diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,116897 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 12.49529544599172,
+  "eval_steps": 200,
+  "global_step": 16600,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0007527286413248024,
+      "grad_norm": 1.2880518436431885,
+      "learning_rate": 1.0000000000000001e-07,
+      "loss": 0.7823,
+      "step": 1
+    },
+    {
+      "epoch": 0.0015054572826496049,
+      "grad_norm": 1.8163784742355347,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 0.8738,
+      "step": 2
+    },
+    {
+      "epoch": 0.002258185923974407,
+      "grad_norm": 2.494739294052124,
+      "learning_rate": 3.0000000000000004e-07,
+      "loss": 0.9478,
+      "step": 3
+    },
+    {
+      "epoch": 0.0030109145652992097,
+      "grad_norm": 2.0457160472869873,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 0.9511,
+      "step": 4
+    },
+    {
+      "epoch": 0.003763643206624012,
+      "grad_norm": 1.032782793045044,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 0.8914,
+      "step": 5
+    },
+    {
+      "epoch": 0.004516371847948814,
+      "grad_norm": 1.7965949773788452,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 0.8714,
+      "step": 6
+    },
+    {
+      "epoch": 0.005269100489273617,
+      "grad_norm": 0.7104043364524841,
+      "learning_rate": 7.000000000000001e-07,
+      "loss": 0.8028,
+      "step": 7
+    },
+    {
+      "epoch": 0.0060218291305984195,
+      "grad_norm": 0.5705575346946716,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 0.8202,
+      "step": 8
+    },
+    {
+      "epoch": 0.006774557771923222,
+      "grad_norm": 1.9678016901016235,
+      "learning_rate": 9e-07,
+      "loss": 0.991,
+      "step": 9
+    },
+    {
+      "epoch": 0.007527286413248024,
+      "grad_norm": 0.40648579597473145,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.786,
+      "step": 10
+    },
+    {
+      "epoch": 0.008280015054572827,
+      "grad_norm": 0.5186575055122375,
+      "learning_rate": 1.1e-06,
+      "loss": 0.835,
+      "step": 11
+    },
+    {
+      "epoch": 0.009032743695897629,
+      "grad_norm": 1.6391487121582031,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 0.9946,
+      "step": 12
+    },
+    {
+      "epoch": 0.00978547233722243,
+      "grad_norm": 0.3144531846046448,
+      "learning_rate": 1.3e-06,
+      "loss": 0.7438,
+      "step": 13
+    },
+    {
+      "epoch": 0.010538200978547234,
+      "grad_norm": 0.3235611021518707,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 0.8498,
+      "step": 14
+    },
+    {
+      "epoch": 0.011290929619872036,
+      "grad_norm": 0.3014790117740631,
+      "learning_rate": 1.5e-06,
+      "loss": 0.8944,
+      "step": 15
+    },
+    {
+      "epoch": 0.012043658261196839,
+      "grad_norm": 0.2660958468914032,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 0.8206,
+      "step": 16
+    },
+    {
+      "epoch": 0.01279638690252164,
+      "grad_norm": 0.2764747738838196,
+      "learning_rate": 1.7000000000000002e-06,
+      "loss": 0.8607,
+      "step": 17
+    },
+    {
+      "epoch": 0.013549115543846444,
+      "grad_norm": 0.22859959304332733,
+      "learning_rate": 1.8e-06,
+      "loss": 0.821,
+      "step": 18
+    },
+    {
+      "epoch": 0.014301844185171246,
+      "grad_norm": 0.24396902322769165,
+      "learning_rate": 1.9e-06,
+      "loss": 0.8708,
+      "step": 19
+    },
+    {
+      "epoch": 0.015054572826496047,
+      "grad_norm": 0.6231729388237,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.9384,
+      "step": 20
+    },
+    {
+      "epoch": 0.01580730146782085,
+      "grad_norm": 0.22251251339912415,
+      "learning_rate": 2.1000000000000002e-06,
+      "loss": 0.838,
+      "step": 21
+    },
+    {
+      "epoch": 0.016560030109145654,
+      "grad_norm": 0.20208397507667542,
+      "learning_rate": 2.2e-06,
+      "loss": 0.7812,
+      "step": 22
+    },
+    {
+      "epoch": 0.017312758750470454,
+      "grad_norm": 0.5645434856414795,
+      "learning_rate": 2.3e-06,
+      "loss": 0.8663,
+      "step": 23
+    },
+    {
+      "epoch": 0.018065487391795258,
+      "grad_norm": 0.6268227696418762,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 0.9997,
+      "step": 24
+    },
+    {
+      "epoch": 0.01881821603312006,
+      "grad_norm": 0.39490875601768494,
+      "learning_rate": 2.5e-06,
+      "loss": 0.8503,
+      "step": 25
+    },
+    {
+      "epoch": 0.01957094467444486,
+      "grad_norm": 0.18386338651180267,
+      "learning_rate": 2.6e-06,
+      "loss": 0.7921,
+      "step": 26
+    },
+    {
+      "epoch": 0.020323673315769664,
+      "grad_norm": 0.17685608565807343,
+      "learning_rate": 2.7e-06,
+      "loss": 0.7881,
+      "step": 27
+    },
+    {
+      "epoch": 0.021076401957094468,
+      "grad_norm": 0.18325509130954742,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 0.8395,
+      "step": 28
+    },
+    {
+      "epoch": 0.02182913059841927,
+      "grad_norm": 1.7831342220306396,
+      "learning_rate": 2.9e-06,
+      "loss": 0.9414,
+      "step": 29
+    },
+    {
+      "epoch": 0.02258185923974407,
+      "grad_norm": 0.28467094898223877,
+      "learning_rate": 3e-06,
+      "loss": 0.8875,
+      "step": 30
+    },
+    {
+      "epoch": 0.023334587881068874,
+      "grad_norm": 0.16373880207538605,
+      "learning_rate": 3.1e-06,
+      "loss": 0.7999,
+      "step": 31
+    },
+    {
+      "epoch": 0.024087316522393678,
+      "grad_norm": 0.16707804799079895,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 0.7456,
+      "step": 32
+    },
+    {
+      "epoch": 0.024840045163718478,
+      "grad_norm": 0.14742392301559448,
+      "learning_rate": 3.3e-06,
+      "loss": 0.774,
+      "step": 33
+    },
+    {
+      "epoch": 0.02559277380504328,
+      "grad_norm": 0.6995845437049866,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 0.8997,
+      "step": 34
+    },
+    {
+      "epoch": 0.026345502446368085,
+      "grad_norm": 0.16255199909210205,
+      "learning_rate": 3.5000000000000004e-06,
+      "loss": 0.8128,
+      "step": 35
+    },
+    {
+      "epoch": 0.027098231087692888,
+      "grad_norm": 0.1572292149066925,
+      "learning_rate": 3.6e-06,
+      "loss": 0.82,
+      "step": 36
+    },
+    {
+      "epoch": 0.027850959729017688,
+      "grad_norm": 0.699761688709259,
+      "learning_rate": 3.7e-06,
+      "loss": 0.8373,
+      "step": 37
+    },
+    {
+      "epoch": 0.02860368837034249,
+      "grad_norm": 0.3444806635379791,
+      "learning_rate": 3.8e-06,
+      "loss": 0.8785,
+      "step": 38
+    },
+    {
+      "epoch": 0.029356417011667295,
+      "grad_norm": 0.1429043412208557,
+      "learning_rate": 3.9e-06,
+      "loss": 0.7459,
+      "step": 39
+    },
+    {
+      "epoch": 0.030109145652992095,
+      "grad_norm": 0.14159370958805084,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.7186,
+      "step": 40
+    },
+    {
+      "epoch": 0.030861874294316898,
+      "grad_norm": 0.1396176964044571,
+      "learning_rate": 4.1000000000000006e-06,
+      "loss": 0.7969,
+      "step": 41
+    },
+    {
+      "epoch": 0.0316146029356417,
+      "grad_norm": 0.15050068497657776,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 0.7257,
+      "step": 42
+    },
+    {
+      "epoch": 0.0323673315769665,
+      "grad_norm": 0.1629023402929306,
+      "learning_rate": 4.2999999999999995e-06,
+      "loss": 0.7728,
+      "step": 43
+    },
+    {
+      "epoch": 0.03312006021829131,
+      "grad_norm": 0.14176440238952637,
+      "learning_rate": 4.4e-06,
+      "loss": 0.7238,
+      "step": 44
+    },
+    {
+      "epoch": 0.03387278885961611,
+      "grad_norm": 0.288448691368103,
+      "learning_rate": 4.5e-06,
+      "loss": 0.8687,
+      "step": 45
+    },
+    {
+      "epoch": 0.03462551750094091,
+      "grad_norm": 0.25777673721313477,
+      "learning_rate": 4.6e-06,
+      "loss": 0.833,
+      "step": 46
+    },
+    {
+      "epoch": 0.035378246142265715,
+      "grad_norm": 0.1412581354379654,
+      "learning_rate": 4.7e-06,
+      "loss": 0.8374,
+      "step": 47
+    },
+    {
+      "epoch": 0.036130974783590515,
+      "grad_norm": 0.18266607820987701,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 0.8651,
+      "step": 48
+    },
+    {
+      "epoch": 0.036883703424915315,
+      "grad_norm": 0.15940433740615845,
+      "learning_rate": 4.9000000000000005e-06,
+      "loss": 0.8278,
+      "step": 49
+    },
+    {
+      "epoch": 0.03763643206624012,
+      "grad_norm": 0.13876453042030334,
+      "learning_rate": 5e-06,
+      "loss": 0.7565,
+      "step": 50
+    },
+    {
+      "epoch": 0.03838916070756492,
+      "grad_norm": 0.5756680369377136,
+      "learning_rate": 5.1e-06,
+      "loss": 0.8546,
+      "step": 51
+    },
+    {
+      "epoch": 0.03914188934888972,
+      "grad_norm": 0.140261709690094,
+      "learning_rate": 5.2e-06,
+      "loss": 0.7957,
+      "step": 52
+    },
+    {
+      "epoch": 0.03989461799021453,
+      "grad_norm": 0.1516708880662918,
+      "learning_rate": 5.3e-06,
+      "loss": 0.8793,
+      "step": 53
+    },
+    {
+      "epoch": 0.04064734663153933,
+      "grad_norm": 0.41485652327537537,
+      "learning_rate": 5.4e-06,
+      "loss": 0.9336,
+      "step": 54
+    },
+    {
+      "epoch": 0.041400075272864136,
+      "grad_norm": 0.14740435779094696,
+      "learning_rate": 5.500000000000001e-06,
+      "loss": 0.8412,
+      "step": 55
+    },
+    {
+      "epoch": 0.042152803914188935,
+      "grad_norm": 0.14479561150074005,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 0.8324,
+      "step": 56
+    },
+    {
+      "epoch": 0.042905532555513735,
+      "grad_norm": 0.15082165598869324,
+      "learning_rate": 5.7000000000000005e-06,
+      "loss": 0.8145,
+      "step": 57
+    },
+    {
+      "epoch": 0.04365826119683854,
+      "grad_norm": 0.14537198841571808,
+      "learning_rate": 5.8e-06,
+      "loss": 0.7889,
+      "step": 58
+    },
+    {
+      "epoch": 0.04441098983816334,
+      "grad_norm": 0.14270208775997162,
+      "learning_rate": 5.9e-06,
+      "loss": 0.8102,
+      "step": 59
+    },
+    {
+      "epoch": 0.04516371847948814,
+      "grad_norm": 0.13857746124267578,
+      "learning_rate": 6e-06,
+      "loss": 0.8533,
+      "step": 60
+    },
+    {
+      "epoch": 0.04591644712081295,
+      "grad_norm": 0.14320848882198334,
+      "learning_rate": 6.1e-06,
+      "loss": 0.8438,
+      "step": 61
+    },
+    {
+      "epoch": 0.04666917576213775,
+      "grad_norm": 0.1336829662322998,
+      "learning_rate": 6.2e-06,
+      "loss": 0.7643,
+      "step": 62
+    },
+    {
+      "epoch": 0.04742190440346255,
+      "grad_norm": 0.14195971190929413,
+      "learning_rate": 6.300000000000001e-06,
+      "loss": 0.7709,
+      "step": 63
+    },
+    {
+      "epoch": 0.048174633044787356,
+      "grad_norm": 0.16386157274246216,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 0.8631,
+      "step": 64
+    },
+    {
+      "epoch": 0.048927361686112156,
+      "grad_norm": 0.15252570807933807,
+      "learning_rate": 6.5000000000000004e-06,
+      "loss": 0.7381,
+      "step": 65
+    },
+    {
+      "epoch": 0.049680090327436956,
+      "grad_norm": 0.14671289920806885,
+      "learning_rate": 6.6e-06,
+      "loss": 0.8669,
+      "step": 66
+    },
+    {
+      "epoch": 0.05043281896876176,
+      "grad_norm": 0.17090877890586853,
+      "learning_rate": 6.700000000000001e-06,
+      "loss": 0.8606,
+      "step": 67
+    },
+    {
+      "epoch": 0.05118554761008656,
+      "grad_norm": 0.308032751083374,
+      "learning_rate": 6.800000000000001e-06,
+      "loss": 0.8722,
+      "step": 68
+    },
+    {
+      "epoch": 0.05193827625141137,
+      "grad_norm": 0.14494512975215912,
+      "learning_rate": 6.900000000000001e-06,
+      "loss": 0.7873,
+      "step": 69
+    },
+    {
+      "epoch": 0.05269100489273617,
+      "grad_norm": 0.17014320194721222,
+      "learning_rate": 7.000000000000001e-06,
+      "loss": 0.8577,
+      "step": 70
+    },
+    {
+      "epoch": 0.05344373353406097,
+      "grad_norm": 0.1427452713251114,
+      "learning_rate": 7.1e-06,
+      "loss": 0.7335,
+      "step": 71
+    },
+    {
+      "epoch": 0.054196462175385776,
+      "grad_norm": 0.14351849257946014,
+      "learning_rate": 7.2e-06,
+      "loss": 0.7942,
+      "step": 72
+    },
+    {
+      "epoch": 0.054949190816710576,
+      "grad_norm": 0.1352851539850235,
+      "learning_rate": 7.2999999999999996e-06,
+      "loss": 0.7531,
+      "step": 73
+    },
+    {
+      "epoch": 0.055701919458035376,
+      "grad_norm": 0.12690991163253784,
+      "learning_rate": 7.4e-06,
+      "loss": 0.7935,
+      "step": 74
+    },
+    {
+      "epoch": 0.05645464809936018,
+      "grad_norm": 0.16269463300704956,
+      "learning_rate": 7.5e-06,
+      "loss": 0.8128,
+      "step": 75
+    },
+    {
+      "epoch": 0.05720737674068498,
+      "grad_norm": 0.14380714297294617,
+      "learning_rate": 7.6e-06,
+      "loss": 0.8429,
+      "step": 76
+    },
+    {
+      "epoch": 0.05796010538200978,
+      "grad_norm": 0.22281956672668457,
+      "learning_rate": 7.7e-06,
+      "loss": 0.7622,
+      "step": 77
+    },
+    {
+      "epoch": 0.05871283402333459,
+      "grad_norm": 0.42543408274650574,
+      "learning_rate": 7.8e-06,
+      "loss": 0.8851,
+      "step": 78
+    },
+    {
+      "epoch": 0.05946556266465939,
+      "grad_norm": 0.15329134464263916,
+      "learning_rate": 7.9e-06,
+      "loss": 0.8114,
+      "step": 79
+    },
+    {
+      "epoch": 0.06021829130598419,
+      "grad_norm": 0.12981769442558289,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.7169,
+      "step": 80
+    },
+    {
+      "epoch": 0.060971019947308996,
+      "grad_norm": 0.25072187185287476,
+      "learning_rate": 8.1e-06,
+      "loss": 0.8372,
+      "step": 81
+    },
+    {
+      "epoch": 0.061723748588633796,
+      "grad_norm": 0.14205990731716156,
+      "learning_rate": 8.200000000000001e-06,
+      "loss": 0.8284,
+      "step": 82
+    },
+    {
+      "epoch": 0.0624764772299586,
+      "grad_norm": 0.3990846574306488,
+      "learning_rate": 8.3e-06,
+      "loss": 0.9115,
+      "step": 83
+    },
+    {
+      "epoch": 0.0632292058712834,
+      "grad_norm": 0.15232448279857635,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 0.7717,
+      "step": 84
+    },
+    {
+      "epoch": 0.06398193451260821,
+      "grad_norm": 0.15215279161930084,
+      "learning_rate": 8.500000000000002e-06,
+      "loss": 0.7581,
+      "step": 85
+    },
+    {
+      "epoch": 0.064734663153933,
+      "grad_norm": 0.38551467657089233,
+      "learning_rate": 8.599999999999999e-06,
+      "loss": 0.8764,
+      "step": 86
+    },
+    {
+      "epoch": 0.06548739179525781,
+      "grad_norm": 0.2729915976524353,
+      "learning_rate": 8.7e-06,
+      "loss": 0.781,
+      "step": 87
+    },
+    {
+      "epoch": 0.06624012043658262,
+      "grad_norm": 0.25350648164749146,
+      "learning_rate": 8.8e-06,
+      "loss": 0.8897,
+      "step": 88
+    },
+    {
+      "epoch": 0.06699284907790741,
+      "grad_norm": 0.18596331775188446,
+      "learning_rate": 8.9e-06,
+      "loss": 0.7855,
+      "step": 89
+    },
+    {
+      "epoch": 0.06774557771923222,
+      "grad_norm": 0.14812487363815308,
+      "learning_rate": 9e-06,
+      "loss": 0.8229,
+      "step": 90
+    },
+    {
+      "epoch": 0.06849830636055702,
+      "grad_norm": 0.14137974381446838,
+      "learning_rate": 9.100000000000001e-06,
+      "loss": 0.7723,
+      "step": 91
+    },
+    {
+      "epoch": 0.06925103500188182,
+      "grad_norm": 0.15291085839271545,
+      "learning_rate": 9.2e-06,
+      "loss": 0.7826,
+      "step": 92
+    },
+    {
+      "epoch": 0.07000376364320662,
+      "grad_norm": 0.16587583720684052,
+      "learning_rate": 9.3e-06,
+      "loss": 0.8022,
+      "step": 93
+    },
+    {
+      "epoch": 0.07075649228453143,
+      "grad_norm": 0.21662834286689758,
+      "learning_rate": 9.4e-06,
+      "loss": 0.8172,
+      "step": 94
+    },
+    {
+      "epoch": 0.07150922092585622,
+      "grad_norm": 0.2848075032234192,
+      "learning_rate": 9.5e-06,
+      "loss": 0.8201,
+      "step": 95
+    },
+    {
+      "epoch": 0.07226194956718103,
+      "grad_norm": 0.13457900285720825,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 0.7187,
+      "step": 96
+    },
+    {
+      "epoch": 0.07301467820850584,
+      "grad_norm": 0.15222664177417755,
+      "learning_rate": 9.7e-06,
+      "loss": 0.8344,
+      "step": 97
+    },
+    {
+      "epoch": 0.07376740684983063,
+      "grad_norm": 0.17529496550559998,
+      "learning_rate": 9.800000000000001e-06,
+      "loss": 0.758,
+      "step": 98
+    },
+    {
+      "epoch": 0.07452013549115544,
+      "grad_norm": 0.4716678261756897,
+      "learning_rate": 9.900000000000002e-06,
+      "loss": 0.8636,
+      "step": 99
+    },
+    {
+      "epoch": 0.07527286413248024,
+      "grad_norm": 0.14991092681884766,
+      "learning_rate": 1e-05,
+      "loss": 0.7215,
+      "step": 100
+    },
+    {
+      "epoch": 0.07602559277380504,
+      "grad_norm": 0.3012496829032898,
+      "learning_rate": 1.0100000000000002e-05,
+      "loss": 0.8816,
+      "step": 101
+    },
+    {
+      "epoch": 0.07677832141512984,
+      "grad_norm": 0.1505427211523056,
+      "learning_rate": 1.02e-05,
+      "loss": 0.7679,
+      "step": 102
+    },
+    {
+      "epoch": 0.07753105005645465,
+      "grad_norm": 0.23428232967853546,
+      "learning_rate": 1.03e-05,
+      "loss": 0.8094,
+      "step": 103
+    },
+    {
+      "epoch": 0.07828377869777944,
+      "grad_norm": 0.3106619715690613,
+      "learning_rate": 1.04e-05,
+      "loss": 0.77,
+      "step": 104
+    },
+    {
+      "epoch": 0.07903650733910425,
+      "grad_norm": 0.16180823743343353,
+      "learning_rate": 1.05e-05,
+      "loss": 0.7591,
+      "step": 105
+    },
+    {
+      "epoch": 0.07978923598042906,
+      "grad_norm": 0.27914944291114807,
+      "learning_rate": 1.06e-05,
+      "loss": 0.8302,
+      "step": 106
+    },
+    {
+      "epoch": 0.08054196462175386,
+      "grad_norm": 0.1637905240058899,
+      "learning_rate": 1.0700000000000001e-05,
+      "loss": 0.8439,
+      "step": 107
+    },
+    {
+      "epoch": 0.08129469326307866,
+      "grad_norm": 0.24072307348251343,
+      "learning_rate": 1.08e-05,
+      "loss": 0.8366,
+      "step": 108
+    },
+    {
+      "epoch": 0.08204742190440346,
+      "grad_norm": 0.21605928242206573,
+      "learning_rate": 1.09e-05,
+      "loss": 0.7957,
+      "step": 109
+    },
+    {
+      "epoch": 0.08280015054572827,
+      "grad_norm": 0.19771215319633484,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 0.8043,
+      "step": 110
+    },
+    {
+      "epoch": 0.08355287918705306,
+      "grad_norm": 0.15618528425693512,
+      "learning_rate": 1.11e-05,
+      "loss": 0.7549,
+      "step": 111
+    },
+    {
+      "epoch": 0.08430560782837787,
+      "grad_norm": 0.14586134254932404,
+      "learning_rate": 1.1200000000000001e-05,
+      "loss": 0.7454,
+      "step": 112
+    },
+    {
+      "epoch": 0.08505833646970268,
+      "grad_norm": 0.16426309943199158,
+      "learning_rate": 1.13e-05,
+      "loss": 0.7442,
+      "step": 113
+    },
+    {
+      "epoch": 0.08581106511102747,
+      "grad_norm": 0.14691542088985443,
+      "learning_rate": 1.1400000000000001e-05,
+      "loss": 0.7676,
+      "step": 114
+    },
+    {
+      "epoch": 0.08656379375235228,
+      "grad_norm": 0.19522573053836823,
+      "learning_rate": 1.1500000000000002e-05,
+      "loss": 0.7879,
+      "step": 115
+    },
+    {
+      "epoch": 0.08731652239367708,
+      "grad_norm": 0.1404934972524643,
+      "learning_rate": 1.16e-05,
+      "loss": 0.77,
+      "step": 116
+    },
+    {
+      "epoch": 0.08806925103500188,
+      "grad_norm": 0.3941081464290619,
+      "learning_rate": 1.1700000000000001e-05,
+      "loss": 0.8153,
+      "step": 117
+    },
+    {
+      "epoch": 0.08882197967632668,
+      "grad_norm": 0.17220336198806763,
+      "learning_rate": 1.18e-05,
+      "loss": 0.8306,
+      "step": 118
+    },
+    {
+      "epoch": 0.08957470831765149,
+      "grad_norm": 0.1557903289794922,
+      "learning_rate": 1.19e-05,
+      "loss": 0.8241,
+      "step": 119
+    },
+    {
+      "epoch": 0.09032743695897628,
+      "grad_norm": 0.16852201521396637,
+      "learning_rate": 1.2e-05,
+      "loss": 0.758,
+      "step": 120
+    },
+    {
+      "epoch": 0.09108016560030109,
+      "grad_norm": 0.14205065369606018,
+      "learning_rate": 1.2100000000000001e-05,
+      "loss": 0.7208,
+      "step": 121
+    },
+    {
+      "epoch": 0.0918328942416259,
+      "grad_norm": 0.16026628017425537,
+      "learning_rate": 1.22e-05,
+      "loss": 0.794,
+      "step": 122
+    },
+    {
+      "epoch": 0.09258562288295069,
+      "grad_norm": 0.13958723843097687,
+      "learning_rate": 1.23e-05,
+      "loss": 0.7186,
+      "step": 123
+    },
+    {
+      "epoch": 0.0933383515242755,
+      "grad_norm": 0.1458817571401596,
+      "learning_rate": 1.24e-05,
+      "loss": 0.7142,
+      "step": 124
+    },
+    {
+      "epoch": 0.0940910801656003,
+      "grad_norm": 0.14147421717643738,
+      "learning_rate": 1.25e-05,
+      "loss": 0.6896,
+      "step": 125
+    },
+    {
+      "epoch": 0.0948438088069251,
+      "grad_norm": 0.15755358338356018,
+      "learning_rate": 1.2600000000000001e-05,
+      "loss": 0.8019,
+      "step": 126
+    },
+    {
+      "epoch": 0.0955965374482499,
+      "grad_norm": 0.5085943937301636,
+      "learning_rate": 1.27e-05,
+      "loss": 0.8993,
+      "step": 127
+    },
+    {
+      "epoch": 0.09634926608957471,
+      "grad_norm": 0.15199662744998932,
+      "learning_rate": 1.2800000000000001e-05,
+      "loss": 0.7047,
+      "step": 128
+    },
+    {
+      "epoch": 0.0971019947308995,
+      "grad_norm": 0.17794065177440643,
+      "learning_rate": 1.29e-05,
+      "loss": 0.7882,
+      "step": 129
+    },
+    {
+      "epoch": 0.09785472337222431,
+      "grad_norm": 0.147978276014328,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 0.7284,
+      "step": 130
+    },
+    {
+      "epoch": 0.09860745201354912,
+      "grad_norm": 0.14350520074367523,
+      "learning_rate": 1.3100000000000002e-05,
+      "loss": 0.7707,
+      "step": 131
+    },
+    {
+      "epoch": 0.09936018065487391,
+      "grad_norm": 0.15746036171913147,
+      "learning_rate": 1.32e-05,
+      "loss": 0.7557,
+      "step": 132
+    },
+    {
+      "epoch": 0.10011290929619872,
+      "grad_norm": 0.14988337457180023,
+      "learning_rate": 1.3300000000000001e-05,
+      "loss": 0.834,
+      "step": 133
+    },
+    {
+      "epoch": 0.10086563793752353,
+      "grad_norm": 0.15569047629833221,
+      "learning_rate": 1.3400000000000002e-05,
+      "loss": 0.6771,
+      "step": 134
+    },
+    {
+      "epoch": 0.10161836657884833,
+      "grad_norm": 0.8520597815513611,
+      "learning_rate": 1.3500000000000001e-05,
+      "loss": 0.7913,
+      "step": 135
+    },
+    {
+      "epoch": 0.10237109522017313,
+      "grad_norm": 0.1925939917564392,
+      "learning_rate": 1.3600000000000002e-05,
+      "loss": 0.773,
+      "step": 136
+    },
+    {
+      "epoch": 0.10312382386149793,
+      "grad_norm": 0.15118294954299927,
+      "learning_rate": 1.3700000000000001e-05,
+      "loss": 0.7545,
+      "step": 137
+    },
+    {
+      "epoch": 0.10387655250282274,
+      "grad_norm": 0.16856883466243744,
+      "learning_rate": 1.3800000000000002e-05,
+      "loss": 0.8053,
+      "step": 138
+    },
+    {
+      "epoch": 0.10462928114414753,
+      "grad_norm": 0.16572517156600952,
+      "learning_rate": 1.3900000000000002e-05,
+      "loss": 0.8149,
+      "step": 139
+    },
+    {
+      "epoch": 0.10538200978547234,
+      "grad_norm": 0.3106241524219513,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 0.8135,
+      "step": 140
+    },
+    {
+      "epoch": 0.10613473842679715,
+      "grad_norm": 0.1630554497241974,
+      "learning_rate": 1.4099999999999999e-05,
+      "loss": 0.7533,
+      "step": 141
+    },
+    {
+      "epoch": 0.10688746706812194,
+      "grad_norm": 0.4348670542240143,
+      "learning_rate": 1.42e-05,
+      "loss": 0.7416,
+      "step": 142
+    },
+    {
+      "epoch": 0.10764019570944675,
+      "grad_norm": 0.15953247249126434,
+      "learning_rate": 1.43e-05,
+      "loss": 0.7811,
+      "step": 143
+    },
+    {
+      "epoch": 0.10839292435077155,
+      "grad_norm": 0.22225669026374817,
+      "learning_rate": 1.44e-05,
+      "loss": 0.7465,
+      "step": 144
+    },
+    {
+      "epoch": 0.10914565299209635,
+      "grad_norm": 0.5987780690193176,
+      "learning_rate": 1.45e-05,
+      "loss": 0.8198,
+      "step": 145
+    },
+    {
+      "epoch": 0.10989838163342115,
+      "grad_norm": 0.18760189414024353,
+      "learning_rate": 1.4599999999999999e-05,
+      "loss": 0.8116,
+      "step": 146
+    },
+    {
+      "epoch": 0.11065111027474596,
+      "grad_norm": 0.15552322566509247,
+      "learning_rate": 1.47e-05,
+      "loss": 0.8181,
+      "step": 147
+    },
+    {
+      "epoch": 0.11140383891607075,
+      "grad_norm": 0.26814016699790955,
+      "learning_rate": 1.48e-05,
+      "loss": 0.8368,
+      "step": 148
+    },
+    {
+      "epoch": 0.11215656755739556,
+      "grad_norm": 0.1618763655424118,
+      "learning_rate": 1.49e-05,
+      "loss": 0.766,
+      "step": 149
+    },
+    {
+      "epoch": 0.11290929619872037,
+      "grad_norm": 0.21285472810268402,
+      "learning_rate": 1.5e-05,
+      "loss": 0.8691,
+      "step": 150
+    },
+    {
+      "epoch": 0.11366202484004516,
+      "grad_norm": 0.19716131687164307,
+      "learning_rate": 1.51e-05,
+      "loss": 0.7853,
+      "step": 151
+    },
+    {
+      "epoch": 0.11441475348136997,
+      "grad_norm": 0.2098015695810318,
+      "learning_rate": 1.52e-05,
+      "loss": 0.695,
+      "step": 152
+    },
+    {
+      "epoch": 0.11516748212269477,
+      "grad_norm": 0.30609405040740967,
+      "learning_rate": 1.53e-05,
+      "loss": 0.8289,
+      "step": 153
+    },
+    {
+      "epoch": 0.11592021076401957,
+      "grad_norm": 0.22784261405467987,
+      "learning_rate": 1.54e-05,
+      "loss": 0.7764,
+      "step": 154
+    },
+    {
+      "epoch": 0.11667293940534437,
+      "grad_norm": 0.16312339901924133,
+      "learning_rate": 1.55e-05,
+      "loss": 0.6869,
+      "step": 155
+    },
+    {
+      "epoch": 0.11742566804666918,
+      "grad_norm": 0.19938555359840393,
+      "learning_rate": 1.56e-05,
+      "loss": 0.8284,
+      "step": 156
+    },
+    {
+      "epoch": 0.11817839668799397,
+      "grad_norm": 0.1370309740304947,
+      "learning_rate": 1.5700000000000002e-05,
+      "loss": 0.7271,
+      "step": 157
+    },
+    {
+      "epoch": 0.11893112532931878,
+      "grad_norm": 0.21091033518314362,
+      "learning_rate": 1.58e-05,
+      "loss": 0.7073,
+      "step": 158
+    },
+    {
+      "epoch": 0.11968385397064359,
+      "grad_norm": 0.24756523966789246,
+      "learning_rate": 1.59e-05,
+      "loss": 0.6774,
+      "step": 159
+    },
+    {
+      "epoch": 0.12043658261196838,
+      "grad_norm": 0.8255650401115417,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.8784,
+      "step": 160
+    },
+    {
+      "epoch": 0.12118931125329319,
+      "grad_norm": 0.19278767704963684,
+      "learning_rate": 1.6100000000000002e-05,
+      "loss": 0.7375,
+      "step": 161
+    },
+    {
+      "epoch": 0.12194203989461799,
+      "grad_norm": 0.16455627977848053,
+      "learning_rate": 1.62e-05,
+      "loss": 0.7121,
+      "step": 162
+    },
+    {
+      "epoch": 0.12269476853594279,
+      "grad_norm": 0.18811045587062836,
+      "learning_rate": 1.63e-05,
+      "loss": 0.7323,
+      "step": 163
+    },
+    {
+      "epoch": 0.12344749717726759,
+      "grad_norm": 0.17033560574054718,
+      "learning_rate": 1.6400000000000002e-05,
+      "loss": 0.8154,
+      "step": 164
+    },
+    {
+      "epoch": 0.1242002258185924,
+      "grad_norm": 0.19222772121429443,
+      "learning_rate": 1.65e-05,
+      "loss": 0.7196,
+      "step": 165
+    },
+    {
+      "epoch": 0.1249529544599172,
+      "grad_norm": 0.16930893063545227,
+      "learning_rate": 1.66e-05,
+      "loss": 0.693,
+      "step": 166
+    },
+    {
+      "epoch": 0.125705683101242,
+      "grad_norm": 0.1914365291595459,
+      "learning_rate": 1.6700000000000003e-05,
+      "loss": 0.7652,
+      "step": 167
+    },
+    {
+      "epoch": 0.1264584117425668,
+      "grad_norm": 0.17743763327598572,
+      "learning_rate": 1.6800000000000002e-05,
+      "loss": 0.7536,
+      "step": 168
+    },
+    {
+      "epoch": 0.1272111403838916,
+      "grad_norm": 0.20731930434703827,
+      "learning_rate": 1.69e-05,
+      "loss": 0.7396,
+      "step": 169
+    },
+    {
+      "epoch": 0.12796386902521642,
+      "grad_norm": 0.15932516753673553,
+      "learning_rate": 1.7000000000000003e-05,
+      "loss": 0.7266,
+      "step": 170
+    },
+    {
+      "epoch": 0.1287165976665412,
+      "grad_norm": 0.18592903017997742,
+      "learning_rate": 1.7100000000000002e-05,
+      "loss": 0.7141,
+      "step": 171
+    },
+    {
+      "epoch": 0.129469326307866,
+      "grad_norm": 0.6350232362747192,
+      "learning_rate": 1.7199999999999998e-05,
+      "loss": 0.8,
+      "step": 172
+    },
+    {
+      "epoch": 0.1302220549491908,
+      "grad_norm": 0.2283705770969391,
+      "learning_rate": 1.73e-05,
+      "loss": 0.7777,
+      "step": 173
+    },
+    {
+      "epoch": 0.13097478359051562,
+      "grad_norm": 0.21548736095428467,
+      "learning_rate": 1.74e-05,
+      "loss": 0.807,
+      "step": 174
+    },
+    {
+      "epoch": 0.13172751223184043,
+      "grad_norm": 0.2221834510564804,
+      "learning_rate": 1.75e-05,
+      "loss": 0.6913,
+      "step": 175
+    },
+    {
+      "epoch": 0.13248024087316523,
+      "grad_norm": 0.20740960538387299,
+      "learning_rate": 1.76e-05,
+      "loss": 0.7712,
+      "step": 176
+    },
+    {
+      "epoch": 0.13323296951449,
+      "grad_norm": 0.17932602763175964,
+      "learning_rate": 1.77e-05,
+      "loss": 0.7791,
+      "step": 177
+    },
+    {
+      "epoch": 0.13398569815581482,
+      "grad_norm": 0.20678380131721497,
+      "learning_rate": 1.78e-05,
+      "loss": 0.8095,
+      "step": 178
+    },
+    {
+      "epoch": 0.13473842679713963,
+      "grad_norm": 0.18143022060394287,
+      "learning_rate": 1.79e-05,
+      "loss": 0.7204,
+      "step": 179
+    },
+    {
+      "epoch": 0.13549115543846443,
+      "grad_norm": 0.20877574384212494,
+      "learning_rate": 1.8e-05,
+      "loss": 0.7318,
+      "step": 180
+    },
+    {
+      "epoch": 0.13624388407978924,
+      "grad_norm": 0.49373695254325867,
+      "learning_rate": 1.81e-05,
+      "loss": 0.8068,
+      "step": 181
+    },
+    {
+      "epoch": 0.13699661272111405,
+      "grad_norm": 0.22442682087421417,
+      "learning_rate": 1.8200000000000002e-05,
+      "loss": 0.7826,
+      "step": 182
+    },
+    {
+      "epoch": 0.13774934136243885,
+      "grad_norm": 0.2147788256406784,
+      "learning_rate": 1.83e-05,
+      "loss": 0.7872,
+      "step": 183
+    },
+    {
+      "epoch": 0.13850207000376363,
+      "grad_norm": 0.21269994974136353,
+      "learning_rate": 1.84e-05,
+      "loss": 0.7469,
+      "step": 184
+    },
+    {
+      "epoch": 0.13925479864508844,
+      "grad_norm": 0.16783151030540466,
+      "learning_rate": 1.85e-05,
+      "loss": 0.7086,
+      "step": 185
+    },
+    {
+      "epoch": 0.14000752728641325,
+      "grad_norm": 0.18291029334068298,
+      "learning_rate": 1.86e-05,
+      "loss": 0.7892,
+      "step": 186
+    },
+    {
+      "epoch": 0.14076025592773805,
+      "grad_norm": 0.15406300127506256,
+      "learning_rate": 1.87e-05,
+      "loss": 0.7796,
+      "step": 187
+    },
+    {
+      "epoch": 0.14151298456906286,
+      "grad_norm": 0.21659930050373077,
+      "learning_rate": 1.88e-05,
+      "loss": 0.7545,
+      "step": 188
+    },
+    {
+      "epoch": 0.14226571321038767,
+      "grad_norm": 0.19015911221504211,
+      "learning_rate": 1.8900000000000002e-05,
+      "loss": 0.7593,
+      "step": 189
+    },
+    {
+      "epoch": 0.14301844185171245,
+      "grad_norm": 0.2072194665670395,
+      "learning_rate": 1.9e-05,
+      "loss": 0.6939,
+      "step": 190
+    },
+    {
+      "epoch": 0.14377117049303725,
+      "grad_norm": 0.4294150471687317,
+      "learning_rate": 1.91e-05,
+      "loss": 0.7961,
+      "step": 191
+    },
+    {
+      "epoch": 0.14452389913436206,
+      "grad_norm": 0.2609142065048218,
+      "learning_rate": 1.9200000000000003e-05,
+      "loss": 0.8123,
+      "step": 192
+    },
+    {
+      "epoch": 0.14527662777568687,
+      "grad_norm": 0.17543676495552063,
+      "learning_rate": 1.93e-05,
+      "loss": 0.6475,
+      "step": 193
+    },
+    {
+      "epoch": 0.14602935641701167,
+      "grad_norm": 0.1896914392709732,
+      "learning_rate": 1.94e-05,
+      "loss": 0.7135,
+      "step": 194
+    },
+    {
+      "epoch": 0.14678208505833648,
+      "grad_norm": 0.41836807131767273,
+      "learning_rate": 1.9500000000000003e-05,
+      "loss": 0.7376,
+      "step": 195
+    },
+    {
+      "epoch": 0.14753481369966126,
+      "grad_norm": 0.15788064897060394,
+      "learning_rate": 1.9600000000000002e-05,
+      "loss": 0.6803,
+      "step": 196
+    },
+    {
+      "epoch": 0.14828754234098607,
+      "grad_norm": 0.23239544034004211,
+      "learning_rate": 1.97e-05,
+      "loss": 0.7856,
+      "step": 197
+    },
+    {
+      "epoch": 0.14904027098231087,
+      "grad_norm": 0.22440968453884125,
+      "learning_rate": 1.9800000000000004e-05,
+      "loss": 0.7233,
+      "step": 198
+    },
+    {
+      "epoch": 0.14979299962363568,
+      "grad_norm": 0.3315947949886322,
+      "learning_rate": 1.9900000000000003e-05,
+      "loss": 0.7357,
+      "step": 199
+    },
+    {
+      "epoch": 0.1505457282649605,
+      "grad_norm": 0.2221190631389618,
+      "learning_rate": 2e-05,
+      "loss": 0.7165,
+      "step": 200
+    },
+    {
+      "epoch": 0.1505457282649605,
+      "eval_loss": 0.7371044158935547,
+      "eval_runtime": 452.379,
+      "eval_samples_per_second": 21.281,
+      "eval_steps_per_second": 0.665,
+      "step": 200
+    },
+    {
+      "epoch": 0.1512984569062853,
+      "grad_norm": 0.24009129405021667,
+      "learning_rate": 2.01e-05,
+      "loss": 0.8662,
+      "step": 201
+    },
+    {
+      "epoch": 0.15205118554761007,
+      "grad_norm": 0.19073617458343506,
+      "learning_rate": 2.0200000000000003e-05,
+      "loss": 0.7552,
+      "step": 202
+    },
+    {
+      "epoch": 0.15280391418893488,
+      "grad_norm": 0.18954817950725555,
+      "learning_rate": 2.0300000000000002e-05,
+      "loss": 0.7761,
+      "step": 203
+    },
+    {
+      "epoch": 0.1535566428302597,
+      "grad_norm": 0.2223181128501892,
+      "learning_rate": 2.04e-05,
+      "loss": 0.7146,
+      "step": 204
+    },
+    {
+      "epoch": 0.1543093714715845,
+      "grad_norm": 0.1961170881986618,
+      "learning_rate": 2.05e-05,
+      "loss": 0.7324,
+      "step": 205
+    },
+    {
+      "epoch": 0.1550621001129093,
+      "grad_norm": 0.22145985066890717,
+      "learning_rate": 2.06e-05,
+      "loss": 0.8165,
+      "step": 206
+    },
+    {
+      "epoch": 0.1558148287542341,
+      "grad_norm": 0.2125367522239685,
+      "learning_rate": 2.07e-05,
+      "loss": 0.8038,
+      "step": 207
+    },
+    {
+      "epoch": 0.1565675573955589,
+      "grad_norm": 0.20995981991291046,
+      "learning_rate": 2.08e-05,
+      "loss": 0.7708,
+      "step": 208
+    },
+    {
+      "epoch": 0.1573202860368837,
+      "grad_norm": 0.19881388545036316,
+      "learning_rate": 2.09e-05,
+      "loss": 0.7643,
+      "step": 209
+    },
+    {
+      "epoch": 0.1580730146782085,
+      "grad_norm": 0.28778818249702454,
+      "learning_rate": 2.1e-05,
+      "loss": 0.6933,
+      "step": 210
+    },
+    {
+      "epoch": 0.1588257433195333,
+      "grad_norm": 0.27259284257888794,
+      "learning_rate": 2.11e-05,
+      "loss": 0.8056,
+      "step": 211
+    },
+    {
+      "epoch": 0.15957847196085811,
+      "grad_norm": 0.19166716933250427,
+      "learning_rate": 2.12e-05,
+      "loss": 0.8022,
+      "step": 212
+    },
+    {
+      "epoch": 0.16033120060218292,
+      "grad_norm": 0.18588681519031525,
+      "learning_rate": 2.13e-05,
+      "loss": 0.7806,
+      "step": 213
+    },
+    {
+      "epoch": 0.16108392924350773,
+      "grad_norm": 0.183628648519516,
+      "learning_rate": 2.1400000000000002e-05,
+      "loss": 0.6891,
+      "step": 214
+    },
+    {
+      "epoch": 0.1618366578848325,
+      "grad_norm": 0.1857215315103531,
+      "learning_rate": 2.15e-05,
+      "loss": 0.7207,
+      "step": 215
+    },
+    {
+      "epoch": 0.16258938652615731,
+      "grad_norm": 0.20978286862373352,
+      "learning_rate": 2.16e-05,
+      "loss": 0.7401,
+      "step": 216
+    },
+    {
+      "epoch": 0.16334211516748212,
+      "grad_norm": 0.23566651344299316,
+      "learning_rate": 2.1700000000000002e-05,
+      "loss": 0.7046,
+      "step": 217
+    },
+    {
+      "epoch": 0.16409484380880693,
+      "grad_norm": 0.1831624060869217,
+      "learning_rate": 2.18e-05,
+      "loss": 0.6956,
+      "step": 218
+    },
+    {
+      "epoch": 0.16484757245013174,
+      "grad_norm": 0.22268442809581757,
+      "learning_rate": 2.19e-05,
+      "loss": 0.7848,
+      "step": 219
+    },
+    {
+      "epoch": 0.16560030109145654,
+      "grad_norm": 0.2552737295627594,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 0.7905,
+      "step": 220
+    },
+    {
+      "epoch": 0.16635302973278132,
+      "grad_norm": 0.23639385402202606,
+      "learning_rate": 2.2100000000000002e-05,
+      "loss": 0.6735,
+      "step": 221
+    },
+    {
+      "epoch": 0.16710575837410613,
+      "grad_norm": 0.1930161714553833,
+      "learning_rate": 2.22e-05,
+      "loss": 0.7089,
+      "step": 222
+    },
+    {
+      "epoch": 0.16785848701543093,
+      "grad_norm": 0.21174860000610352,
+      "learning_rate": 2.23e-05,
+      "loss": 0.7545,
+      "step": 223
+    },
+    {
+      "epoch": 0.16861121565675574,
+      "grad_norm": 0.20672214031219482,
+      "learning_rate": 2.2400000000000002e-05,
+      "loss": 0.7607,
+      "step": 224
+    },
+    {
+      "epoch": 0.16936394429808055,
+      "grad_norm": 0.2115974724292755,
+      "learning_rate": 2.25e-05,
+      "loss": 0.7577,
+      "step": 225
+    },
+    {
+      "epoch": 0.17011667293940536,
+      "grad_norm": 0.2118208408355713,
+      "learning_rate": 2.26e-05,
+      "loss": 0.7582,
+      "step": 226
+    },
+    {
+      "epoch": 0.17086940158073013,
+      "grad_norm": 0.32249540090560913,
+      "learning_rate": 2.2700000000000003e-05,
+      "loss": 0.8349,
+      "step": 227
+    },
+    {
+      "epoch": 0.17162213022205494,
+      "grad_norm": 0.23135042190551758,
+      "learning_rate": 2.2800000000000002e-05,
+      "loss": 0.7329,
+      "step": 228
+    },
+    {
+      "epoch": 0.17237485886337975,
+      "grad_norm": 0.20479941368103027,
+      "learning_rate": 2.29e-05,
+      "loss": 0.7158,
+      "step": 229
+    },
+    {
+      "epoch": 0.17312758750470456,
+      "grad_norm": 0.1859694868326187,
+      "learning_rate": 2.3000000000000003e-05,
+      "loss": 0.6773,
+      "step": 230
+    },
+    {
+      "epoch": 0.17388031614602936,
+      "grad_norm": 0.2070022076368332,
+      "learning_rate": 2.3100000000000002e-05,
+      "loss": 0.7142,
+      "step": 231
+    },
+    {
+      "epoch": 0.17463304478735417,
+      "grad_norm": 0.2626158893108368,
+      "learning_rate": 2.32e-05,
+      "loss": 0.7538,
+      "step": 232
+    },
+    {
+      "epoch": 0.17538577342867895,
+      "grad_norm": 0.20770740509033203,
+      "learning_rate": 2.3300000000000004e-05,
+      "loss": 0.7163,
+      "step": 233
+    },
+    {
+      "epoch": 0.17613850207000376,
+      "grad_norm": 0.26602864265441895,
+      "learning_rate": 2.3400000000000003e-05,
+      "loss": 0.7,
+      "step": 234
+    },
+    {
+      "epoch": 0.17689123071132856,
+      "grad_norm": 0.3950898349285126,
+      "learning_rate": 2.35e-05,
+      "loss": 0.7412,
+      "step": 235
+    },
+    {
+      "epoch": 0.17764395935265337,
+      "grad_norm": 0.26385870575904846,
+      "learning_rate": 2.36e-05,
+      "loss": 0.7065,
+      "step": 236
+    },
+    {
+      "epoch": 0.17839668799397818,
+      "grad_norm": 0.3582912087440491,
+      "learning_rate": 2.37e-05,
+      "loss": 0.6751,
+      "step": 237
+    },
+    {
+      "epoch": 0.17914941663530298,
+      "grad_norm": 0.39984986186027527,
+      "learning_rate": 2.38e-05,
+      "loss": 0.7505,
+      "step": 238
+    },
+    {
+      "epoch": 0.1799021452766278,
+      "grad_norm": 0.25496816635131836,
+      "learning_rate": 2.39e-05,
+      "loss": 0.7376,
+      "step": 239
+    },
+    {
+      "epoch": 0.18065487391795257,
+      "grad_norm": 0.35112592577934265,
+      "learning_rate": 2.4e-05,
+      "loss": 0.6409,
+      "step": 240
+    },
+    {
+      "epoch": 0.18140760255927738,
+      "grad_norm": 0.2467922419309616,
+      "learning_rate": 2.41e-05,
+      "loss": 0.7484,
+      "step": 241
+    },
+    {
+      "epoch": 0.18216033120060218,
+      "grad_norm": 0.2587096393108368,
+      "learning_rate": 2.4200000000000002e-05,
+      "loss": 0.7498,
+      "step": 242
+    },
+    {
+      "epoch": 0.182913059841927,
+      "grad_norm": 0.1878439337015152,
+      "learning_rate": 2.43e-05,
+      "loss": 0.7245,
+      "step": 243
+    },
+    {
+      "epoch": 0.1836657884832518,
+      "grad_norm": 0.35808852314949036,
+      "learning_rate": 2.44e-05,
+      "loss": 0.7227,
+      "step": 244
+    },
+    {
+      "epoch": 0.1844185171245766,
+      "grad_norm": 0.2932528257369995,
+      "learning_rate": 2.45e-05,
+      "loss": 0.7091,
+      "step": 245
+    },
+    {
+      "epoch": 0.18517124576590138,
+      "grad_norm": 0.17734315991401672,
+      "learning_rate": 2.46e-05,
+      "loss": 0.7625,
+      "step": 246
+    },
+    {
+      "epoch": 0.1859239744072262,
+      "grad_norm": 0.21011503040790558,
+      "learning_rate": 2.47e-05,
+      "loss": 0.6622,
+      "step": 247
+    },
+    {
+      "epoch": 0.186676703048551,
+      "grad_norm": 0.2446037083864212,
+      "learning_rate": 2.48e-05,
+      "loss": 0.747,
+      "step": 248
+    },
+    {
+      "epoch": 0.1874294316898758,
+      "grad_norm": 0.2160220891237259,
+      "learning_rate": 2.4900000000000002e-05,
+      "loss": 0.7271,
+      "step": 249
+    },
+    {
+      "epoch": 0.1881821603312006,
+      "grad_norm": 0.21387308835983276,
+      "learning_rate": 2.5e-05,
+      "loss": 0.7224,
+      "step": 250
+    },
+    {
+      "epoch": 0.18893488897252542,
+      "grad_norm": 0.275083988904953,
+      "learning_rate": 2.51e-05,
+      "loss": 0.7733,
+      "step": 251
+    },
+    {
+      "epoch": 0.1896876176138502,
+      "grad_norm": 0.24984300136566162,
+      "learning_rate": 2.5200000000000003e-05,
+      "loss": 0.7649,
+      "step": 252
+    },
+    {
+      "epoch": 0.190440346255175,
+      "grad_norm": 0.20318986475467682,
+      "learning_rate": 2.5300000000000002e-05,
+      "loss": 0.6969,
+      "step": 253
+    },
+    {
+      "epoch": 0.1911930748964998,
+      "grad_norm": 0.23086310923099518,
+      "learning_rate": 2.54e-05,
+      "loss": 0.6951,
+      "step": 254
+    },
+    {
+      "epoch": 0.19194580353782462,
+      "grad_norm": 0.2896444499492645,
+      "learning_rate": 2.5500000000000003e-05,
+      "loss": 0.6976,
+      "step": 255
+    },
+    {
+      "epoch": 0.19269853217914942,
+      "grad_norm": 0.18938231468200684,
+      "learning_rate": 2.5600000000000002e-05,
+      "loss": 0.7253,
+      "step": 256
+    },
+    {
+      "epoch": 0.19345126082047423,
+      "grad_norm": 0.3274862468242645,
+      "learning_rate": 2.57e-05,
+      "loss": 0.7294,
+      "step": 257
+    },
+    {
+      "epoch": 0.194203989461799,
+      "grad_norm": 0.36922314763069153,
+      "learning_rate": 2.58e-05,
+      "loss": 0.7067,
+      "step": 258
+    },
+    {
+      "epoch": 0.19495671810312382,
+      "grad_norm": 0.23391789197921753,
+      "learning_rate": 2.5900000000000003e-05,
+      "loss": 0.7125,
+      "step": 259
+    },
+    {
+      "epoch": 0.19570944674444862,
+      "grad_norm": 0.5645421743392944,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 0.7298,
+      "step": 260
+    },
+    {
+      "epoch": 0.19646217538577343,
+      "grad_norm": 0.30430150032043457,
+      "learning_rate": 2.61e-05,
+      "loss": 0.7142,
+      "step": 261
+    },
+    {
+      "epoch": 0.19721490402709824,
+      "grad_norm": 0.2254580557346344,
+      "learning_rate": 2.6200000000000003e-05,
+      "loss": 0.7338,
+      "step": 262
+    },
+    {
+      "epoch": 0.19796763266842304,
+      "grad_norm": 0.22907628118991852,
+      "learning_rate": 2.6300000000000002e-05,
+      "loss": 0.7288,
+      "step": 263
+    },
+    {
+      "epoch": 0.19872036130974782,
+      "grad_norm": 0.3144034445285797,
+      "learning_rate": 2.64e-05,
+      "loss": 0.7184,
+      "step": 264
+    },
+    {
+      "epoch": 0.19947308995107263,
+      "grad_norm": 0.3524106740951538,
+      "learning_rate": 2.6500000000000004e-05,
+      "loss": 0.7162,
+      "step": 265
+    },
+    {
+      "epoch": 0.20022581859239744,
+      "grad_norm": 0.24677534401416779,
+      "learning_rate": 2.6600000000000003e-05,
+      "loss": 0.7309,
+      "step": 266
+    },
+    {
+      "epoch": 0.20097854723372224,
+      "grad_norm": 0.3870452642440796,
+      "learning_rate": 2.6700000000000002e-05,
+      "loss": 0.8167,
+      "step": 267
+    },
+    {
+      "epoch": 0.20173127587504705,
+      "grad_norm": 0.25113770365715027,
+      "learning_rate": 2.6800000000000004e-05,
+      "loss": 0.7375,
+      "step": 268
+    },
+    {
+      "epoch": 0.20248400451637186,
+      "grad_norm": 0.24697715044021606,
+      "learning_rate": 2.6900000000000003e-05,
+      "loss": 0.7312,
+      "step": 269
+    },
+    {
+      "epoch": 0.20323673315769666,
+      "grad_norm": 0.3623698651790619,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 0.7563,
+      "step": 270
+    },
+    {
+      "epoch": 0.20398946179902144,
+      "grad_norm": 0.2567082643508911,
+      "learning_rate": 2.7100000000000005e-05,
+      "loss": 0.8219,
+      "step": 271
+    },
+    {
+      "epoch": 0.20474219044034625,
+      "grad_norm": 0.32500094175338745,
+      "learning_rate": 2.7200000000000004e-05,
+      "loss": 0.6903,
+      "step": 272
+    },
+    {
+      "epoch": 0.20549491908167106,
+      "grad_norm": 0.259027361869812,
+      "learning_rate": 2.7300000000000003e-05,
+      "loss": 0.7072,
+      "step": 273
+    },
+    {
+      "epoch": 0.20624764772299586,
+      "grad_norm": 0.3389589190483093,
+      "learning_rate": 2.7400000000000002e-05,
+      "loss": 0.736,
+      "step": 274
+    },
+    {
+      "epoch": 0.20700037636432067,
+      "grad_norm": 0.24815364181995392,
+      "learning_rate": 2.7500000000000004e-05,
+      "loss": 0.7518,
+      "step": 275
+    },
+    {
+      "epoch": 0.20775310500564548,
+      "grad_norm": 0.24750302731990814,
+      "learning_rate": 2.7600000000000003e-05,
+      "loss": 0.7083,
+      "step": 276
+    },
+    {
+      "epoch": 0.20850583364697026,
+      "grad_norm": 0.3570554554462433,
+      "learning_rate": 2.7700000000000002e-05,
+      "loss": 0.7808,
+      "step": 277
+    },
+    {
+      "epoch": 0.20925856228829506,
+      "grad_norm": 0.3397029936313629,
+      "learning_rate": 2.7800000000000005e-05,
+      "loss": 0.8382,
+      "step": 278
+    },
+    {
+      "epoch": 0.21001129092961987,
+      "grad_norm": 0.27232542634010315,
+      "learning_rate": 2.7900000000000004e-05,
+      "loss": 0.7072,
+      "step": 279
+    },
+    {
+      "epoch": 0.21076401957094468,
+      "grad_norm": 0.3069293797016144,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 0.7354,
+      "step": 280
+    },
+    {
+      "epoch": 0.21151674821226948,
+      "grad_norm": 0.29552361369132996,
+      "learning_rate": 2.8100000000000005e-05,
+      "loss": 0.7444,
+      "step": 281
+    },
+    {
+      "epoch": 0.2122694768535943,
+      "grad_norm": 0.25239408016204834,
+      "learning_rate": 2.8199999999999998e-05,
+      "loss": 0.7651,
+      "step": 282
+    },
+    {
+      "epoch": 0.21302220549491907,
+      "grad_norm": 0.19612056016921997,
+      "learning_rate": 2.83e-05,
+      "loss": 0.7402,
+      "step": 283
+    },
+    {
+      "epoch": 0.21377493413624388,
+      "grad_norm": 0.2703988254070282,
+      "learning_rate": 2.84e-05,
+      "loss": 0.6318,
+      "step": 284
+    },
+    {
+      "epoch": 0.21452766277756868,
+      "grad_norm": 0.26631176471710205,
+      "learning_rate": 2.8499999999999998e-05,
+      "loss": 0.7349,
+      "step": 285
+    },
+    {
+      "epoch": 0.2152803914188935,
+      "grad_norm": 0.2037457674741745,
+      "learning_rate": 2.86e-05,
+      "loss": 0.742,
+      "step": 286
+    },
+    {
+      "epoch": 0.2160331200602183,
+      "grad_norm": 0.20361584424972534,
+      "learning_rate": 2.87e-05,
+      "loss": 0.6837,
+      "step": 287
+    },
+    {
+      "epoch": 0.2167858487015431,
+      "grad_norm": 0.3454204499721527,
+      "learning_rate": 2.88e-05,
+      "loss": 0.794,
+      "step": 288
+    },
+    {
+      "epoch": 0.21753857734286788,
+      "grad_norm": 0.20436468720436096,
+      "learning_rate": 2.8899999999999998e-05,
+      "loss": 0.7023,
+      "step": 289
+    },
+    {
+      "epoch": 0.2182913059841927,
+      "grad_norm": 0.3148562014102936,
+      "learning_rate": 2.9e-05,
+      "loss": 0.6997,
+      "step": 290
+    },
+    {
+      "epoch": 0.2190440346255175,
+      "grad_norm": 0.7893130779266357,
+      "learning_rate": 2.91e-05,
+      "loss": 0.7352,
+      "step": 291
+    },
+    {
+      "epoch": 0.2197967632668423,
+      "grad_norm": 0.6444323658943176,
+      "learning_rate": 2.9199999999999998e-05,
+      "loss": 0.7196,
+      "step": 292
+    },
+    {
+      "epoch": 0.2205494919081671,
+      "grad_norm": 0.301099568605423,
+      "learning_rate": 2.93e-05,
+      "loss": 0.7169,
+      "step": 293
+    },
+    {
+      "epoch": 0.22130222054949192,
+      "grad_norm": 0.3642562925815582,
+      "learning_rate": 2.94e-05,
+      "loss": 0.7344,
+      "step": 294
+    },
+    {
+      "epoch": 0.2220549491908167,
+      "grad_norm": 0.453366219997406,
+      "learning_rate": 2.95e-05,
+      "loss": 0.735,
+      "step": 295
+    },
+    {
+      "epoch": 0.2228076778321415,
+      "grad_norm": 0.25017601251602173,
+      "learning_rate": 2.96e-05,
+      "loss": 0.7458,
+      "step": 296
+    },
+    {
+      "epoch": 0.2235604064734663,
+      "grad_norm": 0.41115790605545044,
+      "learning_rate": 2.97e-05,
+      "loss": 0.7588,
+      "step": 297
+    },
+    {
+      "epoch": 0.22431313511479112,
+      "grad_norm": 0.25116226077079773,
+      "learning_rate": 2.98e-05,
+      "loss": 0.7201,
+      "step": 298
+    },
+    {
+      "epoch": 0.22506586375611592,
+      "grad_norm": 0.2897501587867737,
+      "learning_rate": 2.9900000000000002e-05,
+      "loss": 0.6736,
+      "step": 299
+    },
+    {
+      "epoch": 0.22581859239744073,
+      "grad_norm": 0.2898609936237335,
+      "learning_rate": 3e-05,
+      "loss": 0.7126,
+      "step": 300
+    },
+    {
+      "epoch": 0.22657132103876554,
+      "grad_norm": 0.24281033873558044,
+      "learning_rate": 3.01e-05,
+      "loss": 0.7235,
+      "step": 301
+    },
+    {
+      "epoch": 0.22732404968009032,
+      "grad_norm": 0.24791251122951508,
+      "learning_rate": 3.02e-05,
+      "loss": 0.829,
+      "step": 302
+    },
+    {
+      "epoch": 0.22807677832141512,
+      "grad_norm": 0.23729638755321503,
+      "learning_rate": 3.03e-05,
+      "loss": 0.7159,
+      "step": 303
+    },
+    {
+      "epoch": 0.22882950696273993,
+      "grad_norm": 0.1961555927991867,
+      "learning_rate": 3.04e-05,
+      "loss": 0.7089,
+      "step": 304
+    },
+    {
+      "epoch": 0.22958223560406474,
+      "grad_norm": 0.2686220705509186,
+      "learning_rate": 3.05e-05,
+      "loss": 0.7039,
+      "step": 305
+    },
+    {
+      "epoch": 0.23033496424538955,
+      "grad_norm": 0.23274317383766174,
+      "learning_rate": 3.06e-05,
+      "loss": 0.7519,
+      "step": 306
+    },
+    {
+      "epoch": 0.23108769288671435,
+      "grad_norm": 0.18943428993225098,
+      "learning_rate": 3.07e-05,
+      "loss": 0.6814,
+      "step": 307
+    },
+    {
+      "epoch": 0.23184042152803913,
+      "grad_norm": 0.22373147308826447,
+      "learning_rate": 3.08e-05,
+      "loss": 0.6882,
+      "step": 308
+    },
+    {
+      "epoch": 0.23259315016936394,
+      "grad_norm": 0.27525392174720764,
+      "learning_rate": 3.09e-05,
+      "loss": 0.6901,
+      "step": 309
+    },
+    {
+      "epoch": 0.23334587881068874,
+      "grad_norm": 0.2217240333557129,
+      "learning_rate": 3.1e-05,
+      "loss": 0.7744,
+      "step": 310
+    },
+    {
+      "epoch": 0.23409860745201355,
+      "grad_norm": 0.19920943677425385,
+      "learning_rate": 3.1100000000000004e-05,
+      "loss": 0.713,
+      "step": 311
+    },
+    {
+      "epoch": 0.23485133609333836,
+      "grad_norm": 0.22240842878818512,
+      "learning_rate": 3.12e-05,
+      "loss": 0.6832,
+      "step": 312
+    },
+    {
+      "epoch": 0.23560406473466317,
+      "grad_norm": 0.16849708557128906,
+      "learning_rate": 3.13e-05,
+      "loss": 0.7034,
+      "step": 313
+    },
+    {
+      "epoch": 0.23635679337598794,
+      "grad_norm": 0.25105270743370056,
+      "learning_rate": 3.1400000000000004e-05,
+      "loss": 0.6979,
+      "step": 314
+    },
+    {
+      "epoch": 0.23710952201731275,
+      "grad_norm": 0.18835586309432983,
+      "learning_rate": 3.15e-05,
+      "loss": 0.7178,
+      "step": 315
+    },
+    {
+      "epoch": 0.23786225065863756,
+      "grad_norm": 0.2040720134973526,
+      "learning_rate": 3.16e-05,
+      "loss": 0.669,
+      "step": 316
+    },
+    {
+      "epoch": 0.23861497929996237,
+      "grad_norm": 0.2377663552761078,
+      "learning_rate": 3.1700000000000005e-05,
+      "loss": 0.7516,
+      "step": 317
+    },
+    {
+      "epoch": 0.23936770794128717,
+      "grad_norm": 0.23407985270023346,
+      "learning_rate": 3.18e-05,
+      "loss": 0.7292,
+      "step": 318
+    },
+    {
+      "epoch": 0.24012043658261198,
+      "grad_norm": 0.22421889007091522,
+      "learning_rate": 3.19e-05,
+      "loss": 0.7141,
+      "step": 319
+    },
+    {
+      "epoch": 0.24087316522393676,
+      "grad_norm": 0.3384268879890442,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 0.741,
+      "step": 320
+    },
+    {
+      "epoch": 0.24162589386526157,
+      "grad_norm": 0.3508033752441406,
+      "learning_rate": 3.21e-05,
+      "loss": 0.7134,
+      "step": 321
+    },
+    {
+      "epoch": 0.24237862250658637,
+      "grad_norm": 0.40767964720726013,
+      "learning_rate": 3.2200000000000003e-05,
+      "loss": 0.7707,
+      "step": 322
+    },
+    {
+      "epoch": 0.24313135114791118,
+      "grad_norm": 0.21702691912651062,
+      "learning_rate": 3.2300000000000006e-05,
+      "loss": 0.7123,
+      "step": 323
+    },
+    {
+      "epoch": 0.24388407978923599,
+      "grad_norm": 0.3439074456691742,
+      "learning_rate": 3.24e-05,
+      "loss": 0.7002,
+      "step": 324
+    },
+    {
+      "epoch": 0.2446368084305608,
+      "grad_norm": 0.7880777716636658,
+      "learning_rate": 3.2500000000000004e-05,
+      "loss": 0.7744,
+      "step": 325
+    },
+    {
+      "epoch": 0.24538953707188557,
+      "grad_norm": 0.2364712804555893,
+      "learning_rate": 3.26e-05,
+      "loss": 0.6912,
+      "step": 326
+    },
+    {
+      "epoch": 0.24614226571321038,
+      "grad_norm": 0.24973264336585999,
+      "learning_rate": 3.27e-05,
+      "loss": 0.6676,
+      "step": 327
+    },
+    {
+      "epoch": 0.24689499435453519,
+      "grad_norm": 0.2864416837692261,
+      "learning_rate": 3.2800000000000004e-05,
+      "loss": 0.8033,
+      "step": 328
+    },
+    {
+      "epoch": 0.24764772299586,
+      "grad_norm": 0.30503344535827637,
+      "learning_rate": 3.29e-05,
+      "loss": 0.7436,
+      "step": 329
+    },
+    {
+      "epoch": 0.2484004516371848,
+      "grad_norm": 0.40133169293403625,
+      "learning_rate": 3.3e-05,
+      "loss": 0.7449,
+      "step": 330
+    },
+    {
+      "epoch": 0.2491531802785096,
+      "grad_norm": 0.5105140805244446,
+      "learning_rate": 3.3100000000000005e-05,
+      "loss": 0.755,
+      "step": 331
+    },
+    {
+      "epoch": 0.2499059089198344,
+      "grad_norm": 0.3283163607120514,
+      "learning_rate": 3.32e-05,
+      "loss": 0.7566,
+      "step": 332
+    },
+    {
+      "epoch": 0.2506586375611592,
+      "grad_norm": 0.581061065196991,
+      "learning_rate": 3.33e-05,
+      "loss": 0.6701,
+      "step": 333
+    },
+    {
+      "epoch": 0.251411366202484,
+      "grad_norm": 0.4370225667953491,
+      "learning_rate": 3.3400000000000005e-05,
+      "loss": 0.709,
+      "step": 334
+    },
+    {
+      "epoch": 0.2521640948438088,
+      "grad_norm": 0.35738813877105713,
+      "learning_rate": 3.35e-05,
+      "loss": 0.6914,
+      "step": 335
+    },
+    {
+      "epoch": 0.2529168234851336,
+      "grad_norm": 0.4989299774169922,
+      "learning_rate": 3.3600000000000004e-05,
+      "loss": 0.8071,
+      "step": 336
+    },
+    {
+      "epoch": 0.2536695521264584,
+      "grad_norm": 0.6059109568595886,
+      "learning_rate": 3.3700000000000006e-05,
+      "loss": 0.6323,
+      "step": 337
+    },
+    {
+      "epoch": 0.2544222807677832,
+      "grad_norm": 0.3837607502937317,
+      "learning_rate": 3.38e-05,
+      "loss": 0.7006,
+      "step": 338
+    },
+    {
+      "epoch": 0.25517500940910803,
+      "grad_norm": 0.5447167158126831,
+      "learning_rate": 3.3900000000000004e-05,
+      "loss": 0.646,
+      "step": 339
+    },
+    {
+      "epoch": 0.25592773805043284,
+      "grad_norm": 0.5344168543815613,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 0.677,
+      "step": 340
+    },
+    {
+      "epoch": 0.25668046669175765,
+      "grad_norm": 0.41303715109825134,
+      "learning_rate": 3.41e-05,
+      "loss": 0.6527,
+      "step": 341
+    },
+    {
+      "epoch": 0.2574331953330824,
+      "grad_norm": 0.42507702112197876,
+      "learning_rate": 3.4200000000000005e-05,
+      "loss": 0.6974,
+      "step": 342
+    },
+    {
+      "epoch": 0.2581859239744072,
+      "grad_norm": 0.33432236313819885,
+      "learning_rate": 3.430000000000001e-05,
+      "loss": 0.6985,
+      "step": 343
+    },
+    {
+      "epoch": 0.258938652615732,
+      "grad_norm": 0.3904542624950409,
+      "learning_rate": 3.4399999999999996e-05,
+      "loss": 0.6119,
+      "step": 344
+    },
+    {
+      "epoch": 0.2596913812570568,
+      "grad_norm": 0.28473809361457825,
+      "learning_rate": 3.45e-05,
+      "loss": 0.688,
+      "step": 345
+    },
+    {
+      "epoch": 0.2604441098983816,
+      "grad_norm": 0.3012646734714508,
+      "learning_rate": 3.46e-05,
+      "loss": 0.8099,
+      "step": 346
+    },
+    {
+      "epoch": 0.26119683853970643,
+      "grad_norm": 0.7255945801734924,
+      "learning_rate": 3.4699999999999996e-05,
+      "loss": 0.7401,
+      "step": 347
+    },
+    {
+      "epoch": 0.26194956718103124,
+      "grad_norm": 0.2889805734157562,
+      "learning_rate": 3.48e-05,
+      "loss": 0.6566,
+      "step": 348
+    },
+    {
+      "epoch": 0.26270229582235605,
+      "grad_norm": 0.3256964385509491,
+      "learning_rate": 3.49e-05,
+      "loss": 0.7319,
+      "step": 349
+    },
+    {
+      "epoch": 0.26345502446368085,
+      "grad_norm": 0.2728472352027893,
+      "learning_rate": 3.5e-05,
+      "loss": 0.5832,
+      "step": 350
+    },
+    {
+      "epoch": 0.26420775310500566,
+      "grad_norm": 0.2984137535095215,
+      "learning_rate": 3.51e-05,
+      "loss": 0.7307,
+      "step": 351
+    },
+    {
+      "epoch": 0.26496048174633047,
+      "grad_norm": 0.3634515106678009,
+      "learning_rate": 3.52e-05,
+      "loss": 0.7945,
+      "step": 352
+    },
+    {
+      "epoch": 0.2657132103876553,
+      "grad_norm": 0.25072675943374634,
+      "learning_rate": 3.53e-05,
+      "loss": 0.7998,
+      "step": 353
+    },
+    {
+      "epoch": 0.26646593902898,
+      "grad_norm": 0.24649253487586975,
+      "learning_rate": 3.54e-05,
+      "loss": 0.6374,
+      "step": 354
+    },
+    {
+      "epoch": 0.26721866767030483,
+      "grad_norm": 0.3220970332622528,
+      "learning_rate": 3.55e-05,
+      "loss": 0.6967,
+      "step": 355
+    },
+    {
+      "epoch": 0.26797139631162964,
+      "grad_norm": 0.3779277801513672,
+      "learning_rate": 3.56e-05,
+      "loss": 0.6951,
+      "step": 356
+    },
+    {
+      "epoch": 0.26872412495295445,
+      "grad_norm": 0.4565024673938751,
+      "learning_rate": 3.57e-05,
+      "loss": 0.6542,
+      "step": 357
+    },
+    {
+      "epoch": 0.26947685359427925,
+      "grad_norm": 0.33353057503700256,
+      "learning_rate": 3.58e-05,
+      "loss": 0.6972,
+      "step": 358
+    },
+    {
+      "epoch": 0.27022958223560406,
+      "grad_norm": 0.4957676827907562,
+      "learning_rate": 3.59e-05,
+      "loss": 0.6382,
+      "step": 359
+    },
+    {
+      "epoch": 0.27098231087692887,
+      "grad_norm": 0.32771801948547363,
+      "learning_rate": 3.6e-05,
+      "loss": 0.6512,
+      "step": 360
+    },
+    {
+      "epoch": 0.2717350395182537,
+      "grad_norm": 0.23294134438037872,
+      "learning_rate": 3.61e-05,
+      "loss": 0.7273,
+      "step": 361
+    },
+    {
+      "epoch": 0.2724877681595785,
+      "grad_norm": 0.3823862075805664,
+      "learning_rate": 3.62e-05,
+      "loss": 0.7382,
+      "step": 362
+    },
+    {
+      "epoch": 0.2732404968009033,
+      "grad_norm": 0.37054651975631714,
+      "learning_rate": 3.63e-05,
+      "loss": 0.6779,
+      "step": 363
+    },
+    {
+      "epoch": 0.2739932254422281,
+      "grad_norm": 0.2064339518547058,
+      "learning_rate": 3.6400000000000004e-05,
+      "loss": 0.6934,
+      "step": 364
+    },
+    {
+      "epoch": 0.2747459540835529,
+      "grad_norm": 0.3629573583602905,
+      "learning_rate": 3.65e-05,
+      "loss": 0.6267,
+      "step": 365
+    },
+    {
+      "epoch": 0.2754986827248777,
+      "grad_norm": 0.29032593965530396,
+      "learning_rate": 3.66e-05,
+      "loss": 0.6429,
+      "step": 366
+    },
+    {
+      "epoch": 0.27625141136620246,
+      "grad_norm": 0.22050155699253082,
+      "learning_rate": 3.6700000000000004e-05,
+      "loss": 0.6658,
+      "step": 367
+    },
+    {
+      "epoch": 0.27700414000752727,
+      "grad_norm": 0.2628399431705475,
+      "learning_rate": 3.68e-05,
+      "loss": 0.7378,
+      "step": 368
+    },
+    {
+      "epoch": 0.2777568686488521,
+      "grad_norm": 0.3074728548526764,
+      "learning_rate": 3.69e-05,
+      "loss": 0.7041,
+      "step": 369
+    },
+    {
+      "epoch": 0.2785095972901769,
+      "grad_norm": 0.23266035318374634,
+      "learning_rate": 3.7e-05,
+      "loss": 0.7271,
+      "step": 370
+    },
+    {
+      "epoch": 0.2792623259315017,
+      "grad_norm": 0.2957321107387543,
+      "learning_rate": 3.71e-05,
+      "loss": 0.6223,
+      "step": 371
+    },
+    {
+      "epoch": 0.2800150545728265,
+      "grad_norm": 0.215256467461586,
+      "learning_rate": 3.72e-05,
+      "loss": 0.6677,
+      "step": 372
+    },
+    {
+      "epoch": 0.2807677832141513,
+      "grad_norm": 0.3338688015937805,
+      "learning_rate": 3.73e-05,
+      "loss": 0.6694,
+      "step": 373
+    },
+    {
+      "epoch": 0.2815205118554761,
+      "grad_norm": 0.2900616228580475,
+      "learning_rate": 3.74e-05,
+      "loss": 0.7235,
+      "step": 374
+    },
+    {
+      "epoch": 0.2822732404968009,
+      "grad_norm": 0.27830448746681213,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.686,
+      "step": 375
+    },
+    {
+      "epoch": 0.2830259691381257,
+      "grad_norm": 0.31134238839149475,
+      "learning_rate": 3.76e-05,
+      "loss": 0.7177,
+      "step": 376
+    },
+    {
+      "epoch": 0.28377869777945053,
+      "grad_norm": 0.20832011103630066,
+      "learning_rate": 3.77e-05,
+      "loss": 0.6302,
+      "step": 377
+    },
+    {
+      "epoch": 0.28453142642077534,
+      "grad_norm": 0.24844639003276825,
+      "learning_rate": 3.7800000000000004e-05,
+      "loss": 0.7132,
+      "step": 378
+    },
+    {
+      "epoch": 0.2852841550621001,
+      "grad_norm": 0.4205746054649353,
+      "learning_rate": 3.79e-05,
+      "loss": 0.699,
+      "step": 379
+    },
+    {
+      "epoch": 0.2860368837034249,
+      "grad_norm": 0.2722179591655731,
+      "learning_rate": 3.8e-05,
+      "loss": 0.6298,
+      "step": 380
+    },
+    {
+      "epoch": 0.2867896123447497,
+      "grad_norm": 0.5374658703804016,
+      "learning_rate": 3.8100000000000005e-05,
+      "loss": 0.7162,
+      "step": 381
+    },
+    {
+      "epoch": 0.2875423409860745,
+      "grad_norm": 0.7044588923454285,
+      "learning_rate": 3.82e-05,
+      "loss": 0.6253,
+      "step": 382
+    },
+    {
+      "epoch": 0.2882950696273993,
+      "grad_norm": 0.39629724621772766,
+      "learning_rate": 3.83e-05,
+      "loss": 0.6739,
+      "step": 383
+    },
+    {
+      "epoch": 0.2890477982687241,
+      "grad_norm": 0.9906265735626221,
+      "learning_rate": 3.8400000000000005e-05,
+      "loss": 0.7478,
+      "step": 384
+    },
+    {
+      "epoch": 0.28980052691004893,
+      "grad_norm": 0.6014965772628784,
+      "learning_rate": 3.85e-05,
+      "loss": 0.6215,
+      "step": 385
+    },
+    {
+      "epoch": 0.29055325555137373,
+      "grad_norm": 0.5241867303848267,
+      "learning_rate": 3.86e-05,
+      "loss": 0.6619,
+      "step": 386
+    },
+    {
+      "epoch": 0.29130598419269854,
+      "grad_norm": 0.43592947721481323,
+      "learning_rate": 3.8700000000000006e-05,
+      "loss": 0.6769,
+      "step": 387
+    },
+    {
+      "epoch": 0.29205871283402335,
+      "grad_norm": 0.7054777145385742,
+      "learning_rate": 3.88e-05,
+      "loss": 0.6867,
+      "step": 388
+    },
+    {
+      "epoch": 0.29281144147534816,
+      "grad_norm": 0.3662691116333008,
+      "learning_rate": 3.8900000000000004e-05,
+      "loss": 0.7396,
+      "step": 389
+    },
+    {
+      "epoch": 0.29356417011667296,
+      "grad_norm": 0.32213476300239563,
+      "learning_rate": 3.9000000000000006e-05,
+      "loss": 0.7144,
+      "step": 390
+    },
+    {
+      "epoch": 0.29431689875799777,
+      "grad_norm": 0.3397921025753021,
+      "learning_rate": 3.91e-05,
+      "loss": 0.6997,
+      "step": 391
+    },
+    {
+      "epoch": 0.2950696273993225,
+      "grad_norm": 0.3259410560131073,
+      "learning_rate": 3.9200000000000004e-05,
+      "loss": 0.7483,
+      "step": 392
+    },
+    {
+      "epoch": 0.2958223560406473,
+      "grad_norm": 0.32705387473106384,
+      "learning_rate": 3.9300000000000007e-05,
+      "loss": 0.7266,
+      "step": 393
+    },
+    {
+      "epoch": 0.29657508468197213,
+      "grad_norm": 0.2648999094963074,
+      "learning_rate": 3.94e-05,
+      "loss": 0.7223,
+      "step": 394
+    },
+    {
+      "epoch": 0.29732781332329694,
+      "grad_norm": 0.27103105187416077,
+      "learning_rate": 3.9500000000000005e-05,
+      "loss": 0.7121,
+      "step": 395
+    },
+    {
+      "epoch": 0.29808054196462175,
+      "grad_norm": 0.29011270403862,
+      "learning_rate": 3.960000000000001e-05,
+      "loss": 0.6592,
+      "step": 396
+    },
+    {
+      "epoch": 0.29883327060594655,
+      "grad_norm": 0.2648636996746063,
+      "learning_rate": 3.97e-05,
+      "loss": 0.6726,
+      "step": 397
+    },
+    {
+      "epoch": 0.29958599924727136,
+      "grad_norm": 0.30544421076774597,
+      "learning_rate": 3.9800000000000005e-05,
+      "loss": 0.7361,
+      "step": 398
+    },
+    {
+      "epoch": 0.30033872788859617,
+      "grad_norm": 0.27379918098449707,
+      "learning_rate": 3.99e-05,
+      "loss": 0.7188,
+      "step": 399
+    },
+    {
+      "epoch": 0.301091456529921,
+      "grad_norm": 0.2808031141757965,
+      "learning_rate": 4e-05,
+      "loss": 0.6579,
+      "step": 400
+    },
+    {
+      "epoch": 0.301091456529921,
+      "eval_loss": 0.6637915372848511,
+      "eval_runtime": 452.5217,
+      "eval_samples_per_second": 21.274,
+      "eval_steps_per_second": 0.665,
+      "step": 400
+    },
+    {
+      "epoch": 0.3018441851712458,
+      "grad_norm": 0.20910316705703735,
+      "learning_rate": 4.0100000000000006e-05,
+      "loss": 0.673,
+      "step": 401
+    },
+    {
+      "epoch": 0.3025969138125706,
+      "grad_norm": 0.26251021027565,
+      "learning_rate": 4.02e-05,
+      "loss": 0.5719,
+      "step": 402
+    },
+    {
+      "epoch": 0.3033496424538954,
+      "grad_norm": 0.2821217477321625,
+      "learning_rate": 4.0300000000000004e-05,
+      "loss": 0.6904,
+      "step": 403
+    },
+    {
+      "epoch": 0.30410237109522015,
+      "grad_norm": 0.3193609118461609,
+      "learning_rate": 4.0400000000000006e-05,
+      "loss": 0.6832,
+      "step": 404
+    },
+    {
+      "epoch": 0.30485509973654495,
+      "grad_norm": 0.2418869286775589,
+      "learning_rate": 4.05e-05,
+      "loss": 0.7142,
+      "step": 405
+    },
+    {
+      "epoch": 0.30560782837786976,
+      "grad_norm": 0.21576568484306335,
+      "learning_rate": 4.0600000000000004e-05,
+      "loss": 0.7342,
+      "step": 406
+    },
+    {
+      "epoch": 0.30636055701919457,
+      "grad_norm": 0.24549145996570587,
+      "learning_rate": 4.07e-05,
+      "loss": 0.619,
+      "step": 407
+    },
+    {
+      "epoch": 0.3071132856605194,
+      "grad_norm": 0.2634813189506531,
+      "learning_rate": 4.08e-05,
+      "loss": 0.7263,
+      "step": 408
+    },
+    {
+      "epoch": 0.3078660143018442,
+      "grad_norm": 0.21867918968200684,
+      "learning_rate": 4.09e-05,
+      "loss": 0.6296,
+      "step": 409
+    },
+    {
+      "epoch": 0.308618742943169,
+      "grad_norm": 0.22789400815963745,
+      "learning_rate": 4.1e-05,
+      "loss": 0.6141,
+      "step": 410
+    },
+    {
+      "epoch": 0.3093714715844938,
+      "grad_norm": 0.3099987208843231,
+      "learning_rate": 4.11e-05,
+      "loss": 0.6141,
+      "step": 411
+    },
+    {
+      "epoch": 0.3101242002258186,
+      "grad_norm": 0.23625628650188446,
+      "learning_rate": 4.12e-05,
+      "loss": 0.755,
+      "step": 412
+    },
+    {
+      "epoch": 0.3108769288671434,
+      "grad_norm": 0.4678927958011627,
+      "learning_rate": 4.13e-05,
+      "loss": 0.6951,
+      "step": 413
+    },
+    {
+      "epoch": 0.3116296575084682,
+      "grad_norm": 0.33995985984802246,
+      "learning_rate": 4.14e-05,
+      "loss": 0.6446,
+      "step": 414
+    },
+    {
+      "epoch": 0.312382386149793,
+      "grad_norm": 0.302704781293869,
+      "learning_rate": 4.15e-05,
+      "loss": 0.7921,
+      "step": 415
+    },
+    {
+      "epoch": 0.3131351147911178,
+      "grad_norm": 0.3012588620185852,
+      "learning_rate": 4.16e-05,
+      "loss": 0.6384,
+      "step": 416
+    },
+    {
+      "epoch": 0.3138878434324426,
+      "grad_norm": 0.2886503338813782,
+      "learning_rate": 4.17e-05,
+      "loss": 0.7456,
+      "step": 417
+    },
+    {
+      "epoch": 0.3146405720737674,
+      "grad_norm": 0.24612435698509216,
+      "learning_rate": 4.18e-05,
+      "loss": 0.6552,
+      "step": 418
+    },
+    {
+      "epoch": 0.3153933007150922,
+      "grad_norm": 0.3362456262111664,
+      "learning_rate": 4.19e-05,
+      "loss": 0.606,
+      "step": 419
+    },
+    {
+      "epoch": 0.316146029356417,
+      "grad_norm": 0.24216234683990479,
+      "learning_rate": 4.2e-05,
+      "loss": 0.6677,
+      "step": 420
+    },
+    {
+      "epoch": 0.3168987579977418,
+      "grad_norm": 0.2626403272151947,
+      "learning_rate": 4.21e-05,
+      "loss": 0.7394,
+      "step": 421
+    },
+    {
+      "epoch": 0.3176514866390666,
+      "grad_norm": 0.300325483083725,
+      "learning_rate": 4.22e-05,
+      "loss": 0.7499,
+      "step": 422
+    },
+    {
+      "epoch": 0.3184042152803914,
+      "grad_norm": 0.3593023121356964,
+      "learning_rate": 4.23e-05,
+      "loss": 0.7903,
+      "step": 423
+    },
+    {
+      "epoch": 0.31915694392171623,
+      "grad_norm": 0.267685204744339,
+      "learning_rate": 4.24e-05,
+      "loss": 0.6782,
+      "step": 424
+    },
+    {
+      "epoch": 0.31990967256304104,
+      "grad_norm": 0.26711738109588623,
+      "learning_rate": 4.25e-05,
+      "loss": 0.6666,
+      "step": 425
+    },
+    {
+      "epoch": 0.32066240120436584,
+      "grad_norm": 0.20883305370807648,
+      "learning_rate": 4.26e-05,
+      "loss": 0.7091,
+      "step": 426
+    },
+    {
+      "epoch": 0.32141512984569065,
+      "grad_norm": 0.2823686897754669,
+      "learning_rate": 4.27e-05,
+      "loss": 0.7508,
+      "step": 427
+    },
+    {
+      "epoch": 0.32216785848701546,
+      "grad_norm": 0.34003427624702454,
+      "learning_rate": 4.2800000000000004e-05,
+      "loss": 0.6402,
+      "step": 428
+    },
+    {
+      "epoch": 0.3229205871283402,
+      "grad_norm": 0.2812309265136719,
+      "learning_rate": 4.29e-05,
+      "loss": 0.7103,
+      "step": 429
+    },
+    {
+      "epoch": 0.323673315769665,
+      "grad_norm": 0.437812864780426,
+      "learning_rate": 4.3e-05,
+      "loss": 0.7043,
+      "step": 430
+    },
+    {
+      "epoch": 0.3244260444109898,
+      "grad_norm": 0.310994416475296,
+      "learning_rate": 4.3100000000000004e-05,
+      "loss": 0.618,
+      "step": 431
+    },
+    {
+      "epoch": 0.32517877305231463,
+      "grad_norm": 0.35913288593292236,
+      "learning_rate": 4.32e-05,
+      "loss": 0.5754,
+      "step": 432
+    },
+    {
+      "epoch": 0.32593150169363944,
+      "grad_norm": 0.6798489093780518,
+      "learning_rate": 4.33e-05,
+      "loss": 0.6714,
+      "step": 433
+    },
+    {
+      "epoch": 0.32668423033496424,
+      "grad_norm": 0.41850361227989197,
+      "learning_rate": 4.3400000000000005e-05,
+      "loss": 0.6107,
+      "step": 434
+    },
+    {
+      "epoch": 0.32743695897628905,
+      "grad_norm": 0.3068406283855438,
+      "learning_rate": 4.35e-05,
+      "loss": 0.718,
+      "step": 435
+    },
+    {
+      "epoch": 0.32818968761761386,
+      "grad_norm": 0.762353777885437,
+      "learning_rate": 4.36e-05,
+      "loss": 0.7022,
+      "step": 436
+    },
+    {
+      "epoch": 0.32894241625893866,
+      "grad_norm": 0.5267826914787292,
+      "learning_rate": 4.3700000000000005e-05,
+      "loss": 0.6498,
+      "step": 437
+    },
+    {
+      "epoch": 0.32969514490026347,
+      "grad_norm": 0.5997746586799622,
+      "learning_rate": 4.38e-05,
+      "loss": 0.6676,
+      "step": 438
+    },
+    {
+      "epoch": 0.3304478735415883,
+      "grad_norm": 0.5417837500572205,
+      "learning_rate": 4.39e-05,
+      "loss": 0.6559,
+      "step": 439
+    },
+    {
+      "epoch": 0.3312006021829131,
+      "grad_norm": 0.41214191913604736,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.5963,
+      "step": 440
+    },
+    {
+      "epoch": 0.33195333082423784,
+      "grad_norm": 0.4180992543697357,
+      "learning_rate": 4.41e-05,
+      "loss": 0.6178,
+      "step": 441
+    },
+    {
+      "epoch": 0.33270605946556264,
+      "grad_norm": 0.33256128430366516,
+      "learning_rate": 4.4200000000000004e-05,
+      "loss": 0.6811,
+      "step": 442
+    },
+    {
+      "epoch": 0.33345878810688745,
+      "grad_norm": 0.3526419699192047,
+      "learning_rate": 4.43e-05,
+      "loss": 0.6285,
+      "step": 443
+    },
+    {
+      "epoch": 0.33421151674821226,
+      "grad_norm": 0.3703833818435669,
+      "learning_rate": 4.44e-05,
+      "loss": 0.7578,
+      "step": 444
+    },
+    {
+      "epoch": 0.33496424538953706,
+      "grad_norm": 0.2921695411205292,
+      "learning_rate": 4.4500000000000004e-05,
+      "loss": 0.6552,
+      "step": 445
+    },
+    {
+      "epoch": 0.33571697403086187,
+      "grad_norm": 0.3625754117965698,
+      "learning_rate": 4.46e-05,
+      "loss": 0.6283,
+      "step": 446
+    },
+    {
+      "epoch": 0.3364697026721867,
+      "grad_norm": 0.35638731718063354,
+      "learning_rate": 4.47e-05,
+      "loss": 0.6448,
+      "step": 447
+    },
+    {
+      "epoch": 0.3372224313135115,
+      "grad_norm": 0.27190133929252625,
+      "learning_rate": 4.4800000000000005e-05,
+      "loss": 0.6659,
+      "step": 448
+    },
+    {
+      "epoch": 0.3379751599548363,
+      "grad_norm": 0.4058379530906677,
+      "learning_rate": 4.49e-05,
+      "loss": 0.6184,
+      "step": 449
+    },
+    {
+      "epoch": 0.3387278885961611,
+      "grad_norm": 0.29619985818862915,
+      "learning_rate": 4.5e-05,
+      "loss": 0.6974,
+      "step": 450
+    },
+    {
+      "epoch": 0.3394806172374859,
+      "grad_norm": 0.3406849205493927,
+      "learning_rate": 4.5100000000000005e-05,
+      "loss": 0.6472,
+      "step": 451
+    },
+    {
+      "epoch": 0.3402333458788107,
+      "grad_norm": 0.4144502580165863,
+      "learning_rate": 4.52e-05,
+      "loss": 0.6073,
+      "step": 452
+    },
+    {
+      "epoch": 0.3409860745201355,
+      "grad_norm": 0.3010427951812744,
+      "learning_rate": 4.53e-05,
+      "loss": 0.6414,
+      "step": 453
+    },
+    {
+      "epoch": 0.34173880316146027,
+      "grad_norm": 0.26276376843452454,
+      "learning_rate": 4.5400000000000006e-05,
+      "loss": 0.5825,
+      "step": 454
+    },
+    {
+      "epoch": 0.3424915318027851,
+      "grad_norm": 0.5613912343978882,
+      "learning_rate": 4.55e-05,
+      "loss": 0.6899,
+      "step": 455
+    },
+    {
+      "epoch": 0.3432442604441099,
+      "grad_norm": 0.4157487452030182,
+      "learning_rate": 4.5600000000000004e-05,
+      "loss": 0.5864,
+      "step": 456
+    },
+    {
+      "epoch": 0.3439969890854347,
+      "grad_norm": 0.47082406282424927,
+      "learning_rate": 4.5700000000000006e-05,
+      "loss": 0.6701,
+      "step": 457
+    },
+    {
+      "epoch": 0.3447497177267595,
+      "grad_norm": 0.7979834675788879,
+      "learning_rate": 4.58e-05,
+      "loss": 0.698,
+      "step": 458
+    },
+    {
+      "epoch": 0.3455024463680843,
+      "grad_norm": 0.3836079239845276,
+      "learning_rate": 4.5900000000000004e-05,
+      "loss": 0.6274,
+      "step": 459
+    },
+    {
+      "epoch": 0.3462551750094091,
+      "grad_norm": 2.087455987930298,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.6858,
+      "step": 460
+    },
+    {
+      "epoch": 0.3470079036507339,
+      "grad_norm": 1.3945508003234863,
+      "learning_rate": 4.61e-05,
+      "loss": 0.6721,
+      "step": 461
+    },
+    {
+      "epoch": 0.3477606322920587,
+      "grad_norm": 0.46851420402526855,
+      "learning_rate": 4.6200000000000005e-05,
+      "loss": 0.6685,
+      "step": 462
+    },
+    {
+      "epoch": 0.34851336093338353,
+      "grad_norm": 0.39278343319892883,
+      "learning_rate": 4.630000000000001e-05,
+      "loss": 0.6758,
+      "step": 463
+    },
+    {
+      "epoch": 0.34926608957470834,
+      "grad_norm": 0.5558626055717468,
+      "learning_rate": 4.64e-05,
+      "loss": 0.7173,
+      "step": 464
+    },
+    {
+      "epoch": 0.35001881821603315,
+      "grad_norm": 0.35553762316703796,
+      "learning_rate": 4.6500000000000005e-05,
+      "loss": 0.6056,
+      "step": 465
+    },
+    {
+      "epoch": 0.3507715468573579,
+      "grad_norm": 0.9027112722396851,
+      "learning_rate": 4.660000000000001e-05,
+      "loss": 0.716,
+      "step": 466
+    },
+    {
+      "epoch": 0.3515242754986827,
+      "grad_norm": 0.3927246034145355,
+      "learning_rate": 4.6700000000000003e-05,
+      "loss": 0.7781,
+      "step": 467
+    },
+    {
+      "epoch": 0.3522770041400075,
+      "grad_norm": 0.3221803605556488,
+      "learning_rate": 4.6800000000000006e-05,
+      "loss": 0.7038,
+      "step": 468
+    },
+    {
+      "epoch": 0.3530297327813323,
+      "grad_norm": 0.4120510220527649,
+      "learning_rate": 4.69e-05,
+      "loss": 0.5748,
+      "step": 469
+    },
+    {
+      "epoch": 0.3537824614226571,
+      "grad_norm": 0.43108487129211426,
+      "learning_rate": 4.7e-05,
+      "loss": 0.6652,
+      "step": 470
+    },
+    {
+      "epoch": 0.35453519006398193,
+      "grad_norm": 0.32627058029174805,
+      "learning_rate": 4.71e-05,
+      "loss": 0.6593,
+      "step": 471
+    },
+    {
+      "epoch": 0.35528791870530674,
+      "grad_norm": 0.328811913728714,
+      "learning_rate": 4.72e-05,
+      "loss": 0.6016,
+      "step": 472
+    },
+    {
+      "epoch": 0.35604064734663154,
+      "grad_norm": 0.30061453580856323,
+      "learning_rate": 4.73e-05,
+      "loss": 0.6662,
+      "step": 473
+    },
+    {
+      "epoch": 0.35679337598795635,
+      "grad_norm": 0.29043376445770264,
+      "learning_rate": 4.74e-05,
+      "loss": 0.6707,
+      "step": 474
+    },
+    {
+      "epoch": 0.35754610462928116,
+      "grad_norm": 0.23478028178215027,
+      "learning_rate": 4.75e-05,
+      "loss": 0.6986,
+      "step": 475
+    },
+    {
+      "epoch": 0.35829883327060597,
+      "grad_norm": 0.3239274024963379,
+      "learning_rate": 4.76e-05,
+      "loss": 0.681,
+      "step": 476
+    },
+    {
+      "epoch": 0.35905156191193077,
+      "grad_norm": 0.21814733743667603,
+      "learning_rate": 4.77e-05,
+      "loss": 0.7161,
+      "step": 477
+    },
+    {
+      "epoch": 0.3598042905532556,
+      "grad_norm": 0.23871944844722748,
+      "learning_rate": 4.78e-05,
+      "loss": 0.6911,
+      "step": 478
+    },
+    {
+      "epoch": 0.36055701919458033,
+      "grad_norm": 0.21428927779197693,
+      "learning_rate": 4.79e-05,
+      "loss": 0.6509,
+      "step": 479
+    },
+    {
+      "epoch": 0.36130974783590514,
+      "grad_norm": 0.2249574512243271,
+      "learning_rate": 4.8e-05,
+      "loss": 0.6548,
+      "step": 480
+    },
+    {
+      "epoch": 0.36206247647722994,
+      "grad_norm": 0.23803046345710754,
+      "learning_rate": 4.8100000000000004e-05,
+      "loss": 0.6165,
+      "step": 481
+    },
+    {
+      "epoch": 0.36281520511855475,
+      "grad_norm": 0.5661165714263916,
+      "learning_rate": 4.82e-05,
+      "loss": 0.5994,
+      "step": 482
+    },
+    {
+      "epoch": 0.36356793375987956,
+      "grad_norm": 0.24691824615001678,
+      "learning_rate": 4.83e-05,
+      "loss": 0.6923,
+      "step": 483
+    },
+    {
+      "epoch": 0.36432066240120436,
+      "grad_norm": 0.2493002712726593,
+      "learning_rate": 4.8400000000000004e-05,
+      "loss": 0.6474,
+      "step": 484
+    },
+    {
+      "epoch": 0.36507339104252917,
+      "grad_norm": 0.27785494923591614,
+      "learning_rate": 4.85e-05,
+      "loss": 0.6467,
+      "step": 485
+    },
+    {
+      "epoch": 0.365826119683854,
+      "grad_norm": 0.319180428981781,
+      "learning_rate": 4.86e-05,
+      "loss": 0.6665,
+      "step": 486
+    },
+    {
+      "epoch": 0.3665788483251788,
+      "grad_norm": 0.19893111288547516,
+      "learning_rate": 4.87e-05,
+      "loss": 0.737,
+      "step": 487
+    },
+    {
+      "epoch": 0.3673315769665036,
+      "grad_norm": 0.3182182013988495,
+      "learning_rate": 4.88e-05,
+      "loss": 0.691,
+      "step": 488
+    },
+    {
+      "epoch": 0.3680843056078284,
+      "grad_norm": 0.2840924859046936,
+      "learning_rate": 4.89e-05,
+      "loss": 0.6585,
+      "step": 489
+    },
+    {
+      "epoch": 0.3688370342491532,
+      "grad_norm": 0.40775710344314575,
+      "learning_rate": 4.9e-05,
+      "loss": 0.6151,
+      "step": 490
+    },
+    {
+      "epoch": 0.36958976289047796,
+      "grad_norm": 0.36615923047065735,
+      "learning_rate": 4.91e-05,
+      "loss": 0.6423,
+      "step": 491
+    },
+    {
+      "epoch": 0.37034249153180276,
+      "grad_norm": 0.20584237575531006,
+      "learning_rate": 4.92e-05,
+      "loss": 0.6418,
+      "step": 492
+    },
+    {
+      "epoch": 0.37109522017312757,
+      "grad_norm": 0.33265167474746704,
+      "learning_rate": 4.93e-05,
+      "loss": 0.6474,
+      "step": 493
+    },
+    {
+      "epoch": 0.3718479488144524,
+      "grad_norm": 0.43340879678726196,
+      "learning_rate": 4.94e-05,
+      "loss": 0.5585,
+      "step": 494
+    },
+    {
+      "epoch": 0.3726006774557772,
+      "grad_norm": 0.42346668243408203,
+      "learning_rate": 4.9500000000000004e-05,
+      "loss": 0.7245,
+      "step": 495
+    },
+    {
+      "epoch": 0.373353406097102,
+      "grad_norm": 0.21770606935024261,
+      "learning_rate": 4.96e-05,
+      "loss": 0.5908,
+      "step": 496
+    },
+    {
+      "epoch": 0.3741061347384268,
+      "grad_norm": 0.3974795341491699,
+      "learning_rate": 4.97e-05,
+      "loss": 0.7228,
+      "step": 497
+    },
+    {
+      "epoch": 0.3748588633797516,
+      "grad_norm": 0.5115483403205872,
+      "learning_rate": 4.9800000000000004e-05,
+      "loss": 0.662,
+      "step": 498
+    },
+    {
+      "epoch": 0.3756115920210764,
+      "grad_norm": 0.4660809636116028,
+      "learning_rate": 4.99e-05,
+      "loss": 0.681,
+      "step": 499
+    },
+    {
+      "epoch": 0.3763643206624012,
+      "grad_norm": 0.39443767070770264,
+      "learning_rate": 5e-05,
+      "loss": 0.7165,
+      "step": 500
+    },
+    {
+      "epoch": 0.377117049303726,
+      "grad_norm": 0.25847744941711426,
+      "learning_rate": 4.9999998769716915e-05,
+      "loss": 0.7147,
+      "step": 501
+    },
+    {
+      "epoch": 0.37786977794505083,
+      "grad_norm": 0.30618685483932495,
+      "learning_rate": 4.999999507886779e-05,
+      "loss": 0.6325,
+      "step": 502
+    },
+    {
+      "epoch": 0.3786225065863756,
+      "grad_norm": 0.43728944659233093,
+      "learning_rate": 4.9999988927453036e-05,
+      "loss": 0.6509,
+      "step": 503
+    },
+    {
+      "epoch": 0.3793752352277004,
+      "grad_norm": 0.3647564649581909,
+      "learning_rate": 4.9999980315473316e-05,
+      "loss": 0.6342,
+      "step": 504
+    },
+    {
+      "epoch": 0.3801279638690252,
+      "grad_norm": 0.2866964042186737,
+      "learning_rate": 4.999996924292958e-05,
+      "loss": 0.6978,
+      "step": 505
+    },
+    {
+      "epoch": 0.38088069251035,
+      "grad_norm": 0.24305477738380432,
+      "learning_rate": 4.9999955709823035e-05,
+      "loss": 0.7269,
+      "step": 506
+    },
+    {
+      "epoch": 0.3816334211516748,
+      "grad_norm": 0.32219618558883667,
+      "learning_rate": 4.9999939716155165e-05,
+      "loss": 0.6613,
+      "step": 507
+    },
+    {
+      "epoch": 0.3823861497929996,
+      "grad_norm": 0.3449519872665405,
+      "learning_rate": 4.999992126192771e-05,
+      "loss": 0.5976,
+      "step": 508
+    },
+    {
+      "epoch": 0.3831388784343244,
+      "grad_norm": 0.3057398498058319,
+      "learning_rate": 4.99999003471427e-05,
+      "loss": 0.639,
+      "step": 509
+    },
+    {
+      "epoch": 0.38389160707564923,
+      "grad_norm": 0.26359298825263977,
+      "learning_rate": 4.9999876971802415e-05,
+      "loss": 0.6185,
+      "step": 510
+    },
+    {
+      "epoch": 0.38464433571697404,
+      "grad_norm": 0.31316515803337097,
+      "learning_rate": 4.9999851135909406e-05,
+      "loss": 0.5709,
+      "step": 511
+    },
+    {
+      "epoch": 0.38539706435829885,
+      "grad_norm": 0.3129493296146393,
+      "learning_rate": 4.999982283946651e-05,
+      "loss": 0.5992,
+      "step": 512
+    },
+    {
+      "epoch": 0.38614979299962365,
+      "grad_norm": 0.29396021366119385,
+      "learning_rate": 4.999979208247682e-05,
+      "loss": 0.7078,
+      "step": 513
+    },
+    {
+      "epoch": 0.38690252164094846,
+      "grad_norm": 0.2548888325691223,
+      "learning_rate": 4.999975886494369e-05,
+      "loss": 0.7154,
+      "step": 514
+    },
+    {
+      "epoch": 0.38765525028227327,
+      "grad_norm": 0.3401060700416565,
+      "learning_rate": 4.999972318687075e-05,
+      "loss": 0.6135,
+      "step": 515
+    },
+    {
+      "epoch": 0.388407978923598,
+      "grad_norm": 0.31648850440979004,
+      "learning_rate": 4.9999685048261925e-05,
+      "loss": 0.6002,
+      "step": 516
+    },
+    {
+      "epoch": 0.3891607075649228,
+      "grad_norm": 0.2970479130744934,
+      "learning_rate": 4.999964444912137e-05,
+      "loss": 0.5893,
+      "step": 517
+    },
+    {
+      "epoch": 0.38991343620624763,
+      "grad_norm": 0.2698556184768677,
+      "learning_rate": 4.9999601389453516e-05,
+      "loss": 0.6222,
+      "step": 518
+    },
+    {
+      "epoch": 0.39066616484757244,
+      "grad_norm": 0.284527450799942,
+      "learning_rate": 4.999955586926309e-05,
+      "loss": 0.5824,
+      "step": 519
+    },
+    {
+      "epoch": 0.39141889348889725,
+      "grad_norm": 0.35957175493240356,
+      "learning_rate": 4.999950788855506e-05,
+      "loss": 0.6247,
+      "step": 520
+    },
+    {
+      "epoch": 0.39217162213022205,
+      "grad_norm": 0.37566396594047546,
+      "learning_rate": 4.999945744733467e-05,
+      "loss": 0.5754,
+      "step": 521
+    },
+    {
+      "epoch": 0.39292435077154686,
+      "grad_norm": 0.32953745126724243,
+      "learning_rate": 4.999940454560744e-05,
+      "loss": 0.6423,
+      "step": 522
+    },
+    {
+      "epoch": 0.39367707941287167,
+      "grad_norm": 0.2475993037223816,
+      "learning_rate": 4.9999349183379165e-05,
+      "loss": 0.696,
+      "step": 523
+    },
+    {
+      "epoch": 0.3944298080541965,
+      "grad_norm": 0.325882226228714,
+      "learning_rate": 4.9999291360655885e-05,
+      "loss": 0.6225,
+      "step": 524
+    },
+    {
+      "epoch": 0.3951825366955213,
+      "grad_norm": 0.37260207533836365,
+      "learning_rate": 4.999923107744393e-05,
+      "loss": 0.6376,
+      "step": 525
+    },
+    {
+      "epoch": 0.3959352653368461,
+      "grad_norm": 0.3236551284790039,
+      "learning_rate": 4.9999168333749894e-05,
+      "loss": 0.5709,
+      "step": 526
+    },
+    {
+      "epoch": 0.3966879939781709,
+      "grad_norm": 0.265729159116745,
+      "learning_rate": 4.9999103129580634e-05,
+      "loss": 0.5936,
+      "step": 527
+    },
+    {
+      "epoch": 0.39744072261949565,
+      "grad_norm": 0.23996983468532562,
+      "learning_rate": 4.999903546494328e-05,
+      "loss": 0.6085,
+      "step": 528
+    },
+    {
+      "epoch": 0.39819345126082045,
+      "grad_norm": 0.36054539680480957,
+      "learning_rate": 4.9998965339845235e-05,
+      "loss": 0.7014,
+      "step": 529
+    },
+    {
+      "epoch": 0.39894617990214526,
+      "grad_norm": 0.6531224250793457,
+      "learning_rate": 4.999889275429417e-05,
+      "loss": 0.5941,
+      "step": 530
+    },
+    {
+      "epoch": 0.39969890854347007,
+      "grad_norm": 0.3937850296497345,
+      "learning_rate": 4.9998817708298024e-05,
+      "loss": 0.702,
+      "step": 531
+    },
+    {
+      "epoch": 0.4004516371847949,
+      "grad_norm": 0.482532262802124,
+      "learning_rate": 4.999874020186499e-05,
+      "loss": 0.5918,
+      "step": 532
+    },
+    {
+      "epoch": 0.4012043658261197,
+      "grad_norm": 0.5132527947425842,
+      "learning_rate": 4.9998660235003555e-05,
+      "loss": 0.6203,
+      "step": 533
+    },
+    {
+      "epoch": 0.4019570944674445,
+      "grad_norm": 0.3084738254547119,
+      "learning_rate": 4.999857780772247e-05,
+      "loss": 0.7374,
+      "step": 534
+    },
+    {
+      "epoch": 0.4027098231087693,
+      "grad_norm": 0.38917598128318787,
+      "learning_rate": 4.9998492920030745e-05,
+      "loss": 0.7196,
+      "step": 535
+    },
+    {
+      "epoch": 0.4034625517500941,
+      "grad_norm": 0.2816764712333679,
+      "learning_rate": 4.9998405571937645e-05,
+      "loss": 0.6529,
+      "step": 536
+    },
+    {
+      "epoch": 0.4042152803914189,
+      "grad_norm": 0.43532678484916687,
+      "learning_rate": 4.9998315763452743e-05,
+      "loss": 0.6208,
+      "step": 537
+    },
+    {
+      "epoch": 0.4049680090327437,
+      "grad_norm": 0.3351176679134369,
+      "learning_rate": 4.9998223494585866e-05,
+      "loss": 0.6376,
+      "step": 538
+    },
+    {
+      "epoch": 0.4057207376740685,
+      "grad_norm": 0.32272958755493164,
+      "learning_rate": 4.9998128765347084e-05,
+      "loss": 0.6343,
+      "step": 539
+    },
+    {
+      "epoch": 0.40647346631539333,
+      "grad_norm": 0.2737825810909271,
+      "learning_rate": 4.999803157574677e-05,
+      "loss": 0.6563,
+      "step": 540
+    },
+    {
+      "epoch": 0.4072261949567181,
+      "grad_norm": 0.31499072909355164,
+      "learning_rate": 4.999793192579554e-05,
+      "loss": 0.6011,
+      "step": 541
+    },
+    {
+      "epoch": 0.4079789235980429,
+      "grad_norm": 0.3173752725124359,
+      "learning_rate": 4.999782981550431e-05,
+      "loss": 0.5915,
+      "step": 542
+    },
+    {
+      "epoch": 0.4087316522393677,
+      "grad_norm": 0.29127031564712524,
+      "learning_rate": 4.999772524488423e-05,
+      "loss": 0.5977,
+      "step": 543
+    },
+    {
+      "epoch": 0.4094843808806925,
+      "grad_norm": 0.2802198529243469,
+      "learning_rate": 4.999761821394674e-05,
+      "loss": 0.6371,
+      "step": 544
+    },
+    {
+      "epoch": 0.4102371095220173,
+      "grad_norm": 0.22397077083587646,
+      "learning_rate": 4.999750872270356e-05,
+      "loss": 0.5802,
+      "step": 545
+    },
+    {
+      "epoch": 0.4109898381633421,
+      "grad_norm": 0.32689180970191956,
+      "learning_rate": 4.999739677116663e-05,
+      "loss": 0.5819,
+      "step": 546
+    },
+    {
+      "epoch": 0.4117425668046669,
+      "grad_norm": 0.45752614736557007,
+      "learning_rate": 4.9997282359348226e-05,
+      "loss": 0.5955,
+      "step": 547
+    },
+    {
+      "epoch": 0.4124952954459917,
+      "grad_norm": 0.4560548961162567,
+      "learning_rate": 4.999716548726085e-05,
+      "loss": 0.627,
+      "step": 548
+    },
+    {
+      "epoch": 0.41324802408731653,
+      "grad_norm": 0.3152342438697815,
+      "learning_rate": 4.999704615491727e-05,
+      "loss": 0.6701,
+      "step": 549
+    },
+    {
+      "epoch": 0.41400075272864134,
+      "grad_norm": 0.27433106303215027,
+      "learning_rate": 4.9996924362330555e-05,
+      "loss": 0.6027,
+      "step": 550
+    },
+    {
+      "epoch": 0.41475348136996615,
+      "grad_norm": 0.39566731452941895,
+      "learning_rate": 4.999680010951401e-05,
+      "loss": 0.6144,
+      "step": 551
+    },
+    {
+      "epoch": 0.41550621001129096,
+      "grad_norm": 0.3024057149887085,
+      "learning_rate": 4.999667339648123e-05,
+      "loss": 0.6441,
+      "step": 552
+    },
+    {
+      "epoch": 0.4162589386526157,
+      "grad_norm": 0.2917120158672333,
+      "learning_rate": 4.999654422324607e-05,
+      "loss": 0.7917,
+      "step": 553
+    },
+    {
+      "epoch": 0.4170116672939405,
+      "grad_norm": 0.3837706744670868,
+      "learning_rate": 4.999641258982266e-05,
+      "loss": 0.6514,
+      "step": 554
+    },
+    {
+      "epoch": 0.4177643959352653,
+      "grad_norm": 0.2621983289718628,
+      "learning_rate": 4.999627849622539e-05,
+      "loss": 0.7225,
+      "step": 555
+    },
+    {
+      "epoch": 0.4185171245765901,
+      "grad_norm": 0.2871634066104889,
+      "learning_rate": 4.999614194246893e-05,
+      "loss": 0.6665,
+      "step": 556
+    },
+    {
+      "epoch": 0.41926985321791493,
+      "grad_norm": 0.2743696868419647,
+      "learning_rate": 4.99960029285682e-05,
+      "loss": 0.589,
+      "step": 557
+    },
+    {
+      "epoch": 0.42002258185923974,
+      "grad_norm": 0.23471172153949738,
+      "learning_rate": 4.9995861454538425e-05,
+      "loss": 0.6076,
+      "step": 558
+    },
+    {
+      "epoch": 0.42077531050056455,
+      "grad_norm": 0.24136625230312347,
+      "learning_rate": 4.9995717520395053e-05,
+      "loss": 0.5651,
+      "step": 559
+    },
+    {
+      "epoch": 0.42152803914188935,
+      "grad_norm": 0.25181466341018677,
+      "learning_rate": 4.999557112615384e-05,
+      "loss": 0.5488,
+      "step": 560
+    },
+    {
+      "epoch": 0.42228076778321416,
+      "grad_norm": 0.22217592597007751,
+      "learning_rate": 4.9995422271830795e-05,
+      "loss": 0.6528,
+      "step": 561
+    },
+    {
+      "epoch": 0.42303349642453897,
+      "grad_norm": 0.2549673318862915,
+      "learning_rate": 4.999527095744218e-05,
+      "loss": 0.5694,
+      "step": 562
+    },
+    {
+      "epoch": 0.4237862250658638,
+      "grad_norm": 0.21260491013526917,
+      "learning_rate": 4.9995117183004556e-05,
+      "loss": 0.6708,
+      "step": 563
+    },
+    {
+      "epoch": 0.4245389537071886,
+      "grad_norm": 0.30507221817970276,
+      "learning_rate": 4.9994960948534744e-05,
+      "loss": 0.6816,
+      "step": 564
+    },
+    {
+      "epoch": 0.42529168234851333,
+      "grad_norm": 0.2624015510082245,
+      "learning_rate": 4.999480225404982e-05,
+      "loss": 0.621,
+      "step": 565
+    },
+    {
+      "epoch": 0.42604441098983814,
+      "grad_norm": 0.30238965153694153,
+      "learning_rate": 4.999464109956714e-05,
+      "loss": 0.6534,
+      "step": 566
+    },
+    {
+      "epoch": 0.42679713963116295,
+      "grad_norm": 0.4034108519554138,
+      "learning_rate": 4.999447748510434e-05,
+      "loss": 0.6433,
+      "step": 567
+    },
+    {
+      "epoch": 0.42754986827248775,
+      "grad_norm": 0.22751101851463318,
+      "learning_rate": 4.999431141067928e-05,
+      "loss": 0.7097,
+      "step": 568
+    },
+    {
+      "epoch": 0.42830259691381256,
+      "grad_norm": 0.33946219086647034,
+      "learning_rate": 4.999414287631017e-05,
+      "loss": 0.6284,
+      "step": 569
+    },
+    {
+      "epoch": 0.42905532555513737,
+      "grad_norm": 0.46905529499053955,
+      "learning_rate": 4.99939718820154e-05,
+      "loss": 0.6739,
+      "step": 570
+    },
+    {
+      "epoch": 0.4298080541964622,
+      "grad_norm": 0.4806632101535797,
+      "learning_rate": 4.9993798427813685e-05,
+      "loss": 0.6228,
+      "step": 571
+    },
+    {
+      "epoch": 0.430560782837787,
+      "grad_norm": 0.4093143343925476,
+      "learning_rate": 4.9993622513724e-05,
+      "loss": 0.4863,
+      "step": 572
+    },
+    {
+      "epoch": 0.4313135114791118,
+      "grad_norm": 0.34644851088523865,
+      "learning_rate": 4.999344413976557e-05,
+      "loss": 0.6749,
+      "step": 573
+    },
+    {
+      "epoch": 0.4320662401204366,
+      "grad_norm": 0.24832235276699066,
+      "learning_rate": 4.99932633059579e-05,
+      "loss": 0.5861,
+      "step": 574
+    },
+    {
+      "epoch": 0.4328189687617614,
+      "grad_norm": 0.35261133313179016,
+      "learning_rate": 4.9993080012320785e-05,
+      "loss": 0.6071,
+      "step": 575
+    },
+    {
+      "epoch": 0.4335716974030862,
+      "grad_norm": 0.4568660855293274,
+      "learning_rate": 4.999289425887425e-05,
+      "loss": 0.6591,
+      "step": 576
+    },
+    {
+      "epoch": 0.434324426044411,
+      "grad_norm": 0.3738342225551605,
+      "learning_rate": 4.9992706045638626e-05,
+      "loss": 0.6112,
+      "step": 577
+    },
+    {
+      "epoch": 0.43507715468573577,
+      "grad_norm": 0.3167920708656311,
+      "learning_rate": 4.999251537263448e-05,
+      "loss": 0.568,
+      "step": 578
+    },
+    {
+      "epoch": 0.4358298833270606,
+      "grad_norm": 0.3619365096092224,
+      "learning_rate": 4.9992322239882664e-05,
+      "loss": 0.6157,
+      "step": 579
+    },
+    {
+      "epoch": 0.4365826119683854,
+      "grad_norm": 0.22076496481895447,
+      "learning_rate": 4.9992126647404313e-05,
+      "loss": 0.6748,
+      "step": 580
+    },
+    {
+      "epoch": 0.4373353406097102,
+      "grad_norm": 0.26996174454689026,
+      "learning_rate": 4.99919285952208e-05,
+      "loss": 0.6581,
+      "step": 581
+    },
+    {
+      "epoch": 0.438088069251035,
+      "grad_norm": 0.24129411578178406,
+      "learning_rate": 4.99917280833538e-05,
+      "loss": 0.6649,
+      "step": 582
+    },
+    {
+      "epoch": 0.4388407978923598,
+      "grad_norm": 0.33558812737464905,
+      "learning_rate": 4.9991525111825235e-05,
+      "loss": 0.5453,
+      "step": 583
+    },
+    {
+      "epoch": 0.4395935265336846,
+      "grad_norm": 0.24653978645801544,
+      "learning_rate": 4.999131968065729e-05,
+      "loss": 0.537,
+      "step": 584
+    },
+    {
+      "epoch": 0.4403462551750094,
+      "grad_norm": 0.2900659441947937,
+      "learning_rate": 4.999111178987244e-05,
+      "loss": 0.6402,
+      "step": 585
+    },
+    {
+      "epoch": 0.4410989838163342,
+      "grad_norm": 0.4853473901748657,
+      "learning_rate": 4.9990901439493425e-05,
+      "loss": 0.5875,
+      "step": 586
+    },
+    {
+      "epoch": 0.44185171245765903,
+      "grad_norm": 0.3388232886791229,
+      "learning_rate": 4.999068862954325e-05,
+      "loss": 0.6597,
+      "step": 587
+    },
+    {
+      "epoch": 0.44260444109898384,
+      "grad_norm": 0.23783810436725616,
+      "learning_rate": 4.999047336004517e-05,
+      "loss": 0.6917,
+      "step": 588
+    },
+    {
+      "epoch": 0.44335716974030864,
+      "grad_norm": 0.30284425616264343,
+      "learning_rate": 4.999025563102274e-05,
+      "loss": 0.6691,
+      "step": 589
+    },
+    {
+      "epoch": 0.4441098983816334,
+      "grad_norm": 0.3457498550415039,
+      "learning_rate": 4.999003544249976e-05,
+      "loss": 0.6855,
+      "step": 590
+    },
+    {
+      "epoch": 0.4448626270229582,
+      "grad_norm": 0.2762914001941681,
+      "learning_rate": 4.998981279450032e-05,
+      "loss": 0.617,
+      "step": 591
+    },
+    {
+      "epoch": 0.445615355664283,
+      "grad_norm": 0.2486667037010193,
+      "learning_rate": 4.998958768704878e-05,
+      "loss": 0.6539,
+      "step": 592
+    },
+    {
+      "epoch": 0.4463680843056078,
+      "grad_norm": 0.31929078698158264,
+      "learning_rate": 4.998936012016974e-05,
+      "loss": 0.5682,
+      "step": 593
+    },
+    {
+      "epoch": 0.4471208129469326,
+      "grad_norm": 0.2948927581310272,
+      "learning_rate": 4.998913009388807e-05,
+      "loss": 0.6155,
+      "step": 594
+    },
+    {
+      "epoch": 0.44787354158825743,
+      "grad_norm": 0.24306873977184296,
+      "learning_rate": 4.9988897608228966e-05,
+      "loss": 0.6322,
+      "step": 595
+    },
+    {
+      "epoch": 0.44862627022958224,
+      "grad_norm": 0.39831748604774475,
+      "learning_rate": 4.998866266321782e-05,
+      "loss": 0.5934,
+      "step": 596
+    },
+    {
+      "epoch": 0.44937899887090704,
+      "grad_norm": 0.3042472302913666,
+      "learning_rate": 4.998842525888034e-05,
+      "loss": 0.6677,
+      "step": 597
+    },
+    {
+      "epoch": 0.45013172751223185,
+      "grad_norm": 0.23539121448993683,
+      "learning_rate": 4.998818539524248e-05,
+      "loss": 0.5415,
+      "step": 598
+    },
+    {
+      "epoch": 0.45088445615355666,
+      "grad_norm": 0.4198962450027466,
+      "learning_rate": 4.998794307233048e-05,
+      "loss": 0.5613,
+      "step": 599
+    },
+    {
+      "epoch": 0.45163718479488146,
+      "grad_norm": 0.7627137303352356,
+      "learning_rate": 4.998769829017084e-05,
+      "loss": 0.5555,
+      "step": 600
+    },
+    {
+      "epoch": 0.45163718479488146,
+      "eval_loss": 0.6112717986106873,
+      "eval_runtime": 453.0314,
+      "eval_samples_per_second": 21.25,
+      "eval_steps_per_second": 0.664,
+      "step": 600
+    },
+    {
+      "epoch": 0.45238991343620627,
+      "grad_norm": 0.6538589000701904,
+      "learning_rate": 4.998745104879031e-05,
+      "loss": 0.659,
+      "step": 601
+    },
+    {
+      "epoch": 0.4531426420775311,
+      "grad_norm": 0.4283027648925781,
+      "learning_rate": 4.9987201348215954e-05,
+      "loss": 0.6333,
+      "step": 602
+    },
+    {
+      "epoch": 0.45389537071885583,
+      "grad_norm": 0.8239355087280273,
+      "learning_rate": 4.998694918847507e-05,
+      "loss": 0.68,
+      "step": 603
+    },
+    {
+      "epoch": 0.45464809936018064,
+      "grad_norm": 0.4322781264781952,
+      "learning_rate": 4.9986694569595224e-05,
+      "loss": 0.663,
+      "step": 604
+    },
+    {
+      "epoch": 0.45540082800150544,
+      "grad_norm": 0.5108072757720947,
+      "learning_rate": 4.9986437491604276e-05,
+      "loss": 0.6546,
+      "step": 605
+    },
+    {
+      "epoch": 0.45615355664283025,
+      "grad_norm": 0.31788963079452515,
+      "learning_rate": 4.9986177954530335e-05,
+      "loss": 0.6283,
+      "step": 606
+    },
+    {
+      "epoch": 0.45690628528415506,
+      "grad_norm": 0.30972692370414734,
+      "learning_rate": 4.998591595840177e-05,
+      "loss": 0.6257,
+      "step": 607
+    },
+    {
+      "epoch": 0.45765901392547986,
+      "grad_norm": 0.2658657133579254,
+      "learning_rate": 4.998565150324724e-05,
+      "loss": 0.6656,
+      "step": 608
+    },
+    {
+      "epoch": 0.45841174256680467,
+      "grad_norm": 0.34154659509658813,
+      "learning_rate": 4.9985384589095674e-05,
+      "loss": 0.7173,
+      "step": 609
+    },
+    {
+      "epoch": 0.4591644712081295,
+      "grad_norm": 0.35295370221138,
+      "learning_rate": 4.9985115215976255e-05,
+      "loss": 0.5608,
+      "step": 610
+    },
+    {
+      "epoch": 0.4599171998494543,
+      "grad_norm": 0.6042841672897339,
+      "learning_rate": 4.998484338391845e-05,
+      "loss": 0.6667,
+      "step": 611
+    },
+    {
+      "epoch": 0.4606699284907791,
+      "grad_norm": 0.30557700991630554,
+      "learning_rate": 4.998456909295197e-05,
+      "loss": 0.6101,
+      "step": 612
+    },
+    {
+      "epoch": 0.4614226571321039,
+      "grad_norm": 0.36563998460769653,
+      "learning_rate": 4.998429234310682e-05,
+      "loss": 0.5914,
+      "step": 613
+    },
+    {
+      "epoch": 0.4621753857734287,
+      "grad_norm": 0.32732459902763367,
+      "learning_rate": 4.998401313441326e-05,
+      "loss": 0.6234,
+      "step": 614
+    },
+    {
+      "epoch": 0.46292811441475346,
+      "grad_norm": 0.3308684825897217,
+      "learning_rate": 4.998373146690183e-05,
+      "loss": 0.6672,
+      "step": 615
+    },
+    {
+      "epoch": 0.46368084305607826,
+      "grad_norm": 0.33196374773979187,
+      "learning_rate": 4.998344734060333e-05,
+      "loss": 0.627,
+      "step": 616
+    },
+    {
+      "epoch": 0.46443357169740307,
+      "grad_norm": 0.2758849263191223,
+      "learning_rate": 4.998316075554884e-05,
+      "loss": 0.5848,
+      "step": 617
+    },
+    {
+      "epoch": 0.4651863003387279,
+      "grad_norm": 0.4222168028354645,
+      "learning_rate": 4.998287171176968e-05,
+      "loss": 0.6781,
+      "step": 618
+    },
+    {
+      "epoch": 0.4659390289800527,
+      "grad_norm": 0.27342143654823303,
+      "learning_rate": 4.998258020929748e-05,
+      "loss": 0.7125,
+      "step": 619
+    },
+    {
+      "epoch": 0.4666917576213775,
+      "grad_norm": 0.29168641567230225,
+      "learning_rate": 4.998228624816412e-05,
+      "loss": 0.7025,
+      "step": 620
+    },
+    {
+      "epoch": 0.4674444862627023,
+      "grad_norm": 0.3021078109741211,
+      "learning_rate": 4.998198982840172e-05,
+      "loss": 0.6359,
+      "step": 621
+    },
+    {
+      "epoch": 0.4681972149040271,
+      "grad_norm": 0.4678989052772522,
+      "learning_rate": 4.998169095004272e-05,
+      "loss": 0.6391,
+      "step": 622
+    },
+    {
+      "epoch": 0.4689499435453519,
+      "grad_norm": 0.28198936581611633,
+      "learning_rate": 4.99813896131198e-05,
+      "loss": 0.6417,
+      "step": 623
+    },
+    {
+      "epoch": 0.4697026721866767,
+      "grad_norm": 0.3174442648887634,
+      "learning_rate": 4.9981085817665916e-05,
+      "loss": 0.7031,
+      "step": 624
+    },
+    {
+      "epoch": 0.4704554008280015,
+      "grad_norm": 0.2910943925380707,
+      "learning_rate": 4.9980779563714274e-05,
+      "loss": 0.709,
+      "step": 625
+    },
+    {
+      "epoch": 0.47120812946932633,
+      "grad_norm": 0.34364739060401917,
+      "learning_rate": 4.9980470851298387e-05,
+      "loss": 0.6937,
+      "step": 626
+    },
+    {
+      "epoch": 0.47196085811065114,
+      "grad_norm": 0.27657550573349,
+      "learning_rate": 4.9980159680452006e-05,
+      "loss": 0.7097,
+      "step": 627
+    },
+    {
+      "epoch": 0.4727135867519759,
+      "grad_norm": 0.28776273131370544,
+      "learning_rate": 4.9979846051209164e-05,
+      "loss": 0.6393,
+      "step": 628
+    },
+    {
+      "epoch": 0.4734663153933007,
+      "grad_norm": 0.27483728528022766,
+      "learning_rate": 4.9979529963604146e-05,
+      "loss": 0.6139,
+      "step": 629
+    },
+    {
+      "epoch": 0.4742190440346255,
+      "grad_norm": 0.33207690715789795,
+      "learning_rate": 4.9979211417671534e-05,
+      "loss": 0.6577,
+      "step": 630
+    },
+    {
+      "epoch": 0.4749717726759503,
+      "grad_norm": 0.26985740661621094,
+      "learning_rate": 4.997889041344616e-05,
+      "loss": 0.5348,
+      "step": 631
+    },
+    {
+      "epoch": 0.4757245013172751,
+      "grad_norm": 0.23181799054145813,
+      "learning_rate": 4.997856695096312e-05,
+      "loss": 0.6108,
+      "step": 632
+    },
+    {
+      "epoch": 0.4764772299585999,
+      "grad_norm": 0.21733063459396362,
+      "learning_rate": 4.9978241030257813e-05,
+      "loss": 0.5277,
+      "step": 633
+    },
+    {
+      "epoch": 0.47722995859992473,
+      "grad_norm": 0.2762012183666229,
+      "learning_rate": 4.9977912651365844e-05,
+      "loss": 0.6741,
+      "step": 634
+    },
+    {
+      "epoch": 0.47798268724124954,
+      "grad_norm": 0.2916512191295624,
+      "learning_rate": 4.9977581814323146e-05,
+      "loss": 0.591,
+      "step": 635
+    },
+    {
+      "epoch": 0.47873541588257434,
+      "grad_norm": 0.2599298059940338,
+      "learning_rate": 4.99772485191659e-05,
+      "loss": 0.488,
+      "step": 636
+    },
+    {
+      "epoch": 0.47948814452389915,
+      "grad_norm": 0.23463739454746246,
+      "learning_rate": 4.997691276593055e-05,
+      "loss": 0.642,
+      "step": 637
+    },
+    {
+      "epoch": 0.48024087316522396,
+      "grad_norm": 0.24529848992824554,
+      "learning_rate": 4.997657455465381e-05,
+      "loss": 0.6612,
+      "step": 638
+    },
+    {
+      "epoch": 0.48099360180654877,
+      "grad_norm": 0.23981422185897827,
+      "learning_rate": 4.9976233885372675e-05,
+      "loss": 0.6982,
+      "step": 639
+    },
+    {
+      "epoch": 0.4817463304478735,
+      "grad_norm": 0.25977396965026855,
+      "learning_rate": 4.997589075812439e-05,
+      "loss": 0.6699,
+      "step": 640
+    },
+    {
+      "epoch": 0.4824990590891983,
+      "grad_norm": 0.20296330749988556,
+      "learning_rate": 4.997554517294648e-05,
+      "loss": 0.6112,
+      "step": 641
+    },
+    {
+      "epoch": 0.48325178773052313,
+      "grad_norm": 0.24328577518463135,
+      "learning_rate": 4.997519712987674e-05,
+      "loss": 0.6298,
+      "step": 642
+    },
+    {
+      "epoch": 0.48400451637184794,
+      "grad_norm": 0.2663193643093109,
+      "learning_rate": 4.997484662895325e-05,
+      "loss": 0.5493,
+      "step": 643
+    },
+    {
+      "epoch": 0.48475724501317274,
+      "grad_norm": 0.45112091302871704,
+      "learning_rate": 4.997449367021431e-05,
+      "loss": 0.6295,
+      "step": 644
+    },
+    {
+      "epoch": 0.48550997365449755,
+      "grad_norm": 0.27620550990104675,
+      "learning_rate": 4.9974138253698535e-05,
+      "loss": 0.6496,
+      "step": 645
+    },
+    {
+      "epoch": 0.48626270229582236,
+      "grad_norm": 0.35247528553009033,
+      "learning_rate": 4.997378037944478e-05,
+      "loss": 0.5433,
+      "step": 646
+    },
+    {
+      "epoch": 0.48701543093714716,
+      "grad_norm": 0.3622686564922333,
+      "learning_rate": 4.99734200474922e-05,
+      "loss": 0.5429,
+      "step": 647
+    },
+    {
+      "epoch": 0.48776815957847197,
+      "grad_norm": 0.27765002846717834,
+      "learning_rate": 4.997305725788019e-05,
+      "loss": 0.708,
+      "step": 648
+    },
+    {
+      "epoch": 0.4885208882197968,
+      "grad_norm": 0.2766205966472626,
+      "learning_rate": 4.997269201064843e-05,
+      "loss": 0.6311,
+      "step": 649
+    },
+    {
+      "epoch": 0.4892736168611216,
+      "grad_norm": 0.31013283133506775,
+      "learning_rate": 4.997232430583686e-05,
+      "loss": 0.6371,
+      "step": 650
+    },
+    {
+      "epoch": 0.4900263455024464,
+      "grad_norm": 0.3680330216884613,
+      "learning_rate": 4.997195414348569e-05,
+      "loss": 0.6639,
+      "step": 651
+    },
+    {
+      "epoch": 0.49077907414377114,
+      "grad_norm": 0.30095064640045166,
+      "learning_rate": 4.99715815236354e-05,
+      "loss": 0.6502,
+      "step": 652
+    },
+    {
+      "epoch": 0.49153180278509595,
+      "grad_norm": 0.3297145664691925,
+      "learning_rate": 4.9971206446326735e-05,
+      "loss": 0.5553,
+      "step": 653
+    },
+    {
+      "epoch": 0.49228453142642076,
+      "grad_norm": 0.6849196553230286,
+      "learning_rate": 4.9970828911600727e-05,
+      "loss": 0.6028,
+      "step": 654
+    },
+    {
+      "epoch": 0.49303726006774556,
+      "grad_norm": 0.3415966033935547,
+      "learning_rate": 4.997044891949864e-05,
+      "loss": 0.5294,
+      "step": 655
+    },
+    {
+      "epoch": 0.49378998870907037,
+      "grad_norm": 0.3403933346271515,
+      "learning_rate": 4.997006647006205e-05,
+      "loss": 0.6099,
+      "step": 656
+    },
+    {
+      "epoch": 0.4945427173503952,
+      "grad_norm": 0.3147316873073578,
+      "learning_rate": 4.9969681563332785e-05,
+      "loss": 0.5755,
+      "step": 657
+    },
+    {
+      "epoch": 0.49529544599172,
+      "grad_norm": 0.42097970843315125,
+      "learning_rate": 4.9969294199352904e-05,
+      "loss": 0.6484,
+      "step": 658
+    },
+    {
+      "epoch": 0.4960481746330448,
+      "grad_norm": 0.3851207196712494,
+      "learning_rate": 4.996890437816481e-05,
+      "loss": 0.7095,
+      "step": 659
+    },
+    {
+      "epoch": 0.4968009032743696,
+      "grad_norm": 0.2845657467842102,
+      "learning_rate": 4.996851209981111e-05,
+      "loss": 0.6306,
+      "step": 660
+    },
+    {
+      "epoch": 0.4975536319156944,
+      "grad_norm": 0.3427131474018097,
+      "learning_rate": 4.996811736433471e-05,
+      "loss": 0.5907,
+      "step": 661
+    },
+    {
+      "epoch": 0.4983063605570192,
+      "grad_norm": 0.3022371828556061,
+      "learning_rate": 4.996772017177877e-05,
+      "loss": 0.5786,
+      "step": 662
+    },
+    {
+      "epoch": 0.499059089198344,
+      "grad_norm": 0.41582608222961426,
+      "learning_rate": 4.996732052218673e-05,
+      "loss": 0.6118,
+      "step": 663
+    },
+    {
+      "epoch": 0.4998118178396688,
+      "grad_norm": 0.3634607791900635,
+      "learning_rate": 4.9966918415602306e-05,
+      "loss": 0.544,
+      "step": 664
+    },
+    {
+      "epoch": 0.5005645464809936,
+      "grad_norm": 0.2934187948703766,
+      "learning_rate": 4.996651385206945e-05,
+      "loss": 0.5431,
+      "step": 665
+    },
+    {
+      "epoch": 0.5013172751223184,
+      "grad_norm": 0.4482651650905609,
+      "learning_rate": 4.996610683163243e-05,
+      "loss": 0.5123,
+      "step": 666
+    },
+    {
+      "epoch": 0.5020700037636432,
+      "grad_norm": 0.32857245206832886,
+      "learning_rate": 4.9965697354335735e-05,
+      "loss": 0.6112,
+      "step": 667
+    },
+    {
+      "epoch": 0.502822732404968,
+      "grad_norm": 0.24476905167102814,
+      "learning_rate": 4.996528542022416e-05,
+      "loss": 0.6165,
+      "step": 668
+    },
+    {
+      "epoch": 0.5035754610462928,
+      "grad_norm": 0.22776851058006287,
+      "learning_rate": 4.9964871029342747e-05,
+      "loss": 0.5543,
+      "step": 669
+    },
+    {
+      "epoch": 0.5043281896876176,
+      "grad_norm": 0.24143142998218536,
+      "learning_rate": 4.996445418173681e-05,
+      "loss": 0.5137,
+      "step": 670
+    },
+    {
+      "epoch": 0.5050809183289424,
+      "grad_norm": 0.2820020020008087,
+      "learning_rate": 4.9964034877451944e-05,
+      "loss": 0.6565,
+      "step": 671
+    },
+    {
+      "epoch": 0.5058336469702672,
+      "grad_norm": 0.26462996006011963,
+      "learning_rate": 4.996361311653399e-05,
+      "loss": 0.545,
+      "step": 672
+    },
+    {
+      "epoch": 0.506586375611592,
+      "grad_norm": 0.26020970940589905,
+      "learning_rate": 4.9963188899029085e-05,
+      "loss": 0.5952,
+      "step": 673
+    },
+    {
+      "epoch": 0.5073391042529168,
+      "grad_norm": 0.25098127126693726,
+      "learning_rate": 4.996276222498361e-05,
+      "loss": 0.659,
+      "step": 674
+    },
+    {
+      "epoch": 0.5080918328942416,
+      "grad_norm": 0.23919065296649933,
+      "learning_rate": 4.996233309444423e-05,
+      "loss": 0.4934,
+      "step": 675
+    },
+    {
+      "epoch": 0.5088445615355665,
+      "grad_norm": 0.24091127514839172,
+      "learning_rate": 4.996190150745788e-05,
+      "loss": 0.4918,
+      "step": 676
+    },
+    {
+      "epoch": 0.5095972901768913,
+      "grad_norm": 0.3139805793762207,
+      "learning_rate": 4.9961467464071756e-05,
+      "loss": 0.672,
+      "step": 677
+    },
+    {
+      "epoch": 0.5103500188182161,
+      "grad_norm": 0.24265427887439728,
+      "learning_rate": 4.9961030964333316e-05,
+      "loss": 0.6243,
+      "step": 678
+    },
+    {
+      "epoch": 0.5111027474595409,
+      "grad_norm": 0.3516172766685486,
+      "learning_rate": 4.99605920082903e-05,
+      "loss": 0.7398,
+      "step": 679
+    },
+    {
+      "epoch": 0.5118554761008657,
+      "grad_norm": 0.2792483866214752,
+      "learning_rate": 4.9960150595990704e-05,
+      "loss": 0.6315,
+      "step": 680
+    },
+    {
+      "epoch": 0.5126082047421905,
+      "grad_norm": 0.3756444454193115,
+      "learning_rate": 4.995970672748281e-05,
+      "loss": 0.613,
+      "step": 681
+    },
+    {
+      "epoch": 0.5133609333835153,
+      "grad_norm": 0.2870337963104248,
+      "learning_rate": 4.995926040281516e-05,
