475140588841e-07,
+      "loss": 0.1285,
+      "step": 1636
+    },
+    {
+      "epoch": 0.37204545454545457,
+      "grad_norm": 4.534781145673639,
+      "learning_rate": 8.699370238527112e-07,
+      "loss": 0.2099,
+      "step": 1637
+    },
+    {
+      "epoch": 0.37227272727272726,
+      "grad_norm": 2.67697317679986,
+      "learning_rate": 8.695264087792127e-07,
+      "loss": 0.106,
+      "step": 1638
+    },
+    {
+      "epoch": 0.3725,
+      "grad_norm": 3.4567416881218413,
+      "learning_rate": 8.691156690477174e-07,
+      "loss": 0.0987,
+      "step": 1639
+    },
+    {
+      "epoch": 0.37272727272727274,
+      "grad_norm": 2.9912606081360997,
+      "learning_rate": 8.687048048676178e-07,
+      "loss": 0.1505,
+      "step": 1640
+    },
+    {
+      "epoch": 0.3729545454545454,
+      "grad_norm": 2.0171741784043626,
+      "learning_rate": 8.682938164483698e-07,
+      "loss": 0.1053,
+      "step": 1641
+    },
+    {
+      "epoch": 0.37318181818181817,
+      "grad_norm": 2.563725154718749,
+      "learning_rate": 8.678827039994928e-07,
+      "loss": 0.1118,
+      "step": 1642
+    },
+    {
+      "epoch": 0.3734090909090909,
+      "grad_norm": 3.42056148099321,
+      "learning_rate": 8.67471467730569e-07,
+      "loss": 0.2544,
+      "step": 1643
+    },
+    {
+      "epoch": 0.37363636363636366,
+      "grad_norm": 3.128123816379985,
+      "learning_rate": 8.670601078512443e-07,
+      "loss": 0.1302,
+      "step": 1644
+    },
+    {
+      "epoch": 0.37386363636363634,
+      "grad_norm": 2.8618809290040454,
+      "learning_rate": 8.666486245712271e-07,
+      "loss": 0.1568,
+      "step": 1645
+    },
+    {
+      "epoch": 0.3740909090909091,
+      "grad_norm": 2.9165714458822922,
+      "learning_rate": 8.662370181002892e-07,
+      "loss": 0.12,
+      "step": 1646
+    },
+    {
+      "epoch": 0.3743181818181818,
+      "grad_norm": 5.2119111311782795,
+      "learning_rate": 8.658252886482649e-07,
+      "loss": 0.1867,
+      "step": 1647
+    },
+    {
+      "epoch": 0.37454545454545457,
+      "grad_norm": 2.9812244982465588,
+      "learning_rate": 8.654134364250509e-07,
+      "loss": 0.1048,
+      "step": 1648
+    },
+    {
+      "epoch": 0.37477272727272726,
+      "grad_norm": 2.6679254796705547,
+      "learning_rate": 8.650014616406072e-07,
+      "loss": 0.1158,
+      "step": 1649
+    },
+    {
+      "epoch": 0.375,
+      "grad_norm": 2.998883257706244,
+      "learning_rate": 8.64589364504956e-07,
+      "loss": 0.145,
+      "step": 1650
+    },
+    {
+      "epoch": 0.37522727272727274,
+      "grad_norm": 3.247232576738985,
+      "learning_rate": 8.641771452281812e-07,
+      "loss": 0.1063,
+      "step": 1651
+    },
+    {
+      "epoch": 0.37545454545454543,
+      "grad_norm": 5.251379563341244,
+      "learning_rate": 8.637648040204299e-07,
+      "loss": 0.1542,
+      "step": 1652
+    },
+    {
+      "epoch": 0.3756818181818182,
+      "grad_norm": 3.1755208063459675,
+      "learning_rate": 8.633523410919114e-07,
+      "loss": 0.1103,
+      "step": 1653
+    },
+    {
+      "epoch": 0.3759090909090909,
+      "grad_norm": 3.6179192136668186,
+      "learning_rate": 8.62939756652896e-07,
+      "loss": 0.2553,
+      "step": 1654
+    },
+    {
+      "epoch": 0.37613636363636366,
+      "grad_norm": 2.9012368697473927,
+      "learning_rate": 8.625270509137168e-07,
+      "loss": 0.1621,
+      "step": 1655
+    },
+    {
+      "epoch": 0.37636363636363634,
+      "grad_norm": 3.5140373362722825,
+      "learning_rate": 8.621142240847688e-07,
+      "loss": 0.2708,
+      "step": 1656
+    },
+    {
+      "epoch": 0.3765909090909091,
+      "grad_norm": 4.731526353564109,
+      "learning_rate": 8.617012763765084e-07,
+      "loss": 0.1699,
+      "step": 1657
+    },
+    {
+      "epoch": 0.37681818181818183,
+      "grad_norm": 1.0252718968902335,
+      "learning_rate": 8.612882079994534e-07,
+      "loss": 0.0974,
+      "step": 1658
+    },
+    {
+      "epoch": 0.3770454545454546,
+      "grad_norm": 2.850600154059715,
+      "learning_rate": 8.608750191641835e-07,
+      "loss": 0.1701,
+      "step": 1659
+    },
+    {
+      "epoch": 0.37727272727272726,
+      "grad_norm": 2.6884403471388048,
+      "learning_rate": 8.604617100813399e-07,
+      "loss": 0.115,
+      "step": 1660
+    },
+    {
+      "epoch": 0.3775,
+      "grad_norm": 3.117997543003442,
+      "learning_rate": 8.600482809616249e-07,
+      "loss": 0.0987,
+      "step": 1661
+    },
+    {
+      "epoch": 0.37772727272727274,
+      "grad_norm": 2.100593311313672,
+      "learning_rate": 8.59634732015802e-07,
+      "loss": 0.0838,
+      "step": 1662
+    },
+    {
+      "epoch": 0.37795454545454543,
+      "grad_norm": 2.3504668568097564,
+      "learning_rate": 8.592210634546958e-07,
+      "loss": 0.1023,
+      "step": 1663
+    },
+    {
+      "epoch": 0.3781818181818182,
+      "grad_norm": 4.279788078216239,
+      "learning_rate": 8.588072754891917e-07,
+      "loss": 0.1922,
+      "step": 1664
+    },
+    {
+      "epoch": 0.3784090909090909,
+      "grad_norm": 2.4169921973465693,
+      "learning_rate": 8.583933683302364e-07,
+      "loss": 0.1502,
+      "step": 1665
+    },
+    {
+      "epoch": 0.37863636363636366,
+      "grad_norm": 3.0616857616483224,
+      "learning_rate": 8.579793421888369e-07,
+      "loss": 0.1407,
+      "step": 1666
+    },
+    {
+      "epoch": 0.37886363636363635,
+      "grad_norm": 2.5728955591165703,
+      "learning_rate": 8.575651972760613e-07,
+      "loss": 0.1645,
+      "step": 1667
+    },
+    {
+      "epoch": 0.3790909090909091,
+      "grad_norm": 1.9771617604062897,
+      "learning_rate": 8.57150933803038e-07,
+      "loss": 0.133,
+      "step": 1668
+    },
+    {
+      "epoch": 0.37931818181818183,
+      "grad_norm": 2.4037975302778256,
+      "learning_rate": 8.567365519809556e-07,
+      "loss": 0.1006,
+      "step": 1669
+    },
+    {
+      "epoch": 0.3795454545454545,
+      "grad_norm": 2.716602953198209,
+      "learning_rate": 8.563220520210636e-07,
+      "loss": 0.1205,
+      "step": 1670
+    },
+    {
+      "epoch": 0.37977272727272726,
+      "grad_norm": 3.6587808319853723,
+      "learning_rate": 8.559074341346713e-07,
+      "loss": 0.1296,
+      "step": 1671
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.1860437224449463,
+      "learning_rate": 8.554926985331481e-07,
+      "loss": 0.0802,
+      "step": 1672
+    },
+    {
+      "epoch": 0.38022727272727275,
+      "grad_norm": 2.0084198647700613,
+      "learning_rate": 8.550778454279239e-07,
+      "loss": 0.1505,
+      "step": 1673
+    },
+    {
+      "epoch": 0.38045454545454543,
+      "grad_norm": 1.4102952822037504,
+      "learning_rate": 8.546628750304877e-07,
+      "loss": 0.068,
+      "step": 1674
+    },
+    {
+      "epoch": 0.3806818181818182,
+      "grad_norm": 3.4863095693116906,
+      "learning_rate": 8.542477875523893e-07,
+      "loss": 0.109,
+      "step": 1675
+    },
+    {
+      "epoch": 0.3809090909090909,
+      "grad_norm": 3.0200915245295104,
+      "learning_rate": 8.538325832052373e-07,
+      "loss": 0.1294,
+      "step": 1676
+    },
+    {
+      "epoch": 0.38113636363636366,
+      "grad_norm": 1.9189334579171182,
+      "learning_rate": 8.534172622007e-07,
+      "loss": 0.1625,
+      "step": 1677
+    },
+    {
+      "epoch": 0.38136363636363635,
+      "grad_norm": 3.6255034237522707,
+      "learning_rate": 8.530018247505058e-07,
+      "loss": 0.1187,
+      "step": 1678
+    },
+    {
+      "epoch": 0.3815909090909091,
+      "grad_norm": 2.0614951570031548,
+      "learning_rate": 8.525862710664421e-07,
+      "loss": 0.0579,
+      "step": 1679
+    },
+    {
+      "epoch": 0.38181818181818183,
+      "grad_norm": 3.3518729646049006,
+      "learning_rate": 8.521706013603551e-07,
+      "loss": 0.1298,
+      "step": 1680
+    },
+    {
+      "epoch": 0.3820454545454545,
+      "grad_norm": 3.4956717820967955,
+      "learning_rate": 8.517548158441506e-07,
+      "loss": 0.1185,
+      "step": 1681
+    },
+    {
+      "epoch": 0.38227272727272726,
+      "grad_norm": 5.411974539743721,
+      "learning_rate": 8.51338914729794e-07,
+      "loss": 0.1622,
+      "step": 1682
+    },
+    {
+      "epoch": 0.3825,
+      "grad_norm": 2.3311483101799246,
+      "learning_rate": 8.509228982293083e-07,
+      "loss": 0.0946,
+      "step": 1683
+    },
+    {
+      "epoch": 0.38272727272727275,
+      "grad_norm": 2.950424330562554,
+      "learning_rate": 8.505067665547761e-07,
+      "loss": 0.1235,
+      "step": 1684
+    },
+    {
+      "epoch": 0.38295454545454544,
+      "grad_norm": 2.1311936682695847,
+      "learning_rate": 8.500905199183393e-07,
+      "loss": 0.1244,
+      "step": 1685
+    },
+    {
+      "epoch": 0.3831818181818182,
+      "grad_norm": 2.8695776245905686,
+      "learning_rate": 8.496741585321971e-07,
+      "loss": 0.0989,
+      "step": 1686
+    },
+    {
+      "epoch": 0.3834090909090909,
+      "grad_norm": 3.0220125381614933,
+      "learning_rate": 8.492576826086083e-07,
+      "loss": 0.1203,
+      "step": 1687
+    },
+    {
+      "epoch": 0.3836363636363636,
+      "grad_norm": 2.977288254895569,
+      "learning_rate": 8.488410923598893e-07,
+      "loss": 0.0967,
+      "step": 1688
+    },
+    {
+      "epoch": 0.38386363636363635,
+      "grad_norm": 2.054542953188505,
+      "learning_rate": 8.484243879984154e-07,
+      "loss": 0.1158,
+      "step": 1689
+    },
+    {
+      "epoch": 0.3840909090909091,
+      "grad_norm": 3.4701412557932216,
+      "learning_rate": 8.480075697366197e-07,
+      "loss": 0.233,
+      "step": 1690
+    },
+    {
+      "epoch": 0.38431818181818184,
+      "grad_norm": 2.589871468128496,
+      "learning_rate": 8.475906377869936e-07,
+      "loss": 0.077,
+      "step": 1691
+    },
+    {
+      "epoch": 0.3845454545454545,
+      "grad_norm": 3.5775687843338813,
+      "learning_rate": 8.471735923620861e-07,
+      "loss": 0.1599,
+      "step": 1692
+    },
+    {
+      "epoch": 0.38477272727272727,
+      "grad_norm": 1.9546790793489908,
+      "learning_rate": 8.467564336745046e-07,
+      "loss": 0.161,
+      "step": 1693
+    },
+    {
+      "epoch": 0.385,
+      "grad_norm": 3.7514081089397,
+      "learning_rate": 8.463391619369136e-07,
+      "loss": 0.1686,
+      "step": 1694
+    },
+    {
+      "epoch": 0.38522727272727275,
+      "grad_norm": 3.5238085041760936,
+      "learning_rate": 8.459217773620358e-07,
+      "loss": 0.16,
+      "step": 1695
+    },
+    {
+      "epoch": 0.38545454545454544,
+      "grad_norm": 2.8100281524678548,
+      "learning_rate": 8.45504280162651e-07,
+      "loss": 0.0869,
+      "step": 1696
+    },
+    {
+      "epoch": 0.3856818181818182,
+      "grad_norm": 2.913452041893612,
+      "learning_rate": 8.45086670551597e-07,
+      "loss": 0.1294,
+      "step": 1697
+    },
+    {
+      "epoch": 0.3859090909090909,
+      "grad_norm": 2.673596809206283,
+      "learning_rate": 8.44668948741768e-07,
+      "loss": 0.2498,
+      "step": 1698
+    },
+    {
+      "epoch": 0.3861363636363636,
+      "grad_norm": 2.4890022002784122,
+      "learning_rate": 8.442511149461159e-07,
+      "loss": 0.1381,
+      "step": 1699
+    },
+    {
+      "epoch": 0.38636363636363635,
+      "grad_norm": 1.7822678250741903,
+      "learning_rate": 8.438331693776504e-07,
+      "loss": 0.0737,
+      "step": 1700
+    },
+    {
+      "epoch": 0.3865909090909091,
+      "grad_norm": 2.7265445227573575,
+      "learning_rate": 8.434151122494366e-07,
+      "loss": 0.2178,
+      "step": 1701
+    },
+    {
+      "epoch": 0.38681818181818184,
+      "grad_norm": 3.765729372105945,
+      "learning_rate": 8.429969437745977e-07,
+      "loss": 0.1222,
+      "step": 1702
+    },
+    {
+      "epoch": 0.3870454545454545,
+      "grad_norm": 2.0622839775905697,
+      "learning_rate": 8.425786641663135e-07,
+      "loss": 0.1161,
+      "step": 1703
+    },
+    {
+      "epoch": 0.38727272727272727,
+      "grad_norm": 2.224974984769739,
+      "learning_rate": 8.421602736378201e-07,
+      "loss": 0.1017,
+      "step": 1704
+    },
+    {
+      "epoch": 0.3875,
+      "grad_norm": 3.5250532892711632,
+      "learning_rate": 8.417417724024101e-07,
+      "loss": 0.1404,
+      "step": 1705
+    },
+    {
+      "epoch": 0.38772727272727275,
+      "grad_norm": 3.6227501490656424,
+      "learning_rate": 8.413231606734332e-07,
+      "loss": 0.1006,
+      "step": 1706
+    },
+    {
+      "epoch": 0.38795454545454544,