+      "loss": 0.6025,
+      "step": 682
+    },
+    {
+      "epoch": 0.51411366202484,
+      "grad_norm": 0.42255210876464844,
+      "learning_rate": 4.995881162203656e-05,
+      "loss": 0.636,
+      "step": 683
+    },
+    {
+      "epoch": 0.5148663906661648,
+      "grad_norm": 0.2785627245903015,
+      "learning_rate": 4.995836038519608e-05,
+      "loss": 0.6123,
+      "step": 684
+    },
+    {
+      "epoch": 0.5156191193074896,
+      "grad_norm": 0.519024133682251,
+      "learning_rate": 4.995790669234308e-05,
+      "loss": 0.6761,
+      "step": 685
+    },
+    {
+      "epoch": 0.5163718479488144,
+      "grad_norm": 0.3842090964317322,
+      "learning_rate": 4.9957450543527165e-05,
+      "loss": 0.5823,
+      "step": 686
+    },
+    {
+      "epoch": 0.5171245765901392,
+      "grad_norm": 0.29842182993888855,
+      "learning_rate": 4.9956991938798225e-05,
+      "loss": 0.5511,
+      "step": 687
+    },
+    {
+      "epoch": 0.517877305231464,
+      "grad_norm": 0.3234037458896637,
+      "learning_rate": 4.995653087820641e-05,
+      "loss": 0.511,
+      "step": 688
+    },
+    {
+      "epoch": 0.5186300338727888,
+      "grad_norm": 0.41808995604515076,
+      "learning_rate": 4.9956067361802145e-05,
+      "loss": 0.608,
+      "step": 689
+    },
+    {
+      "epoch": 0.5193827625141136,
+      "grad_norm": 0.392333984375,
+      "learning_rate": 4.9955601389636114e-05,
+      "loss": 0.6364,
+      "step": 690
+    },
+    {
+      "epoch": 0.5201354911554384,
+      "grad_norm": 0.3020949363708496,
+      "learning_rate": 4.995513296175927e-05,
+      "loss": 0.7087,
+      "step": 691
+    },
+    {
+      "epoch": 0.5208882197967633,
+      "grad_norm": 0.2431594729423523,
+      "learning_rate": 4.9954662078222856e-05,
+      "loss": 0.5814,
+      "step": 692
+    },
+    {
+      "epoch": 0.5216409484380881,
+      "grad_norm": 0.2813843786716461,
+      "learning_rate": 4.9954188739078345e-05,
+      "loss": 0.5857,
+      "step": 693
+    },
+    {
+      "epoch": 0.5223936770794129,
+      "grad_norm": 0.21805472671985626,
+      "learning_rate": 4.9953712944377526e-05,
+      "loss": 0.6172,
+      "step": 694
+    },
+    {
+      "epoch": 0.5231464057207377,
+      "grad_norm": 0.23222240805625916,
+      "learning_rate": 4.9953234694172416e-05,
+      "loss": 0.6446,
+      "step": 695
+    },
+    {
+      "epoch": 0.5238991343620625,
+      "grad_norm": 0.24114561080932617,
+      "learning_rate": 4.995275398851531e-05,
+      "loss": 0.6943,
+      "step": 696
+    },
+    {
+      "epoch": 0.5246518630033873,
+      "grad_norm": 0.29992929100990295,
+      "learning_rate": 4.9952270827458786e-05,
+      "loss": 0.6027,
+      "step": 697
+    },
+    {
+      "epoch": 0.5254045916447121,
+      "grad_norm": 0.2846331000328064,
+      "learning_rate": 4.9951785211055685e-05,
+      "loss": 0.6081,
+      "step": 698
+    },
+    {
+      "epoch": 0.5261573202860369,
+      "grad_norm": 0.3010183870792389,
+      "learning_rate": 4.995129713935911e-05,
+      "loss": 0.6225,
+      "step": 699
+    },
+    {
+      "epoch": 0.5269100489273617,
+      "grad_norm": 0.34363988041877747,
+      "learning_rate": 4.995080661242243e-05,
+      "loss": 0.5799,
+      "step": 700
+    },
+    {
+      "epoch": 0.5276627775686865,
+      "grad_norm": 0.392147421836853,
+      "learning_rate": 4.99503136302993e-05,
+      "loss": 0.5269,
+      "step": 701
+    },
+    {
+      "epoch": 0.5284155062100113,
+      "grad_norm": 0.38512519001960754,
+      "learning_rate": 4.994981819304361e-05,
+      "loss": 0.6255,
+      "step": 702
+    },
+    {
+      "epoch": 0.5291682348513361,
+      "grad_norm": 0.2538129389286041,
+      "learning_rate": 4.994932030070957e-05,
+      "loss": 0.6071,
+      "step": 703
+    },
+    {
+      "epoch": 0.5299209634926609,
+      "grad_norm": 0.24356362223625183,
+      "learning_rate": 4.994881995335161e-05,
+      "loss": 0.629,
+      "step": 704
+    },
+    {
+      "epoch": 0.5306736921339857,
+      "grad_norm": 0.33041203022003174,
+      "learning_rate": 4.9948317151024454e-05,
+      "loss": 0.6143,
+      "step": 705
+    },
+    {
+      "epoch": 0.5314264207753105,
+      "grad_norm": 0.3716380000114441,
+      "learning_rate": 4.994781189378308e-05,
+      "loss": 0.5685,
+      "step": 706
+    },
+    {
+      "epoch": 0.5321791494166354,
+      "grad_norm": 0.36931291222572327,
+      "learning_rate": 4.9947304181682746e-05,
+      "loss": 0.6084,
+      "step": 707
+    },
+    {
+      "epoch": 0.53293187805796,
+      "grad_norm": 0.2867973744869232,
+      "learning_rate": 4.994679401477899e-05,
+      "loss": 0.5874,
+      "step": 708
+    },
+    {
+      "epoch": 0.5336846066992849,
+      "grad_norm": 0.28219693899154663,
+      "learning_rate": 4.994628139312757e-05,
+      "loss": 0.5378,
+      "step": 709
+    },
+    {
+      "epoch": 0.5344373353406097,
+      "grad_norm": 0.32944077253341675,
+      "learning_rate": 4.994576631678457e-05,
+      "loss": 0.6984,
+      "step": 710
+    },
+    {
+      "epoch": 0.5351900639819345,
+      "grad_norm": 0.2578486204147339,
+      "learning_rate": 4.994524878580632e-05,
+      "loss": 0.6188,
+      "step": 711
+    },
+    {
+      "epoch": 0.5359427926232593,
+      "grad_norm": 0.2731911540031433,
+      "learning_rate": 4.99447288002494e-05,
+      "loss": 0.7179,
+      "step": 712
+    },
+    {
+      "epoch": 0.5366955212645841,
+      "grad_norm": 0.3032434582710266,
+      "learning_rate": 4.9944206360170686e-05,
+      "loss": 0.5694,
+      "step": 713
+    },
+    {
+      "epoch": 0.5374482499059089,
+      "grad_norm": 0.35488519072532654,
+      "learning_rate": 4.994368146562731e-05,
+      "loss": 0.5584,
+      "step": 714
+    },
+    {
+      "epoch": 0.5382009785472337,
+      "grad_norm": 0.3443223834037781,
+      "learning_rate": 4.994315411667667e-05,
+      "loss": 0.6475,
+      "step": 715
+    },
+    {
+      "epoch": 0.5389537071885585,
+      "grad_norm": 0.267353355884552,
+      "learning_rate": 4.994262431337644e-05,
+      "loss": 0.538,
+      "step": 716
+    },
+    {
+      "epoch": 0.5397064358298833,
+      "grad_norm": 0.31123363971710205,
+      "learning_rate": 4.9942092055784555e-05,
+      "loss": 0.6593,
+      "step": 717
+    },
+    {
+      "epoch": 0.5404591644712081,
+      "grad_norm": 0.3176477551460266,
+      "learning_rate": 4.994155734395923e-05,
+      "loss": 0.5627,
+      "step": 718
+    },
+    {
+      "epoch": 0.5412118931125329,
+      "grad_norm": 0.24147869646549225,
+      "learning_rate": 4.994102017795893e-05,
+      "loss": 0.6058,
+      "step": 719
+    },
+    {
+      "epoch": 0.5419646217538577,
+      "grad_norm": 0.26703304052352905,
+      "learning_rate": 4.99404805578424e-05,
+      "loss": 0.5773,
+      "step": 720
+    },
+    {
+      "epoch": 0.5427173503951825,
+      "grad_norm": 0.27235594391822815,
+      "learning_rate": 4.993993848366866e-05,
+      "loss": 0.6202,
+      "step": 721
+    },
+    {
+      "epoch": 0.5434700790365073,
+      "grad_norm": 0.2618643045425415,
+      "learning_rate": 4.9939393955496986e-05,
+      "loss": 0.6022,
+      "step": 722
+    },
+    {
+      "epoch": 0.5442228076778322,
+      "grad_norm": 0.24146534502506256,
+      "learning_rate": 4.993884697338692e-05,
+      "loss": 0.6013,
+      "step": 723
+    },
+    {
+      "epoch": 0.544975536319157,
+      "grad_norm": 0.2572357952594757,
+      "learning_rate": 4.9938297537398286e-05,
+      "loss": 0.6807,
+      "step": 724
+    },
+    {
+      "epoch": 0.5457282649604818,
+      "grad_norm": 0.2959592342376709,
+      "learning_rate": 4.993774564759118e-05,
+      "loss": 0.5583,
+      "step": 725
+    },
+    {
+      "epoch": 0.5464809936018066,
+      "grad_norm": 0.26018428802490234,
+      "learning_rate": 4.993719130402593e-05,
+      "loss": 0.644,
+      "step": 726
+    },
+    {
+      "epoch": 0.5472337222431314,
+      "grad_norm": 0.2789685130119324,
+      "learning_rate": 4.993663450676319e-05,
+      "loss": 0.5916,
+      "step": 727
+    },
+    {
+      "epoch": 0.5479864508844562,
+      "grad_norm": 0.33344516158103943,
+      "learning_rate": 4.9936075255863814e-05,
+      "loss": 0.5427,
+      "step": 728
+    },
+    {
+      "epoch": 0.548739179525781,
+      "grad_norm": 0.39214611053466797,
+      "learning_rate": 4.993551355138899e-05,
+      "loss": 0.6256,
+      "step": 729
+    },
+    {
+      "epoch": 0.5494919081671058,
+      "grad_norm": 0.3895111083984375,
+      "learning_rate": 4.993494939340014e-05,
+      "loss": 0.6532,
+      "step": 730
+    },
+    {
+      "epoch": 0.5502446368084306,
+      "grad_norm": 0.26267316937446594,
+      "learning_rate": 4.993438278195894e-05,
+      "loss": 0.4946,
+      "step": 731
+    },
+    {
+      "epoch": 0.5509973654497554,
+      "grad_norm": 0.2939053177833557,
+      "learning_rate": 4.993381371712738e-05,
+      "loss": 0.642,
+      "step": 732
+    },
+    {
+      "epoch": 0.5517500940910801,
+      "grad_norm": 0.3337375223636627,
+      "learning_rate": 4.993324219896768e-05,
+      "loss": 0.6466,
+      "step": 733
+    },
+    {
+      "epoch": 0.5525028227324049,
+      "grad_norm": 0.2528407573699951,
+      "learning_rate": 4.993266822754233e-05,
+      "loss": 0.5295,
+      "step": 734
+    },
+    {
+      "epoch": 0.5532555513737297,
+      "grad_norm": 0.34322038292884827,
+      "learning_rate": 4.993209180291413e-05,
+      "loss": 0.6461,
+      "step": 735
+    },
+    {
+      "epoch": 0.5540082800150545,
+      "grad_norm": 0.3924941420555115,
+      "learning_rate": 4.993151292514608e-05,
+      "loss": 0.6588,
+      "step": 736
+    },
+    {
+      "epoch": 0.5547610086563793,
+      "grad_norm": 0.28860417008399963,
+      "learning_rate": 4.993093159430151e-05,
+      "loss": 0.5731,
+      "step": 737
+    },
+    {
+      "epoch": 0.5555137372977041,
+      "grad_norm": 0.20253914594650269,
+      "learning_rate": 4.993034781044398e-05,
+      "loss": 0.5654,
+      "step": 738
+    },
+    {
+      "epoch": 0.556266465939029,
+      "grad_norm": 0.2564024031162262,
+      "learning_rate": 4.992976157363734e-05,
+      "loss": 0.6363,
+      "step": 739
+    },
+    {
+      "epoch": 0.5570191945803538,
+      "grad_norm": 0.3209322988986969,
+      "learning_rate": 4.9929172883945694e-05,
+      "loss": 0.6138,
+      "step": 740
+    },
+    {
+      "epoch": 0.5577719232216786,
+      "grad_norm": 0.30397796630859375,
+      "learning_rate": 4.992858174143343e-05,
+      "loss": 0.6249,
+      "step": 741
+    },
+    {
+      "epoch": 0.5585246518630034,
+      "grad_norm": 0.26049843430519104,
+      "learning_rate": 4.992798814616519e-05,
+      "loss": 0.6679,
+      "step": 742
+    },
+    {
+      "epoch": 0.5592773805043282,
+      "grad_norm": 0.2424962967634201,
+      "learning_rate": 4.992739209820587e-05,
+      "loss": 0.5625,
+      "step": 743
+    },
+    {
+      "epoch": 0.560030109145653,
+      "grad_norm": 0.2983078062534332,
+      "learning_rate": 4.992679359762068e-05,
+      "loss": 0.5844,
+      "step": 744
+    },
+    {
+      "epoch": 0.5607828377869778,
+      "grad_norm": 0.23673667013645172,
+      "learning_rate": 4.992619264447506e-05,
+      "loss": 0.5469,
+      "step": 745
+    },
+    {
+      "epoch": 0.5615355664283026,
+      "grad_norm": 0.2536585032939911,
+      "learning_rate": 4.992558923883473e-05,
+      "loss": 0.5213,
+      "step": 746
+    },
+    {
+      "epoch": 0.5622882950696274,
+      "grad_norm": 0.23826384544372559,
+      "learning_rate": 4.992498338076567e-05,
+      "loss": 0.5326,
+      "step": 747
+    },
+    {
+      "epoch": 0.5630410237109522,
+      "grad_norm": 0.2310909777879715,
+      "learning_rate": 4.992437507033416e-05,
+      "loss": 0.5412,
+      "step": 748
+    },
+    {
+      "epoch": 0.563793752352277,
+      "grad_norm": 0.2436756044626236,
+      "learning_rate": 4.992376430760669e-05,
+      "loss": 0.5176,
+      "step": 749
+    },
+    {
+      "epoch": 0.5645464809936018,
+      "grad_norm": 0.26019564270973206,
+      "learning_rate": 4.992315109265007e-05,
+      "loss": 0.6613,
+      "step": 750
+    },
+    {
+      "epoch": 0.5652992096349266,
+      "grad_norm": 0.23265880346298218,
+      "learning_rate": 4.9922535425531366e-05,
+      "loss": 0.5766,
+      "step": 751
+    },
+    {
+      "epoch": 0.5660519382762514,
+      "grad_norm": 0.2525985836982727,
+      "learning_rate": 4.99219173063179e-05,
+      "loss": 0.6087,
+      "step": 752
+    },
+    {
+      "epoch": 0.5668046669175762,
+      "grad_norm": 0.30216917395591736,
+      "learning_rate": 4.9921296735077264e-05,
+      "loss": 0.6189,
+      "step": 753
+    },
+    {
+      "epoch": 0.5675573955589011,
+      "grad_norm": 0.29891300201416016,
+      "learning_rate": 4.992067371187733e-05,
+      "loss": 0.6192,
+      "step": 754
+    },
+    {
+      "epoch": 0.5683101242002259,
+      "grad_norm": 0.23613415658473969,
+      "learning_rate": 4.992004823678622e-05,
+      "loss": 0.498,
+      "step": 755
+    },
+    {
+      "epoch": 0.5690628528415507,
+      "grad_norm": 0.24794124066829681,
+      "learning_rate": 4.991942030987236e-05,
+      "loss": 0.6356,
+      "step": 756
+    },
+    {
+      "epoch": 0.5698155814828755,
+      "grad_norm": 0.20576821267604828,
+      "learning_rate": 4.9918789931204384e-05,
+      "loss": 0.6149,
+      "step": 757
+    },
+    {
+      "epoch": 0.5705683101242002,
+      "grad_norm": 0.2602978050708771,
+      "learning_rate": 4.991815710085126e-05,
+      "loss": 0.5742,
+      "step": 758
+    },
+    {
+      "epoch": 0.571321038765525,
+      "grad_norm": 0.29132944345474243,
+      "learning_rate": 4.9917521818882175e-05,
+      "loss": 0.5751,
+      "step": 759
+    },
+    {
+      "epoch": 0.5720737674068498,
+      "grad_norm": 0.26420921087265015,
+      "learning_rate": 4.99168840853666e-05,
+      "loss": 0.5091,
+      "step": 760
+    },
+    {
+      "epoch": 0.5728264960481746,
+      "grad_norm": 0.2652202546596527,
+      "learning_rate": 4.99162439003743e-05,
+      "loss": 0.6544,
+      "step": 761
+    },
+    {
+      "epoch": 0.5735792246894994,
+      "grad_norm": 0.3250247836112976,
+      "learning_rate": 4.9915601263975266e-05,
+      "loss": 0.5759,
+      "step": 762
+    },
+    {
+      "epoch": 0.5743319533308242,
+      "grad_norm": 0.2520033121109009,
+      "learning_rate": 4.991495617623977e-05,
+      "loss": 0.6625,
+      "step": 763
+    },
+    {
+      "epoch": 0.575084681972149,
+      "grad_norm": 0.28177329897880554,
+      "learning_rate": 4.9914308637238376e-05,
+      "loss": 0.6038,
+      "step": 764
+    },
+    {
+      "epoch": 0.5758374106134738,
+      "grad_norm": 0.27470695972442627,
+      "learning_rate": 4.991365864704188e-05,
+      "loss": 0.6542,
+      "step": 765
+    },
+    {
+      "epoch": 0.5765901392547986,
+      "grad_norm": 0.33808833360671997,
+      "learning_rate": 4.991300620572138e-05,
+      "loss": 0.5344,
+      "step": 766
+    },
+    {
+      "epoch": 0.5773428678961234,
+      "grad_norm": 0.24920520186424255,
+      "learning_rate": 4.991235131334822e-05,
+      "loss": 0.5125,
+      "step": 767
+    },
+    {
+      "epoch": 0.5780955965374482,
+      "grad_norm": 0.4014914929866791,
+      "learning_rate": 4.991169396999401e-05,
+      "loss": 0.5366,
+      "step": 768
+    },
+    {
+      "epoch": 0.578848325178773,
+      "grad_norm": 0.2772243618965149,
+      "learning_rate": 4.9911034175730656e-05,
+      "loss": 0.622,
+      "step": 769
+    },
+    {
+      "epoch": 0.5796010538200979,
+      "grad_norm": 0.28865575790405273,
+      "learning_rate": 4.9910371930630284e-05,
+      "loss": 0.6477,
+      "step": 770
+    },
+    {
+      "epoch": 0.5803537824614227,
+      "grad_norm": 0.31725746393203735,
+      "learning_rate": 4.990970723476533e-05,
+      "loss": 0.5982,
+      "step": 771
+    },
+    {
+      "epoch": 0.5811065111027475,
+      "grad_norm": 0.29398849606513977,
+      "learning_rate": 4.9909040088208495e-05,
+      "loss": 0.5759,
+      "step": 772
+    },
+    {
+      "epoch": 0.5818592397440723,
+      "grad_norm": 0.37383005023002625,
+      "learning_rate": 4.990837049103273e-05,
+      "loss": 0.5415,
+      "step": 773
+    },
+    {
+      "epoch": 0.5826119683853971,
+      "grad_norm": 0.3882606625556946,
+      "learning_rate": 4.990769844331125e-05,
+      "loss": 0.5473,
+      "step": 774
+    },
+    {
+      "epoch": 0.5833646970267219,
+      "grad_norm": 0.42454516887664795,
+      "learning_rate": 4.990702394511756e-05,
+      "loss": 0.5174,
+      "step": 775
+    },
+    {
+      "epoch": 0.5841174256680467,
+      "grad_norm": 0.2961101531982422,
+      "learning_rate": 4.990634699652542e-05,
+      "loss": 0.5517,
+      "step": 776
+    },
+    {
+      "epoch": 0.5848701543093715,
+      "grad_norm": 0.34278208017349243,
+      "learning_rate": 4.9905667597608865e-05,
+      "loss": 0.5839,
+      "step": 777
+    },
+    {
+      "epoch": 0.5856228829506963,
+      "grad_norm": 0.4956200420856476,
+      "learning_rate": 4.9904985748442185e-05,
+      "loss": 0.5961,
+      "step": 778
+    },
+    {
+      "epoch": 0.5863756115920211,
+      "grad_norm": 0.30803361535072327,
+      "learning_rate": 4.9904301449099944e-05,
+      "loss": 0.6112,
+      "step": 779
+    },
+    {
+      "epoch": 0.5871283402333459,
+      "grad_norm": 0.2918480634689331,
+      "learning_rate": 4.9903614699656986e-05,
+      "loss": 0.6047,
+      "step": 780
+    },
+    {
+      "epoch": 0.5878810688746707,
+      "grad_norm": 0.3440723717212677,
+      "learning_rate": 4.9902925500188407e-05,
+      "loss": 0.5308,
+      "step": 781
+    },
+    {
+      "epoch": 0.5886337975159955,
+      "grad_norm": 0.31706976890563965,
+      "learning_rate": 4.9902233850769584e-05,
+      "loss": 0.5933,
+      "step": 782
+    },
+    {
+      "epoch": 0.5893865261573202,
+      "grad_norm": 0.26432937383651733,
+      "learning_rate": 4.990153975147614e-05,
+      "loss": 0.6861,
+      "step": 783
+    },
+    {
+      "epoch": 0.590139254798645,
+      "grad_norm": 0.32479196786880493,
+      "learning_rate": 4.990084320238399e-05,
+      "loss": 0.6289,
+      "step": 784
+    },
+    {
+      "epoch": 0.5908919834399698,
+      "grad_norm": 0.34259679913520813,
+      "learning_rate": 4.9900144203569306e-05,
+      "loss": 0.5573,
+      "step": 785
+    },
+    {
+      "epoch": 0.5916447120812947,
+      "grad_norm": 0.3657112419605255,
+      "learning_rate": 4.989944275510853e-05,
+      "loss": 0.5773,
+      "step": 786
+    },
+    {
+      "epoch": 0.5923974407226195,
+      "grad_norm": 0.23250730335712433,
+      "learning_rate": 4.989873885707837e-05,
+      "loss": 0.6048,
+      "step": 787
+    },
+    {
+      "epoch": 0.5931501693639443,
+      "grad_norm": 0.3902212083339691,
+      "learning_rate": 4.98980325095558e-05,
+      "loss": 0.6407,
+      "step": 788
+    },
+    {
+      "epoch": 0.5939028980052691,
+      "grad_norm": 0.31707993149757385,
+      "learning_rate": 4.9897323712618085e-05,
+      "loss": 0.5125,
+      "step": 789
+    },
+    {
+      "epoch": 0.5946556266465939,
+      "grad_norm": 0.27391141653060913,
+      "learning_rate": 4.9896612466342716e-05,
+      "loss": 0.5705,
+      "step": 790
+    },
+    {
+      "epoch": 0.5954083552879187,
+      "grad_norm": 0.3777892291545868,
+      "learning_rate": 4.989589877080747e-05,
+      "loss": 0.6616,
+      "step": 791
+    },
+    {
+      "epoch": 0.5961610839292435,
+      "grad_norm": 0.27415400743484497,
+      "learning_rate": 4.989518262609042e-05,
+      "loss": 0.5558,
+      "step": 792
+    },
+    {
+      "epoch": 0.5969138125705683,
+      "grad_norm": 0.37164580821990967,
+      "learning_rate": 4.9894464032269864e-05,
+      "loss": 0.6003,
+      "step": 793
+    },
+    {
+      "epoch": 0.5976665412118931,
+      "grad_norm": 0.331026166677475,
+      "learning_rate": 4.989374298942439e-05,
+      "loss": 0.5521,
+      "step": 794
+    },
+    {
+      "epoch": 0.5984192698532179,
+      "grad_norm": 0.2725556194782257,
+      "learning_rate": 4.989301949763285e-05,
+      "loss": 0.5633,
+      "step": 795
+    },
+    {
+      "epoch": 0.5991719984945427,
+      "grad_norm": 0.37769103050231934,
+      "learning_rate": 4.9892293556974366e-05,
+      "loss": 0.6453,
+      "step": 796
+    },
+    {
+      "epoch": 0.5999247271358675,
+      "grad_norm": 0.3136664032936096,
+      "learning_rate": 4.9891565167528324e-05,
+      "loss": 0.4859,
+      "step": 797
+    },
+    {
+      "epoch": 0.6006774557771923,
+      "grad_norm": 0.36988160014152527,
+      "learning_rate": 4.989083432937439e-05,
+      "loss": 0.5692,
+      "step": 798
+    },
+    {
+      "epoch": 0.6014301844185171,
+      "grad_norm": 0.42174071073532104,
+      "learning_rate": 4.9890101042592464e-05,
+      "loss": 0.5847,
+      "step": 799
+    },
+    {
+      "epoch": 0.602182913059842,
+      "grad_norm": 0.380263090133667,
+      "learning_rate": 4.988936530726276e-05,
+      "loss": 0.5744,
+      "step": 800
+    },
+    {
+      "epoch": 0.602182913059842,
+      "eval_loss": 0.5559351444244385,
+      "eval_runtime": 452.9694,
+      "eval_samples_per_second": 21.253,
+      "eval_steps_per_second": 0.665,
+      "step": 800
+    },
+    {
+      "epoch": 0.6029356417011668,
+      "grad_norm": 0.3048160970211029,
+      "learning_rate": 4.988862712346573e-05,
+      "loss": 0.5085,
+      "step": 801
+    },
+    {
+      "epoch": 0.6036883703424916,
+      "grad_norm": 0.4458189904689789,
+      "learning_rate": 4.9887886491282086e-05,
+      "loss": 0.499,
+      "step": 802
+    },
+    {
+      "epoch": 0.6044410989838164,
+      "grad_norm": 0.3002050817012787,
+      "learning_rate": 4.988714341079284e-05,
+      "loss": 0.5657,
+      "step": 803
+    },
+    {
+      "epoch": 0.6051938276251412,
+      "grad_norm": 0.36826997995376587,
+      "learning_rate": 4.988639788207926e-05,
+      "loss": 0.5046,
+      "step": 804
+    },
+    {
+      "epoch": 0.605946556266466,
+      "grad_norm": 0.3873305916786194,
+      "learning_rate": 4.988564990522285e-05,
+      "loss": 0.4856,
+      "step": 805
+    },
+    {
+      "epoch": 0.6066992849077908,
+      "grad_norm": 0.38649117946624756,
+      "learning_rate": 4.988489948030543e-05,
+      "loss": 0.4526,
+      "step": 806
+    },
+    {
+      "epoch": 0.6074520135491156,
+      "grad_norm": 0.30856895446777344,
+      "learning_rate": 4.988414660740906e-05,
+      "loss": 0.6455,
+      "step": 807
+    },
+    {
+      "epoch": 0.6082047421904403,
+      "grad_norm": 0.3108121454715729,
+      "learning_rate": 4.988339128661607e-05,
+      "loss": 0.5731,
+      "step": 808
+    },
+    {
+      "epoch": 0.6089574708317651,
+      "grad_norm": 0.3016495108604431,
+      "learning_rate": 4.9882633518009056e-05,
+      "loss": 0.5611,
+      "step": 809
+    },
+    {
+      "epoch": 0.6097101994730899,
+      "grad_norm": 0.39752209186553955,
+      "learning_rate": 4.98818733016709e-05,
+      "loss": 0.6059,
+      "step": 810
+    },
+    {
+      "epoch": 0.6104629281144147,
+      "grad_norm": 0.2507995367050171,
+      "learning_rate": 4.988111063768472e-05,
+      "loss": 0.5689,
+      "step": 811
+    },
+    {
+      "epoch": 0.6112156567557395,
+      "grad_norm": 0.3176596164703369,
+      "learning_rate": 4.988034552613394e-05,
+      "loss": 0.5447,
+      "step": 812
+    },
+    {
+      "epoch": 0.6119683853970643,
+      "grad_norm": 0.21182480454444885,
+      "learning_rate": 4.987957796710222e-05,
+      "loss": 0.4901,
+      "step": 813
+    },
+    {
+      "epoch": 0.6127211140383891,
+      "grad_norm": 0.40150612592697144,
+      "learning_rate": 4.9878807960673496e-05,
+      "loss": 0.7709,
+      "step": 814
+    },
+    {
+      "epoch": 0.6134738426797139,
+      "grad_norm": 0.3013867437839508,
+      "learning_rate": 4.987803550693199e-05,
+      "loss": 0.601,
+      "step": 815
+    },
+    {
+      "epoch": 0.6142265713210388,
+      "grad_norm": 0.28132203221321106,
+      "learning_rate": 4.9877260605962154e-05,
+      "loss": 0.6153,
+      "step": 816
+    },
+    {
+      "epoch": 0.6149792999623636,
+      "grad_norm": 0.31224560737609863,
+      "learning_rate": 4.987648325784875e-05,
+      "loss": 0.6395,
+      "step": 817
+    },
+    {
+      "epoch": 0.6157320286036884,
+      "grad_norm": 0.25310251116752625,
+      "learning_rate": 4.987570346267677e-05,
+      "loss": 0.5478,
+      "step": 818
+    },
+    {
+      "epoch": 0.6164847572450132,
+      "grad_norm": 0.26151570677757263,
+      "learning_rate": 4.987492122053151e-05,
+      "loss": 0.5473,
+      "step": 819
+    },
+    {
+      "epoch": 0.617237485886338,
+      "grad_norm": 0.2458895891904831,
+      "learning_rate": 4.9874136531498506e-05,
+      "loss": 0.6581,
+      "step": 820
+    },
+    {
+      "epoch": 0.6179902145276628,
+      "grad_norm": 0.23778685927391052,
+      "learning_rate": 4.987334939566357e-05,
+      "loss": 0.6123,
+      "step": 821
+    },
+    {
+      "epoch": 0.6187429431689876,
+      "grad_norm": 0.24921543896198273,
+      "learning_rate": 4.9872559813112776e-05,
+      "loss": 0.6618,
+      "step": 822
+    },
+    {
+      "epoch": 0.6194956718103124,
+      "grad_norm": 0.2635401785373688,
+      "learning_rate": 4.987176778393248e-05,
+      "loss": 0.6677,
+      "step": 823
+    },
+    {
+      "epoch": 0.6202484004516372,
+      "grad_norm": 0.25828787684440613,
+      "learning_rate": 4.987097330820929e-05,
+      "loss": 0.503,
+      "step": 824
+    },
+    {
+      "epoch": 0.621001129092962,
+      "grad_norm": 0.3049829304218292,
+      "learning_rate": 4.9870176386030106e-05,
+      "loss": 0.599,
+      "step": 825
+    },
+    {
+      "epoch": 0.6217538577342868,
+      "grad_norm": 0.24836526811122894,
+      "learning_rate": 4.986937701748205e-05,
+      "loss": 0.5544,
+      "step": 826
+    },
+    {
+      "epoch": 0.6225065863756116,
+      "grad_norm": 0.3026723563671112,
+      "learning_rate": 4.986857520265257e-05,
+      "loss": 0.5562,
+      "step": 827
+    },
+    {
+      "epoch": 0.6232593150169364,
+      "grad_norm": 0.35048767924308777,
+      "learning_rate": 4.9867770941629324e-05,
+      "loss": 0.6941,
+      "step": 828
+    },
+    {
+      "epoch": 0.6240120436582612,
+      "grad_norm": 0.269937664270401,
+      "learning_rate": 4.986696423450028e-05,
+      "loss": 0.5443,
+      "step": 829
+    },
+    {
+      "epoch": 0.624764772299586,
+      "grad_norm": 0.36172038316726685,
+      "learning_rate": 4.9866155081353656e-05,
+      "loss": 0.6188,
+      "step": 830
+    },
+    {
+      "epoch": 0.6255175009409109,
+      "grad_norm": 0.29021888971328735,
+      "learning_rate": 4.9865343482277944e-05,
+      "loss": 0.6555,
+      "step": 831
+    },
+    {
+      "epoch": 0.6262702295822355,
+      "grad_norm": 0.23215846717357635,
+      "learning_rate": 4.986452943736189e-05,
+      "loss": 0.5643,
+      "step": 832
+    },
+    {
+      "epoch": 0.6270229582235604,
+      "grad_norm": 0.28057652711868286,
+      "learning_rate": 4.9863712946694527e-05,
+      "loss": 0.4709,
+      "step": 833
+    },
+    {
+      "epoch": 0.6277756868648852,
+      "grad_norm": 0.27879589796066284,
+      "learning_rate": 4.986289401036513e-05,
+      "loss": 0.5646,
+      "step": 834
+    },
+    {
+      "epoch": 0.62852841550621,
+      "grad_norm": 0.2556646764278412,
+      "learning_rate": 4.9862072628463266e-05,
+      "loss": 0.569,
+      "step": 835
+    },
+    {
+      "epoch": 0.6292811441475348,
+      "grad_norm": 0.2581430971622467,
+      "learning_rate": 4.986124880107877e-05,
+      "loss": 0.5902,
+      "step": 836
+    },
+    {
+      "epoch": 0.6300338727888596,
+      "grad_norm": 0.2789875268936157,
+      "learning_rate": 4.986042252830172e-05,
+      "loss": 0.6068,
+      "step": 837
+    },
+    {
+      "epoch": 0.6307866014301844,
+      "grad_norm": 0.249715656042099,
+      "learning_rate": 4.985959381022248e-05,
+      "loss": 0.6006,
+      "step": 838
+    },
+    {
+      "epoch": 0.6315393300715092,
+      "grad_norm": 0.31504085659980774,
+      "learning_rate": 4.985876264693168e-05,
+      "loss": 0.4296,
+      "step": 839
+    },
+    {
+      "epoch": 0.632292058712834,
+      "grad_norm": 0.36688122153282166,
+      "learning_rate": 4.9857929038520214e-05,
+      "loss": 0.4651,
+      "step": 840
+    },
+    {
+      "epoch": 0.6330447873541588,
+      "grad_norm": 0.237411767244339,
+      "learning_rate": 4.985709298507924e-05,
+      "loss": 0.5042,
+      "step": 841
+    },
+    {
+      "epoch": 0.6337975159954836,
+      "grad_norm": 0.2880602777004242,
+      "learning_rate": 4.985625448670019e-05,
+      "loss": 0.6696,
+      "step": 842
+    },
+    {
+      "epoch": 0.6345502446368084,
+      "grad_norm": 0.28828442096710205,
+      "learning_rate": 4.985541354347476e-05,
+      "loss": 0.6347,
+      "step": 843
+    },
+    {
+      "epoch": 0.6353029732781332,
+      "grad_norm": 0.22117064893245697,
+      "learning_rate": 4.985457015549492e-05,
+      "loss": 0.6163,
+      "step": 844
+    },
+    {
+      "epoch": 0.636055701919458,
+      "grad_norm": 0.3070262372493744,
+      "learning_rate": 4.98537243228529e-05,
+      "loss": 0.4938,
+      "step": 845
+    },
+    {
+      "epoch": 0.6368084305607828,
+      "grad_norm": 0.32178089022636414,
+      "learning_rate": 4.985287604564119e-05,
+      "loss": 0.5311,
+      "step": 846
+    },
+    {
+      "epoch": 0.6375611592021077,
+      "grad_norm": 0.26897361874580383,
+      "learning_rate": 4.985202532395257e-05,
+      "loss": 0.5241,
+      "step": 847
+    },
+    {
+      "epoch": 0.6383138878434325,
+      "grad_norm": 0.39192524552345276,
+      "learning_rate": 4.9851172157880055e-05,
+      "loss": 0.5682,
+      "step": 848
+    },
+    {
+      "epoch": 0.6390666164847573,
+      "grad_norm": 0.40348803997039795,
+      "learning_rate": 4.985031654751697e-05,
+      "loss": 0.5149,
+      "step": 849
+    },
+    {
+      "epoch": 0.6398193451260821,
+      "grad_norm": 0.25793150067329407,
+      "learning_rate": 4.984945849295686e-05,
+      "loss": 0.4665,
+      "step": 850
+    },
+    {
+      "epoch": 0.6405720737674069,
+      "grad_norm": 0.31535837054252625,
+      "learning_rate": 4.9848597994293577e-05,
+      "loss": 0.6089,
+      "step": 851
+    },
+    {
+      "epoch": 0.6413248024087317,
+      "grad_norm": 0.2916658818721771,
+      "learning_rate": 4.984773505162122e-05,
+      "loss": 0.5373,
+      "step": 852
+    },
+    {
+      "epoch": 0.6420775310500565,
+      "grad_norm": 0.25033658742904663,
+      "learning_rate": 4.984686966503415e-05,
+      "loss": 0.5879,
+      "step": 853
+    },
+    {
+      "epoch": 0.6428302596913813,
+      "grad_norm": 0.24179618060588837,
+      "learning_rate": 4.984600183462702e-05,
+      "loss": 0.6944,
+      "step": 854
+    },
+    {
+      "epoch": 0.6435829883327061,
+      "grad_norm": 0.23500098288059235,
+      "learning_rate": 4.984513156049471e-05,
+      "loss": 0.5774,
+      "step": 855
+    },
+    {
+      "epoch": 0.6443357169740309,
+      "grad_norm": 0.23763346672058105,
+      "learning_rate": 4.984425884273243e-05,
+      "loss": 0.4638,
+      "step": 856
+    },
+    {
+      "epoch": 0.6450884456153556,
+      "grad_norm": 0.23448865115642548,
+      "learning_rate": 4.984338368143558e-05,
+      "loss": 0.5603,
+      "step": 857
+    },
+    {
+      "epoch": 0.6458411742566804,
+      "grad_norm": 0.23594379425048828,
+      "learning_rate": 4.984250607669989e-05,
+      "loss": 0.5811,
+      "step": 858
+    },
+    {
+      "epoch": 0.6465939028980052,
+      "grad_norm": 0.2194521129131317,
+      "learning_rate": 4.984162602862132e-05,
+      "loss": 0.6168,
+      "step": 859
+    },
+    {
+      "epoch": 0.64734663153933,
+      "grad_norm": 0.23793937265872955,
+      "learning_rate": 4.984074353729613e-05,
+      "loss": 0.5512,
+      "step": 860
+    },
+    {
+      "epoch": 0.6480993601806548,
+      "grad_norm": 0.23750007152557373,
+      "learning_rate": 4.983985860282081e-05,
+      "loss": 0.6245,
+      "step": 861
+    },
+    {
+      "epoch": 0.6488520888219796,
+      "grad_norm": 0.24104493856430054,
+      "learning_rate": 4.983897122529214e-05,
+      "loss": 0.6169,
+      "step": 862
+    },
+    {
+      "epoch": 0.6496048174633045,
+      "grad_norm": 0.2066459059715271,
+      "learning_rate": 4.9838081404807164e-05,
+      "loss": 0.6129,
+      "step": 863
+    },
+    {
+      "epoch": 0.6503575461046293,
+      "grad_norm": 0.22828885912895203,
+      "learning_rate": 4.9837189141463184e-05,
+      "loss": 0.6146,
+      "step": 864
+    },
+    {
+      "epoch": 0.6511102747459541,
+      "grad_norm": 0.22520457208156586,
+      "learning_rate": 4.98362944353578e-05,
+      "loss": 0.4632,
+      "step": 865
+    },
+    {
+      "epoch": 0.6518630033872789,
+      "grad_norm": 0.20578528940677643,
+      "learning_rate": 4.983539728658882e-05,
+      "loss": 0.6034,
+      "step": 866
+    },
+    {
+      "epoch": 0.6526157320286037,
+      "grad_norm": 0.22334690392017365,
+      "learning_rate": 4.9834497695254384e-05,
+      "loss": 0.5899,
+      "step": 867
+    },
+    {
+      "epoch": 0.6533684606699285,
+      "grad_norm": 0.2305312603712082,
+      "learning_rate": 4.983359566145285e-05,
+      "loss": 0.5866,
+      "step": 868
+    },
+    {
+      "epoch": 0.6541211893112533,
+      "grad_norm": 0.20383617281913757,
+      "learning_rate": 4.9832691185282884e-05,
+      "loss": 0.5466,
+      "step": 869
+    },
+    {
+      "epoch": 0.6548739179525781,
+      "grad_norm": 0.21882759034633636,
+      "learning_rate": 4.983178426684339e-05,
+      "loss": 0.5967,
+      "step": 870
+    },
+    {
+      "epoch": 0.6556266465939029,
+      "grad_norm": 0.20372353494167328,
+      "learning_rate": 4.983087490623352e-05,
+      "loss": 0.5676,
+      "step": 871
+    },
+    {
+      "epoch": 0.6563793752352277,
+      "grad_norm": 0.2478175163269043,
+      "learning_rate": 4.982996310355276e-05,
+      "loss": 0.5644,
+      "step": 872
+    },
+    {
+      "epoch": 0.6571321038765525,
+      "grad_norm": 0.22109344601631165,
+      "learning_rate": 4.982904885890081e-05,
+      "loss": 0.6154,
+      "step": 873
+    },
+    {
+      "epoch": 0.6578848325178773,
+      "grad_norm": 0.21225804090499878,
+      "learning_rate": 4.982813217237765e-05,
+      "loss": 0.6068,
+      "step": 874
+    },
+    {
+      "epoch": 0.6586375611592021,
+      "grad_norm": 0.23010121285915375,
+      "learning_rate": 4.9827213044083516e-05,
+      "loss": 0.5052,
+      "step": 875
+    },
+    {
+      "epoch": 0.6593902898005269,
+      "grad_norm": 0.3452666103839874,
+      "learning_rate": 4.9826291474118936e-05,
+      "loss": 0.4947,
+      "step": 876
+    },
+    {
+      "epoch": 0.6601430184418517,
+      "grad_norm": 0.45795926451683044,
+      "learning_rate": 4.9825367462584685e-05,
+      "loss": 0.4802,
+      "step": 877
+    },
+    {
+      "epoch": 0.6608957470831766,
+      "grad_norm": 0.518143355846405,
+      "learning_rate": 4.982444100958182e-05,
+      "loss": 0.5811,
+      "step": 878
+    },
+    {
+      "epoch": 0.6616484757245014,
+      "grad_norm": 0.4170481562614441,
+      "learning_rate": 4.9823512115211654e-05,
+      "loss": 0.5114,
+      "step": 879
+    },
+    {
+      "epoch": 0.6624012043658262,
+      "grad_norm": 0.42540156841278076,
+      "learning_rate": 4.9822580779575757e-05,
+      "loss": 0.561,
+      "step": 880
+    },
+    {
+      "epoch": 0.663153933007151,
+      "grad_norm": 0.35975801944732666,
+      "learning_rate": 4.9821647002775984e-05,
+      "loss": 0.5736,
+      "step": 881
+    },
+    {
+      "epoch": 0.6639066616484757,
+      "grad_norm": 0.40613359212875366,
+      "learning_rate": 4.982071078491447e-05,
+      "loss": 0.5484,
+      "step": 882
+    },
+    {
+      "epoch": 0.6646593902898005,
+      "grad_norm": 0.31350812315940857,
+      "learning_rate": 4.9819772126093563e-05,
+      "loss": 0.5695,
+      "step": 883
+    },
+    {
+      "epoch": 0.6654121189311253,
+      "grad_norm": 0.4069611728191376,
+      "learning_rate": 4.981883102641595e-05,
+      "loss": 0.4612,
+      "step": 884
+    },
+    {
+      "epoch": 0.6661648475724501,
+      "grad_norm": 0.48988521099090576,
+      "learning_rate": 4.9817887485984524e-05,
+      "loss": 0.5606,
+      "step": 885
+    },
+    {
+      "epoch": 0.6669175762137749,
+      "grad_norm": 0.29497969150543213,
+      "learning_rate": 4.9816941504902475e-05,
+      "loss": 0.6759,
+      "step": 886
+    },
+    {
+      "epoch": 0.6676703048550997,
+      "grad_norm": 0.40058502554893494,
+      "learning_rate": 4.9815993083273266e-05,
+      "loss": 0.4526,
+      "step": 887
+    },
+    {
+      "epoch": 0.6684230334964245,
+      "grad_norm": 0.3285306692123413,
+      "learning_rate": 4.98150422212006e-05,
+      "loss": 0.5668,
+      "step": 888
+    },
+    {
+      "epoch": 0.6691757621377493,
+      "grad_norm": 0.3429281413555145,
+      "learning_rate": 4.9814088918788464e-05,
+      "loss": 0.5272,
+      "step": 889
+    },
+    {
+      "epoch": 0.6699284907790741,
+      "grad_norm": 0.3809550702571869,
+      "learning_rate": 4.9813133176141114e-05,
+      "loss": 0.5126,
+      "step": 890
+    },
+    {
+      "epoch": 0.6706812194203989,
+      "grad_norm": 0.27859362959861755,
+      "learning_rate": 4.981217499336307e-05,
+      "loss": 0.5991,
+      "step": 891
+    },
+    {
+      "epoch": 0.6714339480617237,
+      "grad_norm": 0.34026578068733215,
+      "learning_rate": 4.981121437055911e-05,
+      "loss": 0.4553,
+      "step": 892
+    },
+    {
+      "epoch": 0.6721866767030485,
+      "grad_norm": 0.3274897038936615,
+      "learning_rate": 4.9810251307834296e-05,
+      "loss": 0.5523,
+      "step": 893
+    },
+    {
+      "epoch": 0.6729394053443734,
+      "grad_norm": 0.23340466618537903,
+      "learning_rate": 4.9809285805293945e-05,
+      "loss": 0.6083,
+      "step": 894
+    },
+    {
+      "epoch": 0.6736921339856982,
+      "grad_norm": 0.3063945770263672,
+      "learning_rate": 4.980831786304363e-05,
+      "loss": 0.5355,
+      "step": 895
+    },
+    {
+      "epoch": 0.674444862627023,
+      "grad_norm": 0.2833549976348877,
+      "learning_rate": 4.980734748118922e-05,
+      "loss": 0.5071,
+      "step": 896
+    },
+    {
+      "epoch": 0.6751975912683478,
+      "grad_norm": 0.24599292874336243,
+      "learning_rate": 4.980637465983682e-05,
+      "loss": 0.5824,
+      "step": 897
+    },
+    {
+      "epoch": 0.6759503199096726,
+      "grad_norm": 0.29285815358161926,
+      "learning_rate": 4.9805399399092834e-05,
+      "loss": 0.5779,
+      "step": 898
+    },
+    {
+      "epoch": 0.6767030485509974,
+      "grad_norm": 0.22775061428546906,
+      "learning_rate": 4.9804421699063906e-05,
+      "loss": 0.5907,
+      "step": 899
+    },
+    {
+      "epoch": 0.6774557771923222,
+      "grad_norm": 0.2828820049762726,
+      "learning_rate": 4.980344155985695e-05,
+      "loss": 0.4947,
+      "step": 900
+    },
+    {
+      "epoch": 0.678208505833647,
+      "grad_norm": 0.24243590235710144,
+      "learning_rate": 4.980245898157915e-05,
+      "loss": 0.6168,
+      "step": 901
+    },
+    {
+      "epoch": 0.6789612344749718,
+      "grad_norm": 0.23842909932136536,
+      "learning_rate": 4.980147396433798e-05,
+      "loss": 0.5883,
+      "step": 902
+    },
+    {
+      "epoch": 0.6797139631162966,
+      "grad_norm": 0.2535601258277893,
+      "learning_rate": 4.980048650824115e-05,
+      "loss": 0.5886,
+      "step": 903
+    },
+    {
+      "epoch": 0.6804666917576214,
+      "grad_norm": 0.26751720905303955,
+      "learning_rate": 4.979949661339663e-05,
+      "loss": 0.562,
+      "step": 904
+    },
+    {
+      "epoch": 0.6812194203989462,
+      "grad_norm": 0.2696024477481842,
+      "learning_rate": 4.979850427991269e-05,
+      "loss": 0.5584,
+      "step": 905
+    },
+    {
+      "epoch": 0.681972149040271,
+      "grad_norm": 0.2487579882144928,
+      "learning_rate": 4.979750950789785e-05,
+      "loss": 0.5276,
+      "step": 906
+    },
+    {
+      "epoch": 0.6827248776815957,
+      "grad_norm": 0.32148414850234985,
+      "learning_rate": 4.97965122974609e-05,
+      "loss": 0.5505,
+      "step": 907
+    },
+    {
+      "epoch": 0.6834776063229205,
+      "grad_norm": 0.4224289655685425,
+      "learning_rate": 4.9795512648710876e-05,
+      "loss": 0.5474,
+      "step": 908
+    },
+    {
+      "epoch": 0.6842303349642453,
+      "grad_norm": 0.46435248851776123,
+      "learning_rate": 4.9794510561757116e-05,
+      "loss": 0.4998,
+      "step": 909
+    },
+    {
+      "epoch": 0.6849830636055702,
+      "grad_norm": 0.25489723682403564,
+      "learning_rate": 4.97935060367092e-05,
+      "loss": 0.537,
+      "step": 910
+    },
+    {
+      "epoch": 0.685735792246895,
+      "grad_norm": 0.2830134928226471,
+      "learning_rate": 4.9792499073676984e-05,
+      "loss": 0.5259,
+      "step": 911
+    },
+    {
+      "epoch": 0.6864885208882198,
+      "grad_norm": 0.38646063208580017,
+      "learning_rate": 4.9791489672770585e-05,
+      "loss": 0.5595,
+      "step": 912
+    },
+    {
+      "epoch": 0.6872412495295446,
+      "grad_norm": 0.2797878682613373,
+      "learning_rate": 4.979047783410039e-05,
+      "loss": 0.5882,
+      "step": 913
+    },
+    {
+      "epoch": 0.6879939781708694,
+      "grad_norm": 0.2551473081111908,
+      "learning_rate": 4.9789463557777044e-05,
+      "loss": 0.5692,
+      "step": 914
+    },
+    {
+      "epoch": 0.6887467068121942,
+      "grad_norm": 0.25795918703079224,
+      "learning_rate": 4.978844684391149e-05,
+      "loss": 0.6149,
+      "step": 915
+    },
+    {
+      "epoch": 0.689499435453519,
+      "grad_norm": 0.2703177034854889,
+      "learning_rate": 4.978742769261489e-05,
+      "loss": 0.5435,
+      "step": 916
+    },
+    {
+      "epoch": 0.6902521640948438,
+      "grad_norm": 0.2773463726043701,
+      "learning_rate": 4.97864061039987e-05,
+      "loss": 0.5596,
+      "step": 917
+    },
+    {
+      "epoch": 0.6910048927361686,
+      "grad_norm": 0.2579146921634674,
+      "learning_rate": 4.978538207817465e-05,
+      "loss": 0.5876,
+      "step": 918
+    },
+    {
+      "epoch": 0.6917576213774934,
+      "grad_norm": 0.2293855994939804,
+      "learning_rate": 4.978435561525473e-05,
+      "loss": 0.4764,
+      "step": 919
+    },
+    {
+      "epoch": 0.6925103500188182,
+      "grad_norm": 0.31076544523239136,
+      "learning_rate": 4.978332671535117e-05,
+      "loss": 0.5414,
+      "step": 920
+    },
+    {
+      "epoch": 0.693263078660143,
+      "grad_norm": 0.31082287430763245,
+      "learning_rate": 4.978229537857652e-05,
+      "loss": 0.6013,
+      "step": 921
+    },
+    {
+      "epoch": 0.6940158073014678,
+      "grad_norm": 0.21954171359539032,
+      "learning_rate": 4.978126160504353e-05,
+      "loss": 0.6073,
+      "step": 922
+    },
+    {
+      "epoch": 0.6947685359427926,
+      "grad_norm": 0.34903034567832947,
+      "learning_rate": 4.978022539486529e-05,
+      "loss": 0.5952,
+      "step": 923
+    },
+    {
+      "epoch": 0.6955212645841174,
+      "grad_norm": 0.21976274251937866,
+      "learning_rate": 4.977918674815508e-05,
+      "loss": 0.5989,
+      "step": 924
+    },
+    {
+      "epoch": 0.6962739932254423,
+      "grad_norm": 0.28309038281440735,
+      "learning_rate": 4.977814566502651e-05,
+      "loss": 0.5408,
+      "step": 925
+    },
+    {
+      "epoch": 0.6970267218667671,
+      "grad_norm": 0.29695427417755127,
+      "learning_rate": 4.977710214559344e-05,
+      "loss": 0.5549,
+      "step": 926
+    },
+    {
+      "epoch": 0.6977794505080919,
+      "grad_norm": 0.32648470997810364,
+      "learning_rate": 4.9776056189969946e-05,
+      "loss": 0.4226,
+      "step": 927
+    },
+    {
+      "epoch": 0.6985321791494167,
+      "grad_norm": 0.28222760558128357,
+      "learning_rate": 4.9775007798270456e-05,
+      "loss": 0.5198,
+      "step": 928
+    },
+    {
+      "epoch": 0.6992849077907415,
+      "grad_norm": 0.35585933923721313,
+      "learning_rate": 4.97739569706096e-05,
+      "loss": 0.5087,
+      "step": 929
+    },
+    {
+      "epoch": 0.7000376364320663,
+      "grad_norm": 0.26782849431037903,
+      "learning_rate": 4.97729037071023e-05,
+      "loss": 0.5139,
+      "step": 930
+    },
+    {
+      "epoch": 0.7007903650733911,
+      "grad_norm": 0.3336419463157654,
+      "learning_rate": 4.977184800786373e-05,
+      "loss": 0.5865,
+      "step": 931
+    },
+    {
+      "epoch": 0.7015430937147158,
+      "grad_norm": 0.3032510578632355,
+      "learning_rate": 4.977078987300935e-05,
+      "loss": 0.5762,
+      "step": 932
+    },
+    {
+      "epoch": 0.7022958223560406,
+      "grad_norm": 0.2849379777908325,
+      "learning_rate": 4.976972930265488e-05,
+      "loss": 0.6358,
+      "step": 933
+    },
+    {
+      "epoch": 0.7030485509973654,
+      "grad_norm": 0.3834438920021057,
+      "learning_rate": 4.976866629691629e-05,
+      "loss": 0.6413,
+      "step": 934
+    },
+    {
+      "epoch": 0.7038012796386902,
+      "grad_norm": 0.27010586857795715,
+      "learning_rate": 4.9767600855909826e-05,
+      "loss": 0.5653,
+      "step": 935
+    },
+    {
+      "epoch": 0.704554008280015,
+      "grad_norm": 0.29109370708465576,
+      "learning_rate": 4.9766532979752025e-05,
+      "loss": 0.5788,
+      "step": 936
+    },
+    {
+      "epoch": 0.7053067369213398,
+      "grad_norm": 0.28314492106437683,
+      "learning_rate": 4.976546266855965e-05,
+      "loss": 0.5967,
+      "step": 937
+    },
+    {
+      "epoch": 0.7060594655626646,
+      "grad_norm": 0.3314724564552307,
+      "learning_rate": 4.9764389922449756e-05,
+      "loss": 0.5715,
+      "step": 938
+    },
+    {
+      "epoch": 0.7068121942039894,
+      "grad_norm": 0.26889464259147644,
+      "learning_rate": 4.976331474153965e-05,
+      "loss": 0.5828,
+      "step": 939
+    },
+    {
+      "epoch": 0.7075649228453142,
+      "grad_norm": 0.30621853470802307,
+      "learning_rate": 4.976223712594691e-05,
+      "loss": 0.5072,
+      "step": 940
+    },
+    {
+      "epoch": 0.708317651486639,
+      "grad_norm": 0.32690006494522095,
+      "learning_rate": 4.97611570757894e-05,
+      "loss": 0.5823,
+      "step": 941
+    },
+    {
+      "epoch": 0.7090703801279639,
+      "grad_norm": 0.23807530105113983,
+      "learning_rate": 4.9760074591185216e-05,
+      "loss": 0.6298,
+      "step": 942
+    },
+    {
+      "epoch": 0.7098231087692887,
+      "grad_norm": 0.26979222893714905,
+      "learning_rate": 4.975898967225274e-05,
+      "loss": 0.6504,
+      "step": 943
+    },
+    {
+      "epoch": 0.7105758374106135,
+      "grad_norm": 0.2578220069408417,
+      "learning_rate": 4.975790231911062e-05,
+      "loss": 0.5724,
+      "step": 944
+    },
+    {
+      "epoch": 0.7113285660519383,
+      "grad_norm": 0.2688755691051483,
+      "learning_rate": 4.9756812531877775e-05,
+      "loss": 0.5549,
+      "step": 945
+    },
+    {
+      "epoch": 0.7120812946932631,
+      "grad_norm": 0.3879950940608978,
+      "learning_rate": 4.9755720310673365e-05,
+      "loss": 0.5165,
+      "step": 946
+    },
+    {
+      "epoch": 0.7128340233345879,
+      "grad_norm": 0.3370249271392822,
+      "learning_rate": 4.9754625655616845e-05,
+      "loss": 0.4587,
+      "step": 947
+    },
+    {
+      "epoch": 0.7135867519759127,
+      "grad_norm": 0.20515058934688568,
+      "learning_rate": 4.975352856682792e-05,
+      "loss": 0.6057,
+      "step": 948
+    },
+    {
+      "epoch": 0.7143394806172375,
+      "grad_norm": 0.2762977182865143,
+      "learning_rate": 4.9752429044426574e-05,
+      "loss": 0.5841,
+      "step": 949
+    },
+    {
+      "epoch": 0.7150922092585623,
+      "grad_norm": 0.2584463953971863,
+      "learning_rate": 4.975132708853304e-05,
+      "loss": 0.5602,
+      "step": 950
+    },
+    {
+      "epoch": 0.7158449378998871,
+      "grad_norm": 0.252142071723938,
+      "learning_rate": 4.9750222699267836e-05,
+      "loss": 0.5896,
+      "step": 951
+    },
+    {
+      "epoch": 0.7165976665412119,
+      "grad_norm": 0.32410532236099243,
+      "learning_rate": 4.9749115876751724e-05,
+      "loss": 0.5341,
+      "step": 952
+    },
+    {
+      "epoch": 0.7173503951825367,
+      "grad_norm": 0.294111430644989,
+      "learning_rate": 4.9748006621105755e-05,
+      "loss": 0.4804,
+      "step": 953
+    },
+    {
+      "epoch": 0.7181031238238615,
+      "grad_norm": 0.19691382348537445,
+      "learning_rate": 4.9746894932451226e-05,
+      "loss": 0.4578,
+      "step": 954
+    },
+    {
+      "epoch": 0.7188558524651864,
+      "grad_norm": 0.27295902371406555,
+      "learning_rate": 4.974578081090972e-05,
+      "loss": 0.4907,
+      "step": 955
+    },
+    {
+      "epoch": 0.7196085811065112,
+      "grad_norm": 0.25252723693847656,
+      "learning_rate": 4.974466425660307e-05,
+      "loss": 0.511,
+      "step": 956
+    },
+    {
+      "epoch": 0.7203613097478359,
+      "grad_norm": 0.2193632572889328,
+      "learning_rate": 4.974354526965337e-05,
+      "loss": 0.5185,
+      "step": 957
+    },
+    {
+      "epoch": 0.7211140383891607,
+      "grad_norm": 0.27126839756965637,
+      "learning_rate": 4.9742423850183025e-05,
+      "loss": 0.4424,
+      "step": 958
+    },
+    {
+      "epoch": 0.7218667670304855,
+      "grad_norm": 0.2911543846130371,
+      "learning_rate": 4.9741299998314626e-05,
+      "loss": 0.5109,
+      "step": 959
+    },
+    {
+      "epoch": 0.7226194956718103,
+      "grad_norm": 0.23142144083976746,
+      "learning_rate": 4.9740173714171115e-05,
+      "loss": 0.5614,
+      "step": 960
+    },
+    {
+      "epoch": 0.7233722243131351,
+      "grad_norm": 0.411619633436203,
+      "learning_rate": 4.973904499787564e-05,
+      "loss": 0.542,
+      "step": 961
+    },
+    {
+      "epoch": 0.7241249529544599,
+      "grad_norm": 0.43144890666007996,
+      "learning_rate": 4.973791384955163e-05,
+      "loss": 0.405,
+      "step": 962
+    },
+    {
+      "epoch": 0.7248776815957847,
+      "grad_norm": 0.33030205965042114,
+      "learning_rate": 4.97367802693228e-05,
+      "loss": 0.5107,
+      "step": 963
+    },
+    {
+      "epoch": 0.7256304102371095,
+      "grad_norm": 0.7295933365821838,
+      "learning_rate": 4.973564425731312e-05,
+      "loss": 0.6161,
+      "step": 964
+    },
+    {
+      "epoch": 0.7263831388784343,
+      "grad_norm": 0.3703036904335022,
+      "learning_rate": 4.97345058136468e-05,
+      "loss": 0.4888,
+      "step": 965
+    },
+    {
+      "epoch": 0.7271358675197591,
+      "grad_norm": 0.3630495071411133,
+      "learning_rate": 4.973336493844837e-05,
+      "loss": 0.5146,
+      "step": 966
+    },
+    {
+      "epoch": 0.7278885961610839,
+      "grad_norm": 0.2896798551082611,
+      "learning_rate": 4.9732221631842576e-05,
+      "loss": 0.6421,
+      "step": 967
+    },
+    {
+      "epoch": 0.7286413248024087,
+      "grad_norm": 0.38964855670928955,
+      "learning_rate": 4.973107589395444e-05,
+      "loss": 0.5452,
+      "step": 968
+    },
+    {
+      "epoch": 0.7293940534437335,
+      "grad_norm": 0.3577474355697632,
+      "learning_rate": 4.972992772490927e-05,
+      "loss": 0.5643,
+      "step": 969
+    },
+    {
+      "epoch": 0.7301467820850583,
+      "grad_norm": 0.29420387744903564,
+      "learning_rate": 4.9728777124832634e-05,
+      "loss": 0.5525,
+      "step": 970
+    },
+    {
+      "epoch": 0.7308995107263832,
+      "grad_norm": 0.2968292832374573,
+      "learning_rate": 4.972762409385034e-05,
+      "loss": 0.5162,
+      "step": 971
+    },
+    {
+      "epoch": 0.731652239367708,
+      "grad_norm": 0.4148855209350586,
+      "learning_rate": 4.972646863208851e-05,
+      "loss": 0.4895,
+      "step": 972
+    },
+    {
+      "epoch": 0.7324049680090328,
+      "grad_norm": 0.3184104859828949,
+      "learning_rate": 4.972531073967347e-05,
+      "loss": 0.4872,
+      "step": 973
+    },
+    {
+      "epoch": 0.7331576966503576,
+      "grad_norm": 0.3124586343765259,
+      "learning_rate": 4.972415041673187e-05,
+      "loss": 0.4634,
+      "step": 974
+    },
+    {
+      "epoch": 0.7339104252916824,
+      "grad_norm": 0.3398424983024597,
+      "learning_rate": 4.97229876633906e-05,
+      "loss": 0.6005,
+      "step": 975
+    },
+    {
+      "epoch": 0.7346631539330072,
+      "grad_norm": 0.2630004286766052,
+      "learning_rate": 4.972182247977681e-05,
+      "loss": 0.5551,
+      "step": 976
+    },
+    {
+      "epoch": 0.735415882574332,
+      "grad_norm": 0.32554134726524353,
+      "learning_rate": 4.972065486601792e-05,
+      "loss": 0.6351,
+      "step": 977
+    },
+    {
+      "epoch": 0.7361686112156568,
+      "grad_norm": 0.27939605712890625,
+      "learning_rate": 4.971948482224162e-05,
+      "loss": 0.5528,
+      "step": 978
+    },
+    {
+      "epoch": 0.7369213398569816,
+      "grad_norm": 0.3210045099258423,
+      "learning_rate": 4.971831234857587e-05,
+      "loss": 0.5482,
+      "step": 979
+    },
+    {
+      "epoch": 0.7376740684983064,
+      "grad_norm": 0.2987395226955414,
+      "learning_rate": 4.971713744514888e-05,
+      "loss": 0.582,
+      "step": 980
+    },
+    {
+      "epoch": 0.7384267971396311,
+      "grad_norm": 0.3088783323764801,
+      "learning_rate": 4.9715960112089144e-05,
+      "loss": 0.658,
+      "step": 981
+    },
+    {
+      "epoch": 0.7391795257809559,
+      "grad_norm": 0.29171115159988403,
+      "learning_rate": 4.9714780349525414e-05,
+      "loss": 0.4633,
+      "step": 982
+    },
+    {
+      "epoch": 0.7399322544222807,
+      "grad_norm": 0.30490362644195557,
+      "learning_rate": 4.9713598157586704e-05,
+      "loss": 0.572,
+      "step": 983
+    },
+    {
+      "epoch": 0.7406849830636055,
+      "grad_norm": 0.24807530641555786,
+      "learning_rate": 4.97124135364023e-05,
+      "loss": 0.4952,
+      "step": 984
+    },
+    {
+      "epoch": 0.7414377117049303,
+      "grad_norm": 0.2491171807050705,
+      "learning_rate": 4.971122648610175e-05,
+      "loss": 0.6175,
+      "step": 985
+    },
+    {
+      "epoch": 0.7421904403462551,
+      "grad_norm": 0.24763499200344086,
+      "learning_rate": 4.971003700681485e-05,
+      "loss": 0.4876,
+      "step": 986
+    },
+    {
+      "epoch": 0.74294316898758,
+      "grad_norm": 0.36980950832366943,
+      "learning_rate": 4.9708845098671706e-05,
+      "loss": 0.5414,
+      "step": 987
+    },
+    {
+      "epoch": 0.7436958976289048,
+      "grad_norm": 0.3768077492713928,
+      "learning_rate": 4.970765076180266e-05,
+      "loss": 0.5248,
+      "step": 988
+    },
+    {
+      "epoch": 0.7444486262702296,
+      "grad_norm": 0.29031676054000854,
+      "learning_rate": 4.970645399633831e-05,
+      "loss": 0.4588,
+      "step": 989
+    },
+    {
+      "epoch": 0.7452013549115544,
+      "grad_norm": 0.3854856491088867,
+      "learning_rate": 4.970525480240953e-05,
+      "loss": 0.6721,
+      "step": 990
+    },
+    {
+      "epoch": 0.7459540835528792,
+      "grad_norm": 0.31408604979515076,
+      "learning_rate": 4.970405318014748e-05,
+      "loss": 0.5911,
+      "step": 991
+    },
+    {
+      "epoch": 0.746706812194204,
+      "grad_norm": 0.29321709275245667,
+      "learning_rate": 4.9702849129683545e-05,
+      "loss": 0.5496,
+      "step": 992
+    },
+    {
+      "epoch": 0.7474595408355288,
+      "grad_norm": 0.2680532932281494,
+      "learning_rate": 4.9701642651149416e-05,
+      "loss": 0.4715,
+      "step": 993
+    },
+    {
+      "epoch": 0.7482122694768536,
+      "grad_norm": 0.33550432324409485,
+      "learning_rate": 4.970043374467703e-05,
+      "loss": 0.6389,
+      "step": 994
+    },
+    {
+      "epoch": 0.7489649981181784,
+      "grad_norm": 0.28727245330810547,
+      "learning_rate": 4.969922241039859e-05,
+      "loss": 0.5492,
+      "step": 995
+    },
+    {
+      "epoch": 0.7497177267595032,
+      "grad_norm": 0.24397510290145874,
+      "learning_rate": 4.969800864844656e-05,
+      "loss": 0.5524,
+      "step": 996
+    },
+    {
+      "epoch": 0.750470455400828,
+      "grad_norm": 0.3295447826385498,
+      "learning_rate": 4.9696792458953676e-05,
+      "loss": 0.5377,
+      "step": 997
+    },
+    {
+      "epoch": 0.7512231840421528,
+      "grad_norm": 0.27701735496520996,
+      "learning_rate": 4.969557384205295e-05,
+      "loss": 0.473,
+      "step": 998
+    },
+    {
+      "epoch": 0.7519759126834776,
+      "grad_norm": 0.36793816089630127,
+      "learning_rate": 4.969435279787763e-05,
+      "loss": 0.5054,
+      "step": 999
+    },
+    {
+      "epoch": 0.7527286413248024,
+      "grad_norm": 0.38624241948127747,
+      "learning_rate": 4.9693129326561254e-05,
+      "loss": 0.5081,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7527286413248024,
+      "eval_loss": 0.5175215005874634,
+      "eval_runtime": 452.6483,
+      "eval_samples_per_second": 21.268,
+      "eval_steps_per_second": 0.665,
+      "step": 1000
+    },
+    {
+      "epoch": 0.7534813699661272,
+      "grad_norm": 0.2883687913417816,
+      "learning_rate": 4.9691903428237624e-05,
+      "loss": 0.5725,
+      "step": 1001
+    },
+    {
+      "epoch": 0.754234098607452,
+      "grad_norm": 0.5059789419174194,
+      "learning_rate": 4.969067510304081e-05,
+      "loss": 0.4546,
+      "step": 1002
+    },
+    {
+      "epoch": 0.7549868272487769,
+      "grad_norm": 0.5033829212188721,
+      "learning_rate": 4.968944435110512e-05,
+      "loss": 0.5821,
+      "step": 1003
+    },
+    {
+      "epoch": 0.7557395558901017,
+      "grad_norm": 0.3189866542816162,
+      "learning_rate": 4.968821117256516e-05,
+      "loss": 0.6065,
+      "step": 1004
+    },
+    {
+      "epoch": 0.7564922845314265,
+      "grad_norm": 0.32727116346359253,
+      "learning_rate": 4.968697556755578e-05,
+      "loss": 0.5854,
+      "step": 1005
+    },
+    {
+      "epoch": 0.7572450131727512,
+      "grad_norm": 0.31486010551452637,
+      "learning_rate": 4.968573753621212e-05,
+      "loss": 0.4721,
+      "step": 1006
+    },
+    {
+      "epoch": 0.757997741814076,
+      "grad_norm": 0.2902809977531433,
+      "learning_rate": 4.9684497078669554e-05,
+      "loss": 0.5688,
+      "step": 1007
+    },
+    {
+      "epoch": 0.7587504704554008,
+      "grad_norm": 0.32889416813850403,
+      "learning_rate": 4.9683254195063746e-05,
+      "loss": 0.4599,
+      "step": 1008
+    },
+    {
+      "epoch": 0.7595031990967256,
+      "grad_norm": 0.2895301580429077,
+      "learning_rate": 4.96820088855306e-05,
+      "loss": 0.5059,
+      "step": 1009
+    },
+    {
+      "epoch": 0.7602559277380504,
+      "grad_norm": 0.301022469997406,
+      "learning_rate": 4.968076115020632e-05,
+      "loss": 0.5305,
+      "step": 1010
+    },
+    {
+      "epoch": 0.7610086563793752,
+      "grad_norm": 0.3726276457309723,
+      "learning_rate": 4.9679510989227355e-05,
+      "loss": 0.496,
+      "step": 1011
+    },
+    {
+      "epoch": 0.7617613850207,
+      "grad_norm": 0.2959136664867401,
+      "learning_rate": 4.96782584027304e-05,
+      "loss": 0.6012,
+      "step": 1012
+    },
+    {
+      "epoch": 0.7625141136620248,
+      "grad_norm": 0.2398798167705536,
+      "learning_rate": 4.967700339085246e-05,
+      "loss": 0.4928,
+      "step": 1013
+    },
+    {
+      "epoch": 0.7632668423033496,
+      "grad_norm": 0.2528080940246582,
+      "learning_rate": 4.9675745953730766e-05,
+      "loss": 0.5563,
+      "step": 1014
+    },
+    {
+      "epoch": 0.7640195709446744,
+      "grad_norm": 0.25908389687538147,
+      "learning_rate": 4.9674486091502843e-05,
+      "loss": 0.5896,
+      "step": 1015
+    },
+    {
+      "epoch": 0.7647722995859992,
+      "grad_norm": 0.2761833965778351,
+      "learning_rate": 4.967322380430646e-05,
+      "loss": 0.4849,
+      "step": 1016
+    },
+    {
+      "epoch": 0.765525028227324,
+      "grad_norm": 0.2416018843650818,
+      "learning_rate": 4.9671959092279655e-05,
+      "loss": 0.5168,
+      "step": 1017
+    },
+    {
+      "epoch": 0.7662777568686489,
+      "grad_norm": 0.22142943739891052,
+      "learning_rate": 4.9670691955560725e-05,
+      "loss": 0.5615,
+      "step": 1018
+    },
+    {
+      "epoch": 0.7670304855099737,
+      "grad_norm": 0.19668330252170563,
+      "learning_rate": 4.9669422394288276e-05,
+      "loss": 0.5165,
+      "step": 1019
+    },
+    {
+      "epoch": 0.7677832141512985,
+      "grad_norm": 0.2484157830476761,
+      "learning_rate": 4.966815040860112e-05,
+      "loss": 0.3461,
+      "step": 1020
+    },
+    {
+      "epoch": 0.7685359427926233,
+      "grad_norm": 0.2753776013851166,
+      "learning_rate": 4.966687599863836e-05,
+      "loss": 0.5023,
+      "step": 1021
+    },
+    {
+      "epoch": 0.7692886714339481,
+      "grad_norm": 0.21628433465957642,
+      "learning_rate": 4.966559916453937e-05,
+      "loss": 0.5921,
+      "step": 1022
+    },
+    {
+      "epoch": 0.7700414000752729,
+      "grad_norm": 0.21107886731624603,
+      "learning_rate": 4.966431990644379e-05,
+      "loss": 0.6285,
+      "step": 1023
+    },
+    {
+      "epoch": 0.7707941287165977,
+      "grad_norm": 0.3167427182197571,
+      "learning_rate": 4.966303822449149e-05,
+      "loss": 0.4862,
+      "step": 1024
+    },
+    {
+      "epoch": 0.7715468573579225,
+      "grad_norm": 0.306448370218277,
+      "learning_rate": 4.966175411882267e-05,
+      "loss": 0.5101,
+      "step": 1025
+    },
+    {
+      "epoch": 0.7722995859992473,
+      "grad_norm": 0.25023317337036133,
+      "learning_rate": 4.966046758957773e-05,
+      "loss": 0.6461,
+      "step": 1026
+    },
+    {
+      "epoch": 0.7730523146405721,
+      "grad_norm": 0.2553001940250397,
+      "learning_rate": 4.965917863689737e-05,
+      "loss": 0.5059,
+      "step": 1027
+    },
+    {
+      "epoch": 0.7738050432818969,
+      "grad_norm": 0.3020748198032379,
+      "learning_rate": 4.965788726092256e-05,
+      "loss": 0.6352,
+      "step": 1028
+    },
+    {
+      "epoch": 0.7745577719232217,
+      "grad_norm": 0.22905421257019043,
+      "learning_rate": 4.9656593461794505e-05,
+      "loss": 0.5135,
+      "step": 1029
+    },
+    {
+      "epoch": 0.7753105005645465,
+      "grad_norm": 0.2810943126678467,
+      "learning_rate": 4.96552972396547e-05,
+      "loss": 0.425,
+      "step": 1030
+    },
+    {
+      "epoch": 0.7760632292058712,
+      "grad_norm": 0.35539254546165466,
+      "learning_rate": 4.96539985946449e-05,
+      "loss": 0.4737,
+      "step": 1031
+    },
+    {
+      "epoch": 0.776815957847196,
+      "grad_norm": 0.25615307688713074,
+      "learning_rate": 4.965269752690713e-05,
+      "loss": 0.563,
+      "step": 1032
+    },
+    {
+      "epoch": 0.7775686864885208,
+      "grad_norm": 0.3419210612773895,
+      "learning_rate": 4.9651394036583645e-05,
+      "loss": 0.6773,
+      "step": 1033
+    },
+    {
+      "epoch": 0.7783214151298457,
+      "grad_norm": 0.37567681074142456,
+      "learning_rate": 4.965008812381703e-05,
+      "loss": 0.5499,
+      "step": 1034
+    },
+    {
+      "epoch": 0.7790741437711705,
+      "grad_norm": 0.24405786395072937,
+      "learning_rate": 4.9648779788750074e-05,
+      "loss": 0.5563,
+      "step": 1035
+    },
+    {
+      "epoch": 0.7798268724124953,
+      "grad_norm": 0.26948314905166626,
+      "learning_rate": 4.9647469031525865e-05,
+      "loss": 0.4767,
+      "step": 1036
+    },
+    {
+      "epoch": 0.7805796010538201,
+      "grad_norm": 0.27862128615379333,
+      "learning_rate": 4.964615585228773e-05,
+      "loss": 0.5577,
+      "step": 1037
+    },
+    {
+      "epoch": 0.7813323296951449,
+      "grad_norm": 0.3796156644821167,
+      "learning_rate": 4.9644840251179305e-05,
+      "loss": 0.5682,
+      "step": 1038
+    },
+    {
+      "epoch": 0.7820850583364697,
+      "grad_norm": 0.33897826075553894,
+      "learning_rate": 4.964352222834443e-05,
+      "loss": 0.6418,
+      "step": 1039
+    },
+    {
+      "epoch": 0.7828377869777945,
+      "grad_norm": 0.28057190775871277,
+      "learning_rate": 4.9642201783927265e-05,
+      "loss": 0.4904,
+      "step": 1040
+    },
+    {
+      "epoch": 0.7835905156191193,
+      "grad_norm": 0.35910746455192566,
+      "learning_rate": 4.964087891807219e-05,
+      "loss": 0.6355,
+      "step": 1041
+    },
+    {
+      "epoch": 0.7843432442604441,
+      "grad_norm": 0.2763392925262451,
+      "learning_rate": 4.96395536309239e-05,
+      "loss": 0.4713,
+      "step": 1042
+    },
+    {
+      "epoch": 0.7850959729017689,
+      "grad_norm": 0.3383961617946625,
+      "learning_rate": 4.963822592262731e-05,
+      "loss": 0.5456,
+      "step": 1043
+    },
+    {
+      "epoch": 0.7858487015430937,
+      "grad_norm": 0.29178303480148315,
+      "learning_rate": 4.96368957933276e-05,
+      "loss": 0.5679,
+      "step": 1044
+    },
+    {
+      "epoch": 0.7866014301844185,
+      "grad_norm": 0.3238036334514618,
+      "learning_rate": 4.9635563243170266e-05,
+      "loss": 0.604,
+      "step": 1045
+    },
+    {
+      "epoch": 0.7873541588257433,
+      "grad_norm": 0.3299280107021332,
+      "learning_rate": 4.963422827230101e-05,
+      "loss": 0.5213,
+      "step": 1046
+    },
+    {
+      "epoch": 0.7881068874670681,
+      "grad_norm": 0.26834118366241455,
+      "learning_rate": 4.963289088086583e-05,
+      "loss": 0.5896,
+      "step": 1047
+    },
+    {
+      "epoch": 0.788859616108393,
+      "grad_norm": 0.27177783846855164,
+      "learning_rate": 4.963155106901098e-05,
+      "loss": 0.5676,
+      "step": 1048
+    },
+    {
+      "epoch": 0.7896123447497178,
+      "grad_norm": 0.290766179561615,
+      "learning_rate": 4.963020883688297e-05,
+      "loss": 0.5426,
+      "step": 1049
+    },
+    {
+      "epoch": 0.7903650733910426,
+      "grad_norm": 0.2960302531719208,
+      "learning_rate": 4.96288641846286e-05,
+      "loss": 0.5379,
+      "step": 1050
+    },
+    {
+      "epoch": 0.7911178020323674,
+      "grad_norm": 0.2903648018836975,
+      "learning_rate": 4.962751711239492e-05,
+      "loss": 0.4933,
+      "step": 1051
+    },
+    {
+      "epoch": 0.7918705306736922,
+      "grad_norm": 0.3923456072807312,
+      "learning_rate": 4.962616762032923e-05,
+      "loss": 0.5451,
+      "step": 1052
+    },
+    {
+      "epoch": 0.792623259315017,
+      "grad_norm": 0.34004461765289307,
+      "learning_rate": 4.962481570857912e-05,
+      "loss": 0.5076,
+      "step": 1053
+    },
+    {
+      "epoch": 0.7933759879563418,
+      "grad_norm": 0.32796475291252136,
+      "learning_rate": 4.9623461377292426e-05,
+      "loss": 0.4074,
+      "step": 1054
+    },
+    {
+      "epoch": 0.7941287165976666,
+      "grad_norm": 0.22613635659217834,
+      "learning_rate": 4.962210462661725e-05,
+      "loss": 0.5727,
+      "step": 1055
+    },
+    {
+      "epoch": 0.7948814452389913,
+      "grad_norm": 0.3188192546367645,
+      "learning_rate": 4.9620745456701985e-05,
+      "loss": 0.5018,
+      "step": 1056
+    },
+    {
+      "epoch": 0.7956341738803161,
+      "grad_norm": 0.32683590054512024,
+      "learning_rate": 4.961938386769525e-05,
+      "loss": 0.5405,
+      "step": 1057
+    },
+    {
+      "epoch": 0.7963869025216409,
+      "grad_norm": 0.2933056056499481,
+      "learning_rate": 4.9618019859745946e-05,
+      "loss": 0.5904,
+      "step": 1058
+    },
+    {
+      "epoch": 0.7971396311629657,
+      "grad_norm": 0.2852746248245239,
+      "learning_rate": 4.961665343300325e-05,
+      "loss": 0.6317,
+      "step": 1059
+    },
+    {
+      "epoch": 0.7978923598042905,
+      "grad_norm": 0.30428969860076904,
+      "learning_rate": 4.961528458761659e-05,
+      "loss": 0.4694,
+      "step": 1060
+    },
+    {
+      "epoch": 0.7986450884456153,
+      "grad_norm": 0.26382648944854736,
+      "learning_rate": 4.9613913323735645e-05,
+      "loss": 0.5015,
+      "step": 1061
+    },
+    {
+      "epoch": 0.7993978170869401,
+      "grad_norm": 0.25007516145706177,
+      "learning_rate": 4.96125396415104e-05,
+      "loss": 0.4837,
+      "step": 1062
+    },
+    {
+      "epoch": 0.8001505457282649,
+      "grad_norm": 0.3709091246128082,
+      "learning_rate": 4.961116354109105e-05,
+      "loss": 0.4954,
+      "step": 1063
+    },
+    {
+      "epoch": 0.8009032743695897,
+      "grad_norm": 0.32119905948638916,
+      "learning_rate": 4.960978502262811e-05,
+      "loss": 0.5909,
+      "step": 1064
+    },
+    {
+      "epoch": 0.8016560030109146,
+      "grad_norm": 0.2598002254962921,
+      "learning_rate": 4.960840408627233e-05,
+      "loss": 0.5175,
+      "step": 1065
+    },
+    {
+      "epoch": 0.8024087316522394,
+      "grad_norm": 0.3404490649700165,
+      "learning_rate": 4.96070207321747e-05,
+      "loss": 0.5673,
+      "step": 1066
+    },
+    {
+      "epoch": 0.8031614602935642,
+      "grad_norm": 0.3339163064956665,
+      "learning_rate": 4.960563496048653e-05,
+      "loss": 0.6282,
+      "step": 1067
+    },
+    {
+      "epoch": 0.803914188934889,
+      "grad_norm": 0.2878272533416748,
+      "learning_rate": 4.960424677135935e-05,
+      "loss": 0.4558,
+      "step": 1068
+    },
+    {
+      "epoch": 0.8046669175762138,
+      "grad_norm": 0.3069824278354645,
+      "learning_rate": 4.960285616494498e-05,
+      "loss": 0.535,
+      "step": 1069
+    },
+    {
+      "epoch": 0.8054196462175386,
+      "grad_norm": 0.2436538189649582,
+      "learning_rate": 4.9601463141395496e-05,
+      "loss": 0.4869,
+      "step": 1070
+    },
+    {
+      "epoch": 0.8061723748588634,
+      "grad_norm": 0.32743293046951294,
+      "learning_rate": 4.9600067700863225e-05,
+      "loss": 0.4633,
+      "step": 1071
+    },
+    {
+      "epoch": 0.8069251035001882,
+      "grad_norm": 0.43734633922576904,
+      "learning_rate": 4.9598669843500784e-05,
+      "loss": 0.4894,
+      "step": 1072
+    },
+    {
+      "epoch": 0.807677832141513,
+      "grad_norm": 0.2927147150039673,
+      "learning_rate": 4.959726956946104e-05,
+      "loss": 0.4273,
+      "step": 1073
+    },
+    {
+      "epoch": 0.8084305607828378,
+      "grad_norm": 0.2244957536458969,
+      "learning_rate": 4.95958668788971e-05,
+      "loss": 0.5592,
+      "step": 1074
+    },
+    {
+      "epoch": 0.8091832894241626,
+      "grad_norm": 0.22839751839637756,
+      "learning_rate": 4.959446177196238e-05,
+      "loss": 0.4632,
+      "step": 1075
+    },
+    {
+      "epoch": 0.8099360180654874,
+      "grad_norm": 0.22157199680805206,
+      "learning_rate": 4.959305424881055e-05,
+      "loss": 0.4638,
+      "step": 1076
+    },
+    {
+      "epoch": 0.8106887467068122,
+      "grad_norm": 0.2382076382637024,
+      "learning_rate": 4.959164430959552e-05,
+      "loss": 0.5403,
+      "step": 1077
+    },
+    {
+      "epoch": 0.811441475348137,
+      "grad_norm": 0.23634155094623566,
+      "learning_rate": 4.9590231954471484e-05,
+      "loss": 0.5202,
+      "step": 1078
+    },
+    {
+      "epoch": 0.8121942039894618,
+      "grad_norm": 0.33705419301986694,
+      "learning_rate": 4.9588817183592896e-05,
+      "loss": 0.523,
+      "step": 1079
+    },
+    {
+      "epoch": 0.8129469326307867,
+      "grad_norm": 0.23948392271995544,
+      "learning_rate": 4.958739999711446e-05,
+      "loss": 0.6059,
+      "step": 1080
+    },
+    {
+      "epoch": 0.8136996612721114,
+      "grad_norm": 0.3251425325870514,
+      "learning_rate": 4.9585980395191185e-05,
+      "loss": 0.5897,
+      "step": 1081
+    },
+    {
+      "epoch": 0.8144523899134362,
+      "grad_norm": 0.2876942455768585,
+      "learning_rate": 4.958455837797828e-05,
+      "loss": 0.5757,
+      "step": 1082
+    },
+    {
+      "epoch": 0.815205118554761,
+      "grad_norm": 0.3007805347442627,
+      "learning_rate": 4.958313394563128e-05,
+      "loss": 0.519,
+      "step": 1083
+    },
+    {
+      "epoch": 0.8159578471960858,
+      "grad_norm": 0.4013361632823944,
+      "learning_rate": 4.9581707098305965e-05,
+      "loss": 0.5193,
+      "step": 1084
+    },
+    {
+      "epoch": 0.8167105758374106,
+      "grad_norm": 0.22883665561676025,
+      "learning_rate": 4.958027783615835e-05,
+      "loss": 0.4353,
+      "step": 1085
+    },
+    {
+      "epoch": 0.8174633044787354,
+      "grad_norm": 0.3011414408683777,
+      "learning_rate": 4.957884615934475e-05,
+      "loss": 0.5786,
+      "step": 1086
+    },
+    {
+      "epoch": 0.8182160331200602,
+      "grad_norm": 0.2889673411846161,
+      "learning_rate": 4.9577412068021737e-05,
+      "loss": 0.5052,
+      "step": 1087
+    },
+    {
+      "epoch": 0.818968761761385,
+      "grad_norm": 0.2107551395893097,
+      "learning_rate": 4.957597556234613e-05,
+      "loss": 0.5212,
+      "step": 1088
+    },
+    {
+      "epoch": 0.8197214904027098,
+      "grad_norm": 0.27481529116630554,
+      "learning_rate": 4.9574536642475015e-05,
+      "loss": 0.4459,
+      "step": 1089
+    },
+    {
+      "epoch": 0.8204742190440346,
+      "grad_norm": 0.23361338675022125,
+      "learning_rate": 4.957309530856577e-05,
+      "loss": 0.5233,
+      "step": 1090
+    },
+    {
+      "epoch": 0.8212269476853594,
+      "grad_norm": 0.257376492023468,
+      "learning_rate": 4.957165156077601e-05,
+      "loss": 0.5519,
+      "step": 1091
+    },
+    {
+      "epoch": 0.8219796763266842,
+      "grad_norm": 0.35294678807258606,
+      "learning_rate": 4.957020539926362e-05,
+      "loss": 0.6035,
+      "step": 1092
+    },
+    {
+      "epoch": 0.822732404968009,
+      "grad_norm": 0.31586316227912903,
+      "learning_rate": 4.9568756824186746e-05,
+      "loss": 0.5302,
+      "step": 1093
+    },
+    {
+      "epoch": 0.8234851336093338,
+      "grad_norm": 0.20815007388591766,
+      "learning_rate": 4.95673058357038e-05,
+      "loss": 0.5965,
+      "step": 1094
+    },
+    {
+      "epoch": 0.8242378622506586,
+      "grad_norm": 0.2636430263519287,
+      "learning_rate": 4.956585243397348e-05,
+      "loss": 0.5096,
+      "step": 1095
+    },
+    {
+      "epoch": 0.8249905908919835,
+      "grad_norm": 0.2865442633628845,
+      "learning_rate": 4.95643966191547e-05,
+      "loss": 0.4891,
+      "step": 1096
+    },
+    {
+      "epoch": 0.8257433195333083,
+      "grad_norm": 0.2785143554210663,
+      "learning_rate": 4.956293839140669e-05,
+      "loss": 0.6145,
+      "step": 1097
+    },
+    {
+      "epoch": 0.8264960481746331,
+      "grad_norm": 0.2441985011100769,
+      "learning_rate": 4.95614777508889e-05,
+      "loss": 0.571,
+      "step": 1098
+    },
+    {
+      "epoch": 0.8272487768159579,
+      "grad_norm": 0.29185763001441956,
+      "learning_rate": 4.9560014697761074e-05,
+      "loss": 0.5002,
+      "step": 1099
+    },
+    {
+      "epoch": 0.8280015054572827,
+      "grad_norm": 0.2426265925168991,
+      "learning_rate": 4.955854923218321e-05,
+      "loss": 0.5413,
+      "step": 1100
+    },
+    {
+      "epoch": 0.8287542340986075,
+      "grad_norm": 0.3618340790271759,
+      "learning_rate": 4.9557081354315564e-05,
+      "loss": 0.3622,
+      "step": 1101
+    },
+    {
+      "epoch": 0.8295069627399323,
+      "grad_norm": 0.41981735825538635,
+      "learning_rate": 4.955561106431867e-05,
+      "loss": 0.4388,
+      "step": 1102
+    },
+    {
+      "epoch": 0.8302596913812571,
+      "grad_norm": 0.2735825181007385,
+      "learning_rate": 4.95541383623533e-05,
+      "loss": 0.5364,
+      "step": 1103
+    },
+    {
+      "epoch": 0.8310124200225819,
+      "grad_norm": 0.41078656911849976,
+      "learning_rate": 4.9552663248580524e-05,
+      "loss": 0.5081,
+      "step": 1104
+    },
+    {
+      "epoch": 0.8317651486639067,
+      "grad_norm": 0.40181565284729004,
+      "learning_rate": 4.955118572316165e-05,
+      "loss": 0.5308,
+      "step": 1105
+    },
+    {
+      "epoch": 0.8325178773052314,
+      "grad_norm": 0.2651938796043396,
+      "learning_rate": 4.954970578625826e-05,
+      "loss": 0.4421,
+      "step": 1106
+    },
+    {
+      "epoch": 0.8332706059465562,
+      "grad_norm": 0.38710781931877136,
+      "learning_rate": 4.954822343803219e-05,
+      "loss": 0.4705,
+      "step": 1107
+    },
+    {
+      "epoch": 0.834023334587881,
+      "grad_norm": 0.25301316380500793,
+      "learning_rate": 4.954673867864556e-05,
+      "loss": 0.4975,
+      "step": 1108
+    },
+    {
+      "epoch": 0.8347760632292058,
+      "grad_norm": 0.35520708560943604,
+      "learning_rate": 4.9545251508260725e-05,
+      "loss": 0.5629,
+      "step": 1109
+    },
+    {
+      "epoch": 0.8355287918705306,
+      "grad_norm": 0.3409935534000397,
+      "learning_rate": 4.954376192704034e-05,
+      "loss": 0.466,
+      "step": 1110
+    },
+    {
+      "epoch": 0.8362815205118554,
+      "grad_norm": 0.3673972487449646,
+      "learning_rate": 4.9542269935147295e-05,
+      "loss": 0.465,
+      "step": 1111
+    },
+    {
+      "epoch": 0.8370342491531803,
+      "grad_norm": 0.3539561629295349,
+      "learning_rate": 4.954077553274474e-05,
+      "loss": 0.5082,
+      "step": 1112
+    },
+    {
+      "epoch": 0.8377869777945051,
+      "grad_norm": 0.24657347798347473,
+      "learning_rate": 4.953927871999611e-05,
+      "loss": 0.556,
+      "step": 1113
+    },
+    {
+      "epoch": 0.8385397064358299,
+      "grad_norm": 0.39673328399658203,
+      "learning_rate": 4.95377794970651e-05,
+      "loss": 0.5583,
+      "step": 1114
+    },
+    {
+      "epoch": 0.8392924350771547,
+      "grad_norm": 0.2787242829799652,
+      "learning_rate": 4.953627786411566e-05,
+      "loss": 0.5286,
+      "step": 1115
+    },
+    {
+      "epoch": 0.8400451637184795,
+      "grad_norm": 0.288324773311615,
+      "learning_rate": 4.9534773821312004e-05,
+      "loss": 0.4355,
+      "step": 1116
+    },
+    {
+      "epoch": 0.8407978923598043,
+      "grad_norm": 0.3526324927806854,
+      "learning_rate": 4.95332673688186e-05,
+      "loss": 0.518,
+      "step": 1117
+    },
+    {
+      "epoch": 0.8415506210011291,
+      "grad_norm": 0.2872489094734192,
+      "learning_rate": 4.953175850680022e-05,
+      "loss": 0.5329,
+      "step": 1118
+    },
+    {
+      "epoch": 0.8423033496424539,
+      "grad_norm": 0.3045298159122467,
+      "learning_rate": 4.953024723542185e-05,
+      "loss": 0.4951,
+      "step": 1119
+    },
+    {
+      "epoch": 0.8430560782837787,
+      "grad_norm": 0.2534947395324707,
+      "learning_rate": 4.952873355484876e-05,
+      "loss": 0.524,
+      "step": 1120
+    },
+    {
+      "epoch": 0.8438088069251035,
+      "grad_norm": 0.2876877188682556,
+      "learning_rate": 4.952721746524648e-05,
+      "loss": 0.3804,
+      "step": 1121
+    },
+    {
+      "epoch": 0.8445615355664283,
+      "grad_norm": 0.3373505771160126,
+      "learning_rate": 4.9525698966780835e-05,
+      "loss": 0.4974,
+      "step": 1122
+    },
+    {
+      "epoch": 0.8453142642077531,
+      "grad_norm": 0.2768784165382385,
+      "learning_rate": 4.9524178059617854e-05,
+      "loss": 0.5799,
+      "step": 1123
+    },
+    {
+      "epoch": 0.8460669928490779,
+      "grad_norm": 0.32221612334251404,
+      "learning_rate": 4.952265474392387e-05,
+      "loss": 0.452,
+      "step": 1124
+    },
+    {
+      "epoch": 0.8468197214904027,
+      "grad_norm": 0.31510940194129944,
+      "learning_rate": 4.952112901986547e-05,
+      "loss": 0.6227,
+      "step": 1125
+    },
+    {
+      "epoch": 0.8475724501317276,
+      "grad_norm": 0.2869284749031067,
+      "learning_rate": 4.951960088760953e-05,
+      "loss": 0.4337,
+      "step": 1126
+    },
+    {
+      "epoch": 0.8483251787730524,
+      "grad_norm": 0.4039064049720764,
+      "learning_rate": 4.951807034732312e-05,
+      "loss": 0.5403,
+      "step": 1127
+    },
+    {
+      "epoch": 0.8490779074143772,
+      "grad_norm": 0.29176437854766846,
+      "learning_rate": 4.9516537399173655e-05,
+      "loss": 0.6675,
+      "step": 1128
+    },
+    {
+      "epoch": 0.849830636055702,
+      "grad_norm": 0.30433785915374756,
+      "learning_rate": 4.951500204332876e-05,
+      "loss": 0.6421,
+      "step": 1129
+    },
+    {
+      "epoch": 0.8505833646970267,
+      "grad_norm": 0.31294238567352295,
+      "learning_rate": 4.951346427995633e-05,
+      "loss": 0.5219,
+      "step": 1130
+    },
+    {
+      "epoch": 0.8513360933383515,
+      "grad_norm": 0.23071126639842987,
+      "learning_rate": 4.9511924109224555e-05,
+      "loss": 0.5446,
+      "step": 1131
+    },
+    {
+      "epoch": 0.8520888219796763,
+      "grad_norm": 0.36387231945991516,
+      "learning_rate": 4.951038153130185e-05,
+      "loss": 0.5063,
+      "step": 1132
+    },
+    {
+      "epoch": 0.8528415506210011,
+      "grad_norm": 0.3036707937717438,
+      "learning_rate": 4.950883654635691e-05,
+      "loss": 0.5086,
+      "step": 1133
+    },
+    {
+      "epoch": 0.8535942792623259,
+      "grad_norm": 0.27850112318992615,
+      "learning_rate": 4.9507289154558694e-05,
+      "loss": 0.5438,
+      "step": 1134
+    },
+    {
+      "epoch": 0.8543470079036507,
+      "grad_norm": 0.2896309196949005,
+      "learning_rate": 4.950573935607643e-05,
+      "loss": 0.4236,
+      "step": 1135
+    },
+    {
+      "epoch": 0.8550997365449755,
+      "grad_norm": 0.24845333397388458,
+      "learning_rate": 4.950418715107959e-05,
+      "loss": 0.4975,
+      "step": 1136
+    },
+    {
+      "epoch": 0.8558524651863003,
+      "grad_norm": 0.23392173647880554,
+      "learning_rate": 4.950263253973792e-05,
+      "loss": 0.5162,
+      "step": 1137
+    },
+    {
+      "epoch": 0.8566051938276251,
+      "grad_norm": 0.22547271847724915,
+      "learning_rate": 4.950107552222145e-05,
+      "loss": 0.5156,
+      "step": 1138
+    },
+    {
+      "epoch": 0.8573579224689499,
+      "grad_norm": 0.28993135690689087,
+      "learning_rate": 4.949951609870043e-05,
+      "loss": 0.5752,
+      "step": 1139
+    },
+    {
+      "epoch": 0.8581106511102747,
+      "grad_norm": 0.23005807399749756,
+      "learning_rate": 4.949795426934541e-05,
+      "loss": 0.4282,
+      "step": 1140
+    },
+    {
+      "epoch": 0.8588633797515995,
+      "grad_norm": 0.26891854405403137,
+      "learning_rate": 4.9496390034327184e-05,
+      "loss": 0.4568,
+      "step": 1141
+    },
+    {
+      "epoch": 0.8596161083929243,
+      "grad_norm": 0.22370155155658722,
+      "learning_rate": 4.949482339381681e-05,
+      "loss": 0.4595,
+      "step": 1142
+    },
+    {
+      "epoch": 0.8603688370342492,
+      "grad_norm": 0.39190706610679626,
+      "learning_rate": 4.949325434798562e-05,
+      "loss": 0.5103,
+      "step": 1143
+    },
+    {
+      "epoch": 0.861121565675574,
+      "grad_norm": 0.28151118755340576,
+      "learning_rate": 4.949168289700521e-05,
+      "loss": 0.5515,
+      "step": 1144
+    },
+    {
+      "epoch": 0.8618742943168988,
+      "grad_norm": 0.238261416554451,
+      "learning_rate": 4.9490109041047413e-05,
+      "loss": 0.5369,
+      "step": 1145
+    },
+    {
+      "epoch": 0.8626270229582236,
+      "grad_norm": 0.22208376228809357,
+      "learning_rate": 4.948853278028436e-05,
+      "loss": 0.535,
+      "step": 1146
+    },
+    {
+      "epoch": 0.8633797515995484,
+      "grad_norm": 0.23863206803798676,
+      "learning_rate": 4.948695411488841e-05,
+      "loss": 0.4031,
+      "step": 1147
+    },
+    {
+      "epoch": 0.8641324802408732,
+      "grad_norm": 0.21684838831424713,
+      "learning_rate": 4.948537304503222e-05,
+      "loss": 0.4593,
+      "step": 1148
+    },
+    {
+      "epoch": 0.864885208882198,
+      "grad_norm": 0.2647448480129242,
+      "learning_rate": 4.948378957088869e-05,
+      "loss": 0.5013,
+      "step": 1149
+    },
+    {
+      "epoch": 0.8656379375235228,
+      "grad_norm": 0.2307969629764557,
+      "learning_rate": 4.948220369263099e-05,
+      "loss": 0.4838,
+      "step": 1150
+    },
+    {
+      "epoch": 0.8663906661648476,
+      "grad_norm": 0.24615851044654846,
+      "learning_rate": 4.9480615410432535e-05,
+      "loss": 0.457,
+      "step": 1151
+    },
+    {
+      "epoch": 0.8671433948061724,
+      "grad_norm": 0.2506524920463562,
+      "learning_rate": 4.947902472446703e-05,
+      "loss": 0.5473,
+      "step": 1152
+    },
+    {
+      "epoch": 0.8678961234474972,
+      "grad_norm": 0.23276449739933014,
+      "learning_rate": 4.947743163490842e-05,
+      "loss": 0.5257,
+      "step": 1153
+    },
+    {
+      "epoch": 0.868648852088822,
+      "grad_norm": 0.25481903553009033,
+      "learning_rate": 4.947583614193094e-05,
+      "loss": 0.5094,
+      "step": 1154
+    },
+    {
+      "epoch": 0.8694015807301467,
+      "grad_norm": 0.2583901286125183,
+      "learning_rate": 4.947423824570905e-05,
+      "loss": 0.5165,
+      "step": 1155
+    },
+    {
+      "epoch": 0.8701543093714715,
+      "grad_norm": 0.2211441546678543,
+      "learning_rate": 4.9472637946417504e-05,
+      "loss": 0.405,
+      "step": 1156
+    },
+    {
+      "epoch": 0.8709070380127963,
+      "grad_norm": 0.2817108631134033,
+      "learning_rate": 4.9471035244231315e-05,
+      "loss": 0.5154,
+      "step": 1157
+    },
+    {
+      "epoch": 0.8716597666541211,
+      "grad_norm": 0.2779005467891693,
+      "learning_rate": 4.946943013932574e-05,
+      "loss": 0.5114,
+      "step": 1158
+    },
+    {
+      "epoch": 0.872412495295446,
+      "grad_norm": 0.31431448459625244,
+      "learning_rate": 4.9467822631876314e-05,
+      "loss": 0.5444,
+      "step": 1159
+    },
+    {
+      "epoch": 0.8731652239367708,
+      "grad_norm": 0.27327725291252136,
+      "learning_rate": 4.9466212722058844e-05,
+      "loss": 0.3497,
+      "step": 1160
+    },
+    {
+      "epoch": 0.8739179525780956,
+      "grad_norm": 0.27097684144973755,
+      "learning_rate": 4.946460041004936e-05,
+      "loss": 0.4256,
+      "step": 1161
+    },
+    {
+      "epoch": 0.8746706812194204,
+      "grad_norm": 0.24044917523860931,
+      "learning_rate": 4.946298569602421e-05,
+      "loss": 0.5014,
+      "step": 1162
+    },
+    {
+      "epoch": 0.8754234098607452,
+      "grad_norm": 0.22893773019313812,
+      "learning_rate": 4.946136858015996e-05,
+      "loss": 0.4604,
+      "step": 1163
+    },
+    {
+      "epoch": 0.87617613850207,
+      "grad_norm": 0.26195216178894043,
+      "learning_rate": 4.9459749062633456e-05,
+      "loss": 0.4232,
+      "step": 1164
+    },
+    {
+      "epoch": 0.8769288671433948,
+      "grad_norm": 0.3020746409893036,
+      "learning_rate": 4.9458127143621816e-05,
+      "loss": 0.4589,
+      "step": 1165
+    },
+    {
+      "epoch": 0.8776815957847196,
+      "grad_norm": 0.2393629401922226,
+      "learning_rate": 4.9456502823302407e-05,
+      "loss": 0.5256,
+      "step": 1166
+    },
+    {
+      "epoch": 0.8784343244260444,
+      "grad_norm": 0.22941642999649048,
+      "learning_rate": 4.945487610185286e-05,
+      "loss": 0.5828,
+      "step": 1167
+    },
+    {
+      "epoch": 0.8791870530673692,
+      "grad_norm": 0.3992302715778351,
+      "learning_rate": 4.945324697945107e-05,
+      "loss": 0.5017,
+      "step": 1168
+    },
+    {
+      "epoch": 0.879939781708694,
+      "grad_norm": 0.3161161243915558,
+      "learning_rate": 4.9451615456275195e-05,
+      "loss": 0.4228,
+      "step": 1169
+    },
+    {
+      "epoch": 0.8806925103500188,
+      "grad_norm": 0.21980033814907074,
+      "learning_rate": 4.944998153250366e-05,
+      "loss": 0.5826,
+      "step": 1170
+    },
+    {
+      "epoch": 0.8814452389913436,
+      "grad_norm": 0.45348861813545227,
+      "learning_rate": 4.9448345208315145e-05,
+      "loss": 0.5379,
+      "step": 1171
+    },
+    {
+      "epoch": 0.8821979676326684,
+      "grad_norm": 0.3154127299785614,
+      "learning_rate": 4.9446706483888596e-05,
+      "loss": 0.4748,
+      "step": 1172
+    },
+    {
+      "epoch": 0.8829506962739933,
+      "grad_norm": 0.3891940414905548,
+      "learning_rate": 4.9445065359403225e-05,
+      "loss": 0.5079,
+      "step": 1173
+    },
+    {
+      "epoch": 0.8837034249153181,
+      "grad_norm": 0.34833216667175293,
+      "learning_rate": 4.944342183503849e-05,
+      "loss": 0.4987,
+      "step": 1174
+    },
+    {
+      "epoch": 0.8844561535566429,
+      "grad_norm": 0.6995524168014526,
+      "learning_rate": 4.9441775910974146e-05,
+      "loss": 0.471,
+      "step": 1175
+    },
+    {
+      "epoch": 0.8852088821979677,
+      "grad_norm": 0.3556317090988159,
+      "learning_rate": 4.944012758739017e-05,
+      "loss": 0.4588,
+      "step": 1176
+    },
+    {
+      "epoch": 0.8859616108392925,
+      "grad_norm": 0.40076133608818054,
+      "learning_rate": 4.9438476864466836e-05,
+      "loss": 0.5216,
+      "step": 1177
+    },
+    {
+      "epoch": 0.8867143394806173,
+      "grad_norm": 0.3474874496459961,
+      "learning_rate": 4.9436823742384646e-05,
+      "loss": 0.596,
+      "step": 1178
+    },
+    {
+      "epoch": 0.8874670681219421,
+      "grad_norm": 0.3700205981731415,
+      "learning_rate": 4.94351682213244e-05,
+      "loss": 0.5817,
+      "step": 1179
+    },
+    {
+      "epoch": 0.8882197967632668,
+      "grad_norm": 0.3750340938568115,
+      "learning_rate": 4.943351030146714e-05,
+      "loss": 0.5313,
+      "step": 1180
+    },
+    {
+      "epoch": 0.8889725254045916,
+      "grad_norm": 0.28182438015937805,
+      "learning_rate": 4.943184998299416e-05,
+      "loss": 0.4767,
+      "step": 1181
+    },
+    {
+      "epoch": 0.8897252540459164,
+      "grad_norm": 0.3245156705379486,
+      "learning_rate": 4.943018726608705e-05,
+      "loss": 0.4663,
+      "step": 1182
+    },
+    {
+      "epoch": 0.8904779826872412,
+      "grad_norm": 0.27659034729003906,
+      "learning_rate": 4.942852215092763e-05,
+      "loss": 0.4687,
+      "step": 1183
+    },
+    {
+      "epoch": 0.891230711328566,
+      "grad_norm": 0.3289303183555603,
+      "learning_rate": 4.942685463769799e-05,
+      "loss": 0.4289,
+      "step": 1184
+    },
+    {
+      "epoch": 0.8919834399698908,
+      "grad_norm": 0.30149373412132263,
+      "learning_rate": 4.942518472658049e-05,
+      "loss": 0.5705,
+      "step": 1185
+    },
+    {
+      "epoch": 0.8927361686112156,
+      "grad_norm": 0.26159071922302246,
+      "learning_rate": 4.9423512417757766e-05,
+      "loss": 0.4848,
+      "step": 1186
+    },
+    {
+      "epoch": 0.8934888972525404,
+      "grad_norm": 0.252643883228302,
+      "learning_rate": 4.942183771141267e-05,
+      "loss": 0.5828,
+      "step": 1187
+    },
+    {
+      "epoch": 0.8942416258938652,
+      "grad_norm": 0.3003123104572296,
+      "learning_rate": 4.9420160607728374e-05,
+      "loss": 0.5535,
+      "step": 1188
+    },
+    {
+      "epoch": 0.89499435453519,
+      "grad_norm": 0.27188578248023987,
+      "learning_rate": 4.941848110688826e-05,
+      "loss": 0.5008,
+      "step": 1189
+    },
+    {
+      "epoch": 0.8957470831765149,
+      "grad_norm": 0.26913055777549744,
+      "learning_rate": 4.941679920907602e-05,
+      "loss": 0.5585,
+      "step": 1190
+    },
+    {
+      "epoch": 0.8964998118178397,
+      "grad_norm": 0.3241141140460968,
+      "learning_rate": 4.9415114914475566e-05,
+      "loss": 0.4285,
+      "step": 1191
+    },
+    {
+      "epoch": 0.8972525404591645,
+      "grad_norm": 0.3463382422924042,
+      "learning_rate": 4.941342822327109e-05,
+      "loss": 0.4801,
+      "step": 1192
+    },
+    {
+      "epoch": 0.8980052691004893,
+      "grad_norm": 0.25921064615249634,
+      "learning_rate": 4.941173913564705e-05,
+      "loss": 0.5124,
+      "step": 1193
+    },
+    {
+      "epoch": 0.8987579977418141,
+      "grad_norm": 0.29228320717811584,
+      "learning_rate": 4.941004765178816e-05,
+      "loss": 0.5114,
+      "step": 1194
+    },
+    {
+      "epoch": 0.8995107263831389,
+      "grad_norm": 0.30513665080070496,
+      "learning_rate": 4.94083537718794e-05,
+      "loss": 0.4361,
+      "step": 1195
+    },
+    {
+      "epoch": 0.9002634550244637,
+      "grad_norm": 0.24389171600341797,
+      "learning_rate": 4.940665749610601e-05,
+      "loss": 0.5428,
+      "step": 1196
+    },
+    {
+      "epoch": 0.9010161836657885,
+      "grad_norm": 0.2890320420265198,
+      "learning_rate": 4.9404958824653494e-05,
+      "loss": 0.5129,
+      "step": 1197
+    },
+    {
+      "epoch": 0.9017689123071133,
+      "grad_norm": 0.26276320219039917,
+      "learning_rate": 4.9403257757707615e-05,
+      "loss": 0.4338,
+      "step": 1198
+    },
+    {
+      "epoch": 0.9025216409484381,
+      "grad_norm": 0.26051250100135803,
+      "learning_rate": 4.94015542954544e-05,
+      "loss": 0.5773,
+      "step": 1199
+    },
+    {
+      "epoch": 0.9032743695897629,
+      "grad_norm": 0.25636762380599976,
+      "learning_rate": 4.939984843808013e-05,
+      "loss": 0.5355,
+      "step": 1200
+    },
+    {
+      "epoch": 0.9032743695897629,
+      "eval_loss": 0.48609477281570435,
+      "eval_runtime": 452.8354,
+      "eval_samples_per_second": 21.259,
+      "eval_steps_per_second": 0.665,
+      "step": 1200
+    },
+    {
+      "epoch": 0.9040270982310877,
+      "grad_norm": 0.2833421230316162,
+      "learning_rate": 4.939814018577137e-05,
+      "loss": 0.5801,
+      "step": 1201
+    },
+    {
+      "epoch": 0.9047798268724125,
+      "grad_norm": 0.23592475056648254,
+      "learning_rate": 4.939642953871492e-05,
+      "loss": 0.4367,
+      "step": 1202
+    },
+    {
+      "epoch": 0.9055325555137373,
+      "grad_norm": 0.36143290996551514,
+      "learning_rate": 4.9394716497097853e-05,
+      "loss": 0.4536,
+      "step": 1203
+    },
+    {
+      "epoch": 0.9062852841550622,
+      "grad_norm": 0.3903560936450958,
+      "learning_rate": 4.93930010611075e-05,
+      "loss": 0.6157,
+      "step": 1204
+    },
+    {
+      "epoch": 0.9070380127963868,
+      "grad_norm": 0.3298417925834656,
+      "learning_rate": 4.939128323093149e-05,
+      "loss": 0.4545,
+      "step": 1205
+    },
+    {
+      "epoch": 0.9077907414377117,
+      "grad_norm": 0.37502530217170715,
+      "learning_rate": 4.938956300675763e-05,
+      "loss": 0.5126,
+      "step": 1206
+    },
+    {
+      "epoch": 0.9085434700790365,
+      "grad_norm": 0.29354894161224365,
+      "learning_rate": 4.9387840388774094e-05,
+      "loss": 0.457,
+      "step": 1207
+    },
+    {
+      "epoch": 0.9092961987203613,
+      "grad_norm": 0.32724031805992126,
+      "learning_rate": 4.938611537716922e-05,
+      "loss": 0.5053,
+      "step": 1208
+    },
+    {
+      "epoch": 0.9100489273616861,
+      "grad_norm": 0.360363632440567,
+      "learning_rate": 4.9384387972131696e-05,
+      "loss": 0.5809,
+      "step": 1209
+    },
+    {
+      "epoch": 0.9108016560030109,
+      "grad_norm": 0.32081925868988037,
+      "learning_rate": 4.938265817385039e-05,
+      "loss": 0.4809,
+      "step": 1210
+    },
+    {
+      "epoch": 0.9115543846443357,
+      "grad_norm": 0.3068097233772278,
+      "learning_rate": 4.9380925982514493e-05,
+      "loss": 0.4463,
+      "step": 1211
+    },
+    {
+      "epoch": 0.9123071132856605,
+      "grad_norm": 0.28134477138519287,
+      "learning_rate": 4.937919139831343e-05,
+      "loss": 0.5003,
+      "step": 1212
+    },
+    {
+      "epoch": 0.9130598419269853,
+      "grad_norm": 0.26453739404678345,
+      "learning_rate": 4.9377454421436886e-05,
+      "loss": 0.5045,
+      "step": 1213
+    },
+    {
+      "epoch": 0.9138125705683101,
+      "grad_norm": 0.3290517330169678,
+      "learning_rate": 4.937571505207482e-05,
+      "loss": 0.4746,
+      "step": 1214
+    },
+    {
+      "epoch": 0.9145652992096349,
+      "grad_norm": 0.28278452157974243,
+      "learning_rate": 4.937397329041744e-05,
+      "loss": 0.5707,
+      "step": 1215
+    },
+    {
+      "epoch": 0.9153180278509597,
+      "grad_norm": 0.29762253165245056,
+      "learning_rate": 4.9372229136655235e-05,
+      "loss": 0.6243,
+      "step": 1216
+    },
+    {
+      "epoch": 0.9160707564922845,
+      "grad_norm": 0.27343621850013733,
+      "learning_rate": 4.937048259097894e-05,
+      "loss": 0.4469,
+      "step": 1217
+    },
+    {
+      "epoch": 0.9168234851336093,
+      "grad_norm": 0.3225628137588501,
+      "learning_rate": 4.9368733653579544e-05,
+      "loss": 0.3895,
+      "step": 1218
+    },
+    {
+      "epoch": 0.9175762137749341,
+      "grad_norm": 0.24389779567718506,
+      "learning_rate": 4.9366982324648316e-05,
+      "loss": 0.4951,
+      "step": 1219
+    },
+    {
+      "epoch": 0.918328942416259,
+      "grad_norm": 0.2527689039707184,
+      "learning_rate": 4.936522860437678e-05,
+      "loss": 0.531,
+      "step": 1220
+    },
+    {
+      "epoch": 0.9190816710575838,
+      "grad_norm": 0.2563411295413971,
+      "learning_rate": 4.936347249295671e-05,
+      "loss": 0.559,
+      "step": 1221
+    },
+    {
+      "epoch": 0.9198343996989086,
+      "grad_norm": 0.2849944829940796,
+      "learning_rate": 4.9361713990580164e-05,
+      "loss": 0.4136,
+      "step": 1222
+    },
+    {
+      "epoch": 0.9205871283402334,
+      "grad_norm": 0.27929002046585083,
+      "learning_rate": 4.935995309743945e-05,
+      "loss": 0.4697,
+      "step": 1223
+    },
+    {
+      "epoch": 0.9213398569815582,
+      "grad_norm": 0.28022071719169617,
+      "learning_rate": 4.9358189813727125e-05,
+      "loss": 0.5245,
+      "step": 1224
+    },
+    {
+      "epoch": 0.922092585622883,
+      "grad_norm": 0.22791215777397156,
+      "learning_rate": 4.9356424139636024e-05,
+      "loss": 0.5776,
+      "step": 1225
+    },
+    {
+      "epoch": 0.9228453142642078,
+      "grad_norm": 0.2593214511871338,
+      "learning_rate": 4.9354656075359244e-05,
+      "loss": 0.4691,
+      "step": 1226
+    },
+    {
+      "epoch": 0.9235980429055326,
+      "grad_norm": 0.2540011703968048,
+      "learning_rate": 4.935288562109013e-05,
+      "loss": 0.4755,
+      "step": 1227
+    },
+    {
+      "epoch": 0.9243507715468574,
+      "grad_norm": 0.23081199824810028,
+      "learning_rate": 4.93511127770223e-05,
+      "loss": 0.5214,
+      "step": 1228
+    },
+    {
+      "epoch": 0.9251035001881822,
+      "grad_norm": 0.24854926764965057,
+      "learning_rate": 4.934933754334963e-05,
+      "loss": 0.5718,
+      "step": 1229
+    },
+    {
+      "epoch": 0.9258562288295069,
+      "grad_norm": 0.2791685461997986,
+      "learning_rate": 4.934755992026626e-05,
+      "loss": 0.4839,
+      "step": 1230
+    },
+    {
+      "epoch": 0.9266089574708317,
+      "grad_norm": 0.2338806688785553,
+      "learning_rate": 4.934577990796659e-05,
+      "loss": 0.5037,
+      "step": 1231
+    },
+    {
+      "epoch": 0.9273616861121565,
+      "grad_norm": 0.21399447321891785,
+      "learning_rate": 4.9343997506645255e-05,
+      "loss": 0.4861,
+      "step": 1232
+    },
+    {
+      "epoch": 0.9281144147534813,
+      "grad_norm": 0.25547200441360474,
+      "learning_rate": 4.934221271649721e-05,
+      "loss": 0.4201,
+      "step": 1233
+    },
+    {
+      "epoch": 0.9288671433948061,
+      "grad_norm": 0.23391731083393097,
+      "learning_rate": 4.934042553771762e-05,
+      "loss": 0.4765,
+      "step": 1234
+    },
+    {
+      "epoch": 0.929619872036131,
+      "grad_norm": 0.26428666710853577,
+      "learning_rate": 4.933863597050192e-05,
+      "loss": 0.4388,
+      "step": 1235
+    },
+    {
+      "epoch": 0.9303726006774558,
+      "grad_norm": 0.25605109333992004,
+      "learning_rate": 4.933684401504583e-05,
+      "loss": 0.5103,
+      "step": 1236
+    },
+    {
+      "epoch": 0.9311253293187806,
+      "grad_norm": 0.3222599923610687,
+      "learning_rate": 4.9335049671545314e-05,
+      "loss": 0.5468,
+      "step": 1237
+    },
+    {
+      "epoch": 0.9318780579601054,
+      "grad_norm": 0.28262409567832947,
+      "learning_rate": 4.933325294019659e-05,
+      "loss": 0.4484,
+      "step": 1238
+    },
+    {
+      "epoch": 0.9326307866014302,
+      "grad_norm": 0.22664524614810944,
+      "learning_rate": 4.933145382119615e-05,
+      "loss": 0.4669,
+      "step": 1239
+    },
+    {
+      "epoch": 0.933383515242755,
+      "grad_norm": 0.28448450565338135,
+      "learning_rate": 4.932965231474075e-05,
+      "loss": 0.3635,
+      "step": 1240
+    },
+    {
+      "epoch": 0.9341362438840798,
+      "grad_norm": 0.31614425778388977,
+      "learning_rate": 4.9327848421027386e-05,
+      "loss": 0.5578,
+      "step": 1241
+    },
+    {
+      "epoch": 0.9348889725254046,
+      "grad_norm": 0.23316273093223572,
+      "learning_rate": 4.932604214025334e-05,
+      "loss": 0.4997,
+      "step": 1242
+    },
+    {
+      "epoch": 0.9356417011667294,
+      "grad_norm": 0.3959042429924011,
+      "learning_rate": 4.9324233472616136e-05,
+      "loss": 0.4748,
+      "step": 1243
+    },
+    {
+      "epoch": 0.9363944298080542,
+      "grad_norm": 0.436623752117157,
+      "learning_rate": 4.932242241831357e-05,
+      "loss": 0.5147,
+      "step": 1244
+    },
+    {
+      "epoch": 0.937147158449379,
+      "grad_norm": 0.2841193675994873,
+      "learning_rate": 4.932060897754371e-05,
+      "loss": 0.4758,
+      "step": 1245
+    },
+    {
+      "epoch": 0.9378998870907038,
+      "grad_norm": 0.42983853816986084,
+      "learning_rate": 4.9318793150504846e-05,
+      "loss": 0.6139,
+      "step": 1246
+    },
+    {
+      "epoch": 0.9386526157320286,
+      "grad_norm": 0.3630640208721161,
+      "learning_rate": 4.931697493739557e-05,
+      "loss": 0.4981,
+      "step": 1247
+    },
+    {
+      "epoch": 0.9394053443733534,
+      "grad_norm": 0.2806040346622467,
+      "learning_rate": 4.931515433841472e-05,
+      "loss": 0.534,
+      "step": 1248
+    },
+    {
+      "epoch": 0.9401580730146782,
+      "grad_norm": 0.40307143330574036,
+      "learning_rate": 4.9313331353761385e-05,
+      "loss": 0.4633,
+      "step": 1249
+    },
+    {
+      "epoch": 0.940910801656003,
+      "grad_norm": 0.3985889256000519,
+      "learning_rate": 4.931150598363494e-05,
+      "loss": 0.396,
+      "step": 1250
+    },
+    {
+      "epoch": 0.9416635302973279,
+      "grad_norm": 0.21852152049541473,
+      "learning_rate": 4.9309678228234986e-05,
+      "loss": 0.5487,
+      "step": 1251
+    },
+    {
+      "epoch": 0.9424162589386527,
+      "grad_norm": 0.371802419424057,
+      "learning_rate": 4.93078480877614e-05,
+      "loss": 0.5429,
+      "step": 1252
+    },
+    {
+      "epoch": 0.9431689875799775,
+      "grad_norm": 0.2886806130409241,
+      "learning_rate": 4.930601556241435e-05,
+      "loss": 0.5804,
+      "step": 1253
+    },
+    {
+      "epoch": 0.9439217162213023,
+      "grad_norm": 0.36343035101890564,
+      "learning_rate": 4.930418065239422e-05,
+      "loss": 0.4949,
+      "step": 1254
+    },
+    {
+      "epoch": 0.944674444862627,
+      "grad_norm": 0.34280601143836975,
+      "learning_rate": 4.930234335790168e-05,
+      "loss": 0.5746,
+      "step": 1255
+    },
+    {
+      "epoch": 0.9454271735039518,
+      "grad_norm": 0.26453283429145813,
+      "learning_rate": 4.930050367913764e-05,
+      "loss": 0.3678,
+      "step": 1256
+    },
+    {
+      "epoch": 0.9461799021452766,
+      "grad_norm": 0.27096667885780334,
+      "learning_rate": 4.92986616163033e-05,
+      "loss": 0.5049,
+      "step": 1257
+    },
+    {
+      "epoch": 0.9469326307866014,
+      "grad_norm": 0.38201063871383667,
+      "learning_rate": 4.9296817169600105e-05,
+      "loss": 0.4735,
+      "step": 1258
+    },
+    {
+      "epoch": 0.9476853594279262,
+      "grad_norm": 0.251199334859848,
+      "learning_rate": 4.929497033922975e-05,
+      "loss": 0.4376,
+      "step": 1259
+    },
+    {
+      "epoch": 0.948438088069251,
+      "grad_norm": 0.3646220862865448,
+      "learning_rate": 4.9293121125394203e-05,
+      "loss": 0.6256,
+      "step": 1260
+    },
+    {
+      "epoch": 0.9491908167105758,
+      "grad_norm": 0.32793837785720825,
+      "learning_rate": 4.92912695282957e-05,
+      "loss": 0.4952,
+      "step": 1261
+    },
+    {
+      "epoch": 0.9499435453519006,
+      "grad_norm": 0.3636588454246521,
+      "learning_rate": 4.928941554813672e-05,
+      "loss": 0.5241,
+      "step": 1262
+    },
+    {
+      "epoch": 0.9506962739932254,
+      "grad_norm": 0.3305578827857971,
+      "learning_rate": 4.928755918512001e-05,
+      "loss": 0.5006,
+      "step": 1263
+    },
+    {
+      "epoch": 0.9514490026345502,
+      "grad_norm": 0.2502374053001404,
+      "learning_rate": 4.9285700439448596e-05,
+      "loss": 0.4426,
+      "step": 1264
+    },
+    {
+      "epoch": 0.952201731275875,
+      "grad_norm": 0.32213905453681946,
+      "learning_rate": 4.928383931132574e-05,
+      "loss": 0.4069,
+      "step": 1265
+    },
+    {
+      "epoch": 0.9529544599171998,
+      "grad_norm": 0.3596147894859314,
+      "learning_rate": 4.9281975800954956e-05,
+      "loss": 0.4894,
+      "step": 1266
+    },
+    {
+      "epoch": 0.9537071885585247,
+      "grad_norm": 0.28390005230903625,
+      "learning_rate": 4.928010990854005e-05,
+      "loss": 0.4496,
+      "step": 1267
+    },
+    {
+      "epoch": 0.9544599171998495,
+      "grad_norm": 0.2921454608440399,
+      "learning_rate": 4.9278241634285075e-05,
+      "loss": 0.5233,
+      "step": 1268
+    },
+    {
+      "epoch": 0.9552126458411743,
+      "grad_norm": 0.3605840504169464,
+      "learning_rate": 4.927637097839434e-05,
+      "loss": 0.5955,
+      "step": 1269
+    },
+    {
+      "epoch": 0.9559653744824991,
+      "grad_norm": 0.2959432601928711,
+      "learning_rate": 4.9274497941072404e-05,
+      "loss": 0.5789,
+      "step": 1270
+    },
+    {
+      "epoch": 0.9567181031238239,
+      "grad_norm": 0.30519676208496094,
+      "learning_rate": 4.927262252252412e-05,
+      "loss": 0.497,
+      "step": 1271
+    },
+    {
+      "epoch": 0.9574708317651487,
+      "grad_norm": 0.38935577869415283,
+      "learning_rate": 4.927074472295457e-05,
+      "loss": 0.504,
+      "step": 1272
+    },
+    {
+      "epoch": 0.9582235604064735,
+      "grad_norm": 0.3018273711204529,
+      "learning_rate": 4.926886454256911e-05,
+      "loss": 0.5424,
+      "step": 1273
+    },
+    {
+      "epoch": 0.9589762890477983,
+      "grad_norm": 0.37344890832901,
+      "learning_rate": 4.9266981981573346e-05,
+      "loss": 0.4241,
+      "step": 1274
+    },
+    {
+      "epoch": 0.9597290176891231,
+      "grad_norm": 0.4833126664161682,
+      "learning_rate": 4.926509704017316e-05,
+      "loss": 0.497,
+      "step": 1275
+    },
+    {
+      "epoch": 0.9604817463304479,
+      "grad_norm": 0.25947728753089905,
+      "learning_rate": 4.92632097185747e-05,
+      "loss": 0.5148,
+      "step": 1276
+    },
+    {
+      "epoch": 0.9612344749717727,
+      "grad_norm": 0.29501816630363464,
+      "learning_rate": 4.926132001698433e-05,
+      "loss": 0.569,
+      "step": 1277
+    },
+    {
+      "epoch": 0.9619872036130975,
+      "grad_norm": 0.27252379059791565,
+      "learning_rate": 4.925942793560872e-05,
+      "loss": 0.4907,
+      "step": 1278
+    },
+    {
+      "epoch": 0.9627399322544222,
+      "grad_norm": 0.3257346749305725,
+      "learning_rate": 4.925753347465479e-05,
+      "loss": 0.4397,
+      "step": 1279
+    },
+    {
+      "epoch": 0.963492660895747,
+      "grad_norm": 0.26032552123069763,
+      "learning_rate": 4.925563663432971e-05,
+      "loss": 0.5338,
+      "step": 1280
+    },
+    {
+      "epoch": 0.9642453895370718,
+      "grad_norm": 0.3162387013435364,
+      "learning_rate": 4.925373741484092e-05,
+      "loss": 0.4975,
+      "step": 1281
+    },
+    {
+      "epoch": 0.9649981181783966,
+      "grad_norm": 0.27671578526496887,
+      "learning_rate": 4.925183581639611e-05,
+      "loss": 0.537,
+      "step": 1282
+    },
+    {
+      "epoch": 0.9657508468197215,
+      "grad_norm": 0.3222680985927582,
+      "learning_rate": 4.924993183920324e-05,
+      "loss": 0.5186,
+      "step": 1283
+    },
+    {
+      "epoch": 0.9665035754610463,
+      "grad_norm": 0.2674538195133209,
+      "learning_rate": 4.9248025483470525e-05,
+      "loss": 0.5212,
+      "step": 1284
+    },
+    {
+      "epoch": 0.9672563041023711,
+      "grad_norm": 0.29395246505737305,
+      "learning_rate": 4.9246116749406444e-05,
+      "loss": 0.4985,
+      "step": 1285
+    },
+    {
+      "epoch": 0.9680090327436959,
+      "grad_norm": 0.3178046643733978,
+      "learning_rate": 4.924420563721972e-05,
+      "loss": 0.4628,
+      "step": 1286
+    },
+    {
+      "epoch": 0.9687617613850207,
+      "grad_norm": 0.3250569999217987,
+      "learning_rate": 4.924229214711937e-05,
+      "loss": 0.4844,
+      "step": 1287
+    },
+    {
+      "epoch": 0.9695144900263455,
+      "grad_norm": 0.22645941376686096,
+      "learning_rate": 4.9240376279314635e-05,
+      "loss": 0.4901,
+      "step": 1288
+    },
+    {
+      "epoch": 0.9702672186676703,
+      "grad_norm": 0.2682740390300751,
+      "learning_rate": 4.923845803401503e-05,
+      "loss": 0.4747,
+      "step": 1289
+    },
+    {
+      "epoch": 0.9710199473089951,
+      "grad_norm": 0.23335039615631104,
+      "learning_rate": 4.923653741143035e-05,
+      "loss": 0.5856,
+      "step": 1290
+    },
+    {
+      "epoch": 0.9717726759503199,
+      "grad_norm": 0.4143815040588379,
+      "learning_rate": 4.9234614411770616e-05,
+      "loss": 0.4821,
+      "step": 1291
+    },
+    {
+      "epoch": 0.9725254045916447,
+      "grad_norm": 0.2625824809074402,
+      "learning_rate": 4.9232689035246115e-05,
+      "loss": 0.5408,
+      "step": 1292
+    },
+    {
+      "epoch": 0.9732781332329695,
+      "grad_norm": 0.3109580874443054,
+      "learning_rate": 4.923076128206743e-05,
+      "loss": 0.5861,
+      "step": 1293
+    },
+    {
+      "epoch": 0.9740308618742943,
+      "grad_norm": 0.2838791310787201,
+      "learning_rate": 4.9228831152445355e-05,
+      "loss": 0.536,
+      "step": 1294
+    },
+    {
+      "epoch": 0.9747835905156191,
+      "grad_norm": 0.31006261706352234,
+      "learning_rate": 4.922689864659098e-05,
+      "loss": 0.4342,
+      "step": 1295
+    },
+    {
+      "epoch": 0.9755363191569439,
+      "grad_norm": 0.26192715764045715,
+      "learning_rate": 4.9224963764715625e-05,
+      "loss": 0.5546,
+      "step": 1296
+    },
+    {
+      "epoch": 0.9762890477982687,
+      "grad_norm": 0.3110114634037018,
+      "learning_rate": 4.9223026507030904e-05,
+      "loss": 0.4711,
+      "step": 1297
+    },
+    {
+      "epoch": 0.9770417764395936,
+      "grad_norm": 0.2558371126651764,
+      "learning_rate": 4.922108687374866e-05,
+      "loss": 0.5403,
+      "step": 1298
+    },
+    {
+      "epoch": 0.9777945050809184,
+      "grad_norm": 0.39427313208580017,
+      "learning_rate": 4.921914486508102e-05,
+      "loss": 0.446,
+      "step": 1299
+    },
+    {
+      "epoch": 0.9785472337222432,
+      "grad_norm": 0.2422974556684494,
+      "learning_rate": 4.921720048124035e-05,
+      "loss": 0.4343,
+      "step": 1300
+    },
+    {
+      "epoch": 0.979299962363568,
+      "grad_norm": 0.31024158000946045,
+      "learning_rate": 4.921525372243929e-05,
+      "loss": 0.4916,
+      "step": 1301
+    },
+    {
+      "epoch": 0.9800526910048928,
+      "grad_norm": 0.25948041677474976,
+      "learning_rate": 4.921330458889072e-05,
+      "loss": 0.3963,
+      "step": 1302
+    },
+    {
+      "epoch": 0.9808054196462176,
+      "grad_norm": 0.30222320556640625,
+      "learning_rate": 4.921135308080781e-05,
+      "loss": 0.5139,
+      "step": 1303
+    },
+    {
+      "epoch": 0.9815581482875423,
+      "grad_norm": 0.3028087913990021,
+      "learning_rate": 4.920939919840397e-05,
+      "loss": 0.3916,
+      "step": 1304
+    },
+    {
+      "epoch": 0.9823108769288671,
+      "grad_norm": 0.28354349732398987,
+      "learning_rate": 4.920744294189287e-05,
+      "loss": 0.4075,
+      "step": 1305
+    },
+    {
+      "epoch": 0.9830636055701919,
+      "grad_norm": 0.26186737418174744,
+      "learning_rate": 4.9205484311488456e-05,
+      "loss": 0.4974,
+      "step": 1306
+    },
+    {
+      "epoch": 0.9838163342115167,
+      "grad_norm": 0.2711584270000458,
+      "learning_rate": 4.9203523307404906e-05,
+      "loss": 0.4497,
+      "step": 1307
+    },
+    {
+      "epoch": 0.9845690628528415,
+      "grad_norm": 0.27255159616470337,
+      "learning_rate": 4.920155992985667e-05,
+      "loss": 0.4962,
+      "step": 1308
+    },
+    {
+      "epoch": 0.9853217914941663,
+      "grad_norm": 0.25420933961868286,
+      "learning_rate": 4.9199594179058475e-05,
+      "loss": 0.6059,
+      "step": 1309
+    },
+    {
+      "epoch": 0.9860745201354911,
+      "grad_norm": 0.3226134777069092,
+      "learning_rate": 4.919762605522529e-05,
+      "loss": 0.508,
+      "step": 1310
+    },
+    {
+      "epoch": 0.9868272487768159,
+      "grad_norm": 0.28329384326934814,
+      "learning_rate": 4.919565555857234e-05,
+      "loss": 0.5107,
+      "step": 1311
+    },
+    {
+      "epoch": 0.9875799774181407,
+      "grad_norm": 0.2727423310279846,
+      "learning_rate": 4.9193682689315115e-05,
+      "loss": 0.4806,
+      "step": 1312
+    },
+    {
+      "epoch": 0.9883327060594655,
+      "grad_norm": 0.24810165166854858,
+      "learning_rate": 4.9191707447669365e-05,
+      "loss": 0.423,
+      "step": 1313
+    },
+    {
+      "epoch": 0.9890854347007904,
+      "grad_norm": 0.27025067806243896,
+      "learning_rate": 4.91897298338511e-05,
+      "loss": 0.5735,
+      "step": 1314
+    },
+    {
+      "epoch": 0.9898381633421152,
+      "grad_norm": 0.2512607276439667,
+      "learning_rate": 4.918774984807659e-05,
+      "loss": 0.4028,
+      "step": 1315
+    },
+    {
+      "epoch": 0.99059089198344,
+      "grad_norm": 0.24874146282672882,
+      "learning_rate": 4.918576749056237e-05,
+      "loss": 0.5089,
+      "step": 1316
+    },
+    {
+      "epoch": 0.9913436206247648,
+      "grad_norm": 0.2295101284980774,
+      "learning_rate": 4.918378276152522e-05,
+      "loss": 0.4815,
+      "step": 1317
+    },
+    {
+      "epoch": 0.9920963492660896,
+      "grad_norm": 0.26430410146713257,
+      "learning_rate": 4.918179566118219e-05,
+      "loss": 0.4247,
+      "step": 1318
+    },
+    {
+      "epoch": 0.9928490779074144,
+      "grad_norm": 0.2531023323535919,
+      "learning_rate": 4.9179806189750575e-05,
+      "loss": 0.5375,
+      "step": 1319
+    },
+    {
+      "epoch": 0.9936018065487392,
+      "grad_norm": 0.32334452867507935,
+      "learning_rate": 4.917781434744796e-05,
+      "loss": 0.4514,
+      "step": 1320
+    },
+    {
+      "epoch": 0.994354535190064,
+      "grad_norm": 0.4096810519695282,
+      "learning_rate": 4.9175820134492155e-05,
+      "loss": 0.5072,
+      "step": 1321
+    },
+    {
+      "epoch": 0.9951072638313888,
+      "grad_norm": 0.3252415060997009,
+      "learning_rate": 4.9173823551101255e-05,
+      "loss": 0.488,
+      "step": 1322
+    },
+    {
+      "epoch": 0.9958599924727136,
+      "grad_norm": 0.23065952956676483,
+      "learning_rate": 4.917182459749359e-05,
+      "loss": 0.3715,
+      "step": 1323
+    },
+    {
+      "epoch": 0.9966127211140384,
+      "grad_norm": 0.29444420337677,
+      "learning_rate": 4.916982327388778e-05,
+      "loss": 0.4189,
+      "step": 1324
+    },
+    {
+      "epoch": 0.9973654497553632,
+      "grad_norm": 0.2787327170372009,
+      "learning_rate": 4.9167819580502666e-05,
+      "loss": 0.3405,
+      "step": 1325
+    },
+    {
+      "epoch": 0.998118178396688,
+      "grad_norm": 0.26253893971443176,
+      "learning_rate": 4.9165813517557377e-05,
+      "loss": 0.4271,
+      "step": 1326
+    },
+    {
+      "epoch": 0.9988709070380128,
+      "grad_norm": 0.2670835256576538,
+      "learning_rate": 4.916380508527131e-05,
+      "loss": 0.538,
+      "step": 1327
+    },
+    {
+      "epoch": 0.9996236356793377,
+      "grad_norm": 0.27934324741363525,
+      "learning_rate": 4.916179428386408e-05,
+      "loss": 0.498,
+      "step": 1328
+    },
+    {
+      "epoch": 1.0003763643206625,
+      "grad_norm": 0.2735024392604828,
+      "learning_rate": 4.91597811135556e-05,
+      "loss": 0.4199,
+      "step": 1329
+    },
+    {
+      "epoch": 1.0011290929619872,
+      "grad_norm": 0.27295148372650146,
+      "learning_rate": 4.915776557456602e-05,
+      "loss": 0.4581,
+      "step": 1330
+    },
+    {
+      "epoch": 1.001881821603312,
+      "grad_norm": 0.25653672218322754,
+      "learning_rate": 4.915574766711576e-05,
+      "loss": 0.4748,
+      "step": 1331
+    },
+    {
+      "epoch": 1.0026345502446368,
+      "grad_norm": 0.3309451639652252,
+      "learning_rate": 4.915372739142549e-05,
+      "loss": 0.4336,
+      "step": 1332
+    },
+    {
+      "epoch": 1.0033872788859617,
+      "grad_norm": 0.3333767354488373,
+      "learning_rate": 4.915170474771615e-05,
+      "loss": 0.4854,
+      "step": 1333
+    },
+    {
+      "epoch": 1.0041400075272864,
+      "grad_norm": 0.26879942417144775,
+      "learning_rate": 4.914967973620893e-05,
+      "loss": 0.5326,
+      "step": 1334
+    },
+    {
+      "epoch": 1.0048927361686113,
+      "grad_norm": 0.3158165514469147,
+      "learning_rate": 4.91476523571253e-05,
+      "loss": 0.5141,
+      "step": 1335
+    },
+    {
+      "epoch": 1.005645464809936,
+      "grad_norm": 0.33691859245300293,
+      "learning_rate": 4.914562261068693e-05,
+      "loss": 0.4754,
+      "step": 1336
+    },
+    {
+      "epoch": 1.006398193451261,
+      "grad_norm": 0.26826757192611694,
+      "learning_rate": 4.914359049711583e-05,
+      "loss": 0.4807,
+      "step": 1337
+    },
+    {
+      "epoch": 1.0071509220925856,
+      "grad_norm": 0.357645720243454,
+      "learning_rate": 4.914155601663421e-05,
+      "loss": 0.5445,
+      "step": 1338
+    },
+    {
+      "epoch": 1.0079036507339105,
+      "grad_norm": 0.32207003235816956,
+      "learning_rate": 4.9139519169464565e-05,
+      "loss": 0.5097,
+      "step": 1339
+    },
+    {
+      "epoch": 1.0086563793752352,
+      "grad_norm": 0.2776307463645935,
+      "learning_rate": 4.913747995582964e-05,
+      "loss": 0.4886,
+      "step": 1340
+    },
+    {
+      "epoch": 1.00940910801656,
+      "grad_norm": 0.2769996225833893,
+      "learning_rate": 4.913543837595243e-05,
+      "loss": 0.5346,
+      "step": 1341
+    },
+    {
+      "epoch": 1.0101618366578848,
+      "grad_norm": 0.3009834587574005,
+      "learning_rate": 4.9133394430056215e-05,
+      "loss": 0.5442,
+      "step": 1342
+    },
+    {
+      "epoch": 1.0109145652992095,
+      "grad_norm": 0.3116737902164459,
+      "learning_rate": 4.9131348118364506e-05,
+      "loss": 0.555,
+      "step": 1343
+    },
+    {
+      "epoch": 1.0116672939405345,
+      "grad_norm": 0.3650636672973633,
+      "learning_rate": 4.912929944110108e-05,
+      "loss": 0.4616,
+      "step": 1344
+    },
+    {
+      "epoch": 1.0124200225818591,
+      "grad_norm": 0.446410596370697,
+      "learning_rate": 4.912724839849001e-05,
+      "loss": 0.3788,
+      "step": 1345
+    },
+    {
+      "epoch": 1.013172751223184,
+      "grad_norm": 0.29041045904159546,
+      "learning_rate": 4.912519499075555e-05,
+      "loss": 0.4874,
+      "step": 1346
+    },
+    {
+      "epoch": 1.0139254798645088,
+      "grad_norm": 0.29828765988349915,
+      "learning_rate": 4.912313921812229e-05,
+      "loss": 0.4856,
+      "step": 1347
+    },
+    {
+      "epoch": 1.0146782085058337,
+      "grad_norm": 0.38760906457901,
+      "learning_rate": 4.912108108081503e-05,
+      "loss": 0.5086,
+      "step": 1348
+    },
+    {
+      "epoch": 1.0154309371471584,
+      "grad_norm": 0.2524135112762451,
+      "learning_rate": 4.911902057905886e-05,
+      "loss": 0.4886,
+      "step": 1349
+    },
+    {
+      "epoch": 1.0161836657884833,
+      "grad_norm": 0.35103103518486023,
+      "learning_rate": 4.9116957713079084e-05,
+      "loss": 0.5261,
+      "step": 1350
+    },
+    {
+      "epoch": 1.016936394429808,
+      "grad_norm": 0.3308658301830292,
+      "learning_rate": 4.911489248310133e-05,
+      "loss": 0.5136,
+      "step": 1351
+    },
+    {
+      "epoch": 1.017689123071133,
+      "grad_norm": 0.34888026118278503,
+      "learning_rate": 4.911282488935142e-05,
+      "loss": 0.4579,
+      "step": 1352
+    },
+    {
+      "epoch": 1.0184418517124576,
+      "grad_norm": 0.28628939390182495,
+      "learning_rate": 4.9110754932055484e-05,
+      "loss": 0.3889,
+      "step": 1353
+    },
+    {
+      "epoch": 1.0191945803537825,
+      "grad_norm": 0.3951142728328705,
+      "learning_rate": 4.910868261143987e-05,
+      "loss": 0.4446,
+      "step": 1354
+    },
+    {
+      "epoch": 1.0199473089951072,
+      "grad_norm": 0.35897552967071533,
+      "learning_rate": 4.910660792773122e-05,
+      "loss": 0.5493,
+      "step": 1355
+    },
+    {
+      "epoch": 1.0207000376364321,
+      "grad_norm": 0.2645096778869629,
+      "learning_rate": 4.9104530881156414e-05,
+      "loss": 0.4408,
+      "step": 1356
+    },
+    {
+      "epoch": 1.0214527662777568,
+      "grad_norm": 0.2965589165687561,
+      "learning_rate": 4.9102451471942586e-05,
+      "loss": 0.5378,
+      "step": 1357
+    },
+    {
+      "epoch": 1.0222054949190817,
+      "grad_norm": 0.24591881036758423,
+      "learning_rate": 4.910036970031715e-05,
+      "loss": 0.4595,
+      "step": 1358
+    },
+    {
+      "epoch": 1.0229582235604064,
+      "grad_norm": 0.31659215688705444,
+      "learning_rate": 4.9098285566507754e-05,
+      "loss": 0.4742,
+      "step": 1359
+    },
+    {
+      "epoch": 1.0237109522017314,
+      "grad_norm": 0.2579328715801239,
+      "learning_rate": 4.9096199070742325e-05,
+      "loss": 0.4596,
+      "step": 1360
+    },
+    {
+      "epoch": 1.024463680843056,
+      "grad_norm": 0.2878738045692444,
+      "learning_rate": 4.909411021324904e-05,
+      "loss": 0.4908,
+      "step": 1361
+    },
+    {
+      "epoch": 1.025216409484381,
+      "grad_norm": 0.26085877418518066,
+      "learning_rate": 4.909201899425632e-05,
+      "loss": 0.4761,
+      "step": 1362
+    },
+    {
+      "epoch": 1.0259691381257057,
+      "grad_norm": 0.31090879440307617,
+      "learning_rate": 4.908992541399287e-05,
+      "loss": 0.4745,
+      "step": 1363
+    },
+    {
+      "epoch": 1.0267218667670306,
+      "grad_norm": 0.2627177834510803,
+      "learning_rate": 4.908782947268763e-05,
+      "loss": 0.4379,
+      "step": 1364
+    },
+    {
+      "epoch": 1.0274745954083553,
+      "grad_norm": 0.2535850405693054,
+      "learning_rate": 4.908573117056982e-05,
+      "loss": 0.5431,
+      "step": 1365
+    },
+    {
+      "epoch": 1.0282273240496802,
+      "grad_norm": 0.20867329835891724,
+      "learning_rate": 4.908363050786891e-05,
+      "loss": 0.4741,
+      "step": 1366
+    },
+    {
+      "epoch": 1.028980052691005,
+      "grad_norm": 0.252157986164093,
+      "learning_rate": 4.9081527484814616e-05,
+      "loss": 0.4788,
+      "step": 1367
+    },
+    {
+      "epoch": 1.0297327813323296,
+      "grad_norm": 0.23520304262638092,
+      "learning_rate": 4.9079422101636916e-05,
+      "loss": 0.3808,
+      "step": 1368
+    },
+    {
+      "epoch": 1.0304855099736545,
+      "grad_norm": 0.2522774338722229,
+      "learning_rate": 4.907731435856606e-05,
+      "loss": 0.5253,
+      "step": 1369
+    },
+    {
+      "epoch": 1.0312382386149792,
+      "grad_norm": 0.27403345704078674,
+      "learning_rate": 4.907520425583255e-05,
+      "loss": 0.4107,
+      "step": 1370
+    },
+    {
+      "epoch": 1.0319909672563041,
+      "grad_norm": 0.22261206805706024,
+      "learning_rate": 4.907309179366714e-05,
+      "loss": 0.429,
+      "step": 1371
+    },
+    {
+      "epoch": 1.0327436958976288,
+      "grad_norm": 0.3020976185798645,
+      "learning_rate": 4.907097697230084e-05,
+      "loss": 0.4007,
+      "step": 1372
+    },
+    {
+      "epoch": 1.0334964245389537,
+      "grad_norm": 0.2950030565261841,
+      "learning_rate": 4.9068859791964945e-05,
+      "loss": 0.4413,
+      "step": 1373
+    },
+    {
+      "epoch": 1.0342491531802784,
+      "grad_norm": 0.2885960340499878,
+      "learning_rate": 4.9066740252890955e-05,
+      "loss": 0.5628,
+      "step": 1374
+    },
+    {
+      "epoch": 1.0350018818216034,
+      "grad_norm": 0.26749464869499207,
+      "learning_rate": 4.906461835531069e-05,
+      "loss": 0.4726,
+      "step": 1375
+    },
+    {
+      "epoch": 1.035754610462928,
+      "grad_norm": 0.23322290182113647,
+      "learning_rate": 4.906249409945617e-05,
+      "loss": 0.4533,
+      "step": 1376
+    },
+    {
+      "epoch": 1.036507339104253,
+      "grad_norm": 0.26510295271873474,
+      "learning_rate": 4.9060367485559725e-05,
+      "loss": 0.5191,
+      "step": 1377
+    },
+    {
+      "epoch": 1.0372600677455777,
+      "grad_norm": 0.3355228900909424,
+      "learning_rate": 4.905823851385391e-05,
+      "loss": 0.5743,
+      "step": 1378
+    },
+    {
+      "epoch": 1.0380127963869026,
+      "grad_norm": 0.27616459131240845,
+      "learning_rate": 4.9056107184571534e-05,
+      "loss": 0.4572,
+      "step": 1379
+    },
+    {
+      "epoch": 1.0387655250282273,
+      "grad_norm": 0.28265175223350525,
+      "learning_rate": 4.905397349794568e-05,
+      "loss": 0.4632,
+      "step": 1380
+    },
+    {
+      "epoch": 1.0395182536695522,
+      "grad_norm": 0.44230809807777405,
+      "learning_rate": 4.90518374542097e-05,
+      "loss": 0.4691,
+      "step": 1381
+    },
+    {
+      "epoch": 1.040270982310877,
+      "grad_norm": 0.2739470303058624,
+      "learning_rate": 4.904969905359718e-05,
+      "loss": 0.4685,
+      "step": 1382
+    },
+    {
+      "epoch": 1.0410237109522018,
+      "grad_norm": 0.3265460133552551,
+      "learning_rate": 4.904755829634197e-05,
+      "loss": 0.4573,
+      "step": 1383
+    },
+    {
+      "epoch": 1.0417764395935265,
+      "grad_norm": 0.2918044924736023,
+      "learning_rate": 4.904541518267818e-05,
+      "loss": 0.461,
+      "step": 1384
+    },
+    {
+      "epoch": 1.0425291682348514,
+      "grad_norm": 0.29349851608276367,
+      "learning_rate": 4.9043269712840174e-05,
+      "loss": 0.4383,
+      "step": 1385
+    },
+    {
+      "epoch": 1.0432818968761761,
+      "grad_norm": 0.2400587499141693,
+      "learning_rate": 4.9041121887062584e-05,
+      "loss": 0.5239,
+      "step": 1386
+    },
+    {
+      "epoch": 1.044034625517501,
+      "grad_norm": 0.3218742907047272,
+      "learning_rate": 4.903897170558029e-05,
+      "loss": 0.4614,
+      "step": 1387
+    },
+    {
+      "epoch": 1.0447873541588257,
+      "grad_norm": 0.26169392466545105,
+      "learning_rate": 4.903681916862844e-05,
+      "loss": 0.5751,
+      "step": 1388
+    },
+    {
+      "epoch": 1.0455400828001506,
+      "grad_norm": 0.3074067533016205,
+      "learning_rate": 4.9034664276442424e-05,
+      "loss": 0.4622,
+      "step": 1389
+    },
+    {
+      "epoch": 1.0462928114414753,
+      "grad_norm": 0.3291718363761902,
+      "learning_rate": 4.9032507029257895e-05,
+      "loss": 0.4676,
+      "step": 1390
+    },
+    {
+      "epoch": 1.0470455400828,
+      "grad_norm": 0.31526070833206177,
+      "learning_rate": 4.9030347427310774e-05,
+      "loss": 0.5197,
+      "step": 1391
+    },
+    {
+      "epoch": 1.047798268724125,
+      "grad_norm": 0.2738697826862335,
+      "learning_rate": 4.902818547083723e-05,
+      "loss": 0.4629,
+      "step": 1392
+    },
+    {
+      "epoch": 1.0485509973654497,
+      "grad_norm": 0.28174474835395813,
+      "learning_rate": 4.902602116007369e-05,
+      "loss": 0.4612,
+      "step": 1393
+    },
+    {
+      "epoch": 1.0493037260067746,
+      "grad_norm": 0.31263890862464905,
+      "learning_rate": 4.9023854495256835e-05,
+      "loss": 0.3942,
+      "step": 1394
+    },
+    {
+      "epoch": 1.0500564546480993,
+      "grad_norm": 0.25463953614234924,
+      "learning_rate": 4.902168547662362e-05,
+      "loss": 0.4091,
+      "step": 1395
+    },
+    {
+      "epoch": 1.0508091832894242,
+      "grad_norm": 0.3093293607234955,
+      "learning_rate": 4.9019514104411236e-05,
+      "loss": 0.5213,
+      "step": 1396
+    },
+    {
+      "epoch": 1.0515619119307489,
+      "grad_norm": 0.25776270031929016,
+      "learning_rate": 4.9017340378857134e-05,
+      "loss": 0.3876,
+      "step": 1397
+    },
+    {
+      "epoch": 1.0523146405720738,
+      "grad_norm": 0.2540895342826843,
+      "learning_rate": 4.9015164300199054e-05,
+      "loss": 0.413,
+      "step": 1398
+    },
+    {
+      "epoch": 1.0530673692133985,
+      "grad_norm": 0.2734977900981903,
+      "learning_rate": 4.901298586867494e-05,
+      "loss": 0.461,
+      "step": 1399
+    },
+    {
+      "epoch": 1.0538200978547234,
+      "grad_norm": 0.24747925996780396,
+      "learning_rate": 4.901080508452305e-05,
+      "loss": 0.4681,
+      "step": 1400
+    },
+    {
+      "epoch": 1.0538200978547234,
+      "eval_loss": 0.45823848247528076,
+      "eval_runtime": 452.6552,
+      "eval_samples_per_second": 21.268,
+      "eval_steps_per_second": 0.665,
+      "step": 1400
+    },
+    {
+      "epoch": 1.054572826496048,
+      "grad_norm": 0.24641470611095428,
+      "learning_rate": 4.900862194798185e-05,
+      "loss": 0.374,
+      "step": 1401
+    },
+    {
+      "epoch": 1.055325555137373,
+      "grad_norm": 0.2523903250694275,
+      "learning_rate": 4.9006436459290094e-05,
+      "loss": 0.4789,
+      "step": 1402
+    },
+    {
+      "epoch": 1.0560782837786977,
+      "grad_norm": 0.22377508878707886,
+      "learning_rate": 4.9004248618686775e-05,
+      "loss": 0.4748,
+      "step": 1403
+    },
+    {
+      "epoch": 1.0568310124200226,
+      "grad_norm": 0.2601446807384491,
+      "learning_rate": 4.900205842641116e-05,
+      "loss": 0.4525,
+      "step": 1404
+    },
+    {
+      "epoch": 1.0575837410613473,
+      "grad_norm": 0.25875264406204224,
+      "learning_rate": 4.899986588270277e-05,
+      "loss": 0.512,
+      "step": 1405
+    },
+    {
+      "epoch": 1.0583364697026723,
+      "grad_norm": 0.24411985278129578,
+      "learning_rate": 4.899767098780137e-05,
+      "loss": 0.5138,
+      "step": 1406
+    },
+    {
+      "epoch": 1.059089198343997,
+      "grad_norm": 0.27232104539871216,
+      "learning_rate": 4.8995473741946985e-05,
+      "loss": 0.4898,
+      "step": 1407
+    },
+    {
+      "epoch": 1.0598419269853219,
+      "grad_norm": 0.25845345854759216,
+      "learning_rate": 4.899327414537992e-05,
+      "loss": 0.5534,
+      "step": 1408
+    },
+    {
+      "epoch": 1.0605946556266466,
+      "grad_norm": 0.23812782764434814,
+      "learning_rate": 4.8991072198340696e-05,
+      "loss": 0.5221,
+      "step": 1409
+    },
+    {
+      "epoch": 1.0613473842679715,
+      "grad_norm": 0.2594713270664215,
+      "learning_rate": 4.898886790107014e-05,
+      "loss": 0.4522,
+      "step": 1410
+    },
+    {
+      "epoch": 1.0621001129092962,
+      "grad_norm": 0.2509410083293915,
+      "learning_rate": 4.8986661253809285e-05,
+      "loss": 0.4921,
+      "step": 1411
+    },
+    {
+      "epoch": 1.062852841550621,
+      "grad_norm": 0.24951015412807465,
+      "learning_rate": 4.898445225679947e-05,
+      "loss": 0.4631,
+      "step": 1412
+    },
+    {
+      "epoch": 1.0636055701919458,
+      "grad_norm": 0.31927454471588135,
+      "learning_rate": 4.898224091028225e-05,
+      "loss": 0.409,
+      "step": 1413
+    },
+    {
+      "epoch": 1.0643582988332705,
+      "grad_norm": 0.215189129114151,
+      "learning_rate": 4.898002721449947e-05,
+      "loss": 0.5058,
+      "step": 1414
+    },
+    {
+      "epoch": 1.0651110274745954,
+      "grad_norm": 0.34122318029403687,
+      "learning_rate": 4.897781116969321e-05,
+      "loss": 0.5396,
+      "step": 1415
+    },
+    {
+      "epoch": 1.0658637561159203,
+      "grad_norm": 0.29413262009620667,
+      "learning_rate": 4.897559277610581e-05,
+      "loss": 0.4546,
+      "step": 1416
+    },
+    {
+      "epoch": 1.066616484757245,
+      "grad_norm": 0.25795239210128784,
+      "learning_rate": 4.897337203397986e-05,
+      "loss": 0.5037,
+      "step": 1417
+    },
+    {
+      "epoch": 1.0673692133985697,
+      "grad_norm": 0.3698360025882721,
+      "learning_rate": 4.897114894355824e-05,
+      "loss": 0.5169,
+      "step": 1418
+    },
+    {
+      "epoch": 1.0681219420398946,
+      "grad_norm": 0.20962756872177124,
+      "learning_rate": 4.896892350508406e-05,
+      "loss": 0.4902,
+      "step": 1419
+    },
+    {
+      "epoch": 1.0688746706812193,
+      "grad_norm": 0.34285616874694824,
+      "learning_rate": 4.896669571880067e-05,
+      "loss": 0.4346,
+      "step": 1420
+    },
+    {
+      "epoch": 1.0696273993225442,
+      "grad_norm": 0.3179619610309601,
+      "learning_rate": 4.8964465584951724e-05,
+      "loss": 0.4399,
+      "step": 1421
+    },
+    {
+      "epoch": 1.070380127963869,
+      "grad_norm": 0.2600507438182831,
+      "learning_rate": 4.896223310378108e-05,
+      "loss": 0.397,
+      "step": 1422
+    },
+    {
+      "epoch": 1.0711328566051939,
+      "grad_norm": 0.32125377655029297,
+      "learning_rate": 4.895999827553289e-05,
+      "loss": 0.3879,
+      "step": 1423
+    },
+    {
+      "epoch": 1.0718855852465186,
+      "grad_norm": 0.2736128270626068,
+      "learning_rate": 4.895776110045156e-05,
+      "loss": 0.522,
+      "step": 1424
+    },
+    {
+      "epoch": 1.0726383138878435,
+      "grad_norm": 0.26169565320014954,
+      "learning_rate": 4.8955521578781746e-05,
+      "loss": 0.4076,
+      "step": 1425
+    },
+    {
+      "epoch": 1.0733910425291682,
+      "grad_norm": 0.359401673078537,
+      "learning_rate": 4.8953279710768345e-05,
+      "loss": 0.4705,
+      "step": 1426
+    },
+    {
+      "epoch": 1.074143771170493,
+      "grad_norm": 0.2594001293182373,
+      "learning_rate": 4.895103549665652e-05,
+      "loss": 0.3608,
+      "step": 1427
+    },
+    {
+      "epoch": 1.0748964998118178,
+      "grad_norm": 0.2658233940601349,
+      "learning_rate": 4.894878893669171e-05,
+      "loss": 0.5103,
+      "step": 1428
+    },
+    {
+      "epoch": 1.0756492284531427,
+      "grad_norm": 0.37942439317703247,
+      "learning_rate": 4.894654003111959e-05,
+      "loss": 0.4798,
+      "step": 1429
+    },
+    {
+      "epoch": 1.0764019570944674,
+      "grad_norm": 0.28560012578964233,
+      "learning_rate": 4.89442887801861e-05,
+      "loss": 0.5008,
+      "step": 1430
+    },
+    {
+      "epoch": 1.0771546857357923,
+      "grad_norm": 0.38984936475753784,
+      "learning_rate": 4.894203518413743e-05,
+      "loss": 0.4716,
+      "step": 1431
+    },
+    {
+      "epoch": 1.077907414377117,
+      "grad_norm": 0.4396340847015381,
+      "learning_rate": 4.893977924322002e-05,
+      "loss": 0.5459,
+      "step": 1432
+    },
+    {
+      "epoch": 1.078660143018442,
+      "grad_norm": 0.38536134362220764,
+      "learning_rate": 4.89375209576806e-05,
+      "loss": 0.4763,
+      "step": 1433
+    },
+    {
+      "epoch": 1.0794128716597666,
+      "grad_norm": 0.250325471162796,
+      "learning_rate": 4.8935260327766103e-05,
+      "loss": 0.4068,
+      "step": 1434
+    },
+    {
+      "epoch": 1.0801656003010915,
+      "grad_norm": 0.33363649249076843,
+      "learning_rate": 4.8932997353723783e-05,
+      "loss": 0.4754,
+      "step": 1435
+    },
+    {
+      "epoch": 1.0809183289424162,
+      "grad_norm": 0.28988149762153625,
+      "learning_rate": 4.893073203580108e-05,
+      "loss": 0.534,
+      "step": 1436
+    },
+    {
+      "epoch": 1.0816710575837412,
+      "grad_norm": 0.281730979681015,
+      "learning_rate": 4.892846437424575e-05,
+      "loss": 0.436,
+      "step": 1437
+    },
+    {
+      "epoch": 1.0824237862250659,
+      "grad_norm": 0.3015945255756378,
+      "learning_rate": 4.892619436930577e-05,
+      "loss": 0.533,
+      "step": 1438
+    },
+    {
+      "epoch": 1.0831765148663908,
+      "grad_norm": 0.27621564269065857,
+      "learning_rate": 4.8923922021229384e-05,
+      "loss": 0.4779,
+      "step": 1439
+    },
+    {
+      "epoch": 1.0839292435077155,
+      "grad_norm": 0.3359598219394684,
+      "learning_rate": 4.89216473302651e-05,
+      "loss": 0.4456,
+      "step": 1440
+    },
+    {
+      "epoch": 1.0846819721490402,
+      "grad_norm": 0.29987940192222595,
+      "learning_rate": 4.891937029666168e-05,
+      "loss": 0.5024,
+      "step": 1441
+    },
+    {
+      "epoch": 1.085434700790365,
+      "grad_norm": 0.30538544058799744,
+      "learning_rate": 4.8917090920668116e-05,
+      "loss": 0.4637,
+      "step": 1442
+    },
+    {
+      "epoch": 1.0861874294316898,
+      "grad_norm": 0.2581481635570526,
+      "learning_rate": 4.891480920253369e-05,
+      "loss": 0.4325,
+      "step": 1443
+    },
+    {
+      "epoch": 1.0869401580730147,
+      "grad_norm": 0.273296982049942,
+      "learning_rate": 4.891252514250793e-05,
+      "loss": 0.3676,
+      "step": 1444
+    },
+    {
+      "epoch": 1.0876928867143394,
+      "grad_norm": 0.3785010874271393,
+      "learning_rate": 4.891023874084061e-05,
+      "loss": 0.4593,
+      "step": 1445
+    },
+    {
+      "epoch": 1.0884456153556643,
+      "grad_norm": 0.37302330136299133,
+      "learning_rate": 4.890794999778178e-05,
+      "loss": 0.4953,
+      "step": 1446
+    },
+    {
+      "epoch": 1.089198343996989,
+      "grad_norm": 0.2661270797252655,
+      "learning_rate": 4.890565891358172e-05,
+      "loss": 0.4332,
+      "step": 1447
+    },
+    {
+      "epoch": 1.089951072638314,
+      "grad_norm": 0.3164771795272827,
+      "learning_rate": 4.890336548849098e-05,
+      "loss": 0.4204,
+      "step": 1448
+    },
+    {
+      "epoch": 1.0907038012796386,
+      "grad_norm": 0.3196297585964203,
+      "learning_rate": 4.890106972276037e-05,
+      "loss": 0.5145,
+      "step": 1449
+    },
+    {
+      "epoch": 1.0914565299209635,
+      "grad_norm": 0.32933998107910156,
+      "learning_rate": 4.889877161664096e-05,
+      "loss": 0.4839,
+      "step": 1450
+    },
+    {
+      "epoch": 1.0922092585622882,
+      "grad_norm": 0.2536400854587555,
+      "learning_rate": 4.889647117038405e-05,
+      "loss": 0.4984,
+      "step": 1451
+    },
+    {
+      "epoch": 1.0929619872036132,
+      "grad_norm": 0.336419016122818,
+      "learning_rate": 4.8894168384241226e-05,
+      "loss": 0.4889,
+      "step": 1452
+    },
+    {
+      "epoch": 1.0937147158449378,
+      "grad_norm": 0.25935065746307373,
+      "learning_rate": 4.8891863258464315e-05,
+      "loss": 0.4407,
+      "step": 1453
+    },
+    {
+      "epoch": 1.0944674444862628,
+      "grad_norm": 0.25803494453430176,
+      "learning_rate": 4.888955579330539e-05,
+      "loss": 0.3507,
+      "step": 1454
+    },
+    {
+      "epoch": 1.0952201731275875,
+      "grad_norm": 0.31527355313301086,
+      "learning_rate": 4.888724598901681e-05,
+      "loss": 0.4264,
+      "step": 1455
+    },
+    {
+      "epoch": 1.0959729017689124,
+      "grad_norm": 0.2976307272911072,
+      "learning_rate": 4.888493384585116e-05,
+      "loss": 0.4605,
+      "step": 1456
+    },
+    {
+      "epoch": 1.096725630410237,
+      "grad_norm": 0.305641233921051,
+      "learning_rate": 4.88826193640613e-05,
+      "loss": 0.4186,
+      "step": 1457
+    },
+    {
+      "epoch": 1.097478359051562,
+      "grad_norm": 0.2615106999874115,
+      "learning_rate": 4.888030254390033e-05,
+      "loss": 0.3776,
+      "step": 1458
+    },
+    {
+      "epoch": 1.0982310876928867,
+      "grad_norm": 0.3013743460178375,
+      "learning_rate": 4.887798338562163e-05,
+      "loss": 0.4343,
+      "step": 1459
+    },
+    {
+      "epoch": 1.0989838163342116,
+      "grad_norm": 0.26336807012557983,
+      "learning_rate": 4.8875661889478795e-05,
+      "loss": 0.5651,
+      "step": 1460
+    },
+    {
+      "epoch": 1.0997365449755363,
+      "grad_norm": 0.2756704092025757,
+      "learning_rate": 4.887333805572572e-05,
+      "loss": 0.4831,
+      "step": 1461
+    },
+    {
+      "epoch": 1.1004892736168612,
+      "grad_norm": 0.28474482893943787,
+      "learning_rate": 4.887101188461653e-05,
+      "loss": 0.573,
+      "step": 1462
+    },
+    {
+      "epoch": 1.101242002258186,
+      "grad_norm": 0.25330615043640137,
+      "learning_rate": 4.886868337640561e-05,
+      "loss": 0.4186,
+      "step": 1463
+    },
+    {
+      "epoch": 1.1019947308995106,
+      "grad_norm": 0.34977567195892334,
+      "learning_rate": 4.88663525313476e-05,
+      "loss": 0.4242,
+      "step": 1464
+    },
+    {
+      "epoch": 1.1027474595408355,
+      "grad_norm": 0.2349741905927658,
+      "learning_rate": 4.886401934969741e-05,
+      "loss": 0.4601,
+      "step": 1465
+    },
+    {
+      "epoch": 1.1035001881821604,
+      "grad_norm": 0.23235560953617096,
+      "learning_rate": 4.886168383171017e-05,
+      "loss": 0.4847,
+      "step": 1466
+    },
+    {
+      "epoch": 1.1042529168234851,
+      "grad_norm": 0.23717495799064636,
+      "learning_rate": 4.885934597764132e-05,
+      "loss": 0.3993,
+      "step": 1467
+    },
+    {
+      "epoch": 1.1050056454648098,
+      "grad_norm": 0.2526974678039551,
+      "learning_rate": 4.88570057877465e-05,
+      "loss": 0.3598,
+      "step": 1468
+    },
+    {
+      "epoch": 1.1057583741061348,
+      "grad_norm": 0.2582371234893799,
+      "learning_rate": 4.885466326228163e-05,
+      "loss": 0.5151,
+      "step": 1469
+    },
+    {
+      "epoch": 1.1065111027474595,
+      "grad_norm": 0.20974048972129822,
+      "learning_rate": 4.88523184015029e-05,
+      "loss": 0.4274,
+      "step": 1470
+    },
+    {
+      "epoch": 1.1072638313887844,
+      "grad_norm": 0.2116212248802185,
+      "learning_rate": 4.8849971205666724e-05,
+      "loss": 0.3242,
+      "step": 1471
+    },
+    {
+      "epoch": 1.108016560030109,
+      "grad_norm": 0.25828757882118225,
+      "learning_rate": 4.884762167502981e-05,
+      "loss": 0.4636,
+      "step": 1472
+    },
+    {
+      "epoch": 1.108769288671434,
+      "grad_norm": 0.24669979512691498,
+      "learning_rate": 4.884526980984907e-05,
+      "loss": 0.4914,
+      "step": 1473
+    },
+    {
+      "epoch": 1.1095220173127587,
+      "grad_norm": 0.23985162377357483,
+      "learning_rate": 4.884291561038172e-05,
+      "loss": 0.4111,
+      "step": 1474
+    },
+    {
+      "epoch": 1.1102747459540836,
+      "grad_norm": 0.24433863162994385,
+      "learning_rate": 4.884055907688521e-05,
+      "loss": 0.4009,
+      "step": 1475
+    },
+    {
+      "epoch": 1.1110274745954083,
+      "grad_norm": 0.275066614151001,
+      "learning_rate": 4.8838200209617245e-05,
+      "loss": 0.514,
+      "step": 1476
+    },
+    {
+      "epoch": 1.1117802032367332,
+      "grad_norm": 0.22523608803749084,
+      "learning_rate": 4.883583900883579e-05,
+      "loss": 0.5062,
+      "step": 1477
+    },
+    {
+      "epoch": 1.112532931878058,
+      "grad_norm": 0.28829771280288696,
+      "learning_rate": 4.883347547479906e-05,
+      "loss": 0.4489,
+      "step": 1478
+    },
+    {
+      "epoch": 1.1132856605193828,
+      "grad_norm": 0.27681347727775574,
+      "learning_rate": 4.883110960776552e-05,
+      "loss": 0.4439,
+      "step": 1479
+    },
+    {
+      "epoch": 1.1140383891607075,
+      "grad_norm": 0.2102028876543045,
+      "learning_rate": 4.8828741407993905e-05,
+      "loss": 0.4327,
+      "step": 1480
+    },
+    {
+      "epoch": 1.1147911178020324,
+      "grad_norm": 0.26173853874206543,
+      "learning_rate": 4.882637087574319e-05,
+      "loss": 0.4059,
+      "step": 1481
+    },
+    {
+      "epoch": 1.1155438464433571,
+      "grad_norm": 0.24156397581100464,
+      "learning_rate": 4.882399801127263e-05,
+      "loss": 0.4989,
+      "step": 1482
+    },
+    {
+      "epoch": 1.116296575084682,
+      "grad_norm": 0.2603491246700287,
+      "learning_rate": 4.8821622814841704e-05,
+      "loss": 0.4215,
+      "step": 1483
+    },
+    {
+      "epoch": 1.1170493037260067,
+      "grad_norm": 0.21655279397964478,
+      "learning_rate": 4.8819245286710175e-05,
+      "loss": 0.4692,
+      "step": 1484
+    },
+    {
+      "epoch": 1.1178020323673317,
+      "grad_norm": 0.2575494349002838,
+      "learning_rate": 4.8816865427138015e-05,
+      "loss": 0.4011,
+      "step": 1485
+    },
+    {
+      "epoch": 1.1185547610086564,
+      "grad_norm": 0.3025098145008087,
+      "learning_rate": 4.881448323638551e-05,
+      "loss": 0.434,
+      "step": 1486
+    },
+    {
+      "epoch": 1.1193074896499813,
+      "grad_norm": 0.22964079678058624,
+      "learning_rate": 4.881209871471317e-05,
+      "loss": 0.5159,
+      "step": 1487
+    },
+    {
+      "epoch": 1.120060218291306,
+      "grad_norm": 0.25040963292121887,
+      "learning_rate": 4.880971186238175e-05,
+      "loss": 0.4498,
+      "step": 1488
+    },
+    {
+      "epoch": 1.120812946932631,
+      "grad_norm": 0.30517473816871643,
+      "learning_rate": 4.880732267965228e-05,
+      "loss": 0.4899,
+      "step": 1489
+    },
+    {
+      "epoch": 1.1215656755739556,
+      "grad_norm": 0.2710370719432831,
+      "learning_rate": 4.880493116678604e-05,
+      "loss": 0.4629,
+      "step": 1490
+    },
+    {
+      "epoch": 1.1223184042152803,
+      "grad_norm": 0.2325483113527298,
+      "learning_rate": 4.880253732404456e-05,
+      "loss": 0.4588,
+      "step": 1491
+    },
+    {
+      "epoch": 1.1230711328566052,
+      "grad_norm": 0.29568415880203247,
+      "learning_rate": 4.8800141151689614e-05,
+      "loss": 0.5356,
+      "step": 1492
+    },
+    {
+      "epoch": 1.12382386149793,
+      "grad_norm": 0.30240944027900696,
+      "learning_rate": 4.8797742649983265e-05,
+      "loss": 0.5216,
+      "step": 1493
+    },
+    {
+      "epoch": 1.1245765901392548,
+      "grad_norm": 0.2296067625284195,
+      "learning_rate": 4.87953418191878e-05,
+      "loss": 0.3768,
+      "step": 1494
+    },
+    {
+      "epoch": 1.1253293187805795,
+      "grad_norm": 0.2557816803455353,
+      "learning_rate": 4.879293865956578e-05,
+      "loss": 0.5101,
+      "step": 1495
+    },
+    {
+      "epoch": 1.1260820474219044,
+      "grad_norm": 0.2860860824584961,
+      "learning_rate": 4.879053317137999e-05,
+      "loss": 0.5955,
+      "step": 1496
+    },
+    {
+      "epoch": 1.1268347760632291,
+      "grad_norm": 0.26319876313209534,
+      "learning_rate": 4.87881253548935e-05,
+      "loss": 0.4545,
+      "step": 1497
+    },
+    {
+      "epoch": 1.127587504704554,
+      "grad_norm": 0.22734026610851288,
+      "learning_rate": 4.8785715210369635e-05,
+      "loss": 0.4393,
+      "step": 1498
+    },
+    {
+      "epoch": 1.1283402333458787,
+      "grad_norm": 0.29440852999687195,
+      "learning_rate": 4.8783302738071953e-05,
+      "loss": 0.4358,
+      "step": 1499
+    },
+    {
+      "epoch": 1.1290929619872037,
+      "grad_norm": 0.39815109968185425,
+      "learning_rate": 4.878088793826428e-05,
+      "loss": 0.4262,
+      "step": 1500
+    },
+    {
+      "epoch": 1.1298456906285284,
+      "grad_norm": 0.2651158273220062,
+      "learning_rate": 4.87784708112107e-05,
+      "loss": 0.5157,
+      "step": 1501
+    },
+    {
+      "epoch": 1.1305984192698533,
+      "grad_norm": 0.3067253530025482,
+      "learning_rate": 4.8776051357175554e-05,
+      "loss": 0.4934,
+      "step": 1502
+    },
+    {
+      "epoch": 1.131351147911178,
+      "grad_norm": 0.49952003359794617,
+      "learning_rate": 4.8773629576423406e-05,
+      "loss": 0.4262,
+      "step": 1503
+    },
+    {
+      "epoch": 1.1321038765525029,
+      "grad_norm": 0.41666337847709656,
+      "learning_rate": 4.877120546921911e-05,
+      "loss": 0.4991,
+      "step": 1504
+    },
+    {
+      "epoch": 1.1328566051938276,
+      "grad_norm": 0.25050196051597595,
+      "learning_rate": 4.8768779035827775e-05,
+      "loss": 0.3356,
+      "step": 1505
+    },
+    {
+      "epoch": 1.1336093338351525,
+      "grad_norm": 0.491150826215744,
+      "learning_rate": 4.876635027651474e-05,
+      "loss": 0.4827,
+      "step": 1506
+    },
+    {
+      "epoch": 1.1343620624764772,
+      "grad_norm": 0.46305057406425476,
+      "learning_rate": 4.876391919154561e-05,
+      "loss": 0.4746,
+      "step": 1507
+    },
+    {
+      "epoch": 1.1351147911178021,
+      "grad_norm": 0.2571219503879547,
+      "learning_rate": 4.876148578118624e-05,
+      "loss": 0.4524,
+      "step": 1508
+    },
+    {
+      "epoch": 1.1358675197591268,
+      "grad_norm": 0.41765859723091125,
+      "learning_rate": 4.875905004570277e-05,
+      "loss": 0.3443,
+      "step": 1509
+    },
+    {
+      "epoch": 1.1366202484004517,
+      "grad_norm": 0.44442424178123474,
+      "learning_rate": 4.8756611985361535e-05,
+      "loss": 0.5125,
+      "step": 1510
+    },
+    {
+      "epoch": 1.1373729770417764,
+      "grad_norm": 0.27437901496887207,
+      "learning_rate": 4.875417160042917e-05,
+      "loss": 0.3022,
+      "step": 1511
+    },
+    {
+      "epoch": 1.1381257056831013,
+      "grad_norm": 0.47188687324523926,
+      "learning_rate": 4.875172889117256e-05,
+      "loss": 0.4134,
+      "step": 1512
+    },
+    {
+      "epoch": 1.138878434324426,
+      "grad_norm": 0.24360337853431702,
+      "learning_rate": 4.874928385785883e-05,
+      "loss": 0.6131,
+      "step": 1513
+    },
+    {
+      "epoch": 1.1396311629657507,
+      "grad_norm": 0.356600284576416,
+      "learning_rate": 4.8746836500755366e-05,
+      "loss": 0.3239,
+      "step": 1514
+    },
+    {
+      "epoch": 1.1403838916070757,
+      "grad_norm": 0.3338331878185272,
+      "learning_rate": 4.874438682012981e-05,
+      "loss": 0.3763,
+      "step": 1515
+    },
+    {
+      "epoch": 1.1411366202484006,
+      "grad_norm": 0.35194769501686096,
+      "learning_rate": 4.874193481625004e-05,
+      "loss": 0.4828,
+      "step": 1516
+    },
+    {
+      "epoch": 1.1418893488897253,
+      "grad_norm": 0.3172166645526886,
+      "learning_rate": 4.8739480489384224e-05,
+      "loss": 0.4998,
+      "step": 1517
+    },
+    {
+      "epoch": 1.14264207753105,
+      "grad_norm": 0.2854245901107788,
+      "learning_rate": 4.873702383980076e-05,
+      "loss": 0.4186,
+      "step": 1518
+    },
+    {
+      "epoch": 1.1433948061723749,
+      "grad_norm": 0.34285980463027954,
+      "learning_rate": 4.873456486776829e-05,
+      "loss": 0.4516,
+      "step": 1519
+    },
+    {
+      "epoch": 1.1441475348136996,
+      "grad_norm": 0.32318025827407837,
+      "learning_rate": 4.8732103573555725e-05,
+      "loss": 0.4807,
+      "step": 1520
+    },
+    {
+      "epoch": 1.1449002634550245,
+      "grad_norm": 0.26943957805633545,
+      "learning_rate": 4.872963995743224e-05,
+      "loss": 0.4497,
+      "step": 1521
+    },
+    {
+      "epoch": 1.1456529920963492,
+      "grad_norm": 0.375058650970459,
+      "learning_rate": 4.872717401966726e-05,
+      "loss": 0.5047,
+      "step": 1522
+    },
+    {
+      "epoch": 1.146405720737674,
+      "grad_norm": 0.31144535541534424,
+      "learning_rate": 4.8724705760530424e-05,
+      "loss": 0.4403,
+      "step": 1523
+    },
+    {
+      "epoch": 1.1471584493789988,
+      "grad_norm": 0.34315648674964905,
+      "learning_rate": 4.872223518029169e-05,
+      "loss": 0.4068,
+      "step": 1524
+    },
+    {
+      "epoch": 1.1479111780203237,
+      "grad_norm": 0.23718871176242828,
+      "learning_rate": 4.871976227922122e-05,
+      "loss": 0.4786,
+      "step": 1525
+    },
+    {
+      "epoch": 1.1486639066616484,
+      "grad_norm": 0.3298598527908325,
+      "learning_rate": 4.871728705758946e-05,
+      "loss": 0.527,
+      "step": 1526
+    },
+    {
+      "epoch": 1.1494166353029733,
+      "grad_norm": 0.25268009305000305,
+      "learning_rate": 4.871480951566707e-05,
+      "loss": 0.444,
+      "step": 1527
+    },
+    {
+      "epoch": 1.150169363944298,
+      "grad_norm": 0.24051080644130707,
+      "learning_rate": 4.8712329653725016e-05,
+      "loss": 0.39,
+      "step": 1528
+    },
+    {
+      "epoch": 1.150922092585623,
+      "grad_norm": 0.2505611479282379,
+      "learning_rate": 4.870984747203449e-05,
+      "loss": 0.308,
+      "step": 1529
+    },
+    {
+      "epoch": 1.1516748212269476,
+      "grad_norm": 0.26638805866241455,
+      "learning_rate": 4.870736297086692e-05,
+      "loss": 0.3984,
+      "step": 1530
+    },
+    {
+      "epoch": 1.1524275498682726,
+      "grad_norm": 0.29273390769958496,
+      "learning_rate": 4.870487615049403e-05,
+      "loss": 0.4351,
+      "step": 1531
+    },
+    {
+      "epoch": 1.1531802785095973,
+      "grad_norm": 0.22268079221248627,
+      "learning_rate": 4.870238701118776e-05,
+      "loss": 0.4273,
+      "step": 1532
+    },
+    {
+      "epoch": 1.1539330071509222,
+      "grad_norm": 0.24899499118328094,
+      "learning_rate": 4.869989555322033e-05,
+      "loss": 0.3877,
+      "step": 1533
+    },
+    {
+      "epoch": 1.1546857357922469,
+      "grad_norm": 0.2619277834892273,
+      "learning_rate": 4.869740177686419e-05,
+      "loss": 0.3794,
+      "step": 1534
+    },
+    {
+      "epoch": 1.1554384644335718,
+      "grad_norm": 0.2435348629951477,
+      "learning_rate": 4.869490568239206e-05,
+      "loss": 0.4717,
+      "step": 1535
+    },
+    {
+      "epoch": 1.1561911930748965,
+      "grad_norm": 0.24189864099025726,
+      "learning_rate": 4.869240727007691e-05,
+      "loss": 0.3488,
+      "step": 1536
+    },
+    {
+      "epoch": 1.1569439217162212,
+      "grad_norm": 0.23101884126663208,
+      "learning_rate": 4.868990654019197e-05,
+      "loss": 0.3909,
+      "step": 1537
+    },
+    {
+      "epoch": 1.157696650357546,
+      "grad_norm": 0.25546571612358093,
+      "learning_rate": 4.868740349301071e-05,
+      "loss": 0.4831,
+      "step": 1538
+    },
+    {
+      "epoch": 1.158449378998871,
+      "grad_norm": 0.21023917198181152,
+      "learning_rate": 4.868489812880685e-05,
+      "loss": 0.3868,
+      "step": 1539
+    },
+    {
+      "epoch": 1.1592021076401957,
+      "grad_norm": 0.27782702445983887,
+      "learning_rate": 4.868239044785438e-05,
+      "loss": 0.5235,
+      "step": 1540
+    },
+    {
+      "epoch": 1.1599548362815204,
+      "grad_norm": 0.2753055989742279,
+      "learning_rate": 4.8679880450427544e-05,
+      "loss": 0.591,
+      "step": 1541
+    },
+    {
+      "epoch": 1.1607075649228453,
+      "grad_norm": 0.392404705286026,
+      "learning_rate": 4.867736813680083e-05,
+      "loss": 0.484,
+      "step": 1542
+    },
+    {
+      "epoch": 1.16146029356417,
+      "grad_norm": 0.47108009457588196,
+      "learning_rate": 4.8674853507248976e-05,
+      "loss": 0.5394,
+      "step": 1543
+    },
+    {
+      "epoch": 1.162213022205495,
+      "grad_norm": 0.271151602268219,
+      "learning_rate": 4.8672336562046974e-05,
+      "loss": 0.4916,
+      "step": 1544
+    },
+    {
+      "epoch": 1.1629657508468196,
+      "grad_norm": 0.3963524103164673,
+      "learning_rate": 4.866981730147008e-05,
+      "loss": 0.4101,
+      "step": 1545
+    },
+    {
+      "epoch": 1.1637184794881446,
+      "grad_norm": 0.3486645817756653,
+      "learning_rate": 4.866729572579379e-05,
+      "loss": 0.452,
+      "step": 1546
+    },
+    {
+      "epoch": 1.1644712081294692,
+      "grad_norm": 0.24884331226348877,
+      "learning_rate": 4.866477183529386e-05,
+      "loss": 0.4917,
+      "step": 1547
+    },
+    {
+      "epoch": 1.1652239367707942,
+      "grad_norm": 0.3131965100765228,
+      "learning_rate": 4.866224563024631e-05,
+      "loss": 0.4856,
+      "step": 1548
+    },
+    {
+      "epoch": 1.1659766654121189,
+      "grad_norm": 0.2637426257133484,
+      "learning_rate": 4.86597171109274e-05,
+      "loss": 0.4905,
+      "step": 1549
+    },
+    {
+      "epoch": 1.1667293940534438,
+      "grad_norm": 0.2809723913669586,
+      "learning_rate": 4.865718627761363e-05,
+      "loss": 0.4886,
+      "step": 1550
+    },
+    {
+      "epoch": 1.1674821226947685,
+      "grad_norm": 0.2908109426498413,
+      "learning_rate": 4.865465313058178e-05,
+      "loss": 0.4182,
+      "step": 1551
+    },
+    {
+      "epoch": 1.1682348513360934,
+      "grad_norm": 0.26504647731781006,
+      "learning_rate": 4.8652117670108874e-05,
+      "loss": 0.4362,
+      "step": 1552
+    },
+    {
+      "epoch": 1.168987579977418,
+      "grad_norm": 0.23807573318481445,
+      "learning_rate": 4.864957989647217e-05,
+      "loss": 0.4833,
+      "step": 1553
+    },
+    {
+      "epoch": 1.169740308618743,
+      "grad_norm": 0.3041771948337555,
+      "learning_rate": 4.8647039809949225e-05,
+      "loss": 0.4331,
+      "step": 1554
+    },
+    {
+      "epoch": 1.1704930372600677,
+      "grad_norm": 0.2708297073841095,
+      "learning_rate": 4.8644497410817786e-05,
+      "loss": 0.4336,
+      "step": 1555
+    },
+    {
+      "epoch": 1.1712457659013926,
+      "grad_norm": 0.3626130521297455,
+      "learning_rate": 4.8641952699355905e-05,
+      "loss": 0.4753,
+      "step": 1556
+    },
+    {
+      "epoch": 1.1719984945427173,
+      "grad_norm": 0.3433065712451935,
+      "learning_rate": 4.863940567584187e-05,
+      "loss": 0.3821,
+      "step": 1557
+    },
+    {
+      "epoch": 1.1727512231840422,
+      "grad_norm": 0.23410052061080933,
+      "learning_rate": 4.86368563405542e-05,
+      "loss": 0.4918,
+      "step": 1558
+    },
+    {
+      "epoch": 1.173503951825367,
+      "grad_norm": 0.4164998233318329,
+      "learning_rate": 4.863430469377171e-05,
+      "loss": 0.5541,
+      "step": 1559
+    },
+    {
+      "epoch": 1.1742566804666918,
+      "grad_norm": 0.2733100950717926,
+      "learning_rate": 4.863175073577342e-05,
+      "loss": 0.4334,
+      "step": 1560
+    },
+    {
+      "epoch": 1.1750094091080165,
+      "grad_norm": 0.35937491059303284,
+      "learning_rate": 4.8629194466838653e-05,
+      "loss": 0.5385,
+      "step": 1561
+    },
+    {
+      "epoch": 1.1757621377493415,
+      "grad_norm": 0.3160601258277893,
+      "learning_rate": 4.862663588724694e-05,
+      "loss": 0.5595,
+      "step": 1562
+    },
+    {
+      "epoch": 1.1765148663906662,
+      "grad_norm": 0.3055051267147064,
+      "learning_rate": 4.8624074997278096e-05,
+      "loss": 0.4599,
+      "step": 1563
+    },
+    {
+      "epoch": 1.1772675950319909,
+      "grad_norm": 0.3068599998950958,
+      "learning_rate": 4.862151179721217e-05,
+      "loss": 0.4905,
+      "step": 1564
+    },
+    {
+      "epoch": 1.1780203236733158,
+      "grad_norm": 0.3298959732055664,
+      "learning_rate": 4.861894628732946e-05,
+      "loss": 0.4785,
+      "step": 1565
+    },
+    {
+      "epoch": 1.1787730523146407,
+      "grad_norm": 0.437898188829422,
+      "learning_rate": 4.861637846791055e-05,
+      "loss": 0.4663,
+      "step": 1566
+    },
+    {
+      "epoch": 1.1795257809559654,
+      "grad_norm": 0.2881907522678375,
+      "learning_rate": 4.861380833923623e-05,
+      "loss": 0.4586,
+      "step": 1567
+    },
+    {
+      "epoch": 1.18027850959729,
+      "grad_norm": 0.2953038215637207,
+      "learning_rate": 4.861123590158758e-05,
+      "loss": 0.4215,
+      "step": 1568
+    },
+    {
+      "epoch": 1.181031238238615,
+      "grad_norm": 0.2804637849330902,
+      "learning_rate": 4.86086611552459e-05,
+      "loss": 0.3982,
+      "step": 1569
+    },
+    {
+      "epoch": 1.1817839668799397,
+      "grad_norm": 0.38074254989624023,
+      "learning_rate": 4.860608410049278e-05,
+      "loss": 0.435,
+      "step": 1570
+    },
+    {
+      "epoch": 1.1825366955212646,
+      "grad_norm": 0.2530324161052704,
+      "learning_rate": 4.8603504737610033e-05,
+      "loss": 0.4422,
+      "step": 1571
+    },
+    {
+      "epoch": 1.1832894241625893,
+      "grad_norm": 0.27807003259658813,
+      "learning_rate": 4.8600923066879736e-05,
+      "loss": 0.4484,
+      "step": 1572
+    },
+    {
+      "epoch": 1.1840421528039142,
+      "grad_norm": 0.22299648821353912,
+      "learning_rate": 4.859833908858422e-05,
+      "loss": 0.5302,
+      "step": 1573
+    },
+    {
+      "epoch": 1.184794881445239,
+      "grad_norm": 0.2720320224761963,
+      "learning_rate": 4.859575280300606e-05,
+      "loss": 0.4644,
+      "step": 1574
+    },
+    {
+      "epoch": 1.1855476100865638,
+      "grad_norm": 0.22586044669151306,
+      "learning_rate": 4.85931642104281e-05,
+      "loss": 0.501,
+      "step": 1575
+    },
+    {
+      "epoch": 1.1863003387278885,
+      "grad_norm": 0.27383849024772644,
+      "learning_rate": 4.8590573311133404e-05,
+      "loss": 0.3593,
+      "step": 1576
+    },
+    {
+      "epoch": 1.1870530673692135,
+      "grad_norm": 0.2710656523704529,
+      "learning_rate": 4.8587980105405324e-05,
+      "loss": 0.4171,
+      "step": 1577
+    },
+    {
+      "epoch": 1.1878057960105382,
+      "grad_norm": 0.29857417941093445,
+      "learning_rate": 4.8585384593527445e-05,
+      "loss": 0.4557,
+      "step": 1578
+    },
+    {
+      "epoch": 1.188558524651863,
+      "grad_norm": 0.2639225125312805,
+      "learning_rate": 4.8582786775783606e-05,
+      "loss": 0.4486,
+      "step": 1579
+    },
+    {
+      "epoch": 1.1893112532931878,
+      "grad_norm": 0.2900145649909973,
+      "learning_rate": 4.858018665245791e-05,
+      "loss": 0.4669,
+      "step": 1580
+    },
+    {
+      "epoch": 1.1900639819345127,
+      "grad_norm": 0.22493132948875427,
+      "learning_rate": 4.857758422383468e-05,
+      "loss": 0.4837,
+      "step": 1581
+    },
+    {
+      "epoch": 1.1908167105758374,
+      "grad_norm": 0.2863833010196686,
+      "learning_rate": 4.8574979490198544e-05,
+      "loss": 0.4247,
+      "step": 1582
+    },
+    {
+      "epoch": 1.1915694392171623,
+      "grad_norm": 0.2571309208869934,
+      "learning_rate": 4.857237245183433e-05,
+      "loss": 0.4813,
+      "step": 1583
+    },
+    {
+      "epoch": 1.192322167858487,
+      "grad_norm": 0.3386785387992859,
+      "learning_rate": 4.856976310902715e-05,
+      "loss": 0.4348,
+      "step": 1584
+    },
+    {
+      "epoch": 1.193074896499812,
+      "grad_norm": 0.3432064950466156,
+      "learning_rate": 4.856715146206235e-05,
+      "loss": 0.4061,
+      "step": 1585
+    },
+    {
+      "epoch": 1.1938276251411366,
+      "grad_norm": 0.30997684597969055,
+      "learning_rate": 4.8564537511225536e-05,
+      "loss": 0.4856,
+      "step": 1586
+    },
+    {
+      "epoch": 1.1945803537824613,
+      "grad_norm": 0.25904083251953125,
+      "learning_rate": 4.8561921256802586e-05,
+      "loss": 0.4009,
+      "step": 1587
+    },
+    {
+      "epoch": 1.1953330824237862,
+      "grad_norm": 0.30638980865478516,
+      "learning_rate": 4.855930269907958e-05,
+      "loss": 0.4333,
+      "step": 1588
+    },
+    {
+      "epoch": 1.1960858110651111,
+      "grad_norm": 0.24557241797447205,
+      "learning_rate": 4.8556681838342894e-05,
+      "loss": 0.4389,
+      "step": 1589
+    },
+    {
+      "epoch": 1.1968385397064358,
+      "grad_norm": 0.2704263925552368,
+      "learning_rate": 4.855405867487915e-05,
+      "loss": 0.5352,
+      "step": 1590
+    },
+    {
+      "epoch": 1.1975912683477605,
+      "grad_norm": 0.26796644926071167,
+      "learning_rate": 4.8551433208975186e-05,
+      "loss": 0.4703,
+      "step": 1591
+    },
+    {
+      "epoch": 1.1983439969890854,
+      "grad_norm": 0.24404564499855042,
+      "learning_rate": 4.8548805440918154e-05,
+      "loss": 0.4424,
+      "step": 1592
+    },
+    {
+      "epoch": 1.1990967256304101,
+      "grad_norm": 0.2365448772907257,
+      "learning_rate": 4.8546175370995394e-05,
+      "loss": 0.4304,
+      "step": 1593
+    },
+    {
+      "epoch": 1.199849454271735,
+      "grad_norm": 0.25341713428497314,
+      "learning_rate": 4.854354299949454e-05,
+      "loss": 0.4331,
+      "step": 1594
+    },
+    {
+      "epoch": 1.2006021829130598,
+      "grad_norm": 0.24776121973991394,
+      "learning_rate": 4.8540908326703464e-05,
+      "loss": 0.4061,
+      "step": 1595
+    },
+    {
+      "epoch": 1.2013549115543847,
+      "grad_norm": 0.26136210560798645,
+      "learning_rate": 4.8538271352910284e-05,
+      "loss": 0.4268,
+      "step": 1596
+    },
+    {
+      "epoch": 1.2021076401957094,
+      "grad_norm": 0.2652926743030548,
+      "learning_rate": 4.853563207840339e-05,
+      "loss": 0.4699,
+      "step": 1597
+    },
+    {
+      "epoch": 1.2028603688370343,
+      "grad_norm": 0.26088470220565796,
+      "learning_rate": 4.8532990503471393e-05,
+      "loss": 0.5191,
+      "step": 1598
+    },
+    {
+      "epoch": 1.203613097478359,
+      "grad_norm": 0.21366901695728302,
+      "learning_rate": 4.853034662840317e-05,
+      "loss": 0.4505,
+      "step": 1599
+    },
+    {
+      "epoch": 1.204365826119684,
+      "grad_norm": 0.2801893949508667,
+      "learning_rate": 4.8527700453487873e-05,
+      "loss": 0.3875,
+      "step": 1600
+    },
+    {
+      "epoch": 1.204365826119684,
+      "eval_loss": 0.4342336356639862,
+      "eval_runtime": 452.7079,
+      "eval_samples_per_second": 21.265,
+      "eval_steps_per_second": 0.665,
+      "step": 1600
+    },
+    {
+      "epoch": 1.2051185547610086,
+      "grad_norm": 0.21198582649230957,
+      "learning_rate": 4.8525051979014856e-05,
+      "loss": 0.4345,
+      "step": 1601
+    },
+    {
+      "epoch": 1.2058712834023335,
+      "grad_norm": 0.2325175702571869,
+      "learning_rate": 4.852240120527378e-05,
+      "loss": 0.5004,
+      "step": 1602
+    },
+    {
+      "epoch": 1.2066240120436582,
+      "grad_norm": 0.22203423082828522,
+      "learning_rate": 4.85197481325545e-05,
+      "loss": 0.4003,
+      "step": 1603
+    },
+    {
+      "epoch": 1.2073767406849831,
+      "grad_norm": 0.2364037036895752,
+      "learning_rate": 4.851709276114717e-05,
+      "loss": 0.4581,
+      "step": 1604
+    },
+    {
+      "epoch": 1.2081294693263078,
+      "grad_norm": 0.2340337336063385,
+      "learning_rate": 4.8514435091342176e-05,
+      "loss": 0.5464,
+      "step": 1605
+    },
+    {
+      "epoch": 1.2088821979676327,
+      "grad_norm": 0.21733573079109192,
+      "learning_rate": 4.851177512343015e-05,
+      "loss": 0.4327,
+      "step": 1606
+    },
+    {
+      "epoch": 1.2096349266089574,
+      "grad_norm": 0.2235325127840042,
+      "learning_rate": 4.8509112857702e-05,
+      "loss": 0.4456,
+      "step": 1607
+    },
+    {
+      "epoch": 1.2103876552502824,
+      "grad_norm": 0.22018079459667206,
+      "learning_rate": 4.8506448294448844e-05,
+      "loss": 0.4519,
+      "step": 1608
+    },
+    {
+      "epoch": 1.211140383891607,
+      "grad_norm": 0.22693125903606415,
+      "learning_rate": 4.8503781433962093e-05,
+      "loss": 0.3564,
+      "step": 1609
+    },
+    {
+      "epoch": 1.211893112532932,
+      "grad_norm": 0.261954128742218,
+      "learning_rate": 4.8501112276533376e-05,
+      "loss": 0.3158,
+      "step": 1610
+    },
+    {
+      "epoch": 1.2126458411742567,
+      "grad_norm": 0.35271379351615906,
+      "learning_rate": 4.84984408224546e-05,
+      "loss": 0.4607,
+      "step": 1611
+    },
+    {
+      "epoch": 1.2133985698155816,
+      "grad_norm": 0.31556040048599243,
+      "learning_rate": 4.84957670720179e-05,
+      "loss": 0.4156,
+      "step": 1612
+    },
+    {
+      "epoch": 1.2141512984569063,
+      "grad_norm": 0.23087771236896515,
+      "learning_rate": 4.849309102551569e-05,
+      "loss": 0.518,
+      "step": 1613
+    },
+    {
+      "epoch": 1.214904027098231,
+      "grad_norm": 0.45848220586776733,
+      "learning_rate": 4.84904126832406e-05,
+      "loss": 0.4896,
+      "step": 1614
+    },
+    {
+      "epoch": 1.215656755739556,
+      "grad_norm": 0.2948518693447113,
+      "learning_rate": 4.8487732045485535e-05,
+      "loss": 0.5109,
+      "step": 1615
+    },
+    {
+      "epoch": 1.2164094843808808,
+      "grad_norm": 0.30387043952941895,
+      "learning_rate": 4.848504911254365e-05,
+      "loss": 0.3848,
+      "step": 1616
+    },
+    {
+      "epoch": 1.2171622130222055,
+      "grad_norm": 0.3806714117527008,
+      "learning_rate": 4.848236388470836e-05,
+      "loss": 0.3894,
+      "step": 1617
+    },
+    {
+      "epoch": 1.2179149416635302,
+      "grad_norm": 0.2900158166885376,
+      "learning_rate": 4.847967636227328e-05,
+      "loss": 0.4997,
+      "step": 1618
+    },
+    {
+      "epoch": 1.2186676703048551,
+      "grad_norm": 0.3200798034667969,
+      "learning_rate": 4.847698654553235e-05,
+      "loss": 0.4183,
+      "step": 1619
+    },
+    {
+      "epoch": 1.2194203989461798,
+      "grad_norm": 0.27352914214134216,
+      "learning_rate": 4.8474294434779696e-05,
+      "loss": 0.4215,
+      "step": 1620
+    },
+    {
+      "epoch": 1.2201731275875047,
+      "grad_norm": 0.27912941575050354,
+      "learning_rate": 4.8471600030309744e-05,
+      "loss": 0.5205,
+      "step": 1621
+    },
+    {
+      "epoch": 1.2209258562288294,
+      "grad_norm": 0.2748068869113922,
+      "learning_rate": 4.8468903332417146e-05,
+      "loss": 0.4378,
+      "step": 1622
+    },
+    {
+      "epoch": 1.2216785848701543,
+      "grad_norm": 0.2483839988708496,
+      "learning_rate": 4.846620434139681e-05,
+      "loss": 0.4571,
+      "step": 1623
+    },
+    {
+      "epoch": 1.222431313511479,
+      "grad_norm": 0.266802579164505,
+      "learning_rate": 4.846350305754388e-05,
+      "loss": 0.3883,
+      "step": 1624
+    },
+    {
+      "epoch": 1.223184042152804,
+      "grad_norm": 0.2991360127925873,
+      "learning_rate": 4.846079948115378e-05,
+      "loss": 0.5045,
+      "step": 1625
+    },
+    {
+      "epoch": 1.2239367707941287,
+      "grad_norm": 0.2684039771556854,
+      "learning_rate": 4.8458093612522156e-05,
+      "loss": 0.4525,
+      "step": 1626
+    },
+    {
+      "epoch": 1.2246894994354536,
+      "grad_norm": 0.3375545144081116,
+      "learning_rate": 4.8455385451944935e-05,
+      "loss": 0.4485,
+      "step": 1627
+    },
+    {
+      "epoch": 1.2254422280767783,
+      "grad_norm": 0.2723807394504547,
+      "learning_rate": 4.845267499971826e-05,
+      "loss": 0.3566,
+      "step": 1628
+    },
+    {
+      "epoch": 1.2261949567181032,
+      "grad_norm": 0.3085951507091522,
+      "learning_rate": 4.844996225613855e-05,
+      "loss": 0.41,
+      "step": 1629
+    },
+    {
+      "epoch": 1.2269476853594279,
+      "grad_norm": 0.2425432801246643,
+      "learning_rate": 4.8447247221502464e-05,
+      "loss": 0.3857,
+      "step": 1630
+    },
+    {
+      "epoch": 1.2277004140007528,
+      "grad_norm": 0.24313567578792572,
+      "learning_rate": 4.844452989610692e-05,
+      "loss": 0.3943,
+      "step": 1631
+    },
+    {
+      "epoch": 1.2284531426420775,
+      "grad_norm": 0.29278048872947693,
+      "learning_rate": 4.8441810280249065e-05,
+      "loss": 0.4227,
+      "step": 1632
+    },
+    {
+      "epoch": 1.2292058712834024,
+      "grad_norm": 0.24718037247657776,
+      "learning_rate": 4.8439088374226334e-05,
+      "loss": 0.4664,
+      "step": 1633
+    },
+    {
+      "epoch": 1.2299585999247271,
+      "grad_norm": 0.2487625628709793,
+      "learning_rate": 4.8436364178336374e-05,
+      "loss": 0.4267,
+      "step": 1634
+    },
+    {
+      "epoch": 1.230711328566052,
+      "grad_norm": 0.26330143213272095,
+      "learning_rate": 4.843363769287711e-05,
+      "loss": 0.3742,
+      "step": 1635
+    },
+    {
+      "epoch": 1.2314640572073767,
+      "grad_norm": 0.2534533441066742,
+      "learning_rate": 4.843090891814671e-05,
+      "loss": 0.4732,
+      "step": 1636
+    },
+    {
+      "epoch": 1.2322167858487014,
+      "grad_norm": 0.2777356505393982,
+      "learning_rate": 4.842817785444356e-05,
+      "loss": 0.441,
+      "step": 1637
+    },
+    {
+      "epoch": 1.2329695144900263,
+      "grad_norm": 0.40019387006759644,
+      "learning_rate": 4.842544450206635e-05,
+      "loss": 0.4633,
+      "step": 1638
+    },
+    {
+      "epoch": 1.2337222431313513,
+      "grad_norm": 0.2968333065509796,
+      "learning_rate": 4.8422708861314e-05,
+      "loss": 0.5711,
+      "step": 1639
+    },
+    {
+      "epoch": 1.234474971772676,
+      "grad_norm": 0.2956465482711792,
+      "learning_rate": 4.841997093248566e-05,
+      "loss": 0.516,
+      "step": 1640
+    },
+    {
+      "epoch": 1.2352277004140007,
+      "grad_norm": 0.3770597279071808,
+      "learning_rate": 4.8417230715880754e-05,
+      "loss": 0.4513,
+      "step": 1641
+    },
+    {
+      "epoch": 1.2359804290553256,
+      "grad_norm": 0.30362868309020996,
+      "learning_rate": 4.8414488211798944e-05,
+      "loss": 0.4907,
+      "step": 1642
+    },
+    {
+      "epoch": 1.2367331576966503,
+      "grad_norm": 0.3016539514064789,
+      "learning_rate": 4.841174342054014e-05,
+      "loss": 0.5631,
+      "step": 1643
+    },
+    {
+      "epoch": 1.2374858863379752,
+      "grad_norm": 0.33092495799064636,
+      "learning_rate": 4.840899634240452e-05,
+      "loss": 0.4117,
+      "step": 1644
+    },
+    {
+      "epoch": 1.2382386149792999,
+      "grad_norm": 0.4362671375274658,
+      "learning_rate": 4.84062469776925e-05,
+      "loss": 0.382,
+      "step": 1645
+    },
+    {
+      "epoch": 1.2389913436206248,
+      "grad_norm": 0.29778724908828735,
+      "learning_rate": 4.8403495326704746e-05,
+      "loss": 0.5002,
+      "step": 1646
+    },
+    {
+      "epoch": 1.2397440722619495,
+      "grad_norm": 0.3032190501689911,
+      "learning_rate": 4.8400741389742164e-05,
+      "loss": 0.4092,
+      "step": 1647
+    },
+    {
+      "epoch": 1.2404968009032744,
+      "grad_norm": 0.2984626889228821,
+      "learning_rate": 4.839798516710593e-05,
+      "loss": 0.4463,
+      "step": 1648
+    },
+    {
+      "epoch": 1.241249529544599,
+      "grad_norm": 0.28582480549812317,
+      "learning_rate": 4.8395226659097456e-05,
+      "loss": 0.4765,
+      "step": 1649
+    },
+    {
+      "epoch": 1.242002258185924,
+      "grad_norm": 0.3525241017341614,
+      "learning_rate": 4.839246586601841e-05,
+      "loss": 0.4064,
+      "step": 1650
+    },
+    {
+      "epoch": 1.2427549868272487,
+      "grad_norm": 0.3055661916732788,
+      "learning_rate": 4.838970278817071e-05,
+      "loss": 0.4519,
+      "step": 1651
+    },
+    {
+      "epoch": 1.2435077154685736,
+      "grad_norm": 0.24756765365600586,
+      "learning_rate": 4.8386937425856525e-05,
+      "loss": 0.3918,
+      "step": 1652
+    },
+    {
+      "epoch": 1.2442604441098983,
+      "grad_norm": 0.30075377225875854,
+      "learning_rate": 4.8384169779378256e-05,
+      "loss": 0.4804,
+      "step": 1653
+    },
+    {
+      "epoch": 1.2450131727512233,
+      "grad_norm": 0.29932701587677,
+      "learning_rate": 4.838139984903859e-05,
+      "loss": 0.5568,
+      "step": 1654
+    },
+    {
+      "epoch": 1.245765901392548,
+      "grad_norm": 0.25996220111846924,
+      "learning_rate": 4.837862763514042e-05,
+      "loss": 0.5271,
+      "step": 1655
+    },
+    {
+      "epoch": 1.2465186300338729,
+      "grad_norm": 0.33742815256118774,
+      "learning_rate": 4.837585313798693e-05,
+      "loss": 0.4185,
+      "step": 1656
+    },
+    {
+      "epoch": 1.2472713586751976,
+      "grad_norm": 0.2533934712409973,
+      "learning_rate": 4.8373076357881524e-05,
+      "loss": 0.3479,
+      "step": 1657
+    },
+    {
+      "epoch": 1.2480240873165225,
+      "grad_norm": 0.3317462205886841,
+      "learning_rate": 4.837029729512787e-05,
+      "loss": 0.4371,
+      "step": 1658
+    },
+    {
+      "epoch": 1.2487768159578472,
+      "grad_norm": 0.2789627015590668,
+      "learning_rate": 4.8367515950029885e-05,
+      "loss": 0.3966,
+      "step": 1659
+    },
+    {
+      "epoch": 1.249529544599172,
+      "grad_norm": 0.37347209453582764,
+      "learning_rate": 4.8364732322891726e-05,
+      "loss": 0.4647,
+      "step": 1660
+    },
+    {
+      "epoch": 1.2502822732404968,
+      "grad_norm": 0.2304518222808838,
+      "learning_rate": 4.836194641401781e-05,
+      "loss": 0.4874,
+      "step": 1661
+    },
+    {
+      "epoch": 1.2510350018818217,
+      "grad_norm": 0.2855944335460663,
+      "learning_rate": 4.83591582237128e-05,
+      "loss": 0.4694,
+      "step": 1662
+    },
+    {
+      "epoch": 1.2517877305231464,
+      "grad_norm": 0.2600764036178589,
+      "learning_rate": 4.835636775228161e-05,
+      "loss": 0.5039,
+      "step": 1663
+    },
+    {
+      "epoch": 1.252540459164471,
+      "grad_norm": 0.2804638147354126,
+      "learning_rate": 4.835357500002939e-05,
+      "loss": 0.451,
+      "step": 1664
+    },
+    {
+      "epoch": 1.253293187805796,
+      "grad_norm": 0.22513586282730103,
+      "learning_rate": 4.835077996726157e-05,
+      "loss": 0.3513,
+      "step": 1665
+    },
+    {
+      "epoch": 1.254045916447121,
+      "grad_norm": 0.2733144462108612,
+      "learning_rate": 4.8347982654283806e-05,
+      "loss": 0.3954,
+      "step": 1666
+    },
+    {
+      "epoch": 1.2547986450884456,
+      "grad_norm": 0.24333800375461578,
+      "learning_rate": 4.8345183061402e-05,
+      "loss": 0.4641,
+      "step": 1667
+    },
+    {
+      "epoch": 1.2555513737297703,
+      "grad_norm": 0.23035386204719543,
+      "learning_rate": 4.834238118892231e-05,
+      "loss": 0.3468,
+      "step": 1668
+    },
+    {
+      "epoch": 1.2563041023710952,
+      "grad_norm": 0.22813259065151215,
+      "learning_rate": 4.833957703715115e-05,
+      "loss": 0.446,
+      "step": 1669
+    },
+    {
+      "epoch": 1.25705683101242,
+      "grad_norm": 0.22557973861694336,
+      "learning_rate": 4.8336770606395175e-05,
+      "loss": 0.4706,
+      "step": 1670
+    },
+    {
+      "epoch": 1.2578095596537449,
+      "grad_norm": 0.2635526955127716,
+      "learning_rate": 4.83339618969613e-05,
+      "loss": 0.5218,
+      "step": 1671
+    },
+    {
+      "epoch": 1.2585622882950696,
+      "grad_norm": 0.22086364030838013,
+      "learning_rate": 4.833115090915667e-05,
+      "loss": 0.3461,
+      "step": 1672
+    },
+    {
+      "epoch": 1.2593150169363945,
+      "grad_norm": 0.23217661678791046,
+      "learning_rate": 4.83283376432887e-05,
+      "loss": 0.3379,
+      "step": 1673
+    },
+    {
+      "epoch": 1.2600677455777192,
+      "grad_norm": 0.26854801177978516,
+      "learning_rate": 4.8325522099665044e-05,
+      "loss": 0.3753,
+      "step": 1674
+    },
+    {
+      "epoch": 1.260820474219044,
+      "grad_norm": 0.28426262736320496,
+      "learning_rate": 4.83227042785936e-05,
+      "loss": 0.4145,
+      "step": 1675
+    },
+    {
+      "epoch": 1.2615732028603688,
+      "grad_norm": 0.24341978132724762,
+      "learning_rate": 4.831988418038251e-05,
+      "loss": 0.4103,
+      "step": 1676
+    },
+    {
+      "epoch": 1.2623259315016937,
+      "grad_norm": 0.2338050752878189,
+      "learning_rate": 4.83170618053402e-05,
+      "loss": 0.4795,
+      "step": 1677
+    },
+    {
+      "epoch": 1.2630786601430184,
+      "grad_norm": 0.24185499548912048,
+      "learning_rate": 4.83142371537753e-05,
+      "loss": 0.4343,
+      "step": 1678
+    },
+    {
+      "epoch": 1.2638313887843433,
+      "grad_norm": 0.2720898389816284,
+      "learning_rate": 4.8311410225996725e-05,
+      "loss": 0.4101,
+      "step": 1679
+    },
+    {
+      "epoch": 1.264584117425668,
+      "grad_norm": 0.22411160171031952,
+      "learning_rate": 4.830858102231363e-05,
+      "loss": 0.3893,
+      "step": 1680
+    },
+    {
+      "epoch": 1.265336846066993,
+      "grad_norm": 0.25447142124176025,
+      "learning_rate": 4.8305749543035385e-05,
+      "loss": 0.4297,
+      "step": 1681
+    },
+    {
+      "epoch": 1.2660895747083176,
+      "grad_norm": 0.2689523696899414,
+      "learning_rate": 4.8302915788471646e-05,
+      "loss": 0.4328,
+      "step": 1682
+    },
+    {
+      "epoch": 1.2668423033496423,
+      "grad_norm": 0.33353906869888306,
+      "learning_rate": 4.830007975893232e-05,
+      "loss": 0.4185,
+      "step": 1683
+    },
+    {
+      "epoch": 1.2675950319909672,
+      "grad_norm": 0.2785061001777649,
+      "learning_rate": 4.829724145472755e-05,
+      "loss": 0.3387,
+      "step": 1684
+    },
+    {
+      "epoch": 1.2683477606322922,
+      "grad_norm": 0.2881145179271698,
+      "learning_rate": 4.829440087616772e-05,
+      "loss": 0.4227,
+      "step": 1685
+    },
+    {
+      "epoch": 1.2691004892736168,
+      "grad_norm": 0.26391005516052246,
+      "learning_rate": 4.829155802356346e-05,
+      "loss": 0.4097,
+      "step": 1686
+    },
+    {
+      "epoch": 1.2698532179149415,
+      "grad_norm": 0.36870908737182617,
+      "learning_rate": 4.828871289722569e-05,
+      "loss": 0.446,
+      "step": 1687
+    },
+    {
+      "epoch": 1.2706059465562665,
+      "grad_norm": 0.436827689409256,
+      "learning_rate": 4.8285865497465535e-05,
+      "loss": 0.4702,
+      "step": 1688
+    },
+    {
+      "epoch": 1.2713586751975914,
+      "grad_norm": 0.3384016156196594,
+      "learning_rate": 4.8283015824594366e-05,
+      "loss": 0.4246,
+      "step": 1689
+    },
+    {
+      "epoch": 1.272111403838916,
+      "grad_norm": 0.3855983316898346,
+      "learning_rate": 4.8280163878923844e-05,
+      "loss": 0.3952,
+      "step": 1690
+    },
+    {
+      "epoch": 1.2728641324802408,
+      "grad_norm": 0.33927667140960693,
+      "learning_rate": 4.827730966076584e-05,
+      "loss": 0.54,
+      "step": 1691
+    },
+    {
+      "epoch": 1.2736168611215657,
+      "grad_norm": 0.43647444248199463,
+      "learning_rate": 4.827445317043249e-05,
+      "loss": 0.4199,
+      "step": 1692
+    },
+    {
+      "epoch": 1.2743695897628906,
+      "grad_norm": 0.2825603485107422,
+      "learning_rate": 4.827159440823617e-05,
+      "loss": 0.3973,
+      "step": 1693
+    },
+    {
+      "epoch": 1.2751223184042153,
+      "grad_norm": 0.3180522918701172,
+      "learning_rate": 4.826873337448952e-05,
+      "loss": 0.4743,
+      "step": 1694
+    },
+    {
+      "epoch": 1.27587504704554,
+      "grad_norm": 0.31733280420303345,
+      "learning_rate": 4.8265870069505416e-05,
+      "loss": 0.437,
+      "step": 1695
+    },
+    {
+      "epoch": 1.276627775686865,
+      "grad_norm": 0.2857419550418854,
+      "learning_rate": 4.826300449359698e-05,
+      "loss": 0.5211,
+      "step": 1696
+    },
+    {
+      "epoch": 1.2773805043281896,
+      "grad_norm": 0.342062771320343,
+      "learning_rate": 4.826013664707759e-05,
+      "loss": 0.4601,
+      "step": 1697
+    },
+    {
+      "epoch": 1.2781332329695145,
+      "grad_norm": 0.28143781423568726,
+      "learning_rate": 4.825726653026086e-05,
+      "loss": 0.5214,
+      "step": 1698
+    },
+    {
+      "epoch": 1.2788859616108392,
+      "grad_norm": 0.26057353615760803,
+      "learning_rate": 4.825439414346068e-05,
+      "loss": 0.4663,
+      "step": 1699
+    },
+    {
+      "epoch": 1.2796386902521641,
+      "grad_norm": 0.2451959252357483,
+      "learning_rate": 4.825151948699116e-05,
+      "loss": 0.4239,
+      "step": 1700
+    },
+    {
+      "epoch": 1.2803914188934888,
+      "grad_norm": 0.2596691846847534,
+      "learning_rate": 4.8248642561166665e-05,
+      "loss": 0.4523,
+      "step": 1701
+    },
+    {
+      "epoch": 1.2811441475348138,
+      "grad_norm": 0.29040390253067017,
+      "learning_rate": 4.8245763366301814e-05,
+      "loss": 0.4133,
+      "step": 1702
+    },
+    {
+      "epoch": 1.2818968761761385,
+      "grad_norm": 0.2351270318031311,
+      "learning_rate": 4.824288190271147e-05,
+      "loss": 0.4306,
+      "step": 1703
+    },
+    {
+      "epoch": 1.2826496048174634,
+      "grad_norm": 0.25222089886665344,
+      "learning_rate": 4.8239998170710756e-05,
+      "loss": 0.4215,
+      "step": 1704
+    },
+    {
+      "epoch": 1.283402333458788,
+      "grad_norm": 0.25415679812431335,
+      "learning_rate": 4.823711217061502e-05,
+      "loss": 0.4442,
+      "step": 1705
+    },
+    {
+      "epoch": 1.284155062100113,
+      "grad_norm": 0.26396942138671875,
+      "learning_rate": 4.8234223902739874e-05,
+      "loss": 0.4491,
+      "step": 1706
+    },
+    {
+      "epoch": 1.2849077907414377,
+      "grad_norm": 0.26378777623176575,
+      "learning_rate": 4.8231333367401175e-05,
+      "loss": 0.4641,
+      "step": 1707
+    },
+    {
+      "epoch": 1.2856605193827626,
+      "grad_norm": 0.2695189416408539,
+      "learning_rate": 4.822844056491503e-05,
+      "loss": 0.4212,
+      "step": 1708
+    },
+    {
+      "epoch": 1.2864132480240873,
+      "grad_norm": 0.2374517023563385,
+      "learning_rate": 4.822554549559778e-05,
+      "loss": 0.3694,
+      "step": 1709
+    },
+    {
+      "epoch": 1.287165976665412,
+      "grad_norm": 0.21264934539794922,
+      "learning_rate": 4.822264815976605e-05,
+      "loss": 0.5036,
+      "step": 1710
+    },
+    {
+      "epoch": 1.287918705306737,
+      "grad_norm": 0.33890700340270996,
+      "learning_rate": 4.8219748557736664e-05,
+      "loss": 0.4098,
+      "step": 1711
+    },
+    {
+      "epoch": 1.2886714339480618,
+      "grad_norm": 0.2788410186767578,
+      "learning_rate": 4.821684668982673e-05,
+      "loss": 0.5064,
+      "step": 1712
+    },
+    {
+      "epoch": 1.2894241625893865,
+      "grad_norm": 0.24060732126235962,
+      "learning_rate": 4.821394255635359e-05,
+      "loss": 0.4081,
+      "step": 1713
+    },
+    {
+      "epoch": 1.2901768912307112,
+      "grad_norm": 0.2615283131599426,
+      "learning_rate": 4.821103615763484e-05,
+      "loss": 0.3441,
+      "step": 1714
+    },
+    {
+      "epoch": 1.2909296198720361,
+      "grad_norm": 0.29036039113998413,
+      "learning_rate": 4.8208127493988304e-05,
+      "loss": 0.5241,
+      "step": 1715
+    },
+    {
+      "epoch": 1.291682348513361,
+      "grad_norm": 0.26475173234939575,
+      "learning_rate": 4.820521656573209e-05,
+      "loss": 0.5019,
+      "step": 1716
+    },
+    {
+      "epoch": 1.2924350771546858,
+      "grad_norm": 0.3474060297012329,
+      "learning_rate": 4.820230337318451e-05,
+      "loss": 0.5202,
+      "step": 1717
+    },
+    {
+      "epoch": 1.2931878057960104,
+      "grad_norm": 0.3089390993118286,
+      "learning_rate": 4.819938791666416e-05,
+      "loss": 0.4223,
+      "step": 1718
+    },
+    {
+      "epoch": 1.2939405344373354,
+      "grad_norm": 0.30584388971328735,
+      "learning_rate": 4.819647019648988e-05,
+      "loss": 0.4209,
+      "step": 1719
+    },
+    {
+      "epoch": 1.29469326307866,
+      "grad_norm": 0.35376301407814026,
+      "learning_rate": 4.819355021298073e-05,
+      "loss": 0.4128,
+      "step": 1720
+    },
+    {
+      "epoch": 1.295445991719985,
+      "grad_norm": 0.2654455602169037,
+      "learning_rate": 4.819062796645604e-05,
+      "loss": 0.3699,
+      "step": 1721
+    },
+    {
+      "epoch": 1.2961987203613097,
+      "grad_norm": 0.25730618834495544,
+      "learning_rate": 4.8187703457235386e-05,
+      "loss": 0.4528,
+      "step": 1722
+    },
+    {
+      "epoch": 1.2969514490026346,
+      "grad_norm": 0.30552348494529724,
+      "learning_rate": 4.818477668563858e-05,
+      "loss": 0.4089,
+      "step": 1723
+    },
+    {
+      "epoch": 1.2977041776439593,
+      "grad_norm": 0.23026138544082642,
+      "learning_rate": 4.8181847651985703e-05,
+      "loss": 0.4582,
+      "step": 1724
+    },
+    {
+      "epoch": 1.2984569062852842,
+      "grad_norm": 0.3486420214176178,
+      "learning_rate": 4.8178916356597047e-05,
+      "loss": 0.432,
+      "step": 1725
+    },
+    {
+      "epoch": 1.299209634926609,
+      "grad_norm": 0.29950010776519775,
+      "learning_rate": 4.8175982799793207e-05,
+      "loss": 0.4029,
+      "step": 1726
+    },
+    {
+      "epoch": 1.2999623635679338,
+      "grad_norm": 0.2314315289258957,
+      "learning_rate": 4.817304698189496e-05,
+      "loss": 0.3979,
+      "step": 1727
+    },
+    {
+      "epoch": 1.3007150922092585,
+      "grad_norm": 0.31170156598091125,
+      "learning_rate": 4.817010890322338e-05,
+      "loss": 0.4297,
+      "step": 1728
+    },
+    {
+      "epoch": 1.3014678208505834,
+      "grad_norm": 0.2907775640487671,
+      "learning_rate": 4.816716856409977e-05,
+      "loss": 0.5087,
+      "step": 1729
+    },
+    {
+      "epoch": 1.3022205494919081,
+      "grad_norm": 0.24660342931747437,
+      "learning_rate": 4.816422596484568e-05,
+      "loss": 0.3933,
+      "step": 1730
+    },
+    {
+      "epoch": 1.302973278133233,
+      "grad_norm": 0.3163498640060425,
+      "learning_rate": 4.8161281105782906e-05,
+      "loss": 0.4039,
+      "step": 1731
+    },
+    {
+      "epoch": 1.3037260067745577,
+      "grad_norm": 0.27813416719436646,
+      "learning_rate": 4.815833398723348e-05,
+      "loss": 0.4601,
+      "step": 1732
+    },
+    {
+      "epoch": 1.3044787354158824,
+      "grad_norm": 0.34157320857048035,
+      "learning_rate": 4.8155384609519724e-05,
+      "loss": 0.4697,
+      "step": 1733
+    },
+    {
+      "epoch": 1.3052314640572074,
+      "grad_norm": 0.359152227640152,
+      "learning_rate": 4.8152432972964165e-05,
+      "loss": 0.4175,
+      "step": 1734
+    },
+    {
+      "epoch": 1.3059841926985323,
+      "grad_norm": 0.2508833706378937,
+      "learning_rate": 4.814947907788958e-05,
+      "loss": 0.4045,
+      "step": 1735
+    },
+    {
+      "epoch": 1.306736921339857,
+      "grad_norm": 0.25638729333877563,
+      "learning_rate": 4.814652292461901e-05,
+      "loss": 0.4334,
+      "step": 1736
+    },
+    {
+      "epoch": 1.3074896499811817,
+      "grad_norm": 0.24100621044635773,
+      "learning_rate": 4.814356451347573e-05,
+      "loss": 0.3827,
+      "step": 1737
+    },
+    {
+      "epoch": 1.3082423786225066,
+      "grad_norm": 0.20596997439861298,
+      "learning_rate": 4.8140603844783275e-05,
+      "loss": 0.3943,
+      "step": 1738
+    },
+    {
+      "epoch": 1.3089951072638315,
+      "grad_norm": 0.3172217607498169,
+      "learning_rate": 4.8137640918865415e-05,
+      "loss": 0.4853,
+      "step": 1739
+    },
+    {
+      "epoch": 1.3097478359051562,
+      "grad_norm": 0.3050840198993683,
+      "learning_rate": 4.8134675736046176e-05,
+      "loss": 0.3821,
+      "step": 1740
+    },
+    {
+      "epoch": 1.310500564546481,
+      "grad_norm": 0.286487340927124,
+      "learning_rate": 4.813170829664982e-05,
+      "loss": 0.3572,
+      "step": 1741
+    },
+    {
+      "epoch": 1.3112532931878058,
+      "grad_norm": 0.2984033226966858,
+      "learning_rate": 4.812873860100087e-05,
+      "loss": 0.3975,
+      "step": 1742
+    },
+    {
+      "epoch": 1.3120060218291307,
+      "grad_norm": 0.23136569559574127,
+      "learning_rate": 4.812576664942408e-05,
+      "loss": 0.5222,
+      "step": 1743
+    },
+    {
+      "epoch": 1.3127587504704554,
+      "grad_norm": 0.25539660453796387,
+      "learning_rate": 4.8122792442244444e-05,
+      "loss": 0.3287,
+      "step": 1744
+    },
+    {
+      "epoch": 1.3135114791117801,
+      "grad_norm": 0.3180519938468933,
+      "learning_rate": 4.811981597978725e-05,
+      "loss": 0.5237,
+      "step": 1745
+    },
+    {
+      "epoch": 1.314264207753105,
+      "grad_norm": 0.3439962565898895,
+      "learning_rate": 4.811683726237797e-05,
+      "loss": 0.4552,
+      "step": 1746
+    },
+    {
+      "epoch": 1.3150169363944297,
+      "grad_norm": 0.4758007824420929,
+      "learning_rate": 4.8113856290342376e-05,
+      "loss": 0.4861,
+      "step": 1747
+    },
+    {
+      "epoch": 1.3157696650357547,
+      "grad_norm": 0.5402927994728088,
+      "learning_rate": 4.811087306400644e-05,
+      "loss": 0.4101,
+      "step": 1748
+    },
+    {
+      "epoch": 1.3165223936770794,
+      "grad_norm": 0.2702093720436096,
+      "learning_rate": 4.810788758369642e-05,
+      "loss": 0.4119,
+      "step": 1749
+    },
+    {
+      "epoch": 1.3172751223184043,
+      "grad_norm": 0.2811294198036194,
+      "learning_rate": 4.8104899849738795e-05,
+      "loss": 0.4186,
+      "step": 1750
+    },
+    {
+      "epoch": 1.318027850959729,
+      "grad_norm": 0.3253948986530304,
+      "learning_rate": 4.8101909862460296e-05,
+      "loss": 0.348,
+      "step": 1751
+    },
+    {
+      "epoch": 1.3187805796010539,
+      "grad_norm": 0.28159475326538086,
+      "learning_rate": 4.809891762218792e-05,
+      "loss": 0.409,
+      "step": 1752
+    },
+    {
+      "epoch": 1.3195333082423786,
+      "grad_norm": 0.5138587951660156,
+      "learning_rate": 4.809592312924888e-05,
+      "loss": 0.4136,
+      "step": 1753
+    },
+    {
+      "epoch": 1.3202860368837035,
+      "grad_norm": 0.3242422938346863,
+      "learning_rate": 4.8092926383970646e-05,
+      "loss": 0.4528,
+      "step": 1754
+    },
+    {
+      "epoch": 1.3210387655250282,
+      "grad_norm": 0.28211140632629395,
+      "learning_rate": 4.808992738668095e-05,
+      "loss": 0.4767,
+      "step": 1755
+    },
+    {
+      "epoch": 1.321791494166353,
+      "grad_norm": 0.27275869250297546,
+      "learning_rate": 4.808692613770775e-05,
+      "loss": 0.4364,
+      "step": 1756
+    },
+    {
+      "epoch": 1.3225442228076778,
+      "grad_norm": 0.2563418447971344,
+      "learning_rate": 4.808392263737925e-05,
+      "loss": 0.4488,
+      "step": 1757
+    },
+    {
+      "epoch": 1.3232969514490027,
+      "grad_norm": 0.23904260993003845,
+      "learning_rate": 4.808091688602393e-05,
+      "loss": 0.4255,
+      "step": 1758
+    },
+    {
+      "epoch": 1.3240496800903274,
+      "grad_norm": 0.3174615502357483,
+      "learning_rate": 4.807790888397048e-05,
+      "loss": 0.4951,
+      "step": 1759
+    },
+    {
+      "epoch": 1.3248024087316521,
+      "grad_norm": 0.3760294020175934,
+      "learning_rate": 4.807489863154785e-05,
+      "loss": 0.4122,
+      "step": 1760
+    },
+    {
+      "epoch": 1.325555137372977,
+      "grad_norm": 0.2394106388092041,
+      "learning_rate": 4.8071886129085244e-05,
+      "loss": 0.5379,
+      "step": 1761
+    },
+    {
+      "epoch": 1.326307866014302,
+      "grad_norm": 0.4794088900089264,
+      "learning_rate": 4.8068871376912095e-05,
+      "loss": 0.4267,
+      "step": 1762
+    },
+    {
+      "epoch": 1.3270605946556266,
+      "grad_norm": 0.29244256019592285,
+      "learning_rate": 4.80658543753581e-05,
+      "loss": 0.4232,
+      "step": 1763
+    },
+    {
+      "epoch": 1.3278133232969513,
+      "grad_norm": 0.3649486005306244,
+      "learning_rate": 4.806283512475319e-05,
+      "loss": 0.4231,
+      "step": 1764
+    },
+    {
+      "epoch": 1.3285660519382763,
+      "grad_norm": 0.24506047368049622,
+      "learning_rate": 4.8059813625427545e-05,
+      "loss": 0.3707,
+      "step": 1765
+    },
+    {
+      "epoch": 1.3293187805796012,
+      "grad_norm": 0.5567811727523804,
+      "learning_rate": 4.805678987771159e-05,
+      "loss": 0.5078,
+      "step": 1766
+    },
+    {
+      "epoch": 1.3300715092209259,
+      "grad_norm": 0.32808515429496765,
+      "learning_rate": 4.805376388193601e-05,
+      "loss": 0.3601,
+      "step": 1767
+    },
+    {
+      "epoch": 1.3308242378622506,
+      "grad_norm": 0.37941133975982666,
+      "learning_rate": 4.805073563843171e-05,
+      "loss": 0.4908,
+      "step": 1768
+    },
+    {
+      "epoch": 1.3315769665035755,
+      "grad_norm": 0.36102110147476196,
+      "learning_rate": 4.804770514752986e-05,
+      "loss": 0.4236,
+      "step": 1769
+    },
+    {
+      "epoch": 1.3323296951449002,
+      "grad_norm": 0.34539809823036194,
+      "learning_rate": 4.804467240956187e-05,
+      "loss": 0.4836,
+      "step": 1770
+    },
+    {
+      "epoch": 1.333082423786225,
+      "grad_norm": 0.3204517364501953,
+      "learning_rate": 4.804163742485938e-05,
+      "loss": 0.5848,
+      "step": 1771
+    },
+    {
+      "epoch": 1.3338351524275498,
+      "grad_norm": 0.32392236590385437,
+      "learning_rate": 4.803860019375432e-05,
+      "loss": 0.425,
+      "step": 1772
+    },
+    {
+      "epoch": 1.3345878810688747,
+      "grad_norm": 0.3477310240268707,
+      "learning_rate": 4.8035560716578816e-05,
+      "loss": 0.3921,
+      "step": 1773
+    },
+    {
+      "epoch": 1.3353406097101994,
+      "grad_norm": 0.32809314131736755,
+      "learning_rate": 4.803251899366527e-05,
+      "loss": 0.4878,
+      "step": 1774
+    },
+    {
+      "epoch": 1.3360933383515243,
+      "grad_norm": 0.29575666785240173,
+      "learning_rate": 4.802947502534631e-05,
+      "loss": 0.3431,
+      "step": 1775
+    },
+    {
+      "epoch": 1.336846066992849,
+      "grad_norm": 0.3417726159095764,
+      "learning_rate": 4.8026428811954834e-05,
+      "loss": 0.4438,
+      "step": 1776
+    },
+    {
+      "epoch": 1.337598795634174,
+      "grad_norm": 0.3380669057369232,
+      "learning_rate": 4.8023380353823955e-05,
+      "loss": 0.4667,
+      "step": 1777
+    },
+    {
+      "epoch": 1.3383515242754986,
+      "grad_norm": 0.2662280797958374,
+      "learning_rate": 4.8020329651287066e-05,
+      "loss": 0.355,
+      "step": 1778
+    },
+    {
+      "epoch": 1.3391042529168236,
+      "grad_norm": 0.32158663868904114,
+      "learning_rate": 4.8017276704677784e-05,
+      "loss": 0.4637,
+      "step": 1779
+    },
+    {
+      "epoch": 1.3398569815581483,
+      "grad_norm": 0.3010968863964081,
+      "learning_rate": 4.8014221514329955e-05,
+      "loss": 0.487,
+      "step": 1780
+    },
+    {
+      "epoch": 1.3406097101994732,
+      "grad_norm": 0.33339977264404297,
+      "learning_rate": 4.801116408057771e-05,
+      "loss": 0.3881,
+      "step": 1781
+    },
+    {
+      "epoch": 1.3413624388407979,
+      "grad_norm": 0.28095734119415283,
+      "learning_rate": 4.80081044037554e-05,
+      "loss": 0.3771,
+      "step": 1782
+    },
+    {
+      "epoch": 1.3421151674821226,
+      "grad_norm": 0.36039358377456665,
+      "learning_rate": 4.800504248419762e-05,
+      "loss": 0.4428,
+      "step": 1783
+    },
+    {
+      "epoch": 1.3428678961234475,
+      "grad_norm": 0.2893329858779907,
+      "learning_rate": 4.800197832223923e-05,
+      "loss": 0.4336,
+      "step": 1784
+    },
+    {
+      "epoch": 1.3436206247647724,
+      "grad_norm": 0.3461940586566925,
+      "learning_rate": 4.799891191821531e-05,
+      "loss": 0.5297,
+      "step": 1785
+    },
+    {
+      "epoch": 1.344373353406097,
+      "grad_norm": 0.2763758897781372,
+      "learning_rate": 4.79958432724612e-05,
+      "loss": 0.4402,
+      "step": 1786
+    },
+    {
+      "epoch": 1.3451260820474218,
+      "grad_norm": 0.37101924419403076,
+      "learning_rate": 4.79927723853125e-05,
+      "loss": 0.4181,
+      "step": 1787
+    },
+    {
+      "epoch": 1.3458788106887467,
+      "grad_norm": 0.21966445446014404,
+      "learning_rate": 4.798969925710501e-05,
+      "loss": 0.5562,
+      "step": 1788
+    },
+    {
+      "epoch": 1.3466315393300716,
+      "grad_norm": 0.40903809666633606,
+      "learning_rate": 4.798662388817481e-05,
+      "loss": 0.3787,
+      "step": 1789
+    },
+    {
+      "epoch": 1.3473842679713963,
+      "grad_norm": 0.2546432912349701,
+      "learning_rate": 4.7983546278858233e-05,
+      "loss": 0.4552,
+      "step": 1790
+    },
+    {
+      "epoch": 1.348136996612721,
+      "grad_norm": 0.250741183757782,
+      "learning_rate": 4.798046642949183e-05,
+      "loss": 0.5036,
+      "step": 1791
+    },
+    {
+      "epoch": 1.348889725254046,
+      "grad_norm": 0.3019157648086548,
+      "learning_rate": 4.7977384340412406e-05,
+      "loss": 0.4204,
+      "step": 1792
+    },
+    {
+      "epoch": 1.3496424538953709,
+      "grad_norm": 0.24814705550670624,
+      "learning_rate": 4.797430001195702e-05,
+      "loss": 0.46,
+      "step": 1793
+    },
+    {
+      "epoch": 1.3503951825366955,
+      "grad_norm": 0.3248940408229828,
+      "learning_rate": 4.797121344446296e-05,
+      "loss": 0.3758,
+      "step": 1794
+    },
+    {
+      "epoch": 1.3511479111780202,
+      "grad_norm": 0.2421947419643402,
+      "learning_rate": 4.7968124638267784e-05,
+      "loss": 0.3702,
+      "step": 1795
+    },
+    {
+      "epoch": 1.3519006398193452,
+      "grad_norm": 0.23909826576709747,
+      "learning_rate": 4.7965033593709275e-05,
+      "loss": 0.3666,
+      "step": 1796
+    },
+    {
+      "epoch": 1.3526533684606699,
+      "grad_norm": 0.2590198218822479,
+      "learning_rate": 4.7961940311125456e-05,
+      "loss": 0.4595,
+      "step": 1797
+    },
+    {
+      "epoch": 1.3534060971019948,
+      "grad_norm": 0.2789859175682068,
+      "learning_rate": 4.7958844790854604e-05,
+      "loss": 0.4851,
+      "step": 1798
+    },
+    {
+      "epoch": 1.3541588257433195,
+      "grad_norm": 0.2665332555770874,
+      "learning_rate": 4.7955747033235256e-05,
+      "loss": 0.4977,
+      "step": 1799
+    },
+    {
+      "epoch": 1.3549115543846444,
+      "grad_norm": 0.23944230377674103,
+      "learning_rate": 4.795264703860616e-05,
+      "loss": 0.406,
+      "step": 1800
+    },
+    {
+      "epoch": 1.3549115543846444,
+      "eval_loss": 0.4158343970775604,
+      "eval_runtime": 452.6057,
+      "eval_samples_per_second": 21.27,
+      "eval_steps_per_second": 0.665,
+      "step": 1800
+    },
+    {
+      "epoch": 1.355664283025969,
+      "grad_norm": 0.2235507369041443,
+      "learning_rate": 4.7949544807306336e-05,
+      "loss": 0.3068,
+      "step": 1801
+    },
+    {
+      "epoch": 1.356417011667294,
+      "grad_norm": 0.22426214814186096,
+      "learning_rate": 4.794644033967503e-05,
+      "loss": 0.3175,
+      "step": 1802
+    },
+    {
+      "epoch": 1.3571697403086187,
+      "grad_norm": 0.22608070075511932,
+      "learning_rate": 4.794333363605176e-05,
+      "loss": 0.4835,
+      "step": 1803
+    },
+    {
+      "epoch": 1.3579224689499436,
+      "grad_norm": 0.22601571679115295,
+      "learning_rate": 4.794022469677625e-05,
+      "loss": 0.4848,
+      "step": 1804
+    },
+    {
+      "epoch": 1.3586751975912683,
+      "grad_norm": 0.2467537671327591,
+      "learning_rate": 4.79371135221885e-05,
+      "loss": 0.4732,
+      "step": 1805
+    },
+    {
+      "epoch": 1.3594279262325932,
+      "grad_norm": 0.2272249162197113,
+      "learning_rate": 4.793400011262875e-05,
+      "loss": 0.3651,
+      "step": 1806
+    },
+    {
+      "epoch": 1.360180654873918,
+      "grad_norm": 0.2348145842552185,
+      "learning_rate": 4.793088446843747e-05,
+      "loss": 0.4291,
+      "step": 1807
+    },
+    {
+      "epoch": 1.3609333835152428,
+      "grad_norm": 0.20190638303756714,
+      "learning_rate": 4.792776658995538e-05,
+      "loss": 0.3944,
+      "step": 1808
+    },
+    {
+      "epoch": 1.3616861121565675,
+      "grad_norm": 0.2221103459596634,
+      "learning_rate": 4.792464647752344e-05,
+      "loss": 0.4823,
+      "step": 1809
+    },
+    {
+      "epoch": 1.3624388407978922,
+      "grad_norm": 0.2496221959590912,
+      "learning_rate": 4.7921524131482875e-05,
+      "loss": 0.4186,
+      "step": 1810
+    },
+    {
+      "epoch": 1.3631915694392172,
+      "grad_norm": 0.2144601196050644,
+      "learning_rate": 4.791839955217513e-05,
+      "loss": 0.3098,
+      "step": 1811
+    },
+    {
+      "epoch": 1.363944298080542,
+      "grad_norm": 0.3351675570011139,
+      "learning_rate": 4.7915272739941916e-05,
+      "loss": 0.4732,
+      "step": 1812
+    },
+    {
+      "epoch": 1.3646970267218668,
+      "grad_norm": 0.3107703924179077,
+      "learning_rate": 4.7912143695125167e-05,
+      "loss": 0.3944,
+      "step": 1813
+    },
+    {
+      "epoch": 1.3654497553631915,
+      "grad_norm": 0.23332947492599487,
+      "learning_rate": 4.790901241806707e-05,
+      "loss": 0.3731,
+      "step": 1814
+    },
+    {
+      "epoch": 1.3662024840045164,
+      "grad_norm": 0.3760455548763275,
+      "learning_rate": 4.790587890911006e-05,
+      "loss": 0.3866,
+      "step": 1815
+    },
+    {
+      "epoch": 1.3669552126458413,
+      "grad_norm": 0.26869645714759827,
+      "learning_rate": 4.7902743168596815e-05,
+      "loss": 0.4552,
+      "step": 1816
+    },
+    {
+      "epoch": 1.367707941287166,
+      "grad_norm": 0.2312639057636261,
+      "learning_rate": 4.789960519687025e-05,
+      "loss": 0.3608,
+      "step": 1817
+    },
+    {
+      "epoch": 1.3684606699284907,
+      "grad_norm": 0.28747907280921936,
+      "learning_rate": 4.789646499427353e-05,
+      "loss": 0.4329,
+      "step": 1818
+    },
+    {
+      "epoch": 1.3692133985698156,
+      "grad_norm": 0.39789465069770813,
+      "learning_rate": 4.789332256115007e-05,
+      "loss": 0.46,
+      "step": 1819
+    },
+    {
+      "epoch": 1.3699661272111403,
+      "grad_norm": 0.29374760389328003,
+      "learning_rate": 4.78901778978435e-05,
+      "loss": 0.4546,
+      "step": 1820
+    },
+    {
+      "epoch": 1.3707188558524652,
+      "grad_norm": 0.25589150190353394,
+      "learning_rate": 4.788703100469775e-05,
+      "loss": 0.3488,
+      "step": 1821
+    },
+    {
+      "epoch": 1.37147158449379,
+      "grad_norm": 0.2927248477935791,
+      "learning_rate": 4.788388188205693e-05,
+      "loss": 0.4416,
+      "step": 1822
+    },
+    {
+      "epoch": 1.3722243131351148,
+      "grad_norm": 0.3198675811290741,
+      "learning_rate": 4.788073053026544e-05,
+      "loss": 0.5634,
+      "step": 1823
+    },
+    {
+      "epoch": 1.3729770417764395,
+      "grad_norm": 0.2885627746582031,
+      "learning_rate": 4.7877576949667894e-05,
+      "loss": 0.453,
+      "step": 1824
+    },
+    {
+      "epoch": 1.3737297704177645,
+      "grad_norm": 0.3217317759990692,
+      "learning_rate": 4.787442114060918e-05,
+      "loss": 0.3865,
+      "step": 1825
+    },
+    {
+      "epoch": 1.3744824990590891,
+      "grad_norm": 0.2417670488357544,
+      "learning_rate": 4.7871263103434394e-05,
+      "loss": 0.4032,
+      "step": 1826
+    },
+    {
+      "epoch": 1.375235227700414,
+      "grad_norm": 0.23315630853176117,
+      "learning_rate": 4.78681028384889e-05,
+      "loss": 0.4152,
+      "step": 1827
+    },
+    {
+      "epoch": 1.3759879563417388,
+      "grad_norm": 0.2508890926837921,
+      "learning_rate": 4.786494034611831e-05,
+      "loss": 0.4908,
+      "step": 1828
+    },
+    {
+      "epoch": 1.3767406849830637,
+      "grad_norm": 0.2880721390247345,
+      "learning_rate": 4.786177562666846e-05,
+      "loss": 0.4518,
+      "step": 1829
+    },
+    {
+      "epoch": 1.3774934136243884,
+      "grad_norm": 0.2417977899312973,
+      "learning_rate": 4.7858608680485444e-05,
+      "loss": 0.336,
+      "step": 1830
+    },
+    {
+      "epoch": 1.3782461422657133,
+      "grad_norm": 0.2694697380065918,
+      "learning_rate": 4.7855439507915585e-05,
+      "loss": 0.4313,
+      "step": 1831
+    },
+    {
+      "epoch": 1.378998870907038,
+      "grad_norm": 0.22950713336467743,
+      "learning_rate": 4.785226810930546e-05,
+      "loss": 0.3683,
+      "step": 1832
+    },
+    {
+      "epoch": 1.3797515995483627,
+      "grad_norm": 0.2668108344078064,
+      "learning_rate": 4.784909448500191e-05,
+      "loss": 0.4861,
+      "step": 1833
+    },
+    {
+      "epoch": 1.3805043281896876,
+      "grad_norm": 0.2697150409221649,
+      "learning_rate": 4.784591863535198e-05,
+      "loss": 0.5101,
+      "step": 1834
+    },
+    {
+      "epoch": 1.3812570568310125,
+      "grad_norm": 0.33226901292800903,
+      "learning_rate": 4.7842740560702964e-05,
+      "loss": 0.4246,
+      "step": 1835
+    },
+    {
+      "epoch": 1.3820097854723372,
+      "grad_norm": 0.23642519116401672,
+      "learning_rate": 4.783956026140243e-05,
+      "loss": 0.3952,
+      "step": 1836
+    },
+    {
+      "epoch": 1.382762514113662,
+      "grad_norm": 0.237926185131073,
+      "learning_rate": 4.7836377737798174e-05,
+      "loss": 0.4597,
+      "step": 1837
+    },
+    {
+      "epoch": 1.3835152427549868,
+      "grad_norm": 0.2427898496389389,
+      "learning_rate": 4.783319299023821e-05,
+      "loss": 0.4404,
+      "step": 1838
+    },
+    {
+      "epoch": 1.3842679713963117,
+      "grad_norm": 0.2672206461429596,
+      "learning_rate": 4.783000601907084e-05,
+      "loss": 0.4267,
+      "step": 1839
+    },
+    {
+      "epoch": 1.3850207000376364,
+      "grad_norm": 0.22358034551143646,
+      "learning_rate": 4.782681682464459e-05,
+      "loss": 0.4855,
+      "step": 1840
+    },
+    {
+      "epoch": 1.3857734286789611,
+      "grad_norm": 0.2850174903869629,
+      "learning_rate": 4.7823625407308196e-05,
+      "loss": 0.4351,
+      "step": 1841
+    },
+    {
+      "epoch": 1.386526157320286,
+      "grad_norm": 0.275923490524292,
+      "learning_rate": 4.782043176741068e-05,
+      "loss": 0.439,
+      "step": 1842
+    },
+    {
+      "epoch": 1.387278885961611,
+      "grad_norm": 0.2660992443561554,
+      "learning_rate": 4.781723590530131e-05,
+      "loss": 0.5178,
+      "step": 1843
+    },
+    {
+      "epoch": 1.3880316146029357,
+      "grad_norm": 0.24524115025997162,
+      "learning_rate": 4.7814037821329564e-05,
+      "loss": 0.4551,
+      "step": 1844
+    },
+    {
+      "epoch": 1.3887843432442604,
+      "grad_norm": 0.20587483048439026,
+      "learning_rate": 4.781083751584518e-05,
+      "loss": 0.4318,
+      "step": 1845
+    },
+    {
+      "epoch": 1.3895370718855853,
+      "grad_norm": 0.2641163766384125,
+      "learning_rate": 4.780763498919815e-05,
+      "loss": 0.4383,
+      "step": 1846
+    },
+    {
+      "epoch": 1.39028980052691,
+      "grad_norm": 0.2557491064071655,
+      "learning_rate": 4.780443024173868e-05,
+      "loss": 0.4476,
+      "step": 1847
+    },
+    {
+      "epoch": 1.391042529168235,
+      "grad_norm": 0.23952797055244446,
+      "learning_rate": 4.7801223273817246e-05,
+      "loss": 0.4399,
+      "step": 1848
+    },
+    {
+      "epoch": 1.3917952578095596,
+      "grad_norm": 0.25892484188079834,
+      "learning_rate": 4.7798014085784567e-05,
+      "loss": 0.3544,
+      "step": 1849
+    },
+    {
+      "epoch": 1.3925479864508845,
+      "grad_norm": 0.23997732996940613,
+      "learning_rate": 4.779480267799158e-05,
+      "loss": 0.5099,
+      "step": 1850
+    },
+    {
+      "epoch": 1.3933007150922092,
+      "grad_norm": 0.25026360154151917,
+      "learning_rate": 4.7791589050789486e-05,
+      "loss": 0.3893,
+      "step": 1851
+    },
+    {
+      "epoch": 1.3940534437335341,
+      "grad_norm": 0.2612903416156769,
+      "learning_rate": 4.7788373204529716e-05,
+      "loss": 0.4436,
+      "step": 1852
+    },
+    {
+      "epoch": 1.3948061723748588,
+      "grad_norm": 0.24211107194423676,
+      "learning_rate": 4.778515513956396e-05,
+      "loss": 0.5223,
+      "step": 1853
+    },
+    {
+      "epoch": 1.3955589010161837,
+      "grad_norm": 0.28450435400009155,
+      "learning_rate": 4.7781934856244136e-05,
+      "loss": 0.4495,
+      "step": 1854
+    },
+    {
+      "epoch": 1.3963116296575084,
+      "grad_norm": 0.19208616018295288,
+      "learning_rate": 4.777871235492241e-05,
+      "loss": 0.3647,
+      "step": 1855
+    },
+    {
+      "epoch": 1.3970643582988334,
+      "grad_norm": 0.2199915647506714,
+      "learning_rate": 4.7775487635951186e-05,
+      "loss": 0.5124,
+      "step": 1856
+    },
+    {
+      "epoch": 1.397817086940158,
+      "grad_norm": 0.21090516448020935,
+      "learning_rate": 4.777226069968312e-05,
+      "loss": 0.3363,
+      "step": 1857
+    },
+    {
+      "epoch": 1.398569815581483,
+      "grad_norm": 0.23949627578258514,
+      "learning_rate": 4.7769031546471106e-05,
+      "loss": 0.4233,
+      "step": 1858
+    },
+    {
+      "epoch": 1.3993225442228077,
+      "grad_norm": 0.2613455355167389,
+      "learning_rate": 4.776580017666827e-05,
+      "loss": 0.4386,
+      "step": 1859
+    },
+    {
+      "epoch": 1.4000752728641324,
+      "grad_norm": 0.20716920495033264,
+      "learning_rate": 4.7762566590628006e-05,
+      "loss": 0.3185,
+      "step": 1860
+    },
+    {
+      "epoch": 1.4008280015054573,
+      "grad_norm": 0.22486980259418488,
+      "learning_rate": 4.775933078870391e-05,
+      "loss": 0.3163,
+      "step": 1861
+    },
+    {
+      "epoch": 1.4015807301467822,
+      "grad_norm": 0.3022227883338928,
+      "learning_rate": 4.7756092771249865e-05,
+      "loss": 0.3877,
+      "step": 1862
+    },
+    {
+      "epoch": 1.402333458788107,
+      "grad_norm": 0.2707027196884155,
+      "learning_rate": 4.7752852538619974e-05,
+      "loss": 0.4177,
+      "step": 1863
+    },
+    {
+      "epoch": 1.4030861874294316,
+      "grad_norm": 0.2699073851108551,
+      "learning_rate": 4.774961009116857e-05,
+      "loss": 0.4338,
+      "step": 1864
+    },
+    {
+      "epoch": 1.4038389160707565,
+      "grad_norm": 0.2917603552341461,
+      "learning_rate": 4.7746365429250255e-05,
+      "loss": 0.4015,
+      "step": 1865
+    },
+    {
+      "epoch": 1.4045916447120814,
+      "grad_norm": 0.2711542248725891,
+      "learning_rate": 4.774311855321986e-05,
+      "loss": 0.4522,
+      "step": 1866
+    },
+    {
+      "epoch": 1.4053443733534061,
+      "grad_norm": 0.27297964692115784,
+      "learning_rate": 4.773986946343245e-05,
+      "loss": 0.5292,
+      "step": 1867
+    },
+    {
+      "epoch": 1.4060971019947308,
+      "grad_norm": 0.2511192858219147,
+      "learning_rate": 4.773661816024335e-05,
+      "loss": 0.3412,
+      "step": 1868
+    },
+    {
+      "epoch": 1.4068498306360557,
+      "grad_norm": 0.30797258019447327,
+      "learning_rate": 4.773336464400811e-05,
+      "loss": 0.3906,
+      "step": 1869
+    },
+    {
+      "epoch": 1.4076025592773804,
+      "grad_norm": 0.24745629727840424,
+      "learning_rate": 4.773010891508254e-05,
+      "loss": 0.4001,
+      "step": 1870
+    },
+    {
+      "epoch": 1.4083552879187053,
+      "grad_norm": 0.26375365257263184,
+      "learning_rate": 4.772685097382267e-05,
+      "loss": 0.4198,
+      "step": 1871
+    },
+    {
+      "epoch": 1.40910801656003,
+      "grad_norm": 0.2448989301919937,
+      "learning_rate": 4.772359082058478e-05,
+      "loss": 0.4722,
+      "step": 1872
+    },
+    {
+      "epoch": 1.409860745201355,
+      "grad_norm": 0.2702071964740753,
+      "learning_rate": 4.772032845572541e-05,
+      "loss": 0.3562,
+      "step": 1873
+    },
+    {
+      "epoch": 1.4106134738426797,
+      "grad_norm": 0.23285624384880066,
+      "learning_rate": 4.771706387960132e-05,
+      "loss": 0.3707,
+      "step": 1874
+    },
+    {
+      "epoch": 1.4113662024840046,
+      "grad_norm": 0.28724405169487,
+      "learning_rate": 4.771379709256953e-05,
+      "loss": 0.4737,
+      "step": 1875
+    },
+    {
+      "epoch": 1.4121189311253293,
+      "grad_norm": 0.3606455326080322,
+      "learning_rate": 4.7710528094987264e-05,
+      "loss": 0.3695,
+      "step": 1876
+    },
+    {
+      "epoch": 1.4128716597666542,
+      "grad_norm": 0.2405892163515091,
+      "learning_rate": 4.7707256887212046e-05,
+      "loss": 0.381,
+      "step": 1877
+    },
+    {
+      "epoch": 1.4136243884079789,
+      "grad_norm": 0.30289211869239807,
+      "learning_rate": 4.770398346960159e-05,
+      "loss": 0.4173,
+      "step": 1878
+    },
+    {
+      "epoch": 1.4143771170493038,
+      "grad_norm": 0.2286311835050583,
+      "learning_rate": 4.770070784251387e-05,
+      "loss": 0.4468,
+      "step": 1879
+    },
+    {
+      "epoch": 1.4151298456906285,
+      "grad_norm": 0.31470707058906555,
+      "learning_rate": 4.7697430006307126e-05,
+      "loss": 0.4556,
+      "step": 1880
+    },
+    {
+      "epoch": 1.4158825743319534,
+      "grad_norm": 0.2801443934440613,
+      "learning_rate": 4.76941499613398e-05,
+      "loss": 0.3193,
+      "step": 1881
+    },
+    {
+      "epoch": 1.416635302973278,
+      "grad_norm": 0.2869381606578827,
+      "learning_rate": 4.769086770797059e-05,
+      "loss": 0.3147,
+      "step": 1882
+    },
+    {
+      "epoch": 1.4173880316146028,
+      "grad_norm": 0.31108394265174866,
+      "learning_rate": 4.768758324655845e-05,
+      "loss": 0.4601,
+      "step": 1883
+    },
+    {
+      "epoch": 1.4181407602559277,
+      "grad_norm": 0.22520123422145844,
+      "learning_rate": 4.768429657746255e-05,
+      "loss": 0.3465,
+      "step": 1884
+    },
+    {
+      "epoch": 1.4188934888972526,
+      "grad_norm": 0.23259632289409637,
+      "learning_rate": 4.768100770104233e-05,
+      "loss": 0.4102,
+      "step": 1885
+    },
+    {
+      "epoch": 1.4196462175385773,
+      "grad_norm": 0.29215100407600403,
+      "learning_rate": 4.7677716617657445e-05,
+      "loss": 0.4612,
+      "step": 1886
+    },
+    {
+      "epoch": 1.420398946179902,
+      "grad_norm": 0.28146058320999146,
+      "learning_rate": 4.767442332766781e-05,
+      "loss": 0.3379,
+      "step": 1887
+    },
+    {
+      "epoch": 1.421151674821227,
+      "grad_norm": 0.2877790331840515,
+      "learning_rate": 4.7671127831433574e-05,
+      "loss": 0.2712,
+      "step": 1888
+    },
+    {
+      "epoch": 1.4219044034625519,
+      "grad_norm": 0.28199470043182373,
+      "learning_rate": 4.766783012931512e-05,
+      "loss": 0.4227,
+      "step": 1889
+    },
+    {
+      "epoch": 1.4226571321038766,
+      "grad_norm": 0.2780182659626007,
+      "learning_rate": 4.766453022167308e-05,
+      "loss": 0.3747,
+      "step": 1890
+    },
+    {
+      "epoch": 1.4234098607452013,
+      "grad_norm": 0.2877577543258667,
+      "learning_rate": 4.766122810886834e-05,
+      "loss": 0.3424,
+      "step": 1891
+    },
+    {
+      "epoch": 1.4241625893865262,
+      "grad_norm": 0.24384111166000366,
+      "learning_rate": 4.765792379126201e-05,
+      "loss": 0.4214,
+      "step": 1892
+    },
+    {
+      "epoch": 1.424915318027851,
+      "grad_norm": 0.2924976944923401,
+      "learning_rate": 4.765461726921543e-05,
+      "loss": 0.478,
+      "step": 1893
+    },
+    {
+      "epoch": 1.4256680466691758,
+      "grad_norm": 0.2807604968547821,
+      "learning_rate": 4.7651308543090215e-05,
+      "loss": 0.5507,
+      "step": 1894
+    },
+    {
+      "epoch": 1.4264207753105005,
+      "grad_norm": 0.2368244230747223,
+      "learning_rate": 4.7647997613248195e-05,
+      "loss": 0.356,
+      "step": 1895
+    },
+    {
+      "epoch": 1.4271735039518254,
+      "grad_norm": 0.2535645067691803,
+      "learning_rate": 4.764468448005144e-05,
+      "loss": 0.4267,
+      "step": 1896
+    },
+    {
+      "epoch": 1.42792623259315,
+      "grad_norm": 0.23013842105865479,
+      "learning_rate": 4.764136914386229e-05,
+      "loss": 0.434,
+      "step": 1897
+    },
+    {
+      "epoch": 1.428678961234475,
+      "grad_norm": 0.29250895977020264,
+      "learning_rate": 4.7638051605043284e-05,
+      "loss": 0.3994,
+      "step": 1898
+    },
+    {
+      "epoch": 1.4294316898757997,
+      "grad_norm": 0.2755577266216278,
+      "learning_rate": 4.763473186395724e-05,
+      "loss": 0.2847,
+      "step": 1899
+    },
+    {
+      "epoch": 1.4301844185171246,
+      "grad_norm": 0.23972149193286896,
+      "learning_rate": 4.763140992096718e-05,
+      "loss": 0.4769,
+      "step": 1900
+    },
+    {
+      "epoch": 1.4309371471584493,
+      "grad_norm": 0.2506982684135437,
+      "learning_rate": 4.76280857764364e-05,
+      "loss": 0.3108,
+      "step": 1901
+    },
+    {
+      "epoch": 1.4316898757997742,
+      "grad_norm": 0.24815982580184937,
+      "learning_rate": 4.762475943072843e-05,
+      "loss": 0.4853,
+      "step": 1902
+    },
+    {
+      "epoch": 1.432442604441099,
+      "grad_norm": 0.27620357275009155,
+      "learning_rate": 4.762143088420702e-05,
+      "loss": 0.4308,
+      "step": 1903
+    },
+    {
+      "epoch": 1.4331953330824239,
+      "grad_norm": 0.24741479754447937,
+      "learning_rate": 4.7618100137236174e-05,
+      "loss": 0.403,
+      "step": 1904
+    },
+    {
+      "epoch": 1.4339480617237486,
+      "grad_norm": 0.2166122943162918,
+      "learning_rate": 4.761476719018016e-05,
+      "loss": 0.4424,
+      "step": 1905
+    },
+    {
+      "epoch": 1.4347007903650733,
+      "grad_norm": 0.2830376625061035,
+      "learning_rate": 4.7611432043403436e-05,
+      "loss": 0.3548,
+      "step": 1906
+    },
+    {
+      "epoch": 1.4354535190063982,
+      "grad_norm": 0.3047950267791748,
+      "learning_rate": 4.7608094697270745e-05,
+      "loss": 0.3831,
+      "step": 1907
+    },
+    {
+      "epoch": 1.436206247647723,
+      "grad_norm": 0.2667747139930725,
+      "learning_rate": 4.7604755152147045e-05,
+      "loss": 0.3933,
+      "step": 1908
+    },
+    {
+      "epoch": 1.4369589762890478,
+      "grad_norm": 0.25192126631736755,
+      "learning_rate": 4.760141340839756e-05,
+      "loss": 0.4054,
+      "step": 1909
+    },
+    {
+      "epoch": 1.4377117049303725,
+      "grad_norm": 0.27835485339164734,
+      "learning_rate": 4.759806946638772e-05,
+      "loss": 0.4301,
+      "step": 1910
+    },
+    {
+      "epoch": 1.4384644335716974,
+      "grad_norm": 0.367883563041687,
+      "learning_rate": 4.7594723326483215e-05,
+      "loss": 0.4105,
+      "step": 1911
+    },
+    {
+      "epoch": 1.4392171622130223,
+      "grad_norm": 0.2849632799625397,
+      "learning_rate": 4.759137498904999e-05,
+      "loss": 0.5459,
+      "step": 1912
+    },
+    {
+      "epoch": 1.439969890854347,
+      "grad_norm": 0.34697791934013367,
+      "learning_rate": 4.758802445445419e-05,
+      "loss": 0.4272,
+      "step": 1913
+    },
+    {
+      "epoch": 1.4407226194956717,
+      "grad_norm": 0.36664268374443054,
+      "learning_rate": 4.758467172306225e-05,
+      "loss": 0.3896,
+      "step": 1914
+    },
+    {
+      "epoch": 1.4414753481369966,
+      "grad_norm": 0.47121724486351013,
+      "learning_rate": 4.758131679524081e-05,
+      "loss": 0.4438,
+      "step": 1915
+    },
+    {
+      "epoch": 1.4422280767783215,
+      "grad_norm": 0.3056446611881256,
+      "learning_rate": 4.7577959671356734e-05,
+      "loss": 0.3119,
+      "step": 1916
+    },
+    {
+      "epoch": 1.4429808054196462,
+      "grad_norm": 0.4048618972301483,
+      "learning_rate": 4.75746003517772e-05,
+      "loss": 0.4606,
+      "step": 1917
+    },
+    {
+      "epoch": 1.443733534060971,
+      "grad_norm": 0.37834933400154114,
+      "learning_rate": 4.7571238836869535e-05,
+      "loss": 0.4122,
+      "step": 1918
+    },
+    {
+      "epoch": 1.4444862627022959,
+      "grad_norm": 0.35082554817199707,
+      "learning_rate": 4.7567875127001376e-05,
+      "loss": 0.3453,
+      "step": 1919
+    },
+    {
+      "epoch": 1.4452389913436205,
+      "grad_norm": 0.32709914445877075,
+      "learning_rate": 4.7564509222540566e-05,
+      "loss": 0.3945,
+      "step": 1920
+    },
+    {
+      "epoch": 1.4459917199849455,
+      "grad_norm": 0.27086612582206726,
+      "learning_rate": 4.756114112385519e-05,
+      "loss": 0.322,
+      "step": 1921
+    },
+    {
+      "epoch": 1.4467444486262702,
+      "grad_norm": 0.2779649794101715,
+      "learning_rate": 4.7557770831313586e-05,
+      "loss": 0.5233,
+      "step": 1922
+    },
+    {
+      "epoch": 1.447497177267595,
+      "grad_norm": 0.2597295045852661,
+      "learning_rate": 4.7554398345284315e-05,
+      "loss": 0.3906,
+      "step": 1923
+    },
+    {
+      "epoch": 1.4482499059089198,
+      "grad_norm": 0.24280290305614471,
+      "learning_rate": 4.7551023666136195e-05,
+      "loss": 0.3975,
+      "step": 1924
+    },
+    {
+      "epoch": 1.4490026345502447,
+      "grad_norm": 0.2593470513820648,
+      "learning_rate": 4.7547646794238276e-05,
+      "loss": 0.2591,
+      "step": 1925
+    },
+    {
+      "epoch": 1.4497553631915694,
+      "grad_norm": 0.2744225859642029,
+      "learning_rate": 4.754426772995985e-05,
+      "loss": 0.3936,
+      "step": 1926
+    },
+    {
+      "epoch": 1.4505080918328943,
+      "grad_norm": 0.28252989053726196,
+      "learning_rate": 4.754088647367043e-05,
+      "loss": 0.3684,
+      "step": 1927
+    },
+    {
+      "epoch": 1.451260820474219,
+      "grad_norm": 0.30839794874191284,
+      "learning_rate": 4.75375030257398e-05,
+      "loss": 0.3887,
+      "step": 1928
+    },
+    {
+      "epoch": 1.452013549115544,
+      "grad_norm": 0.28060680627822876,
+      "learning_rate": 4.7534117386537966e-05,
+      "loss": 0.356,
+      "step": 1929
+    },
+    {
+      "epoch": 1.4527662777568686,
+      "grad_norm": 0.23066318035125732,
+      "learning_rate": 4.753072955643517e-05,
+      "loss": 0.429,
+      "step": 1930
+    },
+    {
+      "epoch": 1.4535190063981935,
+      "grad_norm": 0.24543917179107666,
+      "learning_rate": 4.7527339535801914e-05,
+      "loss": 0.3976,
+      "step": 1931
+    },
+    {
+      "epoch": 1.4542717350395182,
+      "grad_norm": 0.20563377439975739,
+      "learning_rate": 4.7523947325008916e-05,
+      "loss": 0.4086,
+      "step": 1932
+    },
+    {
+      "epoch": 1.455024463680843,
+      "grad_norm": 0.2709653079509735,
+      "learning_rate": 4.752055292442714e-05,
+      "loss": 0.4142,
+      "step": 1933
+    },
+    {
+      "epoch": 1.4557771923221678,
+      "grad_norm": 0.2726103663444519,
+      "learning_rate": 4.75171563344278e-05,
+      "loss": 0.35,
+      "step": 1934
+    },
+    {
+      "epoch": 1.4565299209634928,
+      "grad_norm": 0.2658027112483978,
+      "learning_rate": 4.751375755538234e-05,
+      "loss": 0.4441,
+      "step": 1935
+    },
+    {
+      "epoch": 1.4572826496048175,
+      "grad_norm": 0.23752526938915253,
+      "learning_rate": 4.751035658766244e-05,
+      "loss": 0.3469,
+      "step": 1936
+    },
+    {
+      "epoch": 1.4580353782461422,
+      "grad_norm": 0.2185560017824173,
+      "learning_rate": 4.7506953431640024e-05,
+      "loss": 0.4218,
+      "step": 1937
+    },
+    {
+      "epoch": 1.458788106887467,
+      "grad_norm": 0.23909461498260498,
+      "learning_rate": 4.7503548087687264e-05,
+      "loss": 0.3726,
+      "step": 1938
+    },
+    {
+      "epoch": 1.459540835528792,
+      "grad_norm": 0.25888824462890625,
+      "learning_rate": 4.7500140556176564e-05,
+      "loss": 0.5146,
+      "step": 1939
+    },
+    {
+      "epoch": 1.4602935641701167,
+      "grad_norm": 0.28061822056770325,
+      "learning_rate": 4.7496730837480554e-05,
+      "loss": 0.4445,
+      "step": 1940
+    },
+    {
+      "epoch": 1.4610462928114414,
+      "grad_norm": 0.2435227781534195,
+      "learning_rate": 4.7493318931972135e-05,
+      "loss": 0.4968,
+      "step": 1941
+    },
+    {
+      "epoch": 1.4617990214527663,
+      "grad_norm": 0.23771744966506958,
+      "learning_rate": 4.748990484002442e-05,
+      "loss": 0.3887,
+      "step": 1942
+    },
+    {
+      "epoch": 1.4625517500940912,
+      "grad_norm": 0.24737682938575745,
+      "learning_rate": 4.748648856201076e-05,
+      "loss": 0.5345,
+      "step": 1943
+    },
+    {
+      "epoch": 1.463304478735416,
+      "grad_norm": 0.26452207565307617,
+      "learning_rate": 4.748307009830476e-05,
+      "loss": 0.4006,
+      "step": 1944
+    },
+    {
+      "epoch": 1.4640572073767406,
+      "grad_norm": 0.2284427434206009,
+      "learning_rate": 4.747964944928025e-05,
+      "loss": 0.3537,
+      "step": 1945
+    },
+    {
+      "epoch": 1.4648099360180655,
+      "grad_norm": 0.41457515954971313,
+      "learning_rate": 4.7476226615311325e-05,
+      "loss": 0.5675,
+      "step": 1946
+    },
+    {
+      "epoch": 1.4655626646593902,
+      "grad_norm": 0.3799053132534027,
+      "learning_rate": 4.7472801596772286e-05,
+      "loss": 0.3346,
+      "step": 1947
+    },
+    {
+      "epoch": 1.4663153933007151,
+      "grad_norm": 0.29767996072769165,
+      "learning_rate": 4.74693743940377e-05,
+      "loss": 0.4181,
+      "step": 1948
+    },
+    {
+      "epoch": 1.4670681219420398,
+      "grad_norm": 0.29820775985717773,
+      "learning_rate": 4.746594500748235e-05,
+      "loss": 0.3975,
+      "step": 1949
+    },
+    {
+      "epoch": 1.4678208505833648,
+      "grad_norm": 0.362165242433548,
+      "learning_rate": 4.7462513437481266e-05,
+      "loss": 0.375,
+      "step": 1950
+    },
+    {
+      "epoch": 1.4685735792246895,
+      "grad_norm": 0.2877495288848877,
+      "learning_rate": 4.745907968440973e-05,
+      "loss": 0.4946,
+      "step": 1951
+    },
+    {
+      "epoch": 1.4693263078660144,
+      "grad_norm": 0.3169662356376648,
+      "learning_rate": 4.745564374864325e-05,
+      "loss": 0.3856,
+      "step": 1952
+    },
+    {
+      "epoch": 1.470079036507339,
+      "grad_norm": 0.2894074320793152,
+      "learning_rate": 4.745220563055758e-05,
+      "loss": 0.4562,
+      "step": 1953
+    },
+    {
+      "epoch": 1.470831765148664,
+      "grad_norm": 0.33665189146995544,
+      "learning_rate": 4.744876533052869e-05,
+      "loss": 0.3366,
+      "step": 1954
+    },
+    {
+      "epoch": 1.4715844937899887,
+      "grad_norm": 0.3115478456020355,
+      "learning_rate": 4.744532284893281e-05,
+      "loss": 0.4986,
+      "step": 1955
+    },
+    {
+      "epoch": 1.4723372224313134,
+      "grad_norm": 0.26671457290649414,
+      "learning_rate": 4.7441878186146424e-05,
+      "loss": 0.3455,
+      "step": 1956
+    },
+    {
+      "epoch": 1.4730899510726383,
+      "grad_norm": 0.30009737610816956,
+      "learning_rate": 4.743843134254621e-05,
+      "loss": 0.4778,
+      "step": 1957
+    },
+    {
+      "epoch": 1.4738426797139632,
+      "grad_norm": 0.32723456621170044,
+      "learning_rate": 4.7434982318509135e-05,
+      "loss": 0.5105,
+      "step": 1958
+    },
+    {
+      "epoch": 1.474595408355288,
+      "grad_norm": 0.3583339750766754,
+      "learning_rate": 4.743153111441235e-05,
+      "loss": 0.4385,
+      "step": 1959
+    },
+    {
+      "epoch": 1.4753481369966126,
+      "grad_norm": 0.2802211344242096,
+      "learning_rate": 4.74280777306333e-05,
+      "loss": 0.375,
+      "step": 1960
+    },
+    {
+      "epoch": 1.4761008656379375,
+      "grad_norm": 0.2641858756542206,
+      "learning_rate": 4.742462216754963e-05,
+      "loss": 0.4059,
+      "step": 1961
+    },
+    {
+      "epoch": 1.4768535942792624,
+      "grad_norm": 0.2720472812652588,
+      "learning_rate": 4.742116442553923e-05,
+      "loss": 0.3352,
+      "step": 1962
+    },
+    {
+      "epoch": 1.4776063229205871,
+      "grad_norm": 0.2691987454891205,
+      "learning_rate": 4.741770450498024e-05,
+      "loss": 0.4212,
+      "step": 1963
+    },
+    {
+      "epoch": 1.4783590515619118,
+      "grad_norm": 0.26934435963630676,
+      "learning_rate": 4.741424240625104e-05,
+      "loss": 0.4268,
+      "step": 1964
+    },
+    {
+      "epoch": 1.4791117802032367,
+      "grad_norm": 0.26060840487480164,
+      "learning_rate": 4.741077812973022e-05,
+      "loss": 0.3759,
+      "step": 1965
+    },
+    {
+      "epoch": 1.4798645088445617,
+      "grad_norm": 0.2875326871871948,
+      "learning_rate": 4.7407311675796644e-05,
+      "loss": 0.4186,
+      "step": 1966
+    },
+    {
+      "epoch": 1.4806172374858864,
+      "grad_norm": 0.25397130846977234,
+      "learning_rate": 4.74038430448294e-05,
+      "loss": 0.413,
+      "step": 1967
+    },
+    {
+      "epoch": 1.481369966127211,
+      "grad_norm": 0.21964526176452637,
+      "learning_rate": 4.7400372237207793e-05,
+      "loss": 0.3846,
+      "step": 1968
+    },
+    {
+      "epoch": 1.482122694768536,
+      "grad_norm": 0.2488030344247818,
+      "learning_rate": 4.739689925331141e-05,
+      "loss": 0.4864,
+      "step": 1969
+    },
+    {
+      "epoch": 1.4828754234098607,
+      "grad_norm": 0.24984131753444672,
+      "learning_rate": 4.739342409352003e-05,
+      "loss": 0.3713,
+      "step": 1970
+    },
+    {
+      "epoch": 1.4836281520511856,
+      "grad_norm": 0.29343971610069275,
+      "learning_rate": 4.738994675821371e-05,
+      "loss": 0.3886,
+      "step": 1971
+    },
+    {
+      "epoch": 1.4843808806925103,
+      "grad_norm": 0.2671203911304474,
+      "learning_rate": 4.738646724777271e-05,
+      "loss": 0.435,
+      "step": 1972
+    },
+    {
+      "epoch": 1.4851336093338352,
+      "grad_norm": 0.26914241909980774,
+      "learning_rate": 4.738298556257755e-05,
+      "loss": 0.4063,
+      "step": 1973
+    },
+    {
+      "epoch": 1.48588633797516,
+      "grad_norm": 0.24846796691417694,
+      "learning_rate": 4.737950170300898e-05,
+      "loss": 0.3599,
+      "step": 1974
+    },
+    {
+      "epoch": 1.4866390666164848,
+      "grad_norm": 0.2841094732284546,
+      "learning_rate": 4.7376015669448e-05,
+      "loss": 0.349,
+      "step": 1975
+    },
+    {
+      "epoch": 1.4873917952578095,
+      "grad_norm": 0.23203669488430023,
+      "learning_rate": 4.7372527462275815e-05,
+      "loss": 0.3778,
+      "step": 1976
+    },
+    {
+      "epoch": 1.4881445238991344,
+      "grad_norm": 0.26892000436782837,
+      "learning_rate": 4.736903708187392e-05,
+      "loss": 0.3831,
+      "step": 1977
+    },
+    {
+      "epoch": 1.4888972525404591,
+      "grad_norm": 0.29945921897888184,
+      "learning_rate": 4.736554452862399e-05,
+      "loss": 0.4013,
+      "step": 1978
+    },
+    {
+      "epoch": 1.489649981181784,
+      "grad_norm": 0.35003694891929626,
+      "learning_rate": 4.736204980290799e-05,
+      "loss": 0.4344,
+      "step": 1979
+    },
+    {
+      "epoch": 1.4904027098231087,
+      "grad_norm": 0.22860077023506165,
+      "learning_rate": 4.735855290510808e-05,
+      "loss": 0.5346,
+      "step": 1980
+    },
+    {
+      "epoch": 1.4911554384644337,
+      "grad_norm": 0.28355228900909424,
+      "learning_rate": 4.7355053835606684e-05,
+      "loss": 0.4065,
+      "step": 1981
+    },
+    {
+      "epoch": 1.4919081671057584,
+      "grad_norm": 0.26692551374435425,
+      "learning_rate": 4.735155259478645e-05,
+      "loss": 0.4307,
+      "step": 1982
+    },
+    {
+      "epoch": 1.492660895747083,
+      "grad_norm": 0.3319557309150696,
+      "learning_rate": 4.7348049183030276e-05,
+      "loss": 0.364,
+      "step": 1983
+    },
+    {
+      "epoch": 1.493413624388408,
+      "grad_norm": 0.2582451403141022,
+      "learning_rate": 4.734454360072127e-05,
+      "loss": 0.4333,
+      "step": 1984
+    },
+    {
+      "epoch": 1.4941663530297329,
+      "grad_norm": 0.2729854881763458,
+      "learning_rate": 4.7341035848242824e-05,
+      "loss": 0.29,
+      "step": 1985
+    },
+    {
+      "epoch": 1.4949190816710576,
+      "grad_norm": 0.21722765266895294,
+      "learning_rate": 4.7337525925978526e-05,
+      "loss": 0.3249,
+      "step": 1986
+    },
+    {
+      "epoch": 1.4956718103123823,
+      "grad_norm": 0.27228429913520813,
+      "learning_rate": 4.733401383431222e-05,
+      "loss": 0.4166,
+      "step": 1987
+    },
+    {
+      "epoch": 1.4964245389537072,
+      "grad_norm": 0.2524281442165375,
+      "learning_rate": 4.733049957362797e-05,
+      "loss": 0.4305,
+      "step": 1988
+    },
+    {
+      "epoch": 1.4971772675950321,
+      "grad_norm": 0.2524823844432831,
+      "learning_rate": 4.732698314431011e-05,
+      "loss": 0.3951,
+      "step": 1989
+    },
+    {
+      "epoch": 1.4979299962363568,
+      "grad_norm": 0.23338036239147186,
+      "learning_rate": 4.732346454674318e-05,
+      "loss": 0.3383,
+      "step": 1990
+    },
+    {
+      "epoch": 1.4986827248776815,
+      "grad_norm": 0.2847089171409607,
+      "learning_rate": 4.7319943781311984e-05,
+      "loss": 0.4499,
+      "step": 1991
+    },
+    {
+      "epoch": 1.4994354535190064,
+      "grad_norm": 0.2401835322380066,
+      "learning_rate": 4.731642084840152e-05,
+      "loss": 0.414,
+      "step": 1992
+    },
+    {
+      "epoch": 1.5001881821603313,
+      "grad_norm": 0.30210888385772705,
+      "learning_rate": 4.731289574839707e-05,
+      "loss": 0.3952,
+      "step": 1993
+    },
+    {
+      "epoch": 1.500940910801656,
+      "grad_norm": 0.2603033483028412,
+      "learning_rate": 4.730936848168413e-05,
+      "loss": 0.4003,
+      "step": 1994
+    },
+    {
+      "epoch": 1.5016936394429807,
+      "grad_norm": 0.2751663625240326,
+      "learning_rate": 4.730583904864843e-05,
+      "loss": 0.3182,
+      "step": 1995
+    },
+    {
+      "epoch": 1.5024463680843057,
+      "grad_norm": 0.273121178150177,
+      "learning_rate": 4.7302307449675956e-05,
+      "loss": 0.3706,
+      "step": 1996
+    },
+    {
+      "epoch": 1.5031990967256306,
+      "grad_norm": 0.26496613025665283,
+      "learning_rate": 4.729877368515291e-05,
+      "loss": 0.4469,
+      "step": 1997
+    },
+    {
+      "epoch": 1.5039518253669553,