+      "grad_norm": 2.3744953266533013,
+      "learning_rate": 8.409044386642945e-07,
+      "loss": 0.1769,
+      "step": 1707
+    },
+    {
+      "epoch": 0.3881818181818182,
+      "grad_norm": 3.394567994746479,
+      "learning_rate": 8.404856065884564e-07,
+      "loss": 0.161,
+      "step": 1708
+    },
+    {
+      "epoch": 0.3884090909090909,
+      "grad_norm": 2.731705137832566,
+      "learning_rate": 8.400666646594362e-07,
+      "loss": 0.1422,
+      "step": 1709
+    },
+    {
+      "epoch": 0.3886363636363636,
+      "grad_norm": 2.669042102801384,
+      "learning_rate": 8.39647613090808e-07,
+      "loss": 0.1986,
+      "step": 1710
+    },
+    {
+      "epoch": 0.38886363636363636,
+      "grad_norm": 4.827021030076677,
+      "learning_rate": 8.392284520962017e-07,
+      "loss": 0.176,
+      "step": 1711
+    },
+    {
+      "epoch": 0.3890909090909091,
+      "grad_norm": 1.9516090495083367,
+      "learning_rate": 8.388091818893029e-07,
+      "loss": 0.0786,
+      "step": 1712
+    },
+    {
+      "epoch": 0.38931818181818184,
+      "grad_norm": 1.5558661224750525,
+      "learning_rate": 8.383898026838527e-07,
+      "loss": 0.0982,
+      "step": 1713
+    },
+    {
+      "epoch": 0.38954545454545453,
+      "grad_norm": 3.510010089216994,
+      "learning_rate": 8.379703146936482e-07,
+      "loss": 0.1781,
+      "step": 1714
+    },
+    {
+      "epoch": 0.38977272727272727,
+      "grad_norm": 3.161756633966076,
+      "learning_rate": 8.375507181325415e-07,
+      "loss": 0.0891,
+      "step": 1715
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.223212391517564,
+      "learning_rate": 8.371310132144401e-07,
+      "loss": 0.1109,
+      "step": 1716
+    },
+    {
+      "epoch": 0.3902272727272727,
+      "grad_norm": 2.6083285725587935,
+      "learning_rate": 8.367112001533073e-07,
+      "loss": 0.1454,
+      "step": 1717
+    },
+    {
+      "epoch": 0.39045454545454544,
+      "grad_norm": 2.899234798206412,
+      "learning_rate": 8.362912791631607e-07,
+      "loss": 0.1538,
+      "step": 1718
+    },
+    {
+      "epoch": 0.3906818181818182,
+      "grad_norm": 3.7710345716103677,
+      "learning_rate": 8.358712504580735e-07,
+      "loss": 0.1165,
+      "step": 1719
+    },
+    {
+      "epoch": 0.39090909090909093,
+      "grad_norm": 2.1990666503184175,
+      "learning_rate": 8.354511142521738e-07,
+      "loss": 0.146,
+      "step": 1720
+    },
+    {
+      "epoch": 0.3911363636363636,
+      "grad_norm": 2.9958825766392545,
+      "learning_rate": 8.350308707596438e-07,
+      "loss": 0.1297,
+      "step": 1721
+    },
+    {
+      "epoch": 0.39136363636363636,
+      "grad_norm": 2.7433104015626655,
+      "learning_rate": 8.346105201947217e-07,
+      "loss": 0.125,
+      "step": 1722
+    },
+    {
+      "epoch": 0.3915909090909091,
+      "grad_norm": 4.8768423179838525,
+      "learning_rate": 8.34190062771699e-07,
+      "loss": 0.2475,
+      "step": 1723
+    },
+    {
+      "epoch": 0.39181818181818184,
+      "grad_norm": 5.186868002354386,
+      "learning_rate": 8.337694987049225e-07,
+      "loss": 0.1585,
+      "step": 1724
+    },
+    {
+      "epoch": 0.39204545454545453,
+      "grad_norm": 1.1413520996568993,
+      "learning_rate": 8.333488282087926e-07,
+      "loss": 0.0926,
+      "step": 1725
+    },
+    {
+      "epoch": 0.3922727272727273,
+      "grad_norm": 3.639331391674984,
+      "learning_rate": 8.329280514977652e-07,
+      "loss": 0.1883,
+      "step": 1726
+    },
+    {
+      "epoch": 0.3925,
+      "grad_norm": 2.0640109706054126,
+      "learning_rate": 8.325071687863487e-07,
+      "loss": 0.1622,
+      "step": 1727
+    },
+    {
+      "epoch": 0.3927272727272727,
+      "grad_norm": 3.1563675491492504,
+      "learning_rate": 8.320861802891068e-07,
+      "loss": 0.1306,
+      "step": 1728
+    },
+    {
+      "epoch": 0.39295454545454545,
+      "grad_norm": 2.3268633188850942,
+      "learning_rate": 8.316650862206572e-07,
+      "loss": 0.1682,
+      "step": 1729
+    },
+    {
+      "epoch": 0.3931818181818182,
+      "grad_norm": 3.815658501229318,
+      "learning_rate": 8.312438867956702e-07,
+      "loss": 0.1476,
+      "step": 1730
+    },
+    {
+      "epoch": 0.39340909090909093,
+      "grad_norm": 2.468257969807865,
+      "learning_rate": 8.308225822288709e-07,
+      "loss": 0.1107,
+      "step": 1731
+    },
+    {
+      "epoch": 0.3936363636363636,
+      "grad_norm": 1.8717588999311954,
+      "learning_rate": 8.304011727350377e-07,
+      "loss": 0.1812,
+      "step": 1732
+    },
+    {
+      "epoch": 0.39386363636363636,
+      "grad_norm": 2.5663426407449403,
+      "learning_rate": 8.299796585290024e-07,
+      "loss": 0.1197,
+      "step": 1733
+    },
+    {
+      "epoch": 0.3940909090909091,
+      "grad_norm": 3.015060846960058,
+      "learning_rate": 8.295580398256503e-07,
+      "loss": 0.218,
+      "step": 1734
+    },
+    {
+      "epoch": 0.39431818181818185,
+      "grad_norm": 2.7478554031102345,
+      "learning_rate": 8.291363168399198e-07,
+      "loss": 0.2502,
+      "step": 1735
+    },
+    {
+      "epoch": 0.39454545454545453,
+      "grad_norm": 2.5092846827652506,
+      "learning_rate": 8.287144897868027e-07,
+      "loss": 0.1303,
+      "step": 1736
+    },
+    {
+      "epoch": 0.3947727272727273,
+      "grad_norm": 4.421750981170485,
+      "learning_rate": 8.282925588813437e-07,
+      "loss": 0.2236,
+      "step": 1737
+    },
+    {
+      "epoch": 0.395,
+      "grad_norm": 2.8440957484427978,
+      "learning_rate": 8.278705243386404e-07,
+      "loss": 0.1532,
+      "step": 1738
+    },
+    {
+      "epoch": 0.3952272727272727,
+      "grad_norm": 4.359009228111746,
+      "learning_rate": 8.274483863738435e-07,
+      "loss": 0.126,
+      "step": 1739
+    },
+    {
+      "epoch": 0.39545454545454545,
+      "grad_norm": 2.2182368045449268,
+      "learning_rate": 8.270261452021559e-07,
+      "loss": 0.0921,
+      "step": 1740
+    },
+    {
+      "epoch": 0.3956818181818182,
+      "grad_norm": 2.658540354146926,
+      "learning_rate": 8.266038010388339e-07,
+      "loss": 0.093,
+      "step": 1741
+    },
+    {
+      "epoch": 0.39590909090909093,
+      "grad_norm": 2.9481856691724415,
+      "learning_rate": 8.261813540991853e-07,
+      "loss": 0.1523,
+      "step": 1742
+    },
+    {
+      "epoch": 0.3961363636363636,
+      "grad_norm": 3.48452710727713,
+      "learning_rate": 8.257588045985712e-07,
+      "loss": 0.1154,
+      "step": 1743
+    },
+    {
+      "epoch": 0.39636363636363636,
+      "grad_norm": 1.566460228351363,
+      "learning_rate": 8.253361527524047e-07,
+      "loss": 0.1146,
+      "step": 1744
+    },
+    {
+      "epoch": 0.3965909090909091,
+      "grad_norm": 2.7376448675254004,
+      "learning_rate": 8.249133987761507e-07,
+      "loss": 0.0742,
+      "step": 1745
+    },
+    {
+      "epoch": 0.3968181818181818,
+      "grad_norm": 2.7541794831317303,
+      "learning_rate": 8.244905428853267e-07,
+      "loss": 0.1224,
+      "step": 1746
+    },
+    {
+      "epoch": 0.39704545454545453,
+      "grad_norm": 3.3124508890143782,
+      "learning_rate": 8.240675852955019e-07,
+      "loss": 0.1636,
+      "step": 1747
+    },
+    {
+      "epoch": 0.3972727272727273,
+      "grad_norm": 3.179870161522294,
+      "learning_rate": 8.236445262222972e-07,
+      "loss": 0.2675,
+      "step": 1748
+    },
+    {
+      "epoch": 0.3975,
+      "grad_norm": 3.9159891931727673,
+      "learning_rate": 8.232213658813856e-07,
+      "loss": 0.2532,
+      "step": 1749
+    },
+    {
+      "epoch": 0.3977272727272727,
+      "grad_norm": 3.9686776637324472,
+      "learning_rate": 8.227981044884914e-07,
+      "loss": 0.2307,
+      "step": 1750
+    },
+    {
+      "epoch": 0.39795454545454545,
+      "grad_norm": 2.5654151976490143,
+      "learning_rate": 8.223747422593907e-07,
+      "loss": 0.1238,
+      "step": 1751
+    },
+    {
+      "epoch": 0.3981818181818182,
+      "grad_norm": 2.769547969332957,
+      "learning_rate": 8.219512794099108e-07,
+      "loss": 0.1057,
+      "step": 1752
+    },
+    {
+      "epoch": 0.39840909090909093,
+      "grad_norm": 1.9416857769689935,
+      "learning_rate": 8.215277161559301e-07,
+      "loss": 0.1548,
+      "step": 1753
+    },
+    {
+      "epoch": 0.3986363636363636,
+      "grad_norm": 2.173642022251857,
+      "learning_rate": 8.211040527133787e-07,
+      "loss": 0.1086,
+      "step": 1754
+    },
+    {
+      "epoch": 0.39886363636363636,
+      "grad_norm": 4.252128171627704,
+      "learning_rate": 8.206802892982376e-07,
+      "loss": 0.1841,
+      "step": 1755
+    },
+    {
+      "epoch": 0.3990909090909091,
+      "grad_norm": 2.8730243736690837,
+      "learning_rate": 8.202564261265385e-07,
+      "loss": 0.1243,
+      "step": 1756
+    },
+    {
+      "epoch": 0.3993181818181818,
+      "grad_norm": 2.865164624204365,
+      "learning_rate": 8.19832463414364e-07,
+      "loss": 0.1725,
+      "step": 1757
+    },
+    {
+      "epoch": 0.39954545454545454,
+      "grad_norm": 2.6206065773952187,
+      "learning_rate": 8.194084013778478e-07,
+      "loss": 0.1307,
+      "step": 1758
+    },
+    {
+      "epoch": 0.3997727272727273,
+      "grad_norm": 4.029846008988471,
+      "learning_rate": 8.189842402331741e-07,
+      "loss": 0.3361,
+      "step": 1759
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.5962747768831123,
+      "learning_rate": 8.185599801965772e-07,
+      "loss": 0.1271,
+      "step": 1760
+    },
+    {
+      "epoch": 0.4002272727272727,
+      "grad_norm": 2.8867403001793828,
+      "learning_rate": 8.181356214843422e-07,
+      "loss": 0.1441,
+      "step": 1761
+    },
+    {
+      "epoch": 0.40045454545454545,
+      "grad_norm": 2.867375767086447,
+      "learning_rate": 8.177111643128048e-07,
+      "loss": 0.1089,
+      "step": 1762
+    },
+    {
+      "epoch": 0.4006818181818182,
+      "grad_norm": 3.8550150238302328,
+      "learning_rate": 8.172866088983501e-07,
+      "loss": 0.1695,
+      "step": 1763
+    },
+    {
+      "epoch": 0.4009090909090909,
+      "grad_norm": 2.2768437050417463,
+      "learning_rate": 8.168619554574142e-07,
+      "loss": 0.1933,
+      "step": 1764
+    },
+    {
+      "epoch": 0.4011363636363636,
+      "grad_norm": 3.480283165251106,
+      "learning_rate": 8.164372042064825e-07,
+      "loss": 0.1674,
+      "step": 1765
+    },
+    {
+      "epoch": 0.40136363636363637,
+      "grad_norm": 2.8264782939012028,
+      "learning_rate": 8.160123553620901e-07,
+      "loss": 0.2388,
+      "step": 1766
+    },
+    {
+      "epoch": 0.4015909090909091,
+      "grad_norm": 2.2898312822390987,
+      "learning_rate": 8.155874091408228e-07,
+      "loss": 0.1595,
+      "step": 1767
+    },
+    {
+      "epoch": 0.4018181818181818,
+      "grad_norm": 3.431558062155486,
+      "learning_rate": 8.15162365759315e-07,
+      "loss": 0.1942,
+      "step": 1768
+    },
+    {
+      "epoch": 0.40204545454545454,
+      "grad_norm": 2.41935998296563,
+      "learning_rate": 8.147372254342514e-07,
+      "loss": 0.0812,
+      "step": 1769
+    },
+    {
+      "epoch": 0.4022727272727273,
+      "grad_norm": 2.1866773521211758,
+      "learning_rate": 8.143119883823658e-07,
+      "loss": 0.0967,
+      "step": 1770
+    },
+    {
+      "epoch": 0.4025,
+      "grad_norm": 2.8726881783984455,
+      "learning_rate": 8.138866548204412e-07,
+      "loss": 0.2135,
+      "step": 1771
+    },
+    {
+      "epoch": 0.4027272727272727,
+      "grad_norm": 2.7307252510149778,
+      "learning_rate": 8.134612249653098e-07,
+      "loss": 0.1198,
+      "step": 1772
+    },
+    {
+      "epoch": 0.40295454545454545,
+      "grad_norm": 2.1539482422642213,
+      "learning_rate": 8.130356990338535e-07,
+      "loss": 0.16,
+      "step": 1773
+    },
+    {
+      "epoch": 0.4031818181818182,
+      "grad_norm": 2.3810221423914055,
+      "learning_rate": 8.126100772430021e-07,
+      "loss": 0.1189,
+      "step": 1774
+    },
+    {
+      "epoch": 0.4034090909090909,
+      "grad_norm": 3.906492312566225,
+      "learning_rate": 8.121843598097351e-07,
+      "loss": 0.133,
+      "step": 1775
+    },
+    {
+      "epoch": 0.4036363636363636,
+      "grad_norm": 2.968750039200645,
+      "learning_rate": 8.117585469510808e-07,
+      "loss": 0.1127,
+      "step": 1776
+    },
+    {
+      "epoch": 0.40386363636363637,
+      "grad_norm": 2.428579155379299,
+      "learning_rate": 8.113326388841154e-07,
+      "loss": 0.1784,