+      "grad_norm": 0.3022703230381012,
+      "learning_rate": 4.729523775546574e-05,
+      "loss": 0.398,
+      "step": 1998
+    },
+    {
+      "epoch": 1.50470455400828,
+      "grad_norm": 0.26286542415618896,
+      "learning_rate": 4.729169966100114e-05,
+      "loss": 0.3703,
+      "step": 1999
+    },
+    {
+      "epoch": 1.5054572826496049,
+      "grad_norm": 0.27176007628440857,
+      "learning_rate": 4.7288159402146e-05,
+      "loss": 0.3945,
+      "step": 2000
+    },
+    {
+      "epoch": 1.5054572826496049,
+      "eval_loss": 0.3970867693424225,
+      "eval_runtime": 452.5044,
+      "eval_samples_per_second": 21.275,
+      "eval_steps_per_second": 0.665,
+      "step": 2000
+    },
+    {
+      "epoch": 1.5062100112909296,
+      "grad_norm": 0.29814159870147705,
+      "learning_rate": 4.7284616979287515e-05,
+      "loss": 0.3741,
+      "step": 2001
+    },
+    {
+      "epoch": 1.5069627399322543,
+      "grad_norm": 0.21999850869178772,
+      "learning_rate": 4.728107239281306e-05,
+      "loss": 0.3825,
+      "step": 2002
+    },
+    {
+      "epoch": 1.5077154685735792,
+      "grad_norm": 0.2765474021434784,
+      "learning_rate": 4.7277525643110264e-05,
+      "loss": 0.4552,
+      "step": 2003
+    },
+    {
+      "epoch": 1.508468197214904,
+      "grad_norm": 0.24566976726055145,
+      "learning_rate": 4.727397673056701e-05,
+      "loss": 0.4537,
+      "step": 2004
+    },
+    {
+      "epoch": 1.5092209258562288,
+      "grad_norm": 0.2563714385032654,
+      "learning_rate": 4.7270425655571384e-05,
+      "loss": 0.354,
+      "step": 2005
+    },
+    {
+      "epoch": 1.5099736544975535,
+      "grad_norm": 0.2889309823513031,
+      "learning_rate": 4.726687241851173e-05,
+      "loss": 0.427,
+      "step": 2006
+    },
+    {
+      "epoch": 1.5107263831388784,
+      "grad_norm": 0.2546594440937042,
+      "learning_rate": 4.7263317019776625e-05,
+      "loss": 0.4091,
+      "step": 2007
+    },
+    {
+      "epoch": 1.5114791117802033,
+      "grad_norm": 0.21638517081737518,
+      "learning_rate": 4.725975945975489e-05,
+      "loss": 0.4273,
+      "step": 2008
+    },
+    {
+      "epoch": 1.512231840421528,
+      "grad_norm": 0.27159205079078674,
+      "learning_rate": 4.7256199738835575e-05,
+      "loss": 0.4294,
+      "step": 2009
+    },
+    {
+      "epoch": 1.5129845690628527,
+      "grad_norm": 0.24578146636486053,
+      "learning_rate": 4.725263785740795e-05,
+      "loss": 0.2779,
+      "step": 2010
+    },
+    {
+      "epoch": 1.5137372977041776,
+      "grad_norm": 0.22894708812236786,
+      "learning_rate": 4.724907381586155e-05,
+      "loss": 0.4139,
+      "step": 2011
+    },
+    {
+      "epoch": 1.5144900263455026,
+      "grad_norm": 0.2773571014404297,
+      "learning_rate": 4.7245507614586127e-05,
+      "loss": 0.4871,
+      "step": 2012
+    },
+    {
+      "epoch": 1.5152427549868273,
+      "grad_norm": 0.2547982931137085,
+      "learning_rate": 4.724193925397168e-05,
+      "loss": 0.3214,
+      "step": 2013
+    },
+    {
+      "epoch": 1.515995483628152,
+      "grad_norm": 0.19818982481956482,
+      "learning_rate": 4.7238368734408435e-05,
+      "loss": 0.309,
+      "step": 2014
+    },
+    {
+      "epoch": 1.5167482122694769,
+      "grad_norm": 0.25458985567092896,
+      "learning_rate": 4.723479605628687e-05,
+      "loss": 0.416,
+      "step": 2015
+    },
+    {
+      "epoch": 1.5175009409108018,
+      "grad_norm": 0.3080231547355652,
+      "learning_rate": 4.7231221219997665e-05,
+      "loss": 0.4634,
+      "step": 2016
+    },
+    {
+      "epoch": 1.5182536695521265,
+      "grad_norm": 0.30419638752937317,
+      "learning_rate": 4.7227644225931783e-05,
+      "loss": 0.3507,
+      "step": 2017
+    },
+    {
+      "epoch": 1.5190063981934512,
+      "grad_norm": 0.2696613371372223,
+      "learning_rate": 4.7224065074480385e-05,
+      "loss": 0.3598,
+      "step": 2018
+    },
+    {
+      "epoch": 1.519759126834776,
+      "grad_norm": 0.24516119062900543,
+      "learning_rate": 4.7220483766034886e-05,
+      "loss": 0.3088,
+      "step": 2019
+    },
+    {
+      "epoch": 1.520511855476101,
+      "grad_norm": 0.26089537143707275,
+      "learning_rate": 4.7216900300986936e-05,
+      "loss": 0.4922,
+      "step": 2020
+    },
+    {
+      "epoch": 1.5212645841174257,
+      "grad_norm": 0.2208845317363739,
+      "learning_rate": 4.7213314679728405e-05,
+      "loss": 0.426,
+      "step": 2021
+    },
+    {
+      "epoch": 1.5220173127587504,
+      "grad_norm": 0.2181219756603241,
+      "learning_rate": 4.720972690265142e-05,
+      "loss": 0.3513,
+      "step": 2022
+    },
+    {
+      "epoch": 1.5227700414000753,
+      "grad_norm": 0.24645842611789703,
+      "learning_rate": 4.720613697014834e-05,
+      "loss": 0.4324,
+      "step": 2023
+    },
+    {
+      "epoch": 1.5235227700414,
+      "grad_norm": 0.23956134915351868,
+      "learning_rate": 4.720254488261175e-05,
+      "loss": 0.3548,
+      "step": 2024
+    },
+    {
+      "epoch": 1.5242754986827247,
+      "grad_norm": 0.24635784327983856,
+      "learning_rate": 4.719895064043447e-05,
+      "loss": 0.4248,
+      "step": 2025
+    },
+    {
+      "epoch": 1.5250282273240496,
+      "grad_norm": 0.21817663311958313,
+      "learning_rate": 4.719535424400956e-05,
+      "loss": 0.2985,
+      "step": 2026
+    },
+    {
+      "epoch": 1.5257809559653746,
+      "grad_norm": 0.2686237394809723,
+      "learning_rate": 4.7191755693730326e-05,
+      "loss": 0.3634,
+      "step": 2027
+    },
+    {
+      "epoch": 1.5265336846066992,
+      "grad_norm": 0.22132927179336548,
+      "learning_rate": 4.718815498999029e-05,
+      "loss": 0.3861,
+      "step": 2028
+    },
+    {
+      "epoch": 1.527286413248024,
+      "grad_norm": 0.2611072063446045,
+      "learning_rate": 4.7184552133183237e-05,
+      "loss": 0.3995,
+      "step": 2029
+    },
+    {
+      "epoch": 1.5280391418893489,
+      "grad_norm": 0.25977379083633423,
+      "learning_rate": 4.7180947123703154e-05,
+      "loss": 0.5257,
+      "step": 2030
+    },
+    {
+      "epoch": 1.5287918705306738,
+      "grad_norm": 0.3137175142765045,
+      "learning_rate": 4.7177339961944276e-05,
+      "loss": 0.3926,
+      "step": 2031
+    },
+    {
+      "epoch": 1.5295445991719985,
+      "grad_norm": 0.41144293546676636,
+      "learning_rate": 4.7173730648301095e-05,
+      "loss": 0.425,
+      "step": 2032
+    },
+    {
+      "epoch": 1.5302973278133232,
+      "grad_norm": 0.2653573155403137,
+      "learning_rate": 4.71701191831683e-05,
+      "loss": 0.3959,
+      "step": 2033
+    },
+    {
+      "epoch": 1.531050056454648,
+      "grad_norm": 0.3415326178073883,
+      "learning_rate": 4.716650556694085e-05,
+      "loss": 0.4265,
+      "step": 2034
+    },
+    {
+      "epoch": 1.531802785095973,
+      "grad_norm": 0.2520945072174072,
+      "learning_rate": 4.7162889800013917e-05,
+      "loss": 0.4428,
+      "step": 2035
+    },
+    {
+      "epoch": 1.5325555137372977,
+      "grad_norm": 0.6656238436698914,
+      "learning_rate": 4.715927188278292e-05,
+      "loss": 0.5213,
+      "step": 2036
+    },
+    {
+      "epoch": 1.5333082423786224,
+      "grad_norm": 0.4202136695384979,
+      "learning_rate": 4.715565181564351e-05,
+      "loss": 0.4412,
+      "step": 2037
+    },
+    {
+      "epoch": 1.5340609710199473,
+      "grad_norm": 0.4057386815547943,
+      "learning_rate": 4.715202959899156e-05,
+      "loss": 0.4245,
+      "step": 2038
+    },
+    {
+      "epoch": 1.5348136996612722,
+      "grad_norm": 0.30114293098449707,
+      "learning_rate": 4.71484052332232e-05,
+      "loss": 0.4331,
+      "step": 2039
+    },
+    {
+      "epoch": 1.535566428302597,
+      "grad_norm": 0.40116575360298157,
+      "learning_rate": 4.7144778718734794e-05,
+      "loss": 0.3905,
+      "step": 2040
+    },
+    {
+      "epoch": 1.5363191569439216,
+      "grad_norm": 0.3638280928134918,
+      "learning_rate": 4.7141150055922916e-05,
+      "loss": 0.3474,
+      "step": 2041
+    },
+    {
+      "epoch": 1.5370718855852465,
+      "grad_norm": 0.28640294075012207,
+      "learning_rate": 4.7137519245184406e-05,
+      "loss": 0.3864,
+      "step": 2042
+    },
+    {
+      "epoch": 1.5378246142265715,
+      "grad_norm": 0.3453085422515869,
+      "learning_rate": 4.7133886286916305e-05,
+      "loss": 0.3785,
+      "step": 2043
+    },
+    {
+      "epoch": 1.5385773428678962,
+      "grad_norm": 0.30082401633262634,
+      "learning_rate": 4.7130251181515926e-05,
+      "loss": 0.4002,
+      "step": 2044
+    },
+    {
+      "epoch": 1.5393300715092209,
+      "grad_norm": 0.28007936477661133,
+      "learning_rate": 4.712661392938079e-05,
+      "loss": 0.3888,
+      "step": 2045
+    },
+    {
+      "epoch": 1.5400828001505458,
+      "grad_norm": 0.36941519379615784,
+      "learning_rate": 4.712297453090866e-05,
+      "loss": 0.4229,
+      "step": 2046
+    },
+    {
+      "epoch": 1.5408355287918707,
+      "grad_norm": 0.28419381380081177,
+      "learning_rate": 4.711933298649754e-05,
+      "loss": 0.3548,
+      "step": 2047
+    },
+    {
+      "epoch": 1.5415882574331954,
+      "grad_norm": 0.2595309317111969,
+      "learning_rate": 4.711568929654566e-05,
+      "loss": 0.3752,
+      "step": 2048
+    },
+    {
+      "epoch": 1.54234098607452,
+      "grad_norm": 0.24950958788394928,
+      "learning_rate": 4.711204346145149e-05,
+      "loss": 0.5183,
+      "step": 2049
+    },
+    {
+      "epoch": 1.543093714715845,
+      "grad_norm": 0.26852983236312866,
+      "learning_rate": 4.7108395481613736e-05,
+      "loss": 0.3407,
+      "step": 2050
+    },
+    {
+      "epoch": 1.5438464433571697,
+      "grad_norm": 0.23763960599899292,
+      "learning_rate": 4.710474535743134e-05,
+      "loss": 0.3745,
+      "step": 2051
+    },
+    {
+      "epoch": 1.5445991719984944,
+      "grad_norm": 0.25649046897888184,
+      "learning_rate": 4.710109308930346e-05,
+      "loss": 0.4842,
+      "step": 2052
+    },
+    {
+      "epoch": 1.5453519006398193,
+      "grad_norm": 0.3215932548046112,
+      "learning_rate": 4.709743867762951e-05,
+      "loss": 0.3804,
+      "step": 2053
+    },
+    {
+      "epoch": 1.5461046292811442,
+      "grad_norm": 0.3294551372528076,
+      "learning_rate": 4.709378212280913e-05,
+      "loss": 0.4826,
+      "step": 2054
+    },
+    {
+      "epoch": 1.546857357922469,
+      "grad_norm": 0.2877941131591797,
+      "learning_rate": 4.7090123425242186e-05,
+      "loss": 0.383,
+      "step": 2055
+    },
+    {
+      "epoch": 1.5476100865637936,
+      "grad_norm": 0.29043012857437134,
+      "learning_rate": 4.708646258532881e-05,
+      "loss": 0.391,
+      "step": 2056
+    },
+    {
+      "epoch": 1.5483628152051185,
+      "grad_norm": 0.3377133309841156,
+      "learning_rate": 4.708279960346933e-05,
+      "loss": 0.4476,
+      "step": 2057
+    },
+    {
+      "epoch": 1.5491155438464435,
+      "grad_norm": 0.35580769181251526,
+      "learning_rate": 4.7079134480064326e-05,
+      "loss": 0.3339,
+      "step": 2058
+    },
+    {
+      "epoch": 1.5498682724877682,
+      "grad_norm": 0.26918649673461914,
+      "learning_rate": 4.707546721551462e-05,
+      "loss": 0.387,
+      "step": 2059
+    },
+    {
+      "epoch": 1.5506210011290928,
+      "grad_norm": 0.22651629149913788,
+      "learning_rate": 4.7071797810221245e-05,
+      "loss": 0.3884,
+      "step": 2060
+    },
+    {
+      "epoch": 1.5513737297704178,
+      "grad_norm": 0.319133460521698,
+      "learning_rate": 4.706812626458549e-05,
+      "loss": 0.3295,
+      "step": 2061
+    },
+    {
+      "epoch": 1.5521264584117427,
+      "grad_norm": 0.2652226686477661,
+      "learning_rate": 4.706445257900886e-05,
+      "loss": 0.4731,
+      "step": 2062
+    },
+    {
+      "epoch": 1.5528791870530674,
+      "grad_norm": 0.30283185839653015,
+      "learning_rate": 4.706077675389311e-05,
+      "loss": 0.4194,
+      "step": 2063
+    },
+    {
+      "epoch": 1.553631915694392,
+      "grad_norm": 0.23244726657867432,
+      "learning_rate": 4.705709878964023e-05,
+      "loss": 0.3676,
+      "step": 2064
+    },
+    {
+      "epoch": 1.554384644335717,
+      "grad_norm": 0.24378874897956848,
+      "learning_rate": 4.705341868665243e-05,
+      "loss": 0.3479,
+      "step": 2065
+    },
+    {
+      "epoch": 1.555137372977042,
+      "grad_norm": 0.24910041689872742,
+      "learning_rate": 4.7049736445332156e-05,
+      "loss": 0.3738,
+      "step": 2066
+    },
+    {
+      "epoch": 1.5558901016183666,
+      "grad_norm": 0.22115467488765717,
+      "learning_rate": 4.70460520660821e-05,
+      "loss": 0.377,
+      "step": 2067
+    },
+    {
+      "epoch": 1.5566428302596913,
+      "grad_norm": 0.26513227820396423,
+      "learning_rate": 4.7042365549305176e-05,
+      "loss": 0.4084,
+      "step": 2068
+    },
+    {
+      "epoch": 1.5573955589010162,
+      "grad_norm": 0.23996300995349884,
+      "learning_rate": 4.703867689540454e-05,
+      "loss": 0.3691,
+      "step": 2069
+    },
+    {
+      "epoch": 1.5581482875423411,
+      "grad_norm": 0.33642974495887756,
+      "learning_rate": 4.703498610478357e-05,
+      "loss": 0.4118,
+      "step": 2070
+    },
+    {
+      "epoch": 1.5589010161836658,
+      "grad_norm": 0.3925519287586212,
+      "learning_rate": 4.7031293177845896e-05,
+      "loss": 0.459,
+      "step": 2071
+    },
+    {
+      "epoch": 1.5596537448249905,
+      "grad_norm": 0.27415987849235535,
+      "learning_rate": 4.702759811499536e-05,
+      "loss": 0.3885,
+      "step": 2072
+    },
+    {
+      "epoch": 1.5604064734663154,
+      "grad_norm": 0.26700878143310547,
+      "learning_rate": 4.702390091663605e-05,
+      "loss": 0.3648,
+      "step": 2073
+    },
+    {
+      "epoch": 1.5611592021076401,
+      "grad_norm": 0.3387112617492676,
+      "learning_rate": 4.7020201583172294e-05,
+      "loss": 0.3229,
+      "step": 2074
+    },
+    {
+      "epoch": 1.5619119307489648,
+      "grad_norm": 0.31989145278930664,
+      "learning_rate": 4.701650011500864e-05,
+      "loss": 0.456,
+      "step": 2075
+    },
+    {
+      "epoch": 1.5626646593902898,
+      "grad_norm": 0.25235316157341003,
+      "learning_rate": 4.7012796512549876e-05,
+      "loss": 0.454,
+      "step": 2076
+    },
+    {
+      "epoch": 1.5634173880316147,
+      "grad_norm": 0.32430699467658997,
+      "learning_rate": 4.7009090776201026e-05,
+      "loss": 0.5387,
+      "step": 2077
+    },
+    {
+      "epoch": 1.5641701166729394,
+      "grad_norm": 0.2457648068666458,
+      "learning_rate": 4.700538290636734e-05,
+      "loss": 0.4032,
+      "step": 2078
+    },
+    {
+      "epoch": 1.564922845314264,
+      "grad_norm": 0.31657326221466064,
+      "learning_rate": 4.700167290345431e-05,
+      "loss": 0.3999,
+      "step": 2079
+    },
+    {
+      "epoch": 1.565675573955589,
+      "grad_norm": 0.24416033923625946,
+      "learning_rate": 4.699796076786764e-05,
+      "loss": 0.3992,
+      "step": 2080
+    },
+    {
+      "epoch": 1.566428302596914,
+      "grad_norm": 0.2741968631744385,
+      "learning_rate": 4.69942465000133e-05,
+      "loss": 0.3686,
+      "step": 2081
+    },
+    {
+      "epoch": 1.5671810312382386,
+      "grad_norm": 0.3155936300754547,
+      "learning_rate": 4.699053010029747e-05,
+      "loss": 0.3982,
+      "step": 2082
+    },
+    {
+      "epoch": 1.5679337598795633,
+      "grad_norm": 0.22421666979789734,
+      "learning_rate": 4.698681156912657e-05,
+      "loss": 0.3464,
+      "step": 2083
+    },
+    {
+      "epoch": 1.5686864885208882,
+      "grad_norm": 0.2735782265663147,
+      "learning_rate": 4.698309090690726e-05,
+      "loss": 0.3512,
+      "step": 2084
+    },
+    {
+      "epoch": 1.5694392171622131,
+      "grad_norm": 0.2946726381778717,
+      "learning_rate": 4.697936811404642e-05,
+      "loss": 0.4446,
+      "step": 2085
+    },
+    {
+      "epoch": 1.5701919458035378,
+      "grad_norm": 0.3433338403701782,
+      "learning_rate": 4.6975643190951166e-05,
+      "loss": 0.353,
+      "step": 2086
+    },
+    {
+      "epoch": 1.5709446744448625,
+      "grad_norm": 0.2644486129283905,
+      "learning_rate": 4.697191613802886e-05,
+      "loss": 0.4146,
+      "step": 2087
+    },
+    {
+      "epoch": 1.5716974030861874,
+      "grad_norm": 0.35960233211517334,
+      "learning_rate": 4.696818695568708e-05,
+      "loss": 0.4491,
+      "step": 2088
+    },
+    {
+      "epoch": 1.5724501317275124,
+      "grad_norm": 0.34053152799606323,
+      "learning_rate": 4.696445564433364e-05,
+      "loss": 0.4057,
+      "step": 2089
+    },
+    {
+      "epoch": 1.573202860368837,
+      "grad_norm": 0.25701889395713806,
+      "learning_rate": 4.6960722204376595e-05,
+      "loss": 0.4353,
+      "step": 2090
+    },
+    {
+      "epoch": 1.5739555890101617,
+      "grad_norm": 0.3475569188594818,
+      "learning_rate": 4.6956986636224234e-05,
+      "loss": 0.3638,
+      "step": 2091
+    },
+    {
+      "epoch": 1.5747083176514867,
+      "grad_norm": 0.3755210340023041,
+      "learning_rate": 4.695324894028506e-05,
+      "loss": 0.4792,
+      "step": 2092
+    },
+    {
+      "epoch": 1.5754610462928116,
+      "grad_norm": 0.2866958975791931,
+      "learning_rate": 4.6949509116967836e-05,
+      "loss": 0.3547,
+      "step": 2093
+    },
+    {
+      "epoch": 1.5762137749341363,
+      "grad_norm": 0.35855135321617126,
+      "learning_rate": 4.694576716668153e-05,
+      "loss": 0.4105,
+      "step": 2094
+    },
+    {
+      "epoch": 1.576966503575461,
+      "grad_norm": 0.2514727711677551,
+      "learning_rate": 4.6942023089835374e-05,
+      "loss": 0.3514,
+      "step": 2095
+    },
+    {
+      "epoch": 1.577719232216786,
+      "grad_norm": 0.38216209411621094,
+      "learning_rate": 4.693827688683879e-05,
+      "loss": 0.5453,
+      "step": 2096
+    },
+    {
+      "epoch": 1.5784719608581108,
+      "grad_norm": 0.2544863820075989,
+      "learning_rate": 4.693452855810149e-05,
+      "loss": 0.3889,
+      "step": 2097
+    },
+    {
+      "epoch": 1.5792246894994355,
+      "grad_norm": 0.32171300053596497,
+      "learning_rate": 4.6930778104033353e-05,
+      "loss": 0.4381,
+      "step": 2098
+    },
+    {
+      "epoch": 1.5799774181407602,
+      "grad_norm": 0.22158420085906982,
+      "learning_rate": 4.6927025525044534e-05,
+      "loss": 0.369,
+      "step": 2099
+    },
+    {
+      "epoch": 1.5807301467820851,
+      "grad_norm": 0.30896368622779846,
+      "learning_rate": 4.692327082154542e-05,
+      "loss": 0.4745,
+      "step": 2100
+    },
+    {
+      "epoch": 1.5814828754234098,
+      "grad_norm": 0.24818137288093567,
+      "learning_rate": 4.69195139939466e-05,
+      "loss": 0.4234,
+      "step": 2101
+    },
+    {
+      "epoch": 1.5822356040647345,
+      "grad_norm": 0.3150216042995453,
+      "learning_rate": 4.691575504265893e-05,
+      "loss": 0.4511,
+      "step": 2102
+    },
+    {
+      "epoch": 1.5829883327060594,
+      "grad_norm": 0.27837666869163513,
+      "learning_rate": 4.6911993968093486e-05,
+      "loss": 0.5235,
+      "step": 2103
+    },
+    {
+      "epoch": 1.5837410613473843,
+      "grad_norm": 0.29607126116752625,
+      "learning_rate": 4.690823077066156e-05,
+      "loss": 0.3598,
+      "step": 2104
+    },
+    {
+      "epoch": 1.584493789988709,
+      "grad_norm": 0.2828029692173004,
+      "learning_rate": 4.69044654507747e-05,
+      "loss": 0.3842,
+      "step": 2105
+    },
+    {
+      "epoch": 1.5852465186300337,
+      "grad_norm": 0.27370041608810425,
+      "learning_rate": 4.6900698008844675e-05,
+      "loss": 0.3609,
+      "step": 2106
+    },
+    {
+      "epoch": 1.5859992472713587,
+      "grad_norm": 0.23606985807418823,
+      "learning_rate": 4.6896928445283484e-05,
+      "loss": 0.3692,
+      "step": 2107
+    },
+    {
+      "epoch": 1.5867519759126836,
+      "grad_norm": 0.2315112054347992,
+      "learning_rate": 4.689315676050336e-05,
+      "loss": 0.4503,
+      "step": 2108
+    },
+    {
+      "epoch": 1.5875047045540083,
+      "grad_norm": 0.2573155462741852,
+      "learning_rate": 4.6889382954916765e-05,
+      "loss": 0.3374,
+      "step": 2109
+    },
+    {
+      "epoch": 1.588257433195333,
+      "grad_norm": 0.22906139492988586,
+      "learning_rate": 4.688560702893641e-05,
+      "loss": 0.3894,
+      "step": 2110
+    },
+    {
+      "epoch": 1.5890101618366579,
+      "grad_norm": 0.24356085062026978,
+      "learning_rate": 4.688182898297521e-05,
+      "loss": 0.3851,
+      "step": 2111
+    },
+    {
+      "epoch": 1.5897628904779828,
+      "grad_norm": 0.2282952070236206,
+      "learning_rate": 4.687804881744634e-05,
+      "loss": 0.4116,
+      "step": 2112
+    },
+    {
+      "epoch": 1.5905156191193075,
+      "grad_norm": 0.22544428706169128,
+      "learning_rate": 4.687426653276318e-05,
+      "loss": 0.3519,
+      "step": 2113
+    },
+    {
+      "epoch": 1.5912683477606322,
+      "grad_norm": 0.27280446887016296,
+      "learning_rate": 4.687048212933936e-05,
+      "loss": 0.4414,
+      "step": 2114
+    },
+    {
+      "epoch": 1.5920210764019571,
+      "grad_norm": 0.2110767811536789,
+      "learning_rate": 4.6866695607588737e-05,
+      "loss": 0.4053,
+      "step": 2115
+    },
+    {
+      "epoch": 1.592773805043282,
+      "grad_norm": 0.2501530945301056,
+      "learning_rate": 4.6862906967925404e-05,
+      "loss": 0.4062,
+      "step": 2116
+    },
+    {
+      "epoch": 1.5935265336846067,
+      "grad_norm": 0.2540273666381836,
+      "learning_rate": 4.685911621076367e-05,
+      "loss": 0.4069,
+      "step": 2117
+    },
+    {
+      "epoch": 1.5942792623259314,
+      "grad_norm": 0.27665528655052185,
+      "learning_rate": 4.6855323336518094e-05,
+      "loss": 0.4049,
+      "step": 2118
+    },
+    {
+      "epoch": 1.5950319909672563,
+      "grad_norm": 0.24394866824150085,
+      "learning_rate": 4.685152834560346e-05,
+      "loss": 0.48,
+      "step": 2119
+    },
+    {
+      "epoch": 1.5957847196085813,
+      "grad_norm": 0.2340671271085739,
+      "learning_rate": 4.684773123843478e-05,
+      "loss": 0.4343,
+      "step": 2120
+    },
+    {
+      "epoch": 1.596537448249906,
+      "grad_norm": 0.24147120118141174,
+      "learning_rate": 4.684393201542731e-05,
+      "loss": 0.4442,
+      "step": 2121
+    },
+    {
+      "epoch": 1.5972901768912307,
+      "grad_norm": 0.27333080768585205,
+      "learning_rate": 4.6840130676996505e-05,
+      "loss": 0.3892,
+      "step": 2122
+    },
+    {
+      "epoch": 1.5980429055325556,
+      "grad_norm": 0.2476656287908554,
+      "learning_rate": 4.6836327223558095e-05,
+      "loss": 0.4175,
+      "step": 2123
+    },
+    {
+      "epoch": 1.5987956341738803,
+      "grad_norm": 0.2916117310523987,
+      "learning_rate": 4.683252165552801e-05,
+      "loss": 0.3796,
+      "step": 2124
+    },
+    {
+      "epoch": 1.599548362815205,
+      "grad_norm": 0.2548069357872009,
+      "learning_rate": 4.682871397332242e-05,
+      "loss": 0.3597,
+      "step": 2125
+    },
+    {
+      "epoch": 1.6003010914565299,
+      "grad_norm": 0.27154824137687683,
+      "learning_rate": 4.682490417735773e-05,
+      "loss": 0.2816,
+      "step": 2126
+    },
+    {
+      "epoch": 1.6010538200978548,
+      "grad_norm": 0.4037674069404602,
+      "learning_rate": 4.682109226805058e-05,
+      "loss": 0.3549,
+      "step": 2127
+    },
+    {
+      "epoch": 1.6018065487391795,
+      "grad_norm": 0.428207665681839,
+      "learning_rate": 4.681727824581782e-05,
+      "loss": 0.3949,
+      "step": 2128
+    },
+    {
+      "epoch": 1.6025592773805042,
+      "grad_norm": 0.2941756844520569,
+      "learning_rate": 4.681346211107656e-05,
+      "loss": 0.4785,
+      "step": 2129
+    },
+    {
+      "epoch": 1.603312006021829,
+      "grad_norm": 0.3889387845993042,
+      "learning_rate": 4.680964386424412e-05,
+      "loss": 0.3235,
+      "step": 2130
+    },
+    {
+      "epoch": 1.604064734663154,
+      "grad_norm": 0.3366186320781708,
+      "learning_rate": 4.680582350573806e-05,
+      "loss": 0.3682,
+      "step": 2131
+    },
+    {
+      "epoch": 1.6048174633044787,
+      "grad_norm": 0.23813843727111816,
+      "learning_rate": 4.680200103597616e-05,
+      "loss": 0.3531,
+      "step": 2132
+    },
+    {
+      "epoch": 1.6055701919458034,
+      "grad_norm": 0.2937071621417999,
+      "learning_rate": 4.679817645537645e-05,
+      "loss": 0.3663,
+      "step": 2133
+    },
+    {
+      "epoch": 1.6063229205871283,
+      "grad_norm": 0.29814639687538147,
+      "learning_rate": 4.679434976435718e-05,
+      "loss": 0.3419,
+      "step": 2134
+    },
+    {
+      "epoch": 1.6070756492284533,
+      "grad_norm": 0.35698220133781433,
+      "learning_rate": 4.6790520963336825e-05,
+      "loss": 0.4419,
+      "step": 2135
+    },
+    {
+      "epoch": 1.607828377869778,
+      "grad_norm": 0.30727824568748474,
+      "learning_rate": 4.6786690052734106e-05,
+      "loss": 0.4321,
+      "step": 2136
+    },
+    {
+      "epoch": 1.6085811065111026,
+      "grad_norm": 0.29018932580947876,
+      "learning_rate": 4.678285703296795e-05,
+      "loss": 0.4259,
+      "step": 2137
+    },
+    {
+      "epoch": 1.6093338351524276,
+      "grad_norm": 0.32225918769836426,
+      "learning_rate": 4.677902190445755e-05,
+      "loss": 0.4024,
+      "step": 2138
+    },
+    {
+      "epoch": 1.6100865637937525,
+      "grad_norm": 0.22488465905189514,
+      "learning_rate": 4.677518466762229e-05,
+      "loss": 0.3607,
+      "step": 2139
+    },
+    {
+      "epoch": 1.6108392924350772,
+      "grad_norm": 0.2789320945739746,
+      "learning_rate": 4.677134532288183e-05,
+      "loss": 0.4375,
+      "step": 2140
+    },
+    {
+      "epoch": 1.6115920210764019,
+      "grad_norm": 0.2834721505641937,
+      "learning_rate": 4.6767503870656e-05,
+      "loss": 0.4202,
+      "step": 2141
+    },
+    {
+      "epoch": 1.6123447497177268,
+      "grad_norm": 0.251900315284729,
+      "learning_rate": 4.676366031136494e-05,
+      "loss": 0.3727,
+      "step": 2142
+    },
+    {
+      "epoch": 1.6130974783590517,
+      "grad_norm": 0.26108479499816895,
+      "learning_rate": 4.675981464542893e-05,
+      "loss": 0.461,
+      "step": 2143
+    },
+    {
+      "epoch": 1.6138502070003764,
+      "grad_norm": 0.23264721035957336,
+      "learning_rate": 4.675596687326855e-05,
+      "loss": 0.3668,
+      "step": 2144
+    },
+    {
+      "epoch": 1.614602935641701,
+      "grad_norm": 0.2500256896018982,
+      "learning_rate": 4.675211699530459e-05,
+      "loss": 0.3734,
+      "step": 2145
+    },
+    {
+      "epoch": 1.615355664283026,
+      "grad_norm": 0.26815587282180786,
+      "learning_rate": 4.674826501195806e-05,
+      "loss": 0.3554,
+      "step": 2146
+    },
+    {
+      "epoch": 1.616108392924351,
+      "grad_norm": 0.28674671053886414,
+      "learning_rate": 4.67444109236502e-05,
+      "loss": 0.4946,
+      "step": 2147
+    },
+    {
+      "epoch": 1.6168611215656754,
+      "grad_norm": 0.23757445812225342,
+      "learning_rate": 4.67405547308025e-05,
+      "loss": 0.414,
+      "step": 2148
+    },
+    {
+      "epoch": 1.6176138502070003,
+      "grad_norm": 0.3468417227268219,
+      "learning_rate": 4.673669643383667e-05,
+      "loss": 0.4084,
+      "step": 2149
+    },
+    {
+      "epoch": 1.6183665788483252,
+      "grad_norm": 0.28463003039360046,
+      "learning_rate": 4.6732836033174634e-05,
+      "loss": 0.5175,
+      "step": 2150
+    },
+    {
+      "epoch": 1.61911930748965,
+      "grad_norm": 0.2501148581504822,
+      "learning_rate": 4.6728973529238564e-05,
+      "loss": 0.3147,
+      "step": 2151
+    },
+    {
+      "epoch": 1.6198720361309746,
+      "grad_norm": 0.3145959973335266,
+      "learning_rate": 4.672510892245086e-05,
+      "loss": 0.4301,
+      "step": 2152
+    },
+    {
+      "epoch": 1.6206247647722996,
+      "grad_norm": 0.2899034023284912,
+      "learning_rate": 4.6721242213234145e-05,
+      "loss": 0.4167,
+      "step": 2153
+    },
+    {
+      "epoch": 1.6213774934136245,
+      "grad_norm": 0.2740843892097473,
+      "learning_rate": 4.6717373402011277e-05,
+      "loss": 0.4162,
+      "step": 2154
+    },
+    {
+      "epoch": 1.6221302220549492,
+      "grad_norm": 0.26012668013572693,
+      "learning_rate": 4.671350248920536e-05,
+      "loss": 0.3607,
+      "step": 2155
+    },
+    {
+      "epoch": 1.6228829506962739,
+      "grad_norm": 0.2225465029478073,
+      "learning_rate": 4.6709629475239684e-05,
+      "loss": 0.4336,
+      "step": 2156
+    },
+    {
+      "epoch": 1.6236356793375988,
+      "grad_norm": 0.2474265694618225,
+      "learning_rate": 4.670575436053782e-05,
+      "loss": 0.4062,
+      "step": 2157
+    },
+    {
+      "epoch": 1.6243884079789237,
+      "grad_norm": 0.22577890753746033,
+      "learning_rate": 4.670187714552353e-05,
+      "loss": 0.3694,
+      "step": 2158
+    },
+    {
+      "epoch": 1.6251411366202484,
+      "grad_norm": 0.2561524212360382,
+      "learning_rate": 4.669799783062083e-05,
+      "loss": 0.4262,
+      "step": 2159
+    },
+    {
+      "epoch": 1.625893865261573,
+      "grad_norm": 0.286355584859848,
+      "learning_rate": 4.669411641625394e-05,
+      "loss": 0.3896,
+      "step": 2160
+    },
+    {
+      "epoch": 1.626646593902898,
+      "grad_norm": 0.2540460526943207,
+      "learning_rate": 4.6690232902847346e-05,
+      "loss": 0.398,
+      "step": 2161
+    },
+    {
+      "epoch": 1.627399322544223,
+      "grad_norm": 0.24461296200752258,
+      "learning_rate": 4.6686347290825736e-05,
+      "loss": 0.403,
+      "step": 2162
+    },
+    {
+      "epoch": 1.6281520511855476,
+      "grad_norm": 0.25704115629196167,
+      "learning_rate": 4.668245958061403e-05,
+      "loss": 0.3613,
+      "step": 2163
+    },
+    {
+      "epoch": 1.6289047798268723,
+      "grad_norm": 0.2534612715244293,
+      "learning_rate": 4.667856977263738e-05,
+      "loss": 0.3769,
+      "step": 2164
+    },
+    {
+      "epoch": 1.6296575084681972,
+      "grad_norm": 0.2476104348897934,
+      "learning_rate": 4.667467786732118e-05,
+      "loss": 0.3768,
+      "step": 2165
+    },
+    {
+      "epoch": 1.6304102371095222,
+      "grad_norm": 0.2944454550743103,
+      "learning_rate": 4.667078386509103e-05,
+      "loss": 0.4026,
+      "step": 2166
+    },
+    {
+      "epoch": 1.6311629657508468,
+      "grad_norm": 0.3456771969795227,
+      "learning_rate": 4.666688776637279e-05,
+      "loss": 0.4912,
+      "step": 2167
+    },
+    {
+      "epoch": 1.6319156943921715,
+      "grad_norm": 0.2632859945297241,
+      "learning_rate": 4.666298957159251e-05,
+      "loss": 0.3985,
+      "step": 2168
+    },
+    {
+      "epoch": 1.6326684230334965,
+      "grad_norm": 0.3266420364379883,
+      "learning_rate": 4.665908928117651e-05,
+      "loss": 0.3361,
+      "step": 2169
+    },
+    {
+      "epoch": 1.6334211516748214,
+      "grad_norm": 0.36075252294540405,
+      "learning_rate": 4.665518689555131e-05,
+      "loss": 0.3776,
+      "step": 2170
+    },
+    {
+      "epoch": 1.634173880316146,
+      "grad_norm": 0.3682076632976532,
+      "learning_rate": 4.6651282415143664e-05,
+      "loss": 0.3296,
+      "step": 2171
+    },
+    {
+      "epoch": 1.6349266089574708,
+      "grad_norm": 0.2555447816848755,
+      "learning_rate": 4.664737584038057e-05,
+      "loss": 0.2592,
+      "step": 2172
+    },
+    {
+      "epoch": 1.6356793375987957,
+      "grad_norm": 0.38291341066360474,
+      "learning_rate": 4.6643467171689246e-05,
+      "loss": 0.4514,
+      "step": 2173
+    },
+    {
+      "epoch": 1.6364320662401204,
+      "grad_norm": 0.32836002111434937,
+      "learning_rate": 4.663955640949712e-05,
+      "loss": 0.448,
+      "step": 2174
+    },
+    {
+      "epoch": 1.637184794881445,
+      "grad_norm": 0.39856404066085815,
+      "learning_rate": 4.6635643554231894e-05,
+      "loss": 0.4236,
+      "step": 2175
+    },
+    {
+      "epoch": 1.63793752352277,
+      "grad_norm": 0.25669917464256287,
+      "learning_rate": 4.6631728606321463e-05,
+      "loss": 0.3533,
+      "step": 2176
+    },
+    {
+      "epoch": 1.638690252164095,
+      "grad_norm": 0.30679914355278015,
+      "learning_rate": 4.662781156619395e-05,
+      "loss": 0.4086,
+      "step": 2177
+    },
+    {
+      "epoch": 1.6394429808054196,
+      "grad_norm": 0.2772020995616913,
+      "learning_rate": 4.662389243427771e-05,
+      "loss": 0.3886,
+      "step": 2178
+    },
+    {
+      "epoch": 1.6401957094467443,
+      "grad_norm": 0.2514660656452179,
+      "learning_rate": 4.661997121100137e-05,
+      "loss": 0.4082,
+      "step": 2179
+    },
+    {
+      "epoch": 1.6409484380880692,
+      "grad_norm": 0.3144722282886505,
+      "learning_rate": 4.6616047896793704e-05,
+      "loss": 0.3672,
+      "step": 2180
+    },
+    {
+      "epoch": 1.6417011667293941,
+      "grad_norm": 0.2718471884727478,
+      "learning_rate": 4.6612122492083785e-05,
+      "loss": 0.3406,
+      "step": 2181
+    },
+    {
+      "epoch": 1.6424538953707188,
+      "grad_norm": 0.25909388065338135,
+      "learning_rate": 4.660819499730089e-05,
+      "loss": 0.3287,
+      "step": 2182
+    },
+    {
+      "epoch": 1.6432066240120435,
+      "grad_norm": 0.2643587589263916,
+      "learning_rate": 4.660426541287452e-05,
+      "loss": 0.3641,
+      "step": 2183
+    },
+    {
+      "epoch": 1.6439593526533685,
+      "grad_norm": 0.32422858476638794,
+      "learning_rate": 4.66003337392344e-05,
+      "loss": 0.3954,
+      "step": 2184
+    },
+    {
+      "epoch": 1.6447120812946934,
+      "grad_norm": 0.26656433939933777,
+      "learning_rate": 4.659639997681051e-05,
+      "loss": 0.3084,
+      "step": 2185
+    },
+    {
+      "epoch": 1.645464809936018,
+      "grad_norm": 0.1933862864971161,
+      "learning_rate": 4.659246412603302e-05,
+      "loss": 0.3113,
+      "step": 2186
+    },
+    {
+      "epoch": 1.6462175385773428,
+      "grad_norm": 0.29049816727638245,
+      "learning_rate": 4.658852618733236e-05,
+      "loss": 0.4503,
+      "step": 2187
+    },
+    {
+      "epoch": 1.6469702672186677,
+      "grad_norm": 0.22460335493087769,
+      "learning_rate": 4.6584586161139185e-05,
+      "loss": 0.3388,
+      "step": 2188
+    },
+    {
+      "epoch": 1.6477229958599926,
+      "grad_norm": 0.3166655898094177,
+      "learning_rate": 4.6580644047884344e-05,
+      "loss": 0.4136,
+      "step": 2189
+    },
+    {
+      "epoch": 1.6484757245013173,
+      "grad_norm": 0.25450584292411804,
+      "learning_rate": 4.6576699847998966e-05,
+      "loss": 0.4172,
+      "step": 2190
+    },
+    {
+      "epoch": 1.649228453142642,
+      "grad_norm": 0.30552271008491516,
+      "learning_rate": 4.657275356191437e-05,
+      "loss": 0.4253,
+      "step": 2191
+    },
+    {
+      "epoch": 1.649981181783967,
+      "grad_norm": 0.2522934675216675,
+      "learning_rate": 4.6568805190062126e-05,
+      "loss": 0.4064,
+      "step": 2192
+    },
+    {
+      "epoch": 1.6507339104252918,
+      "grad_norm": 0.30206817388534546,
+      "learning_rate": 4.6564854732874014e-05,
+      "loss": 0.426,
+      "step": 2193
+    },
+    {
+      "epoch": 1.6514866390666165,
+      "grad_norm": 0.25373488664627075,
+      "learning_rate": 4.656090219078205e-05,
+      "loss": 0.3384,
+      "step": 2194
+    },
+    {
+      "epoch": 1.6522393677079412,
+      "grad_norm": 0.27366021275520325,
+      "learning_rate": 4.655694756421848e-05,
+      "loss": 0.3521,
+      "step": 2195
+    },
+    {
+      "epoch": 1.6529920963492661,
+      "grad_norm": 0.27646973729133606,
+      "learning_rate": 4.655299085361578e-05,
+      "loss": 0.2884,
+      "step": 2196
+    },
+    {
+      "epoch": 1.653744824990591,
+      "grad_norm": 0.3312423825263977,
+      "learning_rate": 4.654903205940665e-05,
+      "loss": 0.4286,
+      "step": 2197
+    },
+    {
+      "epoch": 1.6544975536319155,
+      "grad_norm": 0.2589821517467499,
+      "learning_rate": 4.6545071182024e-05,
+      "loss": 0.3572,
+      "step": 2198
+    },
+    {
+      "epoch": 1.6552502822732404,
+      "grad_norm": 0.3059285879135132,
+      "learning_rate": 4.654110822190101e-05,
+      "loss": 0.3283,
+      "step": 2199
+    },
+    {
+      "epoch": 1.6560030109145654,
+      "grad_norm": 0.27532756328582764,
+      "learning_rate": 4.653714317947105e-05,
+      "loss": 0.334,
+      "step": 2200
+    },
+    {
+      "epoch": 1.6560030109145654,
+      "eval_loss": 0.38176649808883667,
+      "eval_runtime": 452.4862,
+      "eval_samples_per_second": 21.276,
+      "eval_steps_per_second": 0.665,
+      "step": 2200
+    },
+    {
+      "epoch": 1.65675573955589,
+      "grad_norm": 0.31346577405929565,
+      "learning_rate": 4.6533176055167735e-05,
+      "loss": 0.4405,
+      "step": 2201
+    },
+    {
+      "epoch": 1.6575084681972148,
+      "grad_norm": 0.290630966424942,
+      "learning_rate": 4.65292068494249e-05,
+      "loss": 0.4297,
+      "step": 2202
+    },
+    {
+      "epoch": 1.6582611968385397,
+      "grad_norm": 0.2757396101951599,
+      "learning_rate": 4.652523556267661e-05,
+      "loss": 0.3632,
+      "step": 2203
+    },
+    {
+      "epoch": 1.6590139254798646,
+      "grad_norm": 0.2819664180278778,
+      "learning_rate": 4.652126219535717e-05,
+      "loss": 0.3679,
+      "step": 2204
+    },
+    {
+      "epoch": 1.6597666541211893,
+      "grad_norm": 0.2560328245162964,
+      "learning_rate": 4.6517286747901094e-05,
+      "loss": 0.3586,
+      "step": 2205
+    },
+    {
+      "epoch": 1.660519382762514,
+      "grad_norm": 0.2666890621185303,
+      "learning_rate": 4.651330922074313e-05,
+      "loss": 0.4307,
+      "step": 2206
+    },
+    {
+      "epoch": 1.661272111403839,
+      "grad_norm": 0.2352440059185028,
+      "learning_rate": 4.650932961431825e-05,
+      "loss": 0.3422,
+      "step": 2207
+    },
+    {
+      "epoch": 1.6620248400451638,
+      "grad_norm": 0.34961479902267456,
+      "learning_rate": 4.650534792906167e-05,
+      "loss": 0.3861,
+      "step": 2208
+    },
+    {
+      "epoch": 1.6627775686864885,
+      "grad_norm": 0.21926981210708618,
+      "learning_rate": 4.6501364165408814e-05,
+      "loss": 0.3017,
+      "step": 2209
+    },
+    {
+      "epoch": 1.6635302973278132,
+      "grad_norm": 0.2729896008968353,
+      "learning_rate": 4.6497378323795345e-05,
+      "loss": 0.3259,
+      "step": 2210
+    },
+    {
+      "epoch": 1.6642830259691381,
+      "grad_norm": 0.28954842686653137,
+      "learning_rate": 4.649339040465713e-05,
+      "loss": 0.3837,
+      "step": 2211
+    },
+    {
+      "epoch": 1.665035754610463,
+      "grad_norm": 0.2482672482728958,
+      "learning_rate": 4.648940040843032e-05,
+      "loss": 0.2932,
+      "step": 2212
+    },
+    {
+      "epoch": 1.6657884832517877,
+      "grad_norm": 0.302124559879303,
+      "learning_rate": 4.64854083355512e-05,
+      "loss": 0.4764,
+      "step": 2213
+    },
+    {
+      "epoch": 1.6665412118931124,
+      "grad_norm": 0.2665206789970398,
+      "learning_rate": 4.64814141864564e-05,
+      "loss": 0.3516,
+      "step": 2214
+    },
+    {
+      "epoch": 1.6672939405344374,
+      "grad_norm": 0.24247440695762634,
+      "learning_rate": 4.647741796158267e-05,
+      "loss": 0.4705,
+      "step": 2215
+    },
+    {
+      "epoch": 1.6680466691757623,
+      "grad_norm": 0.2672916352748871,
+      "learning_rate": 4.647341966136704e-05,
+      "loss": 0.2923,
+      "step": 2216
+    },
+    {
+      "epoch": 1.668799397817087,
+      "grad_norm": 0.24230746924877167,
+      "learning_rate": 4.646941928624676e-05,
+      "loss": 0.4025,
+      "step": 2217
+    },
+    {
+      "epoch": 1.6695521264584117,
+      "grad_norm": 0.27964067459106445,
+      "learning_rate": 4.6465416836659316e-05,
+      "loss": 0.3831,
+      "step": 2218
+    },
+    {
+      "epoch": 1.6703048550997366,
+      "grad_norm": 0.24343279004096985,
+      "learning_rate": 4.646141231304239e-05,
+      "loss": 0.3651,
+      "step": 2219
+    },
+    {
+      "epoch": 1.6710575837410615,
+      "grad_norm": 0.2161562144756317,
+      "learning_rate": 4.645740571583393e-05,
+      "loss": 0.3102,
+      "step": 2220
+    },
+    {
+      "epoch": 1.6718103123823862,
+      "grad_norm": 0.31353604793548584,
+      "learning_rate": 4.6453397045472086e-05,
+      "loss": 0.3353,
+      "step": 2221
+    },
+    {
+      "epoch": 1.672563041023711,
+      "grad_norm": 0.26049166917800903,
+      "learning_rate": 4.644938630239523e-05,
+      "loss": 0.4159,
+      "step": 2222
+    },
+    {
+      "epoch": 1.6733157696650358,
+      "grad_norm": 0.2983473241329193,
+      "learning_rate": 4.644537348704198e-05,
+      "loss": 0.475,
+      "step": 2223
+    },
+    {
+      "epoch": 1.6740684983063605,
+      "grad_norm": 0.2458380162715912,
+      "learning_rate": 4.644135859985117e-05,
+      "loss": 0.4351,
+      "step": 2224
+    },
+    {
+      "epoch": 1.6748212269476852,
+      "grad_norm": 0.26906728744506836,
+      "learning_rate": 4.643734164126186e-05,
+      "loss": 0.4552,
+      "step": 2225
+    },
+    {
+      "epoch": 1.6755739555890101,
+      "grad_norm": 0.2947966456413269,
+      "learning_rate": 4.643332261171334e-05,
+      "loss": 0.4263,
+      "step": 2226
+    },
+    {
+      "epoch": 1.676326684230335,
+      "grad_norm": 0.3028991222381592,
+      "learning_rate": 4.642930151164513e-05,
+      "loss": 0.4798,
+      "step": 2227
+    },
+    {
+      "epoch": 1.6770794128716597,
+      "grad_norm": 0.2621539831161499,
+      "learning_rate": 4.6425278341496956e-05,
+      "loss": 0.4266,
+      "step": 2228
+    },
+    {
+      "epoch": 1.6778321415129844,
+      "grad_norm": 0.24724997580051422,
+      "learning_rate": 4.64212531017088e-05,
+      "loss": 0.2991,
+      "step": 2229
+    },
+    {
+      "epoch": 1.6785848701543093,
+      "grad_norm": 0.31985169649124146,
+      "learning_rate": 4.641722579272086e-05,
+      "loss": 0.3579,
+      "step": 2230
+    },
+    {
+      "epoch": 1.6793375987956343,
+      "grad_norm": 0.29953494668006897,
+      "learning_rate": 4.641319641497355e-05,
+      "loss": 0.4138,
+      "step": 2231
+    },
+    {
+      "epoch": 1.680090327436959,
+      "grad_norm": 0.24221882224082947,
+      "learning_rate": 4.64091649689075e-05,
+      "loss": 0.3343,
+      "step": 2232
+    },
+    {
+      "epoch": 1.6808430560782837,
+      "grad_norm": 0.34155258536338806,
+      "learning_rate": 4.640513145496362e-05,
+      "loss": 0.4428,
+      "step": 2233
+    },
+    {
+      "epoch": 1.6815957847196086,
+      "grad_norm": 0.22447866201400757,
+      "learning_rate": 4.640109587358297e-05,
+      "loss": 0.417,
+      "step": 2234
+    },
+    {
+      "epoch": 1.6823485133609335,
+      "grad_norm": 0.7035013437271118,
+      "learning_rate": 4.63970582252069e-05,
+      "loss": 0.3627,
+      "step": 2235
+    },
+    {
+      "epoch": 1.6831012420022582,
+      "grad_norm": 0.23631885647773743,
+      "learning_rate": 4.639301851027695e-05,
+      "loss": 0.3477,
+      "step": 2236
+    },
+    {
+      "epoch": 1.6838539706435829,
+      "grad_norm": 0.2778174877166748,
+      "learning_rate": 4.63889767292349e-05,
+      "loss": 0.4426,
+      "step": 2237
+    },
+    {
+      "epoch": 1.6846066992849078,
+      "grad_norm": 0.3181927800178528,
+      "learning_rate": 4.638493288252276e-05,
+      "loss": 0.3541,
+      "step": 2238
+    },
+    {
+      "epoch": 1.6853594279262327,
+      "grad_norm": 0.2422560304403305,
+      "learning_rate": 4.638088697058275e-05,
+      "loss": 0.359,
+      "step": 2239
+    },
+    {
+      "epoch": 1.6861121565675574,
+      "grad_norm": 0.24568437039852142,
+      "learning_rate": 4.637683899385733e-05,
+      "loss": 0.3483,
+      "step": 2240
+    },
+    {
+      "epoch": 1.6868648852088821,
+      "grad_norm": 0.3595545291900635,
+      "learning_rate": 4.6372788952789176e-05,
+      "loss": 0.4843,
+      "step": 2241
+    },
+    {
+      "epoch": 1.687617613850207,
+      "grad_norm": 0.22748684883117676,
+      "learning_rate": 4.63687368478212e-05,
+      "loss": 0.2584,
+      "step": 2242
+    },
+    {
+      "epoch": 1.688370342491532,
+      "grad_norm": 0.3867950141429901,
+      "learning_rate": 4.6364682679396525e-05,
+      "loss": 0.5101,
+      "step": 2243
+    },
+    {
+      "epoch": 1.6891230711328566,
+      "grad_norm": 0.3470532298088074,
+      "learning_rate": 4.636062644795851e-05,
+      "loss": 0.3871,
+      "step": 2244
+    },
+    {
+      "epoch": 1.6898757997741813,
+      "grad_norm": 0.28428521752357483,
+      "learning_rate": 4.635656815395076e-05,
+      "loss": 0.349,
+      "step": 2245
+    },
+    {
+      "epoch": 1.6906285284155063,
+      "grad_norm": 0.31740066409111023,
+      "learning_rate": 4.635250779781706e-05,
+      "loss": 0.3637,
+      "step": 2246
+    },
+    {
+      "epoch": 1.6913812570568312,
+      "grad_norm": 0.270161509513855,
+      "learning_rate": 4.634844538000145e-05,
+      "loss": 0.3987,
+      "step": 2247
+    },
+    {
+      "epoch": 1.6921339856981557,
+      "grad_norm": 0.3209688663482666,
+      "learning_rate": 4.6344380900948195e-05,
+      "loss": 0.4184,
+      "step": 2248
+    },
+    {
+      "epoch": 1.6928867143394806,
+      "grad_norm": 0.31322458386421204,
+      "learning_rate": 4.6340314361101766e-05,
+      "loss": 0.3854,
+      "step": 2249
+    },
+    {
+      "epoch": 1.6936394429808055,
+      "grad_norm": 0.3249536156654358,
+      "learning_rate": 4.6336245760906896e-05,
+      "loss": 0.4423,
+      "step": 2250
+    },
+    {
+      "epoch": 1.6943921716221302,
+      "grad_norm": 0.2755548059940338,
+      "learning_rate": 4.63321751008085e-05,
+      "loss": 0.5334,
+      "step": 2251
+    },
+    {
+      "epoch": 1.6951449002634549,
+      "grad_norm": 0.28478583693504333,
+      "learning_rate": 4.632810238125176e-05,
+      "loss": 0.3336,
+      "step": 2252
+    },
+    {
+      "epoch": 1.6958976289047798,
+      "grad_norm": 0.2255488932132721,
+      "learning_rate": 4.632402760268205e-05,
+      "loss": 0.3582,
+      "step": 2253
+    },
+    {
+      "epoch": 1.6966503575461047,
+      "grad_norm": 0.2774214446544647,
+      "learning_rate": 4.631995076554497e-05,
+      "loss": 0.3594,
+      "step": 2254
+    },
+    {
+      "epoch": 1.6974030861874294,
+      "grad_norm": 0.393932044506073,
+      "learning_rate": 4.631587187028638e-05,
+      "loss": 0.3836,
+      "step": 2255
+    },
+    {
+      "epoch": 1.698155814828754,
+      "grad_norm": 0.2484443187713623,
+      "learning_rate": 4.6311790917352334e-05,
+      "loss": 0.3953,
+      "step": 2256
+    },
+    {
+      "epoch": 1.698908543470079,
+      "grad_norm": 0.3399399518966675,
+      "learning_rate": 4.630770790718911e-05,
+      "loss": 0.4551,
+      "step": 2257
+    },
+    {
+      "epoch": 1.699661272111404,
+      "grad_norm": 0.24737682938575745,
+      "learning_rate": 4.630362284024322e-05,
+      "loss": 0.4563,
+      "step": 2258
+    },
+    {
+      "epoch": 1.7004140007527286,
+      "grad_norm": 0.3255244195461273,
+      "learning_rate": 4.629953571696142e-05,
+      "loss": 0.3775,
+      "step": 2259
+    },
+    {
+      "epoch": 1.7011667293940533,
+      "grad_norm": 0.27298587560653687,
+      "learning_rate": 4.6295446537790654e-05,
+      "loss": 0.4083,
+      "step": 2260
+    },
+    {
+      "epoch": 1.7019194580353783,
+      "grad_norm": 0.26676708459854126,
+      "learning_rate": 4.6291355303178115e-05,
+      "loss": 0.3424,
+      "step": 2261
+    },
+    {
+      "epoch": 1.7026721866767032,
+      "grad_norm": 0.3207310140132904,
+      "learning_rate": 4.628726201357121e-05,
+      "loss": 0.3945,
+      "step": 2262
+    },
+    {
+      "epoch": 1.7034249153180279,
+      "grad_norm": 0.2617543637752533,
+      "learning_rate": 4.6283166669417576e-05,
+      "loss": 0.4059,
+      "step": 2263
+    },
+    {
+      "epoch": 1.7041776439593526,
+      "grad_norm": 0.2804131805896759,
+      "learning_rate": 4.627906927116508e-05,
+      "loss": 0.4803,
+      "step": 2264
+    },
+    {
+      "epoch": 1.7049303726006775,
+      "grad_norm": 0.3003084361553192,
+      "learning_rate": 4.6274969819261795e-05,
+      "loss": 0.4809,
+      "step": 2265
+    },
+    {
+      "epoch": 1.7056831012420024,
+      "grad_norm": 0.24612300097942352,
+      "learning_rate": 4.627086831415605e-05,
+      "loss": 0.3818,
+      "step": 2266
+    },
+    {
+      "epoch": 1.706435829883327,
+      "grad_norm": 0.2848217487335205,
+      "learning_rate": 4.6266764756296356e-05,
+      "loss": 0.4295,
+      "step": 2267
+    },
+    {
+      "epoch": 1.7071885585246518,
+      "grad_norm": 0.2885759770870209,
+      "learning_rate": 4.626265914613149e-05,
+      "loss": 0.3418,
+      "step": 2268
+    },
+    {
+      "epoch": 1.7079412871659767,
+      "grad_norm": 0.2393244355916977,
+      "learning_rate": 4.625855148411042e-05,
+      "loss": 0.3251,
+      "step": 2269
+    },
+    {
+      "epoch": 1.7086940158073016,
+      "grad_norm": 0.25335177779197693,
+      "learning_rate": 4.6254441770682374e-05,
+      "loss": 0.3893,
+      "step": 2270
+    },
+    {
+      "epoch": 1.7094467444486263,
+      "grad_norm": 0.29222571849823,
+      "learning_rate": 4.6250330006296764e-05,
+      "loss": 0.4316,
+      "step": 2271
+    },
+    {
+      "epoch": 1.710199473089951,
+      "grad_norm": 0.3059636652469635,
+      "learning_rate": 4.6246216191403264e-05,
+      "loss": 0.3826,
+      "step": 2272
+    },
+    {
+      "epoch": 1.710952201731276,
+      "grad_norm": 0.2711617648601532,
+      "learning_rate": 4.624210032645174e-05,
+      "loss": 0.3991,
+      "step": 2273
+    },
+    {
+      "epoch": 1.7117049303726006,
+      "grad_norm": 0.21739782392978668,
+      "learning_rate": 4.6237982411892306e-05,
+      "loss": 0.335,
+      "step": 2274
+    },
+    {
+      "epoch": 1.7124576590139253,
+      "grad_norm": 0.2823951542377472,
+      "learning_rate": 4.6233862448175284e-05,
+      "loss": 0.473,
+      "step": 2275
+    },
+    {
+      "epoch": 1.7132103876552502,
+      "grad_norm": 0.25580480694770813,
+      "learning_rate": 4.6229740435751226e-05,
+      "loss": 0.3806,
+      "step": 2276
+    },
+    {
+      "epoch": 1.7139631162965752,
+      "grad_norm": 0.2669616937637329,
+      "learning_rate": 4.622561637507092e-05,
+      "loss": 0.3859,
+      "step": 2277
+    },
+    {
+      "epoch": 1.7147158449378999,
+      "grad_norm": 0.23198086023330688,
+      "learning_rate": 4.622149026658537e-05,
+      "loss": 0.3089,
+      "step": 2278
+    },
+    {
+      "epoch": 1.7154685735792246,
+      "grad_norm": 0.27069026231765747,
+      "learning_rate": 4.6217362110745775e-05,
+      "loss": 0.3624,
+      "step": 2279
+    },
+    {
+      "epoch": 1.7162213022205495,
+      "grad_norm": 0.2512904107570648,
+      "learning_rate": 4.621323190800361e-05,
+      "loss": 0.3441,
+      "step": 2280
+    },
+    {
+      "epoch": 1.7169740308618744,
+      "grad_norm": 0.2323940545320511,
+      "learning_rate": 4.620909965881054e-05,
+      "loss": 0.314,
+      "step": 2281
+    },
+    {
+      "epoch": 1.717726759503199,
+      "grad_norm": 0.21540698409080505,
+      "learning_rate": 4.620496536361846e-05,
+      "loss": 0.3099,
+      "step": 2282
+    },
+    {
+      "epoch": 1.7184794881445238,
+      "grad_norm": 0.22523480653762817,
+      "learning_rate": 4.6200829022879484e-05,
+      "loss": 0.3867,
+      "step": 2283
+    },
+    {
+      "epoch": 1.7192322167858487,
+      "grad_norm": 0.26673033833503723,
+      "learning_rate": 4.619669063704596e-05,
+      "loss": 0.4717,
+      "step": 2284
+    },
+    {
+      "epoch": 1.7199849454271736,
+      "grad_norm": 0.27050355076789856,
+      "learning_rate": 4.619255020657047e-05,
+      "loss": 0.3865,
+      "step": 2285
+    },
+    {
+      "epoch": 1.7207376740684983,
+      "grad_norm": 0.2434975504875183,
+      "learning_rate": 4.6188407731905787e-05,
+      "loss": 0.3407,
+      "step": 2286
+    },
+    {
+      "epoch": 1.721490402709823,
+      "grad_norm": 0.22297044098377228,
+      "learning_rate": 4.618426321350494e-05,
+      "loss": 0.4288,
+      "step": 2287
+    },
+    {
+      "epoch": 1.722243131351148,
+      "grad_norm": 0.21115157008171082,
+      "learning_rate": 4.6180116651821146e-05,
+      "loss": 0.4401,
+      "step": 2288
+    },
+    {
+      "epoch": 1.7229958599924728,
+      "grad_norm": 0.2249159961938858,
+      "learning_rate": 4.617596804730788e-05,
+      "loss": 0.3814,
+      "step": 2289
+    },
+    {
+      "epoch": 1.7237485886337975,
+      "grad_norm": 0.2504001259803772,
+      "learning_rate": 4.617181740041884e-05,
+      "loss": 0.2977,
+      "step": 2290
+    },
+    {
+      "epoch": 1.7245013172751222,
+      "grad_norm": 0.2570140063762665,
+      "learning_rate": 4.616766471160791e-05,
+      "loss": 0.4191,
+      "step": 2291
+    },
+    {
+      "epoch": 1.7252540459164472,
+      "grad_norm": 0.24738752841949463,
+      "learning_rate": 4.616350998132923e-05,
+      "loss": 0.3751,
+      "step": 2292
+    },
+    {
+      "epoch": 1.726006774557772,
+      "grad_norm": 0.22536700963974,
+      "learning_rate": 4.615935321003717e-05,
+      "loss": 0.3861,
+      "step": 2293
+    },
+    {
+      "epoch": 1.7267595031990968,
+      "grad_norm": 0.2662505805492401,
+      "learning_rate": 4.61551943981863e-05,
+      "loss": 0.3857,
+      "step": 2294
+    },
+    {
+      "epoch": 1.7275122318404215,
+      "grad_norm": 0.25364816188812256,
+      "learning_rate": 4.615103354623141e-05,
+      "loss": 0.3861,
+      "step": 2295
+    },
+    {
+      "epoch": 1.7282649604817464,
+      "grad_norm": 0.2839832007884979,
+      "learning_rate": 4.614687065462753e-05,
+      "loss": 0.4882,
+      "step": 2296
+    },
+    {
+      "epoch": 1.729017689123071,
+      "grad_norm": 0.2602632939815521,
+      "learning_rate": 4.614270572382991e-05,
+      "loss": 0.351,
+      "step": 2297
+    },
+    {
+      "epoch": 1.7297704177643958,
+      "grad_norm": 0.286688894033432,
+      "learning_rate": 4.613853875429403e-05,
+      "loss": 0.4577,
+      "step": 2298
+    },
+    {
+      "epoch": 1.7305231464057207,
+      "grad_norm": 0.3134872019290924,
+      "learning_rate": 4.613436974647557e-05,
+      "loss": 0.3866,
+      "step": 2299
+    },
+    {
+      "epoch": 1.7312758750470456,
+      "grad_norm": 0.2554290294647217,
+      "learning_rate": 4.6130198700830455e-05,
+      "loss": 0.394,
+      "step": 2300
+    },
+    {
+      "epoch": 1.7320286036883703,
+      "grad_norm": 0.20738469064235687,
+      "learning_rate": 4.6126025617814814e-05,
+      "loss": 0.3012,
+      "step": 2301
+    },
+    {
+      "epoch": 1.732781332329695,
+      "grad_norm": 0.27413398027420044,
+      "learning_rate": 4.612185049788502e-05,
+      "loss": 0.3629,
+      "step": 2302
+    },
+    {
+      "epoch": 1.73353406097102,
+      "grad_norm": 0.23334622383117676,
+      "learning_rate": 4.611767334149765e-05,
+      "loss": 0.362,
+      "step": 2303
+    },
+    {
+      "epoch": 1.7342867896123448,
+      "grad_norm": 0.28028830885887146,
+      "learning_rate": 4.6113494149109513e-05,
+      "loss": 0.3481,
+      "step": 2304
+    },
+    {
+      "epoch": 1.7350395182536695,
+      "grad_norm": 0.23703761398792267,
+      "learning_rate": 4.610931292117764e-05,
+      "loss": 0.4098,
+      "step": 2305
+    },
+    {
+      "epoch": 1.7357922468949942,
+      "grad_norm": 0.26964184641838074,
+      "learning_rate": 4.61051296581593e-05,
+      "loss": 0.3876,
+      "step": 2306
+    },
+    {
+      "epoch": 1.7365449755363191,
+      "grad_norm": 0.2805062532424927,
+      "learning_rate": 4.610094436051193e-05,
+      "loss": 0.5035,
+      "step": 2307
+    },
+    {
+      "epoch": 1.737297704177644,
+      "grad_norm": 0.3285451829433441,
+      "learning_rate": 4.609675702869326e-05,
+      "loss": 0.4325,
+      "step": 2308
+    },
+    {
+      "epoch": 1.7380504328189688,
+      "grad_norm": 0.4203299880027771,
+      "learning_rate": 4.6092567663161206e-05,
+      "loss": 0.3835,
+      "step": 2309
+    },
+    {
+      "epoch": 1.7388031614602935,
+      "grad_norm": 0.2614361345767975,
+      "learning_rate": 4.60883762643739e-05,
+      "loss": 0.4049,
+      "step": 2310
+    },
+    {
+      "epoch": 1.7395558901016184,
+      "grad_norm": 0.25212642550468445,
+      "learning_rate": 4.6084182832789715e-05,
+      "loss": 0.3495,
+      "step": 2311
+    },
+    {
+      "epoch": 1.7403086187429433,
+      "grad_norm": 0.22141090035438538,
+      "learning_rate": 4.607998736886724e-05,
+      "loss": 0.3171,
+      "step": 2312
+    },
+    {
+      "epoch": 1.741061347384268,
+      "grad_norm": 0.24542848765850067,
+      "learning_rate": 4.607578987306528e-05,
+      "loss": 0.3289,
+      "step": 2313
+    },
+    {
+      "epoch": 1.7418140760255927,
+      "grad_norm": 0.2744753658771515,
+      "learning_rate": 4.607159034584286e-05,
+      "loss": 0.3764,
+      "step": 2314
+    },
+    {
+      "epoch": 1.7425668046669176,
+      "grad_norm": 0.2489544004201889,
+      "learning_rate": 4.606738878765926e-05,
+      "loss": 0.4,
+      "step": 2315
+    },
+    {
+      "epoch": 1.7433195333082425,
+      "grad_norm": 0.2702884078025818,
+      "learning_rate": 4.6063185198973916e-05,
+      "loss": 0.3465,
+      "step": 2316
+    },
+    {
+      "epoch": 1.7440722619495672,
+      "grad_norm": 0.2598249018192291,
+      "learning_rate": 4.6058979580246554e-05,
+      "loss": 0.3484,
+      "step": 2317
+    },
+    {
+      "epoch": 1.744824990590892,
+      "grad_norm": 0.2704434394836426,
+      "learning_rate": 4.605477193193709e-05,
+      "loss": 0.3127,
+      "step": 2318
+    },
+    {
+      "epoch": 1.7455777192322168,
+      "grad_norm": 0.33671027421951294,
+      "learning_rate": 4.605056225450567e-05,
+      "loss": 0.4319,
+      "step": 2319
+    },
+    {
+      "epoch": 1.7463304478735417,
+      "grad_norm": 0.3115230202674866,
+      "learning_rate": 4.604635054841265e-05,
+      "loss": 0.4238,
+      "step": 2320
+    },
+    {
+      "epoch": 1.7470831765148664,
+      "grad_norm": 0.2963281571865082,
+      "learning_rate": 4.6042136814118605e-05,
+      "loss": 0.2746,
+      "step": 2321
+    },
+    {
+      "epoch": 1.7478359051561911,
+      "grad_norm": 0.3261544406414032,
+      "learning_rate": 4.6037921052084374e-05,
+      "loss": 0.3225,
+      "step": 2322
+    },
+    {
+      "epoch": 1.748588633797516,
+      "grad_norm": 0.24824732542037964,
+      "learning_rate": 4.603370326277096e-05,
+      "loss": 0.3799,
+      "step": 2323
+    },
+    {
+      "epoch": 1.7493413624388408,
+      "grad_norm": 0.26168951392173767,
+      "learning_rate": 4.6029483446639616e-05,
+      "loss": 0.3767,
+      "step": 2324
+    },
+    {
+      "epoch": 1.7500940910801654,
+      "grad_norm": 0.2888537049293518,
+      "learning_rate": 4.6025261604151835e-05,
+      "loss": 0.3625,
+      "step": 2325
+    },
+    {
+      "epoch": 1.7508468197214904,
+      "grad_norm": 0.2613416910171509,
+      "learning_rate": 4.602103773576929e-05,
+      "loss": 0.3925,
+      "step": 2326
+    },
+    {
+      "epoch": 1.7515995483628153,
+      "grad_norm": 0.24124504625797272,
+      "learning_rate": 4.60168118419539e-05,
+      "loss": 0.3539,
+      "step": 2327
+    },
+    {
+      "epoch": 1.75235227700414,
+      "grad_norm": 0.2823038101196289,
+      "learning_rate": 4.601258392316781e-05,
+      "loss": 0.4064,
+      "step": 2328
+    },
+    {
+      "epoch": 1.7531050056454647,
+      "grad_norm": 0.26040756702423096,
+      "learning_rate": 4.600835397987338e-05,
+      "loss": 0.4316,
+      "step": 2329
+    },
+    {
+      "epoch": 1.7538577342867896,
+      "grad_norm": 0.2966538965702057,
+      "learning_rate": 4.6004122012533176e-05,
+      "loss": 0.5054,
+      "step": 2330
+    },
+    {
+      "epoch": 1.7546104629281145,
+      "grad_norm": 0.25666168332099915,
+      "learning_rate": 4.599988802161001e-05,
+      "loss": 0.374,
+      "step": 2331
+    },
+    {
+      "epoch": 1.7553631915694392,
+      "grad_norm": 0.27189478278160095,
+      "learning_rate": 4.599565200756691e-05,
+      "loss": 0.2997,
+      "step": 2332
+    },
+    {
+      "epoch": 1.756115920210764,
+      "grad_norm": 0.2394535094499588,
+      "learning_rate": 4.599141397086712e-05,
+      "loss": 0.3532,
+      "step": 2333
+    },
+    {
+      "epoch": 1.7568686488520888,
+      "grad_norm": 0.2856077253818512,
+      "learning_rate": 4.598717391197409e-05,
+      "loss": 0.337,
+      "step": 2334
+    },
+    {
+      "epoch": 1.7576213774934137,
+      "grad_norm": 0.272046834230423,
+      "learning_rate": 4.5982931831351516e-05,
+      "loss": 0.4084,
+      "step": 2335
+    },
+    {
+      "epoch": 1.7583741061347384,
+      "grad_norm": 0.23759464919567108,
+      "learning_rate": 4.597868772946331e-05,
+      "loss": 0.375,
+      "step": 2336
+    },
+    {
+      "epoch": 1.7591268347760631,
+      "grad_norm": 0.27061927318573,
+      "learning_rate": 4.5974441606773594e-05,
+      "loss": 0.4816,
+      "step": 2337
+    },
+    {
+      "epoch": 1.759879563417388,
+      "grad_norm": 0.2385004311800003,
+      "learning_rate": 4.597019346374673e-05,
+      "loss": 0.382,
+      "step": 2338
+    },
+    {
+      "epoch": 1.760632292058713,
+      "grad_norm": 0.24871304631233215,
+      "learning_rate": 4.5965943300847255e-05,
+      "loss": 0.4894,
+      "step": 2339
+    },
+    {
+      "epoch": 1.7613850207000377,
+      "grad_norm": 0.26726284623146057,
+      "learning_rate": 4.5961691118540004e-05,
+      "loss": 0.4348,
+      "step": 2340
+    },
+    {
+      "epoch": 1.7621377493413624,
+      "grad_norm": 0.22760652005672455,
+      "learning_rate": 4.5957436917289966e-05,
+      "loss": 0.2806,
+      "step": 2341
+    },
+    {
+      "epoch": 1.7628904779826873,
+      "grad_norm": 0.22753581404685974,
+      "learning_rate": 4.595318069756237e-05,
+      "loss": 0.3488,
+      "step": 2342
+    },
+    {
+      "epoch": 1.7636432066240122,
+      "grad_norm": 0.2947016656398773,
+      "learning_rate": 4.594892245982268e-05,
+      "loss": 0.2289,
+      "step": 2343
+    },
+    {
+      "epoch": 1.764395935265337,
+      "grad_norm": 0.27869096398353577,
+      "learning_rate": 4.5944662204536566e-05,
+      "loss": 0.4224,
+      "step": 2344
+    },
+    {
+      "epoch": 1.7651486639066616,
+      "grad_norm": 0.24218086898326874,
+      "learning_rate": 4.5940399932169925e-05,
+      "loss": 0.4657,
+      "step": 2345
+    },
+    {
+      "epoch": 1.7659013925479865,
+      "grad_norm": 0.28593766689300537,
+      "learning_rate": 4.5936135643188876e-05,
+      "loss": 0.3677,
+      "step": 2346
+    },
+    {
+      "epoch": 1.7666541211893112,
+      "grad_norm": 0.27382147312164307,
+      "learning_rate": 4.593186933805975e-05,
+      "loss": 0.4198,
+      "step": 2347
+    },
+    {
+      "epoch": 1.767406849830636,
+      "grad_norm": 0.20806118845939636,
+      "learning_rate": 4.59276010172491e-05,
+      "loss": 0.3103,
+      "step": 2348
+    },
+    {
+      "epoch": 1.7681595784719608,
+      "grad_norm": 0.30040886998176575,
+      "learning_rate": 4.592333068122372e-05,
+      "loss": 0.3232,
+      "step": 2349
+    },
+    {
+      "epoch": 1.7689123071132857,
+      "grad_norm": 0.23971568048000336,
+      "learning_rate": 4.591905833045059e-05,
+      "loss": 0.3308,
+      "step": 2350
+    },
+    {
+      "epoch": 1.7696650357546104,
+      "grad_norm": 0.2692120373249054,
+      "learning_rate": 4.591478396539692e-05,
+      "loss": 0.4631,
+      "step": 2351
+    },
+    {
+      "epoch": 1.7704177643959351,
+      "grad_norm": 0.3033456802368164,
+      "learning_rate": 4.591050758653018e-05,
+      "loss": 0.3231,
+      "step": 2352
+    },
+    {
+      "epoch": 1.77117049303726,
+      "grad_norm": 0.2504591643810272,
+      "learning_rate": 4.5906229194318e-05,
+      "loss": 0.4266,
+      "step": 2353
+    },
+    {
+      "epoch": 1.771923221678585,
+      "grad_norm": 0.24774199724197388,
+      "learning_rate": 4.5901948789228274e-05,
+      "loss": 0.4192,
+      "step": 2354
+    },
+    {
+      "epoch": 1.7726759503199097,
+      "grad_norm": 0.2458532601594925,
+      "learning_rate": 4.589766637172909e-05,
+      "loss": 0.3429,
+      "step": 2355
+    },
+    {
+      "epoch": 1.7734286789612344,
+      "grad_norm": 0.20604251325130463,
+      "learning_rate": 4.589338194228878e-05,
+      "loss": 0.3056,
+      "step": 2356
+    },
+    {
+      "epoch": 1.7741814076025593,
+      "grad_norm": 0.24787637591362,
+      "learning_rate": 4.5889095501375865e-05,
+      "loss": 0.3552,
+      "step": 2357
+    },
+    {
+      "epoch": 1.7749341362438842,
+      "grad_norm": 0.24976995587348938,
+      "learning_rate": 4.588480704945912e-05,
+      "loss": 0.338,
+      "step": 2358
+    },
+    {
+      "epoch": 1.7756868648852089,
+      "grad_norm": 0.2508619427680969,
+      "learning_rate": 4.588051658700751e-05,
+      "loss": 0.3445,
+      "step": 2359
+    },
+    {
+      "epoch": 1.7764395935265336,
+      "grad_norm": 0.32370343804359436,
+      "learning_rate": 4.587622411449025e-05,
+      "loss": 0.4665,
+      "step": 2360
+    },
+    {
+      "epoch": 1.7771923221678585,
+      "grad_norm": 0.3169470727443695,
+      "learning_rate": 4.587192963237675e-05,
+      "loss": 0.4385,
+      "step": 2361
+    },
+    {
+      "epoch": 1.7779450508091834,
+      "grad_norm": 0.31540465354919434,
+      "learning_rate": 4.5867633141136646e-05,
+      "loss": 0.3723,
+      "step": 2362
+    },
+    {
+      "epoch": 1.778697779450508,
+      "grad_norm": 0.29631751775741577,
+      "learning_rate": 4.5863334641239786e-05,
+      "loss": 0.3026,
+      "step": 2363
+    },
+    {
+      "epoch": 1.7794505080918328,
+      "grad_norm": 0.4219616651535034,
+      "learning_rate": 4.585903413315628e-05,
+      "loss": 0.4119,
+      "step": 2364
+    },
+    {
+      "epoch": 1.7802032367331577,
+      "grad_norm": 0.30385881662368774,
+      "learning_rate": 4.585473161735638e-05,
+      "loss": 0.3882,
+      "step": 2365
+    },
+    {
+      "epoch": 1.7809559653744826,
+      "grad_norm": 0.35439008474349976,
+      "learning_rate": 4.585042709431063e-05,
+      "loss": 0.2965,
+      "step": 2366
+    },
+    {
+      "epoch": 1.7817086940158073,
+      "grad_norm": 0.305561363697052,
+      "learning_rate": 4.5846120564489776e-05,
+      "loss": 0.41,
+      "step": 2367
+    },
+    {
+      "epoch": 1.782461422657132,
+      "grad_norm": 0.2389414757490158,
+      "learning_rate": 4.584181202836475e-05,
+      "loss": 0.2934,
+      "step": 2368
+    },
+    {
+      "epoch": 1.783214151298457,
+      "grad_norm": 0.43820086121559143,
+      "learning_rate": 4.583750148640674e-05,
+      "loss": 0.3974,
+      "step": 2369
+    },
+    {
+      "epoch": 1.7839668799397819,
+      "grad_norm": 0.3090749680995941,
+      "learning_rate": 4.5833188939087133e-05,
+      "loss": 0.3521,
+      "step": 2370
+    },
+    {
+      "epoch": 1.7847196085811066,
+      "grad_norm": 0.23702551424503326,
+      "learning_rate": 4.5828874386877555e-05,
+      "loss": 0.3594,
+      "step": 2371
+    },
+    {
+      "epoch": 1.7854723372224313,
+      "grad_norm": 0.2542990446090698,
+      "learning_rate": 4.5824557830249834e-05,
+      "loss": 0.3995,
+      "step": 2372
+    },
+    {
+      "epoch": 1.7862250658637562,
+      "grad_norm": 0.22670115530490875,
+      "learning_rate": 4.582023926967601e-05,
+      "loss": 0.4123,
+      "step": 2373
+    },
+    {
+      "epoch": 1.7869777945050809,
+      "grad_norm": 0.25064897537231445,
+      "learning_rate": 4.5815918705628365e-05,
+      "loss": 0.3729,
+      "step": 2374
+    },
+    {
+      "epoch": 1.7877305231464056,
+      "grad_norm": 0.2561325132846832,
+      "learning_rate": 4.581159613857939e-05,
+      "loss": 0.303,
+      "step": 2375
+    },
+    {
+      "epoch": 1.7884832517877305,
+      "grad_norm": 0.25866997241973877,
+      "learning_rate": 4.580727156900179e-05,
+      "loss": 0.3513,
+      "step": 2376
+    },
+    {
+      "epoch": 1.7892359804290554,
+      "grad_norm": 0.3275188207626343,
+      "learning_rate": 4.58029449973685e-05,
+      "loss": 0.3293,
+      "step": 2377
+    },
+    {
+      "epoch": 1.78998870907038,
+      "grad_norm": 0.33279067277908325,
+      "learning_rate": 4.579861642415266e-05,
+      "loss": 0.3378,
+      "step": 2378
+    },
+    {
+      "epoch": 1.7907414377117048,
+      "grad_norm": 0.28028014302253723,
+      "learning_rate": 4.5794285849827636e-05,
+      "loss": 0.3715,
+      "step": 2379
+    },
+    {
+      "epoch": 1.7914941663530297,
+      "grad_norm": 0.29213613271713257,
+      "learning_rate": 4.578995327486703e-05,
+      "loss": 0.3752,
+      "step": 2380
+    },
+    {
+      "epoch": 1.7922468949943546,
+      "grad_norm": 0.25271281599998474,
+      "learning_rate": 4.578561869974462e-05,
+      "loss": 0.3525,
+      "step": 2381
+    },
+    {
+      "epoch": 1.7929996236356793,
+      "grad_norm": 0.3121004104614258,
+      "learning_rate": 4.578128212493444e-05,
+      "loss": 0.4408,
+      "step": 2382
+    },
+    {
+      "epoch": 1.793752352277004,
+      "grad_norm": 0.3498516082763672,
+      "learning_rate": 4.5776943550910733e-05,
+      "loss": 0.3903,
+      "step": 2383
+    },
+    {
+      "epoch": 1.794505080918329,
+      "grad_norm": 0.28842082619667053,
+      "learning_rate": 4.5772602978147964e-05,
+      "loss": 0.4375,
+      "step": 2384
+    },
+    {
+      "epoch": 1.7952578095596539,
+      "grad_norm": 0.2741520404815674,
+      "learning_rate": 4.57682604071208e-05,
+      "loss": 0.3845,
+      "step": 2385
+    },
+    {
+      "epoch": 1.7960105382009786,
+      "grad_norm": 0.32068711519241333,
+      "learning_rate": 4.576391583830415e-05,
+      "loss": 0.443,
+      "step": 2386
+    },
+    {
+      "epoch": 1.7967632668423033,
+      "grad_norm": 0.3074943721294403,
+      "learning_rate": 4.575956927217312e-05,
+      "loss": 0.3619,
+      "step": 2387
+    },
+    {
+      "epoch": 1.7975159954836282,
+      "grad_norm": 0.2288161963224411,
+      "learning_rate": 4.575522070920305e-05,
+      "loss": 0.3672,
+      "step": 2388
+    },
+    {
+      "epoch": 1.798268724124953,
+      "grad_norm": 0.29897555708885193,
+      "learning_rate": 4.5750870149869484e-05,
+      "loss": 0.3781,
+      "step": 2389
+    },
+    {
+      "epoch": 1.7990214527662778,
+      "grad_norm": 0.24602848291397095,
+      "learning_rate": 4.5746517594648205e-05,
+      "loss": 0.4431,
+      "step": 2390
+    },
+    {
+      "epoch": 1.7997741814076025,
+      "grad_norm": 0.3189345598220825,
+      "learning_rate": 4.574216304401519e-05,
+      "loss": 0.4031,
+      "step": 2391
+    },
+    {
+      "epoch": 1.8005269100489274,
+      "grad_norm": 0.2976755201816559,
+      "learning_rate": 4.573780649844665e-05,
+      "loss": 0.3907,
+      "step": 2392
+    },
+    {
+      "epoch": 1.8012796386902523,
+      "grad_norm": 0.23533383011817932,
+      "learning_rate": 4.573344795841901e-05,
+      "loss": 0.2939,
+      "step": 2393
+    },
+    {
+      "epoch": 1.802032367331577,
+      "grad_norm": 0.27129873633384705,
+      "learning_rate": 4.572908742440892e-05,
+      "loss": 0.3265,
+      "step": 2394
+    },
+    {
+      "epoch": 1.8027850959729017,
+      "grad_norm": 0.27881139516830444,
+      "learning_rate": 4.572472489689324e-05,
+      "loss": 0.343,
+      "step": 2395
+    },
+    {
+      "epoch": 1.8035378246142266,
+      "grad_norm": 0.27811944484710693,
+      "learning_rate": 4.5720360376349045e-05,
+      "loss": 0.3438,
+      "step": 2396
+    },
+    {
+      "epoch": 1.8042905532555513,
+      "grad_norm": 0.28822869062423706,
+      "learning_rate": 4.571599386325362e-05,
+      "loss": 0.4022,
+      "step": 2397
+    },
+    {
+      "epoch": 1.805043281896876,
+      "grad_norm": 0.2799322009086609,
+      "learning_rate": 4.5711625358084506e-05,
+      "loss": 0.4235,
+      "step": 2398
+    },
+    {
+      "epoch": 1.805796010538201,
+      "grad_norm": 0.252463161945343,
+      "learning_rate": 4.570725486131943e-05,
+      "loss": 0.3327,
+      "step": 2399
+    },
+    {
+      "epoch": 1.8065487391795259,
+      "grad_norm": 0.28559190034866333,
+      "learning_rate": 4.570288237343632e-05,
+      "loss": 0.3336,
+      "step": 2400
+    },
+    {
+      "epoch": 1.8065487391795259,
+      "eval_loss": 0.36480680108070374,
+      "eval_runtime": 452.815,
+      "eval_samples_per_second": 21.26,
+      "eval_steps_per_second": 0.665,
+      "step": 2400
+    },
+    {
+      "epoch": 1.8073014678208505,
+      "grad_norm": 0.23379391431808472,
+      "learning_rate": 4.5698507894913364e-05,
+      "loss": 0.4555,
+      "step": 2401
+    },
+    {
+      "epoch": 1.8080541964621752,
+      "grad_norm": 0.25998544692993164,
+      "learning_rate": 4.569413142622895e-05,
+      "loss": 0.4543,
+      "step": 2402
+    },
+    {
+      "epoch": 1.8088069251035002,
+      "grad_norm": 0.2589680850505829,
+      "learning_rate": 4.5689752967861675e-05,
+      "loss": 0.3175,
+      "step": 2403
+    },
+    {
+      "epoch": 1.809559653744825,
+      "grad_norm": 0.34279394149780273,
+      "learning_rate": 4.568537252029037e-05,
+      "loss": 0.3191,
+      "step": 2404
+    },
+    {
+      "epoch": 1.8103123823861498,
+      "grad_norm": 0.2907181978225708,
+      "learning_rate": 4.568099008399406e-05,
+      "loss": 0.4051,
+      "step": 2405
+    },
+    {
+      "epoch": 1.8110651110274745,
+      "grad_norm": 0.3784552216529846,
+      "learning_rate": 4.5676605659452015e-05,
+      "loss": 0.3826,
+      "step": 2406
+    },
+    {
+      "epoch": 1.8118178396687994,
+      "grad_norm": 0.2411859929561615,
+      "learning_rate": 4.56722192471437e-05,
+      "loss": 0.46,
+      "step": 2407
+    },
+    {
+      "epoch": 1.8125705683101243,
+      "grad_norm": 0.3865858018398285,
+      "learning_rate": 4.5667830847548816e-05,
+      "loss": 0.4035,
+      "step": 2408
+    },
+    {
+      "epoch": 1.813323296951449,
+      "grad_norm": 0.2817407250404358,
+      "learning_rate": 4.566344046114726e-05,
+      "loss": 0.3987,
+      "step": 2409
+    },
+    {
+      "epoch": 1.8140760255927737,
+      "grad_norm": 0.249583899974823,
+      "learning_rate": 4.565904808841916e-05,
+      "loss": 0.3887,
+      "step": 2410
+    },
+    {
+      "epoch": 1.8148287542340986,
+      "grad_norm": 0.35986873507499695,
+      "learning_rate": 4.565465372984487e-05,
+      "loss": 0.4277,
+      "step": 2411
+    },
+    {
+      "epoch": 1.8155814828754235,
+      "grad_norm": 0.2520371079444885,
+      "learning_rate": 4.565025738590494e-05,
+      "loss": 0.3564,
+      "step": 2412
+    },
+    {
+      "epoch": 1.8163342115167482,
+      "grad_norm": 0.24413807690143585,
+      "learning_rate": 4.5645859057080145e-05,
+      "loss": 0.3209,
+      "step": 2413
+    },
+    {
+      "epoch": 1.817086940158073,
+      "grad_norm": 0.2906354069709778,
+      "learning_rate": 4.564145874385149e-05,
+      "loss": 0.2622,
+      "step": 2414
+    },
+    {
+      "epoch": 1.8178396687993978,
+      "grad_norm": 0.2434302568435669,
+      "learning_rate": 4.563705644670018e-05,
+      "loss": 0.4375,
+      "step": 2415
+    },
+    {
+      "epoch": 1.8185923974407228,
+      "grad_norm": 0.24004963040351868,
+      "learning_rate": 4.563265216610766e-05,
+      "loss": 0.3374,
+      "step": 2416
+    },
+    {
+      "epoch": 1.8193451260820475,
+      "grad_norm": 0.28764358162879944,
+      "learning_rate": 4.562824590255553e-05,
+      "loss": 0.3465,
+      "step": 2417
+    },
+    {
+      "epoch": 1.8200978547233722,
+      "grad_norm": 0.24592001736164093,
+      "learning_rate": 4.5623837656525706e-05,
+      "loss": 0.3422,
+      "step": 2418
+    },
+    {
+      "epoch": 1.820850583364697,
+      "grad_norm": 0.23703496158123016,
+      "learning_rate": 4.5619427428500246e-05,
+      "loss": 0.3158,
+      "step": 2419
+    },
+    {
+      "epoch": 1.821603312006022,
+      "grad_norm": 0.29672378301620483,
+      "learning_rate": 4.561501521896144e-05,
+      "loss": 0.3984,
+      "step": 2420
+    },
+    {
+      "epoch": 1.8223560406473467,
+      "grad_norm": 0.23757430911064148,
+      "learning_rate": 4.56106010283918e-05,
+      "loss": 0.3685,
+      "step": 2421
+    },
+    {
+      "epoch": 1.8231087692886714,
+      "grad_norm": 0.23834286630153656,
+      "learning_rate": 4.560618485727407e-05,
+      "loss": 0.3966,
+      "step": 2422
+    },
+    {
+      "epoch": 1.8238614979299963,
+      "grad_norm": 0.25064605474472046,
+      "learning_rate": 4.5601766706091186e-05,
+      "loss": 0.3237,
+      "step": 2423
+    },
+    {
+      "epoch": 1.824614226571321,
+      "grad_norm": 0.25657644867897034,
+      "learning_rate": 4.5597346575326296e-05,
+      "loss": 0.4278,
+      "step": 2424
+    },
+    {
+      "epoch": 1.8253669552126457,
+      "grad_norm": 0.23903393745422363,
+      "learning_rate": 4.559292446546281e-05,
+      "loss": 0.3667,
+      "step": 2425
+    },
+    {
+      "epoch": 1.8261196838539706,
+      "grad_norm": 0.23517964780330658,
+      "learning_rate": 4.55885003769843e-05,
+      "loss": 0.3649,
+      "step": 2426
+    },
+    {
+      "epoch": 1.8268724124952955,
+      "grad_norm": 0.22778227925300598,
+      "learning_rate": 4.558407431037458e-05,
+      "loss": 0.3272,
+      "step": 2427
+    },
+    {
+      "epoch": 1.8276251411366202,
+      "grad_norm": 0.2909117639064789,
+      "learning_rate": 4.55796462661177e-05,
+      "loss": 0.4194,
+      "step": 2428
+    },
+    {
+      "epoch": 1.828377869777945,
+      "grad_norm": 0.23981842398643494,
+      "learning_rate": 4.5575216244697865e-05,
+      "loss": 0.3595,
+      "step": 2429
+    },
+    {
+      "epoch": 1.8291305984192698,
+      "grad_norm": 0.3225681185722351,
+      "learning_rate": 4.5570784246599565e-05,
+      "loss": 0.3828,
+      "step": 2430
+    },
+    {
+      "epoch": 1.8298833270605948,
+      "grad_norm": 0.23381750285625458,
+      "learning_rate": 4.5566350272307473e-05,
+      "loss": 0.4172,
+      "step": 2431
+    },
+    {
+      "epoch": 1.8306360557019195,
+      "grad_norm": 0.42586663365364075,
+      "learning_rate": 4.556191432230648e-05,
+      "loss": 0.3704,
+      "step": 2432
+    },
+    {
+      "epoch": 1.8313887843432441,
+      "grad_norm": 0.23547814786434174,
+      "learning_rate": 4.5557476397081686e-05,
+      "loss": 0.4142,
+      "step": 2433
+    },
+    {
+      "epoch": 1.832141512984569,
+      "grad_norm": 0.340466171503067,
+      "learning_rate": 4.555303649711842e-05,
+      "loss": 0.3209,
+      "step": 2434
+    },
+    {
+      "epoch": 1.832894241625894,
+      "grad_norm": 0.2328432947397232,
+      "learning_rate": 4.554859462290223e-05,
+      "loss": 0.415,
+      "step": 2435
+    },
+    {
+      "epoch": 1.8336469702672187,
+      "grad_norm": 0.30553358793258667,
+      "learning_rate": 4.5544150774918866e-05,
+      "loss": 0.3524,
+      "step": 2436
+    },
+    {
+      "epoch": 1.8343996989085434,
+      "grad_norm": 0.26823240518569946,
+      "learning_rate": 4.55397049536543e-05,
+      "loss": 0.3338,
+      "step": 2437
+    },
+    {
+      "epoch": 1.8351524275498683,
+      "grad_norm": 0.20805016160011292,
+      "learning_rate": 4.553525715959472e-05,
+      "loss": 0.3867,
+      "step": 2438
+    },
+    {
+      "epoch": 1.8359051561911932,
+      "grad_norm": 0.25177398324012756,
+      "learning_rate": 4.553080739322654e-05,
+      "loss": 0.4349,
+      "step": 2439
+    },
+    {
+      "epoch": 1.836657884832518,
+      "grad_norm": 0.26362836360931396,
+      "learning_rate": 4.552635565503637e-05,
+      "loss": 0.3804,
+      "step": 2440
+    },
+    {
+      "epoch": 1.8374106134738426,
+      "grad_norm": 0.25089165568351746,
+      "learning_rate": 4.5521901945511056e-05,
+      "loss": 0.4095,
+      "step": 2441
+    },
+    {
+      "epoch": 1.8381633421151675,
+      "grad_norm": 0.28410258889198303,
+      "learning_rate": 4.551744626513763e-05,
+      "loss": 0.334,
+      "step": 2442
+    },
+    {
+      "epoch": 1.8389160707564924,
+      "grad_norm": 0.29136165976524353,
+      "learning_rate": 4.551298861440337e-05,
+      "loss": 0.4106,
+      "step": 2443
+    },
+    {
+      "epoch": 1.8396687993978171,
+      "grad_norm": 0.25480666756629944,
+      "learning_rate": 4.5508528993795754e-05,
+      "loss": 0.3848,
+      "step": 2444
+    },
+    {
+      "epoch": 1.8404215280391418,
+      "grad_norm": 0.2783726155757904,
+      "learning_rate": 4.550406740380249e-05,
+      "loss": 0.3893,
+      "step": 2445
+    },
+    {
+      "epoch": 1.8411742566804667,
+      "grad_norm": 0.24906957149505615,
+      "learning_rate": 4.5499603844911484e-05,
+      "loss": 0.3179,
+      "step": 2446
+    },
+    {
+      "epoch": 1.8419269853217914,
+      "grad_norm": 0.34322652220726013,
+      "learning_rate": 4.5495138317610855e-05,
+      "loss": 0.3797,
+      "step": 2447
+    },
+    {
+      "epoch": 1.8426797139631161,
+      "grad_norm": 0.25854215025901794,
+      "learning_rate": 4.5490670822388957e-05,
+      "loss": 0.3005,
+      "step": 2448
+    },
+    {
+      "epoch": 1.843432442604441,
+      "grad_norm": 0.27320393919944763,
+      "learning_rate": 4.548620135973435e-05,
+      "loss": 0.3404,
+      "step": 2449
+    },
+    {
+      "epoch": 1.844185171245766,
+      "grad_norm": 0.266915887594223,
+      "learning_rate": 4.5481729930135805e-05,
+      "loss": 0.3221,
+      "step": 2450
+    },
+    {
+      "epoch": 1.8449378998870907,
+      "grad_norm": 0.31438151001930237,
+      "learning_rate": 4.547725653408231e-05,
+      "loss": 0.3684,
+      "step": 2451
+    },
+    {
+      "epoch": 1.8456906285284154,
+      "grad_norm": 0.2710534334182739,
+      "learning_rate": 4.547278117206307e-05,
+      "loss": 0.4096,
+      "step": 2452
+    },
+    {
+      "epoch": 1.8464433571697403,
+      "grad_norm": 0.2261495590209961,
+      "learning_rate": 4.546830384456751e-05,
+      "loss": 0.3448,
+      "step": 2453
+    },
+    {
+      "epoch": 1.8471960858110652,
+      "grad_norm": 0.24649173021316528,
+      "learning_rate": 4.5463824552085235e-05,
+      "loss": 0.4812,
+      "step": 2454
+    },
+    {
+      "epoch": 1.84794881445239,
+      "grad_norm": 0.28600916266441345,
+      "learning_rate": 4.5459343295106134e-05,
+      "loss": 0.3515,
+      "step": 2455
+    },
+    {
+      "epoch": 1.8487015430937146,
+      "grad_norm": 0.26085376739501953,
+      "learning_rate": 4.545486007412024e-05,
+      "loss": 0.3929,
+      "step": 2456
+    },
+    {
+      "epoch": 1.8494542717350395,
+      "grad_norm": 0.2856564521789551,
+      "learning_rate": 4.545037488961785e-05,
+      "loss": 0.3697,
+      "step": 2457
+    },
+    {
+      "epoch": 1.8502070003763644,
+      "grad_norm": 0.23737157881259918,
+      "learning_rate": 4.544588774208945e-05,
+      "loss": 0.3108,
+      "step": 2458
+    },
+    {
+      "epoch": 1.8509597290176891,
+      "grad_norm": 0.22564566135406494,
+      "learning_rate": 4.544139863202574e-05,
+      "loss": 0.3939,
+      "step": 2459
+    },
+    {
+      "epoch": 1.8517124576590138,
+      "grad_norm": 0.20749300718307495,
+      "learning_rate": 4.543690755991766e-05,
+      "loss": 0.3096,
+      "step": 2460
+    },
+    {
+      "epoch": 1.8524651863003387,
+      "grad_norm": 0.2988424599170685,
+      "learning_rate": 4.543241452625633e-05,
+      "loss": 0.3678,
+      "step": 2461
+    },
+    {
+      "epoch": 1.8532179149416637,
+      "grad_norm": 0.2508583962917328,
+      "learning_rate": 4.5427919531533115e-05,
+      "loss": 0.3738,
+      "step": 2462
+    },
+    {
+      "epoch": 1.8539706435829884,
+      "grad_norm": 0.348410427570343,
+      "learning_rate": 4.5423422576239566e-05,
+      "loss": 0.364,
+      "step": 2463
+    },
+    {
+      "epoch": 1.854723372224313,
+      "grad_norm": 0.2508332133293152,
+      "learning_rate": 4.541892366086748e-05,
+      "loss": 0.3938,
+      "step": 2464
+    },
+    {
+      "epoch": 1.855476100865638,
+      "grad_norm": 0.3348049819469452,
+      "learning_rate": 4.541442278590884e-05,
+      "loss": 0.3725,
+      "step": 2465
+    },
+    {
+      "epoch": 1.8562288295069629,
+      "grad_norm": 0.32221195101737976,
+      "learning_rate": 4.5409919951855855e-05,
+      "loss": 0.4408,
+      "step": 2466
+    },
+    {
+      "epoch": 1.8569815581482876,
+      "grad_norm": 0.31678327918052673,
+      "learning_rate": 4.540541515920095e-05,
+      "loss": 0.396,
+      "step": 2467
+    },
+    {
+      "epoch": 1.8577342867896123,
+      "grad_norm": 0.25160273909568787,
+      "learning_rate": 4.540090840843677e-05,
+      "loss": 0.3715,
+      "step": 2468
+    },
+    {
+      "epoch": 1.8584870154309372,
+      "grad_norm": 0.2515031397342682,
+      "learning_rate": 4.539639970005616e-05,
+      "loss": 0.3283,
+      "step": 2469
+    },
+    {
+      "epoch": 1.8592397440722621,
+      "grad_norm": 0.27471932768821716,
+      "learning_rate": 4.539188903455218e-05,
+      "loss": 0.4035,
+      "step": 2470
+    },
+    {
+      "epoch": 1.8599924727135866,
+      "grad_norm": 0.2433118224143982,
+      "learning_rate": 4.538737641241812e-05,
+      "loss": 0.3964,
+      "step": 2471
+    },
+    {
+      "epoch": 1.8607452013549115,
+      "grad_norm": 0.35174837708473206,
+      "learning_rate": 4.5382861834147464e-05,
+      "loss": 0.4055,
+      "step": 2472
+    },
+    {
+      "epoch": 1.8614979299962364,
+      "grad_norm": 0.26415443420410156,
+      "learning_rate": 4.537834530023393e-05,
+      "loss": 0.3471,
+      "step": 2473
+    },
+    {
+      "epoch": 1.8622506586375611,
+      "grad_norm": 0.314564973115921,
+      "learning_rate": 4.537382681117143e-05,
+      "loss": 0.4133,
+      "step": 2474
+    },
+    {
+      "epoch": 1.8630033872788858,
+      "grad_norm": 0.2666969299316406,
+      "learning_rate": 4.53693063674541e-05,
+      "loss": 0.4257,
+      "step": 2475
+    },
+    {
+      "epoch": 1.8637561159202107,
+      "grad_norm": 0.2730102837085724,
+      "learning_rate": 4.53647839695763e-05,
+      "loss": 0.4529,
+      "step": 2476
+    },
+    {
+      "epoch": 1.8645088445615356,
+      "grad_norm": 0.2755206823348999,
+      "learning_rate": 4.536025961803258e-05,
+      "loss": 0.4638,
+      "step": 2477
+    },
+    {
+      "epoch": 1.8652615732028603,
+      "grad_norm": 0.2775031328201294,
+      "learning_rate": 4.535573331331772e-05,
+      "loss": 0.3892,
+      "step": 2478
+    },
+    {
+      "epoch": 1.866014301844185,
+      "grad_norm": 0.33663591742515564,
+      "learning_rate": 4.535120505592671e-05,
+      "loss": 0.3549,
+      "step": 2479
+    },
+    {
+      "epoch": 1.86676703048551,
+      "grad_norm": 0.31181421875953674,
+      "learning_rate": 4.534667484635476e-05,
+      "loss": 0.3505,
+      "step": 2480
+    },
+    {
+      "epoch": 1.8675197591268349,
+      "grad_norm": 0.3043401539325714,
+      "learning_rate": 4.534214268509727e-05,
+      "loss": 0.3954,
+      "step": 2481
+    },
+    {
+      "epoch": 1.8682724877681596,
+      "grad_norm": 0.30800607800483704,
+      "learning_rate": 4.533760857264988e-05,
+      "loss": 0.3863,
+      "step": 2482
+    },
+    {
+      "epoch": 1.8690252164094843,
+      "grad_norm": 0.30887898802757263,
+      "learning_rate": 4.533307250950845e-05,
+      "loss": 0.4546,
+      "step": 2483
+    },
+    {
+      "epoch": 1.8697779450508092,
+      "grad_norm": 0.2780291438102722,
+      "learning_rate": 4.5328534496169015e-05,
+      "loss": 0.4242,
+      "step": 2484
+    },
+    {
+      "epoch": 1.870530673692134,
+      "grad_norm": 0.25079989433288574,
+      "learning_rate": 4.532399453312785e-05,
+      "loss": 0.4373,
+      "step": 2485
+    },
+    {
+      "epoch": 1.8712834023334588,
+      "grad_norm": 0.26448047161102295,
+      "learning_rate": 4.531945262088144e-05,
+      "loss": 0.3995,
+      "step": 2486
+    },
+    {
+      "epoch": 1.8720361309747835,
+      "grad_norm": 0.2565653920173645,
+      "learning_rate": 4.531490875992648e-05,
+      "loss": 0.3532,
+      "step": 2487
+    },
+    {
+      "epoch": 1.8727888596161084,
+      "grad_norm": 0.28009411692619324,
+      "learning_rate": 4.5310362950759886e-05,
+      "loss": 0.3756,
+      "step": 2488
+    },
+    {
+      "epoch": 1.8735415882574333,
+      "grad_norm": 0.31238654255867004,
+      "learning_rate": 4.5305815193878784e-05,
+      "loss": 0.369,
+      "step": 2489
+    },
+    {
+      "epoch": 1.874294316898758,
+      "grad_norm": 0.297628790140152,
+      "learning_rate": 4.5301265489780495e-05,
+      "loss": 0.3848,
+      "step": 2490
+    },
+    {
+      "epoch": 1.8750470455400827,
+      "grad_norm": 0.30038782954216003,
+      "learning_rate": 4.529671383896258e-05,
+      "loss": 0.4174,
+      "step": 2491
+    },
+    {
+      "epoch": 1.8757997741814076,
+      "grad_norm": 0.32253333926200867,
+      "learning_rate": 4.529216024192279e-05,
+      "loss": 0.3283,
+      "step": 2492
+    },
+    {
+      "epoch": 1.8765525028227326,
+      "grad_norm": 0.3418746292591095,
+      "learning_rate": 4.528760469915912e-05,
+      "loss": 0.3387,
+      "step": 2493
+    },
+    {
+      "epoch": 1.8773052314640573,
+      "grad_norm": 0.33005082607269287,
+      "learning_rate": 4.528304721116973e-05,
+      "loss": 0.4047,
+      "step": 2494
+    },
+    {
+      "epoch": 1.878057960105382,
+      "grad_norm": 0.2953543961048126,
+      "learning_rate": 4.527848777845304e-05,
+      "loss": 0.3582,
+      "step": 2495
+    },
+    {
+      "epoch": 1.8788106887467069,
+      "grad_norm": 0.3035709261894226,
+      "learning_rate": 4.527392640150765e-05,
+      "loss": 0.322,
+      "step": 2496
+    },
+    {
+      "epoch": 1.8795634173880316,
+      "grad_norm": 0.30239391326904297,
+      "learning_rate": 4.52693630808324e-05,
+      "loss": 0.3534,
+      "step": 2497
+    },
+    {
+      "epoch": 1.8803161460293563,
+      "grad_norm": 0.24796463549137115,
+      "learning_rate": 4.52647978169263e-05,
+      "loss": 0.4597,
+      "step": 2498
+    },
+    {
+      "epoch": 1.8810688746706812,
+      "grad_norm": 0.25708073377609253,
+      "learning_rate": 4.526023061028864e-05,
+      "loss": 0.3776,
+      "step": 2499
+    },
+    {
+      "epoch": 1.881821603312006,
+      "grad_norm": 0.3223981261253357,
+      "learning_rate": 4.5255661461418854e-05,
+      "loss": 0.2995,
+      "step": 2500
+    },
+    {
+      "epoch": 1.8825743319533308,
+      "grad_norm": 0.2552452087402344,
+      "learning_rate": 4.5251090370816635e-05,
+      "loss": 0.3006,
+      "step": 2501
+    },
+    {
+      "epoch": 1.8833270605946555,
+      "grad_norm": 0.2619754672050476,
+      "learning_rate": 4.5246517338981855e-05,
+      "loss": 0.3545,
+      "step": 2502
+    },
+    {
+      "epoch": 1.8840797892359804,
+      "grad_norm": 0.22207389771938324,
+      "learning_rate": 4.524194236641462e-05,
+      "loss": 0.3427,
+      "step": 2503
+    },
+    {
+      "epoch": 1.8848325178773053,
+      "grad_norm": 0.2375354915857315,
+      "learning_rate": 4.5237365453615246e-05,
+      "loss": 0.3003,
+      "step": 2504
+    },
+    {
+      "epoch": 1.88558524651863,
+      "grad_norm": 0.24949049949645996,
+      "learning_rate": 4.5232786601084246e-05,
+      "loss": 0.3965,
+      "step": 2505
+    },
+    {
+      "epoch": 1.8863379751599547,
+      "grad_norm": 0.2445673644542694,
+      "learning_rate": 4.522820580932237e-05,
+      "loss": 0.2974,
+      "step": 2506
+    },
+    {
+      "epoch": 1.8870907038012796,
+      "grad_norm": 0.27825281023979187,
+      "learning_rate": 4.522362307883055e-05,
+      "loss": 0.3847,
+      "step": 2507
+    },
+    {
+      "epoch": 1.8878434324426046,
+      "grad_norm": 0.22370657324790955,
+      "learning_rate": 4.521903841010997e-05,
+      "loss": 0.4322,
+      "step": 2508
+    },
+    {
+      "epoch": 1.8885961610839292,
+      "grad_norm": 0.2620297372341156,
+      "learning_rate": 4.5214451803661985e-05,
+      "loss": 0.3967,
+      "step": 2509
+    },
+    {
+      "epoch": 1.889348889725254,
+      "grad_norm": 0.2314402312040329,
+      "learning_rate": 4.5209863259988185e-05,
+      "loss": 0.3859,
+      "step": 2510
+    },
+    {
+      "epoch": 1.8901016183665789,
+      "grad_norm": 0.24672776460647583,
+      "learning_rate": 4.520527277959037e-05,
+      "loss": 0.2938,
+      "step": 2511
+    },
+    {
+      "epoch": 1.8908543470079038,
+      "grad_norm": 0.215503990650177,
+      "learning_rate": 4.520068036297053e-05,
+      "loss": 0.3881,
+      "step": 2512
+    },
+    {
+      "epoch": 1.8916070756492285,
+      "grad_norm": 0.2901112735271454,
+      "learning_rate": 4.519608601063091e-05,
+      "loss": 0.3631,
+      "step": 2513
+    },
+    {
+      "epoch": 1.8923598042905532,
+      "grad_norm": 0.31295013427734375,
+      "learning_rate": 4.519148972307392e-05,
+      "loss": 0.3889,
+      "step": 2514
+    },
+    {
+      "epoch": 1.893112532931878,
+      "grad_norm": 0.2997682988643646,
+      "learning_rate": 4.518689150080223e-05,
+      "loss": 0.3334,
+      "step": 2515
+    },
+    {
+      "epoch": 1.893865261573203,
+      "grad_norm": 0.3028944134712219,
+      "learning_rate": 4.518229134431866e-05,
+      "loss": 0.3379,
+      "step": 2516
+    },
+    {
+      "epoch": 1.8946179902145277,
+      "grad_norm": 0.3535299003124237,
+      "learning_rate": 4.5177689254126306e-05,
+      "loss": 0.3025,
+      "step": 2517
+    },
+    {
+      "epoch": 1.8953707188558524,
+      "grad_norm": 0.2687862813472748,
+      "learning_rate": 4.5173085230728416e-05,
+      "loss": 0.4456,
+      "step": 2518
+    },
+    {
+      "epoch": 1.8961234474971773,
+      "grad_norm": 0.26870307326316833,
+      "learning_rate": 4.5168479274628514e-05,
+      "loss": 0.3544,
+      "step": 2519
+    },
+    {
+      "epoch": 1.8968761761385022,
+      "grad_norm": 0.26136213541030884,
+      "learning_rate": 4.5163871386330274e-05,
+      "loss": 0.3968,
+      "step": 2520
+    },
+    {
+      "epoch": 1.8976289047798267,
+      "grad_norm": 0.23335504531860352,
+      "learning_rate": 4.5159261566337626e-05,
+      "loss": 0.2731,
+      "step": 2521
+    },
+    {
+      "epoch": 1.8983816334211516,
+      "grad_norm": 0.26961278915405273,
+      "learning_rate": 4.515464981515466e-05,
+      "loss": 0.3385,
+      "step": 2522
+    },
+    {
+      "epoch": 1.8991343620624765,
+      "grad_norm": 0.2817212641239166,
+      "learning_rate": 4.5150036133285755e-05,
+      "loss": 0.3276,
+      "step": 2523
+    },
+    {
+      "epoch": 1.8998870907038012,
+      "grad_norm": 0.24963103234767914,
+      "learning_rate": 4.514542052123543e-05,
+      "loss": 0.4003,
+      "step": 2524
+    },
+    {
+      "epoch": 1.900639819345126,
+      "grad_norm": 0.2332722246646881,
+      "learning_rate": 4.5140802979508445e-05,
+      "loss": 0.3309,
+      "step": 2525
+    },
+    {
+      "epoch": 1.9013925479864509,
+      "grad_norm": 0.26240837574005127,
+      "learning_rate": 4.5136183508609766e-05,
+      "loss": 0.2628,
+      "step": 2526
+    },
+    {
+      "epoch": 1.9021452766277758,
+      "grad_norm": 0.27997204661369324,
+      "learning_rate": 4.513156210904458e-05,
+      "loss": 0.3831,
+      "step": 2527
+    },
+    {
+      "epoch": 1.9028980052691005,
+      "grad_norm": 0.23650097846984863,
+      "learning_rate": 4.512693878131827e-05,
+      "loss": 0.3308,
+      "step": 2528
+    },
+    {
+      "epoch": 1.9036507339104252,
+      "grad_norm": 0.24488934874534607,
+      "learning_rate": 4.5122313525936434e-05,
+      "loss": 0.4243,
+      "step": 2529
+    },
+    {
+      "epoch": 1.90440346255175,
+      "grad_norm": 0.2271670401096344,
+      "learning_rate": 4.511768634340489e-05,
+      "loss": 0.518,
+      "step": 2530
+    },
+    {
+      "epoch": 1.905156191193075,
+      "grad_norm": 0.23009781539440155,
+      "learning_rate": 4.5113057234229644e-05,
+      "loss": 0.3709,
+      "step": 2531
+    },
+    {
+      "epoch": 1.9059089198343997,
+      "grad_norm": 0.26348796486854553,
+      "learning_rate": 4.510842619891695e-05,
+      "loss": 0.4031,
+      "step": 2532
+    },
+    {
+      "epoch": 1.9066616484757244,
+      "grad_norm": 0.21193136274814606,
+      "learning_rate": 4.510379323797323e-05,
+      "loss": 0.3537,
+      "step": 2533
+    },
+    {
+      "epoch": 1.9074143771170493,
+      "grad_norm": 0.2594548761844635,
+      "learning_rate": 4.509915835190516e-05,
+      "loss": 0.4046,
+      "step": 2534
+    },
+    {
+      "epoch": 1.9081671057583742,
+      "grad_norm": 0.23847204446792603,
+      "learning_rate": 4.5094521541219587e-05,
+      "loss": 0.2814,
+      "step": 2535
+    },
+    {
+      "epoch": 1.908919834399699,
+      "grad_norm": 0.3382256329059601,
+      "learning_rate": 4.508988280642359e-05,
+      "loss": 0.3839,
+      "step": 2536
+    },
+    {
+      "epoch": 1.9096725630410236,
+      "grad_norm": 0.3339466452598572,
+      "learning_rate": 4.508524214802446e-05,
+      "loss": 0.4116,
+      "step": 2537
+    },
+    {
+      "epoch": 1.9104252916823485,
+      "grad_norm": 0.2127711921930313,
+      "learning_rate": 4.508059956652969e-05,
+      "loss": 0.3955,
+      "step": 2538
+    },
+    {
+      "epoch": 1.9111780203236735,
+      "grad_norm": 0.26324716210365295,
+      "learning_rate": 4.507595506244698e-05,
+      "loss": 0.4942,
+      "step": 2539
+    },
+    {
+      "epoch": 1.9119307489649982,
+      "grad_norm": 0.2769601345062256,
+      "learning_rate": 4.5071308636284243e-05,
+      "loss": 0.367,
+      "step": 2540
+    },
+    {
+      "epoch": 1.9126834776063228,
+      "grad_norm": 0.24079810082912445,
+      "learning_rate": 4.506666028854962e-05,
+      "loss": 0.3899,
+      "step": 2541
+    },
+    {
+      "epoch": 1.9134362062476478,
+      "grad_norm": 0.2358226329088211,
+      "learning_rate": 4.506201001975143e-05,
+      "loss": 0.2529,
+      "step": 2542
+    },
+    {
+      "epoch": 1.9141889348889727,
+      "grad_norm": 0.24965737760066986,
+      "learning_rate": 4.505735783039824e-05,
+      "loss": 0.4475,
+      "step": 2543
+    },
+    {
+      "epoch": 1.9149416635302974,
+      "grad_norm": 0.22146843373775482,
+      "learning_rate": 4.505270372099879e-05,
+      "loss": 0.3762,
+      "step": 2544
+    },
+    {
+      "epoch": 1.915694392171622,
+      "grad_norm": 0.24325509369373322,
+      "learning_rate": 4.504804769206205e-05,
+      "loss": 0.2774,
+      "step": 2545
+    },
+    {
+      "epoch": 1.916447120812947,
+      "grad_norm": 0.23638299107551575,
+      "learning_rate": 4.50433897440972e-05,
+      "loss": 0.3056,
+      "step": 2546
+    },
+    {
+      "epoch": 1.9171998494542717,
+      "grad_norm": 0.27911555767059326,
+      "learning_rate": 4.5038729877613614e-05,
+      "loss": 0.3736,
+      "step": 2547
+    },
+    {
+      "epoch": 1.9179525780955964,
+      "grad_norm": 0.21964851021766663,
+      "learning_rate": 4.503406809312091e-05,
+      "loss": 0.355,
+      "step": 2548
+    },
+    {
+      "epoch": 1.9187053067369213,
+      "grad_norm": 0.26507332921028137,
+      "learning_rate": 4.502940439112887e-05,
+      "loss": 0.2782,
+      "step": 2549
+    },
+    {
+      "epoch": 1.9194580353782462,
+      "grad_norm": 0.22462914884090424,
+      "learning_rate": 4.502473877214754e-05,
+      "loss": 0.3071,
+      "step": 2550
+    },
+    {
+      "epoch": 1.920210764019571,
+      "grad_norm": 0.41086381673812866,
+      "learning_rate": 4.502007123668711e-05,
+      "loss": 0.3288,
+      "step": 2551
+    },
+    {
+      "epoch": 1.9209634926608956,
+      "grad_norm": 0.29616841673851013,
+      "learning_rate": 4.5015401785258034e-05,
+      "loss": 0.4081,
+      "step": 2552
+    },
+    {
+      "epoch": 1.9217162213022205,
+      "grad_norm": 0.2892850935459137,
+      "learning_rate": 4.5010730418370946e-05,
+      "loss": 0.5056,
+      "step": 2553
+    },
+    {
+      "epoch": 1.9224689499435454,
+      "grad_norm": 0.27476003766059875,
+      "learning_rate": 4.5006057136536715e-05,
+      "loss": 0.3846,
+      "step": 2554
+    },
+    {
+      "epoch": 1.9232216785848701,
+      "grad_norm": 0.3733964264392853,
+      "learning_rate": 4.500138194026639e-05,
+      "loss": 0.393,
+      "step": 2555
+    },
+    {
+      "epoch": 1.9239744072261948,
+      "grad_norm": 0.40330639481544495,
+      "learning_rate": 4.4996704830071255e-05,
+      "loss": 0.3073,
+      "step": 2556
+    },
+    {
+      "epoch": 1.9247271358675198,
+      "grad_norm": 0.2905791699886322,
+      "learning_rate": 4.499202580646279e-05,
+      "loss": 0.4067,
+      "step": 2557
+    },
+    {
+      "epoch": 1.9254798645088447,
+      "grad_norm": 0.427398681640625,
+      "learning_rate": 4.498734486995267e-05,
+      "loss": 0.4171,
+      "step": 2558
+    },
+    {
+      "epoch": 1.9262325931501694,
+      "grad_norm": 0.2759098708629608,
+      "learning_rate": 4.498266202105281e-05,
+      "loss": 0.3836,
+      "step": 2559
+    },
+    {
+      "epoch": 1.926985321791494,
+      "grad_norm": 0.35807445645332336,
+      "learning_rate": 4.4977977260275314e-05,
+      "loss": 0.4088,
+      "step": 2560
+    },
+    {
+      "epoch": 1.927738050432819,
+      "grad_norm": 0.28511810302734375,
+      "learning_rate": 4.4973290588132514e-05,
+      "loss": 0.4161,
+      "step": 2561
+    },
+    {
+      "epoch": 1.928490779074144,
+      "grad_norm": 0.30248719453811646,
+      "learning_rate": 4.4968602005136917e-05,
+      "loss": 0.3909,
+      "step": 2562
+    },
+    {
+      "epoch": 1.9292435077154686,
+      "grad_norm": 0.3491272032260895,
+      "learning_rate": 4.496391151180126e-05,
+      "loss": 0.3395,
+      "step": 2563
+    },
+    {
+      "epoch": 1.9299962363567933,
+      "grad_norm": 0.3228083550930023,
+      "learning_rate": 4.4959219108638515e-05,
+      "loss": 0.4633,
+      "step": 2564
+    },
+    {
+      "epoch": 1.9307489649981182,
+      "grad_norm": 0.3028690218925476,
+      "learning_rate": 4.495452479616181e-05,
+      "loss": 0.4073,
+      "step": 2565
+    },
+    {
+      "epoch": 1.9315016936394431,
+      "grad_norm": 0.32129955291748047,
+      "learning_rate": 4.494982857488451e-05,
+      "loss": 0.2462,
+      "step": 2566
+    },
+    {
+      "epoch": 1.9322544222807678,
+      "grad_norm": 0.22042685747146606,
+      "learning_rate": 4.4945130445320206e-05,
+      "loss": 0.3345,
+      "step": 2567
+    },
+    {
+      "epoch": 1.9330071509220925,
+      "grad_norm": 0.2787827253341675,
+      "learning_rate": 4.494043040798265e-05,
+      "loss": 0.3553,
+      "step": 2568
+    },
+    {
+      "epoch": 1.9337598795634174,
+      "grad_norm": 0.25657424330711365,
+      "learning_rate": 4.4935728463385854e-05,
+      "loss": 0.3269,
+      "step": 2569
+    },
+    {
+      "epoch": 1.9345126082047424,
+      "grad_norm": 0.3171713352203369,
+      "learning_rate": 4.4931024612044005e-05,
+      "loss": 0.3734,
+      "step": 2570
+    },
+    {
+      "epoch": 1.9352653368460668,
+      "grad_norm": 0.23930498957633972,
+      "learning_rate": 4.492631885447151e-05,
+      "loss": 0.3744,
+      "step": 2571
+    },
+    {
+      "epoch": 1.9360180654873917,
+      "grad_norm": 0.2976834177970886,
+      "learning_rate": 4.492161119118299e-05,
+      "loss": 0.3761,
+      "step": 2572
+    },
+    {
+      "epoch": 1.9367707941287167,
+      "grad_norm": 0.24410870671272278,
+      "learning_rate": 4.491690162269326e-05,
+      "loss": 0.4371,
+      "step": 2573
+    },
+    {
+      "epoch": 1.9375235227700414,
+      "grad_norm": 0.2382754534482956,
+      "learning_rate": 4.4912190149517354e-05,
+      "loss": 0.3501,
+      "step": 2574
+    },
+    {
+      "epoch": 1.938276251411366,
+      "grad_norm": 0.3457755744457245,
+      "learning_rate": 4.490747677217051e-05,
+      "loss": 0.4677,
+      "step": 2575
+    },
+    {
+      "epoch": 1.939028980052691,
+      "grad_norm": 0.24767127633094788,
+      "learning_rate": 4.490276149116819e-05,
+      "loss": 0.3386,
+      "step": 2576
+    },
+    {
+      "epoch": 1.939781708694016,
+      "grad_norm": 0.3610161244869232,
+      "learning_rate": 4.489804430702603e-05,
+      "loss": 0.3313,
+      "step": 2577
+    },
+    {
+      "epoch": 1.9405344373353406,
+      "grad_norm": 0.25502926111221313,
+      "learning_rate": 4.48933252202599e-05,
+      "loss": 0.3269,
+      "step": 2578
+    },
+    {
+      "epoch": 1.9412871659766653,
+      "grad_norm": 0.31534433364868164,
+      "learning_rate": 4.4888604231385874e-05,
+      "loss": 0.3568,
+      "step": 2579
+    },
+    {
+      "epoch": 1.9420398946179902,
+      "grad_norm": 0.3743187189102173,
+      "learning_rate": 4.488388134092023e-05,
+      "loss": 0.375,
+      "step": 2580
+    },
+    {
+      "epoch": 1.9427926232593151,
+      "grad_norm": 0.2839333117008209,
+      "learning_rate": 4.4879156549379466e-05,
+      "loss": 0.3206,
+      "step": 2581
+    },
+    {
+      "epoch": 1.9435453519006398,
+      "grad_norm": 0.2772039473056793,
+      "learning_rate": 4.487442985728026e-05,
+      "loss": 0.3088,
+      "step": 2582
+    },
+    {
+      "epoch": 1.9442980805419645,
+      "grad_norm": 0.28743916749954224,
+      "learning_rate": 4.486970126513954e-05,
+      "loss": 0.3597,
+      "step": 2583
+    },
+    {
+      "epoch": 1.9450508091832894,
+      "grad_norm": 0.26391181349754333,
+      "learning_rate": 4.486497077347441e-05,
+      "loss": 0.3402,
+      "step": 2584
+    },
+    {
+      "epoch": 1.9458035378246143,
+      "grad_norm": 0.35856932401657104,
+      "learning_rate": 4.4860238382802167e-05,
+      "loss": 0.3885,
+      "step": 2585
+    },
+    {
+      "epoch": 1.946556266465939,
+      "grad_norm": 0.3089780807495117,
+      "learning_rate": 4.485550409364036e-05,
+      "loss": 0.2896,
+      "step": 2586
+    },
+    {
+      "epoch": 1.9473089951072637,
+      "grad_norm": 0.2882886826992035,
+      "learning_rate": 4.485076790650672e-05,
+      "loss": 0.4089,
+      "step": 2587
+    },
+    {
+      "epoch": 1.9480617237485887,
+      "grad_norm": 0.2784572243690491,
+      "learning_rate": 4.4846029821919196e-05,
+      "loss": 0.3489,
+      "step": 2588
+    },
+    {
+      "epoch": 1.9488144523899136,
+      "grad_norm": 0.3286993205547333,
+      "learning_rate": 4.484128984039592e-05,
+      "loss": 0.4025,
+      "step": 2589
+    },
+    {
+      "epoch": 1.9495671810312383,
+      "grad_norm": 0.25698500871658325,
+      "learning_rate": 4.483654796245526e-05,
+      "loss": 0.2753,
+      "step": 2590
+    },
+    {
+      "epoch": 1.950319909672563,
+      "grad_norm": 0.33044111728668213,
+      "learning_rate": 4.483180418861578e-05,
+      "loss": 0.3888,
+      "step": 2591
+    },
+    {
+      "epoch": 1.9510726383138879,
+      "grad_norm": 0.2773405611515045,
+      "learning_rate": 4.4827058519396254e-05,
+      "loss": 0.4254,
+      "step": 2592
+    },
+    {
+      "epoch": 1.9518253669552128,
+      "grad_norm": 0.380172997713089,
+      "learning_rate": 4.482231095531566e-05,
+      "loss": 0.4513,
+      "step": 2593
+    },
+    {
+      "epoch": 1.9525780955965375,
+      "grad_norm": 0.2914517819881439,
+      "learning_rate": 4.481756149689318e-05,
+      "loss": 0.3627,
+      "step": 2594
+    },
+    {
+      "epoch": 1.9533308242378622,
+      "grad_norm": 0.2923394739627838,
+      "learning_rate": 4.481281014464822e-05,
+      "loss": 0.3735,
+      "step": 2595
+    },
+    {
+      "epoch": 1.9540835528791871,
+      "grad_norm": 0.28755083680152893,
+      "learning_rate": 4.480805689910036e-05,
+      "loss": 0.429,
+      "step": 2596
+    },
+    {
+      "epoch": 1.9548362815205118,
+      "grad_norm": 0.23833055794239044,
+      "learning_rate": 4.4803301760769426e-05,
+      "loss": 0.3878,
+      "step": 2597
+    },
+    {
+      "epoch": 1.9555890101618365,
+      "grad_norm": 0.3247975707054138,
+      "learning_rate": 4.479854473017544e-05,
+      "loss": 0.4327,
+      "step": 2598
+    },
+    {
+      "epoch": 1.9563417388031614,
+      "grad_norm": 0.30785927176475525,
+      "learning_rate": 4.479378580783858e-05,
+      "loss": 0.4348,
+      "step": 2599
+    },
+    {
+      "epoch": 1.9570944674444863,
+      "grad_norm": 0.3064795434474945,
+      "learning_rate": 4.478902499427933e-05,
+      "loss": 0.3997,
+      "step": 2600
+    },
+    {
+      "epoch": 1.9570944674444863,
+      "eval_loss": 0.35215482115745544,
+      "eval_runtime": 452.7147,
+      "eval_samples_per_second": 21.265,
+      "eval_steps_per_second": 0.665,
+      "step": 2600
+    },
+    {
+      "epoch": 1.957847196085811,
+      "grad_norm": 0.2663865387439728,
+      "learning_rate": 4.4784262290018294e-05,
+      "loss": 0.3055,
+      "step": 2601
+    },
+    {
+      "epoch": 1.9585999247271357,
+      "grad_norm": 0.2797293961048126,
+      "learning_rate": 4.477949769557632e-05,
+      "loss": 0.3234,
+      "step": 2602
+    },
+    {
+      "epoch": 1.9593526533684607,
+      "grad_norm": 0.2949141263961792,
+      "learning_rate": 4.4774731211474455e-05,
+      "loss": 0.4533,
+      "step": 2603
+    },
+    {
+      "epoch": 1.9601053820097856,
+      "grad_norm": 0.337153822183609,
+      "learning_rate": 4.476996283823396e-05,
+      "loss": 0.3446,
+      "step": 2604
+    },
+    {
+      "epoch": 1.9608581106511103,
+      "grad_norm": 0.282624214887619,
+      "learning_rate": 4.476519257637629e-05,
+      "loss": 0.328,
+      "step": 2605
+    },
+    {
+      "epoch": 1.961610839292435,
+      "grad_norm": 0.24863150715827942,
+      "learning_rate": 4.476042042642311e-05,
+      "loss": 0.3427,
+      "step": 2606
+    },
+    {
+      "epoch": 1.9623635679337599,
+      "grad_norm": 0.24666699767112732,
+      "learning_rate": 4.4755646388896314e-05,
+      "loss": 0.3537,
+      "step": 2607
+    },
+    {
+      "epoch": 1.9631162965750848,
+      "grad_norm": 0.23721617460250854,
+      "learning_rate": 4.475087046431798e-05,
+      "loss": 0.3875,
+      "step": 2608
+    },
+    {
+      "epoch": 1.9638690252164095,
+      "grad_norm": 0.3292771875858307,
+      "learning_rate": 4.4746092653210375e-05,
+      "loss": 0.3075,
+      "step": 2609
+    },
+    {
+      "epoch": 1.9646217538577342,
+      "grad_norm": 0.2505330741405487,
+      "learning_rate": 4.474131295609601e-05,
+      "loss": 0.4093,
+      "step": 2610
+    },
+    {
+      "epoch": 1.965374482499059,
+      "grad_norm": 0.26865649223327637,
+      "learning_rate": 4.473653137349758e-05,
+      "loss": 0.3799,
+      "step": 2611
+    },
+    {
+      "epoch": 1.966127211140384,
+      "grad_norm": 0.2352716326713562,
+      "learning_rate": 4.4731747905937996e-05,
+      "loss": 0.4121,
+      "step": 2612
+    },
+    {
+      "epoch": 1.9668799397817087,
+      "grad_norm": 0.26174402236938477,
+      "learning_rate": 4.4726962553940374e-05,
+      "loss": 0.4842,
+      "step": 2613
+    },
+    {
+      "epoch": 1.9676326684230334,
+      "grad_norm": 0.2798030972480774,
+      "learning_rate": 4.472217531802802e-05,
+      "loss": 0.3915,
+      "step": 2614
+    },
+    {
+      "epoch": 1.9683853970643583,
+      "grad_norm": 0.20891834795475006,
+      "learning_rate": 4.471738619872446e-05,
+      "loss": 0.2656,
+      "step": 2615
+    },
+    {
+      "epoch": 1.9691381257056833,
+      "grad_norm": 0.27711498737335205,
+      "learning_rate": 4.471259519655345e-05,
+      "loss": 0.3217,
+      "step": 2616
+    },
+    {
+      "epoch": 1.969890854347008,
+      "grad_norm": 0.24378874897956848,
+      "learning_rate": 4.470780231203889e-05,
+      "loss": 0.4564,
+      "step": 2617
+    },
+    {
+      "epoch": 1.9706435829883326,
+      "grad_norm": 0.23132920265197754,
+      "learning_rate": 4.4703007545704955e-05,
+      "loss": 0.3252,
+      "step": 2618
+    },
+    {
+      "epoch": 1.9713963116296576,
+      "grad_norm": 0.25726118683815,
+      "learning_rate": 4.469821089807597e-05,
+      "loss": 0.3754,
+      "step": 2619
+    },
+    {
+      "epoch": 1.9721490402709823,
+      "grad_norm": 0.21875324845314026,
+      "learning_rate": 4.46934123696765e-05,
+      "loss": 0.4253,
+      "step": 2620
+    },
+    {
+      "epoch": 1.972901768912307,
+      "grad_norm": 0.28209254145622253,
+      "learning_rate": 4.4688611961031304e-05,
+      "loss": 0.3115,
+      "step": 2621
+    },
+    {
+      "epoch": 1.9736544975536319,
+      "grad_norm": 0.23085246980190277,
+      "learning_rate": 4.468380967266535e-05,
+      "loss": 0.3649,
+      "step": 2622
+    },
+    {
+      "epoch": 1.9744072261949568,
+      "grad_norm": 0.2095174640417099,
+      "learning_rate": 4.46790055051038e-05,
+      "loss": 0.2699,
+      "step": 2623
+    },
+    {
+      "epoch": 1.9751599548362815,
+      "grad_norm": 0.25025835633277893,
+      "learning_rate": 4.467419945887204e-05,
+      "loss": 0.3693,
+      "step": 2624
+    },
+    {
+      "epoch": 1.9759126834776062,
+      "grad_norm": 0.23681640625,
+      "learning_rate": 4.466939153449565e-05,
+      "loss": 0.2985,
+      "step": 2625
+    },
+    {
+      "epoch": 1.976665412118931,
+      "grad_norm": 0.26568958163261414,
+      "learning_rate": 4.466458173250041e-05,
+      "loss": 0.4485,
+      "step": 2626
+    },
+    {
+      "epoch": 1.977418140760256,
+      "grad_norm": 0.3501785695552826,
+      "learning_rate": 4.465977005341232e-05,
+      "loss": 0.4585,
+      "step": 2627
+    },
+    {
+      "epoch": 1.9781708694015807,
+      "grad_norm": 0.3034513592720032,
+      "learning_rate": 4.4654956497757575e-05,
+      "loss": 0.375,
+      "step": 2628
+    },
+    {
+      "epoch": 1.9789235980429054,
+      "grad_norm": 0.2552505433559418,
+      "learning_rate": 4.4650141066062587e-05,
+      "loss": 0.369,
+      "step": 2629
+    },
+    {
+      "epoch": 1.9796763266842303,
+      "grad_norm": 0.27327805757522583,
+      "learning_rate": 4.4645323758853945e-05,
+      "loss": 0.4199,
+      "step": 2630
+    },
+    {
+      "epoch": 1.9804290553255552,
+      "grad_norm": 0.24290579557418823,
+      "learning_rate": 4.464050457665849e-05,
+      "loss": 0.4331,
+      "step": 2631
+    },
+    {
+      "epoch": 1.98118178396688,
+      "grad_norm": 0.30544665455818176,
+      "learning_rate": 4.4635683520003213e-05,
+      "loss": 0.3327,
+      "step": 2632
+    },
+    {
+      "epoch": 1.9819345126082046,
+      "grad_norm": 0.299951434135437,
+      "learning_rate": 4.4630860589415354e-05,
+      "loss": 0.4366,
+      "step": 2633
+    },
+    {
+      "epoch": 1.9826872412495296,
+      "grad_norm": 0.2673509120941162,
+      "learning_rate": 4.462603578542233e-05,
+      "loss": 0.3271,
+      "step": 2634
+    },
+    {
+      "epoch": 1.9834399698908545,
+      "grad_norm": 0.26434648036956787,
+      "learning_rate": 4.462120910855179e-05,
+      "loss": 0.3208,
+      "step": 2635
+    },
+    {
+      "epoch": 1.9841926985321792,
+      "grad_norm": 0.24575695395469666,
+      "learning_rate": 4.461638055933156e-05,
+      "loss": 0.3314,
+      "step": 2636
+    },
+    {
+      "epoch": 1.9849454271735039,
+      "grad_norm": 0.23227694630622864,
+      "learning_rate": 4.4611550138289684e-05,
+      "loss": 0.339,
+      "step": 2637
+    },
+    {
+      "epoch": 1.9856981558148288,
+      "grad_norm": 0.2660466730594635,
+      "learning_rate": 4.460671784595442e-05,
+      "loss": 0.422,
+      "step": 2638
+    },
+    {
+      "epoch": 1.9864508844561537,
+      "grad_norm": 0.21392959356307983,
+      "learning_rate": 4.4601883682854195e-05,
+      "loss": 0.3464,
+      "step": 2639
+    },
+    {
+      "epoch": 1.9872036130974784,
+      "grad_norm": 0.28745710849761963,
+      "learning_rate": 4.45970476495177e-05,
+      "loss": 0.3729,
+      "step": 2640
+    },
+    {
+      "epoch": 1.987956341738803,
+      "grad_norm": 0.2405296117067337,
+      "learning_rate": 4.459220974647376e-05,
+      "loss": 0.3629,
+      "step": 2641
+    },
+    {
+      "epoch": 1.988709070380128,
+      "grad_norm": 0.2906774580478668,
+      "learning_rate": 4.4587369974251476e-05,
+      "loss": 0.3216,
+      "step": 2642
+    },
+    {
+      "epoch": 1.989461799021453,
+      "grad_norm": 0.2739931344985962,
+      "learning_rate": 4.45825283333801e-05,
+      "loss": 0.4217,
+      "step": 2643
+    },
+    {
+      "epoch": 1.9902145276627776,
+      "grad_norm": 0.2390405386686325,
+      "learning_rate": 4.45776848243891e-05,
+      "loss": 0.3174,
+      "step": 2644
+    },
+    {
+      "epoch": 1.9909672563041023,
+      "grad_norm": 0.29576507210731506,
+      "learning_rate": 4.457283944780818e-05,
+      "loss": 0.3441,
+      "step": 2645
+    },
+    {
+      "epoch": 1.9917199849454272,
+      "grad_norm": 0.29248854517936707,
+      "learning_rate": 4.456799220416719e-05,
+      "loss": 0.3348,
+      "step": 2646
+    },
+    {
+      "epoch": 1.992472713586752,
+      "grad_norm": 0.23983711004257202,
+      "learning_rate": 4.456314309399625e-05,
+      "loss": 0.3135,
+      "step": 2647
+    },
+    {
+      "epoch": 1.9932254422280766,
+      "grad_norm": 0.26757991313934326,
+      "learning_rate": 4.4558292117825625e-05,
+      "loss": 0.2421,
+      "step": 2648
+    },
+    {
+      "epoch": 1.9939781708694015,
+      "grad_norm": 0.2667861878871918,
+      "learning_rate": 4.455343927618582e-05,
+      "loss": 0.3709,
+      "step": 2649
+    },
+    {
+      "epoch": 1.9947308995107265,
+      "grad_norm": 0.2357775866985321,
+      "learning_rate": 4.454858456960754e-05,
+      "loss": 0.3935,
+      "step": 2650
+    },
+    {
+      "epoch": 1.9954836281520512,
+      "grad_norm": 0.2601138651371002,
+      "learning_rate": 4.4543727998621676e-05,
+      "loss": 0.3421,
+      "step": 2651
+    },
+    {
+      "epoch": 1.9962363567933759,
+      "grad_norm": 0.2516321837902069,
+      "learning_rate": 4.453886956375935e-05,
+      "loss": 0.3727,
+      "step": 2652
+    },
+    {
+      "epoch": 1.9969890854347008,
+      "grad_norm": 0.22695325314998627,
+      "learning_rate": 4.4534009265551865e-05,
+      "loss": 0.4036,
+      "step": 2653
+    },
+    {
+      "epoch": 1.9977418140760257,
+      "grad_norm": 0.2018071413040161,
+      "learning_rate": 4.4529147104530725e-05,
+      "loss": 0.4065,
+      "step": 2654
+    },
+    {
+      "epoch": 1.9984945427173504,
+      "grad_norm": 0.2747001647949219,
+      "learning_rate": 4.452428308122767e-05,
+      "loss": 0.3732,
+      "step": 2655
+    },
+    {
+      "epoch": 1.999247271358675,
+      "grad_norm": 0.24833376705646515,
+      "learning_rate": 4.451941719617462e-05,
+      "loss": 0.3539,
+      "step": 2656
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.23048147559165955,
+      "learning_rate": 4.451454944990369e-05,
+      "loss": 0.325,
+      "step": 2657
+    },
+    {
+      "epoch": 2.000752728641325,
+      "grad_norm": 0.25489628314971924,
+      "learning_rate": 4.4509679842947214e-05,
+      "loss": 0.3906,
+      "step": 2658
+    },
+    {
+      "epoch": 2.0015054572826494,
+      "grad_norm": 0.2599509358406067,
+      "learning_rate": 4.450480837583772e-05,
+      "loss": 0.3964,
+      "step": 2659
+    },
+    {
+      "epoch": 2.0022581859239743,
+      "grad_norm": 0.256653368473053,
+      "learning_rate": 4.449993504910794e-05,
+      "loss": 0.3486,
+      "step": 2660
+    },
+    {
+      "epoch": 2.0030109145652992,
+      "grad_norm": 0.285410076379776,
+      "learning_rate": 4.449505986329083e-05,
+      "loss": 0.3668,
+      "step": 2661
+    },
+    {
+      "epoch": 2.003763643206624,
+      "grad_norm": 0.23197583854198456,
+      "learning_rate": 4.449018281891953e-05,
+      "loss": 0.2314,
+      "step": 2662
+    },
+    {
+      "epoch": 2.0045163718479486,
+      "grad_norm": 0.23844823241233826,
+      "learning_rate": 4.448530391652738e-05,
+      "loss": 0.3405,
+      "step": 2663
+    },
+    {
+      "epoch": 2.0052691004892735,
+      "grad_norm": 0.2818591892719269,
+      "learning_rate": 4.448042315664792e-05,
+      "loss": 0.3238,
+      "step": 2664
+    },
+    {
+      "epoch": 2.0060218291305985,
+      "grad_norm": 0.24568669497966766,
+      "learning_rate": 4.4475540539814925e-05,
+      "loss": 0.3104,
+      "step": 2665
+    },
+    {
+      "epoch": 2.0067745577719234,
+      "grad_norm": 0.3165369927883148,
+      "learning_rate": 4.4470656066562336e-05,
+      "loss": 0.3674,
+      "step": 2666
+    },
+    {
+      "epoch": 2.007527286413248,
+      "grad_norm": 0.2616836726665497,
+      "learning_rate": 4.4465769737424315e-05,
+      "loss": 0.3889,
+      "step": 2667
+    },
+    {
+      "epoch": 2.0082800150545728,
+      "grad_norm": 0.28959769010543823,
+      "learning_rate": 4.446088155293522e-05,
+      "loss": 0.3835,
+      "step": 2668
+    },
+    {
+      "epoch": 2.0090327436958977,
+      "grad_norm": 0.35126787424087524,
+      "learning_rate": 4.4455991513629623e-05,
+      "loss": 0.3005,
+      "step": 2669
+    },
+    {
+      "epoch": 2.0097854723372226,
+      "grad_norm": 0.3503277897834778,
+      "learning_rate": 4.445109962004228e-05,
+      "loss": 0.352,
+      "step": 2670
+    },
+    {
+      "epoch": 2.010538200978547,
+      "grad_norm": 0.3173494338989258,
+      "learning_rate": 4.444620587270818e-05,
+      "loss": 0.3382,
+      "step": 2671
+    },
+    {
+      "epoch": 2.011290929619872,
+      "grad_norm": 0.3129061162471771,
+      "learning_rate": 4.444131027216248e-05,
+      "loss": 0.3964,
+      "step": 2672
+    },
+    {
+      "epoch": 2.012043658261197,
+      "grad_norm": 0.34795743227005005,
+      "learning_rate": 4.443641281894055e-05,
+      "loss": 0.342,
+      "step": 2673
+    },
+    {
+      "epoch": 2.012796386902522,
+      "grad_norm": 0.25920388102531433,
+      "learning_rate": 4.443151351357799e-05,
+      "loss": 0.2904,
+      "step": 2674
+    },
+    {
+      "epoch": 2.0135491155438463,
+      "grad_norm": 0.2639693021774292,
+      "learning_rate": 4.442661235661056e-05,
+      "loss": 0.2538,
+      "step": 2675
+    },
+    {
+      "epoch": 2.014301844185171,
+      "grad_norm": 0.27858656644821167,
+      "learning_rate": 4.442170934857425e-05,
+      "loss": 0.3129,
+      "step": 2676
+    },
+    {
+      "epoch": 2.015054572826496,
+      "grad_norm": 0.258526086807251,
+      "learning_rate": 4.4416804490005255e-05,
+      "loss": 0.2305,
+      "step": 2677
+    },
+    {
+      "epoch": 2.015807301467821,
+      "grad_norm": 0.2777453064918518,
+      "learning_rate": 4.441189778143995e-05,
+      "loss": 0.4158,
+      "step": 2678
+    },
+    {
+      "epoch": 2.0165600301091455,
+      "grad_norm": 0.2349364012479782,
+      "learning_rate": 4.440698922341493e-05,
+      "loss": 0.3441,
+      "step": 2679
+    },
+    {
+      "epoch": 2.0173127587504704,
+      "grad_norm": 0.24430236220359802,
+      "learning_rate": 4.440207881646699e-05,
+      "loss": 0.3401,
+      "step": 2680
+    },
+    {
+      "epoch": 2.0180654873917954,
+      "grad_norm": 0.24589836597442627,
+      "learning_rate": 4.4397166561133127e-05,
+      "loss": 0.2892,
+      "step": 2681
+    },
+    {
+      "epoch": 2.01881821603312,
+      "grad_norm": 0.31507471203804016,
+      "learning_rate": 4.4392252457950525e-05,
+      "loss": 0.3876,
+      "step": 2682
+    },
+    {
+      "epoch": 2.0195709446744448,
+      "grad_norm": 0.2999946177005768,
+      "learning_rate": 4.438733650745659e-05,
+      "loss": 0.3775,
+      "step": 2683
+    },
+    {
+      "epoch": 2.0203236733157697,
+      "grad_norm": 0.32921358942985535,
+      "learning_rate": 4.4382418710188926e-05,
+      "loss": 0.4075,
+      "step": 2684
+    },
+    {
+      "epoch": 2.0210764019570946,
+      "grad_norm": 0.2672775387763977,
+      "learning_rate": 4.437749906668534e-05,
+      "loss": 0.3293,
+      "step": 2685
+    },
+    {
+      "epoch": 2.021829130598419,
+      "grad_norm": 0.30355551838874817,
+      "learning_rate": 4.4372577577483834e-05,
+      "loss": 0.3491,
+      "step": 2686
+    },
+    {
+      "epoch": 2.022581859239744,
+      "grad_norm": 0.24766533076763153,
+      "learning_rate": 4.4367654243122606e-05,
+      "loss": 0.3096,
+      "step": 2687
+    },
+    {
+      "epoch": 2.023334587881069,
+      "grad_norm": 0.31202223896980286,
+      "learning_rate": 4.4362729064140066e-05,
+      "loss": 0.3604,
+      "step": 2688
+    },
+    {
+      "epoch": 2.024087316522394,
+      "grad_norm": 0.24273498356342316,
+      "learning_rate": 4.435780204107483e-05,
+      "loss": 0.3731,
+      "step": 2689
+    },
+    {
+      "epoch": 2.0248400451637183,
+      "grad_norm": 0.2343394160270691,
+      "learning_rate": 4.4352873174465714e-05,
+      "loss": 0.2601,
+      "step": 2690
+    },
+    {
+      "epoch": 2.025592773805043,
+      "grad_norm": 0.2360345870256424,
+      "learning_rate": 4.434794246485172e-05,
+      "loss": 0.2644,
+      "step": 2691
+    },
+    {
+      "epoch": 2.026345502446368,
+      "grad_norm": 0.25208988785743713,
+      "learning_rate": 4.4343009912772065e-05,
+      "loss": 0.4337,
+      "step": 2692
+    },
+    {
+      "epoch": 2.027098231087693,
+      "grad_norm": 0.26861611008644104,
+      "learning_rate": 4.433807551876618e-05,
+      "loss": 0.4258,
+      "step": 2693
+    },
+    {
+      "epoch": 2.0278509597290175,
+      "grad_norm": 0.29717791080474854,
+      "learning_rate": 4.433313928337366e-05,
+      "loss": 0.3235,
+      "step": 2694
+    },
+    {
+      "epoch": 2.0286036883703424,
+      "grad_norm": 0.2401815950870514,
+      "learning_rate": 4.432820120713435e-05,
+      "loss": 0.3668,
+      "step": 2695
+    },
+    {
+      "epoch": 2.0293564170116674,
+      "grad_norm": 0.31076523661613464,
+      "learning_rate": 4.432326129058825e-05,
+      "loss": 0.3716,
+      "step": 2696
+    },
+    {
+      "epoch": 2.0301091456529923,
+      "grad_norm": 0.25622841715812683,
+      "learning_rate": 4.4318319534275583e-05,
+      "loss": 0.4644,
+      "step": 2697
+    },
+    {
+      "epoch": 2.0308618742943167,
+      "grad_norm": 0.26311925053596497,
+      "learning_rate": 4.431337593873678e-05,
+      "loss": 0.3158,
+      "step": 2698
+    },
+    {
+      "epoch": 2.0316146029356417,
+      "grad_norm": 0.2838535010814667,
+      "learning_rate": 4.4308430504512474e-05,
+      "loss": 0.2951,
+      "step": 2699
+    },
+    {
+      "epoch": 2.0323673315769666,
+      "grad_norm": 0.21860386431217194,
+      "learning_rate": 4.430348323214347e-05,
+      "loss": 0.2993,
+      "step": 2700
+    },
+    {
+      "epoch": 2.0331200602182915,
+      "grad_norm": 0.26034489274024963,
+      "learning_rate": 4.42985341221708e-05,
+      "loss": 0.2869,
+      "step": 2701
+    },
+    {
+      "epoch": 2.033872788859616,
+      "grad_norm": 0.2407262772321701,
+      "learning_rate": 4.4293583175135696e-05,
+      "loss": 0.2433,
+      "step": 2702
+    },
+    {
+      "epoch": 2.034625517500941,
+      "grad_norm": 0.20586948096752167,
+      "learning_rate": 4.428863039157959e-05,
+      "loss": 0.2337,
+      "step": 2703
+    },
+    {
+      "epoch": 2.035378246142266,
+      "grad_norm": 0.27768731117248535,
+      "learning_rate": 4.42836757720441e-05,
+      "loss": 0.4502,
+      "step": 2704
+    },
+    {
+      "epoch": 2.0361309747835907,
+      "grad_norm": 0.2516113221645355,
+      "learning_rate": 4.427871931707107e-05,
+      "loss": 0.3387,
+      "step": 2705
+    },
+    {
+      "epoch": 2.036883703424915,
+      "grad_norm": 0.2656622529029846,
+      "learning_rate": 4.427376102720251e-05,
+      "loss": 0.4519,
+      "step": 2706
+    },
+    {
+      "epoch": 2.03763643206624,
+      "grad_norm": 0.2736435830593109,
+      "learning_rate": 4.426880090298067e-05,
+      "loss": 0.3577,
+      "step": 2707
+    },
+    {
+      "epoch": 2.038389160707565,
+      "grad_norm": 0.2302788496017456,
+      "learning_rate": 4.426383894494797e-05,
+      "loss": 0.3714,
+      "step": 2708
+    },
+    {
+      "epoch": 2.0391418893488895,
+      "grad_norm": 0.2626882791519165,
+      "learning_rate": 4.4258875153647056e-05,
+      "loss": 0.3961,
+      "step": 2709
+    },
+    {
+      "epoch": 2.0398946179902144,
+      "grad_norm": 0.21057431399822235,
+      "learning_rate": 4.425390952962075e-05,
+      "loss": 0.3938,
+      "step": 2710
+    },
+    {
+      "epoch": 2.0406473466315393,
+      "grad_norm": 0.23042893409729004,
+      "learning_rate": 4.424894207341208e-05,
+      "loss": 0.3112,
+      "step": 2711
+    },
+    {
+      "epoch": 2.0414000752728643,
+      "grad_norm": 0.29034313559532166,
+      "learning_rate": 4.424397278556428e-05,
+      "loss": 0.3702,
+      "step": 2712
+    },
+    {
+      "epoch": 2.0421528039141887,
+      "grad_norm": 0.22088676691055298,
+      "learning_rate": 4.423900166662081e-05,
+      "loss": 0.406,
+      "step": 2713
+    },
+    {
+      "epoch": 2.0429055325555137,
+      "grad_norm": 0.2572060823440552,
+      "learning_rate": 4.423402871712527e-05,
+      "loss": 0.3138,
+      "step": 2714
+    },
+    {
+      "epoch": 2.0436582611968386,
+      "grad_norm": 0.21874462068080902,
+      "learning_rate": 4.422905393762152e-05,
+      "loss": 0.4011,
+      "step": 2715
+    },
+    {
+      "epoch": 2.0444109898381635,
+      "grad_norm": 0.23405887186527252,
+      "learning_rate": 4.422407732865358e-05,
+      "loss": 0.319,
+      "step": 2716
+    },
+    {
+      "epoch": 2.045163718479488,
+      "grad_norm": 0.29254159331321716,
+      "learning_rate": 4.4219098890765685e-05,
+      "loss": 0.2935,
+      "step": 2717
+    },
+    {
+      "epoch": 2.045916447120813,
+      "grad_norm": 0.319754421710968,
+      "learning_rate": 4.4214118624502264e-05,
+      "loss": 0.337,
+      "step": 2718
+    },
+    {
+      "epoch": 2.046669175762138,
+      "grad_norm": 0.26668787002563477,
+      "learning_rate": 4.4209136530407974e-05,
+      "loss": 0.3854,
+      "step": 2719
+    },
+    {
+      "epoch": 2.0474219044034627,
+      "grad_norm": 0.282386839389801,
+      "learning_rate": 4.420415260902763e-05,
+      "loss": 0.3967,
+      "step": 2720
+    },
+    {
+      "epoch": 2.048174633044787,
+      "grad_norm": 0.25886815786361694,
+      "learning_rate": 4.419916686090627e-05,
+      "loss": 0.3122,
+      "step": 2721
+    },
+    {
+      "epoch": 2.048927361686112,
+      "grad_norm": 0.2511456608772278,
+      "learning_rate": 4.4194179286589126e-05,
+      "loss": 0.369,
+      "step": 2722
+    },
+    {
+      "epoch": 2.049680090327437,
+      "grad_norm": 0.2937150299549103,
+      "learning_rate": 4.418918988662164e-05,
+      "loss": 0.4071,
+      "step": 2723
+    },
+    {
+      "epoch": 2.050432818968762,
+      "grad_norm": 0.26544344425201416,
+      "learning_rate": 4.418419866154943e-05,
+      "loss": 0.2971,
+      "step": 2724
+    },
+    {
+      "epoch": 2.0511855476100864,
+      "grad_norm": 0.233181431889534,
+      "learning_rate": 4.417920561191835e-05,
+      "loss": 0.3981,
+      "step": 2725
+    },
+    {
+      "epoch": 2.0519382762514113,
+      "grad_norm": 0.2864612936973572,
+      "learning_rate": 4.4174210738274414e-05,
+      "loss": 0.3248,
+      "step": 2726
+    },
+    {
+      "epoch": 2.0526910048927363,
+      "grad_norm": 0.22413119673728943,
+      "learning_rate": 4.4169214041163865e-05,
+      "loss": 0.3568,
+      "step": 2727
+    },
+    {
+      "epoch": 2.053443733534061,
+      "grad_norm": 0.22709019482135773,
+      "learning_rate": 4.4164215521133125e-05,
+      "loss": 0.3217,
+      "step": 2728
+    },
+    {
+      "epoch": 2.0541964621753857,
+      "grad_norm": 0.23878493905067444,
+      "learning_rate": 4.415921517872883e-05,
+      "loss": 0.3111,
+      "step": 2729
+    },
+    {
+      "epoch": 2.0549491908167106,
+      "grad_norm": 0.2712917625904083,
+      "learning_rate": 4.4154213014497816e-05,
+      "loss": 0.3252,
+      "step": 2730
+    },
+    {
+      "epoch": 2.0557019194580355,
+      "grad_norm": 0.2609381675720215,
+      "learning_rate": 4.41492090289871e-05,
+      "loss": 0.3101,
+      "step": 2731
+    },
+    {
+      "epoch": 2.0564546480993604,
+      "grad_norm": 0.22867238521575928,
+      "learning_rate": 4.414420322274392e-05,
+      "loss": 0.3774,
+      "step": 2732
+    },
+    {
+      "epoch": 2.057207376740685,
+      "grad_norm": 0.29357990622520447,
+      "learning_rate": 4.41391955963157e-05,
+      "loss": 0.2847,
+      "step": 2733
+    },
+    {
+      "epoch": 2.05796010538201,
+      "grad_norm": 0.2601042091846466,
+      "learning_rate": 4.413418615025007e-05,
+      "loss": 0.3433,
+      "step": 2734
+    },
+    {
+      "epoch": 2.0587128340233347,
+      "grad_norm": 0.23271621763706207,
+      "learning_rate": 4.412917488509484e-05,
+      "loss": 0.2705,
+      "step": 2735
+    },
+    {
+      "epoch": 2.059465562664659,
+      "grad_norm": 0.22545187175273895,
+      "learning_rate": 4.412416180139806e-05,
+      "loss": 0.2838,
+      "step": 2736
+    },
+    {
+      "epoch": 2.060218291305984,
+      "grad_norm": 0.23265483975410461,
+      "learning_rate": 4.411914689970793e-05,
+      "loss": 0.3292,
+      "step": 2737
+    },
+    {
+      "epoch": 2.060971019947309,
+      "grad_norm": 0.2582428753376007,
+      "learning_rate": 4.411413018057288e-05,
+      "loss": 0.3775,
+      "step": 2738
+    },
+    {
+      "epoch": 2.061723748588634,
+      "grad_norm": 0.22320091724395752,
+      "learning_rate": 4.410911164454154e-05,
+      "loss": 0.3506,
+      "step": 2739
+    },
+    {
+      "epoch": 2.0624764772299584,
+      "grad_norm": 0.2532583177089691,
+      "learning_rate": 4.4104091292162716e-05,
+      "loss": 0.276,
+      "step": 2740
+    },
+    {
+      "epoch": 2.0632292058712833,
+      "grad_norm": 0.2312317043542862,
+      "learning_rate": 4.409906912398544e-05,
+      "loss": 0.3897,
+      "step": 2741
+    },
+    {
+      "epoch": 2.0639819345126083,
+      "grad_norm": 0.24417926371097565,
+      "learning_rate": 4.4094045140558915e-05,
+      "loss": 0.3231,
+      "step": 2742
+    },
+    {
+      "epoch": 2.064734663153933,
+      "grad_norm": 0.2504076361656189,
+      "learning_rate": 4.408901934243256e-05,
+      "loss": 0.384,
+      "step": 2743
+    },
+    {
+      "epoch": 2.0654873917952576,
+      "grad_norm": 0.24249520897865295,
+      "learning_rate": 4.4083991730155996e-05,
+      "loss": 0.3389,
+      "step": 2744
+    },
+    {
+      "epoch": 2.0662401204365826,
+      "grad_norm": 0.49307185411453247,
+      "learning_rate": 4.407896230427903e-05,
+      "loss": 0.3317,
+      "step": 2745
+    },
+    {
+      "epoch": 2.0669928490779075,
+      "grad_norm": 0.2844165563583374,
+      "learning_rate": 4.407393106535167e-05,
+      "loss": 0.278,
+      "step": 2746
+    },
+    {
+      "epoch": 2.0677455777192324,
+      "grad_norm": 0.26344573497772217,
+      "learning_rate": 4.4068898013924126e-05,
+      "loss": 0.3964,
+      "step": 2747
+    },
+    {
+      "epoch": 2.068498306360557,
+      "grad_norm": 0.32131174206733704,
+      "learning_rate": 4.4063863150546814e-05,
+      "loss": 0.4057,
+      "step": 2748
+    },
+    {
+      "epoch": 2.069251035001882,
+      "grad_norm": 0.2625237703323364,
+      "learning_rate": 4.405882647577032e-05,
+      "loss": 0.371,
+      "step": 2749
+    },
+    {
+      "epoch": 2.0700037636432067,
+      "grad_norm": 0.24295015633106232,
+      "learning_rate": 4.4053787990145465e-05,
+      "loss": 0.3487,
+      "step": 2750
+    },
+    {
+      "epoch": 2.0707564922845316,
+      "grad_norm": 0.27463072538375854,
+      "learning_rate": 4.404874769422324e-05,
+      "loss": 0.3697,
+      "step": 2751
+    },
+    {
+      "epoch": 2.071509220925856,
+      "grad_norm": 0.2337772101163864,
+      "learning_rate": 4.404370558855485e-05,
+      "loss": 0.2013,
+      "step": 2752
+    },
+    {
+      "epoch": 2.072261949567181,
+      "grad_norm": 0.2543869614601135,
+      "learning_rate": 4.403866167369169e-05,
+      "loss": 0.3283,
+      "step": 2753
+    },
+    {
+      "epoch": 2.073014678208506,
+      "grad_norm": 1.6927852630615234,
+      "learning_rate": 4.403361595018536e-05,
+      "loss": 0.4846,
+      "step": 2754
+    },
+    {
+      "epoch": 2.0737674068498304,
+      "grad_norm": 0.29285213351249695,
+      "learning_rate": 4.4028568418587636e-05,
+      "loss": 0.3414,
+      "step": 2755
+    },
+    {
+      "epoch": 2.0745201354911553,
+      "grad_norm": 0.29651182889938354,
+      "learning_rate": 4.4023519079450527e-05,
+      "loss": 0.3534,
+      "step": 2756
+    },
+    {
+      "epoch": 2.0752728641324802,
+      "grad_norm": 0.3554665148258209,
+      "learning_rate": 4.401846793332622e-05,
+      "loss": 0.3563,
+      "step": 2757
+    },
+    {
+      "epoch": 2.076025592773805,
+      "grad_norm": 0.25584876537323,
+      "learning_rate": 4.401341498076709e-05,
+      "loss": 0.26,
+      "step": 2758
+    },
+    {
+      "epoch": 2.0767783214151296,
+      "grad_norm": 0.25800368189811707,
+      "learning_rate": 4.400836022232572e-05,
+      "loss": 0.2674,
+      "step": 2759
+    },
+    {
+      "epoch": 2.0775310500564546,
+      "grad_norm": 0.26815587282180786,
+      "learning_rate": 4.4003303658554904e-05,
+      "loss": 0.309,
+      "step": 2760
+    },
+    {
+      "epoch": 2.0782837786977795,
+      "grad_norm": 0.2535979747772217,
+      "learning_rate": 4.3998245290007606e-05,
+      "loss": 0.3961,
+      "step": 2761
+    },
+    {
+      "epoch": 2.0790365073391044,
+      "grad_norm": 0.5836333632469177,
+      "learning_rate": 4.399318511723701e-05,
+      "loss": 0.3059,
+      "step": 2762
+    },
+    {
+      "epoch": 2.079789235980429,
+      "grad_norm": 0.22383765876293182,
+      "learning_rate": 4.398812314079648e-05,
+      "loss": 0.3608,
+      "step": 2763
+    },
+    {
+      "epoch": 2.080541964621754,
+      "grad_norm": 0.28617700934410095,
+      "learning_rate": 4.3983059361239606e-05,
+      "loss": 0.3427,
+      "step": 2764
+    },
+    {
+      "epoch": 2.0812946932630787,
+      "grad_norm": 0.20276646316051483,
+      "learning_rate": 4.397799377912013e-05,
+      "loss": 0.3721,
+      "step": 2765
+    },
+    {
+      "epoch": 2.0820474219044036,
+      "grad_norm": 0.261525422334671,
+      "learning_rate": 4.3972926394992034e-05,
+      "loss": 0.3456,
+      "step": 2766
+    },
+    {
+      "epoch": 2.082800150545728,
+      "grad_norm": 0.22354057431221008,
+      "learning_rate": 4.396785720940948e-05,
+      "loss": 0.3113,
+      "step": 2767
+    },
+    {
+      "epoch": 2.083552879187053,
+      "grad_norm": 0.2347208559513092,
+      "learning_rate": 4.3962786222926814e-05,
+      "loss": 0.3291,
+      "step": 2768
+    },
+    {
+      "epoch": 2.084305607828378,
+      "grad_norm": 0.2740861177444458,
+      "learning_rate": 4.39577134360986e-05,
+      "loss": 0.2851,
+      "step": 2769
+    },
+    {
+      "epoch": 2.085058336469703,
+      "grad_norm": 0.24930934607982635,
+      "learning_rate": 4.3952638849479586e-05,
+      "loss": 0.3702,
+      "step": 2770
+    },
+    {
+      "epoch": 2.0858110651110273,
+      "grad_norm": 0.24003024399280548,
+      "learning_rate": 4.3947562463624726e-05,
+      "loss": 0.3325,
+      "step": 2771
+    },
+    {
+      "epoch": 2.0865637937523522,
+      "grad_norm": 0.2800670862197876,
+      "learning_rate": 4.394248427908916e-05,
+      "loss": 0.3285,
+      "step": 2772
+    },
+    {
+      "epoch": 2.087316522393677,
+      "grad_norm": 0.258011132478714,
+      "learning_rate": 4.393740429642824e-05,
+      "loss": 0.3651,
+      "step": 2773
+    },
+    {
+      "epoch": 2.088069251035002,
+      "grad_norm": 0.23854978382587433,
+      "learning_rate": 4.39323225161975e-05,
+      "loss": 0.4276,
+      "step": 2774
+    },
+    {
+      "epoch": 2.0888219796763265,
+      "grad_norm": 0.2617243230342865,
+      "learning_rate": 4.392723893895267e-05,
+      "loss": 0.2963,
+      "step": 2775
+    },
+    {
+      "epoch": 2.0895747083176515,
+      "grad_norm": 0.20050552487373352,
+      "learning_rate": 4.392215356524969e-05,
+      "loss": 0.3319,
+      "step": 2776
+    },
+    {
+      "epoch": 2.0903274369589764,
+      "grad_norm": 0.21782904863357544,
+      "learning_rate": 4.391706639564468e-05,
+      "loss": 0.2622,
+      "step": 2777
+    },
+    {
+      "epoch": 2.0910801656003013,
+      "grad_norm": 0.24402092397212982,
+      "learning_rate": 4.3911977430693986e-05,
+      "loss": 0.3609,
+      "step": 2778
+    },
+    {
+      "epoch": 2.0918328942416258,
+      "grad_norm": 0.19673487544059753,
+      "learning_rate": 4.3906886670954096e-05,
+      "loss": 0.3354,
+      "step": 2779
+    },
+    {
+      "epoch": 2.0925856228829507,
+      "grad_norm": 0.22465091943740845,
+      "learning_rate": 4.390179411698176e-05,
+      "loss": 0.2225,
+      "step": 2780
+    },
+    {
+      "epoch": 2.0933383515242756,
+      "grad_norm": 0.24059471487998962,
+      "learning_rate": 4.389669976933388e-05,
+      "loss": 0.2576,
+      "step": 2781
+    },
+    {
+      "epoch": 2.0940910801656,
+      "grad_norm": 0.21805280447006226,
+      "learning_rate": 4.389160362856756e-05,
+      "loss": 0.4277,
+      "step": 2782
+    },
+    {
+      "epoch": 2.094843808806925,
+      "grad_norm": 0.2951591908931732,
+      "learning_rate": 4.3886505695240124e-05,
+      "loss": 0.3794,
+      "step": 2783
+    },
+    {
+      "epoch": 2.09559653744825,
+      "grad_norm": 0.22029300034046173,
+      "learning_rate": 4.3881405969909046e-05,
+      "loss": 0.2903,
+      "step": 2784
+    },
+    {
+      "epoch": 2.096349266089575,
+      "grad_norm": 0.2998780608177185,
+      "learning_rate": 4.387630445313204e-05,
+      "loss": 0.2833,
+      "step": 2785
+    },
+    {
+      "epoch": 2.0971019947308993,
+      "grad_norm": 0.23853681981563568,
+      "learning_rate": 4.387120114546701e-05,
+      "loss": 0.258,
+      "step": 2786
+    },
+    {
+      "epoch": 2.0978547233722242,
+      "grad_norm": 0.29853636026382446,
+      "learning_rate": 4.3866096047472035e-05,
+      "loss": 0.267,
+      "step": 2787
+    },
+    {
+      "epoch": 2.098607452013549,
+      "grad_norm": 0.2429654896259308,
+      "learning_rate": 4.3860989159705395e-05,
+      "loss": 0.2621,
+      "step": 2788
+    },
+    {
+      "epoch": 2.099360180654874,
+      "grad_norm": 0.38576745986938477,
+      "learning_rate": 4.385588048272558e-05,
+      "loss": 0.4433,
+      "step": 2789
+    },
+    {
+      "epoch": 2.1001129092961985,
+      "grad_norm": 0.2631751000881195,
+      "learning_rate": 4.385077001709127e-05,
+      "loss": 0.3182,
+      "step": 2790
+    },
+    {
+      "epoch": 2.1008656379375235,
+      "grad_norm": 0.26945045590400696,
+      "learning_rate": 4.3845657763361334e-05,
+      "loss": 0.3763,
+      "step": 2791
+    },
+    {
+      "epoch": 2.1016183665788484,
+      "grad_norm": 0.22530975937843323,
+      "learning_rate": 4.384054372209484e-05,
+      "loss": 0.3961,
+      "step": 2792
+    },
+    {
+      "epoch": 2.1023710952201733,
+      "grad_norm": 0.22856274247169495,
+      "learning_rate": 4.383542789385106e-05,
+      "loss": 0.3195,
+      "step": 2793
+    },
+    {
+      "epoch": 2.1031238238614978,
+      "grad_norm": 0.2894219160079956,
+      "learning_rate": 4.383031027918943e-05,
+      "loss": 0.3246,
+      "step": 2794
+    },
+    {
+      "epoch": 2.1038765525028227,
+      "grad_norm": 0.23839499056339264,
+      "learning_rate": 4.382519087866963e-05,
+      "loss": 0.2778,
+      "step": 2795
+    },
+    {
+      "epoch": 2.1046292811441476,
+      "grad_norm": 0.23057936131954193,
+      "learning_rate": 4.38200696928515e-05,
+      "loss": 0.3533,
+      "step": 2796
+    },
+    {
+      "epoch": 2.1053820097854725,
+      "grad_norm": 0.20821987092494965,
+      "learning_rate": 4.381494672229508e-05,
+      "loss": 0.2785,
+      "step": 2797
+    },
+    {
+      "epoch": 2.106134738426797,
+      "grad_norm": 0.24100840091705322,
+      "learning_rate": 4.380982196756062e-05,
+      "loss": 0.3264,
+      "step": 2798
+    },
+    {
+      "epoch": 2.106887467068122,
+      "grad_norm": 0.2662712633609772,
+      "learning_rate": 4.380469542920855e-05,
+      "loss": 0.3465,
+      "step": 2799
+    },
+    {
+      "epoch": 2.107640195709447,
+      "grad_norm": 0.2489176094532013,
+      "learning_rate": 4.379956710779951e-05,
+      "loss": 0.2634,
+      "step": 2800
+    },
+    {
+      "epoch": 2.107640195709447,
+      "eval_loss": 0.33923009037971497,
+      "eval_runtime": 452.687,
+      "eval_samples_per_second": 21.266,
+      "eval_steps_per_second": 0.665,
+      "step": 2800
+    },
+    {
+      "epoch": 2.1083929243507717,
+      "grad_norm": 0.21650879085063934,
+      "learning_rate": 4.379443700389431e-05,
+      "loss": 0.4196,
+      "step": 2801
+    },
+    {
+      "epoch": 2.109145652992096,
+      "grad_norm": 0.21893756091594696,
+      "learning_rate": 4.378930511805397e-05,
+      "loss": 0.2281,
+      "step": 2802
+    },
+    {
+      "epoch": 2.109898381633421,
+      "grad_norm": 0.24146457016468048,
+      "learning_rate": 4.378417145083972e-05,
+      "loss": 0.3326,
+      "step": 2803
+    },
+    {
+      "epoch": 2.110651110274746,
+      "grad_norm": 0.2302742898464203,
+      "learning_rate": 4.377903600281296e-05,
+      "loss": 0.3849,
+      "step": 2804
+    },
+    {
+      "epoch": 2.111403838916071,
+      "grad_norm": 0.22440895438194275,
+      "learning_rate": 4.37738987745353e-05,
+      "loss": 0.3181,
+      "step": 2805
+    },
+    {
+      "epoch": 2.1121565675573954,
+      "grad_norm": 0.23284725844860077,
+      "learning_rate": 4.376875976656854e-05,
+      "loss": 0.3082,
+      "step": 2806
+    },
+    {
+      "epoch": 2.1129092961987204,
+      "grad_norm": 0.2475927621126175,
+      "learning_rate": 4.3763618979474665e-05,
+      "loss": 0.3382,
+      "step": 2807
+    },
+    {
+      "epoch": 2.1136620248400453,
+      "grad_norm": 0.29367899894714355,
+      "learning_rate": 4.375847641381587e-05,
+      "loss": 0.3631,
+      "step": 2808
+    },
+    {
+      "epoch": 2.1144147534813698,
+      "grad_norm": 0.25572118163108826,
+      "learning_rate": 4.375333207015454e-05,
+      "loss": 0.2831,
+      "step": 2809
+    },
+    {
+      "epoch": 2.1151674821226947,
+      "grad_norm": 0.24017171561717987,
+      "learning_rate": 4.374818594905327e-05,
+      "loss": 0.2381,
+      "step": 2810
+    },
+    {
+      "epoch": 2.1159202107640196,
+      "grad_norm": 0.2717092037200928,
+      "learning_rate": 4.3743038051074794e-05,
+      "loss": 0.4044,
+      "step": 2811
+    },
+    {
+      "epoch": 2.1166729394053445,
+      "grad_norm": 0.26866596937179565,
+      "learning_rate": 4.37378883767821e-05,
+      "loss": 0.3964,
+      "step": 2812
+    },
+    {
+      "epoch": 2.117425668046669,
+      "grad_norm": 0.2578282058238983,
+      "learning_rate": 4.373273692673835e-05,
+      "loss": 0.4596,
+      "step": 2813
+    },
+    {
+      "epoch": 2.118178396687994,
+      "grad_norm": 0.2866767942905426,
+      "learning_rate": 4.372758370150689e-05,
+      "loss": 0.2892,
+      "step": 2814
+    },
+    {
+      "epoch": 2.118931125329319,
+      "grad_norm": 0.251654714345932,
+      "learning_rate": 4.372242870165128e-05,
+      "loss": 0.3549,
+      "step": 2815
+    },
+    {
+      "epoch": 2.1196838539706437,
+      "grad_norm": 0.31731146574020386,
+      "learning_rate": 4.371727192773526e-05,
+      "loss": 0.359,
+      "step": 2816
+    },
+    {
+      "epoch": 2.120436582611968,
+      "grad_norm": 0.31404754519462585,
+      "learning_rate": 4.3712113380322754e-05,
+      "loss": 0.4055,
+      "step": 2817
+    },
+    {
+      "epoch": 2.121189311253293,
+      "grad_norm": 0.2590123116970062,
+      "learning_rate": 4.3706953059977907e-05,
+      "loss": 0.3013,
+      "step": 2818
+    },
+    {
+      "epoch": 2.121942039894618,
+      "grad_norm": 0.2817533016204834,
+      "learning_rate": 4.370179096726504e-05,
+      "loss": 0.3316,
+      "step": 2819
+    },
+    {
+      "epoch": 2.122694768535943,
+      "grad_norm": 0.26242583990097046,
+      "learning_rate": 4.369662710274867e-05,
+      "loss": 0.3167,
+      "step": 2820
+    },
+    {
+      "epoch": 2.1234474971772674,
+      "grad_norm": 0.2866950035095215,
+      "learning_rate": 4.369146146699352e-05,
+      "loss": 0.3442,
+      "step": 2821
+    },
+    {
+      "epoch": 2.1242002258185924,
+      "grad_norm": 0.2287922501564026,
+      "learning_rate": 4.368629406056448e-05,
+      "loss": 0.3116,
+      "step": 2822
+    },
+    {
+      "epoch": 2.1249529544599173,
+      "grad_norm": 0.3480190336704254,
+      "learning_rate": 4.368112488402667e-05,
+      "loss": 0.1764,
+      "step": 2823
+    },
+    {
+      "epoch": 2.125705683101242,
+      "grad_norm": 0.2640867233276367,
+      "learning_rate": 4.3675953937945344e-05,
+      "loss": 0.2942,
+      "step": 2824
+    },
+    {
+      "epoch": 2.1264584117425667,
+      "grad_norm": 0.2453736513853073,
+      "learning_rate": 4.3670781222886044e-05,
+      "loss": 0.3075,
+      "step": 2825
+    },
+    {
+      "epoch": 2.1272111403838916,
+      "grad_norm": 0.2410285621881485,
+      "learning_rate": 4.36656067394144e-05,
+      "loss": 0.2876,
+      "step": 2826
+    },
+    {
+      "epoch": 2.1279638690252165,
+      "grad_norm": 0.24527841806411743,
+      "learning_rate": 4.366043048809633e-05,
+      "loss": 0.2662,
+      "step": 2827
+    },
+    {
+      "epoch": 2.128716597666541,
+      "grad_norm": 0.26857423782348633,
+      "learning_rate": 4.365525246949786e-05,
+      "loss": 0.321,
+      "step": 2828
+    },
+    {
+      "epoch": 2.129469326307866,
+      "grad_norm": 0.2392696589231491,
+      "learning_rate": 4.3650072684185274e-05,
+      "loss": 0.3825,
+      "step": 2829
+    },
+    {
+      "epoch": 2.130222054949191,
+      "grad_norm": 0.2840482294559479,
+      "learning_rate": 4.3644891132725016e-05,
+      "loss": 0.3271,
+      "step": 2830
+    },
+    {
+      "epoch": 2.1309747835905157,
+      "grad_norm": 0.27713486552238464,
+      "learning_rate": 4.363970781568375e-05,
+      "loss": 0.3367,
+      "step": 2831
+    },
+    {
+      "epoch": 2.1317275122318406,
+      "grad_norm": 0.2501160800457001,
+      "learning_rate": 4.363452273362829e-05,
+      "loss": 0.324,
+      "step": 2832
+    },
+    {
+      "epoch": 2.132480240873165,
+      "grad_norm": 0.23428604006767273,
+      "learning_rate": 4.362933588712569e-05,
+      "loss": 0.364,
+      "step": 2833
+    },
+    {
+      "epoch": 2.13323296951449,
+      "grad_norm": 0.2679921090602875,
+      "learning_rate": 4.3624147276743163e-05,
+      "loss": 0.3166,
+      "step": 2834
+    },
+    {
+      "epoch": 2.133985698155815,
+      "grad_norm": 0.31787270307540894,
+      "learning_rate": 4.361895690304813e-05,
+      "loss": 0.4029,
+      "step": 2835
+    },
+    {
+      "epoch": 2.1347384267971394,
+      "grad_norm": 0.2294486165046692,
+      "learning_rate": 4.361376476660822e-05,
+      "loss": 0.2976,
+      "step": 2836
+    },
+    {
+      "epoch": 2.1354911554384644,
+      "grad_norm": 0.23854365944862366,
+      "learning_rate": 4.36085708679912e-05,
+      "loss": 0.282,
+      "step": 2837
+    },
+    {
+      "epoch": 2.1362438840797893,
+      "grad_norm": 0.23996005952358246,
+      "learning_rate": 4.36033752077651e-05,
+      "loss": 0.2812,
+      "step": 2838
+    },
+    {
+      "epoch": 2.136996612721114,
+      "grad_norm": 0.27016231417655945,
+      "learning_rate": 4.35981777864981e-05,
+      "loss": 0.3194,
+      "step": 2839
+    },
+    {
+      "epoch": 2.1377493413624387,
+      "grad_norm": 0.24798771739006042,
+      "learning_rate": 4.359297860475858e-05,
+      "loss": 0.3491,
+      "step": 2840
+    },
+    {
+      "epoch": 2.1385020700037636,
+      "grad_norm": 0.23530200123786926,
+      "learning_rate": 4.358777766311511e-05,
+      "loss": 0.3835,
+      "step": 2841
+    },
+    {
+      "epoch": 2.1392547986450885,
+      "grad_norm": 0.21979792416095734,
+      "learning_rate": 4.3582574962136465e-05,
+      "loss": 0.3614,
+      "step": 2842
+    },
+    {
+      "epoch": 2.1400075272864134,
+      "grad_norm": 0.2245541214942932,
+      "learning_rate": 4.3577370502391604e-05,
+      "loss": 0.2934,
+      "step": 2843
+    },
+    {
+      "epoch": 2.140760255927738,
+      "grad_norm": 0.24512873589992523,
+      "learning_rate": 4.3572164284449673e-05,
+      "loss": 0.2818,
+      "step": 2844
+    },
+    {
+      "epoch": 2.141512984569063,
+      "grad_norm": 0.25681793689727783,
+      "learning_rate": 4.356695630888002e-05,
+      "loss": 0.3342,
+      "step": 2845
+    },
+    {
+      "epoch": 2.1422657132103877,
+      "grad_norm": 0.2700720727443695,
+      "learning_rate": 4.356174657625218e-05,
+      "loss": 0.2882,
+      "step": 2846
+    },
+    {
+      "epoch": 2.1430184418517126,
+      "grad_norm": 0.2202654480934143,
+      "learning_rate": 4.355653508713589e-05,
+      "loss": 0.3261,
+      "step": 2847
+    },
+    {
+      "epoch": 2.143771170493037,
+      "grad_norm": 0.2694127857685089,
+      "learning_rate": 4.355132184210106e-05,
+      "loss": 0.3761,
+      "step": 2848
+    },
+    {
+      "epoch": 2.144523899134362,
+      "grad_norm": 0.22846969962120056,
+      "learning_rate": 4.354610684171781e-05,
+      "loss": 0.3037,
+      "step": 2849
+    },
+    {
+      "epoch": 2.145276627775687,
+      "grad_norm": 0.25535261631011963,
+      "learning_rate": 4.3540890086556435e-05,
+      "loss": 0.3444,
+      "step": 2850
+    },
+    {
+      "epoch": 2.146029356417012,
+      "grad_norm": 0.25501543283462524,
+      "learning_rate": 4.3535671577187445e-05,
+      "loss": 0.3393,
+      "step": 2851
+    },
+    {
+      "epoch": 2.1467820850583363,
+      "grad_norm": 0.26403388381004333,
+      "learning_rate": 4.353045131418151e-05,
+      "loss": 0.3859,
+      "step": 2852
+    },
+    {
+      "epoch": 2.1475348136996613,
+      "grad_norm": 0.22748221457004547,
+      "learning_rate": 4.3525229298109526e-05,
+      "loss": 0.3518,
+      "step": 2853
+    },
+    {
+      "epoch": 2.148287542340986,
+      "grad_norm": 0.2518693208694458,
+      "learning_rate": 4.352000552954256e-05,
+      "loss": 0.3294,
+      "step": 2854
+    },
+    {
+      "epoch": 2.1490402709823107,
+      "grad_norm": 0.21239973604679108,
+      "learning_rate": 4.351478000905187e-05,
+      "loss": 0.3942,
+      "step": 2855
+    },
+    {
+      "epoch": 2.1497929996236356,
+      "grad_norm": 0.20441406965255737,
+      "learning_rate": 4.350955273720893e-05,
+      "loss": 0.3385,
+      "step": 2856
+    },
+    {
+      "epoch": 2.1505457282649605,
+      "grad_norm": 0.28044071793556213,
+      "learning_rate": 4.350432371458536e-05,
+      "loss": 0.3527,
+      "step": 2857
+    },
+    {
+      "epoch": 2.1512984569062854,
+      "grad_norm": 0.22722291946411133,
+      "learning_rate": 4.3499092941753006e-05,
+      "loss": 0.268,
+      "step": 2858
+    },
+    {
+      "epoch": 2.15205118554761,
+      "grad_norm": 0.2726752460002899,
+      "learning_rate": 4.349386041928391e-05,
+      "loss": 0.3624,
+      "step": 2859
+    },
+    {
+      "epoch": 2.152803914188935,
+      "grad_norm": 0.23964020609855652,
+      "learning_rate": 4.348862614775028e-05,
+      "loss": 0.3912,
+      "step": 2860
+    },
+    {
+      "epoch": 2.1535566428302597,
+      "grad_norm": 0.23426704108715057,
+      "learning_rate": 4.348339012772454e-05,
+      "loss": 0.2743,
+      "step": 2861
+    },
+    {
+      "epoch": 2.1543093714715846,
+      "grad_norm": 0.23135240375995636,
+      "learning_rate": 4.347815235977929e-05,
+      "loss": 0.337,
+      "step": 2862
+    },
+    {
+      "epoch": 2.155062100112909,
+      "grad_norm": 0.23875567317008972,
+      "learning_rate": 4.347291284448731e-05,
+      "loss": 0.3304,
+      "step": 2863
+    },
+    {
+      "epoch": 2.155814828754234,
+      "grad_norm": 0.23768161237239838,
+      "learning_rate": 4.346767158242161e-05,
+      "loss": 0.3182,
+      "step": 2864
+    },
+    {
+      "epoch": 2.156567557395559,
+      "grad_norm": 0.21287725865840912,
+      "learning_rate": 4.3462428574155336e-05,
+      "loss": 0.3434,
+      "step": 2865
+    },
+    {
+      "epoch": 2.157320286036884,
+      "grad_norm": 0.24687443673610687,
+      "learning_rate": 4.345718382026189e-05,
+      "loss": 0.2852,
+      "step": 2866
+    },
+    {
+      "epoch": 2.1580730146782083,
+      "grad_norm": 0.2675294280052185,
+      "learning_rate": 4.34519373213148e-05,
+      "loss": 0.3261,
+      "step": 2867
+    },
+    {
+      "epoch": 2.1588257433195333,
+      "grad_norm": 0.1967107653617859,
+      "learning_rate": 4.344668907788782e-05,
+      "loss": 0.2787,
+      "step": 2868
+    },
+    {
+      "epoch": 2.159578471960858,
+      "grad_norm": 0.2775914669036865,
+      "learning_rate": 4.344143909055492e-05,
+      "loss": 0.3491,
+      "step": 2869
+    },
+    {
+      "epoch": 2.160331200602183,
+      "grad_norm": 0.2910844683647156,
+      "learning_rate": 4.343618735989019e-05,
+      "loss": 0.3491,
+      "step": 2870
+    },
+    {
+      "epoch": 2.1610839292435076,
+      "grad_norm": 0.26071488857269287,
+      "learning_rate": 4.343093388646797e-05,
+      "loss": 0.2935,
+      "step": 2871
+    },
+    {
+      "epoch": 2.1618366578848325,
+      "grad_norm": 0.26960721611976624,
+      "learning_rate": 4.3425678670862786e-05,
+      "loss": 0.3822,
+      "step": 2872
+    },
+    {
+      "epoch": 2.1625893865261574,
+      "grad_norm": 0.31412941217422485,
+      "learning_rate": 4.342042171364932e-05,
+      "loss": 0.3678,
+      "step": 2873
+    },
+    {
+      "epoch": 2.1633421151674823,
+      "grad_norm": 0.24239006638526917,
+      "learning_rate": 4.341516301540247e-05,
+      "loss": 0.2648,
+      "step": 2874
+    },
+    {
+      "epoch": 2.164094843808807,
+      "grad_norm": 0.31101781129837036,
+      "learning_rate": 4.340990257669732e-05,
+      "loss": 0.3602,
+      "step": 2875
+    },
+    {
+      "epoch": 2.1648475724501317,
+      "grad_norm": 0.23393119871616364,
+      "learning_rate": 4.340464039810915e-05,
+      "loss": 0.331,
+      "step": 2876
+    },
+    {
+      "epoch": 2.1656003010914566,
+      "grad_norm": 0.2830849885940552,
+      "learning_rate": 4.339937648021341e-05,
+      "loss": 0.239,
+      "step": 2877
+    },
+    {
+      "epoch": 2.1663530297327815,
+      "grad_norm": 0.34498167037963867,
+      "learning_rate": 4.339411082358577e-05,
+      "loss": 0.3185,
+      "step": 2878
+    },
+    {
+      "epoch": 2.167105758374106,
+      "grad_norm": 0.29829004406929016,
+      "learning_rate": 4.338884342880206e-05,
+      "loss": 0.3826,
+      "step": 2879
+    },
+    {
+      "epoch": 2.167858487015431,
+      "grad_norm": 0.2575177848339081,
+      "learning_rate": 4.338357429643833e-05,
+      "loss": 0.3829,
+      "step": 2880
+    },
+    {
+      "epoch": 2.168611215656756,
+      "grad_norm": 0.3038381338119507,
+      "learning_rate": 4.3378303427070786e-05,
+      "loss": 0.326,
+      "step": 2881
+    },
+    {
+      "epoch": 2.1693639442980803,
+      "grad_norm": 0.23678356409072876,
+      "learning_rate": 4.3373030821275856e-05,
+      "loss": 0.2826,
+      "step": 2882
+    },
+    {
+      "epoch": 2.1701166729394052,
+      "grad_norm": 0.24323734641075134,
+      "learning_rate": 4.336775647963014e-05,
+      "loss": 0.3062,
+      "step": 2883
+    },
+    {
+      "epoch": 2.17086940158073,
+      "grad_norm": 0.25698065757751465,
+      "learning_rate": 4.336248040271043e-05,
+      "loss": 0.4247,
+      "step": 2884
+    },
+    {
+      "epoch": 2.171622130222055,
+      "grad_norm": 0.25727424025535583,
+      "learning_rate": 4.335720259109372e-05,
+      "loss": 0.3057,
+      "step": 2885
+    },
+    {
+      "epoch": 2.1723748588633796,
+      "grad_norm": 0.29216012358665466,
+      "learning_rate": 4.3351923045357175e-05,
+      "loss": 0.3697,
+      "step": 2886
+    },
+    {
+      "epoch": 2.1731275875047045,
+      "grad_norm": 0.22975106537342072,
+      "learning_rate": 4.334664176607815e-05,
+      "loss": 0.4042,
+      "step": 2887
+    },
+    {
+      "epoch": 2.1738803161460294,
+      "grad_norm": 0.2555522322654724,
+      "learning_rate": 4.3341358753834213e-05,
+      "loss": 0.3656,
+      "step": 2888
+    },
+    {
+      "epoch": 2.1746330447873543,
+      "grad_norm": 0.28169527649879456,
+      "learning_rate": 4.3336074009203097e-05,
+      "loss": 0.3345,
+      "step": 2889
+    },
+    {
+      "epoch": 2.175385773428679,
+      "grad_norm": 0.2910304665565491,
+      "learning_rate": 4.333078753276274e-05,
+      "loss": 0.2126,
+      "step": 2890
+    },
+    {
+      "epoch": 2.1761385020700037,
+      "grad_norm": 0.6625583171844482,
+      "learning_rate": 4.332549932509125e-05,
+      "loss": 0.4353,
+      "step": 2891
+    },
+    {
+      "epoch": 2.1768912307113286,
+      "grad_norm": 0.23817738890647888,
+      "learning_rate": 4.332020938676695e-05,
+      "loss": 0.3227,
+      "step": 2892
+    },
+    {
+      "epoch": 2.1776439593526535,
+      "grad_norm": 0.30772241950035095,
+      "learning_rate": 4.3314917718368345e-05,
+      "loss": 0.4306,
+      "step": 2893
+    },
+    {
+      "epoch": 2.178396687993978,
+      "grad_norm": 0.28572091460227966,
+      "learning_rate": 4.330962432047411e-05,
+      "loss": 0.2904,
+      "step": 2894
+    },
+    {
+      "epoch": 2.179149416635303,
+      "grad_norm": 0.30406618118286133,
+      "learning_rate": 4.330432919366313e-05,
+      "loss": 0.3656,
+      "step": 2895
+    },
+    {
+      "epoch": 2.179902145276628,
+      "grad_norm": 0.26745840907096863,
+      "learning_rate": 4.329903233851447e-05,
+      "loss": 0.3075,
+      "step": 2896
+    },
+    {
+      "epoch": 2.1806548739179528,
+      "grad_norm": 0.2867394685745239,
+      "learning_rate": 4.329373375560739e-05,
+      "loss": 0.2717,
+      "step": 2897
+    },
+    {
+      "epoch": 2.1814076025592772,
+      "grad_norm": 0.24460767209529877,
+      "learning_rate": 4.328843344552133e-05,
+      "loss": 0.3054,
+      "step": 2898
+    },
+    {
+      "epoch": 2.182160331200602,
+      "grad_norm": 0.2760620713233948,
+      "learning_rate": 4.328313140883592e-05,
+      "loss": 0.3577,
+      "step": 2899
+    },
+    {
+      "epoch": 2.182913059841927,
+      "grad_norm": 0.2150881141424179,
+      "learning_rate": 4.327782764613099e-05,
+      "loss": 0.2904,
+      "step": 2900
+    },
+    {
+      "epoch": 2.183665788483252,
+      "grad_norm": 0.266054630279541,
+      "learning_rate": 4.327252215798656e-05,
+      "loss": 0.3741,
+      "step": 2901
+    },
+    {
+      "epoch": 2.1844185171245765,
+      "grad_norm": 0.2715311348438263,
+      "learning_rate": 4.326721494498282e-05,
+      "loss": 0.3878,
+      "step": 2902
+    },
+    {
+      "epoch": 2.1851712457659014,
+      "grad_norm": 0.27794256806373596,
+      "learning_rate": 4.3261906007700156e-05,
+      "loss": 0.2558,
+      "step": 2903
+    },
+    {
+      "epoch": 2.1859239744072263,
+      "grad_norm": 0.2654722332954407,
+      "learning_rate": 4.325659534671914e-05,
+      "loss": 0.3494,
+      "step": 2904
+    },
+    {
+      "epoch": 2.186676703048551,
+      "grad_norm": 0.25637856125831604,
+      "learning_rate": 4.325128296262056e-05,
+      "loss": 0.3738,
+      "step": 2905
+    },
+    {
+      "epoch": 2.1874294316898757,
+      "grad_norm": 0.35823550820350647,
+      "learning_rate": 4.324596885598534e-05,
+      "loss": 0.3614,
+      "step": 2906
+    },
+    {
+      "epoch": 2.1881821603312006,
+      "grad_norm": 0.3495146334171295,
+      "learning_rate": 4.324065302739466e-05,
+      "loss": 0.4108,
+      "step": 2907
+    },
+    {
+      "epoch": 2.1889348889725255,
+      "grad_norm": 0.2624599039554596,
+      "learning_rate": 4.3235335477429814e-05,
+      "loss": 0.2821,
+      "step": 2908
+    },
+    {
+      "epoch": 2.18968761761385,
+      "grad_norm": 0.2828291058540344,
+      "learning_rate": 4.323001620667234e-05,
+      "loss": 0.3558,
+      "step": 2909
+    },
+    {
+      "epoch": 2.190440346255175,
+      "grad_norm": 0.24262480437755585,
+      "learning_rate": 4.322469521570395e-05,
+      "loss": 0.2856,
+      "step": 2910
+    },
+    {
+      "epoch": 2.1911930748965,
+      "grad_norm": 0.27486103773117065,
+      "learning_rate": 4.3219372505106534e-05,
+      "loss": 0.3749,
+      "step": 2911
+    },
+    {
+      "epoch": 2.1919458035378248,
+      "grad_norm": 0.26400405168533325,
+      "learning_rate": 4.321404807546217e-05,
+      "loss": 0.2616,
+      "step": 2912
+    },
+    {
+      "epoch": 2.1926985321791492,
+      "grad_norm": 0.2524729073047638,
+      "learning_rate": 4.320872192735314e-05,
+      "loss": 0.2557,
+      "step": 2913
+    },
+    {
+      "epoch": 2.193451260820474,
+      "grad_norm": 0.2487626075744629,
+      "learning_rate": 4.320339406136189e-05,
+      "loss": 0.3062,
+      "step": 2914
+    },
+    {
+      "epoch": 2.194203989461799,
+      "grad_norm": 0.22977083921432495,
+      "learning_rate": 4.319806447807108e-05,
+      "loss": 0.3437,
+      "step": 2915
+    },
+    {
+      "epoch": 2.194956718103124,
+      "grad_norm": 0.30964401364326477,
+      "learning_rate": 4.319273317806354e-05,
+      "loss": 0.354,
+      "step": 2916
+    },
+    {
+      "epoch": 2.1957094467444485,
+      "grad_norm": 0.23919643461704254,
+      "learning_rate": 4.3187400161922294e-05,
+      "loss": 0.3087,
+      "step": 2917
+    },
+    {
+      "epoch": 2.1964621753857734,
+      "grad_norm": 0.266682505607605,
+      "learning_rate": 4.318206543023054e-05,
+      "loss": 0.3837,
+      "step": 2918
+    },
+    {
+      "epoch": 2.1972149040270983,
+      "grad_norm": 0.3097464144229889,
+      "learning_rate": 4.31767289835717e-05,
+      "loss": 0.3951,
+      "step": 2919
+    },
+    {
+      "epoch": 2.197967632668423,
+      "grad_norm": 0.2306070774793625,
+      "learning_rate": 4.317139082252935e-05,
+      "loss": 0.3175,
+      "step": 2920
+    },
+    {
+      "epoch": 2.1987203613097477,
+      "grad_norm": 0.2818644940853119,
+      "learning_rate": 4.316605094768726e-05,
+      "loss": 0.3094,
+      "step": 2921
+    },
+    {
+      "epoch": 2.1994730899510726,
+      "grad_norm": 0.28029245138168335,
+      "learning_rate": 4.3160709359629386e-05,
+      "loss": 0.4103,
+      "step": 2922
+    },
+    {
+      "epoch": 2.2002258185923975,
+      "grad_norm": 0.25596460700035095,
+      "learning_rate": 4.315536605893989e-05,
+      "loss": 0.3309,
+      "step": 2923
+    },
+    {
+      "epoch": 2.2009785472337224,
+      "grad_norm": 0.299027681350708,
+      "learning_rate": 4.3150021046203095e-05,
+      "loss": 0.2629,
+      "step": 2924
+    },
+    {
+      "epoch": 2.201731275875047,
+      "grad_norm": 0.28591641783714294,
+      "learning_rate": 4.314467432200353e-05,
+      "loss": 0.3602,
+      "step": 2925
+    },
+    {
+      "epoch": 2.202484004516372,
+      "grad_norm": 0.23459571599960327,
+      "learning_rate": 4.313932588692589e-05,
+      "loss": 0.2315,
+      "step": 2926
+    },
+    {
+      "epoch": 2.2032367331576967,
+      "grad_norm": 0.25278833508491516,
+      "learning_rate": 4.313397574155511e-05,
+      "loss": 0.3605,
+      "step": 2927
+    },
+    {
+      "epoch": 2.203989461799021,
+      "grad_norm": 0.25798362493515015,
+      "learning_rate": 4.3128623886476224e-05,
+      "loss": 0.4224,
+      "step": 2928
+    },
+    {
+      "epoch": 2.204742190440346,
+      "grad_norm": 0.24354977905750275,
+      "learning_rate": 4.3123270322274534e-05,
+      "loss": 0.3553,
+      "step": 2929
+    },
+    {
+      "epoch": 2.205494919081671,
+      "grad_norm": 0.22489775717258453,
+      "learning_rate": 4.311791504953549e-05,
+      "loss": 0.2933,
+      "step": 2930
+    },
+    {
+      "epoch": 2.206247647722996,
+      "grad_norm": 0.230386883020401,
+      "learning_rate": 4.311255806884475e-05,
+      "loss": 0.3023,
+      "step": 2931
+    },
+    {
+      "epoch": 2.207000376364321,
+      "grad_norm": 0.21614402532577515,
+      "learning_rate": 4.310719938078811e-05,
+      "loss": 0.2809,
+      "step": 2932
+    },
+    {
+      "epoch": 2.2077531050056454,
+      "grad_norm": 0.25742220878601074,
+      "learning_rate": 4.3101838985951635e-05,
+      "loss": 0.2872,
+      "step": 2933
+    },
+    {
+      "epoch": 2.2085058336469703,
+      "grad_norm": 0.24426978826522827,
+      "learning_rate": 4.309647688492149e-05,
+      "loss": 0.3641,
+      "step": 2934
+    },
+    {
+      "epoch": 2.209258562288295,
+      "grad_norm": 0.33699896931648254,
+      "learning_rate": 4.309111307828408e-05,
+      "loss": 0.3417,
+      "step": 2935
+    },
+    {
+      "epoch": 2.2100112909296197,
+      "grad_norm": 0.21474701166152954,
+      "learning_rate": 4.308574756662599e-05,
+      "loss": 0.4018,
+      "step": 2936
+    },
+    {
+      "epoch": 2.2107640195709446,
+      "grad_norm": 0.2908034920692444,
+      "learning_rate": 4.308038035053398e-05,
+      "loss": 0.2695,
+      "step": 2937
+    },
+    {
+      "epoch": 2.2115167482122695,
+      "grad_norm": 0.2380618005990982,
+      "learning_rate": 4.3075011430595e-05,
+      "loss": 0.3785,
+      "step": 2938
+    },
+    {
+      "epoch": 2.2122694768535944,
+      "grad_norm": 0.306344211101532,
+      "learning_rate": 4.3069640807396184e-05,
+      "loss": 0.3044,
+      "step": 2939
+    },
+    {
+      "epoch": 2.213022205494919,
+      "grad_norm": 0.2573668956756592,
+      "learning_rate": 4.306426848152486e-05,
+      "loss": 0.3648,
+      "step": 2940
+    },
+    {
+      "epoch": 2.213774934136244,
+      "grad_norm": 0.30854669213294983,
+      "learning_rate": 4.305889445356854e-05,
+      "loss": 0.3125,
+      "step": 2941
+    },
+    {
+      "epoch": 2.2145276627775687,
+      "grad_norm": 0.3531501293182373,
+      "learning_rate": 4.305351872411491e-05,
+      "loss": 0.4501,
+      "step": 2942
+    },
+    {
+      "epoch": 2.2152803914188937,
+      "grad_norm": 0.25418421626091003,
+      "learning_rate": 4.304814129375185e-05,
+      "loss": 0.328,
+      "step": 2943
+    },
+    {
+      "epoch": 2.216033120060218,
+      "grad_norm": 0.29699358344078064,
+      "learning_rate": 4.304276216306744e-05,
+      "loss": 0.3468,
+      "step": 2944
+    },
+    {
+      "epoch": 2.216785848701543,
+      "grad_norm": 0.21802867949008942,
+      "learning_rate": 4.3037381332649934e-05,
+      "loss": 0.3423,
+      "step": 2945
+    },
+    {
+      "epoch": 2.217538577342868,
+      "grad_norm": 0.23188516497612,
+      "learning_rate": 4.303199880308776e-05,
+      "loss": 0.3086,
+      "step": 2946
+    },
+    {
+      "epoch": 2.218291305984193,
+      "grad_norm": 0.2879343330860138,
+      "learning_rate": 4.3026614574969547e-05,
+      "loss": 0.2822,
+      "step": 2947
+    },
+    {
+      "epoch": 2.2190440346255174,
+      "grad_norm": 0.2692761719226837,
+      "learning_rate": 4.302122864888411e-05,
+      "loss": 0.3312,
+      "step": 2948
+    },
+    {
+      "epoch": 2.2197967632668423,
+      "grad_norm": 0.22303175926208496,
+      "learning_rate": 4.301584102542045e-05,
+      "loss": 0.3681,
+      "step": 2949
+    },
+    {
+      "epoch": 2.220549491908167,
+      "grad_norm": 0.2705843448638916,
+      "learning_rate": 4.301045170516773e-05,
+      "loss": 0.341,
+      "step": 2950
+    },
+    {
+      "epoch": 2.221302220549492,
+      "grad_norm": 0.29207631945610046,
+      "learning_rate": 4.3005060688715344e-05,
+      "loss": 0.3632,
+      "step": 2951
+    },
+    {
+      "epoch": 2.2220549491908166,
+      "grad_norm": 0.24983446300029755,
+      "learning_rate": 4.299966797665283e-05,
+      "loss": 0.3619,
+      "step": 2952
+    },
+    {
+      "epoch": 2.2228076778321415,
+      "grad_norm": 0.28683820366859436,
+      "learning_rate": 4.2994273569569926e-05,
+      "loss": 0.3121,
+      "step": 2953
+    },
+    {
+      "epoch": 2.2235604064734664,
+      "grad_norm": 0.26428231596946716,
+      "learning_rate": 4.298887746805656e-05,
+      "loss": 0.3175,
+      "step": 2954
+    },
+    {
+      "epoch": 2.224313135114791,
+      "grad_norm": 0.2529970705509186,
+      "learning_rate": 4.2983479672702835e-05,
+      "loss": 0.3409,
+      "step": 2955
+    },
+    {
+      "epoch": 2.225065863756116,
+      "grad_norm": 0.32026925683021545,
+      "learning_rate": 4.297808018409906e-05,
+      "loss": 0.3118,
+      "step": 2956
+    },
+    {
+      "epoch": 2.2258185923974407,
+      "grad_norm": 0.23574316501617432,
+      "learning_rate": 4.297267900283571e-05,
+      "loss": 0.3393,
+      "step": 2957
+    },
+    {
+      "epoch": 2.2265713210387656,
+      "grad_norm": 0.3141413629055023,
+      "learning_rate": 4.296727612950344e-05,
+      "loss": 0.348,
+      "step": 2958
+    },
+    {
+      "epoch": 2.22732404968009,
+      "grad_norm": 0.236222505569458,
+      "learning_rate": 4.2961871564693115e-05,
+      "loss": 0.2872,
+      "step": 2959
+    },
+    {
+      "epoch": 2.228076778321415,
+      "grad_norm": 0.2775515913963318,
+      "learning_rate": 4.2956465308995754e-05,
+      "loss": 0.2738,
+      "step": 2960
+    },
+    {
+      "epoch": 2.22882950696274,
+      "grad_norm": 0.2574000656604767,
+      "learning_rate": 4.295105736300259e-05,
+      "loss": 0.3628,
+      "step": 2961
+    },
+    {
+      "epoch": 2.229582235604065,
+      "grad_norm": 0.2589409351348877,
+      "learning_rate": 4.2945647727305016e-05,
+      "loss": 0.4293,
+      "step": 2962
+    },
+    {
+      "epoch": 2.2303349642453894,
+      "grad_norm": 0.2626858651638031,
+      "learning_rate": 4.294023640249463e-05,
+      "loss": 0.4086,
+      "step": 2963
+    },
+    {
+      "epoch": 2.2310876928867143,
+      "grad_norm": 0.281058132648468,
+      "learning_rate": 4.293482338916321e-05,
+      "loss": 0.3844,
+      "step": 2964
+    },
+    {
+      "epoch": 2.231840421528039,
+      "grad_norm": 0.23301038146018982,
+      "learning_rate": 4.2929408687902715e-05,
+      "loss": 0.2167,
+      "step": 2965
+    },
+    {
+      "epoch": 2.232593150169364,
+      "grad_norm": 0.24353772401809692,
+      "learning_rate": 4.2923992299305274e-05,
+      "loss": 0.3739,
+      "step": 2966
+    },
+    {
+      "epoch": 2.2333458788106886,
+      "grad_norm": 0.21337196230888367,
+      "learning_rate": 4.291857422396323e-05,
+      "loss": 0.2582,
+      "step": 2967
+    },
+    {
+      "epoch": 2.2340986074520135,
+      "grad_norm": 0.2766335904598236,
+      "learning_rate": 4.291315446246909e-05,
+      "loss": 0.3351,
+      "step": 2968
+    },
+    {
+      "epoch": 2.2348513360933384,
+      "grad_norm": 0.2539624869823456,
+      "learning_rate": 4.290773301541555e-05,
+      "loss": 0.2799,
+      "step": 2969
+    },
+    {
+      "epoch": 2.2356040647346633,
+      "grad_norm": 0.22498565912246704,
+      "learning_rate": 4.29023098833955e-05,
+      "loss": 0.3469,
+      "step": 2970
+    },
+    {
+      "epoch": 2.236356793375988,
+      "grad_norm": 0.2484288215637207,
+      "learning_rate": 4.289688506700199e-05,
+      "loss": 0.348,
+      "step": 2971
+    },
+    {
+      "epoch": 2.2371095220173127,
+      "grad_norm": 0.23865807056427002,
+      "learning_rate": 4.289145856682828e-05,
+      "loss": 0.356,
+      "step": 2972
+    },
+    {
+      "epoch": 2.2378622506586376,
+      "grad_norm": 0.23613299429416656,
+      "learning_rate": 4.28860303834678e-05,
+      "loss": 0.3018,
+      "step": 2973
+    },
+    {
+      "epoch": 2.2386149792999626,
+      "grad_norm": 0.24200153350830078,
+      "learning_rate": 4.288060051751417e-05,
+      "loss": 0.3947,
+      "step": 2974
+    },
+    {
+      "epoch": 2.239367707941287,
+      "grad_norm": 0.2619052827358246,
+      "learning_rate": 4.2875168969561194e-05,
+      "loss": 0.4011,
+      "step": 2975
+    },
+    {
+      "epoch": 2.240120436582612,
+      "grad_norm": 0.22978299856185913,
+      "learning_rate": 4.286973574020285e-05,
+      "loss": 0.2884,
+      "step": 2976
+    },
+    {
+      "epoch": 2.240873165223937,
+      "grad_norm": 0.2045583426952362,
+      "learning_rate": 4.2864300830033324e-05,
+      "loss": 0.2878,
+      "step": 2977
+    },
+    {
+      "epoch": 2.241625893865262,
+      "grad_norm": 0.2780607342720032,
+      "learning_rate": 4.285886423964695e-05,
+      "loss": 0.363,
+      "step": 2978
+    },
+    {
+      "epoch": 2.2423786225065863,
+      "grad_norm": 0.26181504130363464,
+      "learning_rate": 4.285342596963828e-05,
+      "loss": 0.3447,
+      "step": 2979
+    },
+    {
+      "epoch": 2.243131351147911,
+      "grad_norm": 0.22585821151733398,
+      "learning_rate": 4.2847986020602025e-05,
+      "loss": 0.3104,
+      "step": 2980
+    },
+    {
+      "epoch": 2.243884079789236,
+      "grad_norm": 0.21110042929649353,
+      "learning_rate": 4.284254439313309e-05,
+      "loss": 0.3821,
+      "step": 2981
+    },
+    {
+      "epoch": 2.2446368084305606,
+      "grad_norm": 0.21407635509967804,
+      "learning_rate": 4.2837101087826576e-05,
+      "loss": 0.2592,
+      "step": 2982
+    },
+    {
+      "epoch": 2.2453895370718855,
+      "grad_norm": 0.2394321858882904,
+      "learning_rate": 4.2831656105277744e-05,
+      "loss": 0.2569,
+      "step": 2983
+    },
+    {
+      "epoch": 2.2461422657132104,
+      "grad_norm": 0.2623378038406372,
+      "learning_rate": 4.282620944608205e-05,
+      "loss": 0.2718,
+      "step": 2984
+    },
+    {
+      "epoch": 2.2468949943545353,
+      "grad_norm": 0.24720843136310577,
+      "learning_rate": 4.282076111083514e-05,
+      "loss": 0.3783,
+      "step": 2985
+    },
+    {
+      "epoch": 2.24764772299586,
+      "grad_norm": 0.22729666531085968,
+      "learning_rate": 4.281531110013282e-05,
+      "loss": 0.3291,
+      "step": 2986
+    },
+    {
+      "epoch": 2.2484004516371847,
+      "grad_norm": 0.25572770833969116,
+      "learning_rate": 4.2809859414571114e-05,
+      "loss": 0.3156,
+      "step": 2987
+    },
+    {
+      "epoch": 2.2491531802785096,
+      "grad_norm": 0.2719089388847351,
+      "learning_rate": 4.280440605474619e-05,
+      "loss": 0.3099,
+      "step": 2988
+    },
+    {
+      "epoch": 2.2499059089198346,
+      "grad_norm": 0.23800991475582123,
+      "learning_rate": 4.2798951021254436e-05,
+      "loss": 0.2533,
+      "step": 2989
+    },
+    {
+      "epoch": 2.250658637561159,
+      "grad_norm": 0.3018708825111389,
+      "learning_rate": 4.27934943146924e-05,
+      "loss": 0.3725,
+      "step": 2990
+    },
+    {
+      "epoch": 2.251411366202484,
+      "grad_norm": 0.2286914736032486,
+      "learning_rate": 4.278803593565682e-05,
+      "loss": 0.3719,
+      "step": 2991
+    },
+    {
+      "epoch": 2.252164094843809,
+      "grad_norm": 0.2652614414691925,
+      "learning_rate": 4.27825758847446e-05,
+      "loss": 0.2735,
+      "step": 2992
+    },
+    {
+      "epoch": 2.252916823485134,
+      "grad_norm": 0.22484079003334045,
+      "learning_rate": 4.277711416255288e-05,
+      "loss": 0.3372,
+      "step": 2993
+    },
+    {
+      "epoch": 2.2536695521264583,
+      "grad_norm": 0.223599374294281,
+      "learning_rate": 4.277165076967891e-05,
+      "loss": 0.3349,
+      "step": 2994
+    },
+    {
+      "epoch": 2.254422280767783,
+      "grad_norm": 0.2737419605255127,
+      "learning_rate": 4.2766185706720174e-05,
+      "loss": 0.387,
+      "step": 2995
+    },
+    {
+      "epoch": 2.255175009409108,
+      "grad_norm": 0.25657084584236145,
+      "learning_rate": 4.2760718974274326e-05,
+      "loss": 0.3679,
+      "step": 2996
+    },
+    {
+      "epoch": 2.255927738050433,
+      "grad_norm": 0.21851791441440582,
+      "learning_rate": 4.27552505729392e-05,
+      "loss": 0.3253,
+      "step": 2997
+    },
+    {
+      "epoch": 2.2566804666917575,
+      "grad_norm": 0.24634405970573425,
+      "learning_rate": 4.27497805033128e-05,
+      "loss": 0.3277,
+      "step": 2998
+    },
+    {
+      "epoch": 2.2574331953330824,
+      "grad_norm": 0.2602645456790924,
+      "learning_rate": 4.2744308765993326e-05,
+      "loss": 0.3363,
+      "step": 2999
+    },
+    {
+      "epoch": 2.2581859239744073,
+      "grad_norm": 0.23306314647197723,
+      "learning_rate": 4.2738835361579175e-05,
+      "loss": 0.3621,
+      "step": 3000
+    },
+    {
+      "epoch": 2.2581859239744073,
+      "eval_loss": 0.32739749550819397,
+      "eval_runtime": 452.6492,
+      "eval_samples_per_second": 21.268,
+      "eval_steps_per_second": 0.665,
+      "step": 3000
+    },
+    {
+      "epoch": 2.258938652615732,
+      "grad_norm": 0.20242488384246826,
+      "learning_rate": 4.2733360290668894e-05,
+      "loss": 0.3158,
+      "step": 3001
+    },
+    {
+      "epoch": 2.2596913812570567,
+      "grad_norm": 0.22816257178783417,
+      "learning_rate": 4.272788355386124e-05,
+      "loss": 0.3841,
+      "step": 3002
+    },
+    {
+      "epoch": 2.2604441098983816,
+      "grad_norm": 0.25860634446144104,
+      "learning_rate": 4.272240515175514e-05,
+      "loss": 0.3578,
+      "step": 3003
+    },
+    {
+      "epoch": 2.2611968385397065,
+      "grad_norm": 0.25685566663742065,
+      "learning_rate": 4.2716925084949686e-05,
+      "loss": 0.2971,
+      "step": 3004
+    },
+    {
+      "epoch": 2.2619495671810315,
+      "grad_norm": 0.24581988155841827,
+      "learning_rate": 4.271144335404419e-05,
+      "loss": 0.333,
+      "step": 3005
+    },
+    {
+      "epoch": 2.262702295822356,
+      "grad_norm": 0.24747689068317413,
+      "learning_rate": 4.270595995963812e-05,
+      "loss": 0.2722,
+      "step": 3006
+    },
+    {
+      "epoch": 2.263455024463681,
+      "grad_norm": 0.22951442003250122,
+      "learning_rate": 4.2700474902331125e-05,
+      "loss": 0.2161,
+      "step": 3007
+    },
+    {
+      "epoch": 2.2642077531050058,
+      "grad_norm": 0.2235272079706192,
+      "learning_rate": 4.269498818272305e-05,
+      "loss": 0.3545,
+      "step": 3008
+    },
+    {
+      "epoch": 2.2649604817463302,
+      "grad_norm": 0.32005494832992554,
+      "learning_rate": 4.268949980141392e-05,
+      "loss": 0.3605,
+      "step": 3009
+    },
+    {
+      "epoch": 2.265713210387655,
+      "grad_norm": 0.29939424991607666,
+      "learning_rate": 4.268400975900392e-05,
+      "loss": 0.2962,
+      "step": 3010
+    },
+    {
+      "epoch": 2.26646593902898,
+      "grad_norm": 0.30765214562416077,
+      "learning_rate": 4.267851805609344e-05,
+      "loss": 0.3799,
+      "step": 3011
+    },
+    {
+      "epoch": 2.267218667670305,
+      "grad_norm": 0.23683764040470123,
+      "learning_rate": 4.267302469328305e-05,
+      "loss": 0.2006,
+      "step": 3012
+    },
+    {
+      "epoch": 2.2679713963116295,
+      "grad_norm": 0.381918728351593,
+      "learning_rate": 4.2667529671173495e-05,
+      "loss": 0.3519,
+      "step": 3013
+    },
+    {
+      "epoch": 2.2687241249529544,
+      "grad_norm": 0.29179614782333374,
+      "learning_rate": 4.26620329903657e-05,
+      "loss": 0.3805,
+      "step": 3014
+    },
+    {
+      "epoch": 2.2694768535942793,
+      "grad_norm": 0.25847098231315613,
+      "learning_rate": 4.2656534651460765e-05,
+      "loss": 0.3784,
+      "step": 3015
+    },
+    {
+      "epoch": 2.2702295822356042,
+      "grad_norm": 0.2615436911582947,
+      "learning_rate": 4.265103465506e-05,
+      "loss": 0.2428,
+      "step": 3016
+    },
+    {
+      "epoch": 2.2709823108769287,
+      "grad_norm": 0.2500426471233368,
+      "learning_rate": 4.264553300176485e-05,
+      "loss": 0.2184,
+      "step": 3017
+    },
+    {
+      "epoch": 2.2717350395182536,
+      "grad_norm": 0.29261577129364014,
+      "learning_rate": 4.264002969217699e-05,
+      "loss": 0.2777,
+      "step": 3018
+    },
+    {
+      "epoch": 2.2724877681595785,
+      "grad_norm": 0.2849266529083252,
+      "learning_rate": 4.263452472689825e-05,
+      "loss": 0.2871,
+      "step": 3019
+    },
+    {
+      "epoch": 2.2732404968009035,
+      "grad_norm": 0.26664355397224426,
+      "learning_rate": 4.2629018106530626e-05,
+      "loss": 0.321,
+      "step": 3020
+    },
+    {
+      "epoch": 2.273993225442228,
+      "grad_norm": 0.30110591650009155,
+      "learning_rate": 4.262350983167634e-05,
+      "loss": 0.3763,
+      "step": 3021
+    },
+    {
+      "epoch": 2.274745954083553,
+      "grad_norm": 0.2907387614250183,
+      "learning_rate": 4.2617999902937753e-05,
+      "loss": 0.3628,
+      "step": 3022
+    },
+    {
+      "epoch": 2.2754986827248778,
+      "grad_norm": 0.24361242353916168,
+      "learning_rate": 4.261248832091743e-05,
+      "loss": 0.3684,
+      "step": 3023
+    },
+    {
+      "epoch": 2.2762514113662027,
+      "grad_norm": 0.3154049813747406,
+      "learning_rate": 4.260697508621811e-05,
+      "loss": 0.3132,
+      "step": 3024
+    },
+    {
+      "epoch": 2.277004140007527,
+      "grad_norm": 0.25080323219299316,
+      "learning_rate": 4.2601460199442695e-05,
+      "loss": 0.2917,
+      "step": 3025
+    },
+    {
+      "epoch": 2.277756868648852,
+      "grad_norm": 0.2894185483455658,
+      "learning_rate": 4.25959436611943e-05,
+      "loss": 0.2845,
+      "step": 3026
+    },
+    {
+      "epoch": 2.278509597290177,
+      "grad_norm": 0.34587323665618896,
+      "learning_rate": 4.25904254720762e-05,
+      "loss": 0.3367,
+      "step": 3027
+    },
+    {
+      "epoch": 2.2792623259315015,
+      "grad_norm": 0.26134321093559265,
+      "learning_rate": 4.258490563269187e-05,
+      "loss": 0.2461,
+      "step": 3028
+    },
+    {
+      "epoch": 2.2800150545728264,
+      "grad_norm": 0.2658021152019501,
+      "learning_rate": 4.2579384143644927e-05,
+      "loss": 0.2524,
+      "step": 3029
+    },
+    {
+      "epoch": 2.2807677832141513,
+      "grad_norm": 0.32299289107322693,
+      "learning_rate": 4.257386100553921e-05,
+      "loss": 0.3879,
+      "step": 3030
+    },
+    {
+      "epoch": 2.281520511855476,
+      "grad_norm": 0.2481514811515808,
+      "learning_rate": 4.256833621897872e-05,
+      "loss": 0.2952,
+      "step": 3031
+    },
+    {
+      "epoch": 2.282273240496801,
+      "grad_norm": 0.34030941128730774,
+      "learning_rate": 4.256280978456764e-05,
+      "loss": 0.2739,
+      "step": 3032
+    },
+    {
+      "epoch": 2.2830259691381256,
+      "grad_norm": 0.25724542140960693,
+      "learning_rate": 4.2557281702910326e-05,
+      "loss": 0.4068,
+      "step": 3033
+    },
+    {
+      "epoch": 2.2837786977794505,
+      "grad_norm": 0.26574093103408813,
+      "learning_rate": 4.2551751974611316e-05,
+      "loss": 0.2868,
+      "step": 3034
+    },
+    {
+      "epoch": 2.2845314264207754,
+      "grad_norm": 0.2925585210323334,
+      "learning_rate": 4.254622060027535e-05,
+      "loss": 0.2891,
+      "step": 3035
+    },
+    {
+      "epoch": 2.2852841550621,
+      "grad_norm": 0.254281222820282,
+      "learning_rate": 4.2540687580507304e-05,
+      "loss": 0.3034,
+      "step": 3036
+    },
+    {
+      "epoch": 2.286036883703425,
+      "grad_norm": 0.27471408247947693,
+      "learning_rate": 4.253515291591229e-05,
+      "loss": 0.3035,
+      "step": 3037
+    },
+    {
+      "epoch": 2.2867896123447498,
+      "grad_norm": 0.2550313174724579,
+      "learning_rate": 4.2529616607095555e-05,
+      "loss": 0.2905,
+      "step": 3038
+    },
+    {
+      "epoch": 2.2875423409860747,
+      "grad_norm": 0.2708241641521454,
+      "learning_rate": 4.2524078654662544e-05,
+      "loss": 0.364,
+      "step": 3039
+    },
+    {
+      "epoch": 2.288295069627399,
+      "grad_norm": 0.26389050483703613,
+      "learning_rate": 4.251853905921888e-05,
+      "loss": 0.3222,
+      "step": 3040
+    },
+    {
+      "epoch": 2.289047798268724,
+      "grad_norm": 0.2524468004703522,
+      "learning_rate": 4.251299782137037e-05,
+      "loss": 0.3789,
+      "step": 3041
+    },
+    {
+      "epoch": 2.289800526910049,
+      "grad_norm": 0.29753702878952026,
+      "learning_rate": 4.250745494172299e-05,
+      "loss": 0.3273,
+      "step": 3042
+    },
+    {
+      "epoch": 2.290553255551374,
+      "grad_norm": 0.2450147122144699,
+      "learning_rate": 4.2501910420882893e-05,
+      "loss": 0.301,
+      "step": 3043
+    },
+    {
+      "epoch": 2.2913059841926984,
+      "grad_norm": 0.2778531610965729,
+      "learning_rate": 4.2496364259456434e-05,
+      "loss": 0.3545,
+      "step": 3044
+    },
+    {
+      "epoch": 2.2920587128340233,
+      "grad_norm": 0.2792007327079773,
+      "learning_rate": 4.249081645805012e-05,
+      "loss": 0.2112,
+      "step": 3045
+    },
+    {
+      "epoch": 2.292811441475348,
+      "grad_norm": 0.24656659364700317,
+      "learning_rate": 4.2485267017270664e-05,
+      "loss": 0.3842,
+      "step": 3046
+    },
+    {
+      "epoch": 2.293564170116673,
+      "grad_norm": 0.25025859475135803,
+      "learning_rate": 4.2479715937724935e-05,
+      "loss": 0.3465,
+      "step": 3047
+    },
+    {
+      "epoch": 2.2943168987579976,
+      "grad_norm": 0.28225088119506836,
+      "learning_rate": 4.2474163220019986e-05,
+      "loss": 0.3292,
+      "step": 3048
+    },
+    {
+      "epoch": 2.2950696273993225,
+      "grad_norm": 0.2771015763282776,
+      "learning_rate": 4.2468608864763065e-05,
+      "loss": 0.3984,
+      "step": 3049
+    },
+    {
+      "epoch": 2.2958223560406474,
+      "grad_norm": 0.2394844889640808,
+      "learning_rate": 4.2463052872561584e-05,
+      "loss": 0.293,
+      "step": 3050
+    },
+    {
+      "epoch": 2.2965750846819724,
+      "grad_norm": 0.27458515763282776,
+      "learning_rate": 4.245749524402314e-05,
+      "loss": 0.3169,
+      "step": 3051
+    },
+    {
+      "epoch": 2.297327813323297,
+      "grad_norm": 0.23067070543766022,
+      "learning_rate": 4.24519359797555e-05,
+      "loss": 0.3532,
+      "step": 3052
+    },
+    {
+      "epoch": 2.2980805419646217,
+      "grad_norm": 0.21505315601825714,
+      "learning_rate": 4.244637508036662e-05,
+      "loss": 0.3374,
+      "step": 3053
+    },
+    {
+      "epoch": 2.2988332706059467,
+      "grad_norm": 0.25453248620033264,
+      "learning_rate": 4.244081254646464e-05,
+      "loss": 0.2987,
+      "step": 3054
+    },
+    {
+      "epoch": 2.299585999247271,
+      "grad_norm": 0.24662595987319946,
+      "learning_rate": 4.243524837865787e-05,
+      "loss": 0.3074,
+      "step": 3055
+    },
+    {
+      "epoch": 2.300338727888596,
+      "grad_norm": 0.21292367577552795,
+      "learning_rate": 4.242968257755477e-05,
+      "loss": 0.3191,
+      "step": 3056
+    },
+    {
+      "epoch": 2.301091456529921,
+      "grad_norm": 0.2352483719587326,
+      "learning_rate": 4.242411514376404e-05,
+      "loss": 0.3218,
+      "step": 3057
+    },
+    {
+      "epoch": 2.301844185171246,
+      "grad_norm": 0.27879223227500916,
+      "learning_rate": 4.2418546077894525e-05,
+      "loss": 0.324,
+      "step": 3058
+    },
+    {
+      "epoch": 2.302596913812571,
+      "grad_norm": 0.24772599339485168,
+      "learning_rate": 4.241297538055523e-05,
+      "loss": 0.2621,
+      "step": 3059
+    },
+    {
+      "epoch": 2.3033496424538953,
+      "grad_norm": 0.26267412304878235,
+      "learning_rate": 4.240740305235537e-05,
+      "loss": 0.4105,
+      "step": 3060
+    },
+    {
+      "epoch": 2.30410237109522,
+      "grad_norm": 0.2629970610141754,
+      "learning_rate": 4.2401829093904326e-05,
+      "loss": 0.2125,
+      "step": 3061
+    },
+    {
+      "epoch": 2.304855099736545,
+      "grad_norm": 0.23332351446151733,
+      "learning_rate": 4.2396253505811665e-05,
+      "loss": 0.3238,
+      "step": 3062
+    },
+    {
+      "epoch": 2.3056078283778696,
+      "grad_norm": 0.26663368940353394,
+      "learning_rate": 4.23906762886871e-05,
+      "loss": 0.3175,
+      "step": 3063
+    },
+    {
+      "epoch": 2.3063605570191945,
+      "grad_norm": 0.23615744709968567,
+      "learning_rate": 4.238509744314057e-05,
+      "loss": 0.3086,
+      "step": 3064
+    },
+    {
+      "epoch": 2.3071132856605194,
+      "grad_norm": 0.25053855776786804,
+      "learning_rate": 4.237951696978217e-05,
+      "loss": 0.3389,
+      "step": 3065
+    },
+    {
+      "epoch": 2.3078660143018443,
+      "grad_norm": 0.2631237506866455,
+      "learning_rate": 4.237393486922216e-05,
+      "loss": 0.3436,
+      "step": 3066
+    },
+    {
+      "epoch": 2.308618742943169,
+      "grad_norm": 0.3060251474380493,
+      "learning_rate": 4.2368351142071e-05,
+      "loss": 0.2396,
+      "step": 3067
+    },
+    {
+      "epoch": 2.3093714715844937,
+      "grad_norm": 0.2868037521839142,
+      "learning_rate": 4.2362765788939306e-05,
+      "loss": 0.3816,
+      "step": 3068
+    },
+    {
+      "epoch": 2.3101242002258187,
+      "grad_norm": 0.2642500698566437,
+      "learning_rate": 4.23571788104379e-05,
+      "loss": 0.3805,
+      "step": 3069
+    },
+    {
+      "epoch": 2.3108769288671436,
+      "grad_norm": 0.2822163999080658,
+      "learning_rate": 4.2351590207177756e-05,
+      "loss": 0.3029,
+      "step": 3070
+    },
+    {
+      "epoch": 2.311629657508468,
+      "grad_norm": 0.23417970538139343,
+      "learning_rate": 4.2345999979770025e-05,
+      "loss": 0.3031,
+      "step": 3071
+    },
+    {
+      "epoch": 2.312382386149793,
+      "grad_norm": 0.28136029839515686,
+      "learning_rate": 4.234040812882606e-05,
+      "loss": 0.3378,
+      "step": 3072
+    },
+    {
+      "epoch": 2.313135114791118,
+      "grad_norm": 0.2812299132347107,
+      "learning_rate": 4.2334814654957376e-05,
+      "loss": 0.3087,
+      "step": 3073
+    },
+    {
+      "epoch": 2.3138878434324424,
+      "grad_norm": 0.24952401220798492,
+      "learning_rate": 4.232921955877567e-05,
+      "loss": 0.2696,
+      "step": 3074
+    },
+    {
+      "epoch": 2.3146405720737673,
+      "grad_norm": 0.27567771077156067,
+      "learning_rate": 4.23236228408928e-05,
+      "loss": 0.4155,
+      "step": 3075
+    },
+    {
+      "epoch": 2.315393300715092,
+      "grad_norm": 0.24624356627464294,
+      "learning_rate": 4.2318024501920834e-05,
+      "loss": 0.3153,
+      "step": 3076
+    },
+    {
+      "epoch": 2.316146029356417,
+      "grad_norm": 0.2371608316898346,
+      "learning_rate": 4.231242454247198e-05,
+      "loss": 0.3051,
+      "step": 3077
+    },
+    {
+      "epoch": 2.316898757997742,
+      "grad_norm": 0.33288657665252686,
+      "learning_rate": 4.230682296315865e-05,
+      "loss": 0.4104,
+      "step": 3078
+    },
+    {
+      "epoch": 2.3176514866390665,
+      "grad_norm": 0.3132668137550354,
+      "learning_rate": 4.230121976459342e-05,
+      "loss": 0.3575,
+      "step": 3079
+    },
+    {
+      "epoch": 2.3184042152803914,
+      "grad_norm": 0.3036741614341736,
+      "learning_rate": 4.229561494738906e-05,
+      "loss": 0.3587,
+      "step": 3080
+    },
+    {
+      "epoch": 2.3191569439217163,
+      "grad_norm": 0.29431086778640747,
+      "learning_rate": 4.229000851215848e-05,
+      "loss": 0.3839,
+      "step": 3081
+    },
+    {
+      "epoch": 2.319909672563041,
+      "grad_norm": 0.2460717111825943,
+      "learning_rate": 4.228440045951481e-05,
+      "loss": 0.2579,
+      "step": 3082
+    },
+    {
+      "epoch": 2.3206624012043657,
+      "grad_norm": 0.2666926681995392,
+      "learning_rate": 4.2278790790071335e-05,
+      "loss": 0.3336,
+      "step": 3083
+    },
+    {
+      "epoch": 2.3214151298456907,
+      "grad_norm": 0.24027176201343536,
+      "learning_rate": 4.227317950444152e-05,
+      "loss": 0.3279,
+      "step": 3084
+    },
+    {
+      "epoch": 2.3221678584870156,
+      "grad_norm": 0.2813846170902252,
+      "learning_rate": 4.2267566603239005e-05,
+      "loss": 0.3864,
+      "step": 3085
+    },
+    {
+      "epoch": 2.32292058712834,
+      "grad_norm": 0.2710760533809662,
+      "learning_rate": 4.226195208707761e-05,
+      "loss": 0.268,
+      "step": 3086
+    },
+    {
+      "epoch": 2.323673315769665,
+      "grad_norm": 0.21816232800483704,
+      "learning_rate": 4.225633595657133e-05,
+      "loss": 0.4051,
+      "step": 3087
+    },
+    {
+      "epoch": 2.32442604441099,
+      "grad_norm": 0.21321235597133636,
+      "learning_rate": 4.2250718212334336e-05,
+      "loss": 0.2974,
+      "step": 3088
+    },
+    {
+      "epoch": 2.325178773052315,
+      "grad_norm": 0.261374294757843,
+      "learning_rate": 4.224509885498097e-05,
+      "loss": 0.3464,
+      "step": 3089
+    },
+    {
+      "epoch": 2.3259315016936393,
+      "grad_norm": 0.2646394968032837,
+      "learning_rate": 4.223947788512576e-05,
+      "loss": 0.4077,
+      "step": 3090
+    },
+    {
+      "epoch": 2.326684230334964,
+      "grad_norm": 0.2935434579849243,
+      "learning_rate": 4.2233855303383416e-05,
+      "loss": 0.2697,
+      "step": 3091
+    },
+    {
+      "epoch": 2.327436958976289,
+      "grad_norm": 0.2512994706630707,
+      "learning_rate": 4.222823111036881e-05,
+      "loss": 0.3708,
+      "step": 3092
+    },
+    {
+      "epoch": 2.328189687617614,
+      "grad_norm": 0.31740280985832214,
+      "learning_rate": 4.2222605306696995e-05,
+      "loss": 0.328,
+      "step": 3093
+    },
+    {
+      "epoch": 2.3289424162589385,
+      "grad_norm": 0.2675962448120117,
+      "learning_rate": 4.2216977892983187e-05,
+      "loss": 0.4305,
+      "step": 3094
+    },
+    {
+      "epoch": 2.3296951449002634,
+      "grad_norm": 0.3167666792869568,
+      "learning_rate": 4.221134886984281e-05,
+      "loss": 0.2652,
+      "step": 3095
+    },
+    {
+      "epoch": 2.3304478735415883,
+      "grad_norm": 0.3080410063266754,
+      "learning_rate": 4.220571823789144e-05,
+      "loss": 0.3353,
+      "step": 3096
+    },
+    {
+      "epoch": 2.3312006021829133,
+      "grad_norm": 0.2578239440917969,
+      "learning_rate": 4.220008599774482e-05,
+      "loss": 0.2845,
+      "step": 3097
+    },
+    {
+      "epoch": 2.3319533308242377,
+      "grad_norm": 0.3555237352848053,
+      "learning_rate": 4.219445215001892e-05,
+      "loss": 0.3415,
+      "step": 3098
+    },
+    {
+      "epoch": 2.3327060594655626,
+      "grad_norm": 0.23242589831352234,
+      "learning_rate": 4.2188816695329816e-05,
+      "loss": 0.2723,
+      "step": 3099
+    },
+    {
+      "epoch": 2.3334587881068876,
+      "grad_norm": 0.28230777382850647,
+      "learning_rate": 4.2183179634293794e-05,
+      "loss": 0.3202,
+      "step": 3100
+    },
+    {
+      "epoch": 2.334211516748212,
+      "grad_norm": 0.21596476435661316,
+      "learning_rate": 4.217754096752733e-05,
+      "loss": 0.3067,
+      "step": 3101
+    },
+    {
+      "epoch": 2.334964245389537,
+      "grad_norm": 0.2609899044036865,
+      "learning_rate": 4.217190069564706e-05,
+      "loss": 0.3763,
+      "step": 3102
+    },
+    {
+      "epoch": 2.335716974030862,
+      "grad_norm": 0.23115907609462738,
+      "learning_rate": 4.216625881926978e-05,
+      "loss": 0.3249,
+      "step": 3103
+    },
+    {
+      "epoch": 2.336469702672187,
+      "grad_norm": 0.220550075173378,
+      "learning_rate": 4.216061533901248e-05,
+      "loss": 0.2568,
+      "step": 3104
+    },
+    {
+      "epoch": 2.3372224313135117,
+      "grad_norm": 0.23805654048919678,
+      "learning_rate": 4.215497025549234e-05,
+      "loss": 0.2743,
+      "step": 3105
+    },
+    {
+      "epoch": 2.337975159954836,
+      "grad_norm": 0.24250689148902893,
+      "learning_rate": 4.214932356932668e-05,
+      "loss": 0.3569,
+      "step": 3106
+    },
+    {
+      "epoch": 2.338727888596161,
+      "grad_norm": 0.242852583527565,
+      "learning_rate": 4.214367528113302e-05,
+      "loss": 0.361,
+      "step": 3107
+    },
+    {
+      "epoch": 2.339480617237486,
+      "grad_norm": 0.28402554988861084,
+      "learning_rate": 4.213802539152905e-05,
+      "loss": 0.3353,
+      "step": 3108
+    },
+    {
+      "epoch": 2.3402333458788105,
+      "grad_norm": 0.23511642217636108,
+      "learning_rate": 4.213237390113263e-05,
+      "loss": 0.2506,
+      "step": 3109
+    },
+    {
+      "epoch": 2.3409860745201354,
+      "grad_norm": 0.2352583259344101,
+      "learning_rate": 4.2126720810561795e-05,
+      "loss": 0.3319,
+      "step": 3110
+    },
+    {
+      "epoch": 2.3417388031614603,
+      "grad_norm": 0.23784297704696655,
+      "learning_rate": 4.212106612043477e-05,
+      "loss": 0.3631,
+      "step": 3111
+    },
+    {
+      "epoch": 2.3424915318027852,
+      "grad_norm": 0.283479779958725,
+      "learning_rate": 4.211540983136994e-05,
+      "loss": 0.308,
+      "step": 3112
+    },
+    {
+      "epoch": 2.3432442604441097,
+      "grad_norm": 0.3346084952354431,
+      "learning_rate": 4.210975194398585e-05,
+      "loss": 0.2888,
+      "step": 3113
+    },
+    {
+      "epoch": 2.3439969890854346,
+      "grad_norm": 0.24614332616329193,
+      "learning_rate": 4.210409245890127e-05,
+      "loss": 0.3834,
+      "step": 3114
+    },
+    {
+      "epoch": 2.3447497177267596,
+      "grad_norm": 0.292072057723999,
+      "learning_rate": 4.209843137673509e-05,
+      "loss": 0.3325,
+      "step": 3115
+    },
+    {
+      "epoch": 2.3455024463680845,
+      "grad_norm": 0.3260290324687958,
+      "learning_rate": 4.209276869810639e-05,
+      "loss": 0.326,
+      "step": 3116
+    },
+    {
+      "epoch": 2.346255175009409,
+      "grad_norm": 0.24989046156406403,
+      "learning_rate": 4.208710442363446e-05,
+      "loss": 0.3246,
+      "step": 3117
+    },
+    {
+      "epoch": 2.347007903650734,
+      "grad_norm": 0.2712777554988861,
+      "learning_rate": 4.208143855393872e-05,
+      "loss": 0.4036,
+      "step": 3118
+    },
+    {
+      "epoch": 2.347760632292059,
+      "grad_norm": 0.30657777190208435,
+      "learning_rate": 4.2075771089638775e-05,
+      "loss": 0.3197,
+      "step": 3119
+    },
+    {
+      "epoch": 2.3485133609333837,
+      "grad_norm": 0.2338680624961853,
+      "learning_rate": 4.207010203135442e-05,
+      "loss": 0.4699,
+      "step": 3120
+    },
+    {
+      "epoch": 2.349266089574708,
+      "grad_norm": 0.2519816756248474,
+      "learning_rate": 4.206443137970562e-05,
+      "loss": 0.322,
+      "step": 3121
+    },
+    {
+      "epoch": 2.350018818216033,
+      "grad_norm": 0.2496872842311859,
+      "learning_rate": 4.20587591353125e-05,
+      "loss": 0.318,
+      "step": 3122
+    },
+    {
+      "epoch": 2.350771546857358,
+      "grad_norm": 0.25669655203819275,
+      "learning_rate": 4.205308529879536e-05,
+      "loss": 0.3569,
+      "step": 3123
+    },
+    {
+      "epoch": 2.351524275498683,
+      "grad_norm": 0.3327885568141937,
+      "learning_rate": 4.204740987077471e-05,
+      "loss": 0.3009,
+      "step": 3124
+    },
+    {
+      "epoch": 2.3522770041400074,
+      "grad_norm": 0.27954909205436707,
+      "learning_rate": 4.204173285187117e-05,
+      "loss": 0.2987,
+      "step": 3125
+    },
+    {
+      "epoch": 2.3530297327813323,
+      "grad_norm": 0.2652660310268402,
+      "learning_rate": 4.20360542427056e-05,
+      "loss": 0.254,
+      "step": 3126
+    },
+    {
+      "epoch": 2.3537824614226572,
+      "grad_norm": 0.2905503809452057,
+      "learning_rate": 4.2030374043898986e-05,
+      "loss": 0.3038,
+      "step": 3127
+    },
+    {
+      "epoch": 2.3545351900639817,
+      "grad_norm": 0.22478114068508148,
+      "learning_rate": 4.2024692256072516e-05,
+      "loss": 0.2617,
+      "step": 3128
+    },
+    {
+      "epoch": 2.3552879187053066,
+      "grad_norm": 0.2483111023902893,
+      "learning_rate": 4.201900887984754e-05,
+      "loss": 0.335,
+      "step": 3129
+    },
+    {
+      "epoch": 2.3560406473466315,
+      "grad_norm": 0.28553682565689087,
+      "learning_rate": 4.2013323915845575e-05,
+      "loss": 0.4088,
+      "step": 3130
+    },
+    {
+      "epoch": 2.3567933759879565,
+      "grad_norm": 0.25947245955467224,
+      "learning_rate": 4.200763736468834e-05,
+      "loss": 0.3553,
+      "step": 3131
+    },
+    {
+      "epoch": 2.3575461046292814,
+      "grad_norm": 0.33118462562561035,
+      "learning_rate": 4.2001949226997683e-05,
+      "loss": 0.3088,
+      "step": 3132
+    },
+    {
+      "epoch": 2.358298833270606,
+      "grad_norm": 0.2374979555606842,
+      "learning_rate": 4.1996259503395656e-05,
+      "loss": 0.4123,
+      "step": 3133
+    },
+    {
+      "epoch": 2.3590515619119308,
+      "grad_norm": 0.2525404095649719,
+      "learning_rate": 4.1990568194504504e-05,
+      "loss": 0.3057,
+      "step": 3134
+    },
+    {
+      "epoch": 2.3598042905532557,
+      "grad_norm": 0.2590644359588623,
+      "learning_rate": 4.198487530094658e-05,
+      "loss": 0.3813,
+      "step": 3135
+    },
+    {
+      "epoch": 2.36055701919458,
+      "grad_norm": 0.2250070720911026,
+      "learning_rate": 4.197918082334448e-05,
+      "loss": 0.2836,
+      "step": 3136
+    },
+    {
+      "epoch": 2.361309747835905,
+      "grad_norm": 0.2238345742225647,
+      "learning_rate": 4.1973484762320935e-05,
+      "loss": 0.2953,
+      "step": 3137
+    },
+    {
+      "epoch": 2.36206247647723,
+      "grad_norm": 0.2461184561252594,
+      "learning_rate": 4.196778711849886e-05,
+      "loss": 0.2298,
+      "step": 3138
+    },
+    {
+      "epoch": 2.362815205118555,
+      "grad_norm": 0.3162124752998352,
+      "learning_rate": 4.196208789250134e-05,
+      "loss": 0.3242,
+      "step": 3139
+    },
+    {
+      "epoch": 2.3635679337598794,
+      "grad_norm": 0.23780161142349243,
+      "learning_rate": 4.1956387084951626e-05,
+      "loss": 0.3066,
+      "step": 3140
+    },
+    {
+      "epoch": 2.3643206624012043,
+      "grad_norm": 0.25256669521331787,
+      "learning_rate": 4.1950684696473154e-05,
+      "loss": 0.2741,
+      "step": 3141
+    },
+    {
+      "epoch": 2.3650733910425292,
+      "grad_norm": 0.2862425148487091,
+      "learning_rate": 4.1944980727689526e-05,
+      "loss": 0.3313,
+      "step": 3142
+    },
+    {
+      "epoch": 2.365826119683854,
+      "grad_norm": 0.24092036485671997,
+      "learning_rate": 4.1939275179224526e-05,
+      "loss": 0.2484,
+      "step": 3143
+    },
+    {
+      "epoch": 2.3665788483251786,
+      "grad_norm": 0.2905380427837372,
+      "learning_rate": 4.193356805170211e-05,
+      "loss": 0.3213,
+      "step": 3144
+    },
+    {
+      "epoch": 2.3673315769665035,
+      "grad_norm": 0.25254547595977783,
+      "learning_rate": 4.1927859345746384e-05,
+      "loss": 0.3405,
+      "step": 3145
+    },
+    {
+      "epoch": 2.3680843056078285,
+      "grad_norm": 0.27026841044425964,
+      "learning_rate": 4.192214906198166e-05,
+      "loss": 0.3661,
+      "step": 3146
+    },
+    {
+      "epoch": 2.3688370342491534,
+      "grad_norm": 0.2602153718471527,
+      "learning_rate": 4.1916437201032386e-05,
+      "loss": 0.3728,
+      "step": 3147
+    },
+    {
+      "epoch": 2.369589762890478,
+      "grad_norm": 0.2512684762477875,
+      "learning_rate": 4.1910723763523225e-05,
+      "loss": 0.2735,
+      "step": 3148
+    },
+    {
+      "epoch": 2.3703424915318028,
+      "grad_norm": 0.2823040783405304,
+      "learning_rate": 4.190500875007897e-05,
+      "loss": 0.3077,
+      "step": 3149
+    },
+    {
+      "epoch": 2.3710952201731277,
+      "grad_norm": 0.3025682866573334,
+      "learning_rate": 4.1899292161324627e-05,
+      "loss": 0.2992,
+      "step": 3150
+    },
+    {
+      "epoch": 2.3718479488144526,
+      "grad_norm": 0.2497345209121704,
+      "learning_rate": 4.1893573997885335e-05,
+      "loss": 0.2872,
+      "step": 3151
+    },
+    {
+      "epoch": 2.372600677455777,
+      "grad_norm": 0.26848292350769043,
+      "learning_rate": 4.188785426038643e-05,
+      "loss": 0.3325,
+      "step": 3152
+    },
+    {
+      "epoch": 2.373353406097102,
+      "grad_norm": 0.2516702711582184,
+      "learning_rate": 4.188213294945342e-05,
+      "loss": 0.2941,
+      "step": 3153
+    },
+    {
+      "epoch": 2.374106134738427,
+      "grad_norm": 0.23394234478473663,
+      "learning_rate": 4.187641006571197e-05,
+      "loss": 0.2627,
+      "step": 3154
+    },
+    {
+      "epoch": 2.3748588633797514,
+      "grad_norm": 0.2817922532558441,
+      "learning_rate": 4.187068560978793e-05,
+      "loss": 0.3403,
+      "step": 3155
+    },
+    {
+      "epoch": 2.3756115920210763,
+      "grad_norm": 0.24315859377384186,
+      "learning_rate": 4.1864959582307315e-05,
+      "loss": 0.373,
+      "step": 3156
+    },
+    {
+      "epoch": 2.376364320662401,
+      "grad_norm": 0.2519470751285553,
+      "learning_rate": 4.1859231983896326e-05,
+      "loss": 0.3389,
+      "step": 3157
+    },
+    {
+      "epoch": 2.377117049303726,
+      "grad_norm": 0.26468411087989807,
+      "learning_rate": 4.1853502815181304e-05,
+      "loss": 0.2907,
+      "step": 3158
+    },
+    {
+      "epoch": 2.377869777945051,
+      "grad_norm": 0.23911361396312714,
+      "learning_rate": 4.184777207678881e-05,
+      "loss": 0.3001,
+      "step": 3159
+    },
+    {
+      "epoch": 2.3786225065863755,
+      "grad_norm": 0.23903660476207733,
+      "learning_rate": 4.184203976934552e-05,
+      "loss": 0.2307,
+      "step": 3160
+    },
+    {
+      "epoch": 2.3793752352277004,
+      "grad_norm": 0.2648357152938843,
+      "learning_rate": 4.183630589347833e-05,
+      "loss": 0.3856,
+      "step": 3161
+    },
+    {
+      "epoch": 2.3801279638690254,
+      "grad_norm": 0.2667848765850067,
+      "learning_rate": 4.183057044981428e-05,
+      "loss": 0.257,
+      "step": 3162
+    },
+    {
+      "epoch": 2.38088069251035,
+      "grad_norm": 0.25733259320259094,
+      "learning_rate": 4.18248334389806e-05,
+      "loss": 0.2929,
+      "step": 3163
+    },
+    {
+      "epoch": 2.3816334211516748,
+      "grad_norm": 0.2905304431915283,
+      "learning_rate": 4.1819094861604656e-05,
+      "loss": 0.358,
+      "step": 3164
+    },
+    {
+      "epoch": 2.3823861497929997,
+      "grad_norm": 0.2877093553543091,
+      "learning_rate": 4.1813354718314035e-05,
+      "loss": 0.3365,
+      "step": 3165
+    },
+    {
+      "epoch": 2.3831388784343246,
+      "grad_norm": 0.244704008102417,
+      "learning_rate": 4.180761300973646e-05,
+      "loss": 0.3748,
+      "step": 3166
+    },
+    {
+      "epoch": 2.383891607075649,
+      "grad_norm": 0.2684020400047302,
+      "learning_rate": 4.1801869736499836e-05,
+      "loss": 0.3508,
+      "step": 3167
+    },
+    {
+      "epoch": 2.384644335716974,
+      "grad_norm": 0.32708030939102173,
+      "learning_rate": 4.1796124899232244e-05,
+      "loss": 0.3786,
+      "step": 3168
+    },
+    {
+      "epoch": 2.385397064358299,
+      "grad_norm": 0.2635462284088135,
+      "learning_rate": 4.1790378498561924e-05,
+      "loss": 0.3669,
+      "step": 3169
+    },
+    {
+      "epoch": 2.386149792999624,
+      "grad_norm": 0.3123822808265686,
+      "learning_rate": 4.1784630535117295e-05,
+      "loss": 0.3775,
+      "step": 3170
+    },
+    {
+      "epoch": 2.3869025216409483,
+      "grad_norm": 0.3359375298023224,
+      "learning_rate": 4.177888100952695e-05,
+      "loss": 0.283,
+      "step": 3171
+    },
+    {
+      "epoch": 2.387655250282273,
+      "grad_norm": 0.38472145795822144,
+      "learning_rate": 4.177312992241965e-05,
+      "loss": 0.3558,
+      "step": 3172
+    },
+    {
+      "epoch": 2.388407978923598,
+      "grad_norm": 0.2756100594997406,
+      "learning_rate": 4.1767377274424316e-05,
+      "loss": 0.3119,
+      "step": 3173
+    },
+    {
+      "epoch": 2.3891607075649226,
+      "grad_norm": 0.34288182854652405,
+      "learning_rate": 4.176162306617006e-05,
+      "loss": 0.3501,
+      "step": 3174
+    },
+    {
+      "epoch": 2.3899134362062475,
+      "grad_norm": 0.3112064301967621,
+      "learning_rate": 4.1755867298286136e-05,
+      "loss": 0.2345,
+      "step": 3175
+    },
+    {
+      "epoch": 2.3906661648475724,
+      "grad_norm": 0.3451463580131531,
+      "learning_rate": 4.1750109971402005e-05,
+      "loss": 0.3955,
+      "step": 3176
+    },
+    {
+      "epoch": 2.3914188934888974,
+      "grad_norm": 0.3004118502140045,
+      "learning_rate": 4.174435108614727e-05,
+      "loss": 0.3933,
+      "step": 3177
+    },
+    {
+      "epoch": 2.3921716221302223,
+      "grad_norm": 0.2803616225719452,
+      "learning_rate": 4.1738590643151724e-05,
+      "loss": 0.2782,
+      "step": 3178
+    },
+    {
+      "epoch": 2.3929243507715467,
+      "grad_norm": 0.25280919671058655,
+      "learning_rate": 4.17328286430453e-05,
+      "loss": 0.3067,
+      "step": 3179
+    },
+    {
+      "epoch": 2.3936770794128717,
+      "grad_norm": 0.296789288520813,
+      "learning_rate": 4.1727065086458144e-05,
+      "loss": 0.2849,
+      "step": 3180
+    },
+    {
+      "epoch": 2.3944298080541966,
+      "grad_norm": 0.3289393484592438,
+      "learning_rate": 4.1721299974020534e-05,
+      "loss": 0.3853,
+      "step": 3181
+    },
+    {
+      "epoch": 2.395182536695521,
+      "grad_norm": 0.25462961196899414,
+      "learning_rate": 4.171553330636295e-05,
+      "loss": 0.3783,
+      "step": 3182
+    },
+    {
+      "epoch": 2.395935265336846,
+      "grad_norm": 0.2536754012107849,
+      "learning_rate": 4.1709765084116005e-05,
+      "loss": 0.3761,
+      "step": 3183
+    },
+    {
+      "epoch": 2.396687993978171,
+      "grad_norm": 0.25536155700683594,
+      "learning_rate": 4.1703995307910516e-05,
+      "loss": 0.2883,
+      "step": 3184
+    },
+    {
+      "epoch": 2.397440722619496,
+      "grad_norm": 0.29642927646636963,
+      "learning_rate": 4.1698223978377464e-05,
+      "loss": 0.2858,
+      "step": 3185
+    },
+    {
+      "epoch": 2.3981934512608203,
+      "grad_norm": 0.2972022593021393,
+      "learning_rate": 4.169245109614798e-05,
+      "loss": 0.4151,
+      "step": 3186
+    },
+    {
+      "epoch": 2.398946179902145,
+      "grad_norm": 0.24690578877925873,
+      "learning_rate": 4.168667666185338e-05,
+      "loss": 0.2787,
+      "step": 3187
+    },
+    {
+      "epoch": 2.39969890854347,
+      "grad_norm": 0.29868191480636597,
+      "learning_rate": 4.1680900676125154e-05,
+      "loss": 0.263,
+      "step": 3188
+    },
+    {
+      "epoch": 2.400451637184795,
+      "grad_norm": 0.27446621656417847,
+      "learning_rate": 4.167512313959494e-05,
+      "loss": 0.407,
+      "step": 3189
+    },
+    {
+      "epoch": 2.4012043658261195,
+      "grad_norm": 0.25278693437576294,
+      "learning_rate": 4.166934405289458e-05,
+      "loss": 0.3498,
+      "step": 3190
+    },
+    {
+      "epoch": 2.4019570944674444,
+      "grad_norm": 0.22379599511623383,
+      "learning_rate": 4.166356341665606e-05,
+      "loss": 0.3864,
+      "step": 3191
+    },
+    {
+      "epoch": 2.4027098231087693,
+      "grad_norm": 0.22952228784561157,
+      "learning_rate": 4.1657781231511534e-05,
+      "loss": 0.2799,
+      "step": 3192
+    },
+    {
+      "epoch": 2.4034625517500943,
+      "grad_norm": 0.24805280566215515,
+      "learning_rate": 4.165199749809334e-05,
+      "loss": 0.3221,
+      "step": 3193
+    },
+    {
+      "epoch": 2.4042152803914187,
+      "grad_norm": 0.21981747448444366,
+      "learning_rate": 4.164621221703397e-05,
+      "loss": 0.2874,
+      "step": 3194
+    },
+    {
+      "epoch": 2.4049680090327437,
+      "grad_norm": 0.22315512597560883,
+      "learning_rate": 4.164042538896611e-05,
+      "loss": 0.2726,
+      "step": 3195
+    },
+    {
+      "epoch": 2.4057207376740686,
+      "grad_norm": 0.264536589384079,
+      "learning_rate": 4.1634637014522594e-05,
+      "loss": 0.2401,
+      "step": 3196
+    },
+    {
+      "epoch": 2.4064734663153935,
+      "grad_norm": 0.25538671016693115,
+      "learning_rate": 4.1628847094336416e-05,
+      "loss": 0.3192,
+      "step": 3197
+    },
+    {
+      "epoch": 2.407226194956718,
+      "grad_norm": 0.23844631016254425,
+      "learning_rate": 4.1623055629040766e-05,
+      "loss": 0.3088,
+      "step": 3198
+    },
+    {
+      "epoch": 2.407978923598043,
+      "grad_norm": 0.22284305095672607,
+      "learning_rate": 4.161726261926898e-05,
+      "loss": 0.3092,
+      "step": 3199
+    },
+    {
+      "epoch": 2.408731652239368,
+      "grad_norm": 0.23665332794189453,
+      "learning_rate": 4.1611468065654586e-05,
+      "loss": 0.44,
+      "step": 3200
+    },
+    {
+      "epoch": 2.408731652239368,
+      "eval_loss": 0.317351758480072,
+      "eval_runtime": 452.6304,
+      "eval_samples_per_second": 21.269,
+      "eval_steps_per_second": 0.665,
+      "step": 3200
+    },
+    {
+      "epoch": 2.4094843808806923,
+      "grad_norm": 0.2778990566730499,
+      "learning_rate": 4.160567196883126e-05,
+      "loss": 0.2419,
+      "step": 3201
+    },
+    {
+      "epoch": 2.410237109522017,
+      "grad_norm": 0.26874297857284546,
+      "learning_rate": 4.159987432943285e-05,
+      "loss": 0.3675,
+      "step": 3202
+    },
+    {
+      "epoch": 2.410989838163342,
+      "grad_norm": 0.22376471757888794,
+      "learning_rate": 4.1594075148093384e-05,
+      "loss": 0.2828,
+      "step": 3203
+    },
+    {
+      "epoch": 2.411742566804667,
+      "grad_norm": 0.2500115633010864,
+      "learning_rate": 4.158827442544706e-05,
+      "loss": 0.4305,
+      "step": 3204
+    },
+    {
+      "epoch": 2.412495295445992,
+      "grad_norm": 0.25703921914100647,
+      "learning_rate": 4.1582472162128205e-05,
+      "loss": 0.2631,
+      "step": 3205
+    },
+    {
+      "epoch": 2.4132480240873164,
+      "grad_norm": 0.22110925614833832,
+      "learning_rate": 4.157666835877138e-05,
+      "loss": 0.3855,
+      "step": 3206
+    },
+    {
+      "epoch": 2.4140007527286413,
+      "grad_norm": 0.2949497401714325,
+      "learning_rate": 4.157086301601127e-05,
+      "loss": 0.2298,
+      "step": 3207
+    },
+    {
+      "epoch": 2.4147534813699663,
+      "grad_norm": 0.2286837100982666,
+      "learning_rate": 4.1565056134482736e-05,
+      "loss": 0.3348,
+      "step": 3208
+    },
+    {
+      "epoch": 2.4155062100112907,
+      "grad_norm": 0.2368481308221817,
+      "learning_rate": 4.15592477148208e-05,
+      "loss": 0.3341,
+      "step": 3209
+    },
+    {
+      "epoch": 2.4162589386526157,
+      "grad_norm": 0.2027183324098587,
+      "learning_rate": 4.155343775766068e-05,
+      "loss": 0.2685,
+      "step": 3210
+    },
+    {
+      "epoch": 2.4170116672939406,
+      "grad_norm": 0.2481348067522049,
+      "learning_rate": 4.154762626363773e-05,
+      "loss": 0.3575,
+      "step": 3211
+    },
+    {
+      "epoch": 2.4177643959352655,
+      "grad_norm": 0.2299315184354782,
+      "learning_rate": 4.15418132333875e-05,
+      "loss": 0.2223,
+      "step": 3212
+    },
+    {
+      "epoch": 2.41851712457659,
+      "grad_norm": 0.2119791954755783,
+      "learning_rate": 4.153599866754568e-05,
+      "loss": 0.3196,
+      "step": 3213
+    },
+    {
+      "epoch": 2.419269853217915,
+      "grad_norm": 0.22839368879795074,
+      "learning_rate": 4.153018256674816e-05,
+      "loss": 0.2642,
+      "step": 3214
+    },
+    {
+      "epoch": 2.42002258185924,
+      "grad_norm": 0.18245172500610352,
+      "learning_rate": 4.152436493163095e-05,
+      "loss": 0.2516,
+      "step": 3215
+    },
+    {
+      "epoch": 2.4207753105005647,
+      "grad_norm": 0.23357078433036804,
+      "learning_rate": 4.151854576283029e-05,
+      "loss": 0.4013,
+      "step": 3216
+    },
+    {
+      "epoch": 2.421528039141889,
+      "grad_norm": 0.23127427697181702,
+      "learning_rate": 4.151272506098253e-05,
+      "loss": 0.432,
+      "step": 3217
+    },
+    {
+      "epoch": 2.422280767783214,
+      "grad_norm": 0.21227262914180756,
+      "learning_rate": 4.1506902826724234e-05,
+      "loss": 0.3065,
+      "step": 3218
+    },
+    {
+      "epoch": 2.423033496424539,
+      "grad_norm": 0.2538437247276306,
+      "learning_rate": 4.15010790606921e-05,
+      "loss": 0.4112,
+      "step": 3219
+    },
+    {
+      "epoch": 2.423786225065864,
+      "grad_norm": 0.2108335793018341,
+      "learning_rate": 4.149525376352302e-05,
+      "loss": 0.3373,
+      "step": 3220
+    },
+    {
+      "epoch": 2.4245389537071884,
+      "grad_norm": 0.20231382548809052,
+      "learning_rate": 4.148942693585402e-05,
+      "loss": 0.2477,
+      "step": 3221
+    },
+    {
+      "epoch": 2.4252916823485133,
+      "grad_norm": 0.25028109550476074,
+      "learning_rate": 4.148359857832233e-05,
+      "loss": 0.311,
+      "step": 3222
+    },
+    {
+      "epoch": 2.4260444109898383,
+      "grad_norm": 0.23201178014278412,
+      "learning_rate": 4.147776869156532e-05,
+      "loss": 0.3053,
+      "step": 3223
+    },
+    {
+      "epoch": 2.426797139631163,
+      "grad_norm": 0.20834685862064362,
+      "learning_rate": 4.147193727622054e-05,
+      "loss": 0.2851,
+      "step": 3224
+    },
+    {
+      "epoch": 2.4275498682724876,
+      "grad_norm": 0.234198659658432,
+      "learning_rate": 4.14661043329257e-05,
+      "loss": 0.3142,
+      "step": 3225
+    },
+    {
+      "epoch": 2.4283025969138126,
+      "grad_norm": 0.2151699811220169,
+      "learning_rate": 4.14602698623187e-05,
+      "loss": 0.3284,
+      "step": 3226
+    },
+    {
+      "epoch": 2.4290553255551375,
+      "grad_norm": 0.2291640043258667,
+      "learning_rate": 4.1454433865037565e-05,
+      "loss": 0.2351,
+      "step": 3227
+    },
+    {
+      "epoch": 2.429808054196462,
+      "grad_norm": 0.24303194880485535,
+      "learning_rate": 4.144859634172053e-05,
+      "loss": 0.3118,
+      "step": 3228
+    },
+    {
+      "epoch": 2.430560782837787,
+      "grad_norm": 0.23924165964126587,
+      "learning_rate": 4.144275729300597e-05,
+      "loss": 0.3592,
+      "step": 3229
+    },
+    {
+      "epoch": 2.431313511479112,
+      "grad_norm": 0.26898741722106934,
+      "learning_rate": 4.143691671953244e-05,
+      "loss": 0.3323,
+      "step": 3230
+    },
+    {
+      "epoch": 2.4320662401204367,
+      "grad_norm": 0.22569964826107025,
+      "learning_rate": 4.143107462193864e-05,
+      "loss": 0.3097,
+      "step": 3231
+    },
+    {
+      "epoch": 2.4328189687617616,
+      "grad_norm": 0.25603097677230835,
+      "learning_rate": 4.142523100086347e-05,
+      "loss": 0.2836,
+      "step": 3232
+    },
+    {
+      "epoch": 2.433571697403086,
+      "grad_norm": 0.2132018357515335,
+      "learning_rate": 4.1419385856945975e-05,
+      "loss": 0.2406,
+      "step": 3233
+    },
+    {
+      "epoch": 2.434324426044411,
+      "grad_norm": 0.27112433314323425,
+      "learning_rate": 4.141353919082537e-05,
+      "loss": 0.2662,
+      "step": 3234
+    },
+    {
+      "epoch": 2.435077154685736,
+      "grad_norm": 0.24787260591983795,
+      "learning_rate": 4.1407691003141044e-05,
+      "loss": 0.2462,
+      "step": 3235
+    },
+    {
+      "epoch": 2.4358298833270604,
+      "grad_norm": 0.22894951701164246,
+      "learning_rate": 4.1401841294532525e-05,
+      "loss": 0.2717,
+      "step": 3236
+    },
+    {
+      "epoch": 2.4365826119683853,
+      "grad_norm": 0.2257515788078308,
+      "learning_rate": 4.139599006563955e-05,
+      "loss": 0.2575,
+      "step": 3237
+    },
+    {
+      "epoch": 2.4373353406097102,
+      "grad_norm": 0.280437171459198,
+      "learning_rate": 4.1390137317102e-05,
+      "loss": 0.2978,
+      "step": 3238
+    },
+    {
+      "epoch": 2.438088069251035,
+      "grad_norm": 0.28465306758880615,
+      "learning_rate": 4.1384283049559914e-05,
+      "loss": 0.3196,
+      "step": 3239
+    },
+    {
+      "epoch": 2.4388407978923596,
+      "grad_norm": 0.2515600919723511,
+      "learning_rate": 4.13784272636535e-05,
+      "loss": 0.2674,
+      "step": 3240
+    },
+    {
+      "epoch": 2.4395935265336846,
+      "grad_norm": 0.27387291193008423,
+      "learning_rate": 4.137256996002315e-05,
+      "loss": 0.2686,
+      "step": 3241
+    },
+    {
+      "epoch": 2.4403462551750095,
+      "grad_norm": 0.3659888505935669,
+      "learning_rate": 4.136671113930941e-05,
+      "loss": 0.3031,
+      "step": 3242
+    },
+    {
+      "epoch": 2.4410989838163344,
+      "grad_norm": 0.2420133352279663,
+      "learning_rate": 4.136085080215298e-05,
+      "loss": 0.2808,
+      "step": 3243
+    },
+    {
+      "epoch": 2.441851712457659,
+      "grad_norm": 0.34744948148727417,
+      "learning_rate": 4.1354988949194763e-05,
+      "loss": 0.4364,
+      "step": 3244
+    },
+    {
+      "epoch": 2.442604441098984,
+      "grad_norm": 0.2793281674385071,
+      "learning_rate": 4.134912558107577e-05,
+      "loss": 0.2426,
+      "step": 3245
+    },
+    {
+      "epoch": 2.4433571697403087,
+      "grad_norm": 0.2554541528224945,
+      "learning_rate": 4.134326069843723e-05,
+      "loss": 0.3289,
+      "step": 3246
+    },
+    {
+      "epoch": 2.444109898381633,
+      "grad_norm": 0.3060259222984314,
+      "learning_rate": 4.133739430192051e-05,
+      "loss": 0.277,
+      "step": 3247
+    },
+    {
+      "epoch": 2.444862627022958,
+      "grad_norm": 0.29195812344551086,
+      "learning_rate": 4.133152639216716e-05,
+      "loss": 0.4468,
+      "step": 3248
+    },
+    {
+      "epoch": 2.445615355664283,
+      "grad_norm": 0.27057552337646484,
+      "learning_rate": 4.132565696981888e-05,
+      "loss": 0.4081,
+      "step": 3249
+    },
+    {
+      "epoch": 2.446368084305608,
+      "grad_norm": 0.21082696318626404,
+      "learning_rate": 4.1319786035517534e-05,
+      "loss": 0.2291,
+      "step": 3250
+    },
+    {
+      "epoch": 2.447120812946933,
+      "grad_norm": 0.336929053068161,
+      "learning_rate": 4.131391358990516e-05,
+      "loss": 0.2848,
+      "step": 3251
+    },
+    {
+      "epoch": 2.4478735415882573,
+      "grad_norm": 0.290126234292984,
+      "learning_rate": 4.130803963362398e-05,
+      "loss": 0.3822,
+      "step": 3252
+    },
+    {
+      "epoch": 2.4486262702295822,
+      "grad_norm": 0.2986016869544983,
+      "learning_rate": 4.130216416731633e-05,
+      "loss": 0.2648,
+      "step": 3253
+    },
+    {
+      "epoch": 2.449378998870907,
+      "grad_norm": 0.309140145778656,
+      "learning_rate": 4.129628719162477e-05,
+      "loss": 0.3681,
+      "step": 3254
+    },
+    {
+      "epoch": 2.4501317275122316,
+      "grad_norm": 0.2839290201663971,
+      "learning_rate": 4.129040870719198e-05,
+      "loss": 0.2435,
+      "step": 3255
+    },
+    {
+      "epoch": 2.4508844561535565,
+      "grad_norm": 0.2693958282470703,
+      "learning_rate": 4.128452871466082e-05,
+      "loss": 0.3272,
+      "step": 3256
+    },
+    {
+      "epoch": 2.4516371847948815,
+      "grad_norm": 0.2609303295612335,
+      "learning_rate": 4.127864721467433e-05,
+      "loss": 0.3257,
+      "step": 3257
+    },
+    {
+      "epoch": 2.4523899134362064,
+      "grad_norm": 0.27766337990760803,
+      "learning_rate": 4.1272764207875706e-05,
+      "loss": 0.2291,
+      "step": 3258
+    },
+    {
+      "epoch": 2.4531426420775313,
+      "grad_norm": 0.23948991298675537,
+      "learning_rate": 4.1266879694908275e-05,
+      "loss": 0.3535,
+      "step": 3259
+    },
+    {
+      "epoch": 2.4538953707188558,
+      "grad_norm": 0.2659685015678406,
+      "learning_rate": 4.12609936764156e-05,
+      "loss": 0.2592,
+      "step": 3260
+    },
+    {
+      "epoch": 2.4546480993601807,
+      "grad_norm": 0.23549874126911163,
+      "learning_rate": 4.125510615304134e-05,
+      "loss": 0.2733,
+      "step": 3261
+    },
+    {
+      "epoch": 2.4554008280015056,
+      "grad_norm": 0.23151278495788574,
+      "learning_rate": 4.124921712542934e-05,
+      "loss": 0.4141,
+      "step": 3262
+    },
+    {
+      "epoch": 2.45615355664283,
+      "grad_norm": 0.27645736932754517,
+      "learning_rate": 4.124332659422364e-05,
+      "loss": 0.2772,
+      "step": 3263
+    },
+    {
+      "epoch": 2.456906285284155,
+      "grad_norm": 0.262482225894928,
+      "learning_rate": 4.12374345600684e-05,
+      "loss": 0.2393,
+      "step": 3264
+    },
+    {
+      "epoch": 2.45765901392548,
+      "grad_norm": 0.2591935694217682,
+      "learning_rate": 4.123154102360797e-05,
+      "loss": 0.228,
+      "step": 3265
+    },
+    {
+      "epoch": 2.458411742566805,
+      "grad_norm": 0.23518197238445282,
+      "learning_rate": 4.122564598548686e-05,
+      "loss": 0.2491,
+      "step": 3266
+    },
+    {
+      "epoch": 2.4591644712081293,
+      "grad_norm": 0.21327349543571472,
+      "learning_rate": 4.1219749446349746e-05,
+      "loss": 0.3001,
+      "step": 3267
+    },
+    {
+      "epoch": 2.4599171998494542,
+      "grad_norm": 0.2522350251674652,
+      "learning_rate": 4.121385140684146e-05,
+      "loss": 0.3116,
+      "step": 3268
+    },
+    {
+      "epoch": 2.460669928490779,
+      "grad_norm": 0.21622125804424286,
+      "learning_rate": 4.1207951867607e-05,
+      "loss": 0.2945,
+      "step": 3269
+    },
+    {
+      "epoch": 2.461422657132104,
+      "grad_norm": 0.25707659125328064,
+      "learning_rate": 4.120205082929153e-05,
+      "loss": 0.2865,
+      "step": 3270
+    },
+    {
+      "epoch": 2.4621753857734285,
+      "grad_norm": 0.2532758414745331,
+      "learning_rate": 4.119614829254039e-05,
+      "loss": 0.3324,
+      "step": 3271
+    },
+    {
+      "epoch": 2.4629281144147535,
+      "grad_norm": 0.2574339210987091,
+      "learning_rate": 4.119024425799905e-05,
+      "loss": 0.3065,
+      "step": 3272
+    },
+    {
+      "epoch": 2.4636808430560784,
+      "grad_norm": 0.23535795509815216,
+      "learning_rate": 4.118433872631319e-05,
+      "loss": 0.2308,
+      "step": 3273
+    },
+    {
+      "epoch": 2.464433571697403,
+      "grad_norm": 0.2414553314447403,
+      "learning_rate": 4.1178431698128624e-05,
+      "loss": 0.3758,
+      "step": 3274
+    },
+    {
+      "epoch": 2.4651863003387278,
+      "grad_norm": 0.21439288556575775,
+      "learning_rate": 4.1172523174091334e-05,
+      "loss": 0.3827,
+      "step": 3275
+    },
+    {
+      "epoch": 2.4659390289800527,
+      "grad_norm": 0.22451356053352356,
+      "learning_rate": 4.116661315484746e-05,
+      "loss": 0.3145,
+      "step": 3276
+    },
+    {
+      "epoch": 2.4666917576213776,
+      "grad_norm": 0.21148543059825897,
+      "learning_rate": 4.116070164104332e-05,
+      "loss": 0.3099,
+      "step": 3277
+    },
+    {
+      "epoch": 2.4674444862627025,
+      "grad_norm": 0.22086060047149658,
+      "learning_rate": 4.1154788633325384e-05,
+      "loss": 0.3037,
+      "step": 3278
+    },
+    {
+      "epoch": 2.468197214904027,
+      "grad_norm": 0.22676995396614075,
+      "learning_rate": 4.1148874132340296e-05,
+      "loss": 0.2457,
+      "step": 3279
+    },
+    {
+      "epoch": 2.468949943545352,
+      "grad_norm": 0.2714582681655884,
+      "learning_rate": 4.1142958138734856e-05,
+      "loss": 0.3238,
+      "step": 3280
+    },
+    {
+      "epoch": 2.469702672186677,
+      "grad_norm": 0.2434893250465393,
+      "learning_rate": 4.113704065315602e-05,
+      "loss": 0.2737,
+      "step": 3281
+    },
+    {
+      "epoch": 2.4704554008280013,
+      "grad_norm": 0.2354852706193924,
+      "learning_rate": 4.1131121676250926e-05,
+      "loss": 0.2739,
+      "step": 3282
+    },
+    {
+      "epoch": 2.471208129469326,
+      "grad_norm": 0.2452690601348877,
+      "learning_rate": 4.1125201208666865e-05,
+      "loss": 0.2546,
+      "step": 3283
+    },
+    {
+      "epoch": 2.471960858110651,
+      "grad_norm": 0.21485179662704468,
+      "learning_rate": 4.111927925105128e-05,
+      "loss": 0.1942,
+      "step": 3284
+    },
+    {
+      "epoch": 2.472713586751976,
+      "grad_norm": 0.2808873951435089,
+      "learning_rate": 4.111335580405179e-05,
+      "loss": 0.3312,
+      "step": 3285
+    },
+    {
+      "epoch": 2.4734663153933005,
+      "grad_norm": 0.24158687889575958,
+      "learning_rate": 4.110743086831617e-05,
+      "loss": 0.398,
+      "step": 3286
+    },
+    {
+      "epoch": 2.4742190440346254,
+      "grad_norm": 0.2703055143356323,
+      "learning_rate": 4.110150444449238e-05,
+      "loss": 0.2974,
+      "step": 3287
+    },
+    {
+      "epoch": 2.4749717726759504,
+      "grad_norm": 0.23031310737133026,
+      "learning_rate": 4.1095576533228505e-05,
+      "loss": 0.3827,
+      "step": 3288
+    },
+    {
+      "epoch": 2.4757245013172753,
+      "grad_norm": 0.36769384145736694,
+      "learning_rate": 4.108964713517283e-05,
+      "loss": 0.2999,
+      "step": 3289
+    },
+    {
+      "epoch": 2.4764772299585998,
+      "grad_norm": 0.24883553385734558,
+      "learning_rate": 4.1083716250973767e-05,
+      "loss": 0.3671,
+      "step": 3290
+    },
+    {
+      "epoch": 2.4772299585999247,
+      "grad_norm": 0.2676035165786743,
+      "learning_rate": 4.107778388127992e-05,
+      "loss": 0.3687,
+      "step": 3291
+    },
+    {
+      "epoch": 2.4779826872412496,
+      "grad_norm": 0.2953171133995056,
+      "learning_rate": 4.107185002674004e-05,
+      "loss": 0.3157,
+      "step": 3292
+    },
+    {
+      "epoch": 2.4787354158825745,
+      "grad_norm": 0.3281657099723816,
+      "learning_rate": 4.106591468800306e-05,
+      "loss": 0.2397,
+      "step": 3293
+    },
+    {
+      "epoch": 2.479488144523899,
+      "grad_norm": 0.2684437334537506,
+      "learning_rate": 4.105997786571803e-05,
+      "loss": 0.2844,
+      "step": 3294
+    },
+    {
+      "epoch": 2.480240873165224,
+      "grad_norm": 0.23442411422729492,
+      "learning_rate": 4.105403956053422e-05,
+      "loss": 0.3402,
+      "step": 3295
+    },
+    {
+      "epoch": 2.480993601806549,
+      "grad_norm": 0.2681843936443329,
+      "learning_rate": 4.104809977310101e-05,
+      "loss": 0.3179,
+      "step": 3296
+    },
+    {
+      "epoch": 2.4817463304478737,
+      "grad_norm": 0.24901899695396423,
+      "learning_rate": 4.104215850406798e-05,
+      "loss": 0.2226,
+      "step": 3297
+    },
+    {
+      "epoch": 2.482499059089198,
+      "grad_norm": 0.25668439269065857,
+      "learning_rate": 4.1036215754084864e-05,
+      "loss": 0.3691,
+      "step": 3298
+    },
+    {
+      "epoch": 2.483251787730523,
+      "grad_norm": 0.30761560797691345,
+      "learning_rate": 4.103027152380154e-05,
+      "loss": 0.3309,
+      "step": 3299
+    },
+    {
+      "epoch": 2.484004516371848,
+      "grad_norm": 0.2648111581802368,
+      "learning_rate": 4.1024325813868065e-05,
+      "loss": 0.213,
+      "step": 3300
+    },
+    {
+      "epoch": 2.4847572450131725,
+      "grad_norm": 0.2212122678756714,
+      "learning_rate": 4.1018378624934654e-05,
+      "loss": 0.2984,
+      "step": 3301
+    },
+    {
+      "epoch": 2.4855099736544974,
+      "grad_norm": 0.2662200629711151,
+      "learning_rate": 4.1012429957651676e-05,
+      "loss": 0.2993,
+      "step": 3302
+    },
+    {
+      "epoch": 2.4862627022958224,
+      "grad_norm": 0.22049419581890106,
+      "learning_rate": 4.100647981266968e-05,
+      "loss": 0.2513,
+      "step": 3303
+    },
+    {
+      "epoch": 2.4870154309371473,
+      "grad_norm": 0.2637242376804352,
+      "learning_rate": 4.100052819063935e-05,
+      "loss": 0.3027,
+      "step": 3304
+    },
+    {
+      "epoch": 2.487768159578472,
+      "grad_norm": 0.2636163830757141,
+      "learning_rate": 4.0994575092211565e-05,
+      "loss": 0.2763,
+      "step": 3305
+    },
+    {
+      "epoch": 2.4885208882197967,
+      "grad_norm": 0.21163271367549896,
+      "learning_rate": 4.098862051803734e-05,
+      "loss": 0.2297,
+      "step": 3306
+    },
+    {
+      "epoch": 2.4892736168611216,
+      "grad_norm": 0.2573690414428711,
+      "learning_rate": 4.098266446876784e-05,
+      "loss": 0.291,
+      "step": 3307
+    },
+    {
+      "epoch": 2.4900263455024465,
+      "grad_norm": 0.28161823749542236,
+      "learning_rate": 4.097670694505443e-05,
+      "loss": 0.3165,
+      "step": 3308
+    },
+    {
+      "epoch": 2.490779074143771,
+      "grad_norm": 0.2649877369403839,
+      "learning_rate": 4.097074794754862e-05,
+      "loss": 0.3956,
+      "step": 3309
+    },
+    {
+      "epoch": 2.491531802785096,
+      "grad_norm": 0.2355857640504837,
+      "learning_rate": 4.0964787476902056e-05,
+      "loss": 0.3587,
+      "step": 3310
+    },
+    {
+      "epoch": 2.492284531426421,
+      "grad_norm": 0.22356441617012024,
+      "learning_rate": 4.0958825533766573e-05,
+      "loss": 0.2683,
+      "step": 3311
+    },
+    {