+      "step": 1777
+    },
+    {
+      "epoch": 0.4040909090909091,
+      "grad_norm": 1.6890657741090138,
+      "learning_rate": 8.109066358259642e-07,
+      "loss": 0.1611,
+      "step": 1778
+    },
+    {
+      "epoch": 0.4043181818181818,
+      "grad_norm": 2.1863623152961065,
+      "learning_rate": 8.104805379938013e-07,
+      "loss": 0.0726,
+      "step": 1779
+    },
+    {
+      "epoch": 0.40454545454545454,
+      "grad_norm": 2.1603290938092625,
+      "learning_rate": 8.100543456048482e-07,
+      "loss": 0.0832,
+      "step": 1780
+    },
+    {
+      "epoch": 0.4047727272727273,
+      "grad_norm": 4.960432977601387,
+      "learning_rate": 8.096280588763751e-07,
+      "loss": 0.2917,
+      "step": 1781
+    },
+    {
+      "epoch": 0.405,
+      "grad_norm": 1.3730897798892996,
+      "learning_rate": 8.092016780257003e-07,
+      "loss": 0.1199,
+      "step": 1782
+    },
+    {
+      "epoch": 0.4052272727272727,
+      "grad_norm": 2.368700658428993,
+      "learning_rate": 8.0877520327019e-07,
+      "loss": 0.1516,
+      "step": 1783
+    },
+    {
+      "epoch": 0.40545454545454546,
+      "grad_norm": 2.1112200988250898,
+      "learning_rate": 8.083486348272584e-07,
+      "loss": 0.1189,
+      "step": 1784
+    },
+    {
+      "epoch": 0.4056818181818182,
+      "grad_norm": 2.2190398748759783,
+      "learning_rate": 8.079219729143672e-07,
+      "loss": 0.1409,
+      "step": 1785
+    },
+    {
+      "epoch": 0.4059090909090909,
+      "grad_norm": 3.131597998681549,
+      "learning_rate": 8.07495217749026e-07,
+      "loss": 0.2122,
+      "step": 1786
+    },
+    {
+      "epoch": 0.40613636363636363,
+      "grad_norm": 1.954305204698961,
+      "learning_rate": 8.070683695487919e-07,
+      "loss": 0.1605,
+      "step": 1787
+    },
+    {
+      "epoch": 0.40636363636363637,
+      "grad_norm": 3.444444213212602,
+      "learning_rate": 8.066414285312693e-07,
+      "loss": 0.1587,
+      "step": 1788
+    },
+    {
+      "epoch": 0.4065909090909091,
+      "grad_norm": 4.191460244388047,
+      "learning_rate": 8.062143949141099e-07,
+      "loss": 0.2011,
+      "step": 1789
+    },
+    {
+      "epoch": 0.4068181818181818,
+      "grad_norm": 4.033025281181491,
+      "learning_rate": 8.057872689150129e-07,
+      "loss": 0.1384,
+      "step": 1790
+    },
+    {
+      "epoch": 0.40704545454545454,
+      "grad_norm": 3.1949052176082677,
+      "learning_rate": 8.053600507517244e-07,
+      "loss": 0.195,
+      "step": 1791
+    },
+    {
+      "epoch": 0.4072727272727273,
+      "grad_norm": 2.7036090443941534,
+      "learning_rate": 8.049327406420371e-07,
+      "loss": 0.1102,
+      "step": 1792
+    },
+    {
+      "epoch": 0.4075,
+      "grad_norm": 2.2003096929935064,
+      "learning_rate": 8.045053388037916e-07,
+      "loss": 0.2013,
+      "step": 1793
+    },
+    {
+      "epoch": 0.4077272727272727,
+      "grad_norm": 2.5928240347408766,
+      "learning_rate": 8.040778454548741e-07,
+      "loss": 0.2221,
+      "step": 1794
+    },
+    {
+      "epoch": 0.40795454545454546,
+      "grad_norm": 2.483672080364399,
+      "learning_rate": 8.036502608132181e-07,
+      "loss": 0.1451,
+      "step": 1795
+    },
+    {
+      "epoch": 0.4081818181818182,
+      "grad_norm": 2.7771845705683473,
+      "learning_rate": 8.032225850968037e-07,
+      "loss": 0.0862,
+      "step": 1796
+    },
+    {
+      "epoch": 0.4084090909090909,
+      "grad_norm": 3.86845475684359,
+      "learning_rate": 8.027948185236568e-07,
+      "loss": 0.1085,
+      "step": 1797
+    },
+    {
+      "epoch": 0.40863636363636363,
+      "grad_norm": 3.395938813010041,
+      "learning_rate": 8.023669613118507e-07,
+      "loss": 0.1096,
+      "step": 1798
+    },
+    {
+      "epoch": 0.4088636363636364,
+      "grad_norm": 3.376047282459046,
+      "learning_rate": 8.019390136795039e-07,
+      "loss": 0.1461,
+      "step": 1799
+    },
+    {
+      "epoch": 0.4090909090909091,
+      "grad_norm": 2.563044804133937,
+      "learning_rate": 8.01510975844781e-07,
+      "loss": 0.1707,
+      "step": 1800
+    },
+    {
+      "epoch": 0.4093181818181818,
+      "grad_norm": 1.9188560060049273,
+      "learning_rate": 8.010828480258937e-07,
+      "loss": 0.1383,
+      "step": 1801
+    },
+    {
+      "epoch": 0.40954545454545455,
+      "grad_norm": 3.1314802820874568,
+      "learning_rate": 8.006546304410984e-07,
+      "loss": 0.113,
+      "step": 1802
+    },
+    {
+      "epoch": 0.4097727272727273,
+      "grad_norm": 4.675559396805536,
+      "learning_rate": 8.002263233086975e-07,
+      "loss": 0.1476,
+      "step": 1803
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.7154371703715445,
+      "learning_rate": 7.997979268470396e-07,
+      "loss": 0.2886,
+      "step": 1804
+    },
+    {
+      "epoch": 0.4102272727272727,
+      "grad_norm": 2.2005296786754323,
+      "learning_rate": 7.993694412745184e-07,
+      "loss": 0.113,
+      "step": 1805
+    },
+    {
+      "epoch": 0.41045454545454546,
+      "grad_norm": 2.081539778012685,
+      "learning_rate": 7.98940866809573e-07,
+      "loss": 0.1194,
+      "step": 1806
+    },
+    {
+      "epoch": 0.4106818181818182,
+      "grad_norm": 3.9033389053034435,
+      "learning_rate": 7.985122036706879e-07,
+      "loss": 0.1421,
+      "step": 1807
+    },
+    {
+      "epoch": 0.4109090909090909,
+      "grad_norm": 4.572017353781462,
+      "learning_rate": 7.980834520763929e-07,
+      "loss": 0.1549,
+      "step": 1808
+    },
+    {
+      "epoch": 0.41113636363636363,
+      "grad_norm": 2.566760836048433,
+      "learning_rate": 7.976546122452627e-07,
+      "loss": 0.1045,
+      "step": 1809
+    },
+    {
+      "epoch": 0.4113636363636364,
+      "grad_norm": 3.6486952054941137,
+      "learning_rate": 7.972256843959172e-07,
+      "loss": 0.1325,
+      "step": 1810
+    },
+    {
+      "epoch": 0.4115909090909091,
+      "grad_norm": 2.179653511420108,
+      "learning_rate": 7.967966687470211e-07,
+      "loss": 0.1465,
+      "step": 1811
+    },
+    {
+      "epoch": 0.4118181818181818,
+      "grad_norm": 1.8959383449687663,
+      "learning_rate": 7.963675655172839e-07,
+      "loss": 0.0855,
+      "step": 1812
+    },
+    {
+      "epoch": 0.41204545454545455,
+      "grad_norm": 2.249949202671389,
+      "learning_rate": 7.959383749254595e-07,
+      "loss": 0.1215,
+      "step": 1813
+    },
+    {
+      "epoch": 0.4122727272727273,
+      "grad_norm": 2.414288426991572,
+      "learning_rate": 7.955090971903464e-07,
+      "loss": 0.2062,
+      "step": 1814
+    },
+    {
+      "epoch": 0.4125,
+      "grad_norm": 4.440676208855101,
+      "learning_rate": 7.95079732530788e-07,
+      "loss": 0.1569,
+      "step": 1815
+    },
+    {
+      "epoch": 0.4127272727272727,
+      "grad_norm": 2.4574248414887077,
+      "learning_rate": 7.946502811656714e-07,
+      "loss": 0.1052,
+      "step": 1816
+    },
+    {
+      "epoch": 0.41295454545454546,
+      "grad_norm": 4.098125564628173,
+      "learning_rate": 7.942207433139284e-07,
+      "loss": 0.1462,
+      "step": 1817
+    },
+    {
+      "epoch": 0.4131818181818182,
+      "grad_norm": 2.868190608306989,
+      "learning_rate": 7.937911191945342e-07,
+      "loss": 0.2667,
+      "step": 1818
+    },
+    {
+      "epoch": 0.4134090909090909,
+      "grad_norm": 3.0482647223701065,
+      "learning_rate": 7.933614090265089e-07,
+      "loss": 0.0948,
+      "step": 1819
+    },
+    {
+      "epoch": 0.41363636363636364,
+      "grad_norm": 4.10958331208386,
+      "learning_rate": 7.929316130289158e-07,
+      "loss": 0.1024,
+      "step": 1820
+    },
+    {
+      "epoch": 0.4138636363636364,
+      "grad_norm": 2.711564076865291,
+      "learning_rate": 7.925017314208619e-07,
+      "loss": 0.1752,
+      "step": 1821
+    },
+    {
+      "epoch": 0.41409090909090907,
+      "grad_norm": 3.0067467862865405,
+      "learning_rate": 7.920717644214986e-07,
+      "loss": 0.1436,
+      "step": 1822
+    },
+    {
+      "epoch": 0.4143181818181818,
+      "grad_norm": 4.100986119899566,
+      "learning_rate": 7.916417122500199e-07,
+      "loss": 0.1192,
+      "step": 1823
+    },
+    {
+      "epoch": 0.41454545454545455,
+      "grad_norm": 2.8424048381129103,
+      "learning_rate": 7.912115751256637e-07,
+      "loss": 0.1383,
+      "step": 1824
+    },
+    {
+      "epoch": 0.4147727272727273,
+      "grad_norm": 3.006188024103048,
+      "learning_rate": 7.907813532677111e-07,
+      "loss": 0.1207,
+      "step": 1825
+    },
+    {
+      "epoch": 0.415,
+      "grad_norm": 3.345238764308273,
+      "learning_rate": 7.90351046895487e-07,
+      "loss": 0.1394,
+      "step": 1826
+    },
+    {
+      "epoch": 0.4152272727272727,
+      "grad_norm": 2.93299997186385,
+      "learning_rate": 7.899206562283582e-07,
+      "loss": 0.1947,
+      "step": 1827
+    },
+    {
+      "epoch": 0.41545454545454547,
+      "grad_norm": 2.2006824336528203,
+      "learning_rate": 7.894901814857353e-07,
+      "loss": 0.133,
+      "step": 1828
+    },
+    {
+      "epoch": 0.4156818181818182,
+      "grad_norm": 3.1628746082250663,
+      "learning_rate": 7.890596228870716e-07,
+      "loss": 0.244,
+      "step": 1829
+    },
+    {
+      "epoch": 0.4159090909090909,
+      "grad_norm": 1.5276585677326426,
+      "learning_rate": 7.88628980651863e-07,
+      "loss": 0.1128,
+      "step": 1830
+    },
+    {
+      "epoch": 0.41613636363636364,
+      "grad_norm": 2.342268611832613,
+      "learning_rate": 7.881982549996485e-07,
+      "loss": 0.2178,
+      "step": 1831
+    },
+    {
+      "epoch": 0.4163636363636364,
+      "grad_norm": 2.9535563136757976,
+      "learning_rate": 7.87767446150009e-07,
+      "loss": 0.0856,
+      "step": 1832
+    },
+    {
+      "epoch": 0.41659090909090907,
+      "grad_norm": 3.3043910821525504,
+      "learning_rate": 7.873365543225678e-07,
+      "loss": 0.1626,
+      "step": 1833
+    },
+    {
+      "epoch": 0.4168181818181818,
+      "grad_norm": 4.774594944074803,
+      "learning_rate": 7.869055797369915e-07,
+      "loss": 0.1628,
+      "step": 1834
+    },
+    {
+      "epoch": 0.41704545454545455,
+      "grad_norm": 5.463696166929032,
+      "learning_rate": 7.864745226129877e-07,
+      "loss": 0.1738,
+      "step": 1835
+    },
+    {
+      "epoch": 0.4172727272727273,
+      "grad_norm": 2.647871165572832,
+      "learning_rate": 7.860433831703066e-07,
+      "loss": 0.1129,
+      "step": 1836
+    },
+    {
+      "epoch": 0.4175,
+      "grad_norm": 2.5579746431322112,
+      "learning_rate": 7.856121616287403e-07,
+      "loss": 0.1287,
+      "step": 1837
+    },
+    {
+      "epoch": 0.4177272727272727,
+      "grad_norm": 4.471991831053429,
+      "learning_rate": 7.85180858208123e-07,
+      "loss": 0.1042,
+      "step": 1838
+    },
+    {
+      "epoch": 0.41795454545454547,
+      "grad_norm": 2.7072142806056583,
+      "learning_rate": 7.847494731283302e-07,
+      "loss": 0.0935,
+      "step": 1839
+    },
+    {
+      "epoch": 0.41818181818181815,
+      "grad_norm": 3.1394016649151615,
+      "learning_rate": 7.843180066092795e-07,
+      "loss": 0.1098,
+      "step": 1840
+    },
+    {
+      "epoch": 0.4184090909090909,
+      "grad_norm": 3.109567067936425,
+      "learning_rate": 7.838864588709296e-07,
+      "loss": 0.1845,
+      "step": 1841
+    },
+    {
+      "epoch": 0.41863636363636364,
+      "grad_norm": 2.2847958012036655,
+      "learning_rate": 7.834548301332805e-07,
+      "loss": 0.0733,
+      "step": 1842
+    },
+    {
+      "epoch": 0.4188636363636364,
+      "grad_norm": 2.759991571605316,
+      "learning_rate": 7.830231206163742e-07,
+      "loss": 0.1017,
+      "step": 1843
+    },
+    {
+      "epoch": 0.41909090909090907,
+      "grad_norm": 2.445903774773626,
+      "learning_rate": 7.825913305402932e-07,
+      "loss": 0.0856,
+      "step": 1844
+    },
+    {
+      "epoch": 0.4193181818181818,
+      "grad_norm": 2.9301633100306876,
+      "learning_rate": 7.821594601251617e-07,
+      "loss": 0.125,
+      "step": 1845
+    },
+    {
+      "epoch": 0.41954545454545455,
+      "grad_norm": 2.953865859850689,
+      "learning_rate": 7.817275095911439e-07,
+      "loss": 0.1209,
+      "step": 1846
+    },
+    {
+      "epoch": 0.4197727272727273,
+      "grad_norm": 1.9863076558960473,
+      "learning_rate": 7.812954791584458e-07,
+      "loss": 0.0847,
+      "step": 1847
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7259355491292707,
+      "learning_rate": 7.808633690473136e-07,
+      "loss": 0.1571,
+      "step": 1848
+    },
+    {
+      "epoch": 0.4202272727272727,
+      "grad_norm": 3.7332013251607012,
+      "learning_rate": 7.804311794780345e-07,
+      "loss": 0.1166,
+      "step": 1849
+    },
+    {
+      "epoch": 0.42045454545454547,
+      "grad_norm": 3.427181870191958,
+      "learning_rate": 7.799989106709354e-07,
+      "loss": 0.1067,
+      "step": 1850
+    },
+    {
+      "epoch": 0.42068181818181816,
+      "grad_norm": 2.997208064108121,
+      "learning_rate": 7.795665628463847e-07,
+      "loss": 0.1319,
+      "step": 1851
+    },
+    {
+      "epoch": 0.4209090909090909,
+      "grad_norm": 3.677407633760486,
+      "learning_rate": 7.791341362247906e-07,
+      "loss": 0.1098,
+      "step": 1852
+    },
+    {
+      "epoch": 0.42113636363636364,
+      "grad_norm": 2.565029554724448,
+      "learning_rate": 7.78701631026601e-07,
+      "loss": 0.1179,
+      "step": 1853
+    },
+    {
+      "epoch": 0.4213636363636364,
+      "grad_norm": 3.083854499469121,
+      "learning_rate": 7.782690474723045e-07,
+      "loss": 0.1661,
+      "step": 1854
+    },
+    {
+      "epoch": 0.42159090909090907,
+      "grad_norm": 1.4569942393978668,
+      "learning_rate": 7.778363857824295e-07,
+      "loss": 0.0977,
+      "step": 1855
+    },
+    {
+      "epoch": 0.4218181818181818,
+      "grad_norm": 2.490780589067239,
+      "learning_rate": 7.774036461775442e-07,
+      "loss": 0.1988,
+      "step": 1856
+    },
+    {
+      "epoch": 0.42204545454545456,
+      "grad_norm": 2.7487654148573637,
+      "learning_rate": 7.769708288782562e-07,
+      "loss": 0.1116,
+      "step": 1857
+    },
+    {
+      "epoch": 0.4222727272727273,
+      "grad_norm": 2.5713745178176377,
+      "learning_rate": 7.765379341052132e-07,
+      "loss": 0.0982,
+      "step": 1858
+    },
+    {
+      "epoch": 0.4225,
+      "grad_norm": 1.6880631113059341,
+      "learning_rate": 7.761049620791019e-07,
+      "loss": 0.058,
+      "step": 1859
+    },
+    {
+      "epoch": 0.42272727272727273,
+      "grad_norm": 2.761123461720594,
+      "learning_rate": 7.756719130206493e-07,
+      "loss": 0.2357,
+      "step": 1860
+    },
+    {
+      "epoch": 0.42295454545454547,
+      "grad_norm": 2.3360862175678587,
+      "learning_rate": 7.752387871506203e-07,
+      "loss": 0.1812,
+      "step": 1861
+    },
+    {
+      "epoch": 0.42318181818181816,
+      "grad_norm": 3.571387078848797,
+      "learning_rate": 7.748055846898201e-07,
+      "loss": 0.1532,
+      "step": 1862
+    },
+    {
+      "epoch": 0.4234090909090909,
+      "grad_norm": 4.740469783510227,
+      "learning_rate": 7.743723058590926e-07,
+      "loss": 0.2471,
+      "step": 1863
+    },
+    {
+      "epoch": 0.42363636363636364,
+      "grad_norm": 2.7192296487482444,
+      "learning_rate": 7.739389508793204e-07,
+      "loss": 0.2044,
+      "step": 1864
+    },
+    {
+      "epoch": 0.4238636363636364,
+      "grad_norm": 2.4481727251550205,
+      "learning_rate": 7.73505519971425e-07,
+      "loss": 0.0936,
+      "step": 1865
+    },
+    {
+      "epoch": 0.4240909090909091,
+      "grad_norm": 3.1581247279835534,
+      "learning_rate": 7.730720133563671e-07,
+      "loss": 0.1104,
+      "step": 1866
+    },
+    {
+      "epoch": 0.4243181818181818,
+      "grad_norm": 2.2495709202226433,
+      "learning_rate": 7.726384312551452e-07,
+      "loss": 0.2177,
+      "step": 1867
+    },
+    {
+      "epoch": 0.42454545454545456,
+      "grad_norm": 2.7978675200826406,
+      "learning_rate": 7.722047738887967e-07,
+      "loss": 0.1165,
+      "step": 1868
+    },
+    {
+      "epoch": 0.42477272727272725,
+      "grad_norm": 2.5051231178666495,
+      "learning_rate": 7.717710414783974e-07,
+      "loss": 0.1784,
+      "step": 1869
+    },
+    {
+      "epoch": 0.425,
+      "grad_norm": 3.152890457859899,
+      "learning_rate": 7.713372342450616e-07,
+      "loss": 0.168,
+      "step": 1870
+    },
+    {
+      "epoch": 0.42522727272727273,
+      "grad_norm": 2.6422441662031715,
+      "learning_rate": 7.709033524099411e-07,
+      "loss": 0.0695,
+      "step": 1871
+    },
+    {
+      "epoch": 0.4254545454545455,
+      "grad_norm": 2.93048892764854,
+      "learning_rate": 7.704693961942262e-07,
+      "loss": 0.0966,
+      "step": 1872
+    },
+    {
+      "epoch": 0.42568181818181816,
+      "grad_norm": 2.2853375456564526,
+      "learning_rate": 7.700353658191448e-07,
+      "loss": 0.066,
+      "step": 1873
+    },
+    {
+      "epoch": 0.4259090909090909,
+      "grad_norm": 2.316090434081947,
+      "learning_rate": 7.696012615059631e-07,
+      "loss": 0.1466,
+      "step": 1874
+    },
+    {
+      "epoch": 0.42613636363636365,
+      "grad_norm": 2.471443709531704,
+      "learning_rate": 7.691670834759846e-07,
+      "loss": 0.1262,
+      "step": 1875
+    },
+    {
+      "epoch": 0.4263636363636364,
+      "grad_norm": 2.0292988127962017,
+      "learning_rate": 7.687328319505505e-07,
+      "loss": 0.1054,
+      "step": 1876
+    },
+    {
+      "epoch": 0.4265909090909091,
+      "grad_norm": 2.690888089892935,
+      "learning_rate": 7.682985071510392e-07,
+      "loss": 0.1454,
+      "step": 1877
+    },
+    {
+      "epoch": 0.4268181818181818,
+      "grad_norm": 2.92363010650926,
+      "learning_rate": 7.678641092988673e-07,
+      "loss": 0.1288,
+      "step": 1878
+    },
+    {
+      "epoch": 0.42704545454545456,
+      "grad_norm": 2.7575399544512775,
+      "learning_rate": 7.674296386154873e-07,
+      "loss": 0.1801,
+      "step": 1879
+    },
+    {
+      "epoch": 0.42727272727272725,
+      "grad_norm": 3.2178449395524336,
+      "learning_rate": 7.6699509532239e-07,
+      "loss": 0.2002,
+      "step": 1880
+    },
+    {
+      "epoch": 0.4275,
+      "grad_norm": 3.467129989319449,
+      "learning_rate": 7.665604796411029e-07,
+      "loss": 0.1019,
+      "step": 1881
+    },
+    {
+      "epoch": 0.42772727272727273,
+      "grad_norm": 3.2805598221046295,
+      "learning_rate": 7.661257917931899e-07,
+      "loss": 0.1146,
+      "step": 1882
+    },
+    {
+      "epoch": 0.4279545454545455,
+      "grad_norm": 2.3987421029685283,
+      "learning_rate": 7.656910320002522e-07,
+      "loss": 0.0914,
+      "step": 1883
+    },
+    {
+      "epoch": 0.42818181818181816,
+      "grad_norm": 1.4210002263514088,
+      "learning_rate": 7.65256200483928e-07,
+      "loss": 0.0693,
+      "step": 1884
+    },
+    {
+      "epoch": 0.4284090909090909,
+      "grad_norm": 2.6264352598348455,
+      "learning_rate": 7.64821297465891e-07,
+      "loss": 0.1648,
+      "step": 1885
+    },
+    {
+      "epoch": 0.42863636363636365,
+      "grad_norm": 2.5314042051168935,
+      "learning_rate": 7.643863231678523e-07,
+      "loss": 0.1642,
+      "step": 1886
+    },
+    {
+      "epoch": 0.4288636363636364,
+      "grad_norm": 4.402575051876635,
+      "learning_rate": 7.639512778115589e-07,
+      "loss": 0.1334,
+      "step": 1887
+    },
+    {
+      "epoch": 0.4290909090909091,
+      "grad_norm": 3.5067880564238636,
+      "learning_rate": 7.635161616187941e-07,
+      "loss": 0.1068,
+      "step": 1888
+    },
+    {
+      "epoch": 0.4293181818181818,
+      "grad_norm": 2.5755108798385606,
+      "learning_rate": 7.630809748113777e-07,
+      "loss": 0.2242,
+      "step": 1889
+    },
+    {
+      "epoch": 0.42954545454545456,
+      "grad_norm": 2.714218126885066,
+      "learning_rate": 7.626457176111647e-07,
+      "loss": 0.0921,
+      "step": 1890
+    },
+    {
+      "epoch": 0.42977272727272725,
+      "grad_norm": 3.038687258997189,
+      "learning_rate": 7.622103902400466e-07,
+      "loss": 0.0831,
+      "step": 1891
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.94466554554241,
+      "learning_rate": 7.617749929199506e-07,
+      "loss": 0.2477,
+      "step": 1892
+    },
+    {
+      "epoch": 0.43022727272727274,
+      "grad_norm": 3.0269908517669677,
+      "learning_rate": 7.613395258728392e-07,
+      "loss": 0.2078,
+      "step": 1893
+    },
+    {
+      "epoch": 0.4304545454545455,
+      "grad_norm": 3.1251401278055937,
+      "learning_rate": 7.60903989320711e-07,
+      "loss": 0.1749,
+      "step": 1894
+    },
+    {
+      "epoch": 0.43068181818181817,
+      "grad_norm": 2.882407460206729,
+      "learning_rate": 7.604683834855995e-07,
+      "loss": 0.1593,
+      "step": 1895
+    },
+    {
+      "epoch": 0.4309090909090909,
+      "grad_norm": 3.583979873183621,
+      "learning_rate": 7.600327085895742e-07,
+      "loss": 0.1678,
+      "step": 1896
+    },
+    {
+      "epoch": 0.43113636363636365,
+      "grad_norm": 3.3468068637557673,
+      "learning_rate": 7.595969648547388e-07,
+      "loss": 0.1757,
+      "step": 1897
+    },
+    {
+      "epoch": 0.43136363636363634,
+      "grad_norm": 4.2455488451213,
+      "learning_rate": 7.591611525032328e-07,
+      "loss": 0.2802,
+      "step": 1898
+    },
+    {
+      "epoch": 0.4315909090909091,
+      "grad_norm": 1.8757240661453398,
+      "learning_rate": 7.587252717572312e-07,
+      "loss": 0.1815,
+      "step": 1899
+    },
+    {
+      "epoch": 0.4318181818181818,
+      "grad_norm": 3.867721705419927,
+      "learning_rate": 7.582893228389423e-07,
+      "loss": 0.1168,
+      "step": 1900
+    },
+    {
+      "epoch": 0.43204545454545457,
+      "grad_norm": 2.9984096744478195,
+      "learning_rate": 7.578533059706104e-07,
+      "loss": 0.1027,
+      "step": 1901
+    },
+    {
+      "epoch": 0.43227272727272725,
+      "grad_norm": 3.2527534506808053,
+      "learning_rate": 7.574172213745144e-07,
+      "loss": 0.0812,
+      "step": 1902
+    },
+    {
+      "epoch": 0.4325,
+      "grad_norm": 2.0094939161107694,
+      "learning_rate": 7.569810692729672e-07,
+      "loss": 0.1147,
+      "step": 1903
+    },
+    {
+      "epoch": 0.43272727272727274,
+      "grad_norm": 3.296379482619113,
+      "learning_rate": 7.56544849888316e-07,
+      "loss": 0.0959,
+      "step": 1904
+    },
+    {
+      "epoch": 0.4329545454545455,
+      "grad_norm": 3.134757304282939,
+      "learning_rate": 7.561085634429434e-07,
+      "loss": 0.1499,
+      "step": 1905
+    },
+    {
+      "epoch": 0.43318181818181817,
+      "grad_norm": 2.0774149309243177,
+      "learning_rate": 7.556722101592648e-07,
+      "loss": 0.2177,
+      "step": 1906
+    },
+    {
+      "epoch": 0.4334090909090909,
+      "grad_norm": 3.8166454193357895,
+      "learning_rate": 7.552357902597306e-07,
+      "loss": 0.1147,
+      "step": 1907
+    },
+    {
+      "epoch": 0.43363636363636365,
+      "grad_norm": 1.6783244156210106,
+      "learning_rate": 7.547993039668248e-07,
+      "loss": 0.1452,
+      "step": 1908
+    },
+    {
+      "epoch": 0.43386363636363634,
+      "grad_norm": 1.3883878500549327,
+      "learning_rate": 7.543627515030652e-07,
+      "loss": 0.1716,
+      "step": 1909
+    },
+    {
+      "epoch": 0.4340909090909091,
+      "grad_norm": 2.6444573811819234,
+      "learning_rate": 7.539261330910038e-07,
+      "loss": 0.0801,
+      "step": 1910
+    },
+    {
+      "epoch": 0.4343181818181818,
+      "grad_norm": 3.383871127329711,
+      "learning_rate": 7.534894489532255e-07,
+      "loss": 0.2652,
+      "step": 1911
+    },
+    {
+      "epoch": 0.43454545454545457,
+      "grad_norm": 4.578452936844604,
+      "learning_rate": 7.530526993123491e-07,
+      "loss": 0.2424,
+      "step": 1912
+    },
+    {
+      "epoch": 0.43477272727272726,
+      "grad_norm": 3.9339229282845714,
+      "learning_rate": 7.526158843910271e-07,
+      "loss": 0.1873,
+      "step": 1913
+    },
+    {
+      "epoch": 0.435,
+      "grad_norm": 3.3021756949882604,
+      "learning_rate": 7.521790044119447e-07,
+      "loss": 0.2172,
+      "step": 1914
+    },
+    {
+      "epoch": 0.43522727272727274,
+      "grad_norm": 3.1195402141642705,
+      "learning_rate": 7.517420595978203e-07,
+      "loss": 0.1236,
+      "step": 1915
+    },
+    {
+      "epoch": 0.4354545454545454,
+      "grad_norm": 1.9248299766743913,
+      "learning_rate": 7.513050501714061e-07,
+      "loss": 0.071,
+      "step": 1916
+    },
+    {
+      "epoch": 0.43568181818181817,
+      "grad_norm": 3.011625494682081,
+      "learning_rate": 7.508679763554865e-07,
+      "loss": 0.2006,
+      "step": 1917
+    },
+    {
+      "epoch": 0.4359090909090909,
+      "grad_norm": 3.497340118491921,
+      "learning_rate": 7.504308383728785e-07,
+      "loss": 0.1842,
+      "step": 1918
+    },