+      "epoch": 2.4930372600677457,
+      "grad_norm": 0.3172590136528015,
+      "learning_rate": 4.095286211879418e-05,
+      "loss": 0.3253,
+      "step": 3312
+    },
+    {
+      "epoch": 2.49378998870907,
+      "grad_norm": 0.24229219555854797,
+      "learning_rate": 4.0946897232637e-05,
+      "loss": 0.2852,
+      "step": 3313
+    },
+    {
+      "epoch": 2.494542717350395,
+      "grad_norm": 0.24951748549938202,
+      "learning_rate": 4.094093087594736e-05,
+      "loss": 0.2961,
+      "step": 3314
+    },
+    {
+      "epoch": 2.49529544599172,
+      "grad_norm": 0.23023803532123566,
+      "learning_rate": 4.093496304937773e-05,
+      "loss": 0.2545,
+      "step": 3315
+    },
+    {
+      "epoch": 2.496048174633045,
+      "grad_norm": 0.2645374536514282,
+      "learning_rate": 4.092899375358074e-05,
+      "loss": 0.4514,
+      "step": 3316
+    },
+    {
+      "epoch": 2.4968009032743694,
+      "grad_norm": 0.26102715730667114,
+      "learning_rate": 4.092302298920918e-05,
+      "loss": 0.2329,
+      "step": 3317
+    },
+    {
+      "epoch": 2.4975536319156943,
+      "grad_norm": 0.23292653262615204,
+      "learning_rate": 4.0917050756916006e-05,
+      "loss": 0.2948,
+      "step": 3318
+    },
+    {
+      "epoch": 2.4983063605570193,
+      "grad_norm": 0.2883833348751068,
+      "learning_rate": 4.0911077057354335e-05,
+      "loss": 0.2222,
+      "step": 3319
+    },
+    {
+      "epoch": 2.499059089198344,
+      "grad_norm": 0.267316073179245,
+      "learning_rate": 4.0905101891177437e-05,
+      "loss": 0.3115,
+      "step": 3320
+    },
+    {
+      "epoch": 2.4998118178396687,
+      "grad_norm": 0.2665969133377075,
+      "learning_rate": 4.089912525903875e-05,
+      "loss": 0.2576,
+      "step": 3321
+    },
+    {
+      "epoch": 2.5005645464809936,
+      "grad_norm": 0.2612282931804657,
+      "learning_rate": 4.089314716159188e-05,
+      "loss": 0.3065,
+      "step": 3322
+    },
+    {
+      "epoch": 2.5013172751223185,
+      "grad_norm": 0.3620324432849884,
+      "learning_rate": 4.0887167599490564e-05,
+      "loss": 0.3983,
+      "step": 3323
+    },
+    {
+      "epoch": 2.5020700037636434,
+      "grad_norm": 0.28006061911582947,
+      "learning_rate": 4.088118657338873e-05,
+      "loss": 0.3487,
+      "step": 3324
+    },
+    {
+      "epoch": 2.502822732404968,
+      "grad_norm": 0.2910841107368469,
+      "learning_rate": 4.087520408394045e-05,
+      "loss": 0.2892,
+      "step": 3325
+    },
+    {
+      "epoch": 2.503575461046293,
+      "grad_norm": 0.22090686857700348,
+      "learning_rate": 4.086922013179996e-05,
+      "loss": 0.2854,
+      "step": 3326
+    },
+    {
+      "epoch": 2.5043281896876177,
+      "grad_norm": 0.2978094518184662,
+      "learning_rate": 4.086323471762166e-05,
+      "loss": 0.287,
+      "step": 3327
+    },
+    {
+      "epoch": 2.505080918328942,
+      "grad_norm": 0.249454066157341,
+      "learning_rate": 4.0857247842060106e-05,
+      "loss": 0.2885,
+      "step": 3328
+    },
+    {
+      "epoch": 2.505833646970267,
+      "grad_norm": 0.24785691499710083,
+      "learning_rate": 4.085125950577e-05,
+      "loss": 0.461,
+      "step": 3329
+    },
+    {
+      "epoch": 2.506586375611592,
+      "grad_norm": 0.3020055890083313,
+      "learning_rate": 4.0845269709406236e-05,
+      "loss": 0.3339,
+      "step": 3330
+    },
+    {
+      "epoch": 2.507339104252917,
+      "grad_norm": 0.23044832050800323,
+      "learning_rate": 4.083927845362383e-05,
+      "loss": 0.4335,
+      "step": 3331
+    },
+    {
+      "epoch": 2.508091832894242,
+      "grad_norm": 0.2757469713687897,
+      "learning_rate": 4.0833285739077995e-05,
+      "loss": 0.3319,
+      "step": 3332
+    },
+    {
+      "epoch": 2.5088445615355663,
+      "grad_norm": 0.2194899469614029,
+      "learning_rate": 4.082729156642409e-05,
+      "loss": 0.2798,
+      "step": 3333
+    },
+    {
+      "epoch": 2.5095972901768913,
+      "grad_norm": 0.27661749720573425,
+      "learning_rate": 4.08212959363176e-05,
+      "loss": 0.287,
+      "step": 3334
+    },
+    {
+      "epoch": 2.510350018818216,
+      "grad_norm": 0.2871897220611572,
+      "learning_rate": 4.081529884941422e-05,
+      "loss": 0.2606,
+      "step": 3335
+    },
+    {
+      "epoch": 2.5111027474595407,
+      "grad_norm": 0.20966477692127228,
+      "learning_rate": 4.080930030636977e-05,
+      "loss": 0.3055,
+      "step": 3336
+    },
+    {
+      "epoch": 2.5118554761008656,
+      "grad_norm": 0.3124898672103882,
+      "learning_rate": 4.0803300307840257e-05,
+      "loss": 0.2489,
+      "step": 3337
+    },
+    {
+      "epoch": 2.5126082047421905,
+      "grad_norm": 0.2795807123184204,
+      "learning_rate": 4.079729885448183e-05,
+      "loss": 0.3389,
+      "step": 3338
+    },
+    {
+      "epoch": 2.5133609333835154,
+      "grad_norm": 0.24112536013126373,
+      "learning_rate": 4.079129594695077e-05,
+      "loss": 0.2889,
+      "step": 3339
+    },
+    {
+      "epoch": 2.51411366202484,
+      "grad_norm": 0.26017850637435913,
+      "learning_rate": 4.0785291585903585e-05,
+      "loss": 0.2494,
+      "step": 3340
+    },
+    {
+      "epoch": 2.514866390666165,
+      "grad_norm": 0.27152737975120544,
+      "learning_rate": 4.0779285771996884e-05,
+      "loss": 0.3056,
+      "step": 3341
+    },
+    {
+      "epoch": 2.5156191193074897,
+      "grad_norm": 0.2372920662164688,
+      "learning_rate": 4.077327850588746e-05,
+      "loss": 0.2257,
+      "step": 3342
+    },
+    {
+      "epoch": 2.5163718479488146,
+      "grad_norm": 0.31273138523101807,
+      "learning_rate": 4.0767269788232246e-05,
+      "loss": 0.3736,
+      "step": 3343
+    },
+    {
+      "epoch": 2.517124576590139,
+      "grad_norm": 0.27212750911712646,
+      "learning_rate": 4.076125961968836e-05,
+      "loss": 0.3318,
+      "step": 3344
+    },
+    {
+      "epoch": 2.517877305231464,
+      "grad_norm": 0.25927576422691345,
+      "learning_rate": 4.075524800091306e-05,
+      "loss": 0.1963,
+      "step": 3345
+    },
+    {
+      "epoch": 2.518630033872789,
+      "grad_norm": 0.2819238305091858,
+      "learning_rate": 4.0749234932563765e-05,
+      "loss": 0.3032,
+      "step": 3346
+    },
+    {
+      "epoch": 2.5193827625141134,
+      "grad_norm": 0.27583834528923035,
+      "learning_rate": 4.074322041529806e-05,
+      "loss": 0.2504,
+      "step": 3347
+    },
+    {
+      "epoch": 2.5201354911554383,
+      "grad_norm": 0.3781076967716217,
+      "learning_rate": 4.073720444977368e-05,
+      "loss": 0.4084,
+      "step": 3348
+    },
+    {
+      "epoch": 2.5208882197967633,
+      "grad_norm": 0.27221617102622986,
+      "learning_rate": 4.0731187036648525e-05,
+      "loss": 0.3299,
+      "step": 3349
+    },
+    {
+      "epoch": 2.521640948438088,
+      "grad_norm": 0.2976672649383545,
+      "learning_rate": 4.072516817658065e-05,
+      "loss": 0.3374,
+      "step": 3350
+    },
+    {
+      "epoch": 2.522393677079413,
+      "grad_norm": 0.287164568901062,
+      "learning_rate": 4.071914787022826e-05,
+      "loss": 0.3715,
+      "step": 3351
+    },
+    {
+      "epoch": 2.5231464057207376,
+      "grad_norm": 0.2521730959415436,
+      "learning_rate": 4.0713126118249736e-05,
+      "loss": 0.3154,
+      "step": 3352
+    },
+    {
+      "epoch": 2.5238991343620625,
+      "grad_norm": 0.25781524181365967,
+      "learning_rate": 4.0707102921303616e-05,
+      "loss": 0.382,
+      "step": 3353
+    },
+    {
+      "epoch": 2.5246518630033874,
+      "grad_norm": 0.278901606798172,
+      "learning_rate": 4.070107828004856e-05,
+      "loss": 0.3356,
+      "step": 3354
+    },
+    {
+      "epoch": 2.525404591644712,
+      "grad_norm": 0.29049819707870483,
+      "learning_rate": 4.069505219514344e-05,
+      "loss": 0.3147,
+      "step": 3355
+    },
+    {
+      "epoch": 2.526157320286037,
+      "grad_norm": 0.27616214752197266,
+      "learning_rate": 4.068902466724726e-05,
+      "loss": 0.3479,
+      "step": 3356
+    },
+    {
+      "epoch": 2.5269100489273617,
+      "grad_norm": 0.24578142166137695,
+      "learning_rate": 4.068299569701915e-05,
+      "loss": 0.2277,
+      "step": 3357
+    },
+    {
+      "epoch": 2.5276627775686866,
+      "grad_norm": 0.3318336606025696,
+      "learning_rate": 4.067696528511847e-05,
+      "loss": 0.2859,
+      "step": 3358
+    },
+    {
+      "epoch": 2.5284155062100115,
+      "grad_norm": 0.30211690068244934,
+      "learning_rate": 4.0670933432204674e-05,
+      "loss": 0.288,
+      "step": 3359
+    },
+    {
+      "epoch": 2.529168234851336,
+      "grad_norm": 0.242706298828125,
+      "learning_rate": 4.066490013893739e-05,
+      "loss": 0.2457,
+      "step": 3360
+    },
+    {
+      "epoch": 2.529920963492661,
+      "grad_norm": 0.26594793796539307,
+      "learning_rate": 4.065886540597642e-05,
+      "loss": 0.2569,
+      "step": 3361
+    },
+    {
+      "epoch": 2.530673692133986,
+      "grad_norm": 0.2540222108364105,
+      "learning_rate": 4.065282923398172e-05,
+      "loss": 0.1665,
+      "step": 3362
+    },
+    {
+      "epoch": 2.5314264207753103,
+      "grad_norm": 0.25298216938972473,
+      "learning_rate": 4.06467916236134e-05,
+      "loss": 0.3254,
+      "step": 3363
+    },
+    {
+      "epoch": 2.5321791494166352,
+      "grad_norm": 0.22934319078922272,
+      "learning_rate": 4.0640752575531695e-05,
+      "loss": 0.2574,
+      "step": 3364
+    },
+    {
+      "epoch": 2.53293187805796,
+      "grad_norm": 0.32624831795692444,
+      "learning_rate": 4.0634712090397056e-05,
+      "loss": 0.3493,
+      "step": 3365
+    },
+    {
+      "epoch": 2.5336846066992846,
+      "grad_norm": 0.25302940607070923,
+      "learning_rate": 4.062867016887005e-05,
+      "loss": 0.2836,
+      "step": 3366
+    },
+    {
+      "epoch": 2.5344373353406096,
+      "grad_norm": 0.2548306882381439,
+      "learning_rate": 4.06226268116114e-05,
+      "loss": 0.347,
+      "step": 3367
+    },
+    {
+      "epoch": 2.5351900639819345,
+      "grad_norm": 0.2736336290836334,
+      "learning_rate": 4.061658201928203e-05,
+      "loss": 0.3194,
+      "step": 3368
+    },
+    {
+      "epoch": 2.5359427926232594,
+      "grad_norm": 0.2501259744167328,
+      "learning_rate": 4.061053579254296e-05,
+      "loss": 0.3749,
+      "step": 3369
+    },
+    {
+      "epoch": 2.5366955212645843,
+      "grad_norm": 0.3099696636199951,
+      "learning_rate": 4.060448813205541e-05,
+      "loss": 0.2928,
+      "step": 3370
+    },
+    {
+      "epoch": 2.537448249905909,
+      "grad_norm": 0.223527193069458,
+      "learning_rate": 4.059843903848075e-05,
+      "loss": 0.3023,
+      "step": 3371
+    },
+    {
+      "epoch": 2.5382009785472337,
+      "grad_norm": 0.2930067181587219,
+      "learning_rate": 4.059238851248048e-05,
+      "loss": 0.2917,
+      "step": 3372
+    },
+    {
+      "epoch": 2.5389537071885586,
+      "grad_norm": 0.27338191866874695,
+      "learning_rate": 4.0586336554716287e-05,
+      "loss": 0.2867,
+      "step": 3373
+    },
+    {
+      "epoch": 2.539706435829883,
+      "grad_norm": 0.3196203410625458,
+      "learning_rate": 4.0580283165849995e-05,
+      "loss": 0.2901,
+      "step": 3374
+    },
+    {
+      "epoch": 2.540459164471208,
+      "grad_norm": 0.2689710259437561,
+      "learning_rate": 4.0574228346543613e-05,
+      "loss": 0.3773,
+      "step": 3375
+    },
+    {
+      "epoch": 2.541211893112533,
+      "grad_norm": 0.25013887882232666,
+      "learning_rate": 4.056817209745927e-05,
+      "loss": 0.3698,
+      "step": 3376
+    },
+    {
+      "epoch": 2.541964621753858,
+      "grad_norm": 0.2847682237625122,
+      "learning_rate": 4.056211441925928e-05,
+      "loss": 0.3348,
+      "step": 3377
+    },
+    {
+      "epoch": 2.5427173503951828,
+      "grad_norm": 0.20919334888458252,
+      "learning_rate": 4.0556055312606094e-05,
+      "loss": 0.3245,
+      "step": 3378
+    },
+    {
+      "epoch": 2.5434700790365072,
+      "grad_norm": 0.24247343838214874,
+      "learning_rate": 4.054999477816233e-05,
+      "loss": 0.3402,
+      "step": 3379
+    },
+    {
+      "epoch": 2.544222807677832,
+      "grad_norm": 0.3024549186229706,
+      "learning_rate": 4.054393281659074e-05,
+      "loss": 0.4285,
+      "step": 3380
+    },
+    {
+      "epoch": 2.544975536319157,
+      "grad_norm": 0.27722403407096863,
+      "learning_rate": 4.053786942855429e-05,
+      "loss": 0.2919,
+      "step": 3381
+    },
+    {
+      "epoch": 2.5457282649604815,
+      "grad_norm": 0.2997664511203766,
+      "learning_rate": 4.053180461471604e-05,
+      "loss": 0.2824,
+      "step": 3382
+    },
+    {
+      "epoch": 2.5464809936018065,
+      "grad_norm": 0.33107420802116394,
+      "learning_rate": 4.052573837573922e-05,
+      "loss": 0.2753,
+      "step": 3383
+    },
+    {
+      "epoch": 2.5472337222431314,
+      "grad_norm": 0.2949436604976654,
+      "learning_rate": 4.0519670712287246e-05,
+      "loss": 0.3055,
+      "step": 3384
+    },
+    {
+      "epoch": 2.5479864508844563,
+      "grad_norm": 0.2661774754524231,
+      "learning_rate": 4.051360162502365e-05,
+      "loss": 0.3257,
+      "step": 3385
+    },
+    {
+      "epoch": 2.548739179525781,
+      "grad_norm": 0.24765995144844055,
+      "learning_rate": 4.050753111461215e-05,
+      "loss": 0.3519,
+      "step": 3386
+    },
+    {
+      "epoch": 2.5494919081671057,
+      "grad_norm": 0.2949119508266449,
+      "learning_rate": 4.05014591817166e-05,
+      "loss": 0.2641,
+      "step": 3387
+    },
+    {
+      "epoch": 2.5502446368084306,
+      "grad_norm": 0.2650965750217438,
+      "learning_rate": 4.049538582700102e-05,
+      "loss": 0.3484,
+      "step": 3388
+    },
+    {
+      "epoch": 2.5509973654497555,
+      "grad_norm": 0.23756815493106842,
+      "learning_rate": 4.0489311051129594e-05,
+      "loss": 0.3015,
+      "step": 3389
+    },
+    {
+      "epoch": 2.55175009409108,
+      "grad_norm": 0.2903708219528198,
+      "learning_rate": 4.048323485476663e-05,
+      "loss": 0.2414,
+      "step": 3390
+    },
+    {
+      "epoch": 2.552502822732405,
+      "grad_norm": 0.2623025178909302,
+      "learning_rate": 4.047715723857662e-05,
+      "loss": 0.2784,
+      "step": 3391
+    },
+    {
+      "epoch": 2.55325555137373,
+      "grad_norm": 0.24479876458644867,
+      "learning_rate": 4.047107820322422e-05,
+      "loss": 0.2347,
+      "step": 3392
+    },
+    {
+      "epoch": 2.5540082800150543,
+      "grad_norm": 0.25648969411849976,
+      "learning_rate": 4.04649977493742e-05,
+      "loss": 0.294,
+      "step": 3393
+    },
+    {
+      "epoch": 2.5547610086563792,
+      "grad_norm": 0.2448810189962387,
+      "learning_rate": 4.0458915877691516e-05,
+      "loss": 0.2938,
+      "step": 3394
+    },
+    {
+      "epoch": 2.555513737297704,
+      "grad_norm": 0.27155205607414246,
+      "learning_rate": 4.045283258884128e-05,
+      "loss": 0.2335,
+      "step": 3395
+    },
+    {
+      "epoch": 2.556266465939029,
+      "grad_norm": 0.252228707075119,
+      "learning_rate": 4.0446747883488746e-05,
+      "loss": 0.2812,
+      "step": 3396
+    },
+    {
+      "epoch": 2.557019194580354,
+      "grad_norm": 0.2518555819988251,
+      "learning_rate": 4.044066176229932e-05,
+      "loss": 0.3798,
+      "step": 3397
+    },
+    {
+      "epoch": 2.5577719232216785,
+      "grad_norm": 0.2715579867362976,
+      "learning_rate": 4.0434574225938595e-05,
+      "loss": 0.2993,
+      "step": 3398
+    },
+    {
+      "epoch": 2.5585246518630034,
+      "grad_norm": 0.241278275847435,
+      "learning_rate": 4.042848527507227e-05,
+      "loss": 0.3224,
+      "step": 3399
+    },
+    {
+      "epoch": 2.5592773805043283,
+      "grad_norm": 0.24968315660953522,
+      "learning_rate": 4.0422394910366236e-05,
+      "loss": 0.1883,
+      "step": 3400
+    },
+    {
+      "epoch": 2.5592773805043283,
+      "eval_loss": 0.30738940834999084,
+      "eval_runtime": 452.7949,
+      "eval_samples_per_second": 21.261,
+      "eval_steps_per_second": 0.665,
+      "step": 3400
+    },
+    {
+      "epoch": 2.5600301091456528,
+      "grad_norm": 0.2175167053937912,
+      "learning_rate": 4.041630313248652e-05,
+      "loss": 0.3376,
+      "step": 3401
+    },
+    {
+      "epoch": 2.5607828377869777,
+      "grad_norm": 0.23830164968967438,
+      "learning_rate": 4.041020994209931e-05,
+      "loss": 0.3703,
+      "step": 3402
+    },
+    {
+      "epoch": 2.5615355664283026,
+      "grad_norm": 0.2131662219762802,
+      "learning_rate": 4.040411533987096e-05,
+      "loss": 0.3146,
+      "step": 3403
+    },
+    {
+      "epoch": 2.5622882950696275,
+      "grad_norm": 0.24609409272670746,
+      "learning_rate": 4.039801932646795e-05,
+      "loss": 0.3423,
+      "step": 3404
+    },
+    {
+      "epoch": 2.5630410237109524,
+      "grad_norm": 0.23526887595653534,
+      "learning_rate": 4.0391921902556946e-05,
+      "loss": 0.3676,
+      "step": 3405
+    },
+    {
+      "epoch": 2.563793752352277,
+      "grad_norm": 0.25700607895851135,
+      "learning_rate": 4.038582306880474e-05,
+      "loss": 0.2125,
+      "step": 3406
+    },
+    {
+      "epoch": 2.564546480993602,
+      "grad_norm": 0.2821401059627533,
+      "learning_rate": 4.0379722825878295e-05,
+      "loss": 0.3447,
+      "step": 3407
+    },
+    {
+      "epoch": 2.5652992096349267,
+      "grad_norm": 0.24637527763843536,
+      "learning_rate": 4.0373621174444735e-05,
+      "loss": 0.2263,
+      "step": 3408
+    },
+    {
+      "epoch": 2.566051938276251,
+      "grad_norm": 0.2481406182050705,
+      "learning_rate": 4.036751811517131e-05,
+      "loss": 0.2672,
+      "step": 3409
+    },
+    {
+      "epoch": 2.566804666917576,
+      "grad_norm": 0.3134644627571106,
+      "learning_rate": 4.036141364872545e-05,
+      "loss": 0.2774,
+      "step": 3410
+    },
+    {
+      "epoch": 2.567557395558901,
+      "grad_norm": 0.2698027789592743,
+      "learning_rate": 4.035530777577473e-05,
+      "loss": 0.3381,
+      "step": 3411
+    },
+    {
+      "epoch": 2.568310124200226,
+      "grad_norm": 0.2522766590118408,
+      "learning_rate": 4.034920049698688e-05,
+      "loss": 0.3106,
+      "step": 3412
+    },
+    {
+      "epoch": 2.569062852841551,
+      "grad_norm": 0.27410557866096497,
+      "learning_rate": 4.0343091813029784e-05,
+      "loss": 0.3242,
+      "step": 3413
+    },
+    {
+      "epoch": 2.5698155814828754,
+      "grad_norm": 0.3116827607154846,
+      "learning_rate": 4.033698172457148e-05,
+      "loss": 0.2705,
+      "step": 3414
+    },
+    {
+      "epoch": 2.5705683101242003,
+      "grad_norm": 0.23684969544410706,
+      "learning_rate": 4.0330870232280146e-05,
+      "loss": 0.2396,
+      "step": 3415
+    },
+    {
+      "epoch": 2.571321038765525,
+      "grad_norm": 0.23399679362773895,
+      "learning_rate": 4.032475733682414e-05,
+      "loss": 0.1833,
+      "step": 3416
+    },
+    {
+      "epoch": 2.5720737674068497,
+      "grad_norm": 0.2851743698120117,
+      "learning_rate": 4.031864303887196e-05,
+      "loss": 0.297,
+      "step": 3417
+    },
+    {
+      "epoch": 2.5728264960481746,
+      "grad_norm": 0.21854588389396667,
+      "learning_rate": 4.031252733909224e-05,
+      "loss": 0.3507,
+      "step": 3418
+    },
+    {
+      "epoch": 2.5735792246894995,
+      "grad_norm": 0.26796090602874756,
+      "learning_rate": 4.03064102381538e-05,
+      "loss": 0.2646,
+      "step": 3419
+    },
+    {
+      "epoch": 2.574331953330824,
+      "grad_norm": 0.21894748508930206,
+      "learning_rate": 4.030029173672558e-05,
+      "loss": 0.2507,
+      "step": 3420
+    },
+    {
+      "epoch": 2.575084681972149,
+      "grad_norm": 0.24670766294002533,
+      "learning_rate": 4.029417183547671e-05,
+      "loss": 0.2681,
+      "step": 3421
+    },
+    {
+      "epoch": 2.575837410613474,
+      "grad_norm": 0.26532241702079773,
+      "learning_rate": 4.028805053507644e-05,
+      "loss": 0.3448,
+      "step": 3422
+    },
+    {
+      "epoch": 2.5765901392547987,
+      "grad_norm": 0.3074951469898224,
+      "learning_rate": 4.02819278361942e-05,
+      "loss": 0.3005,
+      "step": 3423
+    },
+    {
+      "epoch": 2.5773428678961237,
+      "grad_norm": 0.21249240636825562,
+      "learning_rate": 4.027580373949953e-05,
+      "loss": 0.271,
+      "step": 3424
+    },
+    {
+      "epoch": 2.578095596537448,
+      "grad_norm": 0.26529285311698914,
+      "learning_rate": 4.0269678245662186e-05,
+      "loss": 0.3087,
+      "step": 3425
+    },
+    {
+      "epoch": 2.578848325178773,
+      "grad_norm": 0.22197489440441132,
+      "learning_rate": 4.0263551355352024e-05,
+      "loss": 0.356,
+      "step": 3426
+    },
+    {
+      "epoch": 2.579601053820098,
+      "grad_norm": 0.25850042700767517,
+      "learning_rate": 4.025742306923908e-05,
+      "loss": 0.2466,
+      "step": 3427
+    },
+    {
+      "epoch": 2.5803537824614224,
+      "grad_norm": 0.23061314225196838,
+      "learning_rate": 4.025129338799352e-05,
+      "loss": 0.1798,
+      "step": 3428
+    },
+    {
+      "epoch": 2.5811065111027474,
+      "grad_norm": 0.23760411143302917,
+      "learning_rate": 4.024516231228569e-05,
+      "loss": 0.3344,
+      "step": 3429
+    },
+    {
+      "epoch": 2.5818592397440723,
+      "grad_norm": 0.2205663025379181,
+      "learning_rate": 4.023902984278607e-05,
+      "loss": 0.2725,
+      "step": 3430
+    },
+    {
+      "epoch": 2.582611968385397,
+      "grad_norm": 0.25118958950042725,
+      "learning_rate": 4.02328959801653e-05,
+      "loss": 0.4201,
+      "step": 3431
+    },
+    {
+      "epoch": 2.583364697026722,
+      "grad_norm": 0.2640766203403473,
+      "learning_rate": 4.0226760725094175e-05,
+      "loss": 0.2621,
+      "step": 3432
+    },
+    {
+      "epoch": 2.5841174256680466,
+      "grad_norm": 0.32326650619506836,
+      "learning_rate": 4.0220624078243637e-05,
+      "loss": 0.3699,
+      "step": 3433
+    },
+    {
+      "epoch": 2.5848701543093715,
+      "grad_norm": 0.29907652735710144,
+      "learning_rate": 4.021448604028476e-05,
+      "loss": 0.407,
+      "step": 3434
+    },
+    {
+      "epoch": 2.5856228829506964,
+      "grad_norm": 0.2997153401374817,
+      "learning_rate": 4.020834661188881e-05,
+      "loss": 0.3083,
+      "step": 3435
+    },
+    {
+      "epoch": 2.586375611592021,
+      "grad_norm": 0.27375248074531555,
+      "learning_rate": 4.0202205793727185e-05,
+      "loss": 0.3384,
+      "step": 3436
+    },
+    {
+      "epoch": 2.587128340233346,
+      "grad_norm": 0.25724977254867554,
+      "learning_rate": 4.019606358647142e-05,
+      "loss": 0.3179,
+      "step": 3437
+    },
+    {
+      "epoch": 2.5878810688746707,
+      "grad_norm": 0.30648544430732727,
+      "learning_rate": 4.018991999079325e-05,
+      "loss": 0.3509,
+      "step": 3438
+    },
+    {
+      "epoch": 2.5886337975159956,
+      "grad_norm": 0.26143530011177063,
+      "learning_rate": 4.0183775007364495e-05,
+      "loss": 0.3398,
+      "step": 3439
+    },
+    {
+      "epoch": 2.58938652615732,
+      "grad_norm": 0.2614624798297882,
+      "learning_rate": 4.017762863685718e-05,
+      "loss": 0.3795,
+      "step": 3440
+    },
+    {
+      "epoch": 2.590139254798645,
+      "grad_norm": 0.23439352214336395,
+      "learning_rate": 4.017148087994345e-05,
+      "loss": 0.2207,
+      "step": 3441
+    },
+    {
+      "epoch": 2.59089198343997,
+      "grad_norm": 0.2395840287208557,
+      "learning_rate": 4.0165331737295626e-05,
+      "loss": 0.3969,
+      "step": 3442
+    },
+    {
+      "epoch": 2.591644712081295,
+      "grad_norm": 0.24835245311260223,
+      "learning_rate": 4.015918120958617e-05,
+      "loss": 0.2821,
+      "step": 3443
+    },
+    {
+      "epoch": 2.5923974407226194,
+      "grad_norm": 0.27254509925842285,
+      "learning_rate": 4.015302929748769e-05,
+      "loss": 0.2974,
+      "step": 3444
+    },
+    {
+      "epoch": 2.5931501693639443,
+      "grad_norm": 0.24243034422397614,
+      "learning_rate": 4.014687600167294e-05,
+      "loss": 0.2314,
+      "step": 3445
+    },
+    {
+      "epoch": 2.593902898005269,
+      "grad_norm": 0.2551737427711487,
+      "learning_rate": 4.014072132281486e-05,
+      "loss": 0.2983,
+      "step": 3446
+    },
+    {
+      "epoch": 2.5946556266465937,
+      "grad_norm": 0.25770553946495056,
+      "learning_rate": 4.013456526158648e-05,
+      "loss": 0.2465,
+      "step": 3447
+    },
+    {
+      "epoch": 2.5954083552879186,
+      "grad_norm": 0.3081754148006439,
+      "learning_rate": 4.0128407818661056e-05,
+      "loss": 0.2967,
+      "step": 3448
+    },
+    {
+      "epoch": 2.5961610839292435,
+      "grad_norm": 0.24311798810958862,
+      "learning_rate": 4.0122248994711934e-05,
+      "loss": 0.2893,
+      "step": 3449
+    },
+    {
+      "epoch": 2.5969138125705684,
+      "grad_norm": 0.21341527998447418,
+      "learning_rate": 4.0116088790412645e-05,
+      "loss": 0.2539,
+      "step": 3450
+    },
+    {
+      "epoch": 2.5976665412118933,
+      "grad_norm": 0.2641579210758209,
+      "learning_rate": 4.0109927206436845e-05,
+      "loss": 0.2787,
+      "step": 3451
+    },
+    {
+      "epoch": 2.598419269853218,
+      "grad_norm": 0.2223798930644989,
+      "learning_rate": 4.0103764243458366e-05,
+      "loss": 0.2958,
+      "step": 3452
+    },
+    {
+      "epoch": 2.5991719984945427,
+      "grad_norm": 0.26034289598464966,
+      "learning_rate": 4.009759990215118e-05,
+      "loss": 0.2787,
+      "step": 3453
+    },
+    {
+      "epoch": 2.5999247271358676,
+      "grad_norm": 0.28103014826774597,
+      "learning_rate": 4.009143418318941e-05,
+      "loss": 0.2359,
+      "step": 3454
+    },
+    {
+      "epoch": 2.600677455777192,
+      "grad_norm": 0.2895321249961853,
+      "learning_rate": 4.008526708724733e-05,
+      "loss": 0.347,
+      "step": 3455
+    },
+    {
+      "epoch": 2.601430184418517,
+      "grad_norm": 0.2274211347103119,
+      "learning_rate": 4.007909861499936e-05,
+      "loss": 0.2426,
+      "step": 3456
+    },
+    {
+      "epoch": 2.602182913059842,
+      "grad_norm": 0.22240139544010162,
+      "learning_rate": 4.007292876712008e-05,
+      "loss": 0.2876,
+      "step": 3457
+    },
+    {
+      "epoch": 2.602935641701167,
+      "grad_norm": 0.2530076503753662,
+      "learning_rate": 4.006675754428421e-05,
+      "loss": 0.347,
+      "step": 3458
+    },
+    {
+      "epoch": 2.603688370342492,
+      "grad_norm": 0.24382919073104858,
+      "learning_rate": 4.006058494716664e-05,
+      "loss": 0.3176,
+      "step": 3459
+    },
+    {
+      "epoch": 2.6044410989838163,
+      "grad_norm": 0.23030689358711243,
+      "learning_rate": 4.005441097644237e-05,
+      "loss": 0.2944,
+      "step": 3460
+    },
+    {
+      "epoch": 2.605193827625141,
+      "grad_norm": 0.2478858381509781,
+      "learning_rate": 4.00482356327866e-05,
+      "loss": 0.2438,
+      "step": 3461
+    },
+    {
+      "epoch": 2.605946556266466,
+      "grad_norm": 0.2246301770210266,
+      "learning_rate": 4.004205891687464e-05,
+      "loss": 0.3737,
+      "step": 3462
+    },
+    {
+      "epoch": 2.6066992849077906,
+      "grad_norm": 0.2667984366416931,
+      "learning_rate": 4.003588082938197e-05,
+      "loss": 0.3335,
+      "step": 3463
+    },
+    {
+      "epoch": 2.6074520135491155,
+      "grad_norm": 0.24492165446281433,
+      "learning_rate": 4.002970137098423e-05,
+      "loss": 0.2816,
+      "step": 3464
+    },
+    {
+      "epoch": 2.6082047421904404,
+      "grad_norm": 0.26918309926986694,
+      "learning_rate": 4.0023520542357184e-05,
+      "loss": 0.2585,
+      "step": 3465
+    },
+    {
+      "epoch": 2.608957470831765,
+      "grad_norm": 0.27562081813812256,
+      "learning_rate": 4.0017338344176765e-05,
+      "loss": 0.2318,
+      "step": 3466
+    },
+    {
+      "epoch": 2.60971019947309,
+      "grad_norm": 0.2549436688423157,
+      "learning_rate": 4.001115477711904e-05,
+      "loss": 0.3093,
+      "step": 3467
+    },
+    {
+      "epoch": 2.6104629281144147,
+      "grad_norm": 0.22586102783679962,
+      "learning_rate": 4.0004969841860235e-05,
+      "loss": 0.238,
+      "step": 3468
+    },
+    {
+      "epoch": 2.6112156567557396,
+      "grad_norm": 0.2371741086244583,
+      "learning_rate": 3.9998783539076725e-05,
+      "loss": 0.3445,
+      "step": 3469
+    },
+    {
+      "epoch": 2.6119683853970646,
+      "grad_norm": 0.22480237483978271,
+      "learning_rate": 3.9992595869445046e-05,
+      "loss": 0.2451,
+      "step": 3470
+    },
+    {
+      "epoch": 2.612721114038389,
+      "grad_norm": 0.24147768318653107,
+      "learning_rate": 3.998640683364187e-05,
+      "loss": 0.4074,
+      "step": 3471
+    },
+    {
+      "epoch": 2.613473842679714,
+      "grad_norm": 0.21943730115890503,
+      "learning_rate": 3.998021643234401e-05,
+      "loss": 0.2833,
+      "step": 3472
+    },
+    {
+      "epoch": 2.614226571321039,
+      "grad_norm": 0.20892123878002167,
+      "learning_rate": 3.997402466622845e-05,
+      "loss": 0.2205,
+      "step": 3473
+    },
+    {
+      "epoch": 2.6149792999623633,
+      "grad_norm": 0.25796136260032654,
+      "learning_rate": 3.99678315359723e-05,
+      "loss": 0.3776,
+      "step": 3474
+    },
+    {
+      "epoch": 2.6157320286036883,
+      "grad_norm": 0.21412737667560577,
+      "learning_rate": 3.9961637042252844e-05,
+      "loss": 0.2874,
+      "step": 3475
+    },
+    {
+      "epoch": 2.616484757245013,
+      "grad_norm": 0.23268014192581177,
+      "learning_rate": 3.995544118574749e-05,
+      "loss": 0.2604,
+      "step": 3476
+    },
+    {
+      "epoch": 2.617237485886338,
+      "grad_norm": 0.29492753744125366,
+      "learning_rate": 3.994924396713382e-05,
+      "loss": 0.3546,
+      "step": 3477
+    },
+    {
+      "epoch": 2.617990214527663,
+      "grad_norm": 0.24745526909828186,
+      "learning_rate": 3.994304538708954e-05,
+      "loss": 0.2315,
+      "step": 3478
+    },
+    {
+      "epoch": 2.6187429431689875,
+      "grad_norm": 0.2502319812774658,
+      "learning_rate": 3.993684544629253e-05,
+      "loss": 0.2041,
+      "step": 3479
+    },
+    {
+      "epoch": 2.6194956718103124,
+      "grad_norm": 0.2286662459373474,
+      "learning_rate": 3.9930644145420806e-05,
+      "loss": 0.2418,
+      "step": 3480
+    },
+    {
+      "epoch": 2.6202484004516373,
+      "grad_norm": 0.23955410718917847,
+      "learning_rate": 3.992444148515251e-05,
+      "loss": 0.2734,
+      "step": 3481
+    },
+    {
+      "epoch": 2.621001129092962,
+      "grad_norm": 0.28931885957717896,
+      "learning_rate": 3.991823746616599e-05,
+      "loss": 0.3522,
+      "step": 3482
+    },
+    {
+      "epoch": 2.6217538577342867,
+      "grad_norm": 0.22793510556221008,
+      "learning_rate": 3.991203208913968e-05,
+      "loss": 0.3192,
+      "step": 3483
+    },
+    {
+      "epoch": 2.6225065863756116,
+      "grad_norm": 0.2513625919818878,
+      "learning_rate": 3.9905825354752206e-05,
+      "loss": 0.2689,
+      "step": 3484
+    },
+    {
+      "epoch": 2.6232593150169365,
+      "grad_norm": 0.2540401816368103,
+      "learning_rate": 3.989961726368232e-05,
+      "loss": 0.3066,
+      "step": 3485
+    },
+    {
+      "epoch": 2.6240120436582615,
+      "grad_norm": 0.30108505487442017,
+      "learning_rate": 3.989340781660894e-05,
+      "loss": 0.3395,
+      "step": 3486
+    },
+    {
+      "epoch": 2.624764772299586,
+      "grad_norm": 0.2451237440109253,
+      "learning_rate": 3.988719701421112e-05,
+      "loss": 0.3324,
+      "step": 3487
+    },
+    {
+      "epoch": 2.625517500940911,
+      "grad_norm": 0.2406981736421585,
+      "learning_rate": 3.988098485716804e-05,
+      "loss": 0.2425,
+      "step": 3488
+    },
+    {
+      "epoch": 2.6262702295822358,
+      "grad_norm": 0.20945517718791962,
+      "learning_rate": 3.987477134615908e-05,
+      "loss": 0.2696,
+      "step": 3489
+    },
+    {
+      "epoch": 2.6270229582235602,
+      "grad_norm": 0.2500711977481842,
+      "learning_rate": 3.986855648186374e-05,
+      "loss": 0.3707,
+      "step": 3490
+    },
+    {
+      "epoch": 2.627775686864885,
+      "grad_norm": 0.2566225230693817,
+      "learning_rate": 3.986234026496165e-05,
+      "loss": 0.2661,
+      "step": 3491
+    },
+    {
+      "epoch": 2.62852841550621,
+      "grad_norm": 0.2383897304534912,
+      "learning_rate": 3.985612269613261e-05,
+      "loss": 0.267,
+      "step": 3492
+    },
+    {
+      "epoch": 2.6292811441475346,
+      "grad_norm": 0.3014627695083618,
+      "learning_rate": 3.9849903776056575e-05,
+      "loss": 0.2333,
+      "step": 3493
+    },
+    {
+      "epoch": 2.6300338727888595,
+      "grad_norm": 0.315241277217865,
+      "learning_rate": 3.984368350541363e-05,
+      "loss": 0.3096,
+      "step": 3494
+    },
+    {
+      "epoch": 2.6307866014301844,
+      "grad_norm": 0.2820964753627777,
+      "learning_rate": 3.9837461884884016e-05,
+      "loss": 0.2654,
+      "step": 3495
+    },
+    {
+      "epoch": 2.6315393300715093,
+      "grad_norm": 0.32629215717315674,
+      "learning_rate": 3.9831238915148114e-05,
+      "loss": 0.3436,
+      "step": 3496
+    },
+    {
+      "epoch": 2.6322920587128342,
+      "grad_norm": 0.26127058267593384,
+      "learning_rate": 3.982501459688647e-05,
+      "loss": 0.2507,
+      "step": 3497
+    },
+    {
+      "epoch": 2.6330447873541587,
+      "grad_norm": 0.3318065106868744,
+      "learning_rate": 3.981878893077976e-05,
+      "loss": 0.2849,
+      "step": 3498
+    },
+    {
+      "epoch": 2.6337975159954836,
+      "grad_norm": 0.3791559338569641,
+      "learning_rate": 3.981256191750882e-05,
+      "loss": 0.3971,
+      "step": 3499
+    },
+    {
+      "epoch": 2.6345502446368085,
+      "grad_norm": 0.2771924138069153,
+      "learning_rate": 3.980633355775461e-05,
+      "loss": 0.3301,
+      "step": 3500
+    },
+    {
+      "epoch": 2.635302973278133,
+      "grad_norm": 0.33888354897499084,
+      "learning_rate": 3.980010385219826e-05,
+      "loss": 0.3782,
+      "step": 3501
+    },
+    {
+      "epoch": 2.636055701919458,
+      "grad_norm": 0.304340124130249,
+      "learning_rate": 3.979387280152106e-05,
+      "loss": 0.4353,
+      "step": 3502
+    },
+    {
+      "epoch": 2.636808430560783,
+      "grad_norm": 0.25690534710884094,
+      "learning_rate": 3.978764040640441e-05,
+      "loss": 0.3015,
+      "step": 3503
+    },
+    {
+      "epoch": 2.6375611592021078,
+      "grad_norm": 0.25201040506362915,
+      "learning_rate": 3.9781406667529876e-05,
+      "loss": 0.2446,
+      "step": 3504
+    },
+    {
+      "epoch": 2.6383138878434327,
+      "grad_norm": 0.24650275707244873,
+      "learning_rate": 3.977517158557917e-05,
+      "loss": 0.2562,
+      "step": 3505
+    },
+    {
+      "epoch": 2.639066616484757,
+      "grad_norm": 0.2960452437400818,
+      "learning_rate": 3.976893516123416e-05,
+      "loss": 0.2728,
+      "step": 3506
+    },
+    {
+      "epoch": 2.639819345126082,
+      "grad_norm": 0.28461337089538574,
+      "learning_rate": 3.976269739517685e-05,
+      "loss": 0.3042,
+      "step": 3507
+    },
+    {
+      "epoch": 2.640572073767407,
+      "grad_norm": 0.2616945803165436,
+      "learning_rate": 3.975645828808938e-05,
+      "loss": 0.382,
+      "step": 3508
+    },
+    {
+      "epoch": 2.6413248024087315,
+      "grad_norm": 0.28477802872657776,
+      "learning_rate": 3.975021784065406e-05,
+      "loss": 0.3986,
+      "step": 3509
+    },
+    {
+      "epoch": 2.6420775310500564,
+      "grad_norm": 0.28344449400901794,
+      "learning_rate": 3.9743976053553345e-05,
+      "loss": 0.3135,
+      "step": 3510
+    },
+    {
+      "epoch": 2.6428302596913813,
+      "grad_norm": 0.22989700734615326,
+      "learning_rate": 3.97377329274698e-05,
+      "loss": 0.2207,
+      "step": 3511
+    },
+    {
+      "epoch": 2.643582988332706,
+      "grad_norm": 0.2747710049152374,
+      "learning_rate": 3.97314884630862e-05,
+      "loss": 0.2862,
+      "step": 3512
+    },
+    {
+      "epoch": 2.644335716974031,
+      "grad_norm": 0.2547527253627777,
+      "learning_rate": 3.97252426610854e-05,
+      "loss": 0.2941,
+      "step": 3513
+    },
+    {
+      "epoch": 2.6450884456153556,
+      "grad_norm": 0.29444754123687744,
+      "learning_rate": 3.971899552215044e-05,
+      "loss": 0.3439,
+      "step": 3514
+    },
+    {
+      "epoch": 2.6458411742566805,
+      "grad_norm": 0.24932143092155457,
+      "learning_rate": 3.971274704696451e-05,
+      "loss": 0.2872,
+      "step": 3515
+    },
+    {
+      "epoch": 2.6465939028980054,
+      "grad_norm": 0.2215253710746765,
+      "learning_rate": 3.970649723621093e-05,
+      "loss": 0.2507,
+      "step": 3516
+    },
+    {
+      "epoch": 2.64734663153933,
+      "grad_norm": 0.35267016291618347,
+      "learning_rate": 3.9700246090573154e-05,
+      "loss": 0.2896,
+      "step": 3517
+    },
+    {
+      "epoch": 2.648099360180655,
+      "grad_norm": 0.219065323472023,
+      "learning_rate": 3.969399361073481e-05,
+      "loss": 0.3298,
+      "step": 3518
+    },
+    {
+      "epoch": 2.6488520888219798,
+      "grad_norm": 0.30664679408073425,
+      "learning_rate": 3.968773979737966e-05,
+      "loss": 0.2958,
+      "step": 3519
+    },
+    {
+      "epoch": 2.6496048174633042,
+      "grad_norm": 0.2787836492061615,
+      "learning_rate": 3.9681484651191604e-05,
+      "loss": 0.2851,
+      "step": 3520
+    },
+    {
+      "epoch": 2.650357546104629,
+      "grad_norm": 0.3042992651462555,
+      "learning_rate": 3.9675228172854704e-05,
+      "loss": 0.3463,
+      "step": 3521
+    },
+    {
+      "epoch": 2.651110274745954,
+      "grad_norm": 0.22463855147361755,
+      "learning_rate": 3.966897036305316e-05,
+      "loss": 0.2726,
+      "step": 3522
+    },
+    {
+      "epoch": 2.651863003387279,
+      "grad_norm": 0.320002943277359,
+      "learning_rate": 3.966271122247131e-05,
+      "loss": 0.2733,
+      "step": 3523
+    },
+    {
+      "epoch": 2.652615732028604,
+      "grad_norm": 0.25215479731559753,
+      "learning_rate": 3.965645075179365e-05,
+      "loss": 0.2925,
+      "step": 3524
+    },
+    {
+      "epoch": 2.6533684606699284,
+      "grad_norm": 0.2223145067691803,
+      "learning_rate": 3.9650188951704816e-05,
+      "loss": 0.3255,
+      "step": 3525
+    },
+    {
+      "epoch": 2.6541211893112533,
+      "grad_norm": 0.24988673627376556,
+      "learning_rate": 3.964392582288958e-05,
+      "loss": 0.2905,
+      "step": 3526
+    },
+    {
+      "epoch": 2.654873917952578,
+      "grad_norm": 0.2308509796857834,
+      "learning_rate": 3.963766136603287e-05,
+      "loss": 0.2452,
+      "step": 3527
+    },
+    {
+      "epoch": 2.6556266465939027,
+      "grad_norm": 0.24484717845916748,
+      "learning_rate": 3.963139558181977e-05,
+      "loss": 0.3051,
+      "step": 3528
+    },
+    {
+      "epoch": 2.6563793752352276,
+      "grad_norm": 0.20186135172843933,
+      "learning_rate": 3.962512847093548e-05,
+      "loss": 0.2037,
+      "step": 3529
+    },
+    {
+      "epoch": 2.6571321038765525,
+      "grad_norm": 0.2637765109539032,
+      "learning_rate": 3.961886003406537e-05,
+      "loss": 0.2019,
+      "step": 3530
+    },
+    {
+      "epoch": 2.6578848325178774,
+      "grad_norm": 0.33848053216934204,
+      "learning_rate": 3.961259027189496e-05,
+      "loss": 0.2319,
+      "step": 3531
+    },
+    {
+      "epoch": 2.6586375611592024,
+      "grad_norm": 0.24014240503311157,
+      "learning_rate": 3.960631918510988e-05,
+      "loss": 0.4178,
+      "step": 3532
+    },
+    {
+      "epoch": 2.659390289800527,
+      "grad_norm": 0.21702338755130768,
+      "learning_rate": 3.960004677439594e-05,
+      "loss": 0.2589,
+      "step": 3533
+    },
+    {
+      "epoch": 2.6601430184418517,
+      "grad_norm": 0.2685118615627289,
+      "learning_rate": 3.959377304043907e-05,
+      "loss": 0.2595,
+      "step": 3534
+    },
+    {
+      "epoch": 2.6608957470831767,
+      "grad_norm": 0.26310834288597107,
+      "learning_rate": 3.958749798392536e-05,
+      "loss": 0.2778,
+      "step": 3535
+    },
+    {
+      "epoch": 2.661648475724501,
+      "grad_norm": 0.2805001139640808,
+      "learning_rate": 3.958122160554106e-05,
+      "loss": 0.374,
+      "step": 3536
+    },
+    {
+      "epoch": 2.662401204365826,
+      "grad_norm": 0.26619958877563477,
+      "learning_rate": 3.957494390597251e-05,
+      "loss": 0.3031,
+      "step": 3537
+    },
+    {
+      "epoch": 2.663153933007151,
+      "grad_norm": 0.2531306743621826,
+      "learning_rate": 3.956866488590626e-05,
+      "loss": 0.2894,
+      "step": 3538
+    },
+    {
+      "epoch": 2.6639066616484754,
+      "grad_norm": 0.25079235434532166,
+      "learning_rate": 3.9562384546028956e-05,
+      "loss": 0.245,
+      "step": 3539
+    },
+    {
+      "epoch": 2.6646593902898004,
+      "grad_norm": 0.3299691081047058,
+      "learning_rate": 3.955610288702742e-05,
+      "loss": 0.3236,
+      "step": 3540
+    },
+    {
+      "epoch": 2.6654121189311253,
+      "grad_norm": 0.207356795668602,
+      "learning_rate": 3.9549819909588605e-05,
+      "loss": 0.261,
+      "step": 3541
+    },
+    {
+      "epoch": 2.66616484757245,
+      "grad_norm": 0.24611522257328033,
+      "learning_rate": 3.954353561439959e-05,
+      "loss": 0.2413,
+      "step": 3542
+    },
+    {
+      "epoch": 2.666917576213775,
+      "grad_norm": 0.21910633146762848,
+      "learning_rate": 3.953725000214763e-05,
+      "loss": 0.2744,
+      "step": 3543
+    },
+    {
+      "epoch": 2.6676703048550996,
+      "grad_norm": 0.2576257288455963,
+      "learning_rate": 3.953096307352011e-05,
+      "loss": 0.3098,
+      "step": 3544
+    },
+    {
+      "epoch": 2.6684230334964245,
+      "grad_norm": 0.2917858064174652,
+      "learning_rate": 3.952467482920456e-05,
+      "loss": 0.2407,
+      "step": 3545
+    },
+    {
+      "epoch": 2.6691757621377494,
+      "grad_norm": 0.23574508726596832,
+      "learning_rate": 3.951838526988864e-05,
+      "loss": 0.2937,
+      "step": 3546
+    },
+    {
+      "epoch": 2.669928490779074,
+      "grad_norm": 0.3329881727695465,
+      "learning_rate": 3.951209439626018e-05,
+      "loss": 0.3385,
+      "step": 3547
+    },
+    {
+      "epoch": 2.670681219420399,
+      "grad_norm": 0.3058014512062073,
+      "learning_rate": 3.950580220900714e-05,
+      "loss": 0.3466,
+      "step": 3548
+    },
+    {
+      "epoch": 2.6714339480617237,
+      "grad_norm": 0.3382442593574524,
+      "learning_rate": 3.949950870881761e-05,
+      "loss": 0.328,
+      "step": 3549
+    },
+    {
+      "epoch": 2.6721866767030487,
+      "grad_norm": 0.36945024132728577,
+      "learning_rate": 3.949321389637986e-05,
+      "loss": 0.3576,
+      "step": 3550
+    },
+    {
+      "epoch": 2.6729394053443736,
+      "grad_norm": 0.25221025943756104,
+      "learning_rate": 3.948691777238227e-05,
+      "loss": 0.2062,
+      "step": 3551
+    },
+    {
+      "epoch": 2.673692133985698,
+      "grad_norm": 0.24682921171188354,
+      "learning_rate": 3.948062033751337e-05,
+      "loss": 0.3304,
+      "step": 3552
+    },
+    {
+      "epoch": 2.674444862627023,
+      "grad_norm": 0.2646491229534149,
+      "learning_rate": 3.9474321592461843e-05,
+      "loss": 0.2887,
+      "step": 3553
+    },
+    {
+      "epoch": 2.675197591268348,
+      "grad_norm": 0.31028321385383606,
+      "learning_rate": 3.946802153791651e-05,
+      "loss": 0.3725,
+      "step": 3554
+    },
+    {
+      "epoch": 2.6759503199096724,
+      "grad_norm": 0.24947962164878845,
+      "learning_rate": 3.946172017456634e-05,
+      "loss": 0.2574,
+      "step": 3555
+    },
+    {
+      "epoch": 2.6767030485509973,
+      "grad_norm": 0.31837379932403564,
+      "learning_rate": 3.9455417503100425e-05,
+      "loss": 0.3505,
+      "step": 3556
+    },
+    {
+      "epoch": 2.677455777192322,
+      "grad_norm": 0.27314189076423645,
+      "learning_rate": 3.9449113524208034e-05,
+      "loss": 0.3195,
+      "step": 3557
+    },
+    {
+      "epoch": 2.678208505833647,
+      "grad_norm": 0.2725825011730194,
+      "learning_rate": 3.9442808238578556e-05,
+      "loss": 0.2453,
+      "step": 3558
+    },
+    {
+      "epoch": 2.678961234474972,
+      "grad_norm": 0.22572451829910278,
+      "learning_rate": 3.943650164690152e-05,
+      "loss": 0.35,
+      "step": 3559
+    },
+    {
+      "epoch": 2.6797139631162965,
+      "grad_norm": 0.23380547761917114,
+      "learning_rate": 3.9430193749866615e-05,
+      "loss": 0.3077,
+      "step": 3560
+    },
+    {
+      "epoch": 2.6804666917576214,
+      "grad_norm": 0.24851225316524506,
+      "learning_rate": 3.942388454816365e-05,
+      "loss": 0.3476,
+      "step": 3561
+    },
+    {
+      "epoch": 2.6812194203989463,
+      "grad_norm": 0.2365153729915619,
+      "learning_rate": 3.9417574042482594e-05,
+      "loss": 0.2491,
+      "step": 3562
+    },
+    {
+      "epoch": 2.681972149040271,
+      "grad_norm": 0.2743854820728302,
+      "learning_rate": 3.9411262233513576e-05,
+      "loss": 0.2976,
+      "step": 3563
+    },
+    {
+      "epoch": 2.6827248776815957,
+      "grad_norm": 0.23201635479927063,
+      "learning_rate": 3.940494912194682e-05,
+      "loss": 0.2136,
+      "step": 3564
+    },
+    {
+      "epoch": 2.6834776063229206,
+      "grad_norm": 0.2081533819437027,
+      "learning_rate": 3.939863470847272e-05,
+      "loss": 0.2307,
+      "step": 3565
+    },
+    {
+      "epoch": 2.684230334964245,
+      "grad_norm": 0.2651883661746979,
+      "learning_rate": 3.939231899378183e-05,
+      "loss": 0.374,
+      "step": 3566
+    },
+    {
+      "epoch": 2.68498306360557,
+      "grad_norm": 0.23451437056064606,
+      "learning_rate": 3.9386001978564805e-05,
+      "loss": 0.3186,
+      "step": 3567
+    },
+    {
+      "epoch": 2.685735792246895,
+      "grad_norm": 0.2105848640203476,
+      "learning_rate": 3.937968366351248e-05,
+      "loss": 0.3855,
+      "step": 3568
+    },
+    {
+      "epoch": 2.68648852088822,
+      "grad_norm": 0.2501204311847687,
+      "learning_rate": 3.937336404931581e-05,
+      "loss": 0.3109,
+      "step": 3569
+    },
+    {
+      "epoch": 2.687241249529545,
+      "grad_norm": 0.23140175640583038,
+      "learning_rate": 3.93670431366659e-05,
+      "loss": 0.4037,
+      "step": 3570
+    },
+    {
+      "epoch": 2.6879939781708693,
+      "grad_norm": 0.3992088735103607,
+      "learning_rate": 3.9360720926253986e-05,
+      "loss": 0.3829,
+      "step": 3571
+    },
+    {
+      "epoch": 2.688746706812194,
+      "grad_norm": 0.23513486981391907,
+      "learning_rate": 3.9354397418771465e-05,
+      "loss": 0.3478,
+      "step": 3572
+    },
+    {
+      "epoch": 2.689499435453519,
+      "grad_norm": 0.20594018697738647,
+      "learning_rate": 3.934807261490988e-05,
+      "loss": 0.2372,
+      "step": 3573
+    },
+    {
+      "epoch": 2.6902521640948436,
+      "grad_norm": 0.29265984892845154,
+      "learning_rate": 3.934174651536087e-05,
+      "loss": 0.3062,
+      "step": 3574
+    },
+    {
+      "epoch": 2.6910048927361685,
+      "grad_norm": 0.2577134668827057,
+      "learning_rate": 3.933541912081627e-05,
+      "loss": 0.3366,
+      "step": 3575
+    },
+    {
+      "epoch": 2.6917576213774934,
+      "grad_norm": 0.23314069211483002,
+      "learning_rate": 3.932909043196803e-05,
+      "loss": 0.2704,
+      "step": 3576
+    },
+    {
+      "epoch": 2.6925103500188183,
+      "grad_norm": 0.26079604029655457,
+      "learning_rate": 3.932276044950824e-05,
+      "loss": 0.343,
+      "step": 3577
+    },
+    {
+      "epoch": 2.6932630786601433,
+      "grad_norm": 0.22584302723407745,
+      "learning_rate": 3.931642917412916e-05,
+      "loss": 0.2587,
+      "step": 3578
+    },
+    {
+      "epoch": 2.6940158073014677,
+      "grad_norm": 0.2576099932193756,
+      "learning_rate": 3.9310096606523137e-05,
+      "loss": 0.282,
+      "step": 3579
+    },
+    {
+      "epoch": 2.6947685359427926,
+      "grad_norm": 0.23436400294303894,
+      "learning_rate": 3.930376274738271e-05,
+      "loss": 0.28,
+      "step": 3580
+    },
+    {
+      "epoch": 2.6955212645841176,
+      "grad_norm": 0.31533944606781006,
+      "learning_rate": 3.9297427597400535e-05,
+      "loss": 0.2697,
+      "step": 3581
+    },
+    {
+      "epoch": 2.696273993225442,
+      "grad_norm": 0.23239733278751373,
+      "learning_rate": 3.929109115726941e-05,
+      "loss": 0.2815,
+      "step": 3582
+    },
+    {
+      "epoch": 2.697026721866767,
+      "grad_norm": 0.23369288444519043,
+      "learning_rate": 3.928475342768229e-05,
+      "loss": 0.2815,
+      "step": 3583
+    },
+    {
+      "epoch": 2.697779450508092,
+      "grad_norm": 0.22463206946849823,
+      "learning_rate": 3.927841440933224e-05,
+      "loss": 0.2819,
+      "step": 3584
+    },
+    {
+      "epoch": 2.698532179149417,
+      "grad_norm": 0.2795897424221039,
+      "learning_rate": 3.927207410291251e-05,
+      "loss": 0.2964,
+      "step": 3585
+    },
+    {
+      "epoch": 2.6992849077907417,
+      "grad_norm": 0.2576623260974884,
+      "learning_rate": 3.926573250911646e-05,
+      "loss": 0.2225,
+      "step": 3586
+    },
+    {
+      "epoch": 2.700037636432066,
+      "grad_norm": 0.24374526739120483,
+      "learning_rate": 3.9259389628637586e-05,
+      "loss": 0.3898,
+      "step": 3587
+    },
+    {
+      "epoch": 2.700790365073391,
+      "grad_norm": 0.2419610470533371,
+      "learning_rate": 3.925304546216954e-05,
+      "loss": 0.2803,
+      "step": 3588
+    },
+    {
+      "epoch": 2.701543093714716,
+      "grad_norm": 0.23057061433792114,
+      "learning_rate": 3.9246700010406116e-05,
+      "loss": 0.3549,
+      "step": 3589
+    },
+    {
+      "epoch": 2.7022958223560405,
+      "grad_norm": 0.23192672431468964,
+      "learning_rate": 3.924035327404124e-05,
+      "loss": 0.3693,
+      "step": 3590
+    },
+    {
+      "epoch": 2.7030485509973654,
+      "grad_norm": 0.23594482243061066,
+      "learning_rate": 3.923400525376898e-05,
+      "loss": 0.2888,
+      "step": 3591
+    },
+    {
+      "epoch": 2.7038012796386903,
+      "grad_norm": 0.25268715620040894,
+      "learning_rate": 3.922765595028355e-05,
+      "loss": 0.37,
+      "step": 3592
+    },
+    {
+      "epoch": 2.704554008280015,
+      "grad_norm": 0.2313959151506424,
+      "learning_rate": 3.9221305364279296e-05,
+      "loss": 0.2702,
+      "step": 3593
+    },
+    {
+      "epoch": 2.7053067369213397,
+      "grad_norm": 0.22732217609882355,
+      "learning_rate": 3.921495349645071e-05,
+      "loss": 0.289,
+      "step": 3594
+    },
+    {
+      "epoch": 2.7060594655626646,
+      "grad_norm": 0.2849101722240448,
+      "learning_rate": 3.920860034749242e-05,
+      "loss": 0.2632,
+      "step": 3595
+    },
+    {
+      "epoch": 2.7068121942039896,
+      "grad_norm": 0.25099506974220276,
+      "learning_rate": 3.920224591809921e-05,
+      "loss": 0.3523,
+      "step": 3596
+    },
+    {
+      "epoch": 2.7075649228453145,
+      "grad_norm": 0.23426488041877747,
+      "learning_rate": 3.9195890208965963e-05,
+      "loss": 0.2116,
+      "step": 3597
+    },
+    {
+      "epoch": 2.708317651486639,
+      "grad_norm": 0.270913302898407,
+      "learning_rate": 3.918953322078777e-05,
+      "loss": 0.3153,
+      "step": 3598
+    },
+    {
+      "epoch": 2.709070380127964,
+      "grad_norm": 0.2645542323589325,
+      "learning_rate": 3.918317495425978e-05,
+      "loss": 0.2897,
+      "step": 3599
+    },
+    {
+      "epoch": 2.709823108769289,
+      "grad_norm": 0.2176583707332611,
+      "learning_rate": 3.917681541007734e-05,
+      "loss": 0.2825,
+      "step": 3600
+    },
+    {
+      "epoch": 2.709823108769289,
+      "eval_loss": 0.2991047501564026,
+      "eval_runtime": 452.5095,
+      "eval_samples_per_second": 21.275,
+      "eval_steps_per_second": 0.665,
+      "step": 3600
+    },
+    {
+      "epoch": 2.7105758374106133,
+      "grad_norm": 0.2480589598417282,
+      "learning_rate": 3.917045458893593e-05,
+      "loss": 0.3943,
+      "step": 3601
+    },
+    {
+      "epoch": 2.711328566051938,
+      "grad_norm": 0.2548377811908722,
+      "learning_rate": 3.9164092491531164e-05,
+      "loss": 0.3722,
+      "step": 3602
+    },
+    {
+      "epoch": 2.712081294693263,
+      "grad_norm": 0.24777573347091675,
+      "learning_rate": 3.915772911855876e-05,
+      "loss": 0.2804,
+      "step": 3603
+    },
+    {
+      "epoch": 2.712834023334588,
+      "grad_norm": 0.2536385655403137,
+      "learning_rate": 3.915136447071464e-05,
+      "loss": 0.2653,
+      "step": 3604
+    },
+    {
+      "epoch": 2.713586751975913,
+      "grad_norm": 0.246423602104187,
+      "learning_rate": 3.914499854869482e-05,
+      "loss": 0.2339,
+      "step": 3605
+    },
+    {
+      "epoch": 2.7143394806172374,
+      "grad_norm": 0.2367265671491623,
+      "learning_rate": 3.9138631353195456e-05,
+      "loss": 0.2337,
+      "step": 3606
+    },
+    {
+      "epoch": 2.7150922092585623,
+      "grad_norm": 0.24878746271133423,
+      "learning_rate": 3.913226288491287e-05,
+      "loss": 0.298,
+      "step": 3607
+    },
+    {
+      "epoch": 2.7158449378998872,
+      "grad_norm": 0.3010980486869812,
+      "learning_rate": 3.91258931445435e-05,
+      "loss": 0.314,
+      "step": 3608
+    },
+    {
+      "epoch": 2.7165976665412117,
+      "grad_norm": 0.2541608512401581,
+      "learning_rate": 3.9119522132783945e-05,
+      "loss": 0.1939,
+      "step": 3609
+    },
+    {
+      "epoch": 2.7173503951825366,
+      "grad_norm": 0.26385048031806946,
+      "learning_rate": 3.9113149850330905e-05,
+      "loss": 0.3115,
+      "step": 3610
+    },
+    {
+      "epoch": 2.7181031238238615,
+      "grad_norm": 0.27963218092918396,
+      "learning_rate": 3.910677629788128e-05,
+      "loss": 0.3386,
+      "step": 3611
+    },
+    {
+      "epoch": 2.7188558524651865,
+      "grad_norm": 0.2614336609840393,
+      "learning_rate": 3.910040147613203e-05,
+      "loss": 0.3617,
+      "step": 3612
+    },
+    {
+      "epoch": 2.7196085811065114,
+      "grad_norm": 0.2729731798171997,
+      "learning_rate": 3.909402538578031e-05,
+      "loss": 0.3618,
+      "step": 3613
+    },
+    {
+      "epoch": 2.720361309747836,
+      "grad_norm": 0.24161653220653534,
+      "learning_rate": 3.908764802752342e-05,
+      "loss": 0.2938,
+      "step": 3614
+    },
+    {
+      "epoch": 2.7211140383891608,
+      "grad_norm": 0.285560667514801,
+      "learning_rate": 3.908126940205877e-05,
+      "loss": 0.2522,
+      "step": 3615
+    },
+    {
+      "epoch": 2.7218667670304857,
+      "grad_norm": 0.2464088648557663,
+      "learning_rate": 3.9074889510083896e-05,
+      "loss": 0.27,
+      "step": 3616
+    },
+    {
+      "epoch": 2.72261949567181,
+      "grad_norm": 0.3400322496891022,
+      "learning_rate": 3.906850835229651e-05,
+      "loss": 0.3592,
+      "step": 3617
+    },
+    {
+      "epoch": 2.723372224313135,
+      "grad_norm": 0.2657867670059204,
+      "learning_rate": 3.906212592939445e-05,
+      "loss": 0.2413,
+      "step": 3618
+    },
+    {
+      "epoch": 2.72412495295446,
+      "grad_norm": 0.3189414441585541,
+      "learning_rate": 3.905574224207567e-05,
+      "loss": 0.3191,
+      "step": 3619
+    },
+    {
+      "epoch": 2.7248776815957845,
+      "grad_norm": 0.2853959798812866,
+      "learning_rate": 3.9049357291038306e-05,
+      "loss": 0.3204,
+      "step": 3620
+    },
+    {
+      "epoch": 2.7256304102371094,
+      "grad_norm": 0.2690480947494507,
+      "learning_rate": 3.9042971076980597e-05,
+      "loss": 0.2613,
+      "step": 3621
+    },
+    {
+      "epoch": 2.7263831388784343,
+      "grad_norm": 0.32683539390563965,
+      "learning_rate": 3.90365836006009e-05,
+      "loss": 0.3009,
+      "step": 3622
+    },
+    {
+      "epoch": 2.7271358675197592,
+      "grad_norm": 0.25338879227638245,
+      "learning_rate": 3.903019486259778e-05,
+      "loss": 0.2496,
+      "step": 3623
+    },
+    {
+      "epoch": 2.727888596161084,
+      "grad_norm": 0.251648873090744,
+      "learning_rate": 3.9023804863669896e-05,
+      "loss": 0.3161,
+      "step": 3624
+    },
+    {
+      "epoch": 2.7286413248024086,
+      "grad_norm": 0.2211790829896927,
+      "learning_rate": 3.901741360451602e-05,
+      "loss": 0.2653,
+      "step": 3625
+    },
+    {
+      "epoch": 2.7293940534437335,
+      "grad_norm": 0.23974771797657013,
+      "learning_rate": 3.901102108583512e-05,
+      "loss": 0.2282,
+      "step": 3626
+    },
+    {
+      "epoch": 2.7301467820850585,
+      "grad_norm": 0.2746560275554657,
+      "learning_rate": 3.9004627308326264e-05,
+      "loss": 0.265,
+      "step": 3627
+    },
+    {
+      "epoch": 2.730899510726383,
+      "grad_norm": 0.2316400706768036,
+      "learning_rate": 3.8998232272688654e-05,
+      "loss": 0.2368,
+      "step": 3628
+    },
+    {
+      "epoch": 2.731652239367708,
+      "grad_norm": 0.2548673152923584,
+      "learning_rate": 3.899183597962165e-05,
+      "loss": 0.3075,
+      "step": 3629
+    },
+    {
+      "epoch": 2.7324049680090328,
+      "grad_norm": 0.2076699286699295,
+      "learning_rate": 3.898543842982474e-05,
+      "loss": 0.3339,
+      "step": 3630
+    },
+    {
+      "epoch": 2.7331576966503577,
+      "grad_norm": 0.2574560046195984,
+      "learning_rate": 3.897903962399755e-05,
+      "loss": 0.285,
+      "step": 3631
+    },
+    {
+      "epoch": 2.7339104252916826,
+      "grad_norm": 0.28012874722480774,
+      "learning_rate": 3.8972639562839855e-05,
+      "loss": 0.2892,
+      "step": 3632
+    },
+    {
+      "epoch": 2.734663153933007,
+      "grad_norm": 0.2342361956834793,
+      "learning_rate": 3.8966238247051535e-05,
+      "loss": 0.3013,
+      "step": 3633
+    },
+    {
+      "epoch": 2.735415882574332,
+      "grad_norm": 0.28346532583236694,
+      "learning_rate": 3.895983567733264e-05,
+      "loss": 0.2783,
+      "step": 3634
+    },
+    {
+      "epoch": 2.736168611215657,
+      "grad_norm": 0.2925916612148285,
+      "learning_rate": 3.8953431854383354e-05,
+      "loss": 0.2409,
+      "step": 3635
+    },
+    {
+      "epoch": 2.7369213398569814,
+      "grad_norm": 0.24446411430835724,
+      "learning_rate": 3.8947026778903976e-05,
+      "loss": 0.2422,
+      "step": 3636
+    },
+    {
+      "epoch": 2.7376740684983063,
+      "grad_norm": 0.2293063998222351,
+      "learning_rate": 3.8940620451594965e-05,
+      "loss": 0.267,
+      "step": 3637
+    },
+    {
+      "epoch": 2.738426797139631,
+      "grad_norm": 0.2665459215641022,
+      "learning_rate": 3.893421287315689e-05,
+      "loss": 0.3219,
+      "step": 3638
+    },
+    {
+      "epoch": 2.7391795257809557,
+      "grad_norm": 0.2172742635011673,
+      "learning_rate": 3.89278040442905e-05,
+      "loss": 0.2969,
+      "step": 3639
+    },
+    {
+      "epoch": 2.7399322544222806,
+      "grad_norm": 0.21320024132728577,
+      "learning_rate": 3.8921393965696636e-05,
+      "loss": 0.3672,
+      "step": 3640
+    },
+    {
+      "epoch": 2.7406849830636055,
+      "grad_norm": 0.23037122189998627,
+      "learning_rate": 3.8914982638076296e-05,
+      "loss": 0.3696,
+      "step": 3641
+    },
+    {
+      "epoch": 2.7414377117049304,
+      "grad_norm": 0.22028382122516632,
+      "learning_rate": 3.890857006213062e-05,
+      "loss": 0.3406,
+      "step": 3642
+    },
+    {
+      "epoch": 2.7421904403462554,
+      "grad_norm": 0.21022477746009827,
+      "learning_rate": 3.8902156238560875e-05,
+      "loss": 0.2914,
+      "step": 3643
+    },
+    {
+      "epoch": 2.74294316898758,
+      "grad_norm": 0.22415976226329803,
+      "learning_rate": 3.889574116806846e-05,
+      "loss": 0.3027,
+      "step": 3644
+    },
+    {
+      "epoch": 2.7436958976289048,
+      "grad_norm": 0.2871667444705963,
+      "learning_rate": 3.888932485135493e-05,
+      "loss": 0.3677,
+      "step": 3645
+    },
+    {
+      "epoch": 2.7444486262702297,
+      "grad_norm": 0.22124063968658447,
+      "learning_rate": 3.888290728912196e-05,
+      "loss": 0.3049,
+      "step": 3646
+    },
+    {
+      "epoch": 2.745201354911554,
+      "grad_norm": 0.2441488355398178,
+      "learning_rate": 3.887648848207136e-05,
+      "loss": 0.3313,
+      "step": 3647
+    },
+    {
+      "epoch": 2.745954083552879,
+      "grad_norm": 0.22638776898384094,
+      "learning_rate": 3.887006843090508e-05,
+      "loss": 0.27,
+      "step": 3648
+    },
+    {
+      "epoch": 2.746706812194204,
+      "grad_norm": 0.221455380320549,
+      "learning_rate": 3.886364713632522e-05,
+      "loss": 0.2542,
+      "step": 3649
+    },
+    {
+      "epoch": 2.747459540835529,
+      "grad_norm": 0.2531570792198181,
+      "learning_rate": 3.885722459903399e-05,
+      "loss": 0.3153,
+      "step": 3650
+    },
+    {
+      "epoch": 2.748212269476854,
+      "grad_norm": 0.2315714806318283,
+      "learning_rate": 3.8850800819733744e-05,
+      "loss": 0.2657,
+      "step": 3651
+    },
+    {
+      "epoch": 2.7489649981181783,
+      "grad_norm": 0.24352499842643738,
+      "learning_rate": 3.8844375799126995e-05,
+      "loss": 0.3441,
+      "step": 3652
+    },
+    {
+      "epoch": 2.749717726759503,
+      "grad_norm": 0.24344947934150696,
+      "learning_rate": 3.883794953791637e-05,
+      "loss": 0.2456,
+      "step": 3653
+    },
+    {
+      "epoch": 2.750470455400828,
+      "grad_norm": 0.253471314907074,
+      "learning_rate": 3.883152203680462e-05,
+      "loss": 0.3293,
+      "step": 3654
+    },
+    {
+      "epoch": 2.7512231840421526,
+      "grad_norm": 0.2521964907646179,
+      "learning_rate": 3.8825093296494664e-05,
+      "loss": 0.3364,
+      "step": 3655
+    },
+    {
+      "epoch": 2.7519759126834775,
+      "grad_norm": 0.24777156114578247,
+      "learning_rate": 3.8818663317689526e-05,
+      "loss": 0.2622,
+      "step": 3656
+    },
+    {
+      "epoch": 2.7527286413248024,
+      "grad_norm": 0.23125280439853668,
+      "learning_rate": 3.881223210109238e-05,
+      "loss": 0.2832,
+      "step": 3657
+    },
+    {
+      "epoch": 2.7534813699661274,
+      "grad_norm": 0.23260949552059174,
+      "learning_rate": 3.880579964740656e-05,
+      "loss": 0.3551,
+      "step": 3658
+    },
+    {
+      "epoch": 2.7542340986074523,
+      "grad_norm": 0.2624397873878479,
+      "learning_rate": 3.879936595733547e-05,
+      "loss": 0.2739,
+      "step": 3659
+    },
+    {
+      "epoch": 2.7549868272487767,
+      "grad_norm": 0.2725694179534912,
+      "learning_rate": 3.8792931031582705e-05,
+      "loss": 0.3453,
+      "step": 3660
+    },
+    {
+      "epoch": 2.7557395558901017,
+      "grad_norm": 0.24067136645317078,
+      "learning_rate": 3.8786494870852e-05,
+      "loss": 0.3824,
+      "step": 3661
+    },
+    {
+      "epoch": 2.7564922845314266,
+      "grad_norm": 0.26064398884773254,
+      "learning_rate": 3.878005747584717e-05,
+      "loss": 0.2836,
+      "step": 3662
+    },
+    {
+      "epoch": 2.757245013172751,
+      "grad_norm": 0.2591152787208557,
+      "learning_rate": 3.877361884727222e-05,
+      "loss": 0.2376,
+      "step": 3663
+    },
+    {
+      "epoch": 2.757997741814076,
+      "grad_norm": 0.23061545193195343,
+      "learning_rate": 3.876717898583126e-05,
+      "loss": 0.2736,
+      "step": 3664
+    },
+    {
+      "epoch": 2.758750470455401,
+      "grad_norm": 0.279764860868454,
+      "learning_rate": 3.876073789222855e-05,
+      "loss": 0.277,
+      "step": 3665
+    },
+    {
+      "epoch": 2.7595031990967254,
+      "grad_norm": 0.21765218675136566,
+      "learning_rate": 3.8754295567168466e-05,
+      "loss": 0.2978,
+      "step": 3666
+    },
+    {
+      "epoch": 2.7602559277380503,
+      "grad_norm": 0.24668195843696594,
+      "learning_rate": 3.874785201135554e-05,
+      "loss": 0.3601,
+      "step": 3667
+    },
+    {
+      "epoch": 2.761008656379375,
+      "grad_norm": 0.26271045207977295,
+      "learning_rate": 3.874140722549443e-05,
+      "loss": 0.2559,
+      "step": 3668
+    },
+    {
+      "epoch": 2.7617613850207,
+      "grad_norm": 0.28288859128952026,
+      "learning_rate": 3.873496121028993e-05,
+      "loss": 0.2617,
+      "step": 3669
+    },
+    {
+      "epoch": 2.762514113662025,
+      "grad_norm": 0.2979210317134857,
+      "learning_rate": 3.8728513966446964e-05,
+      "loss": 0.3058,
+      "step": 3670
+    },
+    {
+      "epoch": 2.7632668423033495,
+      "grad_norm": 0.27705296874046326,
+      "learning_rate": 3.872206549467059e-05,
+      "loss": 0.3524,
+      "step": 3671
+    },
+    {
+      "epoch": 2.7640195709446744,
+      "grad_norm": 0.21010646224021912,
+      "learning_rate": 3.871561579566601e-05,
+      "loss": 0.2242,
+      "step": 3672
+    },
+    {
+      "epoch": 2.7647722995859993,
+      "grad_norm": 0.2723652124404907,
+      "learning_rate": 3.870916487013854e-05,
+      "loss": 0.2789,
+      "step": 3673
+    },
+    {
+      "epoch": 2.765525028227324,
+      "grad_norm": 0.2524096369743347,
+      "learning_rate": 3.870271271879367e-05,
+      "loss": 0.3205,
+      "step": 3674
+    },
+    {
+      "epoch": 2.7662777568686487,
+      "grad_norm": 0.2502915561199188,
+      "learning_rate": 3.869625934233696e-05,
+      "loss": 0.2689,
+      "step": 3675
+    },
+    {
+      "epoch": 2.7670304855099737,
+      "grad_norm": 0.22075240314006805,
+      "learning_rate": 3.868980474147417e-05,
+      "loss": 0.2276,
+      "step": 3676
+    },
+    {
+      "epoch": 2.7677832141512986,
+      "grad_norm": 0.24334414303302765,
+      "learning_rate": 3.868334891691117e-05,
+      "loss": 0.2729,
+      "step": 3677
+    },
+    {
+      "epoch": 2.7685359427926235,
+      "grad_norm": 0.3428821861743927,
+      "learning_rate": 3.867689186935392e-05,
+      "loss": 0.2674,
+      "step": 3678
+    },
+    {
+      "epoch": 2.769288671433948,
+      "grad_norm": 0.27191048860549927,
+      "learning_rate": 3.86704335995086e-05,
+      "loss": 0.2869,
+      "step": 3679
+    },
+    {
+      "epoch": 2.770041400075273,
+      "grad_norm": 0.22788488864898682,
+      "learning_rate": 3.866397410808145e-05,
+      "loss": 0.2144,
+      "step": 3680
+    },
+    {
+      "epoch": 2.770794128716598,
+      "grad_norm": 0.257222443819046,
+      "learning_rate": 3.865751339577887e-05,
+      "loss": 0.2823,
+      "step": 3681
+    },
+    {
+      "epoch": 2.7715468573579223,
+      "grad_norm": 0.25115543603897095,
+      "learning_rate": 3.865105146330741e-05,
+      "loss": 0.2135,
+      "step": 3682
+    },
+    {
+      "epoch": 2.772299585999247,
+      "grad_norm": 0.24263986945152283,
+      "learning_rate": 3.864458831137372e-05,
+      "loss": 0.2876,
+      "step": 3683
+    },
+    {
+      "epoch": 2.773052314640572,
+      "grad_norm": 0.22858886420726776,
+      "learning_rate": 3.863812394068461e-05,
+      "loss": 0.3066,
+      "step": 3684
+    },
+    {
+      "epoch": 2.773805043281897,
+      "grad_norm": 0.21354332566261292,
+      "learning_rate": 3.863165835194702e-05,
+      "loss": 0.373,
+      "step": 3685
+    },
+    {
+      "epoch": 2.774557771923222,
+      "grad_norm": 0.22343797981739044,
+      "learning_rate": 3.8625191545868e-05,
+      "loss": 0.2408,
+      "step": 3686
+    },
+    {
+      "epoch": 2.7753105005645464,
+      "grad_norm": 0.18390880525112152,
+      "learning_rate": 3.861872352315476e-05,
+      "loss": 0.2679,
+      "step": 3687
+    },
+    {
+      "epoch": 2.7760632292058713,
+      "grad_norm": 0.30156323313713074,
+      "learning_rate": 3.861225428451463e-05,
+      "loss": 0.3585,
+      "step": 3688
+    },
+    {
+      "epoch": 2.7768159578471963,
+      "grad_norm": 0.29958465695381165,
+      "learning_rate": 3.8605783830655087e-05,
+      "loss": 0.2994,
+      "step": 3689
+    },
+    {
+      "epoch": 2.7775686864885207,
+      "grad_norm": 0.3283798396587372,
+      "learning_rate": 3.8599312162283714e-05,
+      "loss": 0.3408,
+      "step": 3690
+    },
+    {
+      "epoch": 2.7783214151298457,
+      "grad_norm": 0.27597111463546753,
+      "learning_rate": 3.8592839280108254e-05,
+      "loss": 0.3166,
+      "step": 3691
+    },
+    {
+      "epoch": 2.7790741437711706,
+      "grad_norm": 0.2642899453639984,
+      "learning_rate": 3.858636518483656e-05,
+      "loss": 0.2909,
+      "step": 3692
+    },
+    {
+      "epoch": 2.779826872412495,
+      "grad_norm": 0.2457503229379654,
+      "learning_rate": 3.8579889877176644e-05,
+      "loss": 0.2739,
+      "step": 3693
+    },
+    {
+      "epoch": 2.78057960105382,
+      "grad_norm": 0.3533354103565216,
+      "learning_rate": 3.857341335783663e-05,
+      "loss": 0.2986,
+      "step": 3694
+    },
+    {
+      "epoch": 2.781332329695145,
+      "grad_norm": 0.2315247356891632,
+      "learning_rate": 3.856693562752478e-05,
+      "loss": 0.2869,
+      "step": 3695
+    },
+    {
+      "epoch": 2.78208505833647,
+      "grad_norm": 0.26878514885902405,
+      "learning_rate": 3.856045668694949e-05,
+      "loss": 0.296,
+      "step": 3696
+    },
+    {
+      "epoch": 2.7828377869777947,
+      "grad_norm": 0.25236862897872925,
+      "learning_rate": 3.8553976536819285e-05,
+      "loss": 0.2671,
+      "step": 3697
+    },
+    {
+      "epoch": 2.783590515619119,
+      "grad_norm": 0.25019967555999756,
+      "learning_rate": 3.854749517784282e-05,
+      "loss": 0.3303,
+      "step": 3698
+    },
+    {
+      "epoch": 2.784343244260444,
+      "grad_norm": 0.21147629618644714,
+      "learning_rate": 3.85410126107289e-05,
+      "loss": 0.3834,
+      "step": 3699
+    },
+    {
+      "epoch": 2.785095972901769,
+      "grad_norm": 0.23724298179149628,
+      "learning_rate": 3.853452883618644e-05,
+      "loss": 0.18,
+      "step": 3700
+    },
+    {
+      "epoch": 2.7858487015430935,
+      "grad_norm": 0.26697710156440735,
+      "learning_rate": 3.852804385492449e-05,
+      "loss": 0.3216,
+      "step": 3701
+    },
+    {
+      "epoch": 2.7866014301844184,
+      "grad_norm": 0.25657081604003906,
+      "learning_rate": 3.852155766765225e-05,
+      "loss": 0.2611,
+      "step": 3702
+    },
+    {
+      "epoch": 2.7873541588257433,
+      "grad_norm": 0.2387162297964096,
+      "learning_rate": 3.851507027507903e-05,
+      "loss": 0.3322,
+      "step": 3703
+    },
+    {
+      "epoch": 2.7881068874670683,
+      "grad_norm": 0.22775495052337646,
+      "learning_rate": 3.850858167791428e-05,
+      "loss": 0.3297,
+      "step": 3704
+    },
+    {
+      "epoch": 2.788859616108393,
+      "grad_norm": 0.2296811193227768,
+      "learning_rate": 3.85020918768676e-05,
+      "loss": 0.3149,
+      "step": 3705
+    },
+    {
+      "epoch": 2.7896123447497176,
+      "grad_norm": 0.25425487756729126,
+      "learning_rate": 3.849560087264869e-05,
+      "loss": 0.2739,
+      "step": 3706
+    },
+    {
+      "epoch": 2.7903650733910426,
+      "grad_norm": 0.2460237741470337,
+      "learning_rate": 3.84891086659674e-05,
+      "loss": 0.3451,
+      "step": 3707
+    },
+    {
+      "epoch": 2.7911178020323675,
+      "grad_norm": 0.30951493978500366,
+      "learning_rate": 3.848261525753371e-05,
+      "loss": 0.2884,
+      "step": 3708
+    },
+    {
+      "epoch": 2.791870530673692,
+      "grad_norm": 0.20940729975700378,
+      "learning_rate": 3.847612064805773e-05,
+      "loss": 0.2388,
+      "step": 3709
+    },
+    {
+      "epoch": 2.792623259315017,
+      "grad_norm": 0.30875781178474426,
+      "learning_rate": 3.84696248382497e-05,
+      "loss": 0.3119,
+      "step": 3710
+    },
+    {
+      "epoch": 2.793375987956342,
+      "grad_norm": 0.2563682496547699,
+      "learning_rate": 3.846312782881998e-05,
+      "loss": 0.3792,
+      "step": 3711
+    },
+    {
+      "epoch": 2.7941287165976667,
+      "grad_norm": 0.2431015521287918,
+      "learning_rate": 3.845662962047909e-05,
+      "loss": 0.27,
+      "step": 3712
+    },
+    {
+      "epoch": 2.794881445238991,
+      "grad_norm": 0.2527167797088623,
+      "learning_rate": 3.8450130213937654e-05,
+      "loss": 0.3285,
+      "step": 3713
+    },
+    {
+      "epoch": 2.795634173880316,
+      "grad_norm": 0.2486974000930786,
+      "learning_rate": 3.844362960990644e-05,
+      "loss": 0.2691,
+      "step": 3714
+    },
+    {
+      "epoch": 2.796386902521641,
+      "grad_norm": 0.214739128947258,
+      "learning_rate": 3.843712780909635e-05,
+      "loss": 0.3038,
+      "step": 3715
+    },
+    {
+      "epoch": 2.797139631162966,
+      "grad_norm": 0.2666488587856293,
+      "learning_rate": 3.8430624812218406e-05,
+      "loss": 0.3194,
+      "step": 3716
+    },
+    {
+      "epoch": 2.7978923598042904,
+      "grad_norm": 0.21701394021511078,
+      "learning_rate": 3.842412061998377e-05,
+      "loss": 0.1925,
+      "step": 3717
+    },
+    {
+      "epoch": 2.7986450884456153,
+      "grad_norm": 0.9907800555229187,
+      "learning_rate": 3.841761523310372e-05,
+      "loss": 0.3547,
+      "step": 3718
+    },
+    {
+      "epoch": 2.7993978170869402,
+      "grad_norm": 0.2496812790632248,
+      "learning_rate": 3.841110865228968e-05,
+      "loss": 0.2934,
+      "step": 3719
+    },
+    {
+      "epoch": 2.8001505457282647,
+      "grad_norm": 0.3547074794769287,
+      "learning_rate": 3.8404600878253214e-05,
+      "loss": 0.2623,
+      "step": 3720
+    },
+    {
+      "epoch": 2.8009032743695896,
+      "grad_norm": 0.2559802532196045,
+      "learning_rate": 3.839809191170598e-05,
+      "loss": 0.3064,
+      "step": 3721
+    },
+    {
+      "epoch": 2.8016560030109146,
+      "grad_norm": 0.25044018030166626,
+      "learning_rate": 3.8391581753359806e-05,
+      "loss": 0.2181,
+      "step": 3722
+    },
+    {
+      "epoch": 2.8024087316522395,
+      "grad_norm": 0.22955913841724396,
+      "learning_rate": 3.8385070403926625e-05,
+      "loss": 0.3413,
+      "step": 3723
+    },
+    {
+      "epoch": 2.8031614602935644,
+      "grad_norm": 0.24086348712444305,
+      "learning_rate": 3.837855786411851e-05,
+      "loss": 0.3072,
+      "step": 3724
+    },
+    {
+      "epoch": 2.803914188934889,
+      "grad_norm": 0.24002593755722046,
+      "learning_rate": 3.837204413464767e-05,
+      "loss": 0.3578,
+      "step": 3725
+    },
+    {
+      "epoch": 2.804666917576214,
+      "grad_norm": 0.2448420226573944,
+      "learning_rate": 3.8365529216226413e-05,
+      "loss": 0.3225,
+      "step": 3726
+    },
+    {
+      "epoch": 2.8054196462175387,
+      "grad_norm": 0.2386665940284729,
+      "learning_rate": 3.8359013109567234e-05,
+      "loss": 0.2464,
+      "step": 3727
+    },
+    {
+      "epoch": 2.806172374858863,
+      "grad_norm": 0.25235044956207275,
+      "learning_rate": 3.835249581538269e-05,
+      "loss": 0.2794,
+      "step": 3728
+    },
+    {
+      "epoch": 2.806925103500188,
+      "grad_norm": 0.24439798295497894,
+      "learning_rate": 3.834597733438553e-05,
+      "loss": 0.3118,
+      "step": 3729
+    },
+    {
+      "epoch": 2.807677832141513,
+      "grad_norm": 0.21557587385177612,
+      "learning_rate": 3.8339457667288595e-05,
+      "loss": 0.3222,
+      "step": 3730
+    },
+    {
+      "epoch": 2.808430560782838,
+      "grad_norm": 0.2560613453388214,
+      "learning_rate": 3.833293681480487e-05,
+      "loss": 0.2845,
+      "step": 3731
+    },
+    {
+      "epoch": 2.809183289424163,
+      "grad_norm": 0.2858146131038666,
+      "learning_rate": 3.8326414777647455e-05,
+      "loss": 0.3369,
+      "step": 3732
+    },
+    {
+      "epoch": 2.8099360180654873,
+      "grad_norm": 0.24554209411144257,
+      "learning_rate": 3.8319891556529595e-05,
+      "loss": 0.3177,
+      "step": 3733
+    },
+    {
+      "epoch": 2.8106887467068122,
+      "grad_norm": 0.2861701250076294,
+      "learning_rate": 3.831336715216467e-05,
+      "loss": 0.2044,
+      "step": 3734
+    },
+    {
+      "epoch": 2.811441475348137,
+      "grad_norm": 0.23408035933971405,
+      "learning_rate": 3.830684156526617e-05,
+      "loss": 0.3483,
+      "step": 3735
+    },
+    {
+      "epoch": 2.8121942039894616,
+      "grad_norm": 0.30416014790534973,
+      "learning_rate": 3.8300314796547725e-05,
+      "loss": 0.2928,
+      "step": 3736
+    },
+    {
+      "epoch": 2.8129469326307865,
+      "grad_norm": 0.23542504012584686,
+      "learning_rate": 3.8293786846723087e-05,
+      "loss": 0.3547,
+      "step": 3737
+    },
+    {
+      "epoch": 2.8136996612721115,
+      "grad_norm": 0.22702156007289886,
+      "learning_rate": 3.828725771650615e-05,
+      "loss": 0.2589,
+      "step": 3738
+    },
+    {
+      "epoch": 2.814452389913436,
+      "grad_norm": 0.25573328137397766,
+      "learning_rate": 3.828072740661093e-05,
+      "loss": 0.3034,
+      "step": 3739
+    },
+    {
+      "epoch": 2.815205118554761,
+      "grad_norm": 0.22040031850337982,
+      "learning_rate": 3.8274195917751565e-05,
+      "loss": 0.2404,
+      "step": 3740
+    },
+    {
+      "epoch": 2.8159578471960858,
+      "grad_norm": 0.24623963236808777,
+      "learning_rate": 3.826766325064234e-05,
+      "loss": 0.1888,
+      "step": 3741
+    },
+    {
+      "epoch": 2.8167105758374107,
+      "grad_norm": 0.295602411031723,
+      "learning_rate": 3.826112940599764e-05,
+      "loss": 0.2223,
+      "step": 3742
+    },
+    {
+      "epoch": 2.8174633044787356,
+      "grad_norm": 0.2960558831691742,
+      "learning_rate": 3.8254594384532014e-05,
+      "loss": 0.2608,
+      "step": 3743
+    },
+    {
+      "epoch": 2.81821603312006,
+      "grad_norm": 0.28901544213294983,
+      "learning_rate": 3.8248058186960116e-05,
+      "loss": 0.4014,
+      "step": 3744
+    },
+    {
+      "epoch": 2.818968761761385,
+      "grad_norm": 0.31668129563331604,
+      "learning_rate": 3.8241520813996736e-05,
+      "loss": 0.3169,
+      "step": 3745
+    },
+    {
+      "epoch": 2.81972149040271,
+      "grad_norm": 0.2848742604255676,
+      "learning_rate": 3.823498226635678e-05,
+      "loss": 0.3033,
+      "step": 3746
+    },
+    {
+      "epoch": 2.8204742190440344,
+      "grad_norm": 0.27320927381515503,
+      "learning_rate": 3.822844254475531e-05,
+      "loss": 0.3169,
+      "step": 3747
+    },
+    {
+      "epoch": 2.8212269476853593,
+      "grad_norm": 0.26798924803733826,
+      "learning_rate": 3.82219016499075e-05,
+      "loss": 0.305,
+      "step": 3748
+    },
+    {
+      "epoch": 2.8219796763266842,
+      "grad_norm": 0.29395097494125366,
+      "learning_rate": 3.821535958252863e-05,
+      "loss": 0.2659,
+      "step": 3749
+    },
+    {
+      "epoch": 2.822732404968009,
+      "grad_norm": 0.21879392862319946,
+      "learning_rate": 3.8208816343334156e-05,
+      "loss": 0.2059,
+      "step": 3750
+    },
+    {
+      "epoch": 2.823485133609334,
+      "grad_norm": 0.44077804684638977,
+      "learning_rate": 3.820227193303962e-05,
+      "loss": 0.3746,
+      "step": 3751
+    },
+    {
+      "epoch": 2.8242378622506585,
+      "grad_norm": 0.26222801208496094,
+      "learning_rate": 3.8195726352360734e-05,
+      "loss": 0.3198,
+      "step": 3752
+    },
+    {
+      "epoch": 2.8249905908919835,
+      "grad_norm": 0.2694430351257324,
+      "learning_rate": 3.818917960201328e-05,
+      "loss": 0.1678,
+      "step": 3753
+    },
+    {
+      "epoch": 2.8257433195333084,
+      "grad_norm": 0.3461925685405731,
+      "learning_rate": 3.818263168271322e-05,
+      "loss": 0.2554,
+      "step": 3754
+    },
+    {
+      "epoch": 2.826496048174633,
+      "grad_norm": 0.25951993465423584,
+      "learning_rate": 3.817608259517662e-05,
+      "loss": 0.2892,
+      "step": 3755
+    },
+    {
+      "epoch": 2.8272487768159578,
+      "grad_norm": 0.26362621784210205,
+      "learning_rate": 3.8169532340119665e-05,
+      "loss": 0.2874,
+      "step": 3756
+    },
+    {
+      "epoch": 2.8280015054572827,
+      "grad_norm": 0.4425596296787262,
+      "learning_rate": 3.816298091825872e-05,
+      "loss": 0.383,
+      "step": 3757
+    },
+    {
+      "epoch": 2.8287542340986076,
+      "grad_norm": 0.32583147287368774,
+      "learning_rate": 3.81564283303102e-05,
+      "loss": 0.3663,
+      "step": 3758
+    },
+    {
+      "epoch": 2.8295069627399325,
+      "grad_norm": 0.309560626745224,
+      "learning_rate": 3.814987457699069e-05,
+      "loss": 0.2597,
+      "step": 3759
+    },
+    {
+      "epoch": 2.830259691381257,
+      "grad_norm": 0.3251565992832184,
+      "learning_rate": 3.8143319659016925e-05,
+      "loss": 0.24,
+      "step": 3760
+    },
+    {
+      "epoch": 2.831012420022582,
+      "grad_norm": 0.27726542949676514,
+      "learning_rate": 3.8136763577105715e-05,
+      "loss": 0.2809,
+      "step": 3761
+    },
+    {
+      "epoch": 2.831765148663907,
+      "grad_norm": 0.3494848906993866,
+      "learning_rate": 3.813020633197403e-05,
+      "loss": 0.2137,
+      "step": 3762
+    },
+    {
+      "epoch": 2.8325178773052313,
+      "grad_norm": 0.3233604431152344,
+      "learning_rate": 3.812364792433897e-05,
+      "loss": 0.2657,
+      "step": 3763
+    },
+    {
+      "epoch": 2.833270605946556,
+      "grad_norm": 0.28585436940193176,
+      "learning_rate": 3.8117088354917744e-05,
+      "loss": 0.2665,
+      "step": 3764
+    },
+    {
+      "epoch": 2.834023334587881,
+      "grad_norm": 0.2441757768392563,
+      "learning_rate": 3.8110527624427696e-05,
+      "loss": 0.3014,
+      "step": 3765
+    },
+    {
+      "epoch": 2.8347760632292056,
+      "grad_norm": 0.26518505811691284,
+      "learning_rate": 3.81039657335863e-05,
+      "loss": 0.331,
+      "step": 3766
+    },
+    {
+      "epoch": 2.8355287918705305,
+      "grad_norm": 0.28802230954170227,
+      "learning_rate": 3.809740268311116e-05,
+      "loss": 0.2373,
+      "step": 3767
+    },
+    {
+      "epoch": 2.8362815205118554,
+      "grad_norm": 0.23980435729026794,
+      "learning_rate": 3.809083847371999e-05,
+      "loss": 0.3275,
+      "step": 3768
+    },
+    {
+      "epoch": 2.8370342491531804,
+      "grad_norm": 0.266666054725647,
+      "learning_rate": 3.808427310613065e-05,
+      "loss": 0.2902,
+      "step": 3769
+    },
+    {
+      "epoch": 2.8377869777945053,
+      "grad_norm": 0.338029146194458,
+      "learning_rate": 3.807770658106111e-05,
+      "loss": 0.343,
+      "step": 3770
+    },
+    {
+      "epoch": 2.8385397064358298,
+      "grad_norm": 0.26167619228363037,
+      "learning_rate": 3.807113889922949e-05,
+      "loss": 0.2667,
+      "step": 3771
+    },
+    {
+      "epoch": 2.8392924350771547,
+      "grad_norm": 0.2681350111961365,
+      "learning_rate": 3.8064570061354e-05,
+      "loss": 0.3415,
+      "step": 3772
+    },
+    {
+      "epoch": 2.8400451637184796,
+      "grad_norm": 0.278018057346344,
+      "learning_rate": 3.8058000068153024e-05,
+      "loss": 0.3064,
+      "step": 3773
+    },
+    {
+      "epoch": 2.840797892359804,
+      "grad_norm": 0.25901395082473755,
+      "learning_rate": 3.805142892034503e-05,
+      "loss": 0.21,
+      "step": 3774
+    },
+    {
+      "epoch": 2.841550621001129,
+      "grad_norm": 0.2433588206768036,
+      "learning_rate": 3.804485661864864e-05,
+      "loss": 0.2604,
+      "step": 3775
+    },
+    {
+      "epoch": 2.842303349642454,
+      "grad_norm": 0.30719175934791565,
+      "learning_rate": 3.803828316378257e-05,
+      "loss": 0.3906,
+      "step": 3776
+    },
+    {
+      "epoch": 2.843056078283779,
+      "grad_norm": 0.2846048176288605,
+      "learning_rate": 3.803170855646571e-05,
+      "loss": 0.2802,
+      "step": 3777
+    },
+    {
+      "epoch": 2.8438088069251037,
+      "grad_norm": 0.23801067471504211,
+      "learning_rate": 3.802513279741702e-05,
+      "loss": 0.2217,
+      "step": 3778
+    },
+    {
+      "epoch": 2.844561535566428,
+      "grad_norm": 0.23296011984348297,
+      "learning_rate": 3.801855588735565e-05,
+      "loss": 0.3178,
+      "step": 3779
+    },
+    {
+      "epoch": 2.845314264207753,
+      "grad_norm": 0.2920059561729431,
+      "learning_rate": 3.801197782700082e-05,
+      "loss": 0.3184,
+      "step": 3780
+    },
+    {
+      "epoch": 2.846066992849078,
+      "grad_norm": 0.25966864824295044,
+      "learning_rate": 3.800539861707189e-05,
+      "loss": 0.4434,
+      "step": 3781
+    },
+    {
+      "epoch": 2.8468197214904025,
+      "grad_norm": 0.30799803137779236,
+      "learning_rate": 3.7998818258288374e-05,
+      "loss": 0.2641,
+      "step": 3782
+    },
+    {
+      "epoch": 2.8475724501317274,
+      "grad_norm": 0.2706291675567627,
+      "learning_rate": 3.799223675136987e-05,
+      "loss": 0.2481,
+      "step": 3783
+    },
+    {
+      "epoch": 2.8483251787730524,
+      "grad_norm": 0.2876772880554199,
+      "learning_rate": 3.7985654097036134e-05,
+      "loss": 0.2451,
+      "step": 3784
+    },
+    {
+      "epoch": 2.8490779074143773,
+      "grad_norm": 0.2542955279350281,
+      "learning_rate": 3.797907029600702e-05,
+      "loss": 0.2147,
+      "step": 3785
+    },
+    {
+      "epoch": 2.849830636055702,
+      "grad_norm": 0.28877636790275574,
+      "learning_rate": 3.797248534900255e-05,
+      "loss": 0.3087,
+      "step": 3786
+    },
+    {
+      "epoch": 2.8505833646970267,
+      "grad_norm": 0.30618011951446533,
+      "learning_rate": 3.7965899256742814e-05,
+      "loss": 0.2765,
+      "step": 3787
+    },
+    {
+      "epoch": 2.8513360933383516,
+      "grad_norm": 0.29410940408706665,
+      "learning_rate": 3.7959312019948074e-05,
+      "loss": 0.3218,
+      "step": 3788
+    },
+    {
+      "epoch": 2.8520888219796765,
+      "grad_norm": 0.2509450614452362,
+      "learning_rate": 3.79527236393387e-05,
+      "loss": 0.2889,
+      "step": 3789
+    },
+    {
+      "epoch": 2.852841550621001,
+      "grad_norm": 0.24902956187725067,
+      "learning_rate": 3.794613411563519e-05,
+      "loss": 0.1777,
+      "step": 3790
+    },
+    {
+      "epoch": 2.853594279262326,
+      "grad_norm": 0.2341965287923813,
+      "learning_rate": 3.7939543449558145e-05,
+      "loss": 0.2302,
+      "step": 3791
+    },
+    {
+      "epoch": 2.854347007903651,
+      "grad_norm": 0.27365297079086304,
+      "learning_rate": 3.793295164182833e-05,
+      "loss": 0.2465,
+      "step": 3792
+    },
+    {
+      "epoch": 2.8550997365449753,
+      "grad_norm": 0.252591997385025,
+      "learning_rate": 3.7926358693166616e-05,
+      "loss": 0.2044,
+      "step": 3793
+    },
+    {
+      "epoch": 2.8558524651863,
+      "grad_norm": 0.29585039615631104,
+      "learning_rate": 3.7919764604293976e-05,
+      "loss": 0.3046,
+      "step": 3794
+    },
+    {
+      "epoch": 2.856605193827625,
+      "grad_norm": 0.29851946234703064,
+      "learning_rate": 3.7913169375931555e-05,
+      "loss": 0.4257,
+      "step": 3795
+    },
+    {
+      "epoch": 2.85735792246895,
+      "grad_norm": 0.28401631116867065,
+      "learning_rate": 3.790657300880059e-05,
+      "loss": 0.2781,
+      "step": 3796
+    },
+    {
+      "epoch": 2.858110651110275,
+      "grad_norm": 0.2585422396659851,
+      "learning_rate": 3.789997550362244e-05,
+      "loss": 0.3702,
+      "step": 3797
+    },
+    {
+      "epoch": 2.8588633797515994,
+      "grad_norm": 0.24971303343772888,
+      "learning_rate": 3.7893376861118603e-05,
+      "loss": 0.299,
+      "step": 3798
+    },
+    {
+      "epoch": 2.8596161083929243,
+      "grad_norm": 0.4010765254497528,
+      "learning_rate": 3.78867770820107e-05,
+      "loss": 0.2683,
+      "step": 3799
+    },
+    {
+      "epoch": 2.8603688370342493,
+      "grad_norm": 0.23192952573299408,
+      "learning_rate": 3.788017616702048e-05,
+      "loss": 0.3013,
+      "step": 3800
+    },
+    {
+      "epoch": 2.8603688370342493,
+      "eval_loss": 0.29296359419822693,
+      "eval_runtime": 452.8158,
+      "eval_samples_per_second": 21.26,
+      "eval_steps_per_second": 0.665,
+      "step": 3800
+    },
+    {
+      "epoch": 2.8611215656755737,
+      "grad_norm": 0.3594442903995514,
+      "learning_rate": 3.787357411686979e-05,
+      "loss": 0.3151,
+      "step": 3801
+    },
+    {
+      "epoch": 2.8618742943168987,
+      "grad_norm": 0.2728123068809509,
+      "learning_rate": 3.786697093228064e-05,
+      "loss": 0.2855,
+      "step": 3802
+    },
+    {
+      "epoch": 2.8626270229582236,
+      "grad_norm": 0.2590806782245636,
+      "learning_rate": 3.7860366613975126e-05,
+      "loss": 0.2012,
+      "step": 3803
+    },
+    {
+      "epoch": 2.8633797515995485,
+      "grad_norm": 0.3667093515396118,
+      "learning_rate": 3.78537611626755e-05,
+      "loss": 0.3068,
+      "step": 3804
+    },
+    {
+      "epoch": 2.8641324802408734,
+      "grad_norm": 0.28225359320640564,
+      "learning_rate": 3.784715457910413e-05,
+      "loss": 0.3474,
+      "step": 3805
+    },
+    {
+      "epoch": 2.864885208882198,
+      "grad_norm": 0.26375091075897217,
+      "learning_rate": 3.784054686398348e-05,
+      "loss": 0.309,
+      "step": 3806
+    },
+    {
+      "epoch": 2.865637937523523,
+      "grad_norm": 0.2658171057701111,
+      "learning_rate": 3.783393801803617e-05,
+      "loss": 0.2004,
+      "step": 3807
+    },
+    {
+      "epoch": 2.8663906661648477,
+      "grad_norm": 0.2972411513328552,
+      "learning_rate": 3.782732804198494e-05,
+      "loss": 0.3244,
+      "step": 3808
+    },
+    {
+      "epoch": 2.867143394806172,
+      "grad_norm": 0.3258616328239441,
+      "learning_rate": 3.7820716936552647e-05,
+      "loss": 0.3626,
+      "step": 3809
+    },
+    {
+      "epoch": 2.867896123447497,
+      "grad_norm": 0.23540711402893066,
+      "learning_rate": 3.7814104702462264e-05,
+      "loss": 0.2605,
+      "step": 3810
+    },
+    {
+      "epoch": 2.868648852088822,
+      "grad_norm": 0.272938996553421,
+      "learning_rate": 3.780749134043689e-05,
+      "loss": 0.3478,
+      "step": 3811
+    },
+    {
+      "epoch": 2.8694015807301465,
+      "grad_norm": 0.2497367411851883,
+      "learning_rate": 3.7800876851199774e-05,
+      "loss": 0.3652,
+      "step": 3812
+    },
+    {
+      "epoch": 2.8701543093714714,
+      "grad_norm": 0.2532345652580261,
+      "learning_rate": 3.7794261235474236e-05,
+      "loss": 0.2673,
+      "step": 3813
+    },
+    {
+      "epoch": 2.8709070380127963,
+      "grad_norm": 0.24775080382823944,
+      "learning_rate": 3.7787644493983785e-05,
+      "loss": 0.2816,
+      "step": 3814
+    },
+    {
+      "epoch": 2.8716597666541213,
+      "grad_norm": 0.3091953694820404,
+      "learning_rate": 3.778102662745198e-05,
+      "loss": 0.3415,
+      "step": 3815
+    },
+    {
+      "epoch": 2.872412495295446,
+      "grad_norm": 0.22297891974449158,
+      "learning_rate": 3.777440763660257e-05,
+      "loss": 0.2586,
+      "step": 3816
+    },
+    {
+      "epoch": 2.8731652239367707,
+      "grad_norm": 0.21219415962696075,
+      "learning_rate": 3.7767787522159395e-05,
+      "loss": 0.2707,
+      "step": 3817
+    },
+    {
+      "epoch": 2.8739179525780956,
+      "grad_norm": 0.24014152586460114,
+      "learning_rate": 3.7761166284846406e-05,
+      "loss": 0.2665,
+      "step": 3818
+    },
+    {
+      "epoch": 2.8746706812194205,
+      "grad_norm": 0.2323296219110489,
+      "learning_rate": 3.775454392538769e-05,
+      "loss": 0.2961,
+      "step": 3819
+    },
+    {
+      "epoch": 2.875423409860745,
+      "grad_norm": 0.2419193685054779,
+      "learning_rate": 3.774792044450748e-05,
+      "loss": 0.3311,
+      "step": 3820
+    },
+    {
+      "epoch": 2.87617613850207,
+      "grad_norm": 0.20629024505615234,
+      "learning_rate": 3.77412958429301e-05,
+      "loss": 0.3011,
+      "step": 3821
+    },
+    {
+      "epoch": 2.876928867143395,
+      "grad_norm": 0.28723224997520447,
+      "learning_rate": 3.773467012137999e-05,
+      "loss": 0.2533,
+      "step": 3822
+    },
+    {
+      "epoch": 2.8776815957847197,
+      "grad_norm": 0.2444298416376114,
+      "learning_rate": 3.7728043280581764e-05,
+      "loss": 0.308,
+      "step": 3823
+    },
+    {
+      "epoch": 2.8784343244260446,
+      "grad_norm": 0.19169318675994873,
+      "learning_rate": 3.772141532126009e-05,
+      "loss": 0.2288,
+      "step": 3824
+    },
+    {
+      "epoch": 2.879187053067369,
+      "grad_norm": 0.2188243418931961,
+      "learning_rate": 3.771478624413981e-05,
+      "loss": 0.2711,
+      "step": 3825
+    },
+    {
+      "epoch": 2.879939781708694,
+      "grad_norm": 0.24597178399562836,
+      "learning_rate": 3.770815604994585e-05,
+      "loss": 0.2844,
+      "step": 3826
+    },
+    {
+      "epoch": 2.880692510350019,
+      "grad_norm": 0.21203172206878662,
+      "learning_rate": 3.770152473940331e-05,
+      "loss": 0.1829,
+      "step": 3827
+    },
+    {
+      "epoch": 2.8814452389913434,
+      "grad_norm": 0.2482914924621582,
+      "learning_rate": 3.769489231323735e-05,
+      "loss": 0.2791,
+      "step": 3828
+    },
+    {
+      "epoch": 2.8821979676326683,
+      "grad_norm": 0.24820305407047272,
+      "learning_rate": 3.7688258772173295e-05,
+      "loss": 0.3201,
+      "step": 3829
+    },
+    {
+      "epoch": 2.8829506962739933,
+      "grad_norm": 0.2191455066204071,
+      "learning_rate": 3.7681624116936586e-05,
+      "loss": 0.1795,
+      "step": 3830
+    },
+    {
+      "epoch": 2.883703424915318,
+      "grad_norm": 0.221062570810318,
+      "learning_rate": 3.7674988348252766e-05,
+      "loss": 0.2643,
+      "step": 3831
+    },
+    {
+      "epoch": 2.884456153556643,
+      "grad_norm": 0.23770277202129364,
+      "learning_rate": 3.766835146684752e-05,
+      "loss": 0.3293,
+      "step": 3832
+    },
+    {
+      "epoch": 2.8852088821979676,
+      "grad_norm": 0.2346632033586502,
+      "learning_rate": 3.766171347344665e-05,
+      "loss": 0.2516,
+      "step": 3833
+    },
+    {
+      "epoch": 2.8859616108392925,
+      "grad_norm": 0.22736546397209167,
+      "learning_rate": 3.765507436877607e-05,
+      "loss": 0.2847,
+      "step": 3834
+    },
+    {
+      "epoch": 2.8867143394806174,
+      "grad_norm": 0.2406437247991562,
+      "learning_rate": 3.764843415356181e-05,
+      "loss": 0.2925,
+      "step": 3835
+    },
+    {
+      "epoch": 2.887467068121942,
+      "grad_norm": 0.2787608802318573,
+      "learning_rate": 3.764179282853007e-05,
+      "loss": 0.3311,
+      "step": 3836
+    },
+    {
+      "epoch": 2.888219796763267,
+      "grad_norm": 0.25507089495658875,
+      "learning_rate": 3.76351503944071e-05,
+      "loss": 0.2358,
+      "step": 3837
+    },
+    {
+      "epoch": 2.8889725254045917,
+      "grad_norm": 0.27272188663482666,
+      "learning_rate": 3.7628506851919335e-05,
+      "loss": 0.3676,
+      "step": 3838
+    },
+    {
+      "epoch": 2.889725254045916,
+      "grad_norm": 0.23120377957820892,
+      "learning_rate": 3.7621862201793276e-05,
+      "loss": 0.2315,
+      "step": 3839
+    },
+    {
+      "epoch": 2.890477982687241,
+      "grad_norm": 0.2745876908302307,
+      "learning_rate": 3.761521644475559e-05,
+      "loss": 0.2821,
+      "step": 3840
+    },
+    {
+      "epoch": 2.891230711328566,
+      "grad_norm": 0.26322299242019653,
+      "learning_rate": 3.7608569581533044e-05,
+      "loss": 0.265,
+      "step": 3841
+    },
+    {
+      "epoch": 2.891983439969891,
+      "grad_norm": 0.31340914964675903,
+      "learning_rate": 3.7601921612852516e-05,
+      "loss": 0.2774,
+      "step": 3842
+    },
+    {
+      "epoch": 2.892736168611216,
+      "grad_norm": 0.26828503608703613,
+      "learning_rate": 3.759527253944104e-05,
+      "loss": 0.299,
+      "step": 3843
+    },
+    {
+      "epoch": 2.8934888972525403,
+      "grad_norm": 0.25289469957351685,
+      "learning_rate": 3.758862236202573e-05,
+      "loss": 0.3663,
+      "step": 3844
+    },
+    {
+      "epoch": 2.8942416258938652,
+      "grad_norm": 0.2555590271949768,
+      "learning_rate": 3.7581971081333846e-05,
+      "loss": 0.3045,
+      "step": 3845
+    },
+    {
+      "epoch": 2.89499435453519,
+      "grad_norm": 0.22181902825832367,
+      "learning_rate": 3.757531869809277e-05,
+      "loss": 0.3126,
+      "step": 3846
+    },
+    {
+      "epoch": 2.8957470831765146,
+      "grad_norm": 0.24157610535621643,
+      "learning_rate": 3.756866521302998e-05,
+      "loss": 0.2041,
+      "step": 3847
+    },
+    {
+      "epoch": 2.8964998118178396,
+      "grad_norm": 0.2577713429927826,
+      "learning_rate": 3.7562010626873104e-05,
+      "loss": 0.2989,
+      "step": 3848
+    },
+    {
+      "epoch": 2.8972525404591645,
+      "grad_norm": 0.22516414523124695,
+      "learning_rate": 3.755535494034987e-05,
+      "loss": 0.2537,
+      "step": 3849
+    },
+    {
+      "epoch": 2.8980052691004894,
+      "grad_norm": 0.28421133756637573,
+      "learning_rate": 3.754869815418815e-05,
+      "loss": 0.2671,
+      "step": 3850
+    },
+    {
+      "epoch": 2.8987579977418143,
+      "grad_norm": 0.2934923768043518,
+      "learning_rate": 3.7542040269115896e-05,
+      "loss": 0.3223,
+      "step": 3851
+    },
+    {
+      "epoch": 2.899510726383139,
+      "grad_norm": 0.25759610533714294,
+      "learning_rate": 3.753538128586123e-05,
+      "loss": 0.2708,
+      "step": 3852
+    },
+    {
+      "epoch": 2.9002634550244637,
+      "grad_norm": 0.2393772453069687,
+      "learning_rate": 3.752872120515234e-05,
+      "loss": 0.335,
+      "step": 3853
+    },
+    {
+      "epoch": 2.9010161836657886,
+      "grad_norm": 0.2421187311410904,
+      "learning_rate": 3.7522060027717576e-05,
+      "loss": 0.2761,
+      "step": 3854
+    },
+    {
+      "epoch": 2.901768912307113,
+      "grad_norm": 0.23273757100105286,
+      "learning_rate": 3.75153977542854e-05,
+      "loss": 0.2423,
+      "step": 3855
+    },
+    {
+      "epoch": 2.902521640948438,
+      "grad_norm": 0.223927304148674,
+      "learning_rate": 3.7508734385584395e-05,
+      "loss": 0.2744,
+      "step": 3856
+    },
+    {
+      "epoch": 2.903274369589763,
+      "grad_norm": 0.24173517525196075,
+      "learning_rate": 3.750206992234323e-05,
+      "loss": 0.2205,
+      "step": 3857
+    },
+    {
+      "epoch": 2.904027098231088,
+      "grad_norm": 0.21355031430721283,
+      "learning_rate": 3.749540436529074e-05,
+      "loss": 0.2937,
+      "step": 3858
+    },
+    {
+      "epoch": 2.9047798268724128,
+      "grad_norm": 0.23200422525405884,
+      "learning_rate": 3.748873771515586e-05,
+      "loss": 0.3427,
+      "step": 3859
+    },
+    {
+      "epoch": 2.9055325555137372,
+      "grad_norm": 0.24121959507465363,
+      "learning_rate": 3.7482069972667646e-05,
+      "loss": 0.3748,
+      "step": 3860
+    },
+    {
+      "epoch": 2.906285284155062,
+      "grad_norm": 0.22919073700904846,
+      "learning_rate": 3.747540113855526e-05,
+      "loss": 0.354,
+      "step": 3861
+    },
+    {
+      "epoch": 2.907038012796387,
+      "grad_norm": 0.19618059694766998,
+      "learning_rate": 3.7468731213548005e-05,
+      "loss": 0.2049,
+      "step": 3862
+    },
+    {
+      "epoch": 2.9077907414377115,
+      "grad_norm": 0.23413923382759094,
+      "learning_rate": 3.746206019837529e-05,
+      "loss": 0.2461,
+      "step": 3863
+    },
+    {
+      "epoch": 2.9085434700790365,
+      "grad_norm": 0.21516884863376617,
+      "learning_rate": 3.745538809376665e-05,
+      "loss": 0.2573,
+      "step": 3864
+    },
+    {
+      "epoch": 2.9092961987203614,
+      "grad_norm": 0.22905804216861725,
+      "learning_rate": 3.7448714900451735e-05,
+      "loss": 0.3897,
+      "step": 3865
+    },
+    {
+      "epoch": 2.910048927361686,
+      "grad_norm": 0.23640944063663483,
+      "learning_rate": 3.744204061916032e-05,
+      "loss": 0.24,
+      "step": 3866
+    },
+    {
+      "epoch": 2.9108016560030108,
+      "grad_norm": 0.27932271361351013,
+      "learning_rate": 3.743536525062229e-05,
+      "loss": 0.2327,
+      "step": 3867
+    },
+    {
+      "epoch": 2.9115543846443357,
+      "grad_norm": 0.22933490574359894,
+      "learning_rate": 3.742868879556765e-05,
+      "loss": 0.2613,
+      "step": 3868
+    },
+    {
+      "epoch": 2.9123071132856606,
+      "grad_norm": 0.2410215586423874,
+      "learning_rate": 3.7422011254726536e-05,
+      "loss": 0.2389,
+      "step": 3869
+    },
+    {
+      "epoch": 2.9130598419269855,
+      "grad_norm": 0.20221802592277527,
+      "learning_rate": 3.741533262882918e-05,
+      "loss": 0.3225,
+      "step": 3870
+    },
+    {
+      "epoch": 2.91381257056831,
+      "grad_norm": 0.2522701621055603,
+      "learning_rate": 3.7408652918605966e-05,
+      "loss": 0.191,
+      "step": 3871
+    },
+    {
+      "epoch": 2.914565299209635,
+      "grad_norm": 0.25881248712539673,
+      "learning_rate": 3.7401972124787356e-05,
+      "loss": 0.3055,
+      "step": 3872
+    },
+    {
+      "epoch": 2.91531802785096,
+      "grad_norm": 0.2554793059825897,
+      "learning_rate": 3.739529024810397e-05,
+      "loss": 0.292,
+      "step": 3873
+    },
+    {
+      "epoch": 2.9160707564922843,
+      "grad_norm": 0.2474304437637329,
+      "learning_rate": 3.738860728928651e-05,
+      "loss": 0.3269,
+      "step": 3874
+    },
+    {
+      "epoch": 2.9168234851336092,
+      "grad_norm": 0.2395186722278595,
+      "learning_rate": 3.7381923249065835e-05,
+      "loss": 0.3122,
+      "step": 3875
+    },
+    {
+      "epoch": 2.917576213774934,
+      "grad_norm": 0.2859858572483063,
+      "learning_rate": 3.737523812817288e-05,
+      "loss": 0.3718,
+      "step": 3876
+    },
+    {
+      "epoch": 2.918328942416259,
+      "grad_norm": 0.20073392987251282,
+      "learning_rate": 3.736855192733874e-05,
+      "loss": 0.2369,
+      "step": 3877
+    },
+    {
+      "epoch": 2.919081671057584,
+      "grad_norm": 0.3220904767513275,
+      "learning_rate": 3.7361864647294596e-05,
+      "loss": 0.2319,
+      "step": 3878
+    },
+    {
+      "epoch": 2.9198343996989085,
+      "grad_norm": 0.24144913256168365,
+      "learning_rate": 3.7355176288771765e-05,
+      "loss": 0.326,
+      "step": 3879
+    },
+    {
+      "epoch": 2.9205871283402334,
+      "grad_norm": 0.26581698656082153,
+      "learning_rate": 3.734848685250167e-05,
+      "loss": 0.2333,
+      "step": 3880
+    },
+    {
+      "epoch": 2.9213398569815583,
+      "grad_norm": 0.360148161649704,
+      "learning_rate": 3.734179633921585e-05,
+      "loss": 0.2694,
+      "step": 3881
+    },
+    {
+      "epoch": 2.9220925856228828,
+      "grad_norm": 0.2352619767189026,
+      "learning_rate": 3.7335104749646e-05,
+      "loss": 0.2226,
+      "step": 3882
+    },
+    {
+      "epoch": 2.9228453142642077,
+      "grad_norm": 0.3708992004394531,
+      "learning_rate": 3.732841208452386e-05,
+      "loss": 0.2758,
+      "step": 3883
+    },
+    {
+      "epoch": 2.9235980429055326,
+      "grad_norm": 0.2655598521232605,
+      "learning_rate": 3.7321718344581366e-05,
+      "loss": 0.2189,
+      "step": 3884
+    },
+    {
+      "epoch": 2.9243507715468575,
+      "grad_norm": 0.37196671962738037,
+      "learning_rate": 3.731502353055051e-05,
+      "loss": 0.2398,
+      "step": 3885
+    },
+    {
+      "epoch": 2.9251035001881824,
+      "grad_norm": 0.36188021302223206,
+      "learning_rate": 3.730832764316344e-05,
+      "loss": 0.2697,
+      "step": 3886
+    },
+    {
+      "epoch": 2.925856228829507,
+      "grad_norm": 0.238564595580101,
+      "learning_rate": 3.730163068315241e-05,
+      "loss": 0.2759,
+      "step": 3887
+    },
+    {
+      "epoch": 2.926608957470832,
+      "grad_norm": 0.29676759243011475,
+      "learning_rate": 3.729493265124978e-05,
+      "loss": 0.2505,
+      "step": 3888
+    },
+    {
+      "epoch": 2.9273616861121567,
+      "grad_norm": 0.2875533401966095,
+      "learning_rate": 3.7288233548188045e-05,
+      "loss": 0.2771,
+      "step": 3889
+    },
+    {
+      "epoch": 2.928114414753481,
+      "grad_norm": 0.2976825535297394,
+      "learning_rate": 3.7281533374699806e-05,
+      "loss": 0.223,
+      "step": 3890
+    },
+    {
+      "epoch": 2.928867143394806,
+      "grad_norm": 0.3494449853897095,
+      "learning_rate": 3.727483213151778e-05,
+      "loss": 0.2611,
+      "step": 3891
+    },
+    {
+      "epoch": 2.929619872036131,
+      "grad_norm": 0.3256186246871948,
+      "learning_rate": 3.726812981937482e-05,
+      "loss": 0.2854,
+      "step": 3892
+    },
+    {
+      "epoch": 2.9303726006774555,
+      "grad_norm": 0.2907426059246063,
+      "learning_rate": 3.726142643900385e-05,
+      "loss": 0.237,
+      "step": 3893
+    },
+    {
+      "epoch": 2.9311253293187804,
+      "grad_norm": 0.2617429494857788,
+      "learning_rate": 3.7254721991137977e-05,
+      "loss": 0.2637,
+      "step": 3894
+    },
+    {
+      "epoch": 2.9318780579601054,
+      "grad_norm": 0.28986895084381104,
+      "learning_rate": 3.724801647651036e-05,
+      "loss": 0.3192,
+      "step": 3895
+    },
+    {
+      "epoch": 2.9326307866014303,
+      "grad_norm": 0.24382558465003967,
+      "learning_rate": 3.7241309895854324e-05,
+      "loss": 0.3281,
+      "step": 3896
+    },
+    {
+      "epoch": 2.933383515242755,
+      "grad_norm": 0.26339191198349,
+      "learning_rate": 3.723460224990328e-05,
+      "loss": 0.227,
+      "step": 3897
+    },
+    {
+      "epoch": 2.9341362438840797,
+      "grad_norm": 0.2762465476989746,
+      "learning_rate": 3.7227893539390775e-05,
+      "loss": 0.3273,
+      "step": 3898
+    },
+    {
+      "epoch": 2.9348889725254046,
+      "grad_norm": 0.25157439708709717,
+      "learning_rate": 3.722118376505045e-05,
+      "loss": 0.3176,
+      "step": 3899
+    },
+    {
+      "epoch": 2.9356417011667295,
+      "grad_norm": 0.22970561683177948,
+      "learning_rate": 3.721447292761609e-05,
+      "loss": 0.3114,
+      "step": 3900
+    },
+    {
+      "epoch": 2.936394429808054,
+      "grad_norm": 0.19601380825042725,
+      "learning_rate": 3.720776102782158e-05,
+      "loss": 0.1438,
+      "step": 3901
+    },
+    {
+      "epoch": 2.937147158449379,
+      "grad_norm": 0.23695380985736847,
+      "learning_rate": 3.7201048066400915e-05,
+      "loss": 0.1965,
+      "step": 3902
+    },
+    {
+      "epoch": 2.937899887090704,
+      "grad_norm": 0.27348700165748596,
+      "learning_rate": 3.719433404408822e-05,
+      "loss": 0.3337,
+      "step": 3903
+    },
+    {
+      "epoch": 2.9386526157320287,
+      "grad_norm": 0.22778157889842987,
+      "learning_rate": 3.718761896161773e-05,
+      "loss": 0.2156,
+      "step": 3904
+    },
+    {
+      "epoch": 2.9394053443733537,
+      "grad_norm": 0.256119042634964,
+      "learning_rate": 3.718090281972379e-05,
+      "loss": 0.1996,
+      "step": 3905
+    },
+    {
+      "epoch": 2.940158073014678,
+      "grad_norm": 0.2589349150657654,
+      "learning_rate": 3.717418561914088e-05,
+      "loss": 0.3153,
+      "step": 3906
+    },
+    {
+      "epoch": 2.940910801656003,
+      "grad_norm": 0.20211459696292877,
+      "learning_rate": 3.716746736060357e-05,
+      "loss": 0.305,
+      "step": 3907
+    },
+    {
+      "epoch": 2.941663530297328,
+      "grad_norm": 0.2715764343738556,
+      "learning_rate": 3.716074804484657e-05,
+      "loss": 0.3094,
+      "step": 3908
+    },
+    {
+      "epoch": 2.9424162589386524,
+      "grad_norm": 0.2392229288816452,
+      "learning_rate": 3.715402767260468e-05,
+      "loss": 0.2503,
+      "step": 3909
+    },
+    {
+      "epoch": 2.9431689875799774,
+      "grad_norm": 0.25455981492996216,
+      "learning_rate": 3.714730624461285e-05,
+      "loss": 0.2221,
+      "step": 3910
+    },
+    {
+      "epoch": 2.9439217162213023,
+      "grad_norm": 0.24860888719558716,
+      "learning_rate": 3.71405837616061e-05,
+      "loss": 0.2231,
+      "step": 3911
+    },
+    {
+      "epoch": 2.9446744448626267,
+      "grad_norm": 0.23865605890750885,
+      "learning_rate": 3.713386022431961e-05,
+      "loss": 0.2845,
+      "step": 3912
+    },
+    {
+      "epoch": 2.9454271735039517,
+      "grad_norm": 0.2743493616580963,
+      "learning_rate": 3.712713563348865e-05,
+      "loss": 0.2411,
+      "step": 3913
+    },
+    {
+      "epoch": 2.9461799021452766,
+      "grad_norm": 0.21038907766342163,
+      "learning_rate": 3.7120409989848614e-05,
+      "loss": 0.3415,
+      "step": 3914
+    },
+    {
+      "epoch": 2.9469326307866015,
+      "grad_norm": 0.25602051615715027,
+      "learning_rate": 3.711368329413501e-05,
+      "loss": 0.239,
+      "step": 3915
+    },
+    {
+      "epoch": 2.9476853594279264,
+      "grad_norm": 0.2289072722196579,
+      "learning_rate": 3.710695554708344e-05,
+      "loss": 0.2365,
+      "step": 3916
+    },
+    {
+      "epoch": 2.948438088069251,
+      "grad_norm": 0.276589035987854,
+      "learning_rate": 3.710022674942967e-05,
+      "loss": 0.2548,
+      "step": 3917
+    },
+    {
+      "epoch": 2.949190816710576,
+      "grad_norm": 0.24463525414466858,
+      "learning_rate": 3.7093496901909524e-05,
+      "loss": 0.2706,
+      "step": 3918
+    },
+    {
+      "epoch": 2.9499435453519007,
+      "grad_norm": 0.253016859292984,
+      "learning_rate": 3.708676600525899e-05,
+      "loss": 0.4312,
+      "step": 3919
+    },
+    {
+      "epoch": 2.950696273993225,
+      "grad_norm": 0.2138688564300537,
+      "learning_rate": 3.708003406021414e-05,
+      "loss": 0.2445,
+      "step": 3920
+    },
+    {
+      "epoch": 2.95144900263455,
+      "grad_norm": 0.21737828850746155,
+      "learning_rate": 3.7073301067511154e-05,
+      "loss": 0.2243,
+      "step": 3921
+    },
+    {
+      "epoch": 2.952201731275875,
+      "grad_norm": 0.2552622854709625,
+      "learning_rate": 3.7066567027886376e-05,
+      "loss": 0.2884,
+      "step": 3922
+    },
+    {
+      "epoch": 2.9529544599172,
+      "grad_norm": 0.2242586761713028,
+      "learning_rate": 3.70598319420762e-05,
+      "loss": 0.2329,
+      "step": 3923
+    },
+    {
+      "epoch": 2.953707188558525,
+      "grad_norm": 0.2267129272222519,
+      "learning_rate": 3.705309581081718e-05,
+      "loss": 0.2326,
+      "step": 3924
+    },
+    {
+      "epoch": 2.9544599171998494,
+      "grad_norm": 0.233336940407753,
+      "learning_rate": 3.704635863484596e-05,
+      "loss": 0.3474,
+      "step": 3925
+    },
+    {
+      "epoch": 2.9552126458411743,
+      "grad_norm": 0.2534269392490387,
+      "learning_rate": 3.7039620414899314e-05,
+      "loss": 0.2829,
+      "step": 3926
+    },
+    {
+      "epoch": 2.955965374482499,
+      "grad_norm": 0.21083180606365204,
+      "learning_rate": 3.703288115171413e-05,
+      "loss": 0.2977,
+      "step": 3927
+    },
+    {
+      "epoch": 2.9567181031238237,
+      "grad_norm": 0.2518335282802582,
+      "learning_rate": 3.702614084602739e-05,
+      "loss": 0.3129,
+      "step": 3928
+    },
+    {
+      "epoch": 2.9574708317651486,
+      "grad_norm": 0.21477288007736206,
+      "learning_rate": 3.70193994985762e-05,
+      "loss": 0.305,
+      "step": 3929
+    },
+    {
+      "epoch": 2.9582235604064735,
+      "grad_norm": 0.2546151876449585,
+      "learning_rate": 3.7012657110097805e-05,
+      "loss": 0.2295,
+      "step": 3930
+    },
+    {
+      "epoch": 2.9589762890477984,
+      "grad_norm": 0.19920209050178528,
+      "learning_rate": 3.7005913681329525e-05,
+      "loss": 0.1492,
+      "step": 3931
+    },
+    {
+      "epoch": 2.9597290176891233,
+      "grad_norm": 0.22716285288333893,
+      "learning_rate": 3.699916921300882e-05,
+      "loss": 0.3483,
+      "step": 3932
+    },
+    {
+      "epoch": 2.960481746330448,
+      "grad_norm": 0.19375723600387573,
+      "learning_rate": 3.699242370587325e-05,
+      "loss": 0.1737,
+      "step": 3933
+    },
+    {
+      "epoch": 2.9612344749717727,
+      "grad_norm": 0.24931201338768005,
+      "learning_rate": 3.698567716066049e-05,
+      "loss": 0.253,
+      "step": 3934
+    },
+    {
+      "epoch": 2.9619872036130976,
+      "grad_norm": 0.24639329314231873,
+      "learning_rate": 3.6978929578108354e-05,
+      "loss": 0.2986,
+      "step": 3935
+    },
+    {
+      "epoch": 2.962739932254422,
+      "grad_norm": 0.2693914473056793,
+      "learning_rate": 3.6972180958954714e-05,
+      "loss": 0.298,
+      "step": 3936
+    },
+    {
+      "epoch": 2.963492660895747,
+      "grad_norm": 0.25814253091812134,
+      "learning_rate": 3.6965431303937615e-05,
+      "loss": 0.2682,
+      "step": 3937
+    },
+    {
+      "epoch": 2.964245389537072,
+      "grad_norm": 0.2703171670436859,
+      "learning_rate": 3.6958680613795176e-05,
+      "loss": 0.2695,
+      "step": 3938
+    },
+    {
+      "epoch": 2.9649981181783964,
+      "grad_norm": 0.26814547181129456,
+      "learning_rate": 3.695192888926565e-05,
+      "loss": 0.2204,
+      "step": 3939
+    },
+    {
+      "epoch": 2.9657508468197213,
+      "grad_norm": 0.2926277816295624,
+      "learning_rate": 3.694517613108738e-05,
+      "loss": 0.2807,
+      "step": 3940
+    },
+    {
+      "epoch": 2.9665035754610463,
+      "grad_norm": 0.2283763885498047,
+      "learning_rate": 3.6938422339998854e-05,
+      "loss": 0.282,
+      "step": 3941
+    },
+    {
+      "epoch": 2.967256304102371,
+      "grad_norm": 0.25533735752105713,
+      "learning_rate": 3.693166751673866e-05,
+      "loss": 0.2237,
+      "step": 3942
+    },
+    {
+      "epoch": 2.968009032743696,
+      "grad_norm": 0.28198301792144775,
+      "learning_rate": 3.692491166204548e-05,
+      "loss": 0.3949,
+      "step": 3943
+    },
+    {
+      "epoch": 2.9687617613850206,
+      "grad_norm": 0.2481624186038971,
+      "learning_rate": 3.6918154776658125e-05,
+      "loss": 0.2364,
+      "step": 3944
+    },
+    {
+      "epoch": 2.9695144900263455,
+      "grad_norm": 0.24708886444568634,
+      "learning_rate": 3.6911396861315525e-05,
+      "loss": 0.2554,
+      "step": 3945
+    },
+    {
+      "epoch": 2.9702672186676704,
+      "grad_norm": 0.28066912293434143,
+      "learning_rate": 3.690463791675673e-05,
+      "loss": 0.2999,
+      "step": 3946
+    },
+    {
+      "epoch": 2.971019947308995,
+      "grad_norm": 0.2957609295845032,
+      "learning_rate": 3.689787794372085e-05,
+      "loss": 0.3264,
+      "step": 3947
+    },
+    {
+      "epoch": 2.97177267595032,
+      "grad_norm": 0.2610669732093811,
+      "learning_rate": 3.6891116942947184e-05,
+      "loss": 0.321,
+      "step": 3948
+    },
+    {
+      "epoch": 2.9725254045916447,
+      "grad_norm": 0.2639903724193573,
+      "learning_rate": 3.688435491517508e-05,
+      "loss": 0.3555,
+      "step": 3949
+    },
+    {
+      "epoch": 2.9732781332329696,
+      "grad_norm": 0.27712875604629517,
+      "learning_rate": 3.687759186114403e-05,
+      "loss": 0.2776,
+      "step": 3950
+    },
+    {
+      "epoch": 2.9740308618742946,
+      "grad_norm": 0.2509295344352722,
+      "learning_rate": 3.687082778159364e-05,
+      "loss": 0.2157,
+      "step": 3951
+    },
+    {
+      "epoch": 2.974783590515619,
+      "grad_norm": 0.30518725514411926,
+      "learning_rate": 3.686406267726362e-05,
+      "loss": 0.2528,
+      "step": 3952
+    },
+    {
+      "epoch": 2.975536319156944,
+      "grad_norm": 0.25143536925315857,
+      "learning_rate": 3.685729654889377e-05,
+      "loss": 0.232,
+      "step": 3953
+    },
+    {
+      "epoch": 2.976289047798269,
+      "grad_norm": 0.26190367341041565,
+      "learning_rate": 3.6850529397224044e-05,
+      "loss": 0.3137,
+      "step": 3954
+    },
+    {
+      "epoch": 2.9770417764395933,
+      "grad_norm": 0.2561779022216797,
+      "learning_rate": 3.6843761222994476e-05,
+      "loss": 0.336,
+      "step": 3955
+    },
+    {
+      "epoch": 2.9777945050809183,
+      "grad_norm": 0.21558259427547455,
+      "learning_rate": 3.6836992026945234e-05,
+      "loss": 0.3756,
+      "step": 3956
+    },
+    {
+      "epoch": 2.978547233722243,
+      "grad_norm": 0.22294270992279053,
+      "learning_rate": 3.683022180981658e-05,
+      "loss": 0.2112,
+      "step": 3957
+    },
+    {
+      "epoch": 2.979299962363568,
+      "grad_norm": 0.23488843441009521,
+      "learning_rate": 3.68234505723489e-05,
+      "loss": 0.191,
+      "step": 3958
+    },
+    {
+      "epoch": 2.980052691004893,
+      "grad_norm": 0.2498268038034439,
+      "learning_rate": 3.6816678315282674e-05,
+      "loss": 0.288,
+      "step": 3959
+    },
+    {
+      "epoch": 2.9808054196462175,
+      "grad_norm": 0.2585449814796448,
+      "learning_rate": 3.680990503935852e-05,
+      "loss": 0.2995,
+      "step": 3960
+    },
+    {
+      "epoch": 2.9815581482875424,
+      "grad_norm": 0.22118374705314636,
+      "learning_rate": 3.6803130745317154e-05,
+      "loss": 0.3567,
+      "step": 3961
+    },
+    {
+      "epoch": 2.9823108769288673,
+      "grad_norm": 0.24306391179561615,
+      "learning_rate": 3.679635543389939e-05,
+      "loss": 0.3409,
+      "step": 3962
+    },
+    {
+      "epoch": 2.983063605570192,
+      "grad_norm": 0.26298487186431885,
+      "learning_rate": 3.678957910584617e-05,
+      "loss": 0.2012,
+      "step": 3963
+    },
+    {
+      "epoch": 2.9838163342115167,
+      "grad_norm": 0.2566252648830414,
+      "learning_rate": 3.678280176189855e-05,
+      "loss": 0.2213,
+      "step": 3964
+    },
+    {
+      "epoch": 2.9845690628528416,
+      "grad_norm": 0.24532726407051086,
+      "learning_rate": 3.6776023402797686e-05,
+      "loss": 0.3024,
+      "step": 3965
+    },
+    {
+      "epoch": 2.985321791494166,
+      "grad_norm": 0.21696849167346954,
+      "learning_rate": 3.676924402928486e-05,
+      "loss": 0.3083,
+      "step": 3966
+    },
+    {
+      "epoch": 2.986074520135491,
+      "grad_norm": 0.24228574335575104,
+      "learning_rate": 3.676246364210142e-05,
+      "loss": 0.2314,
+      "step": 3967
+    },
+    {
+      "epoch": 2.986827248776816,
+      "grad_norm": 0.24748697876930237,
+      "learning_rate": 3.675568224198889e-05,
+      "loss": 0.3151,
+      "step": 3968
+    },
+    {
+      "epoch": 2.987579977418141,
+      "grad_norm": 0.21610401570796967,
+      "learning_rate": 3.674889982968888e-05,
+      "loss": 0.2146,
+      "step": 3969
+    },
+    {
+      "epoch": 2.9883327060594658,