+    {
+      "epoch": 0.43613636363636366,
+      "grad_norm": 0.8383609468748778,
+      "learning_rate": 7.499936364464328e-07,
+      "loss": 0.0497,
+      "step": 1919
+    },
+    {
+      "epoch": 0.43636363636363634,
+      "grad_norm": 2.692988233820218,
+      "learning_rate": 7.495563707990318e-07,
+      "loss": 0.1452,
+      "step": 1920
+    },
+    {
+      "epoch": 0.4365909090909091,
+      "grad_norm": 3.527112272075242,
+      "learning_rate": 7.491190416535909e-07,
+      "loss": 0.1441,
+      "step": 1921
+    },
+    {
+      "epoch": 0.4368181818181818,
+      "grad_norm": 1.6663922161339164,
+      "learning_rate": 7.486816492330579e-07,
+      "loss": 0.1185,
+      "step": 1922
+    },
+    {
+      "epoch": 0.43704545454545457,
+      "grad_norm": 1.7709247935676666,
+      "learning_rate": 7.482441937604121e-07,
+      "loss": 0.0881,
+      "step": 1923
+    },
+    {
+      "epoch": 0.43727272727272726,
+      "grad_norm": 3.670965985244109,
+      "learning_rate": 7.478066754586659e-07,
+      "loss": 0.1458,
+      "step": 1924
+    },
+    {
+      "epoch": 0.4375,
+      "grad_norm": 1.8928156760766877,
+      "learning_rate": 7.473690945508636e-07,
+      "loss": 0.1473,
+      "step": 1925
+    },
+    {
+      "epoch": 0.43772727272727274,
+      "grad_norm": 4.093814674520042,
+      "learning_rate": 7.469314512600804e-07,
+      "loss": 0.1859,
+      "step": 1926
+    },
+    {
+      "epoch": 0.43795454545454543,
+      "grad_norm": 3.2909159710990745,
+      "learning_rate": 7.464937458094243e-07,
+      "loss": 0.1563,
+      "step": 1927
+    },
+    {
+      "epoch": 0.4381818181818182,
+      "grad_norm": 3.4630582924124167,
+      "learning_rate": 7.460559784220351e-07,
+      "loss": 0.1629,
+      "step": 1928
+    },
+    {
+      "epoch": 0.4384090909090909,
+      "grad_norm": 2.210582746432757,
+      "learning_rate": 7.456181493210836e-07,
+      "loss": 0.0518,
+      "step": 1929
+    },
+    {
+      "epoch": 0.43863636363636366,
+      "grad_norm": 5.319040981661489,
+      "learning_rate": 7.451802587297721e-07,
+      "loss": 0.1967,
+      "step": 1930
+    },
+    {
+      "epoch": 0.43886363636363634,
+      "grad_norm": 3.880922587593865,
+      "learning_rate": 7.447423068713347e-07,
+      "loss": 0.107,
+      "step": 1931
+    },
+    {
+      "epoch": 0.4390909090909091,
+      "grad_norm": 2.9951566044873696,
+      "learning_rate": 7.443042939690362e-07,
+      "loss": 0.1174,
+      "step": 1932
+    },
+    {
+      "epoch": 0.43931818181818183,
+      "grad_norm": 1.9714404932557235,
+      "learning_rate": 7.438662202461728e-07,
+      "loss": 0.1133,
+      "step": 1933
+    },
+    {
+      "epoch": 0.4395454545454546,
+      "grad_norm": 1.644161048044089,
+      "learning_rate": 7.43428085926072e-07,
+      "loss": 0.0661,
+      "step": 1934
+    },
+    {
+      "epoch": 0.43977272727272726,
+      "grad_norm": 2.976168367768778,
+      "learning_rate": 7.429898912320916e-07,
+      "loss": 0.2082,
+      "step": 1935
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.590516688784981,
+      "learning_rate": 7.425516363876204e-07,
+      "loss": 0.1185,
+      "step": 1936
+    },
+    {
+      "epoch": 0.44022727272727274,
+      "grad_norm": 3.953731594860013,
+      "learning_rate": 7.421133216160781e-07,
+      "loss": 0.1906,
+      "step": 1937
+    },
+    {
+      "epoch": 0.44045454545454543,
+      "grad_norm": 3.6840368779002177,
+      "learning_rate": 7.416749471409144e-07,
+      "loss": 0.1871,
+      "step": 1938
+    },
+    {
+      "epoch": 0.4406818181818182,
+      "grad_norm": 3.040178214163844,
+      "learning_rate": 7.412365131856103e-07,
+      "loss": 0.1956,
+      "step": 1939
+    },
+    {
+      "epoch": 0.4409090909090909,
+      "grad_norm": 2.614751760528119,
+      "learning_rate": 7.407980199736762e-07,
+      "loss": 0.114,
+      "step": 1940
+    },
+    {
+      "epoch": 0.44113636363636366,
+      "grad_norm": 5.324665774301034,
+      "learning_rate": 7.403594677286534e-07,
+      "loss": 0.1383,
+      "step": 1941
+    },
+    {
+      "epoch": 0.44136363636363635,
+      "grad_norm": 2.4768052310395454,
+      "learning_rate": 7.39920856674113e-07,
+      "loss": 0.0718,
+      "step": 1942
+    },
+    {
+      "epoch": 0.4415909090909091,
+      "grad_norm": 1.2132201319901428,
+      "learning_rate": 7.394821870336561e-07,
+      "loss": 0.1034,
+      "step": 1943
+    },
+    {
+      "epoch": 0.44181818181818183,
+      "grad_norm": 2.9134251330010206,
+      "learning_rate": 7.390434590309135e-07,
+      "loss": 0.1778,
+      "step": 1944
+    },
+    {
+      "epoch": 0.4420454545454545,
+      "grad_norm": 3.995701880738082,
+      "learning_rate": 7.386046728895459e-07,
+      "loss": 0.1274,
+      "step": 1945
+    },
+    {
+      "epoch": 0.44227272727272726,
+      "grad_norm": 4.619671516322505,
+      "learning_rate": 7.38165828833244e-07,
+      "loss": 0.1391,
+      "step": 1946
+    },
+    {
+      "epoch": 0.4425,
+      "grad_norm": 3.074952434856251,
+      "learning_rate": 7.377269270857274e-07,
+      "loss": 0.1693,
+      "step": 1947
+    },
+    {
+      "epoch": 0.44272727272727275,
+      "grad_norm": 2.4979339209440012,
+      "learning_rate": 7.372879678707455e-07,
+      "loss": 0.1245,
+      "step": 1948
+    },
+    {
+      "epoch": 0.44295454545454543,
+      "grad_norm": 4.466226759838234,
+      "learning_rate": 7.368489514120768e-07,
+      "loss": 0.157,
+      "step": 1949
+    },
+    {
+      "epoch": 0.4431818181818182,
+      "grad_norm": 2.8214659870058005,
+      "learning_rate": 7.364098779335294e-07,
+      "loss": 0.1066,
+      "step": 1950
+    },
+    {
+      "epoch": 0.4434090909090909,
+      "grad_norm": 3.563774803973594,
+      "learning_rate": 7.359707476589398e-07,
+      "loss": 0.1308,
+      "step": 1951
+    },
+    {
+      "epoch": 0.44363636363636366,
+      "grad_norm": 3.054106229738271,
+      "learning_rate": 7.355315608121736e-07,
+      "loss": 0.1818,
+      "step": 1952
+    },
+    {
+      "epoch": 0.44386363636363635,
+      "grad_norm": 7.267160849996355,
+      "learning_rate": 7.35092317617126e-07,
+      "loss": 0.208,
+      "step": 1953
+    },
+    {
+      "epoch": 0.4440909090909091,
+      "grad_norm": 3.9427206771064958,
+      "learning_rate": 7.346530182977203e-07,
+      "loss": 0.2183,
+      "step": 1954
+    },
+    {
+      "epoch": 0.44431818181818183,
+      "grad_norm": 3.009148905469002,
+      "learning_rate": 7.342136630779082e-07,
+      "loss": 0.2065,
+      "step": 1955
+    },
+    {
+      "epoch": 0.4445454545454545,
+      "grad_norm": 1.9393732038166298,
+      "learning_rate": 7.3377425218167e-07,
+      "loss": 0.0977,
+      "step": 1956
+    },
+    {
+      "epoch": 0.44477272727272726,
+      "grad_norm": 4.084032448387616,
+      "learning_rate": 7.333347858330153e-07,
+      "loss": 0.1706,
+      "step": 1957
+    },
+    {
+      "epoch": 0.445,
+      "grad_norm": 3.4352373855005203,
+      "learning_rate": 7.328952642559804e-07,
+      "loss": 0.2291,
+      "step": 1958
+    },
+    {
+      "epoch": 0.44522727272727275,
+      "grad_norm": 3.403325006643408,
+      "learning_rate": 7.32455687674631e-07,
+      "loss": 0.1478,
+      "step": 1959
+    },
+    {
+      "epoch": 0.44545454545454544,
+      "grad_norm": 3.9891993352782262,
+      "learning_rate": 7.320160563130605e-07,
+      "loss": 0.1592,
+      "step": 1960
+    },
+    {
+      "epoch": 0.4456818181818182,
+      "grad_norm": 3.788494373492283,
+      "learning_rate": 7.3157637039539e-07,
+      "loss": 0.1458,
+      "step": 1961
+    },
+    {
+      "epoch": 0.4459090909090909,
+      "grad_norm": 2.1589433264427944,
+      "learning_rate": 7.311366301457684e-07,
+      "loss": 0.1049,
+      "step": 1962
+    },
+    {
+      "epoch": 0.4461363636363636,
+      "grad_norm": 2.469952680882766,
+      "learning_rate": 7.306968357883729e-07,
+      "loss": 0.1358,
+      "step": 1963
+    },
+    {
+      "epoch": 0.44636363636363635,
+      "grad_norm": 3.043100900833321,
+      "learning_rate": 7.302569875474076e-07,
+      "loss": 0.2296,
+      "step": 1964
+    },
+    {
+      "epoch": 0.4465909090909091,
+      "grad_norm": 1.8866313716679426,
+      "learning_rate": 7.298170856471042e-07,
+      "loss": 0.134,
+      "step": 1965
+    },
+    {
+      "epoch": 0.44681818181818184,
+      "grad_norm": 2.4133781679495176,
+      "learning_rate": 7.29377130311722e-07,
+      "loss": 0.102,
+      "step": 1966
+    },
+    {
+      "epoch": 0.4470454545454545,
+      "grad_norm": 3.344694705931305,
+      "learning_rate": 7.289371217655473e-07,
+      "loss": 0.1023,
+      "step": 1967
+    },
+    {
+      "epoch": 0.44727272727272727,
+      "grad_norm": 3.399873561558365,
+      "learning_rate": 7.28497060232894e-07,
+      "loss": 0.1343,
+      "step": 1968
+    },
+    {
+      "epoch": 0.4475,
+      "grad_norm": 3.198757433265919,
+      "learning_rate": 7.280569459381022e-07,
+      "loss": 0.1814,
+      "step": 1969
+    },
+    {
+      "epoch": 0.44772727272727275,
+      "grad_norm": 1.8099079872575885,
+      "learning_rate": 7.276167791055394e-07,
+      "loss": 0.1186,
+      "step": 1970
+    },
+    {
+      "epoch": 0.44795454545454544,
+      "grad_norm": 2.3240295411261744,
+      "learning_rate": 7.271765599596001e-07,
+      "loss": 0.1782,
+      "step": 1971
+    },
+    {
+      "epoch": 0.4481818181818182,
+      "grad_norm": 3.4714997537766723,
+      "learning_rate": 7.267362887247052e-07,
+      "loss": 0.1446,
+      "step": 1972
+    },
+    {
+      "epoch": 0.4484090909090909,
+      "grad_norm": 2.988588432427571,
+      "learning_rate": 7.262959656253019e-07,
+      "loss": 0.1337,
+      "step": 1973
+    },
+    {
+      "epoch": 0.4486363636363636,
+      "grad_norm": 1.4443971302550012,
+      "learning_rate": 7.258555908858641e-07,
+      "loss": 0.1251,
+      "step": 1974
+    },
+    {
+      "epoch": 0.44886363636363635,
+      "grad_norm": 3.109854900703401,
+      "learning_rate": 7.254151647308924e-07,
+      "loss": 0.1084,
+      "step": 1975
+    },
+    {
+      "epoch": 0.4490909090909091,
+      "grad_norm": 4.012097694453197,
+      "learning_rate": 7.249746873849127e-07,
+      "loss": 0.1734,
+      "step": 1976
+    },
+    {
+      "epoch": 0.44931818181818184,
+      "grad_norm": 2.6981890975033265,
+      "learning_rate": 7.245341590724779e-07,
+      "loss": 0.1887,
+      "step": 1977
+    },
+    {
+      "epoch": 0.4495454545454545,
+      "grad_norm": 2.1028705616038206,
+      "learning_rate": 7.240935800181664e-07,
+      "loss": 0.1417,
+      "step": 1978
+    },
+    {
+      "epoch": 0.44977272727272727,
+      "grad_norm": 2.4302486801697505,
+      "learning_rate": 7.236529504465828e-07,
+      "loss": 0.1359,
+      "step": 1979
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.961063580573303,
+      "learning_rate": 7.232122705823567e-07,
+      "loss": 0.1433,
+      "step": 1980
+    },
+    {
+      "epoch": 0.45022727272727275,
+      "grad_norm": 3.308259895810264,
+      "learning_rate": 7.227715406501444e-07,
+      "loss": 0.2934,
+      "step": 1981
+    },
+    {
+      "epoch": 0.45045454545454544,
+      "grad_norm": 2.318768746444751,
+      "learning_rate": 7.223307608746269e-07,
+      "loss": 0.2105,
+      "step": 1982
+    },
+    {
+      "epoch": 0.4506818181818182,
+      "grad_norm": 2.59795157347462,
+      "learning_rate": 7.21889931480511e-07,
+      "loss": 0.2082,
+      "step": 1983
+    },
+    {
+      "epoch": 0.4509090909090909,
+      "grad_norm": 2.4350953859552362,
+      "learning_rate": 7.214490526925289e-07,
+      "loss": 0.0779,
+      "step": 1984
+    },
+    {
+      "epoch": 0.4511363636363636,
+      "grad_norm": 2.0011209646797714,
+      "learning_rate": 7.210081247354372e-07,
+      "loss": 0.0972,
+      "step": 1985
+    },
+    {
+      "epoch": 0.45136363636363636,
+      "grad_norm": 1.4822685428427618,
+      "learning_rate": 7.205671478340189e-07,
+      "loss": 0.1262,
+      "step": 1986
+    },
+    {
+      "epoch": 0.4515909090909091,
+      "grad_norm": 2.805107079896762,
+      "learning_rate": 7.201261222130807e-07,
+      "loss": 0.1835,
+      "step": 1987
+    },
+    {
+      "epoch": 0.45181818181818184,
+      "grad_norm": 3.762672658405357,
+      "learning_rate": 7.196850480974546e-07,
+      "loss": 0.179,
+      "step": 1988
+    },
+    {
+      "epoch": 0.45204545454545453,
+      "grad_norm": 4.10006244202772,
+      "learning_rate": 7.192439257119976e-07,
+      "loss": 0.1197,
+      "step": 1989
+    },
+    {
+      "epoch": 0.45227272727272727,
+      "grad_norm": 3.013894364636968,
+      "learning_rate": 7.188027552815912e-07,
+      "loss": 0.2218,
+      "step": 1990
+    },
+    {
+      "epoch": 0.4525,
+      "grad_norm": 2.5372500220816008,
+      "learning_rate": 7.183615370311407e-07,
+      "loss": 0.2177,
+      "step": 1991
+    },
+    {
+      "epoch": 0.4527272727272727,
+      "grad_norm": 2.5407539309091907,
+      "learning_rate": 7.179202711855771e-07,
+      "loss": 0.1766,
+      "step": 1992
+    },
+    {
+      "epoch": 0.45295454545454544,
+      "grad_norm": 2.244829257574855,
+      "learning_rate": 7.174789579698542e-07,
+      "loss": 0.1038,
+      "step": 1993
+    },
+    {
+      "epoch": 0.4531818181818182,
+      "grad_norm": 3.1501600532253757,
+      "learning_rate": 7.170375976089509e-07,
+      "loss": 0.1109,
+      "step": 1994
+    },
+    {
+      "epoch": 0.45340909090909093,
+      "grad_norm": 2.2295092878147402,
+      "learning_rate": 7.1659619032787e-07,
+      "loss": 0.1109,
+      "step": 1995
+    },
+    {
+      "epoch": 0.4536363636363636,
+      "grad_norm": 5.78097794801983,
+      "learning_rate": 7.16154736351638e-07,
+      "loss": 0.1539,
+      "step": 1996
+    },
+    {
+      "epoch": 0.45386363636363636,
+      "grad_norm": 3.0487235182478925,
+      "learning_rate": 7.157132359053056e-07,
+      "loss": 0.1539,
+      "step": 1997
+    },
+    {
+      "epoch": 0.4540909090909091,
+      "grad_norm": 2.1592100031156636,
+      "learning_rate": 7.152716892139464e-07,
+      "loss": 0.1151,
+      "step": 1998
+    },
+    {
+      "epoch": 0.45431818181818184,
+      "grad_norm": 2.3096732874108272,
+      "learning_rate": 7.148300965026582e-07,
+      "loss": 0.152,
+      "step": 1999
+    },
+    {
+      "epoch": 0.45454545454545453,
+      "grad_norm": 2.3701913548462934,
+      "learning_rate": 7.143884579965624e-07,
+      "loss": 0.1811,
+      "step": 2000
+    },
+    {
+      "epoch": 0.4547727272727273,
+      "grad_norm": 2.2072330250656083,
+      "learning_rate": 7.139467739208034e-07,
+      "loss": 0.1109,
+      "step": 2001
+    },
+    {
+      "epoch": 0.455,
+      "grad_norm": 2.421704298558306,
+      "learning_rate": 7.135050445005487e-07,
+      "loss": 0.1452,
+      "step": 2002
+    },
+    {
+      "epoch": 0.4552272727272727,
+      "grad_norm": 3.8680326315826434,
+      "learning_rate": 7.130632699609892e-07,
+      "loss": 0.1465,
+      "step": 2003
+    },
+    {
+      "epoch": 0.45545454545454545,
+      "grad_norm": 2.67599201449408,
+      "learning_rate": 7.12621450527339e-07,
+      "loss": 0.2283,
+      "step": 2004
+    },
+    {
+      "epoch": 0.4556818181818182,
+      "grad_norm": 2.1292366674682115,
+      "learning_rate": 7.121795864248345e-07,
+      "loss": 0.0968,
+      "step": 2005
+    },
+    {
+      "epoch": 0.45590909090909093,
+      "grad_norm": 4.050165030199472,
+      "learning_rate": 7.117376778787353e-07,
+      "loss": 0.1485,
+      "step": 2006
+    },
+    {
+      "epoch": 0.4561363636363636,
+      "grad_norm": 2.703185165342989,
+      "learning_rate": 7.112957251143236e-07,
+      "loss": 0.2192,
+      "step": 2007
+    },
+    {
+      "epoch": 0.45636363636363636,
+      "grad_norm": 2.6221057878479077,
+      "learning_rate": 7.10853728356904e-07,
+      "loss": 0.1059,
+      "step": 2008
+    },
+    {
+      "epoch": 0.4565909090909091,
+      "grad_norm": 2.6487703149511836,
+      "learning_rate": 7.104116878318037e-07,
+      "loss": 0.0753,
+      "step": 2009
+    },
+    {
+      "epoch": 0.45681818181818185,
+      "grad_norm": 2.4440481068585687,
+      "learning_rate": 7.09969603764372e-07,
+      "loss": 0.1132,
+      "step": 2010
+    },
+    {
+      "epoch": 0.45704545454545453,
+      "grad_norm": 2.9518855108542015,
+      "learning_rate": 7.095274763799805e-07,
+      "loss": 0.0932,
+      "step": 2011
+    },
+    {
+      "epoch": 0.4572727272727273,
+      "grad_norm": 3.114481397324023,
+      "learning_rate": 7.090853059040234e-07,
+      "loss": 0.1277,
+      "step": 2012
+    },
+    {
+      "epoch": 0.4575,
+      "grad_norm": 2.821323338768452,
+      "learning_rate": 7.086430925619157e-07,
+      "loss": 0.1665,
+      "step": 2013
+    },
+    {
+      "epoch": 0.4577272727272727,
+      "grad_norm": 3.5446553405004573,
+      "learning_rate": 7.082008365790953e-07,
+      "loss": 0.1083,
+      "step": 2014
+    },
+    {
+      "epoch": 0.45795454545454545,
+      "grad_norm": 1.9922237990076304,
+      "learning_rate": 7.077585381810213e-07,
+      "loss": 0.1588,
+      "step": 2015
+    },
+    {
+      "epoch": 0.4581818181818182,
+      "grad_norm": 2.6928699851322664,
+      "learning_rate": 7.073161975931748e-07,
+      "loss": 0.2134,
+      "step": 2016
+    },
+    {
+      "epoch": 0.45840909090909093,
+      "grad_norm": 2.4741148764501295,
+      "learning_rate": 7.068738150410582e-07,
+      "loss": 0.0946,
+      "step": 2017
+    },
+    {
+      "epoch": 0.4586363636363636,
+      "grad_norm": 2.2929471408261093,
+      "learning_rate": 7.064313907501949e-07,
+      "loss": 0.164,
+      "step": 2018
+    },
+    {
+      "epoch": 0.45886363636363636,
+      "grad_norm": 3.19000216377548,
+      "learning_rate": 7.059889249461303e-07,
+      "loss": 0.1243,
+      "step": 2019
+    },
+    {
+      "epoch": 0.4590909090909091,
+      "grad_norm": 3.1597646685496628,
+      "learning_rate": 7.055464178544306e-07,
+      "loss": 0.1888,
+      "step": 2020
+    },
+    {
+      "epoch": 0.4593181818181818,
+      "grad_norm": 3.515705757633226,
+      "learning_rate": 7.051038697006832e-07,
+      "loss": 0.1092,
+      "step": 2021
+    },
+    {
+      "epoch": 0.45954545454545453,
+      "grad_norm": 3.5101373154806206,
+      "learning_rate": 7.04661280710496e-07,
+      "loss": 0.1246,
+      "step": 2022
+    },
+    {
+      "epoch": 0.4597727272727273,
+      "grad_norm": 2.052991433952683,
+      "learning_rate": 7.042186511094981e-07,
+      "loss": 0.1028,
+      "step": 2023
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.592059092861549,
+      "learning_rate": 7.037759811233395e-07,
+      "loss": 0.0925,
+      "step": 2024
+    },
+    {
+      "epoch": 0.4602272727272727,
+      "grad_norm": 2.980901419537529,
+      "learning_rate": 7.033332709776902e-07,
+      "loss": 0.1858,
+      "step": 2025
+    },
+    {
+      "epoch": 0.46045454545454545,
+      "grad_norm": 2.2372928549080515,
+      "learning_rate": 7.028905208982413e-07,
+      "loss": 0.1669,
+      "step": 2026
+    },
+    {
+      "epoch": 0.4606818181818182,
+      "grad_norm": 3.7610013060665124,
+      "learning_rate": 7.024477311107038e-07,
+      "loss": 0.1376,
+      "step": 2027
+    },
+    {
+      "epoch": 0.46090909090909093,
+      "grad_norm": 2.8826226581773917,
+      "learning_rate": 7.02004901840809e-07,
+      "loss": 0.184,
+      "step": 2028
+    },
+    {
+      "epoch": 0.4611363636363636,
+      "grad_norm": 2.4609566478178175,
+      "learning_rate": 7.015620333143085e-07,
+      "loss": 0.1401,
+      "step": 2029
+    },
+    {
+      "epoch": 0.46136363636363636,
+      "grad_norm": 2.9987062782750504,
+      "learning_rate": 7.011191257569739e-07,
+      "loss": 0.1885,
+      "step": 2030
+    },
+    {
+      "epoch": 0.4615909090909091,
+      "grad_norm": 1.5656775718737743,
+      "learning_rate": 7.006761793945968e-07,
+      "loss": 0.0913,
+      "step": 2031
+    },
+    {
+      "epoch": 0.4618181818181818,
+      "grad_norm": 2.7993267961811252,
+      "learning_rate": 7.00233194452988e-07,
+      "loss": 0.1556,
+      "step": 2032
+    },
+    {
+      "epoch": 0.46204545454545454,
+      "grad_norm": 3.090729890733826,
+      "learning_rate": 6.997901711579787e-07,
+      "loss": 0.1311,
+      "step": 2033
+    },
+    {
+      "epoch": 0.4622727272727273,
+      "grad_norm": 2.861838467753398,
+      "learning_rate": 6.993471097354193e-07,
+      "loss": 0.1434,
+      "step": 2034
+    },
+    {
+      "epoch": 0.4625,
+      "grad_norm": 4.870664819498051,
+      "learning_rate": 6.989040104111794e-07,
+      "loss": 0.1254,
+      "step": 2035
+    },
+    {
+      "epoch": 0.4627272727272727,
+      "grad_norm": 3.9114364296854567,
+      "learning_rate": 6.984608734111486e-07,
+      "loss": 0.1579,
+      "step": 2036
+    },
+    {
+      "epoch": 0.46295454545454545,
+      "grad_norm": 1.8842422143052586,
+      "learning_rate": 6.980176989612352e-07,
+      "loss": 0.109,
+      "step": 2037
+    },
+    {
+      "epoch": 0.4631818181818182,
+      "grad_norm": 3.478190777048438,
+      "learning_rate": 6.975744872873663e-07,
+      "loss": 0.1466,
+      "step": 2038
+    },
+    {
+      "epoch": 0.4634090909090909,
+      "grad_norm": 2.3479556941659907,
+      "learning_rate": 6.97131238615489e-07,
+      "loss": 0.1546,
+      "step": 2039
+    },
+    {
+      "epoch": 0.4636363636363636,
+      "grad_norm": 2.6982243197634492,
+      "learning_rate": 6.966879531715681e-07,
+      "loss": 0.1026,
+      "step": 2040
+    },
+    {
+      "epoch": 0.46386363636363637,
+      "grad_norm": 3.226297885342514,
+      "learning_rate": 6.962446311815879e-07,
+      "loss": 0.1093,
+      "step": 2041
+    },
+    {
+      "epoch": 0.4640909090909091,
+      "grad_norm": 1.7843878595628588,
+      "learning_rate": 6.958012728715513e-07,
+      "loss": 0.1224,
+      "step": 2042
+    },
+    {
+      "epoch": 0.4643181818181818,
+      "grad_norm": 3.02702744137971,
+      "learning_rate": 6.953578784674791e-07,
+      "loss": 0.1368,
+      "step": 2043
+    },
+    {
+      "epoch": 0.46454545454545454,
+      "grad_norm": 3.785186481407889,
+      "learning_rate": 6.949144481954114e-07,
+      "loss": 0.2264,
+      "step": 2044
+    },
+    {
+      "epoch": 0.4647727272727273,
+      "grad_norm": 2.0708153531593907,
+      "learning_rate": 6.944709822814058e-07,
+      "loss": 0.1087,
+      "step": 2045
+    },
+    {
+      "epoch": 0.465,
+      "grad_norm": 4.18854237727814,
+      "learning_rate": 6.940274809515384e-07,
+      "loss": 0.1571,
+      "step": 2046
+    },
+    {
+      "epoch": 0.4652272727272727,
+      "grad_norm": 4.542808525708609,
+      "learning_rate": 6.935839444319033e-07,
+      "loss": 0.1496,
+      "step": 2047
+    },
+    {
+      "epoch": 0.46545454545454545,
+      "grad_norm": 4.257316340797779,
+      "learning_rate": 6.931403729486128e-07,
+      "loss": 0.1412,
+      "step": 2048
+    },
+    {
+      "epoch": 0.4656818181818182,
+      "grad_norm": 2.533376866384847,
+      "learning_rate": 6.926967667277967e-07,
+      "loss": 0.1413,
+      "step": 2049
+    },
+    {
+      "epoch": 0.4659090909090909,
+      "grad_norm": 3.112452683306534,
+      "learning_rate": 6.922531259956026e-07,
+      "loss": 0.1704,
+      "step": 2050
+    },
+    {
+      "epoch": 0.4661363636363636,
+      "grad_norm": 4.450806175807696,
+      "learning_rate": 6.918094509781957e-07,
+      "loss": 0.1668,
+      "step": 2051
+    },
+    {
+      "epoch": 0.46636363636363637,
+      "grad_norm": 3.3284912244189155,
+      "learning_rate": 6.913657419017585e-07,
+      "loss": 0.1207,
+      "step": 2052
+    },
+    {
+      "epoch": 0.4665909090909091,
+      "grad_norm": 2.8421533841422586,
+      "learning_rate": 6.909219989924912e-07,
+      "loss": 0.2314,
+      "step": 2053
+    },
+    {
+      "epoch": 0.4668181818181818,
+      "grad_norm": 2.4989909631493625,
+      "learning_rate": 6.904782224766112e-07,
+      "loss": 0.1081,
+      "step": 2054
+    },
+    {
+      "epoch": 0.46704545454545454,
+      "grad_norm": 1.958106033249567,
+      "learning_rate": 6.900344125803529e-07,
+      "loss": 0.1425,
+      "step": 2055
+    },
+    {
+      "epoch": 0.4672727272727273,
+      "grad_norm": 3.954211554838329,
+      "learning_rate": 6.895905695299676e-07,
+      "loss": 0.2188,
+      "step": 2056
+    },
+    {
+      "epoch": 0.4675,
+      "grad_norm": 4.748161542223201,
+      "learning_rate": 6.891466935517237e-07,
+      "loss": 0.1958,
+      "step": 2057
+    },
+    {
+      "epoch": 0.4677272727272727,
+      "grad_norm": 3.255055843820895,
+      "learning_rate": 6.887027848719063e-07,
+      "loss": 0.2633,
+      "step": 2058
+    },
+    {
+      "epoch": 0.46795454545454546,
+      "grad_norm": 2.910345231301499,
+      "learning_rate": 6.882588437168173e-07,
+      "loss": 0.2158,
+      "step": 2059
+    },
+    {
+      "epoch": 0.4681818181818182,