+      "grad_norm": 0.24923428893089294,
+      "learning_rate": 3.674211640594308e-05,
+      "loss": 0.2636,
+      "step": 3970
+    },
+    {
+      "epoch": 2.9890854347007902,
+      "grad_norm": 0.22623568773269653,
+      "learning_rate": 3.673533197149332e-05,
+      "loss": 0.2134,
+      "step": 3971
+    },
+    {
+      "epoch": 2.989838163342115,
+      "grad_norm": 0.224626362323761,
+      "learning_rate": 3.672854652708155e-05,
+      "loss": 0.247,
+      "step": 3972
+    },
+    {
+      "epoch": 2.99059089198344,
+      "grad_norm": 0.2445860207080841,
+      "learning_rate": 3.672176007344979e-05,
+      "loss": 0.3402,
+      "step": 3973
+    },
+    {
+      "epoch": 2.9913436206247646,
+      "grad_norm": 0.2770131826400757,
+      "learning_rate": 3.671497261134023e-05,
+      "loss": 0.3108,
+      "step": 3974
+    },
+    {
+      "epoch": 2.9920963492660895,
+      "grad_norm": 0.22840924561023712,
+      "learning_rate": 3.670818414149512e-05,
+      "loss": 0.3497,
+      "step": 3975
+    },
+    {
+      "epoch": 2.9928490779074144,
+      "grad_norm": 0.21186065673828125,
+      "learning_rate": 3.670139466465683e-05,
+      "loss": 0.2816,
+      "step": 3976
+    },
+    {
+      "epoch": 2.9936018065487393,
+      "grad_norm": 0.2571096122264862,
+      "learning_rate": 3.669460418156786e-05,
+      "loss": 0.3453,
+      "step": 3977
+    },
+    {
+      "epoch": 2.9943545351900642,
+      "grad_norm": 0.21727751195430756,
+      "learning_rate": 3.66878126929708e-05,
+      "loss": 0.2182,
+      "step": 3978
+    },
+    {
+      "epoch": 2.9951072638313887,
+      "grad_norm": 0.2627200782299042,
+      "learning_rate": 3.668102019960834e-05,
+      "loss": 0.285,
+      "step": 3979
+    },
+    {
+      "epoch": 2.9958599924727136,
+      "grad_norm": 0.2439442127943039,
+      "learning_rate": 3.6674226702223334e-05,
+      "loss": 0.2195,
+      "step": 3980
+    },
+    {
+      "epoch": 2.9966127211140385,
+      "grad_norm": 0.25678256154060364,
+      "learning_rate": 3.666743220155869e-05,
+      "loss": 0.25,
+      "step": 3981
+    },
+    {
+      "epoch": 2.997365449755363,
+      "grad_norm": 0.29669272899627686,
+      "learning_rate": 3.6660636698357426e-05,
+      "loss": 0.283,
+      "step": 3982
+    },
+    {
+      "epoch": 2.998118178396688,
+      "grad_norm": 0.21316131949424744,
+      "learning_rate": 3.6653840193362715e-05,
+      "loss": 0.3139,
+      "step": 3983
+    },
+    {
+      "epoch": 2.998870907038013,
+      "grad_norm": 0.24230490624904633,
+      "learning_rate": 3.664704268731781e-05,
+      "loss": 0.3167,
+      "step": 3984
+    },
+    {
+      "epoch": 2.9996236356793378,
+      "grad_norm": 0.23789165914058685,
+      "learning_rate": 3.664024418096606e-05,
+      "loss": 0.3214,
+      "step": 3985
+    },
+    {
+      "epoch": 3.0003763643206622,
+      "grad_norm": 0.2383875995874405,
+      "learning_rate": 3.663344467505095e-05,
+      "loss": 0.3288,
+      "step": 3986
+    },
+    {
+      "epoch": 3.001129092961987,
+      "grad_norm": 0.2605611979961395,
+      "learning_rate": 3.662664417031607e-05,
+      "loss": 0.2816,
+      "step": 3987
+    },
+    {
+      "epoch": 3.001881821603312,
+      "grad_norm": 0.2742141783237457,
+      "learning_rate": 3.661984266750509e-05,
+      "loss": 0.3768,
+      "step": 3988
+    },
+    {
+      "epoch": 3.002634550244637,
+      "grad_norm": 0.22624947130680084,
+      "learning_rate": 3.661304016736184e-05,
+      "loss": 0.334,
+      "step": 3989
+    },
+    {
+      "epoch": 3.0033872788859615,
+      "grad_norm": 0.23834966123104095,
+      "learning_rate": 3.660623667063021e-05,
+      "loss": 0.1593,
+      "step": 3990
+    },
+    {
+      "epoch": 3.0041400075272864,
+      "grad_norm": 0.2423871010541916,
+      "learning_rate": 3.6599432178054235e-05,
+      "loss": 0.2251,
+      "step": 3991
+    },
+    {
+      "epoch": 3.0048927361686113,
+      "grad_norm": 0.22667233645915985,
+      "learning_rate": 3.6592626690378036e-05,
+      "loss": 0.2594,
+      "step": 3992
+    },
+    {
+      "epoch": 3.005645464809936,
+      "grad_norm": 0.2330666035413742,
+      "learning_rate": 3.658582020834586e-05,
+      "loss": 0.2941,
+      "step": 3993
+    },
+    {
+      "epoch": 3.0063981934512607,
+      "grad_norm": 0.24673590064048767,
+      "learning_rate": 3.657901273270204e-05,
+      "loss": 0.3376,
+      "step": 3994
+    },
+    {
+      "epoch": 3.0071509220925856,
+      "grad_norm": 0.24798467755317688,
+      "learning_rate": 3.657220426419103e-05,
+      "loss": 0.2352,
+      "step": 3995
+    },
+    {
+      "epoch": 3.0079036507339105,
+      "grad_norm": 0.2683374881744385,
+      "learning_rate": 3.6565394803557413e-05,
+      "loss": 0.3149,
+      "step": 3996
+    },
+    {
+      "epoch": 3.0086563793752354,
+      "grad_norm": 0.26582881808280945,
+      "learning_rate": 3.655858435154584e-05,
+      "loss": 0.3003,
+      "step": 3997
+    },
+    {
+      "epoch": 3.00940910801656,
+      "grad_norm": 0.2588028907775879,
+      "learning_rate": 3.655177290890111e-05,
+      "loss": 0.2916,
+      "step": 3998
+    },
+    {
+      "epoch": 3.010161836657885,
+      "grad_norm": 0.23969189822673798,
+      "learning_rate": 3.65449604763681e-05,
+      "loss": 0.2532,
+      "step": 3999
+    },
+    {
+      "epoch": 3.0109145652992098,
+      "grad_norm": 0.25628963112831116,
+      "learning_rate": 3.6538147054691817e-05,
+      "loss": 0.2876,
+      "step": 4000
+    },
+    {
+      "epoch": 3.0109145652992098,
+      "eval_loss": 0.28469446301460266,
+      "eval_runtime": 452.8805,
+      "eval_samples_per_second": 21.257,
+      "eval_steps_per_second": 0.665,
+      "step": 4000
+    },
+    {
+      "epoch": 3.0116672939405342,
+      "grad_norm": 0.30473971366882324,
+      "learning_rate": 3.6531332644617356e-05,
+      "loss": 0.1913,
+      "step": 4001
+    },
+    {
+      "epoch": 3.012420022581859,
+      "grad_norm": 0.2756206691265106,
+      "learning_rate": 3.652451724688994e-05,
+      "loss": 0.1978,
+      "step": 4002
+    },
+    {
+      "epoch": 3.013172751223184,
+      "grad_norm": 0.2489887923002243,
+      "learning_rate": 3.651770086225488e-05,
+      "loss": 0.2469,
+      "step": 4003
+    },
+    {
+      "epoch": 3.013925479864509,
+      "grad_norm": 0.2639017403125763,
+      "learning_rate": 3.6510883491457606e-05,
+      "loss": 0.2596,
+      "step": 4004
+    },
+    {
+      "epoch": 3.0146782085058335,
+      "grad_norm": 0.2438778579235077,
+      "learning_rate": 3.650406513524367e-05,
+      "loss": 0.2649,
+      "step": 4005
+    },
+    {
+      "epoch": 3.0154309371471584,
+      "grad_norm": 0.27859583497047424,
+      "learning_rate": 3.649724579435871e-05,
+      "loss": 0.2792,
+      "step": 4006
+    },
+    {
+      "epoch": 3.0161836657884833,
+      "grad_norm": 0.23256948590278625,
+      "learning_rate": 3.6490425469548466e-05,
+      "loss": 0.3195,
+      "step": 4007
+    },
+    {
+      "epoch": 3.016936394429808,
+      "grad_norm": 0.23744861781597137,
+      "learning_rate": 3.6483604161558824e-05,
+      "loss": 0.2111,
+      "step": 4008
+    },
+    {
+      "epoch": 3.0176891230711327,
+      "grad_norm": 0.23651458323001862,
+      "learning_rate": 3.647678187113573e-05,
+      "loss": 0.3082,
+      "step": 4009
+    },
+    {
+      "epoch": 3.0184418517124576,
+      "grad_norm": 0.24159178137779236,
+      "learning_rate": 3.6469958599025265e-05,
+      "loss": 0.2681,
+      "step": 4010
+    },
+    {
+      "epoch": 3.0191945803537825,
+      "grad_norm": 0.2462047040462494,
+      "learning_rate": 3.6463134345973616e-05,
+      "loss": 0.2934,
+      "step": 4011
+    },
+    {
+      "epoch": 3.0199473089951074,
+      "grad_norm": 0.22545668482780457,
+      "learning_rate": 3.645630911272708e-05,
+      "loss": 0.2994,
+      "step": 4012
+    },
+    {
+      "epoch": 3.020700037636432,
+      "grad_norm": 0.2206968069076538,
+      "learning_rate": 3.644948290003203e-05,
+      "loss": 0.253,
+      "step": 4013
+    },
+    {
+      "epoch": 3.021452766277757,
+      "grad_norm": 0.18955200910568237,
+      "learning_rate": 3.644265570863499e-05,
+      "loss": 0.2019,
+      "step": 4014
+    },
+    {
+      "epoch": 3.0222054949190817,
+      "grad_norm": 0.23771551251411438,
+      "learning_rate": 3.643582753928256e-05,
+      "loss": 0.251,
+      "step": 4015
+    },
+    {
+      "epoch": 3.0229582235604067,
+      "grad_norm": 0.205471009016037,
+      "learning_rate": 3.642899839272148e-05,
+      "loss": 0.198,
+      "step": 4016
+    },
+    {
+      "epoch": 3.023710952201731,
+      "grad_norm": 0.22410355508327484,
+      "learning_rate": 3.6422168269698554e-05,
+      "loss": 0.2528,
+      "step": 4017
+    },
+    {
+      "epoch": 3.024463680843056,
+      "grad_norm": 0.2515904903411865,
+      "learning_rate": 3.641533717096072e-05,
+      "loss": 0.3045,
+      "step": 4018
+    },
+    {
+      "epoch": 3.025216409484381,
+      "grad_norm": 0.2276780903339386,
+      "learning_rate": 3.640850509725502e-05,
+      "loss": 0.2851,
+      "step": 4019
+    },
+    {
+      "epoch": 3.025969138125706,
+      "grad_norm": 0.2504514455795288,
+      "learning_rate": 3.640167204932859e-05,
+      "loss": 0.2651,
+      "step": 4020
+    },
+    {
+      "epoch": 3.0267218667670304,
+      "grad_norm": 0.2447642683982849,
+      "learning_rate": 3.639483802792869e-05,
+      "loss": 0.2905,
+      "step": 4021
+    },
+    {
+      "epoch": 3.0274745954083553,
+      "grad_norm": 0.2259979546070099,
+      "learning_rate": 3.638800303380268e-05,
+      "loss": 0.2153,
+      "step": 4022
+    },
+    {
+      "epoch": 3.02822732404968,
+      "grad_norm": 0.2203562706708908,
+      "learning_rate": 3.638116706769801e-05,
+      "loss": 0.2893,
+      "step": 4023
+    },
+    {
+      "epoch": 3.0289800526910047,
+      "grad_norm": 0.23090384900569916,
+      "learning_rate": 3.6374330130362284e-05,
+      "loss": 0.2646,
+      "step": 4024
+    },
+    {
+      "epoch": 3.0297327813323296,
+      "grad_norm": 0.24378718435764313,
+      "learning_rate": 3.636749222254314e-05,
+      "loss": 0.2444,
+      "step": 4025
+    },
+    {
+      "epoch": 3.0304855099736545,
+      "grad_norm": 0.22122325003147125,
+      "learning_rate": 3.6360653344988385e-05,
+      "loss": 0.2867,
+      "step": 4026
+    },
+    {
+      "epoch": 3.0312382386149794,
+      "grad_norm": 0.2564096748828888,
+      "learning_rate": 3.6353813498445896e-05,
+      "loss": 0.3048,
+      "step": 4027
+    },
+    {
+      "epoch": 3.031990967256304,
+      "grad_norm": 0.24084307253360748,
+      "learning_rate": 3.6346972683663676e-05,
+      "loss": 0.252,
+      "step": 4028
+    },
+    {
+      "epoch": 3.032743695897629,
+      "grad_norm": 0.264571875333786,
+      "learning_rate": 3.634013090138983e-05,
+      "loss": 0.2046,
+      "step": 4029
+    },
+    {
+      "epoch": 3.0334964245389537,
+      "grad_norm": 0.27085044980049133,
+      "learning_rate": 3.633328815237256e-05,
+      "loss": 0.2324,
+      "step": 4030
+    },
+    {
+      "epoch": 3.0342491531802787,
+      "grad_norm": 0.25510427355766296,
+      "learning_rate": 3.6326444437360176e-05,
+      "loss": 0.2203,
+      "step": 4031
+    },
+    {
+      "epoch": 3.035001881821603,
+      "grad_norm": 0.24542585015296936,
+      "learning_rate": 3.6319599757101095e-05,
+      "loss": 0.2167,
+      "step": 4032
+    },
+    {
+      "epoch": 3.035754610462928,
+      "grad_norm": 0.2667347490787506,
+      "learning_rate": 3.6312754112343847e-05,
+      "loss": 0.2206,
+      "step": 4033
+    },
+    {
+      "epoch": 3.036507339104253,
+      "grad_norm": 0.24609816074371338,
+      "learning_rate": 3.6305907503837064e-05,
+      "loss": 0.2949,
+      "step": 4034
+    },
+    {
+      "epoch": 3.037260067745578,
+      "grad_norm": 0.2534525692462921,
+      "learning_rate": 3.6299059932329464e-05,
+      "loss": 0.2211,
+      "step": 4035
+    },
+    {
+      "epoch": 3.0380127963869024,
+      "grad_norm": 0.32085245847702026,
+      "learning_rate": 3.6292211398569905e-05,
+      "loss": 0.2719,
+      "step": 4036
+    },
+    {
+      "epoch": 3.0387655250282273,
+      "grad_norm": 0.22961580753326416,
+      "learning_rate": 3.628536190330733e-05,
+      "loss": 0.2949,
+      "step": 4037
+    },
+    {
+      "epoch": 3.039518253669552,
+      "grad_norm": 0.2448877990245819,
+      "learning_rate": 3.627851144729078e-05,
+      "loss": 0.3438,
+      "step": 4038
+    },
+    {
+      "epoch": 3.040270982310877,
+      "grad_norm": 0.2796836495399475,
+      "learning_rate": 3.6271660031269425e-05,
+      "loss": 0.2396,
+      "step": 4039
+    },
+    {
+      "epoch": 3.0410237109522016,
+      "grad_norm": 0.23957069218158722,
+      "learning_rate": 3.6264807655992504e-05,
+      "loss": 0.2857,
+      "step": 4040
+    },
+    {
+      "epoch": 3.0417764395935265,
+      "grad_norm": 0.242221400141716,
+      "learning_rate": 3.62579543222094e-05,
+      "loss": 0.213,
+      "step": 4041
+    },
+    {
+      "epoch": 3.0425291682348514,
+      "grad_norm": 0.42793571949005127,
+      "learning_rate": 3.625110003066958e-05,
+      "loss": 0.2923,
+      "step": 4042
+    },
+    {
+      "epoch": 3.0432818968761763,
+      "grad_norm": 0.2513652443885803,
+      "learning_rate": 3.624424478212262e-05,
+      "loss": 0.3083,
+      "step": 4043
+    },
+    {
+      "epoch": 3.044034625517501,
+      "grad_norm": 0.24435169994831085,
+      "learning_rate": 3.62373885773182e-05,
+      "loss": 0.2692,
+      "step": 4044
+    },
+    {
+      "epoch": 3.0447873541588257,
+      "grad_norm": 0.2882906496524811,
+      "learning_rate": 3.62305314170061e-05,
+      "loss": 0.2801,
+      "step": 4045
+    },
+    {
+      "epoch": 3.0455400828001506,
+      "grad_norm": 0.27083173394203186,
+      "learning_rate": 3.622367330193621e-05,
+      "loss": 0.2627,
+      "step": 4046
+    },
+    {
+      "epoch": 3.046292811441475,
+      "grad_norm": 0.21130917966365814,
+      "learning_rate": 3.6216814232858526e-05,
+      "loss": 0.3239,
+      "step": 4047
+    },
+    {
+      "epoch": 3.0470455400828,
+      "grad_norm": 0.24277999997138977,
+      "learning_rate": 3.620995421052314e-05,
+      "loss": 0.3146,
+      "step": 4048
+    },
+    {
+      "epoch": 3.047798268724125,
+      "grad_norm": 0.23749327659606934,
+      "learning_rate": 3.620309323568025e-05,
+      "loss": 0.2636,
+      "step": 4049
+    },
+    {
+      "epoch": 3.04855099736545,
+      "grad_norm": 0.24618779122829437,
+      "learning_rate": 3.619623130908018e-05,
+      "loss": 0.2077,
+      "step": 4050
+    },
+    {
+      "epoch": 3.0493037260067744,
+      "grad_norm": 0.28758418560028076,
+      "learning_rate": 3.618936843147333e-05,
+      "loss": 0.2558,
+      "step": 4051
+    },
+    {
+      "epoch": 3.0500564546480993,
+      "grad_norm": 0.24061399698257446,
+      "learning_rate": 3.61825046036102e-05,
+      "loss": 0.3466,
+      "step": 4052
+    },
+    {
+      "epoch": 3.050809183289424,
+      "grad_norm": 0.25235438346862793,
+      "learning_rate": 3.6175639826241426e-05,
+      "loss": 0.2268,
+      "step": 4053
+    },
+    {
+      "epoch": 3.051561911930749,
+      "grad_norm": 0.2521819770336151,
+      "learning_rate": 3.616877410011772e-05,
+      "loss": 0.3254,
+      "step": 4054
+    },
+    {
+      "epoch": 3.0523146405720736,
+      "grad_norm": 0.25099533796310425,
+      "learning_rate": 3.616190742598992e-05,
+      "loss": 0.3658,
+      "step": 4055
+    },
+    {
+      "epoch": 3.0530673692133985,
+      "grad_norm": 0.2746138572692871,
+      "learning_rate": 3.6155039804608945e-05,
+      "loss": 0.2224,
+      "step": 4056
+    },
+    {
+      "epoch": 3.0538200978547234,
+      "grad_norm": 0.31982412934303284,
+      "learning_rate": 3.614817123672583e-05,
+      "loss": 0.3066,
+      "step": 4057
+    },
+    {
+      "epoch": 3.0545728264960483,
+      "grad_norm": 0.2395693063735962,
+      "learning_rate": 3.614130172309171e-05,
+      "loss": 0.2442,
+      "step": 4058
+    },
+    {
+      "epoch": 3.055325555137373,
+      "grad_norm": 0.23548652231693268,
+      "learning_rate": 3.613443126445782e-05,
+      "loss": 0.2611,
+      "step": 4059
+    },
+    {
+      "epoch": 3.0560782837786977,
+      "grad_norm": 0.21469540894031525,
+      "learning_rate": 3.612755986157552e-05,
+      "loss": 0.2469,
+      "step": 4060
+    },
+    {
+      "epoch": 3.0568310124200226,
+      "grad_norm": 0.24203552305698395,
+      "learning_rate": 3.6120687515196236e-05,
+      "loss": 0.3108,
+      "step": 4061
+    },
+    {
+      "epoch": 3.0575837410613476,
+      "grad_norm": 0.22485774755477905,
+      "learning_rate": 3.6113814226071526e-05,
+      "loss": 0.2664,
+      "step": 4062
+    },
+    {
+      "epoch": 3.058336469702672,
+      "grad_norm": 0.21790331602096558,
+      "learning_rate": 3.610693999495304e-05,
+      "loss": 0.2176,
+      "step": 4063
+    },
+    {
+      "epoch": 3.059089198343997,
+      "grad_norm": 0.23315246403217316,
+      "learning_rate": 3.610006482259254e-05,
+      "loss": 0.2569,
+      "step": 4064
+    },
+    {
+      "epoch": 3.059841926985322,
+      "grad_norm": 0.22919975221157074,
+      "learning_rate": 3.609318870974188e-05,
+      "loss": 0.2743,
+      "step": 4065
+    },
+    {
+      "epoch": 3.060594655626647,
+      "grad_norm": 0.2723551094532013,
+      "learning_rate": 3.608631165715303e-05,
+      "loss": 0.1796,
+      "step": 4066
+    },
+    {
+      "epoch": 3.0613473842679713,
+      "grad_norm": 0.22856126725673676,
+      "learning_rate": 3.607943366557805e-05,
+      "loss": 0.2311,
+      "step": 4067
+    },
+    {
+      "epoch": 3.062100112909296,
+      "grad_norm": 0.29819172620773315,
+      "learning_rate": 3.607255473576908e-05,
+      "loss": 0.3618,
+      "step": 4068
+    },
+    {
+      "epoch": 3.062852841550621,
+      "grad_norm": 0.26040467619895935,
+      "learning_rate": 3.606567486847844e-05,
+      "loss": 0.2804,
+      "step": 4069
+    },
+    {
+      "epoch": 3.063605570191946,
+      "grad_norm": 0.25827983021736145,
+      "learning_rate": 3.605879406445846e-05,
+      "loss": 0.2337,
+      "step": 4070
+    },
+    {
+      "epoch": 3.0643582988332705,
+      "grad_norm": 0.20515388250350952,
+      "learning_rate": 3.605191232446163e-05,
+      "loss": 0.288,
+      "step": 4071
+    },
+    {
+      "epoch": 3.0651110274745954,
+      "grad_norm": 0.2478419840335846,
+      "learning_rate": 3.604502964924053e-05,
+      "loss": 0.2353,
+      "step": 4072
+    },
+    {
+      "epoch": 3.0658637561159203,
+      "grad_norm": 0.27954572439193726,
+      "learning_rate": 3.6038146039547835e-05,
+      "loss": 0.311,
+      "step": 4073
+    },
+    {
+      "epoch": 3.066616484757245,
+      "grad_norm": 0.2077292501926422,
+      "learning_rate": 3.6031261496136317e-05,
+      "loss": 0.1724,
+      "step": 4074
+    },
+    {
+      "epoch": 3.0673692133985697,
+      "grad_norm": 0.2342333197593689,
+      "learning_rate": 3.602437601975887e-05,
+      "loss": 0.2006,
+      "step": 4075
+    },
+    {
+      "epoch": 3.0681219420398946,
+      "grad_norm": 0.2549091577529907,
+      "learning_rate": 3.6017489611168493e-05,
+      "loss": 0.2416,
+      "step": 4076
+    },
+    {
+      "epoch": 3.0688746706812196,
+      "grad_norm": 0.23401792347431183,
+      "learning_rate": 3.601060227111824e-05,
+      "loss": 0.2427,
+      "step": 4077
+    },
+    {
+      "epoch": 3.069627399322544,
+      "grad_norm": 0.21557414531707764,
+      "learning_rate": 3.6003714000361324e-05,
+      "loss": 0.3264,
+      "step": 4078
+    },
+    {
+      "epoch": 3.070380127963869,
+      "grad_norm": 0.2612052857875824,
+      "learning_rate": 3.5996824799651035e-05,
+      "loss": 0.2959,
+      "step": 4079
+    },
+    {
+      "epoch": 3.071132856605194,
+      "grad_norm": 0.2360296994447708,
+      "learning_rate": 3.598993466974076e-05,
+      "loss": 0.2821,
+      "step": 4080
+    },
+    {
+      "epoch": 3.071885585246519,
+      "grad_norm": 0.2189525067806244,
+      "learning_rate": 3.598304361138398e-05,
+      "loss": 0.2924,
+      "step": 4081
+    },
+    {
+      "epoch": 3.0726383138878433,
+      "grad_norm": 0.22907648980617523,
+      "learning_rate": 3.597615162533432e-05,
+      "loss": 0.2377,
+      "step": 4082
+    },
+    {
+      "epoch": 3.073391042529168,
+      "grad_norm": 0.26757267117500305,
+      "learning_rate": 3.5969258712345466e-05,
+      "loss": 0.2609,
+      "step": 4083
+    },
+    {
+      "epoch": 3.074143771170493,
+      "grad_norm": 0.23703263700008392,
+      "learning_rate": 3.59623648731712e-05,
+      "loss": 0.2433,
+      "step": 4084
+    },
+    {
+      "epoch": 3.074896499811818,
+      "grad_norm": 0.23177854716777802,
+      "learning_rate": 3.5955470108565435e-05,
+      "loss": 0.2341,
+      "step": 4085
+    },
+    {
+      "epoch": 3.0756492284531425,
+      "grad_norm": 0.2682729959487915,
+      "learning_rate": 3.5948574419282175e-05,
+      "loss": 0.2739,
+      "step": 4086
+    },
+    {
+      "epoch": 3.0764019570944674,
+      "grad_norm": 0.22988373041152954,
+      "learning_rate": 3.594167780607552e-05,
+      "loss": 0.2718,
+      "step": 4087
+    },
+    {
+      "epoch": 3.0771546857357923,
+      "grad_norm": 0.1824927031993866,
+      "learning_rate": 3.593478026969967e-05,
+      "loss": 0.2514,
+      "step": 4088
+    },
+    {
+      "epoch": 3.0779074143771172,
+      "grad_norm": 0.2467728853225708,
+      "learning_rate": 3.592788181090894e-05,
+      "loss": 0.2679,
+      "step": 4089
+    },
+    {
+      "epoch": 3.0786601430184417,
+      "grad_norm": 0.19702216982841492,
+      "learning_rate": 3.5920982430457707e-05,
+      "loss": 0.2483,
+      "step": 4090
+    },
+    {
+      "epoch": 3.0794128716597666,
+      "grad_norm": 0.23686936497688293,
+      "learning_rate": 3.5914082129100506e-05,
+      "loss": 0.3225,
+      "step": 4091
+    },
+    {
+      "epoch": 3.0801656003010915,
+      "grad_norm": 0.2483891099691391,
+      "learning_rate": 3.590718090759194e-05,
+      "loss": 0.2465,
+      "step": 4092
+    },
+    {
+      "epoch": 3.0809183289424165,
+      "grad_norm": 0.2191562056541443,
+      "learning_rate": 3.590027876668671e-05,
+      "loss": 0.2234,
+      "step": 4093
+    },
+    {
+      "epoch": 3.081671057583741,
+      "grad_norm": 0.2538817822933197,
+      "learning_rate": 3.589337570713962e-05,
+      "loss": 0.2622,
+      "step": 4094
+    },
+    {
+      "epoch": 3.082423786225066,
+      "grad_norm": 0.22546115517616272,
+      "learning_rate": 3.588647172970558e-05,
+      "loss": 0.3586,
+      "step": 4095
+    },
+    {
+      "epoch": 3.0831765148663908,
+      "grad_norm": 0.28520074486732483,
+      "learning_rate": 3.5879566835139597e-05,
+      "loss": 0.3144,
+      "step": 4096
+    },
+    {
+      "epoch": 3.0839292435077157,
+      "grad_norm": 0.250141441822052,
+      "learning_rate": 3.587266102419678e-05,
+      "loss": 0.2223,
+      "step": 4097
+    },
+    {
+      "epoch": 3.08468197214904,
+      "grad_norm": 0.26582029461860657,
+      "learning_rate": 3.586575429763236e-05,
+      "loss": 0.2355,
+      "step": 4098
+    },
+    {
+      "epoch": 3.085434700790365,
+      "grad_norm": 0.22455231845378876,
+      "learning_rate": 3.585884665620162e-05,
+      "loss": 0.305,
+      "step": 4099
+    },
+    {
+      "epoch": 3.08618742943169,
+      "grad_norm": 0.2740948498249054,
+      "learning_rate": 3.5851938100659964e-05,
+      "loss": 0.3107,
+      "step": 4100
+    },
+    {
+      "epoch": 3.0869401580730145,
+      "grad_norm": 0.19685223698616028,
+      "learning_rate": 3.584502863176293e-05,
+      "loss": 0.2924,
+      "step": 4101
+    },
+    {
+      "epoch": 3.0876928867143394,
+      "grad_norm": 0.2920851409435272,
+      "learning_rate": 3.5838118250266104e-05,
+      "loss": 0.2399,
+      "step": 4102
+    },
+    {
+      "epoch": 3.0884456153556643,
+      "grad_norm": 0.255236953496933,
+      "learning_rate": 3.5831206956925206e-05,
+      "loss": 0.2717,
+      "step": 4103
+    },
+    {
+      "epoch": 3.0891983439969892,
+      "grad_norm": 0.250510036945343,
+      "learning_rate": 3.582429475249604e-05,
+      "loss": 0.2788,
+      "step": 4104
+    },
+    {
+      "epoch": 3.0899510726383137,
+      "grad_norm": 0.22168336808681488,
+      "learning_rate": 3.5817381637734516e-05,
+      "loss": 0.3558,
+      "step": 4105
+    },
+    {
+      "epoch": 3.0907038012796386,
+      "grad_norm": 0.23940861225128174,
+      "learning_rate": 3.581046761339664e-05,
+      "loss": 0.3032,
+      "step": 4106
+    },
+    {
+      "epoch": 3.0914565299209635,
+      "grad_norm": 0.2423814833164215,
+      "learning_rate": 3.580355268023852e-05,
+      "loss": 0.2543,
+      "step": 4107
+    },
+    {
+      "epoch": 3.0922092585622885,
+      "grad_norm": 0.28369438648223877,
+      "learning_rate": 3.5796636839016364e-05,
+      "loss": 0.3349,
+      "step": 4108
+    },
+    {
+      "epoch": 3.092961987203613,
+      "grad_norm": 0.20807339251041412,
+      "learning_rate": 3.578972009048647e-05,
+      "loss": 0.2481,
+      "step": 4109
+    },
+    {
+      "epoch": 3.093714715844938,
+      "grad_norm": 0.2491023689508438,
+      "learning_rate": 3.578280243540526e-05,
+      "loss": 0.2636,
+      "step": 4110
+    },
+    {
+      "epoch": 3.0944674444862628,
+      "grad_norm": 0.24914215505123138,
+      "learning_rate": 3.577588387452923e-05,
+      "loss": 0.295,
+      "step": 4111
+    },
+    {
+      "epoch": 3.0952201731275877,
+      "grad_norm": 0.3419395089149475,
+      "learning_rate": 3.5768964408614966e-05,
+      "loss": 0.319,
+      "step": 4112
+    },
+    {
+      "epoch": 3.095972901768912,
+      "grad_norm": 0.24558810889720917,
+      "learning_rate": 3.576204403841919e-05,
+      "loss": 0.261,
+      "step": 4113
+    },
+    {
+      "epoch": 3.096725630410237,
+      "grad_norm": 0.2745939791202545,
+      "learning_rate": 3.5755122764698715e-05,
+      "loss": 0.3016,
+      "step": 4114
+    },
+    {
+      "epoch": 3.097478359051562,
+      "grad_norm": 0.2545306086540222,
+      "learning_rate": 3.574820058821042e-05,
+      "loss": 0.3064,
+      "step": 4115
+    },
+    {
+      "epoch": 3.098231087692887,
+      "grad_norm": 0.2768802344799042,
+      "learning_rate": 3.57412775097113e-05,
+      "loss": 0.3162,
+      "step": 4116
+    },
+    {
+      "epoch": 3.0989838163342114,
+      "grad_norm": 0.2314392775297165,
+      "learning_rate": 3.5734353529958475e-05,
+      "loss": 0.3278,
+      "step": 4117
+    },
+    {
+      "epoch": 3.0997365449755363,
+      "grad_norm": 0.302063912153244,
+      "learning_rate": 3.572742864970913e-05,
+      "loss": 0.2532,
+      "step": 4118
+    },
+    {
+      "epoch": 3.100489273616861,
+      "grad_norm": 0.23794293403625488,
+      "learning_rate": 3.5720502869720545e-05,
+      "loss": 0.2976,
+      "step": 4119
+    },
+    {
+      "epoch": 3.101242002258186,
+      "grad_norm": 0.2949250042438507,
+      "learning_rate": 3.571357619075013e-05,
+      "loss": 0.2434,
+      "step": 4120
+    },
+    {
+      "epoch": 3.1019947308995106,
+      "grad_norm": 0.23979072272777557,
+      "learning_rate": 3.5706648613555375e-05,
+      "loss": 0.2296,
+      "step": 4121
+    },
+    {
+      "epoch": 3.1027474595408355,
+      "grad_norm": 0.2040252834558487,
+      "learning_rate": 3.569972013889387e-05,
+      "loss": 0.2316,
+      "step": 4122
+    },
+    {
+      "epoch": 3.1035001881821604,
+      "grad_norm": 0.2522539794445038,
+      "learning_rate": 3.569279076752329e-05,
+      "loss": 0.2452,
+      "step": 4123
+    },
+    {
+      "epoch": 3.104252916823485,
+      "grad_norm": 0.27498728036880493,
+      "learning_rate": 3.568586050020143e-05,
+      "loss": 0.2499,
+      "step": 4124
+    },
+    {
+      "epoch": 3.10500564546481,
+      "grad_norm": 0.24378183484077454,
+      "learning_rate": 3.567892933768617e-05,
+      "loss": 0.326,
+      "step": 4125
+    },
+    {
+      "epoch": 3.1057583741061348,
+      "grad_norm": 0.2355974316596985,
+      "learning_rate": 3.56719972807355e-05,
+      "loss": 0.2367,
+      "step": 4126
+    },
+    {
+      "epoch": 3.1065111027474597,
+      "grad_norm": 0.2438228726387024,
+      "learning_rate": 3.5665064330107496e-05,
+      "loss": 0.2906,
+      "step": 4127
+    },
+    {
+      "epoch": 3.107263831388784,
+      "grad_norm": 0.20041614770889282,
+      "learning_rate": 3.5658130486560335e-05,
+      "loss": 0.2535,
+      "step": 4128
+    },
+    {
+      "epoch": 3.108016560030109,
+      "grad_norm": 0.2487453818321228,
+      "learning_rate": 3.5651195750852276e-05,
+      "loss": 0.3559,
+      "step": 4129
+    },
+    {
+      "epoch": 3.108769288671434,
+      "grad_norm": 0.24319089949131012,
+      "learning_rate": 3.564426012374172e-05,
+      "loss": 0.2995,
+      "step": 4130
+    },
+    {
+      "epoch": 3.109522017312759,
+      "grad_norm": 0.24014505743980408,
+      "learning_rate": 3.563732360598711e-05,
+      "loss": 0.2541,
+      "step": 4131
+    },
+    {
+      "epoch": 3.1102747459540834,
+      "grad_norm": 0.22249974310398102,
+      "learning_rate": 3.5630386198347034e-05,
+      "loss": 0.3484,
+      "step": 4132
+    },
+    {
+      "epoch": 3.1110274745954083,
+      "grad_norm": 0.19903342425823212,
+      "learning_rate": 3.5623447901580145e-05,
+      "loss": 0.3228,
+      "step": 4133
+    },
+    {
+      "epoch": 3.111780203236733,
+      "grad_norm": 0.21568480134010315,
+      "learning_rate": 3.56165087164452e-05,
+      "loss": 0.1673,
+      "step": 4134
+    },
+    {
+      "epoch": 3.112532931878058,
+      "grad_norm": 0.22584779560565948,
+      "learning_rate": 3.560956864370107e-05,
+      "loss": 0.2507,
+      "step": 4135
+    },
+    {
+      "epoch": 3.1132856605193826,
+      "grad_norm": 0.19897295534610748,
+      "learning_rate": 3.5602627684106706e-05,
+      "loss": 0.2296,
+      "step": 4136
+    },
+    {
+      "epoch": 3.1140383891607075,
+      "grad_norm": 0.2144831120967865,
+      "learning_rate": 3.5595685838421164e-05,
+      "loss": 0.2736,
+      "step": 4137
+    },
+    {
+      "epoch": 3.1147911178020324,
+      "grad_norm": 0.24475625157356262,
+      "learning_rate": 3.5588743107403586e-05,
+      "loss": 0.2453,
+      "step": 4138
+    },
+    {
+      "epoch": 3.1155438464433574,
+      "grad_norm": 0.21365268528461456,
+      "learning_rate": 3.558179949181322e-05,
+      "loss": 0.336,
+      "step": 4139
+    },
+    {
+      "epoch": 3.116296575084682,
+      "grad_norm": 0.2763289511203766,
+      "learning_rate": 3.557485499240941e-05,
+      "loss": 0.2889,
+      "step": 4140
+    },
+    {
+      "epoch": 3.1170493037260067,
+      "grad_norm": 0.29469984769821167,
+      "learning_rate": 3.5567909609951603e-05,
+      "loss": 0.3386,
+      "step": 4141
+    },
+    {
+      "epoch": 3.1178020323673317,
+      "grad_norm": 0.26894718408584595,
+      "learning_rate": 3.556096334519934e-05,
+      "loss": 0.2367,
+      "step": 4142
+    },
+    {
+      "epoch": 3.1185547610086566,
+      "grad_norm": 0.2808041572570801,
+      "learning_rate": 3.555401619891224e-05,
+      "loss": 0.2914,
+      "step": 4143
+    },
+    {
+      "epoch": 3.119307489649981,
+      "grad_norm": 0.2906474173069,
+      "learning_rate": 3.554706817185003e-05,
+      "loss": 0.2902,
+      "step": 4144
+    },
+    {
+      "epoch": 3.120060218291306,
+      "grad_norm": 0.2494940161705017,
+      "learning_rate": 3.5540119264772546e-05,
+      "loss": 0.3096,
+      "step": 4145
+    },
+    {
+      "epoch": 3.120812946932631,
+      "grad_norm": 0.25555089116096497,
+      "learning_rate": 3.553316947843971e-05,
+      "loss": 0.3006,
+      "step": 4146
+    },
+    {
+      "epoch": 3.1215656755739554,
+      "grad_norm": 0.2495083510875702,
+      "learning_rate": 3.552621881361154e-05,
+      "loss": 0.2615,
+      "step": 4147
+    },
+    {
+      "epoch": 3.1223184042152803,
+      "grad_norm": 0.2299984246492386,
+      "learning_rate": 3.551926727104814e-05,
+      "loss": 0.2352,
+      "step": 4148
+    },
+    {
+      "epoch": 3.123071132856605,
+      "grad_norm": 0.255581796169281,
+      "learning_rate": 3.5512314851509734e-05,
+      "loss": 0.3202,
+      "step": 4149
+    },
+    {
+      "epoch": 3.12382386149793,
+      "grad_norm": 0.20771536231040955,
+      "learning_rate": 3.550536155575662e-05,
+      "loss": 0.292,
+      "step": 4150
+    },
+    {
+      "epoch": 3.1245765901392546,
+      "grad_norm": 0.2360757291316986,
+      "learning_rate": 3.54984073845492e-05,
+      "loss": 0.253,
+      "step": 4151
+    },
+    {
+      "epoch": 3.1253293187805795,
+      "grad_norm": 0.24740858376026154,
+      "learning_rate": 3.5491452338647986e-05,
+      "loss": 0.3276,
+      "step": 4152
+    },
+    {
+      "epoch": 3.1260820474219044,
+      "grad_norm": 0.22477495670318604,
+      "learning_rate": 3.548449641881355e-05,
+      "loss": 0.2362,
+      "step": 4153
+    },
+    {
+      "epoch": 3.1268347760632293,
+      "grad_norm": 0.22011318802833557,
+      "learning_rate": 3.547753962580659e-05,
+      "loss": 0.2588,
+      "step": 4154
+    },
+    {
+      "epoch": 3.127587504704554,
+      "grad_norm": 0.2189411073923111,
+      "learning_rate": 3.547058196038789e-05,
+      "loss": 0.2645,
+      "step": 4155
+    },
+    {
+      "epoch": 3.1283402333458787,
+      "grad_norm": 0.2610968351364136,
+      "learning_rate": 3.546362342331833e-05,
+      "loss": 0.2868,
+      "step": 4156
+    },
+    {
+      "epoch": 3.1290929619872037,
+      "grad_norm": 0.22110888361930847,
+      "learning_rate": 3.5456664015358886e-05,
+      "loss": 0.2201,
+      "step": 4157
+    },
+    {
+      "epoch": 3.1298456906285286,
+      "grad_norm": 0.23754064738750458,
+      "learning_rate": 3.544970373727063e-05,
+      "loss": 0.2918,
+      "step": 4158
+    },
+    {
+      "epoch": 3.130598419269853,
+      "grad_norm": 0.20449262857437134,
+      "learning_rate": 3.5442742589814723e-05,
+      "loss": 0.3047,
+      "step": 4159
+    },
+    {
+      "epoch": 3.131351147911178,
+      "grad_norm": 0.2706467807292938,
+      "learning_rate": 3.543578057375243e-05,
+      "loss": 0.2835,
+      "step": 4160
+    },
+    {
+      "epoch": 3.132103876552503,
+      "grad_norm": 0.2413809597492218,
+      "learning_rate": 3.54288176898451e-05,
+      "loss": 0.3384,
+      "step": 4161
+    },
+    {
+      "epoch": 3.132856605193828,
+      "grad_norm": 0.22295692563056946,
+      "learning_rate": 3.542185393885419e-05,
+      "loss": 0.2275,
+      "step": 4162
+    },
+    {
+      "epoch": 3.1336093338351523,
+      "grad_norm": 0.2457658052444458,
+      "learning_rate": 3.541488932154125e-05,
+      "loss": 0.2751,
+      "step": 4163
+    },
+    {
+      "epoch": 3.134362062476477,
+      "grad_norm": 0.24042803049087524,
+      "learning_rate": 3.5407923838667914e-05,
+      "loss": 0.2547,
+      "step": 4164
+    },
+    {
+      "epoch": 3.135114791117802,
+      "grad_norm": 0.25533750653266907,
+      "learning_rate": 3.540095749099591e-05,
+      "loss": 0.235,
+      "step": 4165
+    },
+    {
+      "epoch": 3.135867519759127,
+      "grad_norm": 0.22975963354110718,
+      "learning_rate": 3.539399027928707e-05,
+      "loss": 0.2381,
+      "step": 4166
+    },
+    {
+      "epoch": 3.1366202484004515,
+      "grad_norm": 0.25097131729125977,
+      "learning_rate": 3.538702220430333e-05,
+      "loss": 0.2078,
+      "step": 4167
+    },
+    {
+      "epoch": 3.1373729770417764,
+      "grad_norm": 0.22966592013835907,
+      "learning_rate": 3.538005326680669e-05,
+      "loss": 0.251,
+      "step": 4168
+    },
+    {
+      "epoch": 3.1381257056831013,
+      "grad_norm": 0.3023170232772827,
+      "learning_rate": 3.537308346755929e-05,
+      "loss": 0.2202,
+      "step": 4169
+    },
+    {
+      "epoch": 3.1388784343244263,
+      "grad_norm": 0.2532676160335541,
+      "learning_rate": 3.5366112807323304e-05,
+      "loss": 0.3339,
+      "step": 4170
+    },
+    {
+      "epoch": 3.1396311629657507,
+      "grad_norm": 0.2231484353542328,
+      "learning_rate": 3.535914128686105e-05,
+      "loss": 0.1628,
+      "step": 4171
+    },
+    {
+      "epoch": 3.1403838916070757,
+      "grad_norm": 0.21488764882087708,
+      "learning_rate": 3.535216890693492e-05,
+      "loss": 0.2172,
+      "step": 4172
+    },
+    {
+      "epoch": 3.1411366202484006,
+      "grad_norm": 0.20217975974082947,
+      "learning_rate": 3.5345195668307404e-05,
+      "loss": 0.1905,
+      "step": 4173
+    },
+    {
+      "epoch": 3.141889348889725,
+      "grad_norm": 0.2107428014278412,
+      "learning_rate": 3.53382215717411e-05,
+      "loss": 0.2444,
+      "step": 4174
+    },
+    {
+      "epoch": 3.14264207753105,
+      "grad_norm": 0.22928081452846527,
+      "learning_rate": 3.533124661799865e-05,
+      "loss": 0.1992,
+      "step": 4175
+    },
+    {
+      "epoch": 3.143394806172375,
+      "grad_norm": 0.2481890469789505,
+      "learning_rate": 3.5324270807842855e-05,
+      "loss": 0.2929,
+      "step": 4176
+    },
+    {
+      "epoch": 3.1441475348137,
+      "grad_norm": 0.2403666079044342,
+      "learning_rate": 3.531729414203656e-05,
+      "loss": 0.2261,
+      "step": 4177
+    },
+    {
+      "epoch": 3.1449002634550243,
+      "grad_norm": 0.22844095528125763,
+      "learning_rate": 3.531031662134274e-05,
+      "loss": 0.2953,
+      "step": 4178
+    },
+    {
+      "epoch": 3.145652992096349,
+      "grad_norm": 0.24557112157344818,
+      "learning_rate": 3.530333824652443e-05,
+      "loss": 0.2665,
+      "step": 4179
+    },
+    {
+      "epoch": 3.146405720737674,
+      "grad_norm": 0.28586795926094055,
+      "learning_rate": 3.529635901834478e-05,
+      "loss": 0.312,
+      "step": 4180
+    },
+    {
+      "epoch": 3.147158449378999,
+      "grad_norm": 0.25375378131866455,
+      "learning_rate": 3.5289378937567035e-05,
+      "loss": 0.2633,
+      "step": 4181
+    },
+    {
+      "epoch": 3.1479111780203235,
+      "grad_norm": 0.25925207138061523,
+      "learning_rate": 3.528239800495452e-05,
+      "loss": 0.3062,
+      "step": 4182
+    },
+    {
+      "epoch": 3.1486639066616484,
+      "grad_norm": 0.24381649494171143,
+      "learning_rate": 3.5275416221270656e-05,
+      "loss": 0.3046,
+      "step": 4183
+    },
+    {
+      "epoch": 3.1494166353029733,
+      "grad_norm": 0.27563148736953735,
+      "learning_rate": 3.5268433587278964e-05,
+      "loss": 0.2422,
+      "step": 4184
+    },
+    {
+      "epoch": 3.1501693639442983,
+      "grad_norm": 0.24201300740242004,
+      "learning_rate": 3.526145010374306e-05,
+      "loss": 0.27,
+      "step": 4185
+    },
+    {
+      "epoch": 3.1509220925856227,
+      "grad_norm": 0.23533004522323608,
+      "learning_rate": 3.5254465771426634e-05,
+      "loss": 0.327,
+      "step": 4186
+    },
+    {
+      "epoch": 3.1516748212269476,
+      "grad_norm": 0.24813827872276306,
+      "learning_rate": 3.5247480591093495e-05,
+      "loss": 0.2472,
+      "step": 4187
+    },
+    {
+      "epoch": 3.1524275498682726,
+      "grad_norm": 0.26624932885169983,
+      "learning_rate": 3.524049456350753e-05,
+      "loss": 0.2607,
+      "step": 4188
+    },
+    {
+      "epoch": 3.1531802785095975,
+      "grad_norm": 0.26426246762275696,
+      "learning_rate": 3.52335076894327e-05,
+      "loss": 0.2342,
+      "step": 4189
+    },
+    {
+      "epoch": 3.153933007150922,
+      "grad_norm": 0.26692017912864685,
+      "learning_rate": 3.5226519969633124e-05,
+      "loss": 0.2098,
+      "step": 4190
+    },
+    {
+      "epoch": 3.154685735792247,
+      "grad_norm": 0.22750146687030792,
+      "learning_rate": 3.521953140487292e-05,
+      "loss": 0.238,
+      "step": 4191
+    },
+    {
+      "epoch": 3.155438464433572,
+      "grad_norm": 0.2432069331407547,
+      "learning_rate": 3.521254199591637e-05,
+      "loss": 0.3514,
+      "step": 4192
+    },
+    {
+      "epoch": 3.1561911930748967,
+      "grad_norm": 0.2042943239212036,
+      "learning_rate": 3.520555174352784e-05,
+      "loss": 0.192,
+      "step": 4193
+    },
+    {
+      "epoch": 3.156943921716221,
+      "grad_norm": 0.253132700920105,
+      "learning_rate": 3.519856064847174e-05,
+      "loss": 0.2759,
+      "step": 4194
+    },
+    {
+      "epoch": 3.157696650357546,
+      "grad_norm": 0.30248159170150757,
+      "learning_rate": 3.519156871151262e-05,
+      "loss": 0.2917,
+      "step": 4195
+    },
+    {
+      "epoch": 3.158449378998871,
+      "grad_norm": 0.2456314116716385,
+      "learning_rate": 3.518457593341512e-05,
+      "loss": 0.2881,
+      "step": 4196
+    },
+    {
+      "epoch": 3.159202107640196,
+      "grad_norm": 0.2789459824562073,
+      "learning_rate": 3.5177582314943944e-05,
+      "loss": 0.2875,
+      "step": 4197
+    },
+    {
+      "epoch": 3.1599548362815204,
+      "grad_norm": 0.24137146770954132,
+      "learning_rate": 3.5170587856863906e-05,
+      "loss": 0.2441,
+      "step": 4198
+    },
+    {
+      "epoch": 3.1607075649228453,
+      "grad_norm": 0.2537572383880615,
+      "learning_rate": 3.516359255993992e-05,
+      "loss": 0.2334,
+      "step": 4199
+    },
+    {
+      "epoch": 3.1614602935641702,
+      "grad_norm": 0.2791388928890228,
+      "learning_rate": 3.515659642493697e-05,
+      "loss": 0.2293,
+      "step": 4200
+    },
+    {
+      "epoch": 3.1614602935641702,
+      "eval_loss": 0.27616873383522034,
+      "eval_runtime": 452.8242,
+      "eval_samples_per_second": 21.26,
+      "eval_steps_per_second": 0.665,
+      "step": 4200
+    },
+    {
+      "epoch": 3.1622130222054947,
+      "grad_norm": 0.2508070170879364,
+      "learning_rate": 3.514959945262015e-05,
+      "loss": 0.2127,
+      "step": 4201
+    },
+    {
+      "epoch": 3.1629657508468196,
+      "grad_norm": 0.2344474047422409,
+      "learning_rate": 3.514260164375463e-05,
+      "loss": 0.2338,
+      "step": 4202
+    },
+    {
+      "epoch": 3.1637184794881446,
+      "grad_norm": 0.21113352477550507,
+      "learning_rate": 3.513560299910569e-05,
+      "loss": 0.2819,
+      "step": 4203
+    },
+    {
+      "epoch": 3.1644712081294695,
+      "grad_norm": 0.2982528507709503,
+      "learning_rate": 3.512860351943869e-05,
+      "loss": 0.3044,
+      "step": 4204
+    },
+    {
+      "epoch": 3.165223936770794,
+      "grad_norm": 0.2511220872402191,
+      "learning_rate": 3.512160320551906e-05,
+      "loss": 0.2502,
+      "step": 4205
+    },
+    {
+      "epoch": 3.165976665412119,
+      "grad_norm": 0.26736944913864136,
+      "learning_rate": 3.511460205811238e-05,
+      "loss": 0.3101,
+      "step": 4206
+    },
+    {
+      "epoch": 3.166729394053444,
+      "grad_norm": 0.26714661717414856,
+      "learning_rate": 3.5107600077984246e-05,
+      "loss": 0.2197,
+      "step": 4207
+    },
+    {
+      "epoch": 3.1674821226947687,
+      "grad_norm": 0.22154010832309723,
+      "learning_rate": 3.510059726590042e-05,
+      "loss": 0.2419,
+      "step": 4208
+    },
+    {
+      "epoch": 3.168234851336093,
+      "grad_norm": 0.24371710419654846,
+      "learning_rate": 3.5093593622626706e-05,
+      "loss": 0.3048,
+      "step": 4209
+    },
+    {
+      "epoch": 3.168987579977418,
+      "grad_norm": 0.2156084030866623,
+      "learning_rate": 3.508658914892901e-05,
+      "loss": 0.2843,
+      "step": 4210
+    },
+    {
+      "epoch": 3.169740308618743,
+      "grad_norm": 0.2803584337234497,
+      "learning_rate": 3.5079583845573336e-05,
+      "loss": 0.1836,
+      "step": 4211
+    },
+    {
+      "epoch": 3.170493037260068,
+      "grad_norm": 0.21650968492031097,
+      "learning_rate": 3.507257771332576e-05,
+      "loss": 0.2807,
+      "step": 4212
+    },
+    {
+      "epoch": 3.1712457659013924,
+      "grad_norm": 0.2967400848865509,
+      "learning_rate": 3.506557075295247e-05,
+      "loss": 0.3191,
+      "step": 4213
+    },
+    {
+      "epoch": 3.1719984945427173,
+      "grad_norm": 0.257382333278656,
+      "learning_rate": 3.505856296521974e-05,
+      "loss": 0.2294,
+      "step": 4214
+    },
+    {
+      "epoch": 3.1727512231840422,
+      "grad_norm": 0.22964023053646088,
+      "learning_rate": 3.505155435089393e-05,
+      "loss": 0.3274,
+      "step": 4215
+    },
+    {
+      "epoch": 3.173503951825367,
+      "grad_norm": 0.23591181635856628,
+      "learning_rate": 3.504454491074149e-05,
+      "loss": 0.2637,
+      "step": 4216
+    },
+    {
+      "epoch": 3.1742566804666916,
+      "grad_norm": 0.21676062047481537,
+      "learning_rate": 3.503753464552895e-05,
+      "loss": 0.2365,
+      "step": 4217
+    },
+    {
+      "epoch": 3.1750094091080165,
+      "grad_norm": 0.26934340596199036,
+      "learning_rate": 3.5030523556022966e-05,
+      "loss": 0.2643,
+      "step": 4218
+    },
+    {
+      "epoch": 3.1757621377493415,
+      "grad_norm": 0.22267571091651917,
+      "learning_rate": 3.502351164299025e-05,
+      "loss": 0.2449,
+      "step": 4219
+    },
+    {
+      "epoch": 3.176514866390666,
+      "grad_norm": 0.20940953493118286,
+      "learning_rate": 3.50164989071976e-05,
+      "loss": 0.3107,
+      "step": 4220
+    },
+    {
+      "epoch": 3.177267595031991,
+      "grad_norm": 0.21713441610336304,
+      "learning_rate": 3.500948534941194e-05,
+      "loss": 0.2719,
+      "step": 4221
+    },
+    {
+      "epoch": 3.1780203236733158,
+      "grad_norm": 0.2613834738731384,
+      "learning_rate": 3.500247097040025e-05,
+      "loss": 0.305,
+      "step": 4222
+    },
+    {
+      "epoch": 3.1787730523146407,
+      "grad_norm": 0.2170284241437912,
+      "learning_rate": 3.4995455770929616e-05,
+      "loss": 0.1659,
+      "step": 4223
+    },
+    {
+      "epoch": 3.1795257809559656,
+      "grad_norm": 0.26366913318634033,
+      "learning_rate": 3.49884397517672e-05,
+      "loss": 0.2947,
+      "step": 4224
+    },
+    {
+      "epoch": 3.18027850959729,
+      "grad_norm": 0.26526251435279846,
+      "learning_rate": 3.498142291368028e-05,
+      "loss": 0.2878,
+      "step": 4225
+    },
+    {
+      "epoch": 3.181031238238615,
+      "grad_norm": 0.23916034400463104,
+      "learning_rate": 3.49744052574362e-05,
+      "loss": 0.2207,
+      "step": 4226
+    },
+    {
+      "epoch": 3.18178396687994,
+      "grad_norm": 0.25780028104782104,
+      "learning_rate": 3.4967386783802393e-05,
+      "loss": 0.2307,
+      "step": 4227
+    },
+    {
+      "epoch": 3.1825366955212644,
+      "grad_norm": 0.2627592086791992,
+      "learning_rate": 3.4960367493546394e-05,
+      "loss": 0.266,
+      "step": 4228
+    },
+    {
+      "epoch": 3.1832894241625893,
+      "grad_norm": 0.2690248489379883,
+      "learning_rate": 3.495334738743582e-05,
+      "loss": 0.2743,
+      "step": 4229
+    },
+    {
+      "epoch": 3.1840421528039142,
+      "grad_norm": 0.24928145110607147,
+      "learning_rate": 3.494632646623839e-05,
+      "loss": 0.3602,
+      "step": 4230
+    },
+    {
+      "epoch": 3.184794881445239,
+      "grad_norm": 0.2537795305252075,
+      "learning_rate": 3.493930473072188e-05,
+      "loss": 0.2907,
+      "step": 4231
+    },
+    {
+      "epoch": 3.1855476100865636,
+      "grad_norm": 0.2252132147550583,
+      "learning_rate": 3.4932282181654206e-05,
+      "loss": 0.2873,
+      "step": 4232
+    },
+    {
+      "epoch": 3.1863003387278885,
+      "grad_norm": 0.21153025329113007,
+      "learning_rate": 3.492525881980332e-05,
+      "loss": 0.1782,
+      "step": 4233
+    },
+    {
+      "epoch": 3.1870530673692135,
+      "grad_norm": 0.26019200682640076,
+      "learning_rate": 3.491823464593729e-05,
+      "loss": 0.2515,
+      "step": 4234
+    },
+    {
+      "epoch": 3.1878057960105384,
+      "grad_norm": 0.21015417575836182,
+      "learning_rate": 3.491120966082428e-05,
+      "loss": 0.2844,
+      "step": 4235
+    },
+    {
+      "epoch": 3.188558524651863,
+      "grad_norm": 0.24161483347415924,
+      "learning_rate": 3.490418386523252e-05,
+      "loss": 0.2678,
+      "step": 4236
+    },
+    {
+      "epoch": 3.1893112532931878,
+      "grad_norm": 0.24035662412643433,
+      "learning_rate": 3.489715725993033e-05,
+      "loss": 0.2803,
+      "step": 4237
+    },
+    {
+      "epoch": 3.1900639819345127,
+      "grad_norm": 0.23852227628231049,
+      "learning_rate": 3.4890129845686164e-05,
+      "loss": 0.2499,
+      "step": 4238
+    },
+    {
+      "epoch": 3.1908167105758376,
+      "grad_norm": 0.20605511963367462,
+      "learning_rate": 3.488310162326851e-05,
+      "loss": 0.3028,
+      "step": 4239
+    },
+    {
+      "epoch": 3.191569439217162,
+      "grad_norm": 0.26826193928718567,
+      "learning_rate": 3.4876072593445946e-05,
+      "loss": 0.3116,
+      "step": 4240
+    },
+    {
+      "epoch": 3.192322167858487,
+      "grad_norm": 0.23713040351867676,
+      "learning_rate": 3.486904275698719e-05,
+      "loss": 0.2747,
+      "step": 4241
+    },
+    {
+      "epoch": 3.193074896499812,
+      "grad_norm": 0.2200765311717987,
+      "learning_rate": 3.4862012114660985e-05,
+      "loss": 0.2038,
+      "step": 4242
+    },
+    {
+      "epoch": 3.193827625141137,
+      "grad_norm": 0.25719040632247925,
+      "learning_rate": 3.48549806672362e-05,
+      "loss": 0.2728,
+      "step": 4243
+    },
+    {
+      "epoch": 3.1945803537824613,
+      "grad_norm": 0.2508452534675598,
+      "learning_rate": 3.4847948415481804e-05,
+      "loss": 0.2777,
+      "step": 4244
+    },
+    {
+      "epoch": 3.195333082423786,
+      "grad_norm": 0.24424058198928833,
+      "learning_rate": 3.4840915360166806e-05,
+      "loss": 0.297,
+      "step": 4245
+    },
+    {
+      "epoch": 3.196085811065111,
+      "grad_norm": 0.2507564425468445,
+      "learning_rate": 3.483388150206035e-05,
+      "loss": 0.2541,
+      "step": 4246
+    },
+    {
+      "epoch": 3.1968385397064356,
+      "grad_norm": 0.23858338594436646,
+      "learning_rate": 3.482684684193163e-05,
+      "loss": 0.2485,
+      "step": 4247
+    },
+    {
+      "epoch": 3.1975912683477605,
+      "grad_norm": 0.19384023547172546,
+      "learning_rate": 3.481981138054995e-05,
+      "loss": 0.1707,
+      "step": 4248
+    },
+    {
+      "epoch": 3.1983439969890854,
+      "grad_norm": 0.22478212416172028,
+      "learning_rate": 3.4812775118684716e-05,
+      "loss": 0.2093,
+      "step": 4249
+    },
+    {
+      "epoch": 3.1990967256304104,
+      "grad_norm": 0.21306730806827545,
+      "learning_rate": 3.480573805710538e-05,
+      "loss": 0.2847,
+      "step": 4250
+    },
+    {
+      "epoch": 3.199849454271735,
+      "grad_norm": 0.23104806244373322,
+      "learning_rate": 3.4798700196581517e-05,
+      "loss": 0.2177,
+      "step": 4251
+    },
+    {
+      "epoch": 3.2006021829130598,
+      "grad_norm": 0.23389798402786255,
+      "learning_rate": 3.479166153788278e-05,
+      "loss": 0.2011,
+      "step": 4252
+    },
+    {
+      "epoch": 3.2013549115543847,
+      "grad_norm": 0.2128008008003235,
+      "learning_rate": 3.4784622081778896e-05,
+      "loss": 0.2177,
+      "step": 4253
+    },
+    {
+      "epoch": 3.2021076401957096,
+      "grad_norm": 0.2543852925300598,
+      "learning_rate": 3.477758182903968e-05,
+      "loss": 0.2782,
+      "step": 4254
+    },
+    {
+      "epoch": 3.202860368837034,
+      "grad_norm": 0.23819124698638916,
+      "learning_rate": 3.4770540780435075e-05,
+      "loss": 0.2617,
+      "step": 4255
+    },
+    {
+      "epoch": 3.203613097478359,
+      "grad_norm": 0.20858234167099,
+      "learning_rate": 3.476349893673505e-05,
+      "loss": 0.1887,
+      "step": 4256
+    },
+    {
+      "epoch": 3.204365826119684,
+      "grad_norm": 0.224604532122612,
+      "learning_rate": 3.475645629870971e-05,
+      "loss": 0.2129,
+      "step": 4257
+    },
+    {
+      "epoch": 3.205118554761009,
+      "grad_norm": 0.2260502576828003,
+      "learning_rate": 3.474941286712921e-05,
+      "loss": 0.2852,
+      "step": 4258
+    },
+    {
+      "epoch": 3.2058712834023333,
+      "grad_norm": 0.23820757865905762,
+      "learning_rate": 3.474236864276383e-05,
+      "loss": 0.2102,
+      "step": 4259
+    },
+    {
+      "epoch": 3.206624012043658,
+      "grad_norm": 0.22064432501792908,
+      "learning_rate": 3.4735323626383896e-05,
+      "loss": 0.2857,
+      "step": 4260
+    },
+    {
+      "epoch": 3.207376740684983,
+      "grad_norm": 0.19906550645828247,
+      "learning_rate": 3.4728277818759856e-05,
+      "loss": 0.2219,
+      "step": 4261
+    },
+    {
+      "epoch": 3.208129469326308,
+      "grad_norm": 0.26420557498931885,
+      "learning_rate": 3.4721231220662214e-05,
+      "loss": 0.2646,
+      "step": 4262
+    },
+    {
+      "epoch": 3.2088821979676325,
+      "grad_norm": 0.25352802872657776,
+      "learning_rate": 3.4714183832861585e-05,
+      "loss": 0.221,
+      "step": 4263
+    },
+    {
+      "epoch": 3.2096349266089574,
+      "grad_norm": 0.2509932518005371,
+      "learning_rate": 3.4707135656128656e-05,
+      "loss": 0.2106,
+      "step": 4264
+    },
+    {
+      "epoch": 3.2103876552502824,
+      "grad_norm": 0.23016680777072906,
+      "learning_rate": 3.4700086691234205e-05,
+      "loss": 0.2184,
+      "step": 4265
+    },
+    {
+      "epoch": 3.2111403838916073,
+      "grad_norm": 0.2556430995464325,
+      "learning_rate": 3.4693036938949114e-05,
+      "loss": 0.2209,
+      "step": 4266
+    },
+    {
+      "epoch": 3.2118931125329317,
+      "grad_norm": 0.22331488132476807,
+      "learning_rate": 3.4685986400044304e-05,
+      "loss": 0.2659,
+      "step": 4267
+    },
+    {
+      "epoch": 3.2126458411742567,
+      "grad_norm": 0.30740198493003845,
+      "learning_rate": 3.467893507529083e-05,
+      "loss": 0.3106,
+      "step": 4268
+    },
+    {
+      "epoch": 3.2133985698155816,
+      "grad_norm": 0.21525688469409943,
+      "learning_rate": 3.467188296545981e-05,
+      "loss": 0.2461,
+      "step": 4269
+    },
+    {
+      "epoch": 3.2141512984569065,
+      "grad_norm": 0.2479272335767746,
+      "learning_rate": 3.4664830071322454e-05,
+      "loss": 0.2433,
+      "step": 4270
+    },
+    {
+      "epoch": 3.214904027098231,
+      "grad_norm": 0.24682079255580902,
+      "learning_rate": 3.465777639365005e-05,
+      "loss": 0.2542,
+      "step": 4271
+    },
+    {
+      "epoch": 3.215656755739556,
+      "grad_norm": 0.2516164779663086,
+      "learning_rate": 3.465072193321397e-05,
+      "loss": 0.2834,
+      "step": 4272
+    },
+    {
+      "epoch": 3.216409484380881,
+      "grad_norm": 0.2666584551334381,
+      "learning_rate": 3.46436666907857e-05,
+      "loss": 0.3251,
+      "step": 4273
+    },
+    {
+      "epoch": 3.2171622130222053,
+      "grad_norm": 0.24124675989151,
+      "learning_rate": 3.463661066713679e-05,
+      "loss": 0.2889,
+      "step": 4274
+    },
+    {
+      "epoch": 3.21791494166353,
+      "grad_norm": 0.2359425276517868,
+      "learning_rate": 3.462955386303885e-05,
+      "loss": 0.268,
+      "step": 4275
+    },
+    {
+      "epoch": 3.218667670304855,
+      "grad_norm": 0.23599910736083984,
+      "learning_rate": 3.462249627926363e-05,
+      "loss": 0.2618,
+      "step": 4276
+    },
+    {
+      "epoch": 3.21942039894618,
+      "grad_norm": 0.24697864055633545,
+      "learning_rate": 3.461543791658292e-05,
+      "loss": 0.2785,
+      "step": 4277
+    },
+    {
+      "epoch": 3.2201731275875045,
+      "grad_norm": 0.2254694551229477,
+      "learning_rate": 3.460837877576862e-05,
+      "loss": 0.2935,
+      "step": 4278
+    },
+    {
+      "epoch": 3.2209258562288294,
+      "grad_norm": 0.21764926612377167,
+      "learning_rate": 3.46013188575927e-05,
+      "loss": 0.2397,
+      "step": 4279
+    },
+    {
+      "epoch": 3.2216785848701543,
+      "grad_norm": 0.21717418730258942,
+      "learning_rate": 3.459425816282724e-05,
+      "loss": 0.1858,
+      "step": 4280
+    },
+    {
+      "epoch": 3.2224313135114793,
+      "grad_norm": 0.27352797985076904,
+      "learning_rate": 3.458719669224436e-05,
+      "loss": 0.3393,
+      "step": 4281
+    },
+    {
+      "epoch": 3.2231840421528037,
+      "grad_norm": 0.2252342253923416,
+      "learning_rate": 3.4580134446616306e-05,
+      "loss": 0.2835,
+      "step": 4282
+    },
+    {
+      "epoch": 3.2239367707941287,
+      "grad_norm": 0.23820818960666656,
+      "learning_rate": 3.45730714267154e-05,
+      "loss": 0.252,
+      "step": 4283
+    },
+    {
+      "epoch": 3.2246894994354536,
+      "grad_norm": 0.2423304319381714,
+      "learning_rate": 3.456600763331403e-05,
+      "loss": 0.236,
+      "step": 4284
+    },
+    {
+      "epoch": 3.2254422280767785,
+      "grad_norm": 0.24676740169525146,
+      "learning_rate": 3.45589430671847e-05,
+      "loss": 0.2547,
+      "step": 4285
+    },
+    {
+      "epoch": 3.226194956718103,
+      "grad_norm": 0.2882114052772522,
+      "learning_rate": 3.455187772909996e-05,
+      "loss": 0.2703,
+      "step": 4286
+    },
+    {
+      "epoch": 3.226947685359428,
+      "grad_norm": 0.24428004026412964,
+      "learning_rate": 3.454481161983247e-05,
+      "loss": 0.3029,
+      "step": 4287
+    },
+    {
+      "epoch": 3.227700414000753,
+      "grad_norm": 0.23090480268001556,
+      "learning_rate": 3.453774474015499e-05,
+      "loss": 0.2595,
+      "step": 4288
+    },
+    {
+      "epoch": 3.2284531426420777,
+      "grad_norm": 0.20903180539608002,
+      "learning_rate": 3.453067709084031e-05,
+      "loss": 0.299,
+      "step": 4289
+    },
+    {
+      "epoch": 3.229205871283402,
+      "grad_norm": 0.21339118480682373,
+      "learning_rate": 3.452360867266136e-05,
+      "loss": 0.3278,
+      "step": 4290
+    },
+    {
+      "epoch": 3.229958599924727,
+      "grad_norm": 0.24371756613254547,
+      "learning_rate": 3.451653948639112e-05,
+      "loss": 0.3553,
+      "step": 4291
+    },
+    {
+      "epoch": 3.230711328566052,
+      "grad_norm": 0.21457231044769287,
+      "learning_rate": 3.450946953280268e-05,
+      "loss": 0.3678,
+      "step": 4292
+    },
+    {
+      "epoch": 3.2314640572073765,
+      "grad_norm": 0.22485870122909546,
+      "learning_rate": 3.450239881266919e-05,
+      "loss": 0.2752,
+      "step": 4293
+    },
+    {
+      "epoch": 3.2322167858487014,
+      "grad_norm": 0.22895142436027527,
+      "learning_rate": 3.449532732676388e-05,
+      "loss": 0.3241,
+      "step": 4294
+    },
+    {
+      "epoch": 3.2329695144900263,
+      "grad_norm": 0.24345633387565613,
+      "learning_rate": 3.4488255075860105e-05,
+      "loss": 0.3054,
+      "step": 4295
+    },
+    {
+      "epoch": 3.2337222431313513,
+      "grad_norm": 0.24971047043800354,
+      "learning_rate": 3.448118206073127e-05,
+      "loss": 0.3001,
+      "step": 4296
+    },
+    {
+      "epoch": 3.234474971772676,
+      "grad_norm": 0.22359171509742737,
+      "learning_rate": 3.447410828215083e-05,
+      "loss": 0.196,
+      "step": 4297
+    },
+    {
+      "epoch": 3.2352277004140007,
+      "grad_norm": 0.23416151106357574,
+      "learning_rate": 3.446703374089242e-05,
+      "loss": 0.2415,
+      "step": 4298
+    },
+    {
+      "epoch": 3.2359804290553256,
+      "grad_norm": 0.20930065214633942,
+      "learning_rate": 3.445995843772967e-05,
+      "loss": 0.2542,
+      "step": 4299
+    },
+    {
+      "epoch": 3.2367331576966505,
+      "grad_norm": 0.2042551189661026,
+      "learning_rate": 3.4452882373436316e-05,
+      "loss": 0.2527,
+      "step": 4300
+    },
+    {
+      "epoch": 3.237485886337975,
+      "grad_norm": 0.25222691893577576,
+      "learning_rate": 3.444580554878621e-05,
+      "loss": 0.2362,
+      "step": 4301
+    },
+    {
+      "epoch": 3.2382386149793,
+      "grad_norm": 0.26191022992134094,
+      "learning_rate": 3.443872796455325e-05,
+      "loss": 0.3028,
+      "step": 4302
+    },
+    {
+      "epoch": 3.238991343620625,
+      "grad_norm": 0.24931985139846802,
+      "learning_rate": 3.443164962151143e-05,
+      "loss": 0.1997,
+      "step": 4303
+    },
+    {
+      "epoch": 3.2397440722619497,
+      "grad_norm": 0.24336226284503937,
+      "learning_rate": 3.442457052043484e-05,
+      "loss": 0.26,
+      "step": 4304
+    },
+    {
+      "epoch": 3.240496800903274,
+      "grad_norm": 0.22032158076763153,
+      "learning_rate": 3.441749066209762e-05,
+      "loss": 0.2642,
+      "step": 4305
+    },
+    {
+      "epoch": 3.241249529544599,
+      "grad_norm": 0.24513062834739685,
+      "learning_rate": 3.441041004727402e-05,
+      "loss": 0.2627,
+      "step": 4306
+    },
+    {
+      "epoch": 3.242002258185924,
+      "grad_norm": 0.22943951189517975,
+      "learning_rate": 3.440332867673837e-05,
+      "loss": 0.1945,
+      "step": 4307
+    },
+    {
+      "epoch": 3.242754986827249,
+      "grad_norm": 0.24166247248649597,
+      "learning_rate": 3.439624655126507e-05,
+      "loss": 0.2244,
+      "step": 4308
+    },
+    {
+      "epoch": 3.2435077154685734,
+      "grad_norm": 0.23216578364372253,
+      "learning_rate": 3.4389163671628624e-05,
+      "loss": 0.2021,
+      "step": 4309
+    },
+    {
+      "epoch": 3.2442604441098983,
+      "grad_norm": 0.2251998484134674,
+      "learning_rate": 3.438208003860359e-05,
+      "loss": 0.244,
+      "step": 4310
+    },
+    {
+      "epoch": 3.2450131727512233,
+      "grad_norm": 0.25227245688438416,
+      "learning_rate": 3.4374995652964636e-05,
+      "loss": 0.2853,
+      "step": 4311
+    },
+    {
+      "epoch": 3.245765901392548,
+      "grad_norm": 0.2345009446144104,
+      "learning_rate": 3.436791051548649e-05,
+      "loss": 0.1993,
+      "step": 4312
+    },
+    {
+      "epoch": 3.2465186300338726,
+      "grad_norm": 0.2782454788684845,
+      "learning_rate": 3.4360824626943976e-05,
+      "loss": 0.2985,
+      "step": 4313
+    },
+    {
+      "epoch": 3.2472713586751976,
+      "grad_norm": 0.26159006357192993,
+      "learning_rate": 3.4353737988111994e-05,
+      "loss": 0.302,
+      "step": 4314
+    },
+    {
+      "epoch": 3.2480240873165225,
+      "grad_norm": 0.2388232946395874,
+      "learning_rate": 3.434665059976554e-05,
+      "loss": 0.1728,
+      "step": 4315
+    },
+    {
+      "epoch": 3.2487768159578474,
+      "grad_norm": 0.2112424075603485,
+      "learning_rate": 3.4339562462679654e-05,
+      "loss": 0.2811,
+      "step": 4316
+    },
+    {
+      "epoch": 3.249529544599172,
+      "grad_norm": 0.234798401594162,
+      "learning_rate": 3.433247357762951e-05,
+      "loss": 0.2258,
+      "step": 4317
+    },
+    {
+      "epoch": 3.250282273240497,
+      "grad_norm": 0.26226380467414856,
+      "learning_rate": 3.4325383945390334e-05,
+      "loss": 0.2231,
+      "step": 4318
+    },
+    {
+      "epoch": 3.2510350018818217,
+      "grad_norm": 0.2538587152957916,
+      "learning_rate": 3.431829356673742e-05,
+      "loss": 0.2514,
+      "step": 4319
+    },
+    {
+      "epoch": 3.251787730523146,
+      "grad_norm": 0.20693328976631165,
+      "learning_rate": 3.431120244244617e-05,
+      "loss": 0.2478,
+      "step": 4320
+    },
+    {
+      "epoch": 3.252540459164471,
+      "grad_norm": 0.2505183219909668,
+      "learning_rate": 3.4304110573292075e-05,
+      "loss": 0.2898,
+      "step": 4321
+    },
+    {
+      "epoch": 3.253293187805796,
+      "grad_norm": 0.2534908652305603,
+      "learning_rate": 3.429701796005066e-05,
+      "loss": 0.2597,
+      "step": 4322
+    },
+    {
+      "epoch": 3.254045916447121,
+      "grad_norm": 0.24066928029060364,
+      "learning_rate": 3.4289924603497596e-05,
+      "loss": 0.1885,
+      "step": 4323
+    },
+    {
+      "epoch": 3.254798645088446,
+      "grad_norm": 0.2931833565235138,
+      "learning_rate": 3.428283050440858e-05,
+      "loss": 0.2826,
+      "step": 4324
+    },
+    {
+      "epoch": 3.2555513737297703,
+      "grad_norm": 0.22506831586360931,
+      "learning_rate": 3.427573566355942e-05,
+      "loss": 0.2421,
+      "step": 4325
+    },
+    {
+      "epoch": 3.2563041023710952,
+      "grad_norm": 0.2232486456632614,
+      "learning_rate": 3.4268640081725996e-05,
+      "loss": 0.4114,
+      "step": 4326
+    },
+    {
+      "epoch": 3.25705683101242,
+      "grad_norm": 0.2331424206495285,
+      "learning_rate": 3.426154375968426e-05,
+      "loss": 0.247,
+      "step": 4327
+    },
+    {
+      "epoch": 3.2578095596537446,
+      "grad_norm": 0.26332470774650574,
+      "learning_rate": 3.4254446698210276e-05,
+      "loss": 0.2539,
+      "step": 4328
+    },
+    {
+      "epoch": 3.2585622882950696,
+      "grad_norm": 0.274596244096756,
+      "learning_rate": 3.4247348898080145e-05,
+      "loss": 0.2428,
+      "step": 4329
+    },
+    {
+      "epoch": 3.2593150169363945,
+      "grad_norm": 0.24902930855751038,
+      "learning_rate": 3.424025036007009e-05,
+      "loss": 0.2363,
+      "step": 4330
+    },
+    {
+      "epoch": 3.2600677455777194,
+      "grad_norm": 0.23229530453681946,
+      "learning_rate": 3.42331510849564e-05,
+      "loss": 0.233,
+      "step": 4331
+    },
+    {
+      "epoch": 3.260820474219044,
+      "grad_norm": 0.19760465621948242,
+      "learning_rate": 3.4226051073515406e-05,
+      "loss": 0.2183,
+      "step": 4332
+    },
+    {
+      "epoch": 3.261573202860369,
+      "grad_norm": 0.19229428470134735,
+      "learning_rate": 3.4218950326523594e-05,
+      "loss": 0.1949,
+      "step": 4333
+    },
+    {
+      "epoch": 3.2623259315016937,
+      "grad_norm": 0.23052626848220825,
+      "learning_rate": 3.421184884475748e-05,
+      "loss": 0.256,
+      "step": 4334
+    },
+    {
+      "epoch": 3.2630786601430186,
+      "grad_norm": 0.2808074653148651,
+      "learning_rate": 3.420474662899366e-05,
+      "loss": 0.2447,
+      "step": 4335
+    },
+    {
+      "epoch": 3.263831388784343,
+      "grad_norm": 0.2385113537311554,
+      "learning_rate": 3.419764368000883e-05,
+      "loss": 0.161,
+      "step": 4336
+    },
+    {
+      "epoch": 3.264584117425668,
+      "grad_norm": 0.20850436389446259,
+      "learning_rate": 3.4190539998579756e-05,
+      "loss": 0.1876,
+      "step": 4337
+    },
+    {
+      "epoch": 3.265336846066993,
+      "grad_norm": 0.23445850610733032,
+      "learning_rate": 3.4183435585483286e-05,
+      "loss": 0.2346,
+      "step": 4338
+    },
+    {
+      "epoch": 3.266089574708318,
+      "grad_norm": 0.22819551825523376,
+      "learning_rate": 3.417633044149635e-05,
+      "loss": 0.2399,
+      "step": 4339
+    },
+    {
+      "epoch": 3.2668423033496423,
+      "grad_norm": 0.19153788685798645,
+      "learning_rate": 3.416922456739596e-05,
+      "loss": 0.2095,
+      "step": 4340
+    },
+    {
+      "epoch": 3.2675950319909672,
+      "grad_norm": 0.24742184579372406,
+      "learning_rate": 3.4162117963959194e-05,
+      "loss": 0.306,
+      "step": 4341
+    },
+    {
+      "epoch": 3.268347760632292,
+      "grad_norm": 0.2662699222564697,
+      "learning_rate": 3.4155010631963226e-05,
+      "loss": 0.2473,
+      "step": 4342
+    },
+    {
+      "epoch": 3.269100489273617,
+      "grad_norm": 0.24540211260318756,
+      "learning_rate": 3.4147902572185295e-05,
+      "loss": 0.3057,
+      "step": 4343
+    },
+    {
+      "epoch": 3.2698532179149415,
+      "grad_norm": 0.22312702238559723,
+      "learning_rate": 3.414079378540274e-05,
+      "loss": 0.3744,
+      "step": 4344
+    },
+    {
+      "epoch": 3.2706059465562665,
+      "grad_norm": 0.23626141250133514,
+      "learning_rate": 3.413368427239296e-05,
+      "loss": 0.2765,
+      "step": 4345
+    },
+    {
+      "epoch": 3.2713586751975914,
+      "grad_norm": 0.24758140742778778,
+      "learning_rate": 3.4126574033933436e-05,
+      "loss": 0.2323,
+      "step": 4346
+    },
+    {
+      "epoch": 3.272111403838916,
+      "grad_norm": 0.23340576887130737,
+      "learning_rate": 3.411946307080175e-05,
+      "loss": 0.2099,
+      "step": 4347
+    },
+    {
+      "epoch": 3.2728641324802408,
+      "grad_norm": 0.24228236079216003,
+      "learning_rate": 3.411235138377552e-05,
+      "loss": 0.2165,
+      "step": 4348
+    },
+    {
+      "epoch": 3.2736168611215657,
+      "grad_norm": 0.22364583611488342,
+      "learning_rate": 3.4105238973632495e-05,
+      "loss": 0.3081,
+      "step": 4349
+    },
+    {
+      "epoch": 3.2743695897628906,
+      "grad_norm": 0.22426077723503113,
+      "learning_rate": 3.4098125841150466e-05,
+      "loss": 0.2032,
+      "step": 4350
+    },
+    {
+      "epoch": 3.2751223184042155,
+      "grad_norm": 0.2325281798839569,
+      "learning_rate": 3.409101198710731e-05,
+      "loss": 0.2501,
+      "step": 4351
+    },
+    {
+      "epoch": 3.27587504704554,
+      "grad_norm": 0.24281547963619232,
+      "learning_rate": 3.408389741228099e-05,
+      "loss": 0.298,
+      "step": 4352
+    },
+    {
+      "epoch": 3.276627775686865,
+      "grad_norm": 0.2334796041250229,
+      "learning_rate": 3.407678211744955e-05,
+      "loss": 0.3272,
+      "step": 4353
+    },
+    {
+      "epoch": 3.27738050432819,
+      "grad_norm": 0.15186311304569244,
+      "learning_rate": 3.406966610339111e-05,
+      "loss": 0.2228,
+      "step": 4354
+    },
+    {
+      "epoch": 3.2781332329695143,
+      "grad_norm": 0.20660027861595154,
+      "learning_rate": 3.406254937088386e-05,
+      "loss": 0.1738,
+      "step": 4355
+    },
+    {
+      "epoch": 3.2788859616108392,
+      "grad_norm": 0.21992743015289307,
+      "learning_rate": 3.405543192070608e-05,
+      "loss": 0.2987,
+      "step": 4356
+    },
+    {
+      "epoch": 3.279638690252164,
+      "grad_norm": 0.22646485269069672,
+      "learning_rate": 3.404831375363611e-05,
+      "loss": 0.2903,
+      "step": 4357
+    },
+    {
+      "epoch": 3.280391418893489,
+      "grad_norm": 0.20550431311130524,
+      "learning_rate": 3.404119487045241e-05,
+      "loss": 0.1767,
+      "step": 4358
+    },
+    {
+      "epoch": 3.2811441475348135,
+      "grad_norm": 0.2229396104812622,
+      "learning_rate": 3.403407527193347e-05,
+      "loss": 0.2315,
+      "step": 4359
+    },
+    {
+      "epoch": 3.2818968761761385,
+      "grad_norm": 0.20946243405342102,
+      "learning_rate": 3.402695495885788e-05,
+      "loss": 0.2619,
+      "step": 4360
+    },
+    {
+      "epoch": 3.2826496048174634,
+      "grad_norm": 0.2421065866947174,
+      "learning_rate": 3.40198339320043e-05,
+      "loss": 0.2696,
+      "step": 4361
+    },
+    {
+      "epoch": 3.2834023334587883,
+      "grad_norm": 0.2484889030456543,
+      "learning_rate": 3.401271219215149e-05,
+      "loss": 0.2212,
+      "step": 4362
+    },
+    {
+      "epoch": 3.2841550621001128,
+      "grad_norm": 0.24150028824806213,
+      "learning_rate": 3.4005589740078275e-05,
+      "loss": 0.2244,
+      "step": 4363
+    },
+    {
+      "epoch": 3.2849077907414377,
+      "grad_norm": 0.25528982281684875,
+      "learning_rate": 3.399846657656353e-05,
+      "loss": 0.2598,
+      "step": 4364
+    },
+    {
+      "epoch": 3.2856605193827626,
+      "grad_norm": 0.21575923264026642,
+      "learning_rate": 3.3991342702386265e-05,
+      "loss": 0.2208,
+      "step": 4365
+    },
+    {
+      "epoch": 3.286413248024087,
+      "grad_norm": 0.22954390943050385,
+      "learning_rate": 3.398421811832552e-05,
+      "loss": 0.3262,
+      "step": 4366
+    },
+    {
+      "epoch": 3.287165976665412,
+      "grad_norm": 0.21661384403705597,
+      "learning_rate": 3.397709282516042e-05,
+      "loss": 0.247,
+      "step": 4367
+    },
+    {
+      "epoch": 3.287918705306737,
+      "grad_norm": 0.24700552225112915,
+      "learning_rate": 3.396996682367019e-05,
+      "loss": 0.2672,
+      "step": 4368
+    },
+    {
+      "epoch": 3.288671433948062,
+      "grad_norm": 0.23790591955184937,
+      "learning_rate": 3.396284011463413e-05,
+      "loss": 0.2133,
+      "step": 4369
+    },
+    {
+      "epoch": 3.2894241625893867,
+      "grad_norm": 0.23531661927700043,
+      "learning_rate": 3.3955712698831574e-05,
+      "loss": 0.221,
+      "step": 4370
+    },
+    {
+      "epoch": 3.290176891230711,
+      "grad_norm": 0.25075089931488037,
+      "learning_rate": 3.3948584577041987e-05,
+      "loss": 0.2251,
+      "step": 4371
+    },
+    {
+      "epoch": 3.290929619872036,
+      "grad_norm": 0.19639582931995392,
+      "learning_rate": 3.394145575004489e-05,
+      "loss": 0.2393,
+      "step": 4372
+    },
+    {
+      "epoch": 3.291682348513361,
+      "grad_norm": 0.2326192557811737,
+      "learning_rate": 3.393432621861987e-05,
+      "loss": 0.2083,
+      "step": 4373
+    },
+    {
+      "epoch": 3.2924350771546855,
+      "grad_norm": 0.2613600194454193,
+      "learning_rate": 3.392719598354662e-05,
+      "loss": 0.2433,
+      "step": 4374
+    },
+    {
+      "epoch": 3.2931878057960104,
+      "grad_norm": 0.21137025952339172,
+      "learning_rate": 3.392006504560487e-05,
+      "loss": 0.3267,
+      "step": 4375
+    },
+    {
+      "epoch": 3.2939405344373354,
+      "grad_norm": 0.19189059734344482,
+      "learning_rate": 3.391293340557447e-05,
+      "loss": 0.1999,
+      "step": 4376
+    },
+    {
+      "epoch": 3.2946932630786603,
+      "grad_norm": 0.24361777305603027,
+      "learning_rate": 3.39058010642353e-05,
+      "loss": 0.2767,
+      "step": 4377
+    },
+    {
+      "epoch": 3.2954459917199848,
+      "grad_norm": 0.2401793748140335,
+      "learning_rate": 3.389866802236737e-05,
+      "loss": 0.3342,
+      "step": 4378
+    },
+    {
+      "epoch": 3.2961987203613097,
+      "grad_norm": 0.28771331906318665,
+      "learning_rate": 3.389153428075072e-05,
+      "loss": 0.273,
+      "step": 4379
+    },
+    {
+      "epoch": 3.2969514490026346,
+      "grad_norm": 0.2652786076068878,
+      "learning_rate": 3.388439984016549e-05,
+      "loss": 0.2781,
+      "step": 4380
+    },
+    {
+      "epoch": 3.2977041776439595,
+      "grad_norm": 0.22487470507621765,
+      "learning_rate": 3.387726470139189e-05,
+      "loss": 0.1935,
+      "step": 4381
+    },
+    {
+      "epoch": 3.298456906285284,
+      "grad_norm": 0.2834269106388092,
+      "learning_rate": 3.3870128865210216e-05,
+      "loss": 0.1868,
+      "step": 4382
+    },
+    {
+      "epoch": 3.299209634926609,
+      "grad_norm": 0.21680234372615814,
+      "learning_rate": 3.3862992332400817e-05,
+      "loss": 0.2843,
+      "step": 4383
+    },
+    {
+      "epoch": 3.299962363567934,
+      "grad_norm": 0.22775641083717346,
+      "learning_rate": 3.3855855103744154e-05,
+      "loss": 0.2091,
+      "step": 4384
+    },
+    {
+      "epoch": 3.3007150922092587,
+      "grad_norm": 0.29582053422927856,
+      "learning_rate": 3.384871718002073e-05,
+      "loss": 0.2893,
+      "step": 4385
+    },
+    {
+      "epoch": 3.301467820850583,
+      "grad_norm": 0.2356960028409958,
+      "learning_rate": 3.3841578562011136e-05,
+      "loss": 0.2064,
+      "step": 4386
+    },
+    {
+      "epoch": 3.302220549491908,
+      "grad_norm": 0.2569403052330017,
+      "learning_rate": 3.383443925049605e-05,
+      "loss": 0.2329,
+      "step": 4387
+    },
+    {
+      "epoch": 3.302973278133233,
+      "grad_norm": 0.208156555891037,
+      "learning_rate": 3.382729924625621e-05,
+      "loss": 0.245,
+      "step": 4388
+    },
+    {
+      "epoch": 3.303726006774558,
+      "grad_norm": 0.258470743894577,
+      "learning_rate": 3.3820158550072436e-05,
+      "loss": 0.3053,
+      "step": 4389
+    },
+    {
+      "epoch": 3.3044787354158824,
+      "grad_norm": 0.2330002635717392,
+      "learning_rate": 3.381301716272562e-05,
+      "loss": 0.2561,
+      "step": 4390
+    },
+    {
+      "epoch": 3.3052314640572074,
+      "grad_norm": 0.26110371947288513,
+      "learning_rate": 3.3805875084996754e-05,
+      "loss": 0.2079,
+      "step": 4391
+    },
+    {
+      "epoch": 3.3059841926985323,
+      "grad_norm": 0.17518791556358337,
+      "learning_rate": 3.379873231766686e-05,
+      "loss": 0.2349,
+      "step": 4392
+    },
+    {
+      "epoch": 3.3067369213398567,
+      "grad_norm": 0.26016706228256226,
+      "learning_rate": 3.3791588861517066e-05,
+      "loss": 0.2708,
+      "step": 4393
+    },
+    {
+      "epoch": 3.3074896499811817,
+      "grad_norm": 0.20838241279125214,
+      "learning_rate": 3.3784444717328576e-05,
+      "loss": 0.1652,
+      "step": 4394
+    },
+    {
+      "epoch": 3.3082423786225066,
+      "grad_norm": 0.23531432449817657,
+      "learning_rate": 3.3777299885882666e-05,
+      "loss": 0.2393,
+      "step": 4395
+    },
+    {
+      "epoch": 3.3089951072638315,
+      "grad_norm": 0.1995512992143631,
+      "learning_rate": 3.377015436796067e-05,
+      "loss": 0.2307,
+      "step": 4396
+    },
+    {
+      "epoch": 3.3097478359051564,
+      "grad_norm": 0.2937281131744385,
+      "learning_rate": 3.3763008164344026e-05,
+      "loss": 0.3907,
+      "step": 4397
+    },
+    {
+      "epoch": 3.310500564546481,
+      "grad_norm": 0.24185779690742493,
+      "learning_rate": 3.3755861275814216e-05,
+      "loss": 0.2778,
+      "step": 4398
+    },
+    {
+      "epoch": 3.311253293187806,
+      "grad_norm": 0.1959562599658966,
+      "learning_rate": 3.3748713703152834e-05,
+      "loss": 0.2137,
+      "step": 4399
+    },
+    {
+      "epoch": 3.3120060218291307,
+      "grad_norm": 0.2420838326215744,
+      "learning_rate": 3.37415654471415e-05,
+      "loss": 0.2475,
+      "step": 4400
+    },
+    {
+      "epoch": 3.3120060218291307,
+      "eval_loss": 0.26956143975257874,
+      "eval_runtime": 452.7914,
+      "eval_samples_per_second": 21.261,
+      "eval_steps_per_second": 0.665,
+      "step": 4400
+    },
+    {
+      "epoch": 3.312758750470455,
+      "grad_norm": 0.25560662150382996,
+      "learning_rate": 3.373441650856197e-05,
+      "loss": 0.3283,
+      "step": 4401
+    },
+    {
+      "epoch": 3.31351147911178,
+      "grad_norm": 0.24067647755146027,
+      "learning_rate": 3.3727266888196016e-05,
+      "loss": 0.2277,
+      "step": 4402
+    },
+    {
+      "epoch": 3.314264207753105,
+      "grad_norm": 0.22620703279972076,
+      "learning_rate": 3.3720116586825517e-05,
+      "loss": 0.2231,
+      "step": 4403
+    },
+    {
+      "epoch": 3.31501693639443,
+      "grad_norm": 0.2534901201725006,
+      "learning_rate": 3.371296560523242e-05,
+      "loss": 0.2648,
+      "step": 4404
+    },
+    {
+      "epoch": 3.3157696650357544,
+      "grad_norm": 0.2342853844165802,
+      "learning_rate": 3.370581394419875e-05,
+      "loss": 0.2238,
+      "step": 4405
+    },
+    {
+      "epoch": 3.3165223936770794,
+      "grad_norm": 0.2611262798309326,
+      "learning_rate": 3.36986616045066e-05,
+      "loss": 0.2008,
+      "step": 4406
+    },
+    {
+      "epoch": 3.3172751223184043,
+      "grad_norm": 0.2356398105621338,
+      "learning_rate": 3.369150858693813e-05,
+      "loss": 0.3029,
+      "step": 4407
+    },
+    {
+      "epoch": 3.318027850959729,
+      "grad_norm": 0.22626996040344238,
+      "learning_rate": 3.3684354892275586e-05,
+      "loss": 0.216,
+      "step": 4408
+    },
+    {
+      "epoch": 3.3187805796010537,
+      "grad_norm": 0.21779298782348633,
+      "learning_rate": 3.3677200521301294e-05,
+      "loss": 0.2968,
+      "step": 4409
+    },
+    {
+      "epoch": 3.3195333082423786,
+      "grad_norm": 0.2144898772239685,
+      "learning_rate": 3.367004547479764e-05,
+      "loss": 0.2423,
+      "step": 4410
+    },
+    {
+      "epoch": 3.3202860368837035,
+      "grad_norm": 0.2062818557024002,
+      "learning_rate": 3.3662889753547096e-05,
+      "loss": 0.2558,
+      "step": 4411
+    },
+    {
+      "epoch": 3.3210387655250284,
+      "grad_norm": 0.22228798270225525,
+      "learning_rate": 3.3655733358332186e-05,
+      "loss": 0.216,
+      "step": 4412
+    },
+    {
+      "epoch": 3.321791494166353,
+      "grad_norm": 0.22566427290439606,
+      "learning_rate": 3.364857628993553e-05,
+      "loss": 0.2775,
+      "step": 4413
+    },
+    {
+      "epoch": 3.322544222807678,
+      "grad_norm": 0.21886609494686127,
+      "learning_rate": 3.364141854913982e-05,
+      "loss": 0.2129,
+      "step": 4414
+    },
+    {
+      "epoch": 3.3232969514490027,
+      "grad_norm": 0.23252087831497192,
+      "learning_rate": 3.363426013672781e-05,
+      "loss": 0.2518,
+      "step": 4415
+    },
+    {
+      "epoch": 3.3240496800903276,
+      "grad_norm": 0.20676042139530182,
+      "learning_rate": 3.3627101053482336e-05,
+      "loss": 0.3293,
+      "step": 4416
+    },
+    {
+      "epoch": 3.324802408731652,
+      "grad_norm": 0.25676387548446655,
+      "learning_rate": 3.36199413001863e-05,
+      "loss": 0.321,
+      "step": 4417
+    },
+    {
+      "epoch": 3.325555137372977,
+      "grad_norm": 0.20490726828575134,
+      "learning_rate": 3.361278087762268e-05,
+      "loss": 0.1677,
+      "step": 4418
+    },
+    {
+      "epoch": 3.326307866014302,
+      "grad_norm": 0.24005365371704102,
+      "learning_rate": 3.360561978657454e-05,
+      "loss": 0.2441,
+      "step": 4419
+    },
+    {
+      "epoch": 3.3270605946556264,
+      "grad_norm": 0.21559499204158783,
+      "learning_rate": 3.3598458027825004e-05,
+      "loss": 0.3003,
+      "step": 4420
+    },
+    {
+      "epoch": 3.3278133232969513,
+      "grad_norm": 0.25588542222976685,
+      "learning_rate": 3.359129560215725e-05,
+      "loss": 0.2237,
+      "step": 4421
+    },
+    {
+      "epoch": 3.3285660519382763,
+      "grad_norm": 0.19732153415679932,
+      "learning_rate": 3.358413251035458e-05,
+      "loss": 0.2431,
+      "step": 4422
+    },
+    {
+      "epoch": 3.329318780579601,
+      "grad_norm": 0.24252808094024658,
+      "learning_rate": 3.3576968753200324e-05,
+      "loss": 0.3151,
+      "step": 4423
+    },
+    {
+      "epoch": 3.330071509220926,
+      "grad_norm": 0.20880044996738434,
+      "learning_rate": 3.35698043314779e-05,
+      "loss": 0.2319,
+      "step": 4424
+    },
+    {
+      "epoch": 3.3308242378622506,
+      "grad_norm": 0.23722043633460999,
+      "learning_rate": 3.356263924597081e-05,
+      "loss": 0.1832,
+      "step": 4425
+    },
+    {
+      "epoch": 3.3315769665035755,
+      "grad_norm": 0.243950754404068,
+      "learning_rate": 3.355547349746259e-05,
+      "loss": 0.311,
+      "step": 4426
+    },
+    {
+      "epoch": 3.3323296951449004,
+      "grad_norm": 0.25246044993400574,
+      "learning_rate": 3.354830708673691e-05,
+      "loss": 0.3065,
+      "step": 4427
+    },
+    {
+      "epoch": 3.333082423786225,
+      "grad_norm": 0.230362668633461,
+      "learning_rate": 3.354114001457744e-05,
+      "loss": 0.2471,
+      "step": 4428
+    },
+    {
+      "epoch": 3.33383515242755,
+      "grad_norm": 0.2339286506175995,
+      "learning_rate": 3.353397228176801e-05,
+      "loss": 0.189,
+      "step": 4429
+    },
+    {
+      "epoch": 3.3345878810688747,
+      "grad_norm": 0.2191702276468277,
+      "learning_rate": 3.3526803889092416e-05,
+      "loss": 0.2598,
+      "step": 4430
+    },
+    {
+      "epoch": 3.3353406097101996,
+      "grad_norm": 0.21600094437599182,
+      "learning_rate": 3.351963483733462e-05,
+      "loss": 0.2983,
+      "step": 4431
+    },
+    {
+      "epoch": 3.336093338351524,
+      "grad_norm": 0.24301683902740479,
+      "learning_rate": 3.351246512727861e-05,
+      "loss": 0.2827,
+      "step": 4432
+    },
+    {
+      "epoch": 3.336846066992849,
+      "grad_norm": 0.22888097167015076,
+      "learning_rate": 3.3505294759708454e-05,
+      "loss": 0.2256,
+      "step": 4433
+    },
+    {
+      "epoch": 3.337598795634174,
+      "grad_norm": 0.2652900218963623,
+      "learning_rate": 3.349812373540829e-05,
+      "loss": 0.3099,
+      "step": 4434
+    },
+    {
+      "epoch": 3.338351524275499,
+      "grad_norm": 0.2860013246536255,
+      "learning_rate": 3.349095205516233e-05,
+      "loss": 0.3054,
+      "step": 4435
+    },
+    {
+      "epoch": 3.3391042529168233,
+      "grad_norm": 0.23592554032802582,
+      "learning_rate": 3.3483779719754865e-05,
+      "loss": 0.2538,
+      "step": 4436
+    },
+    {
+      "epoch": 3.3398569815581483,
+      "grad_norm": 0.30981314182281494,
+      "learning_rate": 3.3476606729970236e-05,
+      "loss": 0.2693,
+      "step": 4437
+    },
+    {
+      "epoch": 3.340609710199473,
+      "grad_norm": 0.23273180425167084,
+      "learning_rate": 3.3469433086592895e-05,
+      "loss": 0.2394,
+      "step": 4438
+    },
+    {
+      "epoch": 3.3413624388407976,
+      "grad_norm": 0.2695595324039459,
+      "learning_rate": 3.346225879040732e-05,
+      "loss": 0.2323,
+      "step": 4439
+    },
+    {
+      "epoch": 3.3421151674821226,
+      "grad_norm": 0.21806611120700836,
+      "learning_rate": 3.345508384219809e-05,
+      "loss": 0.1861,
+      "step": 4440
+    },
+    {
+      "epoch": 3.3428678961234475,
+      "grad_norm": 0.22588451206684113,
+      "learning_rate": 3.344790824274984e-05,
+      "loss": 0.1703,
+      "step": 4441
+    },
+    {
+      "epoch": 3.3436206247647724,
+      "grad_norm": 0.26058393716812134,
+      "learning_rate": 3.344073199284729e-05,
+      "loss": 0.321,
+      "step": 4442
+    },
+    {
+      "epoch": 3.3443733534060973,
+      "grad_norm": 0.2518130838871002,
+      "learning_rate": 3.343355509327522e-05,
+      "loss": 0.2341,
+      "step": 4443
+    },
+    {
+      "epoch": 3.345126082047422,
+      "grad_norm": 0.2144014686346054,
+      "learning_rate": 3.342637754481849e-05,
+      "loss": 0.224,
+      "step": 4444
+    },
+    {
+      "epoch": 3.3458788106887467,
+      "grad_norm": 0.2678791284561157,
+      "learning_rate": 3.341919934826201e-05,
+      "loss": 0.279,
+      "step": 4445
+    },
+    {
+      "epoch": 3.3466315393300716,
+      "grad_norm": 0.23319803178310394,
+      "learning_rate": 3.34120205043908e-05,
+      "loss": 0.2072,
+      "step": 4446
+    },
+    {
+      "epoch": 3.347384267971396,
+      "grad_norm": 0.23095160722732544,
+      "learning_rate": 3.3404841013989916e-05,
+      "loss": 0.2279,
+      "step": 4447
+    },
+    {
+      "epoch": 3.348136996612721,
+      "grad_norm": 0.24624522030353546,
+      "learning_rate": 3.339766087784449e-05,
+      "loss": 0.3295,
+      "step": 4448
+    },
+    {
+      "epoch": 3.348889725254046,
+      "grad_norm": 0.23273321986198425,
+      "learning_rate": 3.339048009673974e-05,
+      "loss": 0.2633,
+      "step": 4449
+    },
+    {
+      "epoch": 3.349642453895371,
+      "grad_norm": 0.1916068196296692,
+      "learning_rate": 3.3383298671460944e-05,
+      "loss": 0.159,
+      "step": 4450
+    },
+    {
+      "epoch": 3.3503951825366958,
+      "grad_norm": 0.2428368777036667,
+      "learning_rate": 3.337611660279345e-05,
+      "loss": 0.2327,
+      "step": 4451
+    },
+    {
+      "epoch": 3.3511479111780202,
+      "grad_norm": 0.22111964225769043,
+      "learning_rate": 3.3368933891522684e-05,
+      "loss": 0.2401,
+      "step": 4452
+    },
+    {
+      "epoch": 3.351900639819345,
+      "grad_norm": 0.25172701478004456,
+      "learning_rate": 3.3361750538434124e-05,
+      "loss": 0.2692,
+      "step": 4453
+    },
+    {
+      "epoch": 3.35265336846067,
+      "grad_norm": 0.23564256727695465,
+      "learning_rate": 3.335456654431334e-05,
+      "loss": 0.2789,
+      "step": 4454
+    },
+    {
+      "epoch": 3.3534060971019946,
+      "grad_norm": 0.24926498532295227,
+      "learning_rate": 3.334738190994596e-05,
+      "loss": 0.3687,
+      "step": 4455
+    },
+    {
+      "epoch": 3.3541588257433195,
+      "grad_norm": 0.2487155646085739,
+      "learning_rate": 3.334019663611769e-05,
+      "loss": 0.3435,
+      "step": 4456
+    },
+    {
+      "epoch": 3.3549115543846444,
+      "grad_norm": 0.1875758171081543,
+      "learning_rate": 3.333301072361429e-05,
+      "loss": 0.2372,
+      "step": 4457
+    },
+    {
+      "epoch": 3.3556642830259693,
+      "grad_norm": 0.19570639729499817,
+      "learning_rate": 3.3325824173221623e-05,
+      "loss": 0.226,
+      "step": 4458
+    },
+    {
+      "epoch": 3.356417011667294,
+      "grad_norm": 0.262664258480072,
+      "learning_rate": 3.331863698572557e-05,
+      "loss": 0.2692,
+      "step": 4459
+    },
+    {
+      "epoch": 3.3571697403086187,
+      "grad_norm": 0.22128160297870636,
+      "learning_rate": 3.3311449161912125e-05,
+      "loss": 0.27,
+      "step": 4460
+    },
+    {
+      "epoch": 3.3579224689499436,
+      "grad_norm": 0.2336128205060959,
+      "learning_rate": 3.330426070256734e-05,
+      "loss": 0.2749,
+      "step": 4461
+    },
+    {
+      "epoch": 3.3586751975912685,
+      "grad_norm": 0.24978044629096985,
+      "learning_rate": 3.329707160847732e-05,
+      "loss": 0.3266,
+      "step": 4462
+    },
+    {
+      "epoch": 3.359427926232593,
+      "grad_norm": 0.2333175539970398,
+      "learning_rate": 3.328988188042828e-05,
+      "loss": 0.2902,
+      "step": 4463
+    },
+    {
+      "epoch": 3.360180654873918,
+      "grad_norm": 0.22683875262737274,
+      "learning_rate": 3.328269151920646e-05,
+      "loss": 0.2445,
+      "step": 4464
+    },
+    {
+      "epoch": 3.360933383515243,
+      "grad_norm": 0.22933819890022278,
+      "learning_rate": 3.3275500525598184e-05,
+      "loss": 0.2319,
+      "step": 4465
+    },
+    {
+      "epoch": 3.3616861121565673,
+      "grad_norm": 0.22729063034057617,
+      "learning_rate": 3.326830890038985e-05,
+      "loss": 0.2163,
+      "step": 4466
+    },
+    {
+      "epoch": 3.3624388407978922,
+      "grad_norm": 0.273760050535202,
+      "learning_rate": 3.326111664436794e-05,
+      "loss": 0.2261,
+      "step": 4467
+    },
+    {
+      "epoch": 3.363191569439217,
+      "grad_norm": 0.250075101852417,
+      "learning_rate": 3.3253923758318965e-05,
+      "loss": 0.2904,
+      "step": 4468
+    },
+    {
+      "epoch": 3.363944298080542,
+      "grad_norm": 0.24466657638549805,
+      "learning_rate": 3.3246730243029545e-05,
+      "loss": 0.2075,
+      "step": 4469
+    },
+    {
+      "epoch": 3.364697026721867,
+      "grad_norm": 0.24638257920742035,
+      "learning_rate": 3.323953609928635e-05,
+      "loss": 0.2054,
+      "step": 4470
+    },
+    {
+      "epoch": 3.3654497553631915,
+      "grad_norm": 0.22150875627994537,
+      "learning_rate": 3.3232341327876095e-05,
+      "loss": 0.3018,
+      "step": 4471
+    },
+    {
+      "epoch": 3.3662024840045164,
+      "grad_norm": 0.3146562874317169,
+      "learning_rate": 3.322514592958563e-05,
+      "loss": 0.2037,
+      "step": 4472
+    },
+    {
+      "epoch": 3.3669552126458413,
+      "grad_norm": 0.21975335478782654,
+      "learning_rate": 3.321794990520181e-05,
+      "loss": 0.2497,
+      "step": 4473
+    },
+    {
+      "epoch": 3.3677079412871658,
+      "grad_norm": 0.29439878463745117,
+      "learning_rate": 3.3210753255511583e-05,
+      "loss": 0.2902,
+      "step": 4474
+    },
+    {
+      "epoch": 3.3684606699284907,
+      "grad_norm": 0.2928523123264313,
+      "learning_rate": 3.3203555981301966e-05,
+      "loss": 0.2754,
+      "step": 4475
+    },
+    {
+      "epoch": 3.3692133985698156,
+      "grad_norm": 0.23594313859939575,
+      "learning_rate": 3.319635808336005e-05,
+      "loss": 0.2368,
+      "step": 4476
+    },
+    {
+      "epoch": 3.3699661272111405,
+      "grad_norm": 0.24637825787067413,
+      "learning_rate": 3.3189159562472966e-05,
+      "loss": 0.2261,
+      "step": 4477
+    },
+    {
+      "epoch": 3.370718855852465,
+      "grad_norm": 0.2650476098060608,
+      "learning_rate": 3.3181960419427946e-05,
+      "loss": 0.2833,
+      "step": 4478
+    },
+    {
+      "epoch": 3.37147158449379,
+      "grad_norm": 0.22983939945697784,
+      "learning_rate": 3.317476065501229e-05,
+      "loss": 0.2395,
+      "step": 4479
+    },
+    {
+      "epoch": 3.372224313135115,
+      "grad_norm": 0.23335875570774078,
+      "learning_rate": 3.3167560270013336e-05,
+      "loss": 0.2588,
+      "step": 4480
+    },
+    {
+      "epoch": 3.3729770417764398,
+      "grad_norm": 0.2323383390903473,
+      "learning_rate": 3.316035926521851e-05,
+      "loss": 0.2433,
+      "step": 4481
+    },
+    {
+      "epoch": 3.3737297704177642,
+      "grad_norm": 0.22165131568908691,
+      "learning_rate": 3.3153157641415295e-05,
+      "loss": 0.2691,
+      "step": 4482
+    },
+    {
+      "epoch": 3.374482499059089,
+      "grad_norm": 0.2589486837387085,
+      "learning_rate": 3.314595539939127e-05,
+      "loss": 0.2531,
+      "step": 4483
+    },
+    {
+      "epoch": 3.375235227700414,
+      "grad_norm": 0.25686752796173096,
+      "learning_rate": 3.3138752539934055e-05,
+      "loss": 0.2432,
+      "step": 4484
+    },
+    {
+      "epoch": 3.375987956341739,
+      "grad_norm": 0.2467522770166397,
+      "learning_rate": 3.313154906383133e-05,
+      "loss": 0.2307,
+      "step": 4485
+    },
+    {
+      "epoch": 3.3767406849830635,
+      "grad_norm": 0.2655801773071289,
+      "learning_rate": 3.3124344971870876e-05,
+      "loss": 0.1846,
+      "step": 4486
+    },
+    {
+      "epoch": 3.3774934136243884,
+      "grad_norm": 0.21505588293075562,
+      "learning_rate": 3.3117140264840516e-05,
+      "loss": 0.2459,
+      "step": 4487
+    },
+    {
+      "epoch": 3.3782461422657133,
+      "grad_norm": 0.24954243004322052,
+      "learning_rate": 3.3109934943528126e-05,
+      "loss": 0.3858,
+      "step": 4488
+    },
+    {
+      "epoch": 3.378998870907038,
+      "grad_norm": 0.2749645709991455,
+      "learning_rate": 3.310272900872169e-05,
+      "loss": 0.2399,
+      "step": 4489
+    },
+    {
+      "epoch": 3.3797515995483627,
+      "grad_norm": 0.2680414915084839,
+      "learning_rate": 3.309552246120924e-05,
+      "loss": 0.1836,
+      "step": 4490
+    },
+    {
+      "epoch": 3.3805043281896876,
+      "grad_norm": 0.27743154764175415,
+      "learning_rate": 3.308831530177885e-05,
+      "loss": 0.1943,
+      "step": 4491
+    },
+    {
+      "epoch": 3.3812570568310125,
+      "grad_norm": 0.31435826420783997,
+      "learning_rate": 3.308110753121872e-05,
+      "loss": 0.2565,
+      "step": 4492
+    },
+    {
+      "epoch": 3.382009785472337,
+      "grad_norm": 0.2583412826061249,
+      "learning_rate": 3.3073899150317055e-05,
+      "loss": 0.2361,
+      "step": 4493
+    },
+    {
+      "epoch": 3.382762514113662,
+      "grad_norm": 0.24619446694850922,
+      "learning_rate": 3.306669015986216e-05,
+      "loss": 0.1467,
+      "step": 4494
+    },
+    {
+      "epoch": 3.383515242754987,
+      "grad_norm": 0.26891234517097473,
+      "learning_rate": 3.305948056064239e-05,
+      "loss": 0.2939,
+      "step": 4495
+    },
+    {
+      "epoch": 3.3842679713963117,
+      "grad_norm": 0.290066123008728,
+      "learning_rate": 3.305227035344619e-05,
+      "loss": 0.2332,
+      "step": 4496
+    },
+    {
+      "epoch": 3.3850207000376367,
+      "grad_norm": 0.2809111177921295,
+      "learning_rate": 3.3045059539062055e-05,
+      "loss": 0.1837,
+      "step": 4497
+    },
+    {
+      "epoch": 3.385773428678961,
+      "grad_norm": 0.26370975375175476,
+      "learning_rate": 3.3037848118278544e-05,
+      "loss": 0.1983,
+      "step": 4498
+    },
+    {
+      "epoch": 3.386526157320286,
+      "grad_norm": 0.21314498782157898,
+      "learning_rate": 3.3030636091884285e-05,
+      "loss": 0.2023,
+      "step": 4499
+    },
+    {
+      "epoch": 3.387278885961611,
+      "grad_norm": 0.2557527422904968,
+      "learning_rate": 3.3023423460667985e-05,
+      "loss": 0.2834,
+      "step": 4500
+    },
+    {
+      "epoch": 3.3880316146029354,
+      "grad_norm": 0.2971329092979431,
+      "learning_rate": 3.301621022541839e-05,
+      "loss": 0.2777,
+      "step": 4501
+    },
+    {
+      "epoch": 3.3887843432442604,
+      "grad_norm": 0.32346343994140625,
+      "learning_rate": 3.300899638692434e-05,
+      "loss": 0.2522,
+      "step": 4502
+    },
+    {
+      "epoch": 3.3895370718855853,
+      "grad_norm": 0.278656005859375,
+      "learning_rate": 3.300178194597474e-05,
+      "loss": 0.2266,
+      "step": 4503
+    },
+    {
+      "epoch": 3.39028980052691,
+      "grad_norm": 0.20448268949985504,
+      "learning_rate": 3.299456690335853e-05,
+      "loss": 0.2031,
+      "step": 4504
+    },
+    {
+      "epoch": 3.3910425291682347,
+      "grad_norm": 0.219923734664917,
+      "learning_rate": 3.298735125986475e-05,
+      "loss": 0.3105,
+      "step": 4505
+    },
+    {
+      "epoch": 3.3917952578095596,
+      "grad_norm": 0.22961227595806122,
+      "learning_rate": 3.2980135016282476e-05,
+      "loss": 0.191,
+      "step": 4506
+    },
+    {
+      "epoch": 3.3925479864508845,
+      "grad_norm": 0.2323451191186905,
+      "learning_rate": 3.297291817340088e-05,
+      "loss": 0.261,
+      "step": 4507
+    },
+    {
+      "epoch": 3.3933007150922094,
+      "grad_norm": 0.23045477271080017,
+      "learning_rate": 3.296570073200918e-05,
+      "loss": 0.2141,
+      "step": 4508
+    },
+    {
+      "epoch": 3.394053443733534,
+      "grad_norm": 0.23621521890163422,
+      "learning_rate": 3.295848269289667e-05,
+      "loss": 0.2711,
+      "step": 4509
+    },
+    {
+      "epoch": 3.394806172374859,
+      "grad_norm": 0.24012619256973267,
+      "learning_rate": 3.2951264056852695e-05,
+      "loss": 0.3336,
+      "step": 4510
+    },
+    {
+      "epoch": 3.3955589010161837,
+      "grad_norm": 0.2143302708864212,
+      "learning_rate": 3.294404482466668e-05,
+      "loss": 0.189,
+      "step": 4511
+    },
+    {
+      "epoch": 3.3963116296575087,
+      "grad_norm": 0.22501996159553528,
+      "learning_rate": 3.2936824997128106e-05,
+      "loss": 0.1979,
+      "step": 4512
+    },
+    {
+      "epoch": 3.397064358298833,
+      "grad_norm": 0.24541762471199036,
+      "learning_rate": 3.292960457502653e-05,
+      "loss": 0.1649,
+      "step": 4513
+    },
+    {
+      "epoch": 3.397817086940158,
+      "grad_norm": 0.28883126378059387,
+      "learning_rate": 3.2922383559151556e-05,
+      "loss": 0.2784,
+      "step": 4514
+    },
+    {
+      "epoch": 3.398569815581483,
+      "grad_norm": 0.2225121259689331,
+      "learning_rate": 3.291516195029288e-05,
+      "loss": 0.2552,
+      "step": 4515
+    },
+    {
+      "epoch": 3.399322544222808,
+      "grad_norm": 0.24010956287384033,
+      "learning_rate": 3.2907939749240226e-05,
+      "loss": 0.1447,
+      "step": 4516
+    },
+    {
+      "epoch": 3.4000752728641324,
+      "grad_norm": 0.23096373677253723,
+      "learning_rate": 3.29007169567834e-05,
+      "loss": 0.2733,
+      "step": 4517
+    },
+    {
+      "epoch": 3.4008280015054573,
+      "grad_norm": 0.2530667781829834,
+      "learning_rate": 3.2893493573712295e-05,
+      "loss": 0.2504,
+      "step": 4518
+    },
+    {
+      "epoch": 3.401580730146782,
+      "grad_norm": 0.25462326407432556,
+      "learning_rate": 3.288626960081685e-05,
+      "loss": 0.2769,
+      "step": 4519
+    },
+    {
+      "epoch": 3.4023334587881067,
+      "grad_norm": 0.20936715602874756,
+      "learning_rate": 3.287904503888704e-05,
+      "loss": 0.1404,
+      "step": 4520
+    },
+    {
+      "epoch": 3.4030861874294316,
+      "grad_norm": 0.29498201608657837,
+      "learning_rate": 3.287181988871298e-05,
+      "loss": 0.3288,
+      "step": 4521
+    },
+    {
+      "epoch": 3.4038389160707565,
+      "grad_norm": 0.23187844455242157,
+      "learning_rate": 3.286459415108475e-05,
+      "loss": 0.3338,
+      "step": 4522
+    },
+    {
+      "epoch": 3.4045916447120814,
+      "grad_norm": 0.20951801538467407,
+      "learning_rate": 3.2857367826792585e-05,
+      "loss": 0.2348,
+      "step": 4523
+    },
+    {
+      "epoch": 3.4053443733534063,
+      "grad_norm": 0.2456183284521103,
+      "learning_rate": 3.285014091662673e-05,
+      "loss": 0.2759,
+      "step": 4524
+    },
+    {
+      "epoch": 3.406097101994731,
+      "grad_norm": 0.22665651142597198,
+      "learning_rate": 3.28429134213775e-05,
+      "loss": 0.2395,
+      "step": 4525
+    },
+    {
+      "epoch": 3.4068498306360557,
+      "grad_norm": 0.2441994547843933,
+      "learning_rate": 3.2835685341835284e-05,
+      "loss": 0.2161,
+      "step": 4526
+    },
+    {
+      "epoch": 3.4076025592773806,
+      "grad_norm": 0.25650861859321594,
+      "learning_rate": 3.282845667879055e-05,
+      "loss": 0.2464,
+      "step": 4527
+    },
+    {
+      "epoch": 3.408355287918705,
+      "grad_norm": 0.25130757689476013,
+      "learning_rate": 3.282122743303381e-05,
+      "loss": 0.2718,
+      "step": 4528
+    },
+    {
+      "epoch": 3.40910801656003,
+      "grad_norm": 0.20662502944469452,
+      "learning_rate": 3.281399760535563e-05,
+      "loss": 0.2072,
+      "step": 4529
+    },
+    {
+      "epoch": 3.409860745201355,
+      "grad_norm": 0.2100837081670761,
+      "learning_rate": 3.280676719654666e-05,
+      "loss": 0.1722,
+      "step": 4530
+    },
+    {
+      "epoch": 3.41061347384268,
+      "grad_norm": 0.2139636129140854,
+      "learning_rate": 3.279953620739762e-05,
+      "loss": 0.2098,
+      "step": 4531
+    },
+    {
+      "epoch": 3.4113662024840044,
+      "grad_norm": 0.24739238619804382,
+      "learning_rate": 3.279230463869926e-05,
+      "loss": 0.2901,
+      "step": 4532
+    },
+    {
+      "epoch": 3.4121189311253293,
+      "grad_norm": 0.23985832929611206,
+      "learning_rate": 3.278507249124241e-05,
+      "loss": 0.183,
+      "step": 4533
+    },
+    {
+      "epoch": 3.412871659766654,
+      "grad_norm": 0.2145209163427353,
+      "learning_rate": 3.2777839765817995e-05,
+      "loss": 0.1341,
+      "step": 4534
+    },
+    {
+      "epoch": 3.413624388407979,
+      "grad_norm": 0.2018221765756607,
+      "learning_rate": 3.277060646321696e-05,
+      "loss": 0.3519,
+      "step": 4535
+    },
+    {
+      "epoch": 3.4143771170493036,
+      "grad_norm": 0.2215980440378189,
+      "learning_rate": 3.276337258423032e-05,
+      "loss": 0.1984,
+      "step": 4536
+    },
+    {
+      "epoch": 3.4151298456906285,
+      "grad_norm": 0.24166131019592285,
+      "learning_rate": 3.275613812964916e-05,
+      "loss": 0.236,
+      "step": 4537
+    },
+    {
+      "epoch": 3.4158825743319534,
+      "grad_norm": 0.2247515320777893,
+      "learning_rate": 3.274890310026465e-05,
+      "loss": 0.312,
+      "step": 4538
+    },
+    {
+      "epoch": 3.416635302973278,
+      "grad_norm": 0.22589103877544403,
+      "learning_rate": 3.274166749686798e-05,
+      "loss": 0.1964,
+      "step": 4539
+    },
+    {
+      "epoch": 3.417388031614603,
+      "grad_norm": 0.21556368470191956,
+      "learning_rate": 3.273443132025044e-05,
+      "loss": 0.1809,
+      "step": 4540
+    },
+    {
+      "epoch": 3.4181407602559277,
+      "grad_norm": 0.24636949598789215,
+      "learning_rate": 3.272719457120335e-05,
+      "loss": 0.3328,
+      "step": 4541
+    },
+    {
+      "epoch": 3.4188934888972526,
+      "grad_norm": 0.2552788555622101,
+      "learning_rate": 3.2719957250518127e-05,
+      "loss": 0.2345,
+      "step": 4542
+    },
+    {
+      "epoch": 3.4196462175385776,
+      "grad_norm": 0.2573532164096832,
+      "learning_rate": 3.271271935898622e-05,
+      "loss": 0.2252,
+      "step": 4543
+    },
+    {
+      "epoch": 3.420398946179902,
+      "grad_norm": 0.23427307605743408,
+      "learning_rate": 3.270548089739918e-05,
+      "loss": 0.2122,
+      "step": 4544
+    },
+    {
+      "epoch": 3.421151674821227,
+      "grad_norm": 0.21819184720516205,
+      "learning_rate": 3.269824186654855e-05,
+      "loss": 0.2405,
+      "step": 4545
+    },
+    {
+      "epoch": 3.421904403462552,
+      "grad_norm": 0.2360946089029312,
+      "learning_rate": 3.269100226722602e-05,
+      "loss": 0.2319,
+      "step": 4546
+    },
+    {
+      "epoch": 3.4226571321038763,
+      "grad_norm": 0.2561880052089691,
+      "learning_rate": 3.268376210022329e-05,
+      "loss": 0.3058,
+      "step": 4547
+    },
+    {
+      "epoch": 3.4234098607452013,
+      "grad_norm": 0.23402509093284607,
+      "learning_rate": 3.267652136633213e-05,
+      "loss": 0.1918,
+      "step": 4548
+    },
+    {
+      "epoch": 3.424162589386526,
+      "grad_norm": 0.26239320635795593,
+      "learning_rate": 3.2669280066344365e-05,
+      "loss": 0.3518,
+      "step": 4549
+    },
+    {
+      "epoch": 3.424915318027851,
+      "grad_norm": 0.29247212409973145,
+      "learning_rate": 3.2662038201051914e-05,
+      "loss": 0.2824,
+      "step": 4550
+    },
+    {
+      "epoch": 3.4256680466691756,
+      "grad_norm": 0.1956847608089447,
+      "learning_rate": 3.265479577124672e-05,
+      "loss": 0.2155,
+      "step": 4551
+    },
+    {
+      "epoch": 3.4264207753105005,
+      "grad_norm": 0.30370083451271057,
+      "learning_rate": 3.2647552777720815e-05,
+      "loss": 0.2601,
+      "step": 4552
+    },
+    {
+      "epoch": 3.4271735039518254,
+      "grad_norm": 0.33797115087509155,
+      "learning_rate": 3.2640309221266274e-05,
+      "loss": 0.1366,
+      "step": 4553
+    },
+    {
+      "epoch": 3.4279262325931503,
+      "grad_norm": 0.24863211810588837,
+      "learning_rate": 3.2633065102675256e-05,
+      "loss": 0.2706,
+      "step": 4554
+    },
+    {
+      "epoch": 3.428678961234475,
+      "grad_norm": 0.24394676089286804,
+      "learning_rate": 3.262582042273994e-05,
+      "loss": 0.214,
+      "step": 4555
+    },
+    {
+      "epoch": 3.4294316898757997,
+      "grad_norm": 0.23905305564403534,
+      "learning_rate": 3.2618575182252615e-05,
+      "loss": 0.2043,
+      "step": 4556
+    },
+    {
+      "epoch": 3.4301844185171246,
+      "grad_norm": 0.2447129338979721,
+      "learning_rate": 3.261132938200561e-05,
+      "loss": 0.2492,
+      "step": 4557
+    },
+    {
+      "epoch": 3.4309371471584496,
+      "grad_norm": 0.2717069089412689,
+      "learning_rate": 3.26040830227913e-05,
+      "loss": 0.2385,
+      "step": 4558
+    },
+    {
+      "epoch": 3.431689875799774,
+      "grad_norm": 0.27040770649909973,
+      "learning_rate": 3.259683610540215e-05,
+      "loss": 0.2282,
+      "step": 4559
+    },
+    {
+      "epoch": 3.432442604441099,
+      "grad_norm": 0.2632172703742981,
+      "learning_rate": 3.258958863063066e-05,
+      "loss": 0.3007,
+      "step": 4560
+    },
+    {
+      "epoch": 3.433195333082424,
+      "grad_norm": 0.22317521274089813,
+      "learning_rate": 3.258234059926941e-05,
+      "loss": 0.2515,
+      "step": 4561
+    },
+    {
+      "epoch": 3.433948061723749,
+      "grad_norm": 0.23810110986232758,
+      "learning_rate": 3.2575092012111044e-05,
+      "loss": 0.2002,
+      "step": 4562
+    },
+    {
+      "epoch": 3.4347007903650733,
+      "grad_norm": 0.23417848348617554,
+      "learning_rate": 3.2567842869948236e-05,
+      "loss": 0.2598,
+      "step": 4563
+    },
+    {
+      "epoch": 3.435453519006398,
+      "grad_norm": 0.263209730386734,
+      "learning_rate": 3.256059317357375e-05,
+      "loss": 0.2864,
+      "step": 4564
+    },
+    {
+      "epoch": 3.436206247647723,
+      "grad_norm": 0.20537708699703217,
+      "learning_rate": 3.255334292378041e-05,
+      "loss": 0.2387,
+      "step": 4565
+    },
+    {
+      "epoch": 3.4369589762890476,
+      "grad_norm": 0.22285690903663635,
+      "learning_rate": 3.254609212136108e-05,
+      "loss": 0.1989,
+      "step": 4566
+    },
+    {
+      "epoch": 3.4377117049303725,
+      "grad_norm": 0.22890333831310272,
+      "learning_rate": 3.2538840767108704e-05,
+      "loss": 0.2117,
+      "step": 4567
+    },
+    {
+      "epoch": 3.4384644335716974,
+      "grad_norm": 0.26774492859840393,
+      "learning_rate": 3.2531588861816273e-05,
+      "loss": 0.2109,
+      "step": 4568
+    },
+    {
+      "epoch": 3.4392171622130223,
+      "grad_norm": 0.21564172208309174,
+      "learning_rate": 3.2524336406276854e-05,
+      "loss": 0.2777,
+      "step": 4569
+    },
+    {
+      "epoch": 3.4399698908543472,
+      "grad_norm": 0.22652292251586914,
+      "learning_rate": 3.251708340128357e-05,
+      "loss": 0.1847,
+      "step": 4570
+    },
+    {
+      "epoch": 3.4407226194956717,
+      "grad_norm": 0.22922997176647186,
+      "learning_rate": 3.250982984762956e-05,
+      "loss": 0.2091,
+      "step": 4571
+    },
+    {
+      "epoch": 3.4414753481369966,
+      "grad_norm": 0.2538823187351227,
+      "learning_rate": 3.250257574610812e-05,
+      "loss": 0.258,
+      "step": 4572
+    },
+    {
+      "epoch": 3.4422280767783215,
+      "grad_norm": 0.20500899851322174,
+      "learning_rate": 3.2495321097512504e-05,
+      "loss": 0.1804,
+      "step": 4573
+    },
+    {
+      "epoch": 3.442980805419646,
+      "grad_norm": 0.2057073712348938,
+      "learning_rate": 3.248806590263608e-05,
+      "loss": 0.3088,
+      "step": 4574
+    },
+    {
+      "epoch": 3.443733534060971,
+      "grad_norm": 0.23014681041240692,
+      "learning_rate": 3.248081016227227e-05,
+      "loss": 0.2925,
+      "step": 4575
+    },
+    {
+      "epoch": 3.444486262702296,
+      "grad_norm": 0.2304486781358719,
+      "learning_rate": 3.247355387721456e-05,
+      "loss": 0.262,
+      "step": 4576
+    },
+    {
+      "epoch": 3.4452389913436208,
+      "grad_norm": 0.19499722123146057,
+      "learning_rate": 3.246629704825646e-05,
+      "loss": 0.3103,
+      "step": 4577
+    },
+    {
+      "epoch": 3.4459917199849452,
+      "grad_norm": 0.25188708305358887,
+      "learning_rate": 3.24590396761916e-05,
+      "loss": 0.2586,
+      "step": 4578
+    },
+    {
+      "epoch": 3.44674444862627,
+      "grad_norm": 0.21051320433616638,
+      "learning_rate": 3.245178176181361e-05,
+      "loss": 0.2576,
+      "step": 4579
+    },
+    {
+      "epoch": 3.447497177267595,
+      "grad_norm": 0.218415305018425,
+      "learning_rate": 3.2444523305916195e-05,
+      "loss": 0.3335,
+      "step": 4580
+    },
+    {
+      "epoch": 3.44824990590892,
+      "grad_norm": 0.2095789611339569,
+      "learning_rate": 3.243726430929316e-05,
+      "loss": 0.2289,
+      "step": 4581
+    },
+    {
+      "epoch": 3.4490026345502445,
+      "grad_norm": 0.20837633311748505,
+      "learning_rate": 3.243000477273833e-05,
+      "loss": 0.2848,
+      "step": 4582
+    },
+    {
+      "epoch": 3.4497553631915694,
+      "grad_norm": 0.197526752948761,
+      "learning_rate": 3.2422744697045575e-05,
+      "loss": 0.3192,
+      "step": 4583
+    },
+    {
+      "epoch": 3.4505080918328943,
+      "grad_norm": 0.22360116243362427,
+      "learning_rate": 3.241548408300886e-05,
+      "loss": 0.2248,
+      "step": 4584
+    },
+    {
+      "epoch": 3.4512608204742192,
+      "grad_norm": 0.1976037472486496,
+      "learning_rate": 3.240822293142221e-05,
+      "loss": 0.1867,
+      "step": 4585
+    },
+    {
+      "epoch": 3.4520135491155437,
+      "grad_norm": 0.20121712982654572,
+      "learning_rate": 3.2400961243079676e-05,
+      "loss": 0.1703,
+      "step": 4586
+    },
+    {
+      "epoch": 3.4527662777568686,
+      "grad_norm": 0.20382685959339142,
+      "learning_rate": 3.2393699018775385e-05,
+      "loss": 0.298,
+      "step": 4587
+    },
+    {
+      "epoch": 3.4535190063981935,
+      "grad_norm": 0.22023646533489227,
+      "learning_rate": 3.238643625930353e-05,
+      "loss": 0.2454,
+      "step": 4588
+    },
+    {
+      "epoch": 3.4542717350395185,
+      "grad_norm": 0.22840559482574463,
+      "learning_rate": 3.237917296545835e-05,
+      "loss": 0.2361,
+      "step": 4589
+    },
+    {
+      "epoch": 3.455024463680843,
+      "grad_norm": 0.20509395003318787,
+      "learning_rate": 3.237190913803415e-05,
+      "loss": 0.2185,
+      "step": 4590
+    },
+    {
+      "epoch": 3.455777192322168,
+      "grad_norm": 0.23574210703372955,
+      "learning_rate": 3.236464477782528e-05,
+      "loss": 0.234,
+      "step": 4591
+    },
+    {
+      "epoch": 3.4565299209634928,
+      "grad_norm": 0.21608391404151917,
+      "learning_rate": 3.235737988562619e-05,
+      "loss": 0.2824,
+      "step": 4592
+    },
+    {
+      "epoch": 3.4572826496048172,
+      "grad_norm": 0.21126538515090942,
+      "learning_rate": 3.2350114462231326e-05,
+      "loss": 0.2117,
+      "step": 4593
+    },
+    {
+      "epoch": 3.458035378246142,
+      "grad_norm": 0.23429341614246368,
+      "learning_rate": 3.234284850843525e-05,
+      "loss": 0.2692,
+      "step": 4594
+    },
+    {
+      "epoch": 3.458788106887467,
+      "grad_norm": 0.19278846681118011,
+      "learning_rate": 3.233558202503253e-05,
+      "loss": 0.2143,
+      "step": 4595
+    },
+    {
+      "epoch": 3.459540835528792,
+      "grad_norm": 0.2370903491973877,
+      "learning_rate": 3.232831501281783e-05,
+      "loss": 0.3198,
+      "step": 4596
+    },
+    {
+      "epoch": 3.460293564170117,
+      "grad_norm": 0.2628404498100281,
+      "learning_rate": 3.232104747258587e-05,
+      "loss": 0.2272,
+      "step": 4597
+    },
+    {
+      "epoch": 3.4610462928114414,
+      "grad_norm": 0.24138009548187256,
+      "learning_rate": 3.23137794051314e-05,
+      "loss": 0.3245,
+      "step": 4598
+    },
+    {
+      "epoch": 3.4617990214527663,
+      "grad_norm": 0.2539466321468353,
+      "learning_rate": 3.230651081124925e-05,
+      "loss": 0.3268,
+      "step": 4599
+    },
+    {
+      "epoch": 3.462551750094091,
+      "grad_norm": 0.20467647910118103,
+      "learning_rate": 3.2299241691734304e-05,
+      "loss": 0.3017,
+      "step": 4600
+    },
+    {
+      "epoch": 3.462551750094091,
+      "eval_loss": 0.2653123736381531,
+      "eval_runtime": 452.6204,
+      "eval_samples_per_second": 21.269,
+      "eval_steps_per_second": 0.665,
+      "step": 4600
+    },
+    {
+      "epoch": 3.4633044787354157,
+      "grad_norm": 0.32926279306411743,
+      "learning_rate": 3.2291972047381504e-05,
+      "loss": 0.2757,
+      "step": 4601
+    },
+    {
+      "epoch": 3.4640572073767406,
+      "grad_norm": 0.2006891667842865,
+      "learning_rate": 3.228470187898585e-05,
+      "loss": 0.2051,
+      "step": 4602
+    },
+    {
+      "epoch": 3.4648099360180655,
+      "grad_norm": 0.23129135370254517,
+      "learning_rate": 3.2277431187342385e-05,
+      "loss": 0.1027,
+      "step": 4603
+    },
+    {
+      "epoch": 3.4655626646593904,
+      "grad_norm": 0.2615605890750885,
+      "learning_rate": 3.227015997324623e-05,
+      "loss": 0.2721,
+      "step": 4604
+    },
+    {
+      "epoch": 3.466315393300715,
+      "grad_norm": 0.21139384806156158,
+      "learning_rate": 3.2262888237492554e-05,
+      "loss": 0.2546,
+      "step": 4605
+    },
+    {
+      "epoch": 3.46706812194204,
+      "grad_norm": 0.24159006774425507,
+      "learning_rate": 3.2255615980876585e-05,
+      "loss": 0.2173,
+      "step": 4606
+    },
+    {
+      "epoch": 3.4678208505833648,
+      "grad_norm": 0.215541809797287,
+      "learning_rate": 3.22483432041936e-05,
+      "loss": 0.2162,
+      "step": 4607
+    },
+    {
+      "epoch": 3.4685735792246897,
+      "grad_norm": 0.23948045074939728,
+      "learning_rate": 3.224106990823894e-05,
+      "loss": 0.2197,
+      "step": 4608
+    },
+    {
+      "epoch": 3.469326307866014,
+      "grad_norm": 0.20460976660251617,
+      "learning_rate": 3.2233796093807996e-05,
+      "loss": 0.214,
+      "step": 4609
+    },
+    {
+      "epoch": 3.470079036507339,
+      "grad_norm": 0.2200765609741211,
+      "learning_rate": 3.222652176169624e-05,
+      "loss": 0.3574,
+      "step": 4610
+    },
+    {
+      "epoch": 3.470831765148664,
+      "grad_norm": 0.25050991773605347,
+      "learning_rate": 3.221924691269916e-05,
+      "loss": 0.3304,
+      "step": 4611
+    },
+    {
+      "epoch": 3.471584493789989,
+      "grad_norm": 0.2427821308374405,
+      "learning_rate": 3.221197154761233e-05,
+      "loss": 0.2975,
+      "step": 4612
+    },
+    {
+      "epoch": 3.4723372224313134,
+      "grad_norm": 0.2215385138988495,
+      "learning_rate": 3.2204695667231386e-05,
+      "loss": 0.32,
+      "step": 4613
+    },
+    {
+      "epoch": 3.4730899510726383,
+      "grad_norm": 0.23463651537895203,
+      "learning_rate": 3.219741927235199e-05,
+      "loss": 0.2669,
+      "step": 4614
+    },
+    {
+      "epoch": 3.473842679713963,
+      "grad_norm": 0.2429659366607666,
+      "learning_rate": 3.219014236376989e-05,
+      "loss": 0.3494,
+      "step": 4615
+    },
+    {
+      "epoch": 3.474595408355288,
+      "grad_norm": 0.24286594986915588,
+      "learning_rate": 3.2182864942280875e-05,
+      "loss": 0.1811,
+      "step": 4616
+    },
+    {
+      "epoch": 3.4753481369966126,
+      "grad_norm": 0.23303590714931488,
+      "learning_rate": 3.217558700868079e-05,
+      "loss": 0.1595,
+      "step": 4617
+    },
+    {
+      "epoch": 3.4761008656379375,
+      "grad_norm": 0.20852313935756683,
+      "learning_rate": 3.2168308563765536e-05,
+      "loss": 0.2472,
+      "step": 4618
+    },
+    {
+      "epoch": 3.4768535942792624,
+      "grad_norm": 0.23222090303897858,
+      "learning_rate": 3.216102960833108e-05,
+      "loss": 0.2251,
+      "step": 4619
+    },
+    {
+      "epoch": 3.477606322920587,
+      "grad_norm": 0.21649880707263947,
+      "learning_rate": 3.215375014317343e-05,
+      "loss": 0.1691,
+      "step": 4620
+    },
+    {
+      "epoch": 3.478359051561912,
+      "grad_norm": 0.2192290872335434,
+      "learning_rate": 3.214647016908867e-05,
+      "loss": 0.1808,
+      "step": 4621
+    },
+    {
+      "epoch": 3.4791117802032367,
+      "grad_norm": 0.23734527826309204,
+      "learning_rate": 3.2139189686872904e-05,
+      "loss": 0.2613,
+      "step": 4622
+    },
+    {
+      "epoch": 3.4798645088445617,
+      "grad_norm": 0.2700372040271759,
+      "learning_rate": 3.213190869732233e-05,
+      "loss": 0.2642,
+      "step": 4623
+    },
+    {
+      "epoch": 3.4806172374858866,
+      "grad_norm": 0.20992548763751984,
+      "learning_rate": 3.212462720123319e-05,
+      "loss": 0.2958,
+      "step": 4624
+    },
+    {
+      "epoch": 3.481369966127211,
+      "grad_norm": 0.21860797703266144,
+      "learning_rate": 3.211734519940177e-05,
+      "loss": 0.2335,
+      "step": 4625
+    },
+    {
+      "epoch": 3.482122694768536,
+      "grad_norm": 0.2343943566083908,