+      "grad_norm": 2.808781233420102,
+      "learning_rate": 6.878148703127751e-07,
+      "loss": 0.2005,
+      "step": 2060
+    },
+    {
+      "epoch": 0.4684090909090909,
+      "grad_norm": 3.603818496878295,
+      "learning_rate": 6.873708648861143e-07,
+      "loss": 0.1378,
+      "step": 2061
+    },
+    {
+      "epoch": 0.46863636363636363,
+      "grad_norm": 3.0208874257134246,
+      "learning_rate": 6.869268276631862e-07,
+      "loss": 0.1471,
+      "step": 2062
+    },
+    {
+      "epoch": 0.46886363636363637,
+      "grad_norm": 1.741127778856493,
+      "learning_rate": 6.86482758870358e-07,
+      "loss": 0.1561,
+      "step": 2063
+    },
+    {
+      "epoch": 0.4690909090909091,
+      "grad_norm": 2.12057623537009,
+      "learning_rate": 6.86038658734013e-07,
+      "loss": 0.1452,
+      "step": 2064
+    },
+    {
+      "epoch": 0.4693181818181818,
+      "grad_norm": 2.265979059672908,
+      "learning_rate": 6.85594527480551e-07,
+      "loss": 0.0977,
+      "step": 2065
+    },
+    {
+      "epoch": 0.46954545454545454,
+      "grad_norm": 2.1248106274236163,
+      "learning_rate": 6.851503653363871e-07,
+      "loss": 0.152,
+      "step": 2066
+    },
+    {
+      "epoch": 0.4697727272727273,
+      "grad_norm": 3.3216661501648193,
+      "learning_rate": 6.847061725279521e-07,
+      "loss": 0.1442,
+      "step": 2067
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.449629920969097,
+      "learning_rate": 6.84261949281693e-07,
+      "loss": 0.159,
+      "step": 2068
+    },
+    {
+      "epoch": 0.4702272727272727,
+      "grad_norm": 2.3489113276186906,
+      "learning_rate": 6.838176958240716e-07,
+      "loss": 0.1064,
+      "step": 2069
+    },
+    {
+      "epoch": 0.47045454545454546,
+      "grad_norm": 1.91536874968937,
+      "learning_rate": 6.833734123815657e-07,
+      "loss": 0.1064,
+      "step": 2070
+    },
+    {
+      "epoch": 0.4706818181818182,
+      "grad_norm": 2.5969586706599257,
+      "learning_rate": 6.829290991806681e-07,
+      "loss": 0.1325,
+      "step": 2071
+    },
+    {
+      "epoch": 0.4709090909090909,
+      "grad_norm": 3.5379325202460574,
+      "learning_rate": 6.824847564478868e-07,
+      "loss": 0.1709,
+      "step": 2072
+    },
+    {
+      "epoch": 0.47113636363636363,
+      "grad_norm": 2.436267540325309,
+      "learning_rate": 6.820403844097449e-07,
+      "loss": 0.1835,
+      "step": 2073
+    },
+    {
+      "epoch": 0.4713636363636364,
+      "grad_norm": 3.7671509503482925,
+      "learning_rate": 6.815959832927805e-07,
+      "loss": 0.1087,
+      "step": 2074
+    },
+    {
+      "epoch": 0.4715909090909091,
+      "grad_norm": 2.377988116578579,
+      "learning_rate": 6.811515533235463e-07,
+      "loss": 0.1269,
+      "step": 2075
+    },
+    {
+      "epoch": 0.4718181818181818,
+      "grad_norm": 3.1066920810909706,
+      "learning_rate": 6.8070709472861e-07,
+      "loss": 0.0904,
+      "step": 2076
+    },
+    {
+      "epoch": 0.47204545454545455,
+      "grad_norm": 3.1080163188785113,
+      "learning_rate": 6.802626077345538e-07,
+      "loss": 0.1421,
+      "step": 2077
+    },
+    {
+      "epoch": 0.4722727272727273,
+      "grad_norm": 2.470266487929162,
+      "learning_rate": 6.798180925679738e-07,
+      "loss": 0.0851,
+      "step": 2078
+    },
+    {
+      "epoch": 0.4725,
+      "grad_norm": 4.033775668036644,
+      "learning_rate": 6.793735494554815e-07,
+      "loss": 0.1643,
+      "step": 2079
+    },
+    {
+      "epoch": 0.4727272727272727,
+      "grad_norm": 3.4229735496188813,
+      "learning_rate": 6.789289786237023e-07,
+      "loss": 0.2344,
+      "step": 2080
+    },
+    {
+      "epoch": 0.47295454545454546,
+      "grad_norm": 4.260003874421788,
+      "learning_rate": 6.78484380299275e-07,
+      "loss": 0.2215,
+      "step": 2081
+    },
+    {
+      "epoch": 0.4731818181818182,
+      "grad_norm": 3.2383530768409,
+      "learning_rate": 6.780397547088531e-07,
+      "loss": 0.3007,
+      "step": 2082
+    },
+    {
+      "epoch": 0.4734090909090909,
+      "grad_norm": 1.9854935494977148,
+      "learning_rate": 6.775951020791043e-07,
+      "loss": 0.1529,
+      "step": 2083
+    },
+    {
+      "epoch": 0.47363636363636363,
+      "grad_norm": 2.564358663395547,
+      "learning_rate": 6.771504226367093e-07,
+      "loss": 0.1685,
+      "step": 2084
+    },
+    {
+      "epoch": 0.4738636363636364,
+      "grad_norm": 3.1362616127462295,
+      "learning_rate": 6.767057166083628e-07,
+      "loss": 0.1546,
+      "step": 2085
+    },
+    {
+      "epoch": 0.4740909090909091,
+      "grad_norm": 2.5637797633065538,
+      "learning_rate": 6.762609842207733e-07,
+      "loss": 0.2369,
+      "step": 2086
+    },
+    {
+      "epoch": 0.4743181818181818,
+      "grad_norm": 2.563902686988228,
+      "learning_rate": 6.758162257006623e-07,
+      "loss": 0.1036,
+      "step": 2087
+    },
+    {
+      "epoch": 0.47454545454545455,
+      "grad_norm": 2.957596045111071,
+      "learning_rate": 6.753714412747652e-07,
+      "loss": 0.2783,
+      "step": 2088
+    },
+    {
+      "epoch": 0.4747727272727273,
+      "grad_norm": 4.306613981763354,
+      "learning_rate": 6.749266311698298e-07,
+      "loss": 0.1225,
+      "step": 2089
+    },
+    {
+      "epoch": 0.475,
+      "grad_norm": 3.6283809298090866,
+      "learning_rate": 6.744817956126176e-07,
+      "loss": 0.1153,
+      "step": 2090
+    },
+    {
+      "epoch": 0.4752272727272727,
+      "grad_norm": 2.6742910964761704,
+      "learning_rate": 6.740369348299033e-07,
+      "loss": 0.1504,
+      "step": 2091
+    },
+    {
+      "epoch": 0.47545454545454546,
+      "grad_norm": 2.012294617288209,
+      "learning_rate": 6.735920490484734e-07,
+      "loss": 0.1372,
+      "step": 2092
+    },
+    {
+      "epoch": 0.4756818181818182,
+      "grad_norm": 2.803522779263299,
+      "learning_rate": 6.731471384951283e-07,
+      "loss": 0.1697,
+      "step": 2093
+    },
+    {
+      "epoch": 0.4759090909090909,
+      "grad_norm": 2.598752325178973,
+      "learning_rate": 6.727022033966804e-07,
+      "loss": 0.12,
+      "step": 2094
+    },
+    {
+      "epoch": 0.47613636363636364,
+      "grad_norm": 1.791834207259415,
+      "learning_rate": 6.722572439799549e-07,
+      "loss": 0.115,
+      "step": 2095
+    },
+    {
+      "epoch": 0.4763636363636364,
+      "grad_norm": 2.428609241805222,
+      "learning_rate": 6.718122604717891e-07,
+      "loss": 0.116,
+      "step": 2096
+    },
+    {
+      "epoch": 0.47659090909090907,
+      "grad_norm": 2.976107693669219,
+      "learning_rate": 6.713672530990327e-07,
+      "loss": 0.1622,
+      "step": 2097
+    },
+    {
+      "epoch": 0.4768181818181818,
+      "grad_norm": 2.7129439645291904,
+      "learning_rate": 6.709222220885478e-07,
+      "loss": 0.0731,
+      "step": 2098
+    },
+    {
+      "epoch": 0.47704545454545455,
+      "grad_norm": 1.602081048711278,
+      "learning_rate": 6.704771676672081e-07,
+      "loss": 0.1274,
+      "step": 2099
+    },
+    {
+      "epoch": 0.4772727272727273,
+      "grad_norm": 4.268225866159453,
+      "learning_rate": 6.700320900618997e-07,
+      "loss": 0.2188,
+      "step": 2100
+    },
+    {
+      "epoch": 0.4775,
+      "grad_norm": 2.477289843646431,
+      "learning_rate": 6.695869894995202e-07,
+      "loss": 0.0919,
+      "step": 2101
+    },
+    {
+      "epoch": 0.4777272727272727,
+      "grad_norm": 3.6264692112432515,
+      "learning_rate": 6.691418662069793e-07,
+      "loss": 0.1038,
+      "step": 2102
+    },
+    {
+      "epoch": 0.47795454545454547,
+      "grad_norm": 3.1138549516047873,
+      "learning_rate": 6.686967204111976e-07,
+      "loss": 0.1755,
+      "step": 2103
+    },
+    {
+      "epoch": 0.4781818181818182,
+      "grad_norm": 5.434394724367757,
+      "learning_rate": 6.682515523391078e-07,
+      "loss": 0.1962,
+      "step": 2104
+    },
+    {
+      "epoch": 0.4784090909090909,
+      "grad_norm": 3.5338537224511306,
+      "learning_rate": 6.678063622176536e-07,
+      "loss": 0.1293,
+      "step": 2105
+    },
+    {
+      "epoch": 0.47863636363636364,
+      "grad_norm": 3.0738832422647464,
+      "learning_rate": 6.673611502737902e-07,
+      "loss": 0.1493,
+      "step": 2106
+    },
+    {
+      "epoch": 0.4788636363636364,
+      "grad_norm": 3.766067028359517,
+      "learning_rate": 6.669159167344839e-07,
+      "loss": 0.1236,
+      "step": 2107
+    },
+    {
+      "epoch": 0.47909090909090907,
+      "grad_norm": 2.910060123178053,
+      "learning_rate": 6.664706618267115e-07,
+      "loss": 0.1421,
+      "step": 2108
+    },
+    {
+      "epoch": 0.4793181818181818,
+      "grad_norm": 2.15471544754446,
+      "learning_rate": 6.660253857774616e-07,
+      "loss": 0.1691,
+      "step": 2109
+    },
+    {
+      "epoch": 0.47954545454545455,
+      "grad_norm": 2.649061810309099,
+      "learning_rate": 6.655800888137327e-07,
+      "loss": 0.1799,
+      "step": 2110
+    },
+    {
+      "epoch": 0.4797727272727273,
+      "grad_norm": 4.3890589534732944,
+      "learning_rate": 6.651347711625344e-07,
+      "loss": 0.1352,
+      "step": 2111
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2636180689350764,
+      "learning_rate": 6.646894330508872e-07,
+      "loss": 0.0801,
+      "step": 2112
+    },
+    {
+      "epoch": 0.4802272727272727,
+      "grad_norm": 3.284650584596919,
+      "learning_rate": 6.642440747058209e-07,
+      "loss": 0.1628,
+      "step": 2113
+    },
+    {
+      "epoch": 0.48045454545454547,
+      "grad_norm": 2.166831121588381,
+      "learning_rate": 6.637986963543769e-07,
+      "loss": 0.0944,
+      "step": 2114
+    },
+    {
+      "epoch": 0.48068181818181815,
+      "grad_norm": 1.5019723355778514,
+      "learning_rate": 6.63353298223606e-07,
+      "loss": 0.0601,
+      "step": 2115
+    },
+    {
+      "epoch": 0.4809090909090909,
+      "grad_norm": 2.037137643018292,
+      "learning_rate": 6.629078805405695e-07,
+      "loss": 0.0899,
+      "step": 2116
+    },
+    {
+      "epoch": 0.48113636363636364,
+      "grad_norm": 2.583109226541136,
+      "learning_rate": 6.62462443532338e-07,
+      "loss": 0.0842,
+      "step": 2117
+    },
+    {
+      "epoch": 0.4813636363636364,
+      "grad_norm": 1.6812393461570792,
+      "learning_rate": 6.62016987425993e-07,
+      "loss": 0.1931,
+      "step": 2118
+    },
+    {
+      "epoch": 0.48159090909090907,
+      "grad_norm": 2.3282473997050617,
+      "learning_rate": 6.615715124486248e-07,
+      "loss": 0.1088,
+      "step": 2119
+    },
+    {
+      "epoch": 0.4818181818181818,
+      "grad_norm": 2.836516803616151,
+      "learning_rate": 6.61126018827334e-07,
+      "loss": 0.2364,
+      "step": 2120
+    },
+    {
+      "epoch": 0.48204545454545455,
+      "grad_norm": 1.5852765789282905,
+      "learning_rate": 6.606805067892301e-07,
+      "loss": 0.1053,
+      "step": 2121
+    },
+    {
+      "epoch": 0.4822727272727273,
+      "grad_norm": 3.224886862209135,
+      "learning_rate": 6.602349765614325e-07,
+      "loss": 0.1559,
+      "step": 2122
+    },
+    {
+      "epoch": 0.4825,
+      "grad_norm": 2.043508349238602,
+      "learning_rate": 6.597894283710692e-07,
+      "loss": 0.1003,
+      "step": 2123
+    },
+    {
+      "epoch": 0.4827272727272727,
+      "grad_norm": 2.40742339247241,
+      "learning_rate": 6.593438624452787e-07,
+      "loss": 0.1317,
+      "step": 2124
+    },
+    {
+      "epoch": 0.48295454545454547,
+      "grad_norm": 2.985094406293537,
+      "learning_rate": 6.588982790112068e-07,
+      "loss": 0.1055,
+      "step": 2125
+    },
+    {
+      "epoch": 0.48318181818181816,
+      "grad_norm": 3.6878170535595842,
+      "learning_rate": 6.584526782960095e-07,
+      "loss": 0.1887,
+      "step": 2126
+    },
+    {
+      "epoch": 0.4834090909090909,
+      "grad_norm": 1.198469763915997,
+      "learning_rate": 6.580070605268513e-07,
+      "loss": 0.0865,
+      "step": 2127
+    },
+    {
+      "epoch": 0.48363636363636364,
+      "grad_norm": 2.127115140406063,
+      "learning_rate": 6.575614259309048e-07,
+      "loss": 0.1585,
+      "step": 2128
+    },
+    {
+      "epoch": 0.4838636363636364,
+      "grad_norm": 2.098470238123538,
+      "learning_rate": 6.571157747353519e-07,
+      "loss": 0.1908,
+      "step": 2129
+    },
+    {
+      "epoch": 0.48409090909090907,
+      "grad_norm": 3.2599457551392534,
+      "learning_rate": 6.56670107167383e-07,
+      "loss": 0.084,
+      "step": 2130
+    },
+    {
+      "epoch": 0.4843181818181818,
+      "grad_norm": 2.676040691342199,
+      "learning_rate": 6.562244234541964e-07,
+      "loss": 0.1002,
+      "step": 2131
+    },
+    {
+      "epoch": 0.48454545454545456,
+      "grad_norm": 2.2656999543542704,
+      "learning_rate": 6.557787238229984e-07,
+      "loss": 0.1267,
+      "step": 2132
+    },
+    {
+      "epoch": 0.4847727272727273,
+      "grad_norm": 5.078632288858154,
+      "learning_rate": 6.553330085010044e-07,
+      "loss": 0.153,
+      "step": 2133
+    },
+    {
+      "epoch": 0.485,
+      "grad_norm": 3.0320603119930105,
+      "learning_rate": 6.548872777154369e-07,
+      "loss": 0.1288,
+      "step": 2134
+    },
+    {
+      "epoch": 0.48522727272727273,
+      "grad_norm": 3.628775296290297,
+      "learning_rate": 6.544415316935267e-07,
+      "loss": 0.1009,
+      "step": 2135
+    },
+    {
+      "epoch": 0.48545454545454547,
+      "grad_norm": 2.09259380302618,
+      "learning_rate": 6.539957706625123e-07,
+      "loss": 0.128,
+      "step": 2136
+    },
+    {
+      "epoch": 0.48568181818181816,
+      "grad_norm": 2.815095245151195,
+      "learning_rate": 6.535499948496396e-07,
+      "loss": 0.115,
+      "step": 2137
+    },
+    {
+      "epoch": 0.4859090909090909,
+      "grad_norm": 1.436132096421439,
+      "learning_rate": 6.531042044821625e-07,
+      "loss": 0.1419,
+      "step": 2138
+    },
+    {
+      "epoch": 0.48613636363636364,
+      "grad_norm": 2.777740481513346,
+      "learning_rate": 6.526583997873422e-07,
+      "loss": 0.1676,
+      "step": 2139
+    },
+    {
+      "epoch": 0.4863636363636364,
+      "grad_norm": 3.2849291416014474,
+      "learning_rate": 6.522125809924465e-07,
+      "loss": 0.1134,
+      "step": 2140
+    },
+    {
+      "epoch": 0.4865909090909091,
+      "grad_norm": 2.5224424050433196,
+      "learning_rate": 6.517667483247517e-07,
+      "loss": 0.1984,
+      "step": 2141
+    },
+    {
+      "epoch": 0.4868181818181818,
+      "grad_norm": 4.704003132588,
+      "learning_rate": 6.513209020115399e-07,
+      "loss": 0.1201,
+      "step": 2142
+    },
+    {
+      "epoch": 0.48704545454545456,
+      "grad_norm": 1.9990667279843302,
+      "learning_rate": 6.508750422801006e-07,
+      "loss": 0.0916,
+      "step": 2143
+    },
+    {
+      "epoch": 0.48727272727272725,
+      "grad_norm": 2.185621124633731,
+      "learning_rate": 6.504291693577309e-07,
+      "loss": 0.0949,
+      "step": 2144
+    },
+    {
+      "epoch": 0.4875,
+      "grad_norm": 2.022060365990564,
+      "learning_rate": 6.499832834717335e-07,
+      "loss": 0.0917,
+      "step": 2145
+    },
+    {
+      "epoch": 0.48772727272727273,
+      "grad_norm": 3.0444255340570074,
+      "learning_rate": 6.495373848494179e-07,
+      "loss": 0.1657,
+      "step": 2146
+    },
+    {
+      "epoch": 0.4879545454545455,
+      "grad_norm": 3.4491154041741354,
+      "learning_rate": 6.49091473718101e-07,
+      "loss": 0.2628,
+      "step": 2147
+    },
+    {
+      "epoch": 0.48818181818181816,
+      "grad_norm": 1.76693877009867,
+      "learning_rate": 6.48645550305105e-07,
+      "loss": 0.1175,
+      "step": 2148
+    },
+    {
+      "epoch": 0.4884090909090909,
+      "grad_norm": 1.58263981456577,
+      "learning_rate": 6.48199614837759e-07,
+      "loss": 0.1439,
+      "step": 2149
+    },
+    {
+      "epoch": 0.48863636363636365,
+      "grad_norm": 3.950401784661815,
+      "learning_rate": 6.477536675433979e-07,
+      "loss": 0.1731,
+      "step": 2150
+    },
+    {
+      "epoch": 0.4888636363636364,
+      "grad_norm": 3.048410820502032,
+      "learning_rate": 6.473077086493629e-07,
+      "loss": 0.1134,
+      "step": 2151
+    },
+    {
+      "epoch": 0.4890909090909091,
+      "grad_norm": 3.635503670727867,
+      "learning_rate": 6.468617383830007e-07,
+      "loss": 0.1778,
+      "step": 2152
+    },
+    {
+      "epoch": 0.4893181818181818,
+      "grad_norm": 2.853489446537876,
+      "learning_rate": 6.464157569716648e-07,
+      "loss": 0.1357,
+      "step": 2153
+    },
+    {
+      "epoch": 0.48954545454545456,
+      "grad_norm": 3.058415419377323,
+      "learning_rate": 6.459697646427128e-07,
+      "loss": 0.164,
+      "step": 2154
+    },
+    {
+      "epoch": 0.48977272727272725,
+      "grad_norm": 2.7048723575232327,
+      "learning_rate": 6.455237616235091e-07,
+      "loss": 0.1036,
+      "step": 2155
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.604178552793881,
+      "learning_rate": 6.450777481414236e-07,
+      "loss": 0.1458,
+      "step": 2156
+    },
+    {
+      "epoch": 0.49022727272727273,
+      "grad_norm": 3.905931365483569,
+      "learning_rate": 6.446317244238302e-07,
+      "loss": 0.2202,
+      "step": 2157
+    },
+    {
+      "epoch": 0.4904545454545455,
+      "grad_norm": 3.494766333323408,
+      "learning_rate": 6.441856906981096e-07,
+      "loss": 0.1417,
+      "step": 2158
+    },
+    {
+      "epoch": 0.49068181818181816,
+      "grad_norm": 2.718387339638051,
+      "learning_rate": 6.437396471916469e-07,
+      "loss": 0.0919,
+      "step": 2159
+    },
+    {
+      "epoch": 0.4909090909090909,
+      "grad_norm": 2.5115677238954226,
+      "learning_rate": 6.432935941318322e-07,
+      "loss": 0.1437,
+      "step": 2160
+    },
+    {
+      "epoch": 0.49113636363636365,
+      "grad_norm": 3.3454972529522986,
+      "learning_rate": 6.428475317460601e-07,
+      "loss": 0.091,
+      "step": 2161
+    },
+    {
+      "epoch": 0.4913636363636364,
+      "grad_norm": 3.8046062413133552,
+      "learning_rate": 6.424014602617309e-07,
+      "loss": 0.1257,
+      "step": 2162
+    },
+    {
+      "epoch": 0.4915909090909091,
+      "grad_norm": 3.160997160357057,
+      "learning_rate": 6.419553799062486e-07,
+      "loss": 0.2776,
+      "step": 2163
+    },
+    {
+      "epoch": 0.4918181818181818,
+      "grad_norm": 2.41082853491261,
+      "learning_rate": 6.415092909070224e-07,
+      "loss": 0.2202,
+      "step": 2164
+    },
+    {
+      "epoch": 0.49204545454545456,
+      "grad_norm": 3.963167415656143,
+      "learning_rate": 6.410631934914653e-07,
+      "loss": 0.2066,
+      "step": 2165
+    },
+    {
+      "epoch": 0.49227272727272725,
+      "grad_norm": 4.0430432582135,
+      "learning_rate": 6.406170878869951e-07,
+      "loss": 0.207,
+      "step": 2166
+    },
+    {
+      "epoch": 0.4925,
+      "grad_norm": 3.6242838272941658,
+      "learning_rate": 6.401709743210335e-07,
+      "loss": 0.1111,
+      "step": 2167
+    },
+    {
+      "epoch": 0.49272727272727274,
+      "grad_norm": 3.285278781776353,
+      "learning_rate": 6.397248530210065e-07,
+      "loss": 0.143,
+      "step": 2168
+    },
+    {
+      "epoch": 0.4929545454545455,
+      "grad_norm": 2.064553384081473,
+      "learning_rate": 6.392787242143437e-07,
+      "loss": 0.1498,
+      "step": 2169
+    },
+    {
+      "epoch": 0.49318181818181817,
+      "grad_norm": 2.657281543656906,
+      "learning_rate": 6.388325881284785e-07,
+      "loss": 0.0935,
+      "step": 2170
+    },
+    {
+      "epoch": 0.4934090909090909,
+      "grad_norm": 3.7433441112894665,
+      "learning_rate": 6.383864449908489e-07,
+      "loss": 0.1125,
+      "step": 2171
+    },
+    {
+      "epoch": 0.49363636363636365,
+      "grad_norm": 3.4124527425872975,
+      "learning_rate": 6.379402950288952e-07,
+      "loss": 0.1225,
+      "step": 2172
+    },
+    {
+      "epoch": 0.49386363636363634,
+      "grad_norm": 2.2147324012996306,
+      "learning_rate": 6.374941384700618e-07,
+      "loss": 0.1618,
+      "step": 2173
+    },
+    {
+      "epoch": 0.4940909090909091,
+      "grad_norm": 4.092493130156925,
+      "learning_rate": 6.37047975541797e-07,
+      "loss": 0.1548,
+      "step": 2174
+    },
+    {
+      "epoch": 0.4943181818181818,
+      "grad_norm": 3.073060205528055,
+      "learning_rate": 6.366018064715512e-07,
+      "loss": 0.1256,
+      "step": 2175
+    },
+    {
+      "epoch": 0.49454545454545457,
+      "grad_norm": 3.507870329446146,
+      "learning_rate": 6.361556314867789e-07,
+      "loss": 0.1372,
+      "step": 2176
+    },
+    {
+      "epoch": 0.49477272727272725,
+      "grad_norm": 2.153954943384018,
+      "learning_rate": 6.35709450814937e-07,
+      "loss": 0.0898,
+      "step": 2177
+    },
+    {
+      "epoch": 0.495,
+      "grad_norm": 2.257308881332642,
+      "learning_rate": 6.352632646834858e-07,
+      "loss": 0.194,
+      "step": 2178
+    },
+    {
+      "epoch": 0.49522727272727274,
+      "grad_norm": 3.7154103180222062,
+      "learning_rate": 6.34817073319888e-07,
+      "loss": 0.2176,
+      "step": 2179
+    },
+    {
+      "epoch": 0.4954545454545455,
+      "grad_norm": 3.276963226186806,
+      "learning_rate": 6.34370876951609e-07,
+      "loss": 0.2527,
+      "step": 2180
+    },
+    {
+      "epoch": 0.49568181818181817,
+      "grad_norm": 3.870410697079271,
+      "learning_rate": 6.339246758061171e-07,
+      "loss": 0.2754,
+      "step": 2181
+    },
+    {
+      "epoch": 0.4959090909090909,
+      "grad_norm": 3.4409553287457872,
+      "learning_rate": 6.334784701108826e-07,
+      "loss": 0.2294,
+      "step": 2182
+    },
+    {
+      "epoch": 0.49613636363636365,
+      "grad_norm": 2.8510841965439035,
+      "learning_rate": 6.330322600933784e-07,
+      "loss": 0.1315,
+      "step": 2183
+    },
+    {
+      "epoch": 0.49636363636363634,
+      "grad_norm": 2.819984510784436,
+      "learning_rate": 6.325860459810793e-07,
+      "loss": 0.2212,
+      "step": 2184
+    },
+    {
+      "epoch": 0.4965909090909091,
+      "grad_norm": 2.270558263838988,
+      "learning_rate": 6.321398280014627e-07,
+      "loss": 0.1409,
+      "step": 2185
+    },
+    {
+      "epoch": 0.4968181818181818,
+      "grad_norm": 3.7327662672036244,
+      "learning_rate": 6.316936063820075e-07,
+      "loss": 0.1242,
+      "step": 2186
+    },
+    {
+      "epoch": 0.49704545454545457,
+      "grad_norm": 2.58124327181106,
+      "learning_rate": 6.312473813501946e-07,
+      "loss": 0.1474,
+      "step": 2187
+    },
+    {
+      "epoch": 0.49727272727272726,
+      "grad_norm": 3.0499079534788125,
+      "learning_rate": 6.308011531335068e-07,
+      "loss": 0.0857,
+      "step": 2188
+    },
+    {
+      "epoch": 0.4975,
+      "grad_norm": 3.64864955260816,
+      "learning_rate": 6.303549219594282e-07,
+      "loss": 0.1425,
+      "step": 2189
+    },
+    {
+      "epoch": 0.49772727272727274,
+      "grad_norm": 2.6681640812353926,
+      "learning_rate": 6.299086880554446e-07,
+      "loss": 0.1201,
+      "step": 2190
+    },
+    {
+      "epoch": 0.4979545454545454,
+      "grad_norm": 3.338120076468721,
+      "learning_rate": 6.294624516490435e-07,
+      "loss": 0.2023,
+      "step": 2191
+    },
+    {
+      "epoch": 0.49818181818181817,
+      "grad_norm": 3.2834731487169413,
+      "learning_rate": 6.290162129677131e-07,
+      "loss": 0.1193,
+      "step": 2192
+    },
+    {
+      "epoch": 0.4984090909090909,
+      "grad_norm": 4.335023471962203,
+      "learning_rate": 6.285699722389428e-07,
+      "loss": 0.1956,
+      "step": 2193
+    },
+    {
+      "epoch": 0.49863636363636366,
+      "grad_norm": 4.21556656957727,
+      "learning_rate": 6.281237296902237e-07,
+      "loss": 0.1772,
+      "step": 2194
+    },
+    {
+      "epoch": 0.49886363636363634,
+      "grad_norm": 3.635062755704252,
+      "learning_rate": 6.276774855490472e-07,
+      "loss": 0.1136,
+      "step": 2195
+    },
+    {
+      "epoch": 0.4990909090909091,
+      "grad_norm": 0.8267094932276462,
+      "learning_rate": 6.272312400429057e-07,
+      "loss": 0.0621,
+      "step": 2196
+    },
+    {
+      "epoch": 0.4993181818181818,
+      "grad_norm": 2.581134328280723,
+      "learning_rate": 6.267849933992926e-07,
+      "loss": 0.162,
+      "step": 2197
+    },
+    {
+      "epoch": 0.49954545454545457,
+      "grad_norm": 1.6839011822659404,
+      "learning_rate": 6.263387458457009e-07,
+      "loss": 0.0722,
+      "step": 2198
+    },
+    {
+      "epoch": 0.49977272727272726,
+      "grad_norm": 2.0901263638690084,
+      "learning_rate": 6.258924976096252e-07,
+      "loss": 0.1791,
+      "step": 2199
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.509523688536928,
+      "learning_rate": 6.254462489185601e-07,
+      "loss": 0.223,
+      "step": 2200
+    }
+  ],
+  "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": 7302513819648.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}