+      "learning_rate": 3.2110062692624423e-05,
+      "loss": 0.3405,
+      "step": 4626
+    },
+    {
+      "epoch": 3.482875423409861,
+      "grad_norm": 0.2845248579978943,
+      "learning_rate": 3.210277968169756e-05,
+      "loss": 0.3277,
+      "step": 4627
+    },
+    {
+      "epoch": 3.4836281520511854,
+      "grad_norm": 0.2662038505077362,
+      "learning_rate": 3.2095496167417614e-05,
+      "loss": 0.2367,
+      "step": 4628
+    },
+    {
+      "epoch": 3.4843808806925103,
+      "grad_norm": 0.23150581121444702,
+      "learning_rate": 3.2088212150581135e-05,
+      "loss": 0.2804,
+      "step": 4629
+    },
+    {
+      "epoch": 3.485133609333835,
+      "grad_norm": 0.21941180527210236,
+      "learning_rate": 3.208092763198466e-05,
+      "loss": 0.332,
+      "step": 4630
+    },
+    {
+      "epoch": 3.48588633797516,
+      "grad_norm": 0.24565307796001434,
+      "learning_rate": 3.207364261242482e-05,
+      "loss": 0.2747,
+      "step": 4631
+    },
+    {
+      "epoch": 3.4866390666164846,
+      "grad_norm": 0.23220492899417877,
+      "learning_rate": 3.206635709269832e-05,
+      "loss": 0.3008,
+      "step": 4632
+    },
+    {
+      "epoch": 3.4873917952578095,
+      "grad_norm": 0.27279746532440186,
+      "learning_rate": 3.205907107360186e-05,
+      "loss": 0.1913,
+      "step": 4633
+    },
+    {
+      "epoch": 3.4881445238991344,
+      "grad_norm": 0.21762283146381378,
+      "learning_rate": 3.2051784555932246e-05,
+      "loss": 0.3123,
+      "step": 4634
+    },
+    {
+      "epoch": 3.4888972525404593,
+      "grad_norm": 0.24064242839813232,
+      "learning_rate": 3.204449754048631e-05,
+      "loss": 0.2894,
+      "step": 4635
+    },
+    {
+      "epoch": 3.489649981181784,
+      "grad_norm": 0.25169047713279724,
+      "learning_rate": 3.203721002806097e-05,
+      "loss": 0.1837,
+      "step": 4636
+    },
+    {
+      "epoch": 3.4904027098231087,
+      "grad_norm": 0.19061623513698578,
+      "learning_rate": 3.202992201945315e-05,
+      "loss": 0.1601,
+      "step": 4637
+    },
+    {
+      "epoch": 3.4911554384644337,
+      "grad_norm": 0.22432774305343628,
+      "learning_rate": 3.202263351545988e-05,
+      "loss": 0.149,
+      "step": 4638
+    },
+    {
+      "epoch": 3.491908167105758,
+      "grad_norm": 0.2371026575565338,
+      "learning_rate": 3.201534451687819e-05,
+      "loss": 0.1744,
+      "step": 4639
+    },
+    {
+      "epoch": 3.492660895747083,
+      "grad_norm": 0.21448653936386108,
+      "learning_rate": 3.200805502450523e-05,
+      "loss": 0.236,
+      "step": 4640
+    },
+    {
+      "epoch": 3.493413624388408,
+      "grad_norm": 0.2697927951812744,
+      "learning_rate": 3.2000765039138145e-05,
+      "loss": 0.2588,
+      "step": 4641
+    },
+    {
+      "epoch": 3.494166353029733,
+      "grad_norm": 0.19746632874011993,
+      "learning_rate": 3.199347456157416e-05,
+      "loss": 0.2655,
+      "step": 4642
+    },
+    {
+      "epoch": 3.494919081671058,
+      "grad_norm": 0.23481103777885437,
+      "learning_rate": 3.198618359261055e-05,
+      "loss": 0.2417,
+      "step": 4643
+    },
+    {
+      "epoch": 3.4956718103123823,
+      "grad_norm": 0.2625901699066162,
+      "learning_rate": 3.1978892133044655e-05,
+      "loss": 0.3318,
+      "step": 4644
+    },
+    {
+      "epoch": 3.496424538953707,
+      "grad_norm": 0.23280900716781616,
+      "learning_rate": 3.197160018367385e-05,
+      "loss": 0.2396,
+      "step": 4645
+    },
+    {
+      "epoch": 3.497177267595032,
+      "grad_norm": 0.23164743185043335,
+      "learning_rate": 3.196430774529557e-05,
+      "loss": 0.2029,
+      "step": 4646
+    },
+    {
+      "epoch": 3.4979299962363566,
+      "grad_norm": 0.22277292609214783,
+      "learning_rate": 3.1957014818707296e-05,
+      "loss": 0.2557,
+      "step": 4647
+    },
+    {
+      "epoch": 3.4986827248776815,
+      "grad_norm": 0.24459075927734375,
+      "learning_rate": 3.19497214047066e-05,
+      "loss": 0.3744,
+      "step": 4648
+    },
+    {
+      "epoch": 3.4994354535190064,
+      "grad_norm": 0.26677289605140686,
+      "learning_rate": 3.1942427504091053e-05,
+      "loss": 0.2444,
+      "step": 4649
+    },
+    {
+      "epoch": 3.5001881821603313,
+      "grad_norm": 0.20881375670433044,
+      "learning_rate": 3.1935133117658306e-05,
+      "loss": 0.2235,
+      "step": 4650
+    },
+    {
+      "epoch": 3.5009409108016563,
+      "grad_norm": 0.22307883203029633,
+      "learning_rate": 3.192783824620608e-05,
+      "loss": 0.222,
+      "step": 4651
+    },
+    {
+      "epoch": 3.5016936394429807,
+      "grad_norm": 0.21375197172164917,
+      "learning_rate": 3.1920542890532116e-05,
+      "loss": 0.179,
+      "step": 4652
+    },
+    {
+      "epoch": 3.5024463680843057,
+      "grad_norm": 0.24992479383945465,
+      "learning_rate": 3.191324705143424e-05,
+      "loss": 0.2397,
+      "step": 4653
+    },
+    {
+      "epoch": 3.5031990967256306,
+      "grad_norm": 0.22332361340522766,
+      "learning_rate": 3.1905950729710284e-05,
+      "loss": 0.2011,
+      "step": 4654
+    },
+    {
+      "epoch": 3.503951825366955,
+      "grad_norm": 0.25609153509140015,
+      "learning_rate": 3.189865392615818e-05,
+      "loss": 0.2426,
+      "step": 4655
+    },
+    {
+      "epoch": 3.50470455400828,
+      "grad_norm": 0.2541739046573639,
+      "learning_rate": 3.1891356641575915e-05,
+      "loss": 0.23,
+      "step": 4656
+    },
+    {
+      "epoch": 3.505457282649605,
+      "grad_norm": 0.20693154633045197,
+      "learning_rate": 3.1884058876761466e-05,
+      "loss": 0.1703,
+      "step": 4657
+    },
+    {
+      "epoch": 3.5062100112909294,
+      "grad_norm": 0.26591697335243225,
+      "learning_rate": 3.1876760632512946e-05,
+      "loss": 0.3001,
+      "step": 4658
+    },
+    {
+      "epoch": 3.5069627399322543,
+      "grad_norm": 0.2647171914577484,
+      "learning_rate": 3.186946190962846e-05,
+      "loss": 0.271,
+      "step": 4659
+    },
+    {
+      "epoch": 3.507715468573579,
+      "grad_norm": 0.24843217432498932,
+      "learning_rate": 3.186216270890619e-05,
+      "loss": 0.2917,
+      "step": 4660
+    },
+    {
+      "epoch": 3.508468197214904,
+      "grad_norm": 0.26441705226898193,
+      "learning_rate": 3.185486303114436e-05,
+      "loss": 0.2737,
+      "step": 4661
+    },
+    {
+      "epoch": 3.509220925856229,
+      "grad_norm": 0.21745693683624268,
+      "learning_rate": 3.1847562877141254e-05,
+      "loss": 0.2115,
+      "step": 4662
+    },
+    {
+      "epoch": 3.5099736544975535,
+      "grad_norm": 0.24864079058170319,
+      "learning_rate": 3.184026224769522e-05,
+      "loss": 0.1716,
+      "step": 4663
+    },
+    {
+      "epoch": 3.5107263831388784,
+      "grad_norm": 0.21622109413146973,
+      "learning_rate": 3.1832961143604616e-05,
+      "loss": 0.1883,
+      "step": 4664
+    },
+    {
+      "epoch": 3.5114791117802033,
+      "grad_norm": 0.21494987607002258,
+      "learning_rate": 3.182565956566791e-05,
+      "loss": 0.1851,
+      "step": 4665
+    },
+    {
+      "epoch": 3.512231840421528,
+      "grad_norm": 0.2578828036785126,
+      "learning_rate": 3.1818357514683574e-05,
+      "loss": 0.2517,
+      "step": 4666
+    },
+    {
+      "epoch": 3.5129845690628527,
+      "grad_norm": 0.2409975528717041,
+      "learning_rate": 3.181105499145015e-05,
+      "loss": 0.1944,
+      "step": 4667
+    },
+    {
+      "epoch": 3.5137372977041776,
+      "grad_norm": 0.23496273159980774,
+      "learning_rate": 3.1803751996766236e-05,
+      "loss": 0.3052,
+      "step": 4668
+    },
+    {
+      "epoch": 3.5144900263455026,
+      "grad_norm": 0.21771270036697388,
+      "learning_rate": 3.179644853143047e-05,
+      "loss": 0.2239,
+      "step": 4669
+    },
+    {
+      "epoch": 3.5152427549868275,
+      "grad_norm": 0.2589648365974426,
+      "learning_rate": 3.178914459624155e-05,
+      "loss": 0.2509,
+      "step": 4670
+    },
+    {
+      "epoch": 3.515995483628152,
+      "grad_norm": 0.21365737915039062,
+      "learning_rate": 3.178184019199824e-05,
+      "loss": 0.2247,
+      "step": 4671
+    },
+    {
+      "epoch": 3.516748212269477,
+      "grad_norm": 0.2617795467376709,
+      "learning_rate": 3.177453531949932e-05,
+      "loss": 0.3052,
+      "step": 4672
+    },
+    {
+      "epoch": 3.517500940910802,
+      "grad_norm": 0.2238275706768036,
+      "learning_rate": 3.176722997954363e-05,
+      "loss": 0.145,
+      "step": 4673
+    },
+    {
+      "epoch": 3.5182536695521263,
+      "grad_norm": 0.2298695296049118,
+      "learning_rate": 3.17599241729301e-05,
+      "loss": 0.2266,
+      "step": 4674
+    },
+    {
+      "epoch": 3.519006398193451,
+      "grad_norm": 0.20343808829784393,
+      "learning_rate": 3.175261790045766e-05,
+      "loss": 0.2282,
+      "step": 4675
+    },
+    {
+      "epoch": 3.519759126834776,
+      "grad_norm": 0.27162376046180725,
+      "learning_rate": 3.1745311162925316e-05,
+      "loss": 0.2324,
+      "step": 4676
+    },
+    {
+      "epoch": 3.520511855476101,
+      "grad_norm": 0.24369500577449799,
+      "learning_rate": 3.173800396113214e-05,
+      "loss": 0.2384,
+      "step": 4677
+    },
+    {
+      "epoch": 3.521264584117426,
+      "grad_norm": 0.2633240520954132,
+      "learning_rate": 3.173069629587721e-05,
+      "loss": 0.2533,
+      "step": 4678
+    },
+    {
+      "epoch": 3.5220173127587504,
+      "grad_norm": 0.25865378975868225,
+      "learning_rate": 3.17233881679597e-05,
+      "loss": 0.2466,
+      "step": 4679
+    },
+    {
+      "epoch": 3.5227700414000753,
+      "grad_norm": 0.19811291992664337,
+      "learning_rate": 3.171607957817881e-05,
+      "loss": 0.2282,
+      "step": 4680
+    },
+    {
+      "epoch": 3.5235227700414002,
+      "grad_norm": 0.23018454015254974,
+      "learning_rate": 3.1708770527333794e-05,
+      "loss": 0.2354,
+      "step": 4681
+    },
+    {
+      "epoch": 3.5242754986827247,
+      "grad_norm": 0.2365209311246872,
+      "learning_rate": 3.1701461016223955e-05,
+      "loss": 0.2237,
+      "step": 4682
+    },
+    {
+      "epoch": 3.5250282273240496,
+      "grad_norm": 0.23074781894683838,
+      "learning_rate": 3.169415104564867e-05,
+      "loss": 0.2416,
+      "step": 4683
+    },
+    {
+      "epoch": 3.5257809559653746,
+      "grad_norm": 0.21103055775165558,
+      "learning_rate": 3.168684061640732e-05,
+      "loss": 0.2341,
+      "step": 4684
+    },
+    {
+      "epoch": 3.526533684606699,
+      "grad_norm": 0.27615639567375183,
+      "learning_rate": 3.167952972929938e-05,
+      "loss": 0.3318,
+      "step": 4685
+    },
+    {
+      "epoch": 3.527286413248024,
+      "grad_norm": 0.20370444655418396,
+      "learning_rate": 3.167221838512435e-05,
+      "loss": 0.2445,
+      "step": 4686
+    },
+    {
+      "epoch": 3.528039141889349,
+      "grad_norm": 0.25770625472068787,
+      "learning_rate": 3.1664906584681784e-05,
+      "loss": 0.323,
+      "step": 4687
+    },
+    {
+      "epoch": 3.528791870530674,
+      "grad_norm": 0.23526985943317413,
+      "learning_rate": 3.165759432877131e-05,
+      "loss": 0.2668,
+      "step": 4688
+    },
+    {
+      "epoch": 3.5295445991719987,
+      "grad_norm": 0.22668442130088806,
+      "learning_rate": 3.165028161819257e-05,
+      "loss": 0.2503,
+      "step": 4689
+    },
+    {
+      "epoch": 3.530297327813323,
+      "grad_norm": 0.22488263249397278,
+      "learning_rate": 3.164296845374526e-05,
+      "loss": 0.2511,
+      "step": 4690
+    },
+    {
+      "epoch": 3.531050056454648,
+      "grad_norm": 0.24676614999771118,
+      "learning_rate": 3.1635654836229164e-05,
+      "loss": 0.2209,
+      "step": 4691
+    },
+    {
+      "epoch": 3.531802785095973,
+      "grad_norm": 0.24630489945411682,
+      "learning_rate": 3.162834076644406e-05,
+      "loss": 0.2,
+      "step": 4692
+    },
+    {
+      "epoch": 3.5325555137372975,
+      "grad_norm": 0.23829898238182068,
+      "learning_rate": 3.162102624518982e-05,
+      "loss": 0.2901,
+      "step": 4693
+    },
+    {
+      "epoch": 3.5333082423786224,
+      "grad_norm": 0.22154086828231812,
+      "learning_rate": 3.1613711273266356e-05,
+      "loss": 0.3456,
+      "step": 4694
+    },
+    {
+      "epoch": 3.5340609710199473,
+      "grad_norm": 0.20844945311546326,
+      "learning_rate": 3.160639585147359e-05,
+      "loss": 0.3353,
+      "step": 4695
+    },
+    {
+      "epoch": 3.5348136996612722,
+      "grad_norm": 0.2364875227212906,
+      "learning_rate": 3.159907998061157e-05,
+      "loss": 0.2527,
+      "step": 4696
+    },
+    {
+      "epoch": 3.535566428302597,
+      "grad_norm": 0.22501994669437408,
+      "learning_rate": 3.1591763661480314e-05,
+      "loss": 0.307,
+      "step": 4697
+    },
+    {
+      "epoch": 3.5363191569439216,
+      "grad_norm": 0.2264687418937683,
+      "learning_rate": 3.1584446894879936e-05,
+      "loss": 0.3692,
+      "step": 4698
+    },
+    {
+      "epoch": 3.5370718855852465,
+      "grad_norm": 0.20822975039482117,
+      "learning_rate": 3.157712968161059e-05,
+      "loss": 0.2364,
+      "step": 4699
+    },
+    {
+      "epoch": 3.5378246142265715,
+      "grad_norm": 0.2214827537536621,
+      "learning_rate": 3.156981202247248e-05,
+      "loss": 0.1938,
+      "step": 4700
+    },
+    {
+      "epoch": 3.538577342867896,
+      "grad_norm": 0.22560828924179077,
+      "learning_rate": 3.156249391826583e-05,
+      "loss": 0.3072,
+      "step": 4701
+    },
+    {
+      "epoch": 3.539330071509221,
+      "grad_norm": 0.21689650416374207,
+      "learning_rate": 3.155517536979096e-05,
+      "loss": 0.2004,
+      "step": 4702
+    },
+    {
+      "epoch": 3.5400828001505458,
+      "grad_norm": 0.2610621452331543,
+      "learning_rate": 3.1547856377848214e-05,
+      "loss": 0.2405,
+      "step": 4703
+    },
+    {
+      "epoch": 3.5408355287918707,
+      "grad_norm": 0.2322438359260559,
+      "learning_rate": 3.154053694323797e-05,
+      "loss": 0.1925,
+      "step": 4704
+    },
+    {
+      "epoch": 3.5415882574331956,
+      "grad_norm": 0.21608270704746246,
+      "learning_rate": 3.153321706676069e-05,
+      "loss": 0.2493,
+      "step": 4705
+    },
+    {
+      "epoch": 3.54234098607452,
+      "grad_norm": 0.23441867530345917,
+      "learning_rate": 3.152589674921685e-05,
+      "loss": 0.3357,
+      "step": 4706
+    },
+    {
+      "epoch": 3.543093714715845,
+      "grad_norm": 0.23080876469612122,
+      "learning_rate": 3.1518575991406994e-05,
+      "loss": 0.3191,
+      "step": 4707
+    },
+    {
+      "epoch": 3.54384644335717,
+      "grad_norm": 0.1915387213230133,
+      "learning_rate": 3.151125479413171e-05,
+      "loss": 0.1739,
+      "step": 4708
+    },
+    {
+      "epoch": 3.5445991719984944,
+      "grad_norm": 0.23302657902240753,
+      "learning_rate": 3.1503933158191635e-05,
+      "loss": 0.2093,
+      "step": 4709
+    },
+    {
+      "epoch": 3.5453519006398193,
+      "grad_norm": 0.21911181509494781,
+      "learning_rate": 3.149661108438744e-05,
+      "loss": 0.2532,
+      "step": 4710
+    },
+    {
+      "epoch": 3.5461046292811442,
+      "grad_norm": 0.23243676126003265,
+      "learning_rate": 3.148928857351989e-05,
+      "loss": 0.1693,
+      "step": 4711
+    },
+    {
+      "epoch": 3.5468573579224687,
+      "grad_norm": 0.23154929280281067,
+      "learning_rate": 3.1481965626389706e-05,
+      "loss": 0.3268,
+      "step": 4712
+    },
+    {
+      "epoch": 3.5476100865637936,
+      "grad_norm": 0.21773704886436462,
+      "learning_rate": 3.147464224379777e-05,
+      "loss": 0.2765,
+      "step": 4713
+    },
+    {
+      "epoch": 3.5483628152051185,
+      "grad_norm": 0.25048738718032837,
+      "learning_rate": 3.146731842654493e-05,
+      "loss": 0.2868,
+      "step": 4714
+    },
+    {
+      "epoch": 3.5491155438464435,
+      "grad_norm": 0.22850368916988373,
+      "learning_rate": 3.1459994175432115e-05,
+      "loss": 0.1764,
+      "step": 4715
+    },
+    {
+      "epoch": 3.5498682724877684,
+      "grad_norm": 0.21296484768390656,
+      "learning_rate": 3.145266949126029e-05,
+      "loss": 0.2687,
+      "step": 4716
+    },
+    {
+      "epoch": 3.550621001129093,
+      "grad_norm": 0.21714502573013306,
+      "learning_rate": 3.144534437483046e-05,
+      "loss": 0.2764,
+      "step": 4717
+    },
+    {
+      "epoch": 3.5513737297704178,
+      "grad_norm": 0.26318982243537903,
+      "learning_rate": 3.143801882694371e-05,
+      "loss": 0.2166,
+      "step": 4718
+    },
+    {
+      "epoch": 3.5521264584117427,
+      "grad_norm": 0.20666277408599854,
+      "learning_rate": 3.1430692848401145e-05,
+      "loss": 0.1988,
+      "step": 4719
+    },
+    {
+      "epoch": 3.552879187053067,
+      "grad_norm": 0.2180001139640808,
+      "learning_rate": 3.142336644000392e-05,
+      "loss": 0.1808,
+      "step": 4720
+    },
+    {
+      "epoch": 3.553631915694392,
+      "grad_norm": 0.21964119374752045,
+      "learning_rate": 3.141603960255323e-05,
+      "loss": 0.2564,
+      "step": 4721
+    },
+    {
+      "epoch": 3.554384644335717,
+      "grad_norm": 0.22958554327487946,
+      "learning_rate": 3.140871233685034e-05,
+      "loss": 0.2251,
+      "step": 4722
+    },
+    {
+      "epoch": 3.555137372977042,
+      "grad_norm": 0.24831566214561462,
+      "learning_rate": 3.140138464369656e-05,
+      "loss": 0.3367,
+      "step": 4723
+    },
+    {
+      "epoch": 3.555890101618367,
+      "grad_norm": 0.20222161710262299,
+      "learning_rate": 3.13940565238932e-05,
+      "loss": 0.244,
+      "step": 4724
+    },
+    {
+      "epoch": 3.5566428302596913,
+      "grad_norm": 0.22380349040031433,
+      "learning_rate": 3.138672797824169e-05,
+      "loss": 0.2741,
+      "step": 4725
+    },
+    {
+      "epoch": 3.557395558901016,
+      "grad_norm": 0.25034862756729126,
+      "learning_rate": 3.137939900754344e-05,
+      "loss": 0.302,
+      "step": 4726
+    },
+    {
+      "epoch": 3.558148287542341,
+      "grad_norm": 0.24525268375873566,
+      "learning_rate": 3.137206961259995e-05,
+      "loss": 0.3782,
+      "step": 4727
+    },
+    {
+      "epoch": 3.5589010161836656,
+      "grad_norm": 0.23584416508674622,
+      "learning_rate": 3.136473979421275e-05,
+      "loss": 0.1828,
+      "step": 4728
+    },
+    {
+      "epoch": 3.5596537448249905,
+      "grad_norm": 0.24111253023147583,
+      "learning_rate": 3.135740955318342e-05,
+      "loss": 0.2799,
+      "step": 4729
+    },
+    {
+      "epoch": 3.5604064734663154,
+      "grad_norm": 0.22044381499290466,
+      "learning_rate": 3.1350078890313576e-05,
+      "loss": 0.2396,
+      "step": 4730
+    },
+    {
+      "epoch": 3.56115920210764,
+      "grad_norm": 0.20934171974658966,
+      "learning_rate": 3.134274780640489e-05,
+      "loss": 0.2031,
+      "step": 4731
+    },
+    {
+      "epoch": 3.561911930748965,
+      "grad_norm": 0.21612851321697235,
+      "learning_rate": 3.1335416302259095e-05,
+      "loss": 0.2473,
+      "step": 4732
+    },
+    {
+      "epoch": 3.5626646593902898,
+      "grad_norm": 0.265924870967865,
+      "learning_rate": 3.132808437867792e-05,
+      "loss": 0.2639,
+      "step": 4733
+    },
+    {
+      "epoch": 3.5634173880316147,
+      "grad_norm": 0.2187691479921341,
+      "learning_rate": 3.1320752036463205e-05,
+      "loss": 0.3033,
+      "step": 4734
+    },
+    {
+      "epoch": 3.5641701166729396,
+      "grad_norm": 0.225166916847229,
+      "learning_rate": 3.13134192764168e-05,
+      "loss": 0.2661,
+      "step": 4735
+    },
+    {
+      "epoch": 3.564922845314264,
+      "grad_norm": 0.21113333106040955,
+      "learning_rate": 3.130608609934058e-05,
+      "loss": 0.2378,
+      "step": 4736
+    },
+    {
+      "epoch": 3.565675573955589,
+      "grad_norm": 0.2388783097267151,
+      "learning_rate": 3.129875250603651e-05,
+      "loss": 0.2421,
+      "step": 4737
+    },
+    {
+      "epoch": 3.566428302596914,
+      "grad_norm": 0.22409674525260925,
+      "learning_rate": 3.129141849730659e-05,
+      "loss": 0.3212,
+      "step": 4738
+    },
+    {
+      "epoch": 3.5671810312382384,
+      "grad_norm": 0.23214958608150482,
+      "learning_rate": 3.128408407395283e-05,
+      "loss": 0.2819,
+      "step": 4739
+    },
+    {
+      "epoch": 3.5679337598795633,
+      "grad_norm": 0.20376986265182495,
+      "learning_rate": 3.1276749236777326e-05,
+      "loss": 0.1564,
+      "step": 4740
+    },
+    {
+      "epoch": 3.568686488520888,
+      "grad_norm": 0.2304265946149826,
+      "learning_rate": 3.1269413986582206e-05,
+      "loss": 0.3457,
+      "step": 4741
+    },
+    {
+      "epoch": 3.569439217162213,
+      "grad_norm": 0.21088165044784546,
+      "learning_rate": 3.126207832416964e-05,
+      "loss": 0.2329,
+      "step": 4742
+    },
+    {
+      "epoch": 3.570191945803538,
+      "grad_norm": 0.23325680196285248,
+      "learning_rate": 3.125474225034185e-05,
+      "loss": 0.4094,
+      "step": 4743
+    },
+    {
+      "epoch": 3.5709446744448625,
+      "grad_norm": 0.19287259876728058,
+      "learning_rate": 3.1247405765901084e-05,
+      "loss": 0.2076,
+      "step": 4744
+    },
+    {
+      "epoch": 3.5716974030861874,
+      "grad_norm": 0.24154983460903168,
+      "learning_rate": 3.124006887164966e-05,
+      "loss": 0.2801,
+      "step": 4745
+    },
+    {
+      "epoch": 3.5724501317275124,
+      "grad_norm": 0.257909893989563,
+      "learning_rate": 3.123273156838993e-05,
+      "loss": 0.2724,
+      "step": 4746
+    },
+    {
+      "epoch": 3.573202860368837,
+      "grad_norm": 0.21960952877998352,
+      "learning_rate": 3.122539385692429e-05,
+      "loss": 0.2221,
+      "step": 4747
+    },
+    {
+      "epoch": 3.5739555890101617,
+      "grad_norm": 0.2408861368894577,
+      "learning_rate": 3.1218055738055174e-05,
+      "loss": 0.3598,
+      "step": 4748
+    },
+    {
+      "epoch": 3.5747083176514867,
+      "grad_norm": 0.24846968054771423,
+      "learning_rate": 3.1210717212585065e-05,
+      "loss": 0.1879,
+      "step": 4749
+    },
+    {
+      "epoch": 3.5754610462928116,
+      "grad_norm": 0.2561246156692505,
+      "learning_rate": 3.1203378281316515e-05,
+      "loss": 0.2634,
+      "step": 4750
+    },
+    {
+      "epoch": 3.5762137749341365,
+      "grad_norm": 0.2329346239566803,
+      "learning_rate": 3.119603894505208e-05,
+      "loss": 0.2616,
+      "step": 4751
+    },
+    {
+      "epoch": 3.576966503575461,
+      "grad_norm": 0.24927261471748352,
+      "learning_rate": 3.118869920459438e-05,
+      "loss": 0.252,
+      "step": 4752
+    },
+    {
+      "epoch": 3.577719232216786,
+      "grad_norm": 0.2625426948070526,
+      "learning_rate": 3.118135906074608e-05,
+      "loss": 0.2073,
+      "step": 4753
+    },
+    {
+      "epoch": 3.578471960858111,
+      "grad_norm": 0.2609004080295563,
+      "learning_rate": 3.1174018514309894e-05,
+      "loss": 0.2791,
+      "step": 4754
+    },
+    {
+      "epoch": 3.5792246894994353,
+      "grad_norm": 0.22738653421401978,
+      "learning_rate": 3.1166677566088556e-05,
+      "loss": 0.2633,
+      "step": 4755
+    },
+    {
+      "epoch": 3.57997741814076,
+      "grad_norm": 0.2210693657398224,
+      "learning_rate": 3.1159336216884884e-05,
+      "loss": 0.3722,
+      "step": 4756
+    },
+    {
+      "epoch": 3.580730146782085,
+      "grad_norm": 0.27977901697158813,
+      "learning_rate": 3.11519944675017e-05,
+      "loss": 0.3906,
+      "step": 4757
+    },
+    {
+      "epoch": 3.5814828754234096,
+      "grad_norm": 0.1966094821691513,
+      "learning_rate": 3.114465231874189e-05,
+      "loss": 0.1623,
+      "step": 4758
+    },
+    {
+      "epoch": 3.5822356040647345,
+      "grad_norm": 0.2783859968185425,
+      "learning_rate": 3.1137309771408384e-05,
+      "loss": 0.2572,
+      "step": 4759
+    },
+    {
+      "epoch": 3.5829883327060594,
+      "grad_norm": 0.24252162873744965,
+      "learning_rate": 3.112996682630415e-05,
+      "loss": 0.1826,
+      "step": 4760
+    },
+    {
+      "epoch": 3.5837410613473843,
+      "grad_norm": 0.25034675002098083,
+      "learning_rate": 3.11226234842322e-05,
+      "loss": 0.2073,
+      "step": 4761
+    },
+    {
+      "epoch": 3.5844937899887093,
+      "grad_norm": 0.22137326002120972,
+      "learning_rate": 3.111527974599559e-05,
+      "loss": 0.2455,
+      "step": 4762
+    },
+    {
+      "epoch": 3.5852465186300337,
+      "grad_norm": 0.2545652687549591,
+      "learning_rate": 3.110793561239743e-05,
+      "loss": 0.3314,
+      "step": 4763
+    },
+    {
+      "epoch": 3.5859992472713587,
+      "grad_norm": 0.2734377086162567,
+      "learning_rate": 3.110059108424086e-05,
+      "loss": 0.2403,
+      "step": 4764
+    },
+    {
+      "epoch": 3.5867519759126836,
+      "grad_norm": 0.2514321208000183,
+      "learning_rate": 3.109324616232905e-05,
+      "loss": 0.2073,
+      "step": 4765
+    },
+    {
+      "epoch": 3.587504704554008,
+      "grad_norm": 0.22296380996704102,
+      "learning_rate": 3.1085900847465246e-05,
+      "loss": 0.292,
+      "step": 4766
+    },
+    {
+      "epoch": 3.588257433195333,
+      "grad_norm": 0.20941561460494995,
+      "learning_rate": 3.107855514045272e-05,
+      "loss": 0.2851,
+      "step": 4767
+    },
+    {
+      "epoch": 3.589010161836658,
+      "grad_norm": 0.2288130670785904,
+      "learning_rate": 3.107120904209478e-05,
+      "loss": 0.2786,
+      "step": 4768
+    },
+    {
+      "epoch": 3.589762890477983,
+      "grad_norm": 0.2196059376001358,
+      "learning_rate": 3.106386255319478e-05,
+      "loss": 0.3043,
+      "step": 4769
+    },
+    {
+      "epoch": 3.5905156191193077,
+      "grad_norm": 0.22183668613433838,
+      "learning_rate": 3.1056515674556146e-05,
+      "loss": 0.2069,
+      "step": 4770
+    },
+    {
+      "epoch": 3.591268347760632,
+      "grad_norm": 0.20780779421329498,
+      "learning_rate": 3.1049168406982296e-05,
+      "loss": 0.2236,
+      "step": 4771
+    },
+    {
+      "epoch": 3.592021076401957,
+      "grad_norm": 0.2356027364730835,
+      "learning_rate": 3.1041820751276726e-05,
+      "loss": 0.3269,
+      "step": 4772
+    },
+    {
+      "epoch": 3.592773805043282,
+      "grad_norm": 0.23140111565589905,
+      "learning_rate": 3.103447270824297e-05,
+      "loss": 0.2737,
+      "step": 4773
+    },
+    {
+      "epoch": 3.5935265336846065,
+      "grad_norm": 0.21066869795322418,
+      "learning_rate": 3.1027124278684586e-05,
+      "loss": 0.1998,
+      "step": 4774
+    },
+    {
+      "epoch": 3.5942792623259314,
+      "grad_norm": 0.21444103121757507,
+      "learning_rate": 3.1019775463405195e-05,
+      "loss": 0.2462,
+      "step": 4775
+    },
+    {
+      "epoch": 3.5950319909672563,
+      "grad_norm": 0.1907918006181717,
+      "learning_rate": 3.101242626320846e-05,
+      "loss": 0.2009,
+      "step": 4776
+    },
+    {
+      "epoch": 3.5957847196085813,
+      "grad_norm": 0.21659192442893982,
+      "learning_rate": 3.100507667889806e-05,
+      "loss": 0.2222,
+      "step": 4777
+    },
+    {
+      "epoch": 3.596537448249906,
+      "grad_norm": 0.19378384947776794,
+      "learning_rate": 3.0997726711277744e-05,
+      "loss": 0.168,
+      "step": 4778
+    },
+    {
+      "epoch": 3.5972901768912307,
+      "grad_norm": 0.21419967710971832,
+      "learning_rate": 3.099037636115129e-05,
+      "loss": 0.1948,
+      "step": 4779
+    },
+    {
+      "epoch": 3.5980429055325556,
+      "grad_norm": 0.2136116623878479,
+      "learning_rate": 3.0983025629322535e-05,
+      "loss": 0.2601,
+      "step": 4780
+    },
+    {
+      "epoch": 3.5987956341738805,
+      "grad_norm": 0.21488958597183228,
+      "learning_rate": 3.097567451659533e-05,
+      "loss": 0.29,
+      "step": 4781
+    },
+    {
+      "epoch": 3.599548362815205,
+      "grad_norm": 0.2574782371520996,
+      "learning_rate": 3.096832302377358e-05,
+      "loss": 0.2536,
+      "step": 4782
+    },
+    {
+      "epoch": 3.60030109145653,
+      "grad_norm": 0.22757631540298462,
+      "learning_rate": 3.096097115166124e-05,
+      "loss": 0.2434,
+      "step": 4783
+    },
+    {
+      "epoch": 3.601053820097855,
+      "grad_norm": 0.21044713258743286,
+      "learning_rate": 3.09536189010623e-05,
+      "loss": 0.2914,
+      "step": 4784
+    },
+    {
+      "epoch": 3.6018065487391793,
+      "grad_norm": 0.26193374395370483,
+      "learning_rate": 3.0946266272780793e-05,
+      "loss": 0.2991,
+      "step": 4785
+    },
+    {
+      "epoch": 3.602559277380504,
+      "grad_norm": 0.20791497826576233,
+      "learning_rate": 3.093891326762078e-05,
+      "loss": 0.2405,
+      "step": 4786
+    },
+    {
+      "epoch": 3.603312006021829,
+      "grad_norm": 0.22689415514469147,
+      "learning_rate": 3.093155988638639e-05,
+      "loss": 0.3241,
+      "step": 4787
+    },
+    {
+      "epoch": 3.604064734663154,
+      "grad_norm": 0.23813918232917786,
+      "learning_rate": 3.092420612988177e-05,
+      "loss": 0.2939,
+      "step": 4788
+    },
+    {
+      "epoch": 3.604817463304479,
+      "grad_norm": 0.20832981169223785,
+      "learning_rate": 3.091685199891112e-05,
+      "loss": 0.2153,
+      "step": 4789
+    },
+    {
+      "epoch": 3.6055701919458034,
+      "grad_norm": 0.23879937827587128,
+      "learning_rate": 3.090949749427866e-05,
+      "loss": 0.1234,
+      "step": 4790
+    },
+    {
+      "epoch": 3.6063229205871283,
+      "grad_norm": 0.19066521525382996,
+      "learning_rate": 3.090214261678869e-05,
+      "loss": 0.2865,
+      "step": 4791
+    },
+    {
+      "epoch": 3.6070756492284533,
+      "grad_norm": 0.2644321620464325,
+      "learning_rate": 3.089478736724551e-05,
+      "loss": 0.2438,
+      "step": 4792
+    },
+    {
+      "epoch": 3.6078283778697777,
+      "grad_norm": 0.2074516862630844,
+      "learning_rate": 3.08874317464535e-05,
+      "loss": 0.1946,
+      "step": 4793
+    },
+    {
+      "epoch": 3.6085811065111026,
+      "grad_norm": 0.21272635459899902,
+      "learning_rate": 3.0880075755217037e-05,
+      "loss": 0.1795,
+      "step": 4794
+    },
+    {
+      "epoch": 3.6093338351524276,
+      "grad_norm": 0.19117318093776703,
+      "learning_rate": 3.087271939434057e-05,
+      "loss": 0.1617,
+      "step": 4795
+    },
+    {
+      "epoch": 3.6100865637937525,
+      "grad_norm": 0.24222403764724731,
+      "learning_rate": 3.08653626646286e-05,
+      "loss": 0.3435,
+      "step": 4796
+    },
+    {
+      "epoch": 3.6108392924350774,
+      "grad_norm": 0.2480611950159073,
+      "learning_rate": 3.08580055668856e-05,
+      "loss": 0.286,
+      "step": 4797
+    },
+    {
+      "epoch": 3.611592021076402,
+      "grad_norm": 0.33622655272483826,
+      "learning_rate": 3.0850648101916173e-05,
+      "loss": 0.2994,
+      "step": 4798
+    },
+    {
+      "epoch": 3.612344749717727,
+      "grad_norm": 0.2411850094795227,
+      "learning_rate": 3.084329027052491e-05,
+      "loss": 0.2801,
+      "step": 4799
+    },
+    {
+      "epoch": 3.6130974783590517,
+      "grad_norm": 0.22896000742912292,
+      "learning_rate": 3.0835932073516444e-05,
+      "loss": 0.2581,
+      "step": 4800
+    },
+    {
+      "epoch": 3.6130974783590517,
+      "eval_loss": 0.25820624828338623,
+      "eval_runtime": 452.9039,
+      "eval_samples_per_second": 21.256,
+      "eval_steps_per_second": 0.665,
+      "step": 4800
+    },
+    {
+      "epoch": 3.613850207000376,
+      "grad_norm": 0.24396610260009766,
+      "learning_rate": 3.082857351169547e-05,
+      "loss": 0.19,
+      "step": 4801
+    },
+    {
+      "epoch": 3.614602935641701,
+      "grad_norm": 0.2545047998428345,
+      "learning_rate": 3.0821214585866696e-05,
+      "loss": 0.2292,
+      "step": 4802
+    },
+    {
+      "epoch": 3.615355664283026,
+      "grad_norm": 0.2523135542869568,
+      "learning_rate": 3.081385529683488e-05,
+      "loss": 0.2744,
+      "step": 4803
+    },
+    {
+      "epoch": 3.616108392924351,
+      "grad_norm": 0.23640425503253937,
+      "learning_rate": 3.080649564540485e-05,
+      "loss": 0.1828,
+      "step": 4804
+    },
+    {
+      "epoch": 3.6168611215656754,
+      "grad_norm": 0.25340649485588074,
+      "learning_rate": 3.079913563238141e-05,
+      "loss": 0.328,
+      "step": 4805
+    },
+    {
+      "epoch": 3.6176138502070003,
+      "grad_norm": 0.22594138979911804,
+      "learning_rate": 3.079177525856947e-05,
+      "loss": 0.2915,
+      "step": 4806
+    },
+    {
+      "epoch": 3.6183665788483252,
+      "grad_norm": 0.23372043669223785,
+      "learning_rate": 3.0784414524773924e-05,
+      "loss": 0.202,
+      "step": 4807
+    },
+    {
+      "epoch": 3.61911930748965,
+      "grad_norm": 0.2676745057106018,
+      "learning_rate": 3.077705343179975e-05,
+      "loss": 0.3056,
+      "step": 4808
+    },
+    {
+      "epoch": 3.6198720361309746,
+      "grad_norm": 0.23161888122558594,
+      "learning_rate": 3.0769691980451946e-05,
+      "loss": 0.2767,
+      "step": 4809
+    },
+    {
+      "epoch": 3.6206247647722996,
+      "grad_norm": 0.2698519229888916,
+      "learning_rate": 3.076233017153553e-05,
+      "loss": 0.2833,
+      "step": 4810
+    },
+    {
+      "epoch": 3.6213774934136245,
+      "grad_norm": 0.2886054515838623,
+      "learning_rate": 3.0754968005855605e-05,
+      "loss": 0.2663,
+      "step": 4811
+    },
+    {
+      "epoch": 3.622130222054949,
+      "grad_norm": 0.26353025436401367,
+      "learning_rate": 3.074760548421727e-05,
+      "loss": 0.2616,
+      "step": 4812
+    },
+    {
+      "epoch": 3.622882950696274,
+      "grad_norm": 0.23660698533058167,
+      "learning_rate": 3.074024260742567e-05,
+      "loss": 0.294,
+      "step": 4813
+    },
+    {
+      "epoch": 3.623635679337599,
+      "grad_norm": 0.2499127984046936,
+      "learning_rate": 3.0732879376286026e-05,
+      "loss": 0.1866,
+      "step": 4814
+    },
+    {
+      "epoch": 3.6243884079789237,
+      "grad_norm": 0.23045708239078522,
+      "learning_rate": 3.072551579160354e-05,
+      "loss": 0.2354,
+      "step": 4815
+    },
+    {
+      "epoch": 3.6251411366202486,
+      "grad_norm": 0.26418015360832214,
+      "learning_rate": 3.071815185418351e-05,
+      "loss": 0.313,
+      "step": 4816
+    },
+    {
+      "epoch": 3.625893865261573,
+      "grad_norm": 0.24416646361351013,
+      "learning_rate": 3.0710787564831224e-05,
+      "loss": 0.2663,
+      "step": 4817
+    },
+    {
+      "epoch": 3.626646593902898,
+      "grad_norm": 0.22831043601036072,
+      "learning_rate": 3.070342292435204e-05,
+      "loss": 0.2714,
+      "step": 4818
+    },
+    {
+      "epoch": 3.627399322544223,
+      "grad_norm": 0.24665769934654236,
+      "learning_rate": 3.0696057933551333e-05,
+      "loss": 0.2833,
+      "step": 4819
+    },
+    {
+      "epoch": 3.6281520511855474,
+      "grad_norm": 0.2145809680223465,
+      "learning_rate": 3.068869259323455e-05,
+      "loss": 0.278,
+      "step": 4820
+    },
+    {
+      "epoch": 3.6289047798268723,
+      "grad_norm": 0.2842905521392822,
+      "learning_rate": 3.068132690420713e-05,
+      "loss": 0.2896,
+      "step": 4821
+    },
+    {
+      "epoch": 3.6296575084681972,
+      "grad_norm": 0.20013706386089325,
+      "learning_rate": 3.067396086727458e-05,
+      "loss": 0.1975,
+      "step": 4822
+    },
+    {
+      "epoch": 3.630410237109522,
+      "grad_norm": 0.24887973070144653,
+      "learning_rate": 3.0666594483242446e-05,
+      "loss": 0.325,
+      "step": 4823
+    },
+    {
+      "epoch": 3.631162965750847,
+      "grad_norm": 0.21456503868103027,
+      "learning_rate": 3.06592277529163e-05,
+      "loss": 0.2233,
+      "step": 4824
+    },
+    {
+      "epoch": 3.6319156943921715,
+      "grad_norm": 0.2539563775062561,
+      "learning_rate": 3.0651860677101754e-05,
+      "loss": 0.2765,
+      "step": 4825
+    },
+    {
+      "epoch": 3.6326684230334965,
+      "grad_norm": 0.22977475821971893,
+      "learning_rate": 3.0644493256604466e-05,
+      "loss": 0.2625,
+      "step": 4826
+    },
+    {
+      "epoch": 3.6334211516748214,
+      "grad_norm": 0.22674930095672607,
+      "learning_rate": 3.063712549223013e-05,
+      "loss": 0.2597,
+      "step": 4827
+    },
+    {
+      "epoch": 3.634173880316146,
+      "grad_norm": 0.22804288566112518,
+      "learning_rate": 3.0629757384784454e-05,
+      "loss": 0.2125,
+      "step": 4828
+    },
+    {
+      "epoch": 3.6349266089574708,
+      "grad_norm": 0.2076561450958252,
+      "learning_rate": 3.062238893507322e-05,
+      "loss": 0.1743,
+      "step": 4829
+    },
+    {
+      "epoch": 3.6356793375987957,
+      "grad_norm": 0.23488694429397583,
+      "learning_rate": 3.061502014390222e-05,
+      "loss": 0.2186,
+      "step": 4830
+    },
+    {
+      "epoch": 3.63643206624012,
+      "grad_norm": 0.2576046288013458,
+      "learning_rate": 3.0607651012077315e-05,
+      "loss": 0.2901,
+      "step": 4831
+    },
+    {
+      "epoch": 3.637184794881445,
+      "grad_norm": 0.23835991322994232,
+      "learning_rate": 3.060028154040436e-05,
+      "loss": 0.2279,
+      "step": 4832
+    },
+    {
+      "epoch": 3.63793752352277,
+      "grad_norm": 0.22578774392604828,
+      "learning_rate": 3.059291172968927e-05,
+      "loss": 0.3388,
+      "step": 4833
+    },
+    {
+      "epoch": 3.638690252164095,
+      "grad_norm": 0.2377738356590271,
+      "learning_rate": 3.058554158073802e-05,
+      "loss": 0.3255,
+      "step": 4834
+    },
+    {
+      "epoch": 3.63944298080542,
+      "grad_norm": 0.2083633989095688,
+      "learning_rate": 3.057817109435657e-05,
+      "loss": 0.2603,
+      "step": 4835
+    },
+    {
+      "epoch": 3.6401957094467443,
+      "grad_norm": 0.2007228434085846,
+      "learning_rate": 3.057080027135096e-05,
+      "loss": 0.1908,
+      "step": 4836
+    },
+    {
+      "epoch": 3.6409484380880692,
+      "grad_norm": 0.2841181755065918,
+      "learning_rate": 3.056342911252725e-05,
+      "loss": 0.261,
+      "step": 4837
+    },
+    {
+      "epoch": 3.641701166729394,
+      "grad_norm": 0.19502191245555878,
+      "learning_rate": 3.055605761869153e-05,
+      "loss": 0.1692,
+      "step": 4838
+    },
+    {
+      "epoch": 3.6424538953707186,
+      "grad_norm": 0.24782055616378784,
+      "learning_rate": 3.0548685790649955e-05,
+      "loss": 0.2503,
+      "step": 4839
+    },
+    {
+      "epoch": 3.6432066240120435,
+      "grad_norm": 0.2586572468280792,
+      "learning_rate": 3.054131362920868e-05,
+      "loss": 0.1752,
+      "step": 4840
+    },
+    {
+      "epoch": 3.6439593526533685,
+      "grad_norm": 0.2266005575656891,
+      "learning_rate": 3.0533941135173924e-05,
+      "loss": 0.2082,
+      "step": 4841
+    },
+    {
+      "epoch": 3.6447120812946934,
+      "grad_norm": 0.23031409084796906,
+      "learning_rate": 3.052656830935192e-05,
+      "loss": 0.2284,
+      "step": 4842
+    },
+    {
+      "epoch": 3.6454648099360183,
+      "grad_norm": 0.22700883448123932,
+      "learning_rate": 3.0519195152548964e-05,
+      "loss": 0.2346,
+      "step": 4843
+    },
+    {
+      "epoch": 3.6462175385773428,
+      "grad_norm": 0.19426539540290833,
+      "learning_rate": 3.051182166557136e-05,
+      "loss": 0.2768,
+      "step": 4844
+    },
+    {
+      "epoch": 3.6469702672186677,
+      "grad_norm": 0.22896550595760345,
+      "learning_rate": 3.0504447849225466e-05,
+      "loss": 0.2595,
+      "step": 4845
+    },
+    {
+      "epoch": 3.6477229958599926,
+      "grad_norm": 0.2140006273984909,
+      "learning_rate": 3.0497073704317673e-05,
+      "loss": 0.1928,
+      "step": 4846
+    },
+    {
+      "epoch": 3.648475724501317,
+      "grad_norm": 0.2582569718360901,
+      "learning_rate": 3.048969923165441e-05,
+      "loss": 0.2261,
+      "step": 4847
+    },
+    {
+      "epoch": 3.649228453142642,
+      "grad_norm": 0.21107441186904907,
+      "learning_rate": 3.048232443204213e-05,
+      "loss": 0.1879,
+      "step": 4848
+    },
+    {
+      "epoch": 3.649981181783967,
+      "grad_norm": 0.19312848150730133,
+      "learning_rate": 3.047494930628733e-05,
+      "loss": 0.2137,
+      "step": 4849
+    },
+    {
+      "epoch": 3.650733910425292,
+      "grad_norm": 0.2076844871044159,
+      "learning_rate": 3.0467573855196558e-05,
+      "loss": 0.2614,
+      "step": 4850
+    },
+    {
+      "epoch": 3.6514866390666167,
+      "grad_norm": 0.20677466690540314,
+      "learning_rate": 3.0460198079576358e-05,
+      "loss": 0.2154,
+      "step": 4851
+    },
+    {
+      "epoch": 3.652239367707941,
+      "grad_norm": 0.2481384575366974,
+      "learning_rate": 3.0452821980233352e-05,
+      "loss": 0.2287,
+      "step": 4852
+    },
+    {
+      "epoch": 3.652992096349266,
+      "grad_norm": 0.1920328140258789,
+      "learning_rate": 3.0445445557974177e-05,
+      "loss": 0.2325,
+      "step": 4853
+    },
+    {
+      "epoch": 3.653744824990591,
+      "grad_norm": 0.2354832887649536,
+      "learning_rate": 3.0438068813605502e-05,
+      "loss": 0.3007,
+      "step": 4854
+    },
+    {
+      "epoch": 3.6544975536319155,
+      "grad_norm": 0.2490939497947693,
+      "learning_rate": 3.0430691747934038e-05,
+      "loss": 0.1881,
+      "step": 4855
+    },
+    {
+      "epoch": 3.6552502822732404,
+      "grad_norm": 0.1936967521905899,
+      "learning_rate": 3.042331436176653e-05,
+      "loss": 0.2193,
+      "step": 4856
+    },
+    {
+      "epoch": 3.6560030109145654,
+      "grad_norm": 0.250393807888031,
+      "learning_rate": 3.0415936655909754e-05,
+      "loss": 0.1926,
+      "step": 4857
+    },
+    {
+      "epoch": 3.65675573955589,
+      "grad_norm": 0.25850898027420044,
+      "learning_rate": 3.0408558631170542e-05,
+      "loss": 0.3052,
+      "step": 4858
+    },
+    {
+      "epoch": 3.6575084681972148,
+      "grad_norm": 0.20357652008533478,
+      "learning_rate": 3.040118028835572e-05,
+      "loss": 0.2642,
+      "step": 4859
+    },
+    {
+      "epoch": 3.6582611968385397,
+      "grad_norm": 0.25979605317115784,
+      "learning_rate": 3.039380162827219e-05,
+      "loss": 0.3118,
+      "step": 4860
+    },
+    {
+      "epoch": 3.6590139254798646,
+      "grad_norm": 0.19522950053215027,
+      "learning_rate": 3.038642265172686e-05,
+      "loss": 0.1808,
+      "step": 4861
+    },
+    {
+      "epoch": 3.6597666541211895,
+      "grad_norm": 0.25541573762893677,
+      "learning_rate": 3.0379043359526695e-05,
+      "loss": 0.1943,
+      "step": 4862
+    },
+    {
+      "epoch": 3.660519382762514,
+      "grad_norm": 0.19946402311325073,
+      "learning_rate": 3.0371663752478668e-05,
+      "loss": 0.1,
+      "step": 4863
+    },
+    {
+      "epoch": 3.661272111403839,
+      "grad_norm": 0.2441890388727188,
+      "learning_rate": 3.0364283831389818e-05,
+      "loss": 0.2516,
+      "step": 4864
+    },
+    {
+      "epoch": 3.662024840045164,
+      "grad_norm": 0.2364385724067688,
+      "learning_rate": 3.0356903597067194e-05,
+      "loss": 0.1686,
+      "step": 4865
+    },
+    {
+      "epoch": 3.6627775686864883,
+      "grad_norm": 0.2499358355998993,
+      "learning_rate": 3.0349523050317885e-05,
+      "loss": 0.2628,
+      "step": 4866
+    },
+    {
+      "epoch": 3.663530297327813,
+      "grad_norm": 0.23725810647010803,
+      "learning_rate": 3.0342142191949013e-05,
+      "loss": 0.2291,
+      "step": 4867
+    },
+    {
+      "epoch": 3.664283025969138,
+      "grad_norm": 0.2781839966773987,
+      "learning_rate": 3.0334761022767754e-05,
+      "loss": 0.2239,
+      "step": 4868
+    },
+    {
+      "epoch": 3.665035754610463,
+      "grad_norm": 0.2656850218772888,
+      "learning_rate": 3.0327379543581293e-05,
+      "loss": 0.1915,
+      "step": 4869
+    },
+    {
+      "epoch": 3.665788483251788,
+      "grad_norm": 0.20144368708133698,
+      "learning_rate": 3.031999775519685e-05,
+      "loss": 0.2031,
+      "step": 4870
+    },
+    {
+      "epoch": 3.6665412118931124,
+      "grad_norm": 0.2398781031370163,
+      "learning_rate": 3.0312615658421688e-05,
+      "loss": 0.1739,
+      "step": 4871
+    },
+    {
+      "epoch": 3.6672939405344374,
+      "grad_norm": 0.2179783582687378,
+      "learning_rate": 3.0305233254063114e-05,
+      "loss": 0.2434,
+      "step": 4872
+    },
+    {
+      "epoch": 3.6680466691757623,
+      "grad_norm": 0.21857503056526184,
+      "learning_rate": 3.0297850542928448e-05,
+      "loss": 0.1302,
+      "step": 4873
+    },
+    {
+      "epoch": 3.6687993978170867,
+      "grad_norm": 0.250797837972641,
+      "learning_rate": 3.029046752582505e-05,
+      "loss": 0.2876,
+      "step": 4874
+    },
+    {
+      "epoch": 3.6695521264584117,
+      "grad_norm": 0.19488811492919922,
+      "learning_rate": 3.0283084203560324e-05,
+      "loss": 0.1953,
+      "step": 4875
+    },
+    {
+      "epoch": 3.6703048550997366,
+      "grad_norm": 0.2371215522289276,
+      "learning_rate": 3.027570057694169e-05,
+      "loss": 0.1823,
+      "step": 4876
+    },
+    {
+      "epoch": 3.6710575837410615,
+      "grad_norm": 0.22229599952697754,
+      "learning_rate": 3.0268316646776618e-05,
+      "loss": 0.2506,
+      "step": 4877
+    },
+    {
+      "epoch": 3.6718103123823864,
+      "grad_norm": 0.1838085800409317,
+      "learning_rate": 3.0260932413872595e-05,
+      "loss": 0.2279,
+      "step": 4878
+    },
+    {
+      "epoch": 3.672563041023711,
+      "grad_norm": 0.24501128494739532,
+      "learning_rate": 3.025354787903715e-05,
+      "loss": 0.1775,
+      "step": 4879
+    },
+    {
+      "epoch": 3.673315769665036,
+      "grad_norm": 0.21434840559959412,
+      "learning_rate": 3.0246163043077853e-05,
+      "loss": 0.1897,
+      "step": 4880
+    },
+    {
+      "epoch": 3.6740684983063607,
+      "grad_norm": 0.21759526431560516,
+      "learning_rate": 3.0238777906802302e-05,
+      "loss": 0.3066,
+      "step": 4881
+    },
+    {
+      "epoch": 3.674821226947685,
+      "grad_norm": 0.2665358781814575,
+      "learning_rate": 3.023139247101811e-05,
+      "loss": 0.2643,
+      "step": 4882
+    },
+    {
+      "epoch": 3.67557395558901,
+      "grad_norm": 0.22974827885627747,
+      "learning_rate": 3.0224006736532946e-05,
+      "loss": 0.2993,
+      "step": 4883
+    },
+    {
+      "epoch": 3.676326684230335,
+      "grad_norm": 0.2290305197238922,
+      "learning_rate": 3.0216620704154497e-05,
+      "loss": 0.2484,
+      "step": 4884
+    },
+    {
+      "epoch": 3.6770794128716595,
+      "grad_norm": 0.2337796986103058,
+      "learning_rate": 3.0209234374690498e-05,
+      "loss": 0.282,
+      "step": 4885
+    },
+    {
+      "epoch": 3.6778321415129844,
+      "grad_norm": 0.2113298773765564,
+      "learning_rate": 3.02018477489487e-05,
+      "loss": 0.1833,
+      "step": 4886
+    },
+    {
+      "epoch": 3.6785848701543093,
+      "grad_norm": 0.2291206568479538,
+      "learning_rate": 3.0194460827736887e-05,
+      "loss": 0.2604,
+      "step": 4887
+    },
+    {
+      "epoch": 3.6793375987956343,
+      "grad_norm": 0.25271928310394287,
+      "learning_rate": 3.0187073611862893e-05,
+      "loss": 0.3043,
+      "step": 4888
+    },
+    {
+      "epoch": 3.680090327436959,
+      "grad_norm": 0.2417176216840744,
+      "learning_rate": 3.0179686102134576e-05,
+      "loss": 0.2429,
+      "step": 4889
+    },
+    {
+      "epoch": 3.6808430560782837,
+      "grad_norm": 0.20943641662597656,
+      "learning_rate": 3.01722982993598e-05,
+      "loss": 0.2158,
+      "step": 4890
+    },
+    {
+      "epoch": 3.6815957847196086,
+      "grad_norm": 0.262931227684021,
+      "learning_rate": 3.0164910204346513e-05,
+      "loss": 0.1905,
+      "step": 4891
+    },
+    {
+      "epoch": 3.6823485133609335,
+      "grad_norm": 0.25925493240356445,
+      "learning_rate": 3.0157521817902652e-05,
+      "loss": 0.1771,
+      "step": 4892
+    },
+    {
+      "epoch": 3.683101242002258,
+      "grad_norm": 0.2180473804473877,
+      "learning_rate": 3.0150133140836197e-05,
+      "loss": 0.1401,
+      "step": 4893
+    },
+    {
+      "epoch": 3.683853970643583,
+      "grad_norm": 0.25075820088386536,
+      "learning_rate": 3.014274417395517e-05,
+      "loss": 0.2744,
+      "step": 4894
+    },
+    {
+      "epoch": 3.684606699284908,
+      "grad_norm": 0.1901656836271286,
+      "learning_rate": 3.013535491806761e-05,
+      "loss": 0.1894,
+      "step": 4895
+    },
+    {
+      "epoch": 3.6853594279262327,
+      "grad_norm": 0.25133615732192993,
+      "learning_rate": 3.012796537398159e-05,
+      "loss": 0.2978,
+      "step": 4896
+    },
+    {
+      "epoch": 3.6861121565675576,
+      "grad_norm": 0.19855251908302307,
+      "learning_rate": 3.0120575542505242e-05,
+      "loss": 0.2825,
+      "step": 4897
+    },
+    {
+      "epoch": 3.686864885208882,
+      "grad_norm": 0.23921585083007812,
+      "learning_rate": 3.011318542444669e-05,
+      "loss": 0.2507,
+      "step": 4898
+    },
+    {
+      "epoch": 3.687617613850207,
+      "grad_norm": 0.22538770735263824,
+      "learning_rate": 3.0105795020614104e-05,
+      "loss": 0.2539,
+      "step": 4899
+    },
+    {
+      "epoch": 3.688370342491532,
+      "grad_norm": 0.2056659609079361,
+      "learning_rate": 3.0098404331815695e-05,
+      "loss": 0.1859,
+      "step": 4900
+    },
+    {
+      "epoch": 3.6891230711328564,
+      "grad_norm": 0.20844072103500366,
+      "learning_rate": 3.0091013358859687e-05,
+      "loss": 0.29,
+      "step": 4901
+    },
+    {
+      "epoch": 3.6898757997741813,
+      "grad_norm": 0.21161197125911713,
+      "learning_rate": 3.008362210255436e-05,
+      "loss": 0.2497,
+      "step": 4902
+    },
+    {
+      "epoch": 3.6906285284155063,
+      "grad_norm": 0.23654623329639435,
+      "learning_rate": 3.0076230563708006e-05,
+      "loss": 0.1881,
+      "step": 4903
+    },
+    {
+      "epoch": 3.691381257056831,
+      "grad_norm": 0.257632315158844,
+      "learning_rate": 3.0068838743128947e-05,
+      "loss": 0.2618,
+      "step": 4904
+    },
+    {
+      "epoch": 3.6921339856981557,
+      "grad_norm": 0.24203605949878693,
+      "learning_rate": 3.0061446641625546e-05,
+      "loss": 0.2545,
+      "step": 4905
+    },
+    {
+      "epoch": 3.6928867143394806,
+      "grad_norm": 0.2524026036262512,
+      "learning_rate": 3.0054054260006194e-05,
+      "loss": 0.2011,
+      "step": 4906
+    },
+    {
+      "epoch": 3.6936394429808055,
+      "grad_norm": 0.2827187776565552,
+      "learning_rate": 3.0046661599079314e-05,
+      "loss": 0.2443,
+      "step": 4907
+    },
+    {
+      "epoch": 3.6943921716221304,
+      "grad_norm": 0.2363114207983017,
+      "learning_rate": 3.0039268659653336e-05,
+      "loss": 0.2686,
+      "step": 4908
+    },
+    {
+      "epoch": 3.695144900263455,
+      "grad_norm": 0.27060961723327637,
+      "learning_rate": 3.0031875442536762e-05,
+      "loss": 0.2317,
+      "step": 4909
+    },
+    {
+      "epoch": 3.69589762890478,
+      "grad_norm": 0.23877963423728943,
+      "learning_rate": 3.00244819485381e-05,
+      "loss": 0.2311,
+      "step": 4910
+    },
+    {
+      "epoch": 3.6966503575461047,
+      "grad_norm": 0.24891892075538635,
+      "learning_rate": 3.0017088178465886e-05,
+      "loss": 0.2177,
+      "step": 4911
+    },
+    {
+      "epoch": 3.697403086187429,
+      "grad_norm": 0.21277309954166412,
+      "learning_rate": 3.0009694133128692e-05,
+      "loss": 0.2205,
+      "step": 4912
+    },
+    {
+      "epoch": 3.698155814828754,
+      "grad_norm": 0.25471729040145874,
+      "learning_rate": 3.0002299813335123e-05,
+      "loss": 0.2521,
+      "step": 4913
+    },
+    {
+      "epoch": 3.698908543470079,
+      "grad_norm": 0.23714998364448547,
+      "learning_rate": 2.999490521989381e-05,
+      "loss": 0.247,
+      "step": 4914
+    },
+    {
+      "epoch": 3.699661272111404,
+      "grad_norm": 0.25781306624412537,
+      "learning_rate": 2.99875103536134e-05,
+      "loss": 0.3442,
+      "step": 4915
+    },
+    {
+      "epoch": 3.700414000752729,
+      "grad_norm": 0.2195185273885727,
+      "learning_rate": 2.998011521530261e-05,
+      "loss": 0.3072,
+      "step": 4916
+    },
+    {
+      "epoch": 3.7011667293940533,
+      "grad_norm": 0.2890910506248474,
+      "learning_rate": 2.997271980577015e-05,
+      "loss": 0.2925,
+      "step": 4917
+    },
+    {
+      "epoch": 3.7019194580353783,
+      "grad_norm": 0.2419637143611908,
+      "learning_rate": 2.996532412582476e-05,
+      "loss": 0.25,
+      "step": 4918
+    },
+    {
+      "epoch": 3.702672186676703,
+      "grad_norm": 0.2434101700782776,
+      "learning_rate": 2.995792817627524e-05,
+      "loss": 0.1982,
+      "step": 4919
+    },
+    {
+      "epoch": 3.7034249153180276,
+      "grad_norm": 0.224112868309021,
+      "learning_rate": 2.9950531957930382e-05,
+      "loss": 0.2325,
+      "step": 4920
+    },
+    {
+      "epoch": 3.7041776439593526,
+      "grad_norm": 0.211380273103714,
+      "learning_rate": 2.9943135471599038e-05,
+      "loss": 0.3208,
+      "step": 4921
+    },
+    {
+      "epoch": 3.7049303726006775,
+      "grad_norm": 0.2249416708946228,
+      "learning_rate": 2.993573871809008e-05,
+      "loss": 0.2901,
+      "step": 4922
+    },
+    {
+      "epoch": 3.7056831012420024,
+      "grad_norm": 0.23972856998443604,
+      "learning_rate": 2.9928341698212387e-05,
+      "loss": 0.2331,
+      "step": 4923
+    },
+    {
+      "epoch": 3.7064358298833273,
+      "grad_norm": 0.24161523580551147,
+      "learning_rate": 2.9920944412774894e-05,
+      "loss": 0.2416,
+      "step": 4924
+    },
+    {
+      "epoch": 3.707188558524652,
+      "grad_norm": 0.2386944442987442,
+      "learning_rate": 2.9913546862586566e-05,
+      "loss": 0.238,
+      "step": 4925
+    },
+    {
+      "epoch": 3.7079412871659767,
+      "grad_norm": 0.2309044897556305,
+      "learning_rate": 2.990614904845638e-05,
+      "loss": 0.2073,
+      "step": 4926
+    },
+    {
+      "epoch": 3.7086940158073016,
+      "grad_norm": 0.20817400515079498,
+      "learning_rate": 2.989875097119335e-05,
+      "loss": 0.272,
+      "step": 4927
+    },
+    {
+      "epoch": 3.709446744448626,
+      "grad_norm": 0.19363300502300262,
+      "learning_rate": 2.9891352631606524e-05,
+      "loss": 0.216,
+      "step": 4928
+    },
+    {
+      "epoch": 3.710199473089951,
+      "grad_norm": 0.21729417145252228,
+      "learning_rate": 2.9883954030504968e-05,
+      "loss": 0.2866,
+      "step": 4929
+    },
+    {
+      "epoch": 3.710952201731276,
+      "grad_norm": 0.20183777809143066,
+      "learning_rate": 2.9876555168697784e-05,
+      "loss": 0.2719,
+      "step": 4930
+    },
+    {
+      "epoch": 3.7117049303726004,
+      "grad_norm": 0.2520635426044464,
+      "learning_rate": 2.9869156046994094e-05,
+      "loss": 0.2304,
+      "step": 4931
+    },
+    {
+      "epoch": 3.7124576590139253,
+      "grad_norm": 0.23991979658603668,
+      "learning_rate": 2.9861756666203074e-05,
+      "loss": 0.213,
+      "step": 4932
+    },
+    {
+      "epoch": 3.7132103876552502,
+      "grad_norm": 0.23540598154067993,
+      "learning_rate": 2.985435702713389e-05,
+      "loss": 0.2093,
+      "step": 4933
+    },
+    {
+      "epoch": 3.713963116296575,
+      "grad_norm": 0.282291054725647,
+      "learning_rate": 2.9846957130595753e-05,
+      "loss": 0.297,
+      "step": 4934
+    },
+    {
+      "epoch": 3.7147158449379,
+      "grad_norm": 0.19732174277305603,
+      "learning_rate": 2.983955697739792e-05,
+      "loss": 0.2316,
+      "step": 4935
+    },
+    {
+      "epoch": 3.7154685735792246,
+      "grad_norm": 0.3023718595504761,
+      "learning_rate": 2.983215656834965e-05,
+      "loss": 0.2275,
+      "step": 4936
+    },
+    {
+      "epoch": 3.7162213022205495,
+      "grad_norm": 0.20999909937381744,
+      "learning_rate": 2.9824755904260245e-05,
+      "loss": 0.1996,
+      "step": 4937
+    },
+    {
+      "epoch": 3.7169740308618744,
+      "grad_norm": 0.21586467325687408,
+      "learning_rate": 2.9817354985939033e-05,
+      "loss": 0.2035,
+      "step": 4938
+    },
+    {
+      "epoch": 3.717726759503199,
+      "grad_norm": 0.18120795488357544,
+      "learning_rate": 2.980995381419537e-05,
+      "loss": 0.2229,
+      "step": 4939
+    },
+    {
+      "epoch": 3.718479488144524,
+      "grad_norm": 0.20566239953041077,
+      "learning_rate": 2.9802552389838617e-05,
+      "loss": 0.2305,
+      "step": 4940
+    },
+    {
+      "epoch": 3.7192322167858487,
+      "grad_norm": 0.20658445358276367,
+      "learning_rate": 2.97951507136782e-05,
+      "loss": 0.196,
+      "step": 4941
+    },
+    {
+      "epoch": 3.7199849454271736,
+      "grad_norm": 0.219600647687912,
+      "learning_rate": 2.9787748786523557e-05,
+      "loss": 0.195,
+      "step": 4942
+    },
+    {
+      "epoch": 3.7207376740684985,
+      "grad_norm": 0.2536700963973999,
+      "learning_rate": 2.978034660918414e-05,
+      "loss": 0.2379,
+      "step": 4943
+    },
+    {
+      "epoch": 3.721490402709823,
+      "grad_norm": 0.2126384675502777,
+      "learning_rate": 2.9772944182469455e-05,
+      "loss": 0.1578,
+      "step": 4944
+    },
+    {
+      "epoch": 3.722243131351148,
+      "grad_norm": 0.2201615869998932,
+      "learning_rate": 2.9765541507189004e-05,
+      "loss": 0.2093,
+      "step": 4945
+    },
+    {
+      "epoch": 3.722995859992473,
+      "grad_norm": 0.2634842097759247,
+      "learning_rate": 2.9758138584152335e-05,
+      "loss": 0.2789,
+      "step": 4946
+    },
+    {
+      "epoch": 3.7237485886337973,
+      "grad_norm": 0.2326439917087555,
+      "learning_rate": 2.9750735414169033e-05,
+      "loss": 0.2195,
+      "step": 4947
+    },
+    {
+      "epoch": 3.7245013172751222,
+      "grad_norm": 0.17543047666549683,
+      "learning_rate": 2.9743331998048695e-05,
+      "loss": 0.1232,
+      "step": 4948
+    },
+    {
+      "epoch": 3.725254045916447,
+      "grad_norm": 0.20578692853450775,
+      "learning_rate": 2.9735928336600933e-05,
+      "loss": 0.166,
+      "step": 4949
+    },
+    {
+      "epoch": 3.726006774557772,
+      "grad_norm": 0.22685043513774872,
+      "learning_rate": 2.9728524430635417e-05,
+      "loss": 0.2356,
+      "step": 4950
+    },
+    {
+      "epoch": 3.726759503199097,
+      "grad_norm": 0.19269147515296936,
+      "learning_rate": 2.972112028096181e-05,
+      "loss": 0.1907,
+      "step": 4951
+    },
+    {
+      "epoch": 3.7275122318404215,
+      "grad_norm": 0.1811269074678421,
+      "learning_rate": 2.9713715888389843e-05,
+      "loss": 0.2113,
+      "step": 4952
+    },
+    {
+      "epoch": 3.7282649604817464,
+      "grad_norm": 0.28195124864578247,
+      "learning_rate": 2.9706311253729218e-05,
+      "loss": 0.2125,
+      "step": 4953
+    },
+    {
+      "epoch": 3.7290176891230713,
+      "grad_norm": 0.2503829598426819,
+      "learning_rate": 2.969890637778972e-05,
+      "loss": 0.3006,
+      "step": 4954
+    },
+    {
+      "epoch": 3.7297704177643958,
+      "grad_norm": 0.24444866180419922,
+      "learning_rate": 2.969150126138113e-05,
+      "loss": 0.2648,
+      "step": 4955
+    },
+    {
+      "epoch": 3.7305231464057207,
+      "grad_norm": 0.24843493103981018,
+      "learning_rate": 2.968409590531325e-05,
+      "loss": 0.2809,
+      "step": 4956
+    },
+    {
+      "epoch": 3.7312758750470456,
+      "grad_norm": 0.20699986815452576,
+      "learning_rate": 2.9676690310395934e-05,
+      "loss": 0.1903,
+      "step": 4957
+    },
+    {
+      "epoch": 3.73202860368837,
+      "grad_norm": 0.2517138421535492,
+      "learning_rate": 2.9669284477439035e-05,
+      "loss": 0.2405,
+      "step": 4958
+    },
+    {
+      "epoch": 3.732781332329695,
+      "grad_norm": 0.19565045833587646,
+      "learning_rate": 2.9661878407252443e-05,
+      "loss": 0.1414,
+      "step": 4959
+    },
+    {
+      "epoch": 3.73353406097102,
+      "grad_norm": 0.2269783318042755,
+      "learning_rate": 2.965447210064608e-05,
+      "loss": 0.1451,
+      "step": 4960
+    },
+    {
+      "epoch": 3.734286789612345,
+      "grad_norm": 0.25252658128738403,
+      "learning_rate": 2.9647065558429886e-05,
+      "loss": 0.327,
+      "step": 4961
+    },
+    {
+      "epoch": 3.7350395182536698,
+      "grad_norm": 0.27433717250823975,
+      "learning_rate": 2.9639658781413832e-05,
+      "loss": 0.2618,
+      "step": 4962
+    },
+    {
+      "epoch": 3.7357922468949942,
+      "grad_norm": 0.32026538252830505,
+      "learning_rate": 2.9632251770407914e-05,
+      "loss": 0.3066,
+      "step": 4963
+    },
+    {
+      "epoch": 3.736544975536319,
+      "grad_norm": 0.2294379472732544,
+      "learning_rate": 2.9624844526222144e-05,
+      "loss": 0.2325,
+      "step": 4964
+    },
+    {
+      "epoch": 3.737297704177644,
+      "grad_norm": 0.24915535748004913,
+      "learning_rate": 2.961743704966658e-05,
+      "loss": 0.2203,
+      "step": 4965
+    },
+    {
+      "epoch": 3.7380504328189685,
+      "grad_norm": 0.219607874751091,
+      "learning_rate": 2.961002934155127e-05,
+      "loss": 0.2433,
+      "step": 4966
+    },
+    {
+      "epoch": 3.7388031614602935,
+      "grad_norm": 0.2994817793369293,
+      "learning_rate": 2.960262140268634e-05,
+      "loss": 0.1823,
+      "step": 4967
+    },
+    {
+      "epoch": 3.7395558901016184,
+      "grad_norm": 0.21685168147087097,
+      "learning_rate": 2.959521323388188e-05,
+      "loss": 0.2357,
+      "step": 4968
+    },
+    {
+      "epoch": 3.7403086187429433,
+      "grad_norm": 0.2582666575908661,
+      "learning_rate": 2.9587804835948053e-05,
+      "loss": 0.2788,
+      "step": 4969
+    },
+    {
+      "epoch": 3.741061347384268,
+      "grad_norm": 0.22557054460048676,
+      "learning_rate": 2.9580396209695032e-05,
+      "loss": 0.2782,
+      "step": 4970
+    },
+    {
+      "epoch": 3.7418140760255927,
+      "grad_norm": 0.2103082686662674,
+      "learning_rate": 2.9572987355933013e-05,
+      "loss": 0.2808,
+      "step": 4971
+    },
+    {
+      "epoch": 3.7425668046669176,
+      "grad_norm": 0.20777970552444458,
+      "learning_rate": 2.9565578275472204e-05,
+      "loss": 0.1667,
+      "step": 4972
+    },
+    {
+      "epoch": 3.7433195333082425,
+      "grad_norm": 0.25421443581581116,
+      "learning_rate": 2.9558168969122864e-05,
+      "loss": 0.2335,
+      "step": 4973
+    },
+    {
+      "epoch": 3.744072261949567,
+      "grad_norm": 0.23191359639167786,
+      "learning_rate": 2.955075943769526e-05,
+      "loss": 0.2212,
+      "step": 4974
+    },
+    {
+      "epoch": 3.744824990590892,
+      "grad_norm": 0.24743898212909698,
+      "learning_rate": 2.9543349681999687e-05,
+      "loss": 0.2646,
+      "step": 4975
+    },
+    {
+      "epoch": 3.745577719232217,
+      "grad_norm": 0.20150445401668549,
+      "learning_rate": 2.9535939702846465e-05,
+      "loss": 0.2051,
+      "step": 4976
+    },
+    {
+      "epoch": 3.7463304478735417,
+      "grad_norm": 0.2648310959339142,
+      "learning_rate": 2.9528529501045944e-05,
+      "loss": 0.3062,
+      "step": 4977
+    },
+    {
+      "epoch": 3.7470831765148667,
+      "grad_norm": 0.234842449426651,
+      "learning_rate": 2.9521119077408477e-05,
+      "loss": 0.2557,
+      "step": 4978
+    },
+    {
+      "epoch": 3.747835905156191,
+      "grad_norm": 0.2285090684890747,
+      "learning_rate": 2.9513708432744474e-05,
+      "loss": 0.2227,
+      "step": 4979
+    },
+    {
+      "epoch": 3.748588633797516,
+      "grad_norm": 0.22200258076190948,
+      "learning_rate": 2.9506297567864343e-05,
+      "loss": 0.2404,
+      "step": 4980
+    },
+    {
+      "epoch": 3.749341362438841,
+      "grad_norm": 0.23074108362197876,
+      "learning_rate": 2.949888648357852e-05,
+      "loss": 0.2364,
+      "step": 4981
+    },
+    {
+      "epoch": 3.7500940910801654,
+      "grad_norm": 0.230948805809021,
+      "learning_rate": 2.949147518069748e-05,
+      "loss": 0.2519,
+      "step": 4982
+    },
+    {
+      "epoch": 3.7508468197214904,
+      "grad_norm": 0.18817424774169922,
+      "learning_rate": 2.9484063660031718e-05,
+      "loss": 0.1974,
+      "step": 4983
+    },
+    {
+      "epoch": 3.7515995483628153,
+      "grad_norm": 0.26656481623649597,
+      "learning_rate": 2.9476651922391728e-05,
+      "loss": 0.2384,
+      "step": 4984
+    },
+    {
+      "epoch": 3.7523522770041398,
+      "grad_norm": 0.23164458572864532,
+      "learning_rate": 2.9469239968588053e-05,
+      "loss": 0.1986,
+      "step": 4985
+    },
+    {
+      "epoch": 3.7531050056454647,
+      "grad_norm": 0.22538228332996368,
+      "learning_rate": 2.9461827799431265e-05,
+      "loss": 0.2871,
+      "step": 4986
+    },
+    {
+      "epoch": 3.7538577342867896,
+      "grad_norm": 0.22592346370220184,
+      "learning_rate": 2.9454415415731927e-05,
+      "loss": 0.2397,
+      "step": 4987
+    },
+    {
+      "epoch": 3.7546104629281145,
+      "grad_norm": 0.22934392094612122,
+      "learning_rate": 2.944700281830066e-05,
+      "loss": 0.2165,
+      "step": 4988
+    },
+    {
+      "epoch": 3.7553631915694394,
+      "grad_norm": 0.21534451842308044,
+      "learning_rate": 2.9439590007948098e-05,
+      "loss": 0.223,
+      "step": 4989
+    },
+    {
+      "epoch": 3.756115920210764,
+      "grad_norm": 0.22100253403186798,
+      "learning_rate": 2.9432176985484887e-05,
+      "loss": 0.2436,
+      "step": 4990
+    },
+    {
+      "epoch": 3.756868648852089,
+      "grad_norm": 0.24248898029327393,
+      "learning_rate": 2.9424763751721702e-05,
+      "loss": 0.2897,
+      "step": 4991
+    },
+    {
+      "epoch": 3.7576213774934137,
+      "grad_norm": 0.1988062560558319,
+      "learning_rate": 2.9417350307469254e-05,
+      "loss": 0.3109,
+      "step": 4992
+    },
+    {
+      "epoch": 3.758374106134738,
+      "grad_norm": 0.21245701611042023,
+      "learning_rate": 2.9409936653538255e-05,
+      "loss": 0.2411,
+      "step": 4993
+    },
+    {
+      "epoch": 3.759126834776063,
+      "grad_norm": 0.22481264173984528,
+      "learning_rate": 2.940252279073945e-05,
+      "loss": 0.3452,
+      "step": 4994
+    },
+    {
+      "epoch": 3.759879563417388,
+      "grad_norm": 0.21850720047950745,
+      "learning_rate": 2.9395108719883623e-05,
+      "loss": 0.2253,
+      "step": 4995
+    },
+    {
+      "epoch": 3.760632292058713,
+      "grad_norm": 0.22655008733272552,
+      "learning_rate": 2.9387694441781553e-05,
+      "loss": 0.2493,
+      "step": 4996
+    },
+    {
+      "epoch": 3.761385020700038,
+      "grad_norm": 0.22760231792926788,
+      "learning_rate": 2.938027995724405e-05,
+      "loss": 0.2142,
+      "step": 4997
+    },
+    {
+      "epoch": 3.7621377493413624,
+      "grad_norm": 0.20316734910011292,
+      "learning_rate": 2.9372865267081966e-05,
+      "loss": 0.2437,
+      "step": 4998
+    },
+    {
+      "epoch": 3.7628904779826873,
+      "grad_norm": 0.22929130494594574,
+      "learning_rate": 2.9365450372106146e-05,
+      "loss": 0.3486,
+      "step": 4999
+    },
+    {
+      "epoch": 3.763643206624012,
+      "grad_norm": 0.2081385999917984,
+      "learning_rate": 2.9358035273127483e-05,
+      "loss": 0.3027,
+      "step": 5000
+    },
+    {
+      "epoch": 3.763643206624012,
+      "eval_loss": 0.25287604331970215,
+      "eval_runtime": 452.5311,
+      "eval_samples_per_second": 21.274,
+      "eval_steps_per_second": 0.665,
+      "step": 5000
+    },
+    {
+      "epoch": 3.7643959352653367,
+      "grad_norm": 0.20297525823116302,
+      "learning_rate": 2.9350619970956868e-05,
+      "loss": 0.2003,
+      "step": 5001
+    },
+    {
+      "epoch": 3.7651486639066616,
+      "grad_norm": 0.2140226513147354,
+      "learning_rate": 2.9343204466405243e-05,
+      "loss": 0.2299,
+      "step": 5002
+    },
+    {
+      "epoch": 3.7659013925479865,
+      "grad_norm": 0.23281987011432648,
+      "learning_rate": 2.933578876028355e-05,
+      "loss": 0.293,
+      "step": 5003
+    },
+    {
+      "epoch": 3.766654121189311,
+      "grad_norm": 0.20945844054222107,
+      "learning_rate": 2.9328372853402752e-05,
+      "loss": 0.2312,
+      "step": 5004
+    },
+    {
+      "epoch": 3.767406849830636,
+      "grad_norm": 0.20274145901203156,
+      "learning_rate": 2.9320956746573853e-05,
+      "loss": 0.2018,
+      "step": 5005
+    },
+    {
+      "epoch": 3.768159578471961,
+      "grad_norm": 0.23777367174625397,
+      "learning_rate": 2.9313540440607863e-05,
+      "loss": 0.2556,
+      "step": 5006
+    },
+    {
+      "epoch": 3.7689123071132857,
+      "grad_norm": 0.2636973261833191,
+      "learning_rate": 2.930612393631581e-05,
+      "loss": 0.2916,
+      "step": 5007
+    },
+    {
+      "epoch": 3.7696650357546106,
+      "grad_norm": 0.2426040917634964,
+      "learning_rate": 2.929870723450876e-05,
+      "loss": 0.1978,
+      "step": 5008
+    },
+    {
+      "epoch": 3.770417764395935,
+      "grad_norm": 0.20711366832256317,
+      "learning_rate": 2.9291290335997802e-05,
+      "loss": 0.2407,
+      "step": 5009
+    },
+    {
+      "epoch": 3.77117049303726,
+      "grad_norm": 0.21846096217632294,
+      "learning_rate": 2.9283873241594005e-05,
+      "loss": 0.2722,
+      "step": 5010
+    },
+    {
+      "epoch": 3.771923221678585,
+      "grad_norm": 0.19396333396434784,
+      "learning_rate": 2.9276455952108526e-05,
+      "loss": 0.1364,
+      "step": 5011
+    },
+    {
+      "epoch": 3.7726759503199094,
+      "grad_norm": 0.20704659819602966,
+      "learning_rate": 2.92690384683525e-05,
+      "loss": 0.2727,
+      "step": 5012
+    },
+    {
+      "epoch": 3.7734286789612344,
+      "grad_norm": 0.2271822988986969,
+      "learning_rate": 2.926162079113708e-05,
+      "loss": 0.2235,
+      "step": 5013
+    },
+    {
+      "epoch": 3.7741814076025593,
+      "grad_norm": 0.230509415268898,
+      "learning_rate": 2.925420292127346e-05,
+      "loss": 0.1955,
+      "step": 5014
+    },
+    {
+      "epoch": 3.774934136243884,
+      "grad_norm": 0.2288891077041626,
+      "learning_rate": 2.9246784859572845e-05,
+      "loss": 0.2334,
+      "step": 5015
+    },
+    {
+      "epoch": 3.775686864885209,
+      "grad_norm": 0.208413228392601,
+      "learning_rate": 2.923936660684647e-05,
+      "loss": 0.1719,
+      "step": 5016
+    },
+    {
+      "epoch": 3.7764395935265336,
+      "grad_norm": 0.20778211951255798,
+      "learning_rate": 2.9231948163905566e-05,
+      "loss": 0.226,
+      "step": 5017
+    },
+    {
+      "epoch": 3.7771923221678585,
+      "grad_norm": 0.24410541355609894,
+      "learning_rate": 2.9224529531561422e-05,
+      "loss": 0.3376,
+      "step": 5018
+    },
+    {
+      "epoch": 3.7779450508091834,
+      "grad_norm": 0.2200707495212555,
+      "learning_rate": 2.9217110710625327e-05,
+      "loss": 0.2388,
+      "step": 5019
+    },
+    {
+      "epoch": 3.778697779450508,
+      "grad_norm": 0.236529141664505,
+      "learning_rate": 2.920969170190857e-05,
+      "loss": 0.2902,
+      "step": 5020
+    },
+    {
+      "epoch": 3.779450508091833,
+      "grad_norm": 0.25804251432418823,
+      "learning_rate": 2.9202272506222512e-05,
+      "loss": 0.2543,
+      "step": 5021
+    },
+    {
+      "epoch": 3.7802032367331577,
+      "grad_norm": 0.22098176181316376,
+      "learning_rate": 2.919485312437849e-05,
+      "loss": 0.2748,
+      "step": 5022
+    },
+    {
+      "epoch": 3.7809559653744826,
+      "grad_norm": 0.20779988169670105,
+      "learning_rate": 2.9187433557187876e-05,
+      "loss": 0.2649,
+      "step": 5023
+    },
+    {
+      "epoch": 3.7817086940158076,
+      "grad_norm": 0.19098080694675446,
+      "learning_rate": 2.9180013805462065e-05,
+      "loss": 0.2733,
+      "step": 5024
+    },
+    {
+      "epoch": 3.782461422657132,
+      "grad_norm": 0.27111443877220154,
+      "learning_rate": 2.9172593870012477e-05,
+      "loss": 0.2354,
+      "step": 5025
+    },
+    {
+      "epoch": 3.783214151298457,
+      "grad_norm": 0.19564269483089447,
+      "learning_rate": 2.9165173751650525e-05,
+      "loss": 0.0992,
+      "step": 5026
+    },
+    {
+      "epoch": 3.783966879939782,
+      "grad_norm": 0.19716283679008484,
+      "learning_rate": 2.9157753451187687e-05,
+      "loss": 0.2022,
+      "step": 5027
+    },
+    {
+      "epoch": 3.7847196085811063,
+      "grad_norm": 0.2327619343996048,
+      "learning_rate": 2.9150332969435417e-05,
+      "loss": 0.211,
+      "step": 5028
+    },
+    {
+      "epoch": 3.7854723372224313,
+      "grad_norm": 0.2564774751663208,
+      "learning_rate": 2.9142912307205217e-05,
+      "loss": 0.1972,
+      "step": 5029
+    },
+    {
+      "epoch": 3.786225065863756,
+      "grad_norm": 0.30185091495513916,
+      "learning_rate": 2.91354914653086e-05,
+      "loss": 0.3712,
+      "step": 5030
+    },
+    {
+      "epoch": 3.7869777945050807,
+      "grad_norm": 0.22205087542533875,
+      "learning_rate": 2.912807044455709e-05,
+      "loss": 0.1422,
+      "step": 5031
+    },
+    {
+      "epoch": 3.7877305231464056,
+      "grad_norm": 0.23346519470214844,
+      "learning_rate": 2.912064924576225e-05,
+      "loss": 0.2448,
+      "step": 5032
+    },
+    {
+      "epoch": 3.7884832517877305,
+      "grad_norm": 0.2858509123325348,
+      "learning_rate": 2.9113227869735644e-05,
+      "loss": 0.2723,
+      "step": 5033
+    },
+    {
+      "epoch": 3.7892359804290554,
+      "grad_norm": 0.22153380513191223,
+      "learning_rate": 2.9105806317288864e-05,
+      "loss": 0.1619,
+      "step": 5034
+    },
+    {
+      "epoch": 3.7899887090703803,
+      "grad_norm": 0.22411642968654633,
+      "learning_rate": 2.909838458923351e-05,
+      "loss": 0.2863,
+      "step": 5035
+    },
+    {
+      "epoch": 3.790741437711705,
+      "grad_norm": 0.24853326380252838,
+      "learning_rate": 2.909096268638123e-05,
+      "loss": 0.2316,
+      "step": 5036
+    },
+    {
+      "epoch": 3.7914941663530297,
+      "grad_norm": 0.22622482478618622,
+      "learning_rate": 2.9083540609543662e-05,
+      "loss": 0.23,
+      "step": 5037
+    },
+    {
+      "epoch": 3.7922468949943546,
+      "grad_norm": 0.23589874804019928,
+      "learning_rate": 2.9076118359532472e-05,
+      "loss": 0.1987,
+      "step": 5038
+    },
+    {
+      "epoch": 3.792999623635679,
+      "grad_norm": 0.2505094110965729,
+      "learning_rate": 2.9068695937159345e-05,
+      "loss": 0.2817,
+      "step": 5039
+    },
+    {
+      "epoch": 3.793752352277004,
+      "grad_norm": 0.23291698098182678,
+      "learning_rate": 2.9061273343235994e-05,
+      "loss": 0.2764,
+      "step": 5040
+    },
+    {
+      "epoch": 3.794505080918329,
+      "grad_norm": 0.21750196814537048,
+      "learning_rate": 2.9053850578574138e-05,
+      "loss": 0.244,
+      "step": 5041
+    },
+    {
+      "epoch": 3.795257809559654,
+      "grad_norm": 0.22086776793003082,
+      "learning_rate": 2.9046427643985513e-05,
+      "loss": 0.1713,
+      "step": 5042
+    },
+    {
+      "epoch": 3.796010538200979,
+      "grad_norm": 0.20527967810630798,
+      "learning_rate": 2.903900454028189e-05,
+      "loss": 0.2469,
+      "step": 5043
+    },
+    {
+      "epoch": 3.7967632668423033,
+      "grad_norm": 0.2774546444416046,
+      "learning_rate": 2.9031581268275043e-05,
+      "loss": 0.2111,
+      "step": 5044
+    },
+    {
+      "epoch": 3.797515995483628,
+      "grad_norm": 0.2190197855234146,
+      "learning_rate": 2.9024157828776766e-05,
+      "loss": 0.2696,
+      "step": 5045
+    },
+    {
+      "epoch": 3.798268724124953,
+      "grad_norm": 0.24375464022159576,
+      "learning_rate": 2.901673422259889e-05,
+      "loss": 0.2358,
+      "step": 5046
+    },
+    {
+      "epoch": 3.7990214527662776,
+      "grad_norm": 0.20910973846912384,
+      "learning_rate": 2.9009310450553228e-05,
+      "loss": 0.2333,
+      "step": 5047
+    },
+    {
+      "epoch": 3.7997741814076025,
+      "grad_norm": 0.25209105014801025,
+      "learning_rate": 2.900188651345166e-05,
+      "loss": 0.2329,
+      "step": 5048
+    },
+    {
+      "epoch": 3.8005269100489274,
+      "grad_norm": 0.2119540423154831,
+      "learning_rate": 2.8994462412106026e-05,
+      "loss": 0.1557,
+      "step": 5049
+    },
+    {
+      "epoch": 3.8012796386902523,
+      "grad_norm": 0.25448599457740784,
+      "learning_rate": 2.8987038147328238e-05,
+      "loss": 0.2807,
+      "step": 5050
+    },
+    {
+      "epoch": 3.8020323673315772,
+      "grad_norm": 0.2518323063850403,
+      "learning_rate": 2.8979613719930194e-05,
+      "loss": 0.195,
+      "step": 5051
+    },
+    {
+      "epoch": 3.8027850959729017,
+      "grad_norm": 0.31249821186065674,
+      "learning_rate": 2.8972189130723815e-05,
+      "loss": 0.4038,
+      "step": 5052
+    },
+    {
+      "epoch": 3.8035378246142266,
+      "grad_norm": 0.24685145914554596,
+      "learning_rate": 2.8964764380521046e-05,
+      "loss": 0.2317,
+      "step": 5053
+    },
+    {
+      "epoch": 3.8042905532555515,
+      "grad_norm": 0.2855522632598877,
+      "learning_rate": 2.8957339470133855e-05,
+      "loss": 0.2483,
+      "step": 5054
+    },
+    {
+      "epoch": 3.805043281896876,
+      "grad_norm": 0.30564913153648376,
+      "learning_rate": 2.8949914400374196e-05,
+      "loss": 0.2769,
+      "step": 5055
+    },
+    {
+      "epoch": 3.805796010538201,
+      "grad_norm": 0.24688883125782013,
+      "learning_rate": 2.894248917205409e-05,
+      "loss": 0.2554,
+      "step": 5056
+    },
+    {
+      "epoch": 3.806548739179526,
+      "grad_norm": 0.26991334557533264,
+      "learning_rate": 2.8935063785985533e-05,
+      "loss": 0.2173,
+      "step": 5057
+    },
+    {
+      "epoch": 3.8073014678208503,
+      "grad_norm": 0.2707289457321167,
+      "learning_rate": 2.8927638242980554e-05,
+      "loss": 0.343,
+      "step": 5058
+    },
+    {
+      "epoch": 3.8080541964621752,
+      "grad_norm": 0.23455655574798584,
+      "learning_rate": 2.892021254385121e-05,
+      "loss": 0.281,
+      "step": 5059
+    },
+    {
+      "epoch": 3.8088069251035,
+      "grad_norm": 0.29732874035835266,
+      "learning_rate": 2.8912786689409556e-05,
+      "loss": 0.2429,
+      "step": 5060
+    },
+    {
+      "epoch": 3.809559653744825,
+      "grad_norm": 0.2976124882698059,
+      "learning_rate": 2.890536068046767e-05,
+      "loss": 0.3489,
+      "step": 5061
+    },
+    {
+      "epoch": 3.81031238238615,
+      "grad_norm": 0.3079575002193451,
+      "learning_rate": 2.8897934517837654e-05,
+      "loss": 0.2199,
+      "step": 5062
+    },
+    {
+      "epoch": 3.8110651110274745,
+      "grad_norm": 0.23774851858615875,
+      "learning_rate": 2.8890508202331623e-05,
+      "loss": 0.149,
+      "step": 5063
+    },
+    {
+      "epoch": 3.8118178396687994,
+      "grad_norm": 0.2558790147304535,
+      "learning_rate": 2.88830817347617e-05,
+      "loss": 0.1894,
+      "step": 5064
+    },
+    {
+      "epoch": 3.8125705683101243,
+      "grad_norm": 0.24791587889194489,
+      "learning_rate": 2.887565511594005e-05,
+      "loss": 0.2539,
+      "step": 5065
+    },
+    {
+      "epoch": 3.813323296951449,
+      "grad_norm": 0.29002317786216736,
+      "learning_rate": 2.8868228346678817e-05,
+      "loss": 0.2236,
+      "step": 5066
+    },
+    {
+      "epoch": 3.8140760255927737,
+      "grad_norm": 0.20201194286346436,
+      "learning_rate": 2.8860801427790184e-05,
+      "loss": 0.1622,
+      "step": 5067
+    },
+    {
+      "epoch": 3.8148287542340986,
+      "grad_norm": 0.2225465476512909,
+      "learning_rate": 2.8853374360086367e-05,
+      "loss": 0.1954,
+      "step": 5068
+    },
+    {
+      "epoch": 3.8155814828754235,
+      "grad_norm": 0.21178816258907318,
+      "learning_rate": 2.8845947144379554e-05,
+      "loss": 0.2281,
+      "step": 5069
+    },
+    {
+      "epoch": 3.8163342115167485,
+      "grad_norm": 0.2161841243505478,
+      "learning_rate": 2.8838519781481983e-05,
+      "loss": 0.2774,
+      "step": 5070
+    },
+    {
+      "epoch": 3.817086940158073,
+      "grad_norm": 0.1842058151960373,
+      "learning_rate": 2.8831092272205907e-05,
+      "loss": 0.1959,
+      "step": 5071
+    },
+    {
+      "epoch": 3.817839668799398,
+      "grad_norm": 0.22901754081249237,
+      "learning_rate": 2.8823664617363577e-05,
+      "loss": 0.1789,
+      "step": 5072
+    },
+    {
+      "epoch": 3.8185923974407228,
+      "grad_norm": 0.20832528173923492,
+      "learning_rate": 2.8816236817767277e-05,
+      "loss": 0.2257,
+      "step": 5073
+    },
+    {
+      "epoch": 3.8193451260820472,
+      "grad_norm": 0.21220485866069794,
+      "learning_rate": 2.8808808874229288e-05,
+      "loss": 0.1493,
+      "step": 5074
+    },
+    {
+      "epoch": 3.820097854723372,
+      "grad_norm": 0.19573749601840973,
+      "learning_rate": 2.8801380787561944e-05,
+      "loss": 0.2465,
+      "step": 5075
+    },
+    {
+      "epoch": 3.820850583364697,
+      "grad_norm": 0.20319585502147675,
+      "learning_rate": 2.879395255857755e-05,
+      "loss": 0.2856,
+      "step": 5076
+    },
+    {
+      "epoch": 3.821603312006022,
+      "grad_norm": 0.23471473157405853,
+      "learning_rate": 2.878652418808844e-05,
+      "loss": 0.3129,
+      "step": 5077
+    },
+    {
+      "epoch": 3.822356040647347,
+      "grad_norm": 0.2766118049621582,
+      "learning_rate": 2.8779095676906993e-05,
+      "loss": 0.233,
+      "step": 5078
+    },
+    {
+      "epoch": 3.8231087692886714,
+      "grad_norm": 0.21900545060634613,
+      "learning_rate": 2.8771667025845554e-05,
+      "loss": 0.2175,
+      "step": 5079
+    },
+    {
+      "epoch": 3.8238614979299963,
+      "grad_norm": 0.23307707905769348,
+      "learning_rate": 2.8764238235716522e-05,
+      "loss": 0.216,
+      "step": 5080
+    },
+    {
+      "epoch": 3.824614226571321,
+      "grad_norm": 0.21210472285747528,
+      "learning_rate": 2.8756809307332305e-05,
+      "loss": 0.2009,
+      "step": 5081
+    },
+    {
+      "epoch": 3.8253669552126457,
+      "grad_norm": 0.18869878351688385,
+      "learning_rate": 2.8749380241505307e-05,
+      "loss": 0.1787,
+      "step": 5082
+    },
+    {
+      "epoch": 3.8261196838539706,
+      "grad_norm": 0.2082013636827469,
+      "learning_rate": 2.8741951039047964e-05,
+      "loss": 0.1778,
+      "step": 5083
+    },
+    {
+      "epoch": 3.8268724124952955,
+      "grad_norm": 0.1925409436225891,
+      "learning_rate": 2.8734521700772727e-05,
+      "loss": 0.1993,
+      "step": 5084
+    },
+    {
+      "epoch": 3.82762514113662,
+      "grad_norm": 0.19932378828525543,
+      "learning_rate": 2.8727092227492053e-05,
+      "loss": 0.2299,
+      "step": 5085
+    },
+    {
+      "epoch": 3.828377869777945,
+      "grad_norm": 0.210063636302948,
+      "learning_rate": 2.8719662620018412e-05,
+      "loss": 0.2677,
+      "step": 5086
+    },
+    {
+      "epoch": 3.82913059841927,
+      "grad_norm": 0.20391707122325897,
+      "learning_rate": 2.8712232879164302e-05,
+      "loss": 0.1412,
+      "step": 5087
+    },
+    {
+      "epoch": 3.8298833270605948,
+      "grad_norm": 0.30928945541381836,
+      "learning_rate": 2.8704803005742224e-05,
+      "loss": 0.2357,
+      "step": 5088
+    },
+    {
+      "epoch": 3.8306360557019197,
+      "grad_norm": 0.20575815439224243,
+      "learning_rate": 2.869737300056471e-05,
+      "loss": 0.1648,
+      "step": 5089
+    },
+    {
+      "epoch": 3.831388784343244,
+      "grad_norm": 0.26886075735092163,
+      "learning_rate": 2.868994286444428e-05,
+      "loss": 0.3022,
+      "step": 5090
+    },
+    {
+      "epoch": 3.832141512984569,
+      "grad_norm": 0.2509852647781372,
+      "learning_rate": 2.8682512598193483e-05,
+      "loss": 0.2445,
+      "step": 5091
+    },
+    {
+      "epoch": 3.832894241625894,
+      "grad_norm": 0.22419552505016327,
+      "learning_rate": 2.867508220262489e-05,
+      "loss": 0.2293,
+      "step": 5092
+    },
+    {
+      "epoch": 3.8336469702672185,
+      "grad_norm": 0.24295414984226227,
+      "learning_rate": 2.8667651678551076e-05,
+      "loss": 0.2449,
+      "step": 5093
+    },
+    {
+      "epoch": 3.8343996989085434,
+      "grad_norm": 0.2324046790599823,
+      "learning_rate": 2.866022102678463e-05,
+      "loss": 0.2679,
+      "step": 5094
+    },
+    {
+      "epoch": 3.8351524275498683,
+      "grad_norm": 0.20904166996479034,
+      "learning_rate": 2.8652790248138147e-05,
+      "loss": 0.2967,
+      "step": 5095
+    },
+    {
+      "epoch": 3.835905156191193,
+      "grad_norm": 0.200456902384758,
+      "learning_rate": 2.864535934342426e-05,
+      "loss": 0.2392,
+      "step": 5096
+    },
+    {
+      "epoch": 3.836657884832518,
+      "grad_norm": 0.21577133238315582,
+      "learning_rate": 2.8637928313455603e-05,
+      "loss": 0.235,
+      "step": 5097
+    },
+    {
+      "epoch": 3.8374106134738426,
+      "grad_norm": 0.19194909930229187,
+      "learning_rate": 2.8630497159044817e-05,
+      "loss": 0.2598,
+      "step": 5098
+    },
+    {
+      "epoch": 3.8381633421151675,
+      "grad_norm": 0.25362008810043335,
+      "learning_rate": 2.8623065881004557e-05,
+      "loss": 0.1702,
+      "step": 5099
+    },
+    {
+      "epoch": 3.8389160707564924,
+      "grad_norm": 0.20863516628742218,
+      "learning_rate": 2.86156344801475e-05,
+      "loss": 0.185,
+      "step": 5100
+    },
+    {
+      "epoch": 3.839668799397817,
+      "grad_norm": 0.20490995049476624,
+      "learning_rate": 2.860820295728634e-05,
+      "loss": 0.2485,
+      "step": 5101
+    },
+    {
+      "epoch": 3.840421528039142,
+      "grad_norm": 0.20513062179088593,
+      "learning_rate": 2.8600771313233766e-05,
+      "loss": 0.277,
+      "step": 5102
+    },
+    {
+      "epoch": 3.8411742566804667,
+      "grad_norm": 0.21700604259967804,
+      "learning_rate": 2.85933395488025e-05,
+      "loss": 0.2011,
+      "step": 5103
+    },
+    {
+      "epoch": 3.841926985321791,
+      "grad_norm": 0.24530494213104248,
+      "learning_rate": 2.858590766480526e-05,
+      "loss": 0.2349,
+      "step": 5104
+    },
+    {
+      "epoch": 3.842679713963116,
+      "grad_norm": 0.1805228590965271,
+      "learning_rate": 2.8578475662054795e-05,
+      "loss": 0.1585,
+      "step": 5105
+    },
+    {
+      "epoch": 3.843432442604441,
+      "grad_norm": 0.20177897810935974,
+      "learning_rate": 2.8571043541363845e-05,
+      "loss": 0.2314,
+      "step": 5106
+    },
+    {
+      "epoch": 3.844185171245766,
+      "grad_norm": 0.22272732853889465,
+      "learning_rate": 2.856361130354519e-05,
+      "loss": 0.1997,
+      "step": 5107
+    },
+    {
+      "epoch": 3.844937899887091,
+      "grad_norm": 0.23982009291648865,
+      "learning_rate": 2.8556178949411606e-05,
+      "loss": 0.2148,
+      "step": 5108
+    },
+    {
+      "epoch": 3.8456906285284154,
+      "grad_norm": 0.23234324157238007,
+      "learning_rate": 2.8548746479775873e-05,
+      "loss": 0.2294,
+      "step": 5109
+    },
+    {
+      "epoch": 3.8464433571697403,
+      "grad_norm": 0.23520337045192719,
+      "learning_rate": 2.85413138954508e-05,
+      "loss": 0.3116,
+      "step": 5110
+    },
+    {
+      "epoch": 3.847196085811065,
+      "grad_norm": 0.23817332088947296,
+      "learning_rate": 2.8533881197249206e-05,
+      "loss": 0.1449,
+      "step": 5111
+    },
+    {
+      "epoch": 3.8479488144523897,
+      "grad_norm": 0.20768016576766968,
+      "learning_rate": 2.8526448385983922e-05,
+      "loss": 0.2135,
+      "step": 5112
+    },
+    {
+      "epoch": 3.8487015430937146,
+      "grad_norm": 0.2392178177833557,
+      "learning_rate": 2.8519015462467792e-05,
+      "loss": 0.1842,
+      "step": 5113
+    },
+    {
+      "epoch": 3.8494542717350395,
+      "grad_norm": 0.24025419354438782,
+      "learning_rate": 2.8511582427513657e-05,
+      "loss": 0.3702,
+      "step": 5114
+    },
+    {
+      "epoch": 3.8502070003763644,
+      "grad_norm": 0.2597820460796356,
+      "learning_rate": 2.850414928193438e-05,
+      "loss": 0.2101,
+      "step": 5115
+    },
+    {
+      "epoch": 3.8509597290176893,
+      "grad_norm": 0.2206566333770752,
+      "learning_rate": 2.8496716026542853e-05,
+      "loss": 0.2718,
+      "step": 5116
+    },
+    {
+      "epoch": 3.851712457659014,
+      "grad_norm": 0.21958042681217194,
+      "learning_rate": 2.8489282662151963e-05,
+      "loss": 0.2821,
+      "step": 5117
+    },
+    {
+      "epoch": 3.8524651863003387,
+      "grad_norm": 0.23492427170276642,
+      "learning_rate": 2.8481849189574606e-05,
+      "loss": 0.1689,
+      "step": 5118
+    },
+    {
+      "epoch": 3.8532179149416637,
+      "grad_norm": 0.2574809789657593,
+      "learning_rate": 2.84744156096237e-05,
+      "loss": 0.2127,
+      "step": 5119
+    },
+    {
+      "epoch": 3.853970643582988,
+      "grad_norm": 0.21407221257686615,
+      "learning_rate": 2.8466981923112168e-05,
+      "loss": 0.2225,
+      "step": 5120
+    },
+    {
+      "epoch": 3.854723372224313,
+      "grad_norm": 0.21752016246318817,
+      "learning_rate": 2.845954813085295e-05,
+      "loss": 0.1244,
+      "step": 5121
+    },
+    {
+      "epoch": 3.855476100865638,
+      "grad_norm": 0.19059991836547852,
+      "learning_rate": 2.845211423365898e-05,
+      "loss": 0.1896,
+      "step": 5122
+    },
+    {
+      "epoch": 3.856228829506963,
+      "grad_norm": 0.19855523109436035,
+      "learning_rate": 2.8444680232343236e-05,
+      "loss": 0.1991,
+      "step": 5123
+    },
+    {
+      "epoch": 3.856981558148288,
+      "grad_norm": 0.2134282886981964,
+      "learning_rate": 2.8437246127718682e-05,
+      "loss": 0.1911,
+      "step": 5124
+    },
+    {
+      "epoch": 3.8577342867896123,
+      "grad_norm": 0.2184550166130066,
+      "learning_rate": 2.8429811920598298e-05,
+      "loss": 0.2636,
+      "step": 5125
+    },
+    {
+      "epoch": 3.858487015430937,
+      "grad_norm": 0.226718008518219,
+      "learning_rate": 2.8422377611795087e-05,
+      "loss": 0.3058,
+      "step": 5126
+    },
+    {
+      "epoch": 3.859239744072262,
+      "grad_norm": 0.1950312852859497,
+      "learning_rate": 2.841494320212204e-05,
+      "loss": 0.2589,
+      "step": 5127
+    },
+    {
+      "epoch": 3.8599924727135866,
+      "grad_norm": 0.2260953038930893,
+      "learning_rate": 2.8407508692392182e-05,
+      "loss": 0.2264,
+      "step": 5128
+    },
+    {
+      "epoch": 3.8607452013549115,
+      "grad_norm": 0.19346271455287933,
+      "learning_rate": 2.8400074083418537e-05,
+      "loss": 0.2448,
+      "step": 5129
+    },
+    {
+      "epoch": 3.8614979299962364,
+      "grad_norm": 0.254910409450531,
+      "learning_rate": 2.8392639376014148e-05,
+      "loss": 0.2484,
+      "step": 5130
+    },
+    {
+      "epoch": 3.862250658637561,
+      "grad_norm": 0.22099614143371582,
+      "learning_rate": 2.8385204570992057e-05,
+      "loss": 0.198,
+      "step": 5131
+    },
+    {
+      "epoch": 3.863003387278886,
+      "grad_norm": 0.22251619398593903,
+      "learning_rate": 2.8377769669165322e-05,
+      "loss": 0.1564,
+      "step": 5132
+    },
+    {
+      "epoch": 3.8637561159202107,
+      "grad_norm": 0.23072871565818787,
+      "learning_rate": 2.837033467134702e-05,
+      "loss": 0.2119,
+      "step": 5133
+    },
+    {
+      "epoch": 3.8645088445615356,
+      "grad_norm": 0.22055558860301971,
+      "learning_rate": 2.836289957835022e-05,
+      "loss": 0.2727,
+      "step": 5134
+    },
+    {
+      "epoch": 3.8652615732028606,
+      "grad_norm": 0.2777969241142273,
+      "learning_rate": 2.8355464390988028e-05,
+      "loss": 0.2838,
+      "step": 5135
+    },
+    {
+      "epoch": 3.866014301844185,
+      "grad_norm": 0.24909964203834534,
+      "learning_rate": 2.834802911007353e-05,
+      "loss": 0.3294,
+      "step": 5136
+    },
+    {
+      "epoch": 3.86676703048551,
+      "grad_norm": 0.2507777810096741,
+      "learning_rate": 2.8340593736419845e-05,
+      "loss": 0.2125,
+      "step": 5137
+    },
+    {
+      "epoch": 3.867519759126835,
+      "grad_norm": 0.2611388862133026,
+      "learning_rate": 2.83331582708401e-05,
+      "loss": 0.1727,
+      "step": 5138
+    },
+    {
+      "epoch": 3.8682724877681594,
+      "grad_norm": 0.22727464139461517,
+      "learning_rate": 2.8325722714147405e-05,
+      "loss": 0.1836,
+      "step": 5139
+    },
+    {
+      "epoch": 3.8690252164094843,
+      "grad_norm": 0.20622748136520386,
+      "learning_rate": 2.8318287067154924e-05,
+      "loss": 0.1749,
+      "step": 5140
+    },
+    {
+      "epoch": 3.869777945050809,
+      "grad_norm": 0.2239706814289093,
+      "learning_rate": 2.8310851330675804e-05,
+      "loss": 0.131,
+      "step": 5141
+    },
+    {
+      "epoch": 3.870530673692134,
+      "grad_norm": 0.24085046350955963,
+      "learning_rate": 2.830341550552319e-05,
+      "loss": 0.2635,
+      "step": 5142
+    },
+    {
+      "epoch": 3.871283402333459,
+      "grad_norm": 0.24871136248111725,
+      "learning_rate": 2.8295979592510274e-05,
+      "loss": 0.2567,
+      "step": 5143
+    },
+    {
+      "epoch": 3.8720361309747835,
+      "grad_norm": 0.24747350811958313,
+      "learning_rate": 2.828854359245022e-05,
+      "loss": 0.27,
+      "step": 5144
+    },
+    {
+      "epoch": 3.8727888596161084,
+      "grad_norm": 0.25158149003982544,
+      "learning_rate": 2.828110750615623e-05,
+      "loss": 0.2453,
+      "step": 5145
+    },
+    {
+      "epoch": 3.8735415882574333,
+      "grad_norm": 0.2190900593996048,
+      "learning_rate": 2.8273671334441493e-05,
+      "loss": 0.2181,
+      "step": 5146
+    },
+    {
+      "epoch": 3.874294316898758,
+      "grad_norm": 0.2048143446445465,
+      "learning_rate": 2.826623507811922e-05,
+      "loss": 0.1778,
+      "step": 5147
+    },
+    {
+      "epoch": 3.8750470455400827,
+      "grad_norm": 0.2932250201702118,
+      "learning_rate": 2.825879873800264e-05,
+      "loss": 0.1648,
+      "step": 5148
+    },
+    {
+      "epoch": 3.8757997741814076,
+      "grad_norm": 0.21717530488967896,
+      "learning_rate": 2.8251362314904967e-05,
+      "loss": 0.2608,
+      "step": 5149
+    },
+    {
+      "epoch": 3.8765525028227326,
+      "grad_norm": 0.22747915983200073,
+      "learning_rate": 2.824392580963944e-05,
+      "loss": 0.226,
+      "step": 5150
+    },
+    {
+      "epoch": 3.8773052314640575,
+      "grad_norm": 0.2449466735124588,
+      "learning_rate": 2.8236489223019302e-05,
+      "loss": 0.2297,
+      "step": 5151
+    },
+    {
+      "epoch": 3.878057960105382,
+      "grad_norm": 0.22408461570739746,
+      "learning_rate": 2.822905255585781e-05,
+      "loss": 0.3004,
+      "step": 5152
+    },
+    {
+      "epoch": 3.878810688746707,
+      "grad_norm": 0.22775417566299438,
+      "learning_rate": 2.8221615808968227e-05,
+      "loss": 0.2985,
+      "step": 5153
+    },
+    {
+      "epoch": 3.879563417388032,
+      "grad_norm": 0.21200519800186157,
+      "learning_rate": 2.8214178983163832e-05,
+      "loss": 0.2247,
+      "step": 5154
+    },
+    {
+      "epoch": 3.8803161460293563,
+      "grad_norm": 0.23513729870319366,
+      "learning_rate": 2.8206742079257886e-05,
+      "loss": 0.2639,
+      "step": 5155
+    },
+    {
+      "epoch": 3.881068874670681,
+      "grad_norm": 0.18240822851657867,
+      "learning_rate": 2.8199305098063705e-05,
+      "loss": 0.1845,
+      "step": 5156
+    },
+    {
+      "epoch": 3.881821603312006,
+      "grad_norm": 0.20865511894226074,
+      "learning_rate": 2.8191868040394553e-05,
+      "loss": 0.2119,
+      "step": 5157
+    },
+    {
+      "epoch": 3.8825743319533306,
+      "grad_norm": 0.2449069619178772,
+      "learning_rate": 2.8184430907063757e-05,
+      "loss": 0.2289,
+      "step": 5158
+    },
+    {
+      "epoch": 3.8833270605946555,
+      "grad_norm": 0.24041078984737396,
+      "learning_rate": 2.817699369888463e-05,
+      "loss": 0.2578,
+      "step": 5159
+    },
+    {
+      "epoch": 3.8840797892359804,
+      "grad_norm": 0.25661009550094604,
+      "learning_rate": 2.8169556416670484e-05,
+      "loss": 0.3244,
+      "step": 5160
+    },
+    {
+      "epoch": 3.8848325178773053,
+      "grad_norm": 0.2601989805698395,
+      "learning_rate": 2.816211906123466e-05,
+      "loss": 0.3733,
+      "step": 5161
+    },
+    {
+      "epoch": 3.8855852465186302,
+      "grad_norm": 0.1847388744354248,
+      "learning_rate": 2.815468163339049e-05,
+      "loss": 0.2485,
+      "step": 5162
+    },
+    {
+      "epoch": 3.8863379751599547,
+      "grad_norm": 0.18649053573608398,
+      "learning_rate": 2.8147244133951316e-05,
+      "loss": 0.203,
+      "step": 5163
+    },
+    {
+      "epoch": 3.8870907038012796,
+      "grad_norm": 0.22029078006744385,
+      "learning_rate": 2.8139806563730504e-05,
+      "loss": 0.1827,
+      "step": 5164
+    },
+    {
+      "epoch": 3.8878434324426046,
+      "grad_norm": 0.21909651160240173,
+      "learning_rate": 2.8132368923541403e-05,
+      "loss": 0.2166,
+      "step": 5165
+    },
+    {
+      "epoch": 3.888596161083929,
+      "grad_norm": 0.20682485401630402,
+      "learning_rate": 2.8124931214197387e-05,
+      "loss": 0.217,
+      "step": 5166
+    },
+    {
+      "epoch": 3.889348889725254,
+      "grad_norm": 0.23113669455051422,
+      "learning_rate": 2.8117493436511833e-05,
+      "loss": 0.1367,
+      "step": 5167
+    },
+    {
+      "epoch": 3.890101618366579,
+      "grad_norm": 0.2224704474210739,
+      "learning_rate": 2.811005559129813e-05,
+      "loss": 0.2215,
+      "step": 5168
+    },
+    {
+      "epoch": 3.890854347007904,
+      "grad_norm": 0.27703362703323364,
+      "learning_rate": 2.8102617679369657e-05,
+      "loss": 0.2094,
+      "step": 5169
+    },
+    {
+      "epoch": 3.8916070756492287,
+      "grad_norm": 0.22835442423820496,
+      "learning_rate": 2.8095179701539827e-05,
+      "loss": 0.1238,
+      "step": 5170
+    },
+    {
+      "epoch": 3.892359804290553,
+      "grad_norm": 0.2551701068878174,
+      "learning_rate": 2.808774165862204e-05,
+      "loss": 0.1391,
+      "step": 5171
+    },
+    {
+      "epoch": 3.893112532931878,
+      "grad_norm": 0.22724920511245728,
+      "learning_rate": 2.808030355142971e-05,
+      "loss": 0.2443,
+      "step": 5172
+    },
+    {
+      "epoch": 3.893865261573203,
+      "grad_norm": 0.2212437391281128,
+      "learning_rate": 2.8072865380776253e-05,
+      "loss": 0.2596,
+      "step": 5173
+    },
+    {
+      "epoch": 3.8946179902145275,
+      "grad_norm": 0.23742999136447906,
+      "learning_rate": 2.80654271474751e-05,
+      "loss": 0.2492,
+      "step": 5174
+    },
+    {
+      "epoch": 3.8953707188558524,
+      "grad_norm": 0.21718014776706696,
+      "learning_rate": 2.805798885233968e-05,
+      "loss": 0.2637,
+      "step": 5175
+    },
+    {
+      "epoch": 3.8961234474971773,
+      "grad_norm": 0.19993548095226288,
+      "learning_rate": 2.8050550496183442e-05,
+      "loss": 0.2285,
+      "step": 5176
+    },
+    {
+      "epoch": 3.8968761761385022,
+      "grad_norm": 0.24054668843746185,
+      "learning_rate": 2.8043112079819827e-05,
+      "loss": 0.2431,
+      "step": 5177
+    },
+    {
+      "epoch": 3.8976289047798267,
+      "grad_norm": 0.28412532806396484,
+      "learning_rate": 2.80356736040623e-05,
+      "loss": 0.1892,
+      "step": 5178
+    },
+    {
+      "epoch": 3.8983816334211516,
+      "grad_norm": 0.19540680944919586,
+      "learning_rate": 2.8028235069724303e-05,
+      "loss": 0.2128,
+      "step": 5179
+    },
+    {
+      "epoch": 3.8991343620624765,
+      "grad_norm": 0.22434371709823608,
+      "learning_rate": 2.8020796477619314e-05,
+      "loss": 0.2555,
+      "step": 5180
+    },
+    {
+      "epoch": 3.8998870907038015,
+      "grad_norm": 0.20730356872081757,
+      "learning_rate": 2.801335782856081e-05,
+      "loss": 0.2293,
+      "step": 5181
+    },
+    {
+      "epoch": 3.900639819345126,
+      "grad_norm": 0.205093652009964,
+      "learning_rate": 2.8005919123362262e-05,
+      "loss": 0.2092,
+      "step": 5182
+    },
+    {
+      "epoch": 3.901392547986451,
+      "grad_norm": 0.22410397231578827,
+      "learning_rate": 2.799848036283716e-05,
+      "loss": 0.2338,
+      "step": 5183
+    },
+    {
+      "epoch": 3.9021452766277758,
+      "grad_norm": 0.21636341512203217,
+      "learning_rate": 2.7991041547798995e-05,
+      "loss": 0.2292,
+      "step": 5184
+    },
+    {
+      "epoch": 3.9028980052691002,
+      "grad_norm": 0.20038799941539764,
+      "learning_rate": 2.7983602679061265e-05,
+      "loss": 0.2797,
+      "step": 5185
+    },
+    {
+      "epoch": 3.903650733910425,
+      "grad_norm": 0.209176167845726,
+      "learning_rate": 2.797616375743748e-05,
+      "loss": 0.24,
+      "step": 5186
+    },
+    {
+      "epoch": 3.90440346255175,
+      "grad_norm": 0.207145094871521,
+      "learning_rate": 2.7968724783741135e-05,
+      "loss": 0.2033,
+      "step": 5187
+    },
+    {
+      "epoch": 3.905156191193075,
+      "grad_norm": 0.25238731503486633,
+      "learning_rate": 2.796128575878575e-05,
+      "loss": 0.3304,
+      "step": 5188
+    },
+    {
+      "epoch": 3.9059089198344,
+      "grad_norm": 0.24730351567268372,
+      "learning_rate": 2.795384668338485e-05,
+      "loss": 0.2628,
+      "step": 5189
+    },
+    {
+      "epoch": 3.9066616484757244,
+      "grad_norm": 0.2127484679222107,
+      "learning_rate": 2.7946407558351955e-05,
+      "loss": 0.2359,
+      "step": 5190
+    },
+    {
+      "epoch": 3.9074143771170493,
+      "grad_norm": 0.2112760692834854,
+      "learning_rate": 2.7938968384500612e-05,
+      "loss": 0.2342,
+      "step": 5191
+    },
+    {
+      "epoch": 3.9081671057583742,
+      "grad_norm": 0.2216774821281433,
+      "learning_rate": 2.7931529162644337e-05,
+      "loss": 0.2275,
+      "step": 5192
+    },
+    {
+      "epoch": 3.9089198343996987,
+      "grad_norm": 0.23816600441932678,
+      "learning_rate": 2.792408989359669e-05,
+      "loss": 0.2366,
+      "step": 5193
+    },
+    {
+      "epoch": 3.9096725630410236,
+      "grad_norm": 0.21431398391723633,
+      "learning_rate": 2.7916650578171205e-05,
+      "loss": 0.1635,
+      "step": 5194
+    },
+    {
+      "epoch": 3.9104252916823485,
+      "grad_norm": 0.19035840034484863,
+      "learning_rate": 2.7909211217181437e-05,
+      "loss": 0.1591,
+      "step": 5195
+    },
+    {
+      "epoch": 3.9111780203236735,
+      "grad_norm": 0.22524571418762207,
+      "learning_rate": 2.7901771811440946e-05,
+      "loss": 0.2315,
+      "step": 5196
+    },
+    {
+      "epoch": 3.9119307489649984,
+      "grad_norm": 0.2331884354352951,
+      "learning_rate": 2.78943323617633e-05,
+      "loss": 0.2791,
+      "step": 5197
+    },
+    {
+      "epoch": 3.912683477606323,
+      "grad_norm": 0.23903335630893707,
+      "learning_rate": 2.788689286896205e-05,
+      "loss": 0.2047,
+      "step": 5198
+    },
+    {
+      "epoch": 3.9134362062476478,
+      "grad_norm": 0.2021699845790863,
+      "learning_rate": 2.787945333385078e-05,
+      "loss": 0.1814,
+      "step": 5199
+    },
+    {
+      "epoch": 3.9141889348889727,
+      "grad_norm": 0.21187584102153778,
+      "learning_rate": 2.787201375724307e-05,
+      "loss": 0.211,
+      "step": 5200
+    },
+    {
+      "epoch": 3.9141889348889727,
+      "eval_loss": 0.2472326159477234,
+      "eval_runtime": 452.6195,
+      "eval_samples_per_second": 21.27,
+      "eval_steps_per_second": 0.665,
+      "step": 5200
+    },
+    {
+      "epoch": 3.914941663530297,
+      "grad_norm": 0.2097022980451584,
+      "learning_rate": 2.786457413995249e-05,
+      "loss": 0.1926,
+      "step": 5201
+    },
+    {
+      "epoch": 3.915694392171622,
+      "grad_norm": 0.24172590672969818,
+      "learning_rate": 2.785713448279263e-05,
+      "loss": 0.2713,
+      "step": 5202
+    },
+    {
+      "epoch": 3.916447120812947,
+      "grad_norm": 0.21279914677143097,
+      "learning_rate": 2.784969478657708e-05,
+      "loss": 0.2593,
+      "step": 5203
+    },
+    {
+      "epoch": 3.9171998494542715,
+      "grad_norm": 0.23038940131664276,
+      "learning_rate": 2.784225505211943e-05,
+      "loss": 0.1877,
+      "step": 5204
+    },
+    {
+      "epoch": 3.9179525780955964,
+      "grad_norm": 0.24554310739040375,
+      "learning_rate": 2.7834815280233296e-05,
+      "loss": 0.1956,
+      "step": 5205
+    },
+    {
+      "epoch": 3.9187053067369213,
+      "grad_norm": 0.19825585186481476,
+      "learning_rate": 2.7827375471732258e-05,
+      "loss": 0.1141,
+      "step": 5206
+    },
+    {
+      "epoch": 3.919458035378246,
+      "grad_norm": 0.2244979590177536,
+      "learning_rate": 2.7819935627429932e-05,
+      "loss": 0.1619,
+      "step": 5207
+    },
+    {
+      "epoch": 3.920210764019571,
+      "grad_norm": 0.20835700631141663,
+      "learning_rate": 2.781249574813992e-05,
+      "loss": 0.2055,
+      "step": 5208
+    },
+    {
+      "epoch": 3.9209634926608956,
+      "grad_norm": 0.22171145677566528,
+      "learning_rate": 2.7805055834675847e-05,
+      "loss": 0.2499,
+      "step": 5209
+    },
+    {
+      "epoch": 3.9217162213022205,
+      "grad_norm": 0.2312268614768982,
+      "learning_rate": 2.779761588785133e-05,
+      "loss": 0.224,
+      "step": 5210
+    },
+    {
+      "epoch": 3.9224689499435454,
+      "grad_norm": 0.23826062679290771,
+      "learning_rate": 2.779017590847999e-05,
+      "loss": 0.2546,
+      "step": 5211
+    },
+    {
+      "epoch": 3.92322167858487,
+      "grad_norm": 0.18400180339813232,
+      "learning_rate": 2.778273589737544e-05,
+      "loss": 0.2325,
+      "step": 5212
+    },
+    {
+      "epoch": 3.923974407226195,
+      "grad_norm": 0.2038690447807312,
+      "learning_rate": 2.777529585535132e-05,
+      "loss": 0.2251,
+      "step": 5213
+    },
+    {
+      "epoch": 3.9247271358675198,
+      "grad_norm": 0.1944538652896881,
+      "learning_rate": 2.776785578322126e-05,
+      "loss": 0.1477,
+      "step": 5214
+    },
+    {
+      "epoch": 3.9254798645088447,
+      "grad_norm": 0.2455904483795166,
+      "learning_rate": 2.776041568179889e-05,
+      "loss": 0.2365,
+      "step": 5215
+    },
+    {
+      "epoch": 3.9262325931501696,
+      "grad_norm": 0.21341755986213684,
+      "learning_rate": 2.7752975551897857e-05,
+      "loss": 0.2391,
+      "step": 5216
+    },
+    {
+      "epoch": 3.926985321791494,
+      "grad_norm": 0.2073490023612976,
+      "learning_rate": 2.7745535394331794e-05,
+      "loss": 0.2014,
+      "step": 5217
+    },
+    {
+      "epoch": 3.927738050432819,
+      "grad_norm": 0.22116345167160034,
+      "learning_rate": 2.7738095209914355e-05,
+      "loss": 0.1847,
+      "step": 5218
+    },
+    {
+      "epoch": 3.928490779074144,
+      "grad_norm": 0.203879252076149,
+      "learning_rate": 2.7730654999459182e-05,
+      "loss": 0.2421,
+      "step": 5219
+    },
+    {
+      "epoch": 3.9292435077154684,
+      "grad_norm": 0.23773352801799774,
+      "learning_rate": 2.772321476377992e-05,
+      "loss": 0.279,
+      "step": 5220
+    },
+    {
+      "epoch": 3.9299962363567933,
+      "grad_norm": 0.22271859645843506,
+      "learning_rate": 2.7715774503690235e-05,
+      "loss": 0.2188,
+      "step": 5221
+    },
+    {
+      "epoch": 3.930748964998118,
+      "grad_norm": 0.22991454601287842,
+      "learning_rate": 2.7708334220003767e-05,
+      "loss": 0.2874,
+      "step": 5222
+    },
+    {
+      "epoch": 3.931501693639443,
+      "grad_norm": 0.19609224796295166,
+      "learning_rate": 2.770089391353418e-05,
+      "loss": 0.165,
+      "step": 5223
+    },
+    {
+      "epoch": 3.932254422280768,
+      "grad_norm": 0.22230276465415955,
+      "learning_rate": 2.7693453585095147e-05,
+      "loss": 0.2197,
+      "step": 5224
+    },
+    {
+      "epoch": 3.9330071509220925,
+      "grad_norm": 0.2185634821653366,
+      "learning_rate": 2.7686013235500314e-05,
+      "loss": 0.1886,
+      "step": 5225
+    },
+    {
+      "epoch": 3.9337598795634174,
+      "grad_norm": 0.22119298577308655,
+      "learning_rate": 2.7678572865563358e-05,
+      "loss": 0.2099,
+      "step": 5226
+    },
+    {
+      "epoch": 3.9345126082047424,
+      "grad_norm": 0.1952393501996994,
+      "learning_rate": 2.7671132476097932e-05,
+      "loss": 0.2322,
+      "step": 5227
+    },
+    {
+      "epoch": 3.935265336846067,
+      "grad_norm": 0.18340881168842316,
+      "learning_rate": 2.7663692067917723e-05,
+      "loss": 0.1692,
+      "step": 5228
+    },
+    {
+      "epoch": 3.9360180654873917,
+      "grad_norm": 0.2330743968486786,
+      "learning_rate": 2.76562516418364e-05,
+      "loss": 0.2392,
+      "step": 5229
+    },
+    {
+      "epoch": 3.9367707941287167,
+      "grad_norm": 0.22727690637111664,
+      "learning_rate": 2.7648811198667623e-05,
+      "loss": 0.2419,
+      "step": 5230
+    },
+    {
+      "epoch": 3.937523522770041,
+      "grad_norm": 0.2039709836244583,
+      "learning_rate": 2.764137073922508e-05,
+      "loss": 0.1878,
+      "step": 5231
+    },
+    {
+      "epoch": 3.938276251411366,
+      "grad_norm": 0.2515558898448944,
+      "learning_rate": 2.763393026432245e-05,
+      "loss": 0.2845,
+      "step": 5232
+    },
+    {
+      "epoch": 3.939028980052691,
+      "grad_norm": 0.19824016094207764,
+      "learning_rate": 2.76264897747734e-05,
+      "loss": 0.3262,
+      "step": 5233
+    },
+    {
+      "epoch": 3.939781708694016,
+      "grad_norm": 0.22061805427074432,
+      "learning_rate": 2.761904927139162e-05,
+      "loss": 0.2303,
+      "step": 5234
+    },
+    {
+      "epoch": 3.940534437335341,
+      "grad_norm": 0.21236543357372284,
+      "learning_rate": 2.7611608754990797e-05,
+      "loss": 0.1779,
+      "step": 5235
+    },
+    {
+      "epoch": 3.9412871659766653,
+      "grad_norm": 0.21420659124851227,
+      "learning_rate": 2.7604168226384603e-05,
+      "loss": 0.1462,
+      "step": 5236
+    },
+    {
+      "epoch": 3.94203989461799,
+      "grad_norm": 0.20456738770008087,
+      "learning_rate": 2.7596727686386732e-05,
+      "loss": 0.1869,
+      "step": 5237
+    },
+    {
+      "epoch": 3.942792623259315,
+      "grad_norm": 0.22511440515518188,
+      "learning_rate": 2.758928713581086e-05,
+      "loss": 0.1599,
+      "step": 5238
+    },
+    {
+      "epoch": 3.9435453519006396,
+      "grad_norm": 0.1975979208946228,
+      "learning_rate": 2.7581846575470688e-05,
+      "loss": 0.2701,
+      "step": 5239
+    },
+    {
+      "epoch": 3.9442980805419645,
+      "grad_norm": 0.1957053393125534,
+      "learning_rate": 2.7574406006179893e-05,
+      "loss": 0.2478,
+      "step": 5240
+    },
+    {
+      "epoch": 3.9450508091832894,
+      "grad_norm": 0.2126346379518509,
+      "learning_rate": 2.7566965428752174e-05,
+      "loss": 0.2092,
+      "step": 5241
+    },
+    {
+      "epoch": 3.9458035378246143,
+      "grad_norm": 0.19817128777503967,
+      "learning_rate": 2.755952484400121e-05,
+      "loss": 0.3272,
+      "step": 5242
+    },
+    {
+      "epoch": 3.9465562664659393,
+      "grad_norm": 0.21198275685310364,
+      "learning_rate": 2.7552084252740702e-05,
+      "loss": 0.1826,
+      "step": 5243
+    },
+    {
+      "epoch": 3.9473089951072637,
+      "grad_norm": 0.20729517936706543,
+      "learning_rate": 2.7544643655784334e-05,
+      "loss": 0.2133,
+      "step": 5244
+    },
+    {
+      "epoch": 3.9480617237485887,
+      "grad_norm": 0.2179228961467743,
+      "learning_rate": 2.7537203053945813e-05,
+      "loss": 0.3173,
+      "step": 5245
+    },
+    {
+      "epoch": 3.9488144523899136,
+      "grad_norm": 0.20608119666576385,
+      "learning_rate": 2.7529762448038814e-05,
+      "loss": 0.168,
+      "step": 5246
+    },
+    {
+      "epoch": 3.949567181031238,
+      "grad_norm": 0.20234349370002747,
+      "learning_rate": 2.7522321838877046e-05,
+      "loss": 0.2455,
+      "step": 5247
+    },
+    {
+      "epoch": 3.950319909672563,
+      "grad_norm": 0.20164379477500916,
+      "learning_rate": 2.751488122727419e-05,
+      "loss": 0.2069,
+      "step": 5248
+    },
+    {
+      "epoch": 3.951072638313888,
+      "grad_norm": 0.1753852665424347,
+      "learning_rate": 2.7507440614043938e-05,
+      "loss": 0.2216,
+      "step": 5249
+    },
+    {
+      "epoch": 3.951825366955213,
+      "grad_norm": 0.20509560406208038,
+      "learning_rate": 2.7500000000000004e-05,
+      "loss": 0.267,
+      "step": 5250
+    },
+    {
+      "epoch": 3.9525780955965377,
+      "grad_norm": 0.20615456998348236,
+      "learning_rate": 2.7492559385956068e-05,
+      "loss": 0.2328,
+      "step": 5251
+    },
+    {
+      "epoch": 3.953330824237862,
+      "grad_norm": 0.24267330765724182,
+      "learning_rate": 2.748511877272582e-05,
+      "loss": 0.2173,
+      "step": 5252
+    },
+    {
+      "epoch": 3.954083552879187,
+      "grad_norm": 0.19328705966472626,
+      "learning_rate": 2.7477678161122972e-05,
+      "loss": 0.1577,
+      "step": 5253
+    },
+    {
+      "epoch": 3.954836281520512,
+      "grad_norm": 0.20610471069812775,
+      "learning_rate": 2.7470237551961188e-05,
+      "loss": 0.2111,
+      "step": 5254
+    },
+    {
+      "epoch": 3.9555890101618365,
+      "grad_norm": 0.22875353693962097,
+      "learning_rate": 2.7462796946054192e-05,
+      "loss": 0.2092,
+      "step": 5255
+    },
+    {
+      "epoch": 3.9563417388031614,
+      "grad_norm": 0.17902930080890656,
+      "learning_rate": 2.7455356344215665e-05,
+      "loss": 0.2924,
+      "step": 5256
+    },
+    {
+      "epoch": 3.9570944674444863,
+      "grad_norm": 0.28893256187438965,
+      "learning_rate": 2.7447915747259306e-05,
+      "loss": 0.2175,
+      "step": 5257
+    },
+    {
+      "epoch": 3.957847196085811,
+      "grad_norm": 0.22897616028785706,
+      "learning_rate": 2.74404751559988e-05,
+      "loss": 0.2614,
+      "step": 5258
+    },
+    {
+      "epoch": 3.9585999247271357,
+      "grad_norm": 0.194962278008461,
+      "learning_rate": 2.7433034571247835e-05,
+      "loss": 0.1457,
+      "step": 5259
+    },
+    {
+      "epoch": 3.9593526533684607,
+      "grad_norm": 0.22722980380058289,
+      "learning_rate": 2.742559399382011e-05,
+      "loss": 0.1338,
+      "step": 5260
+    },
+    {
+      "epoch": 3.9601053820097856,
+      "grad_norm": 0.2401580661535263,
+      "learning_rate": 2.7418153424529318e-05,
+      "loss": 0.1952,
+      "step": 5261
+    },
+    {
+      "epoch": 3.9608581106511105,
+      "grad_norm": 0.22119103372097015,
+      "learning_rate": 2.7410712864189147e-05,
+      "loss": 0.1812,
+      "step": 5262
+    },
+    {
+      "epoch": 3.961610839292435,
+      "grad_norm": 0.24943147599697113,
+      "learning_rate": 2.7403272313613277e-05,
+      "loss": 0.2682,
+      "step": 5263
+    },
+    {
+      "epoch": 3.96236356793376,
+      "grad_norm": 0.20843467116355896,
+      "learning_rate": 2.7395831773615405e-05,
+      "loss": 0.2351,
+      "step": 5264
+    },
+    {
+      "epoch": 3.963116296575085,
+      "grad_norm": 0.2440183460712433,
+      "learning_rate": 2.7388391245009215e-05,
+      "loss": 0.252,
+      "step": 5265
+    },
+    {
+      "epoch": 3.9638690252164093,
+      "grad_norm": 0.2079485058784485,
+      "learning_rate": 2.7380950728608384e-05,
+      "loss": 0.257,
+      "step": 5266
+    },
+    {
+      "epoch": 3.964621753857734,
+      "grad_norm": 0.20726004242897034,
+      "learning_rate": 2.73735102252266e-05,
+      "loss": 0.2192,
+      "step": 5267
+    },
+    {
+      "epoch": 3.965374482499059,
+      "grad_norm": 0.2530066967010498,
+      "learning_rate": 2.7366069735677558e-05,
+      "loss": 0.193,
+      "step": 5268
+    },
+    {
+      "epoch": 3.966127211140384,
+      "grad_norm": 0.2218208760023117,
+      "learning_rate": 2.7358629260774928e-05,
+      "loss": 0.2293,
+      "step": 5269
+    },
+    {
+      "epoch": 3.966879939781709,
+      "grad_norm": 0.211141899228096,
+      "learning_rate": 2.7351188801332383e-05,
+      "loss": 0.2679,
+      "step": 5270
+    },
+    {
+      "epoch": 3.9676326684230334,
+      "grad_norm": 0.2157779037952423,
+      "learning_rate": 2.734374835816361e-05,
+      "loss": 0.1928,
+      "step": 5271
+    },
+    {
+      "epoch": 3.9683853970643583,
+      "grad_norm": 0.24232225120067596,
+      "learning_rate": 2.733630793208229e-05,
+      "loss": 0.2532,
+      "step": 5272
+    },
+    {
+      "epoch": 3.9691381257056833,
+      "grad_norm": 0.19285544753074646,
+      "learning_rate": 2.732886752390207e-05,
+      "loss": 0.1237,
+      "step": 5273
+    },
+    {
+      "epoch": 3.9698908543470077,
+      "grad_norm": 0.22670121490955353,
+      "learning_rate": 2.732142713443665e-05,
+      "loss": 0.208,
+      "step": 5274
+    },
+    {
+      "epoch": 3.9706435829883326,
+      "grad_norm": 0.2511517107486725,
+      "learning_rate": 2.7313986764499695e-05,
+      "loss": 0.2399,
+      "step": 5275
+    },
+    {
+      "epoch": 3.9713963116296576,
+      "grad_norm": 0.2176334112882614,
+      "learning_rate": 2.7306546414904865e-05,
+      "loss": 0.1716,
+      "step": 5276
+    },
+    {
+      "epoch": 3.972149040270982,
+      "grad_norm": 0.1943291872739792,
+      "learning_rate": 2.7299106086465825e-05,
+      "loss": 0.2064,
+      "step": 5277
+    },
+    {
+      "epoch": 3.972901768912307,
+      "grad_norm": 0.18535782396793365,
+      "learning_rate": 2.7291665779996245e-05,
+      "loss": 0.2137,
+      "step": 5278
+    },
+    {
+      "epoch": 3.973654497553632,
+      "grad_norm": 0.21579720079898834,
+      "learning_rate": 2.7284225496309774e-05,
+      "loss": 0.1983,
+      "step": 5279
+    },
+    {
+      "epoch": 3.974407226194957,
+      "grad_norm": 0.21473805606365204,
+      "learning_rate": 2.7276785236220083e-05,
+      "loss": 0.2659,
+      "step": 5280
+    },
+    {
+      "epoch": 3.9751599548362817,
+      "grad_norm": 0.2130424529314041,
+      "learning_rate": 2.7269345000540823e-05,
+      "loss": 0.2793,
+      "step": 5281
+    },
+    {
+      "epoch": 3.975912683477606,
+      "grad_norm": 0.22970011830329895,
+      "learning_rate": 2.7261904790085653e-05,
+      "loss": 0.1821,
+      "step": 5282
+    },
+    {
+      "epoch": 3.976665412118931,
+      "grad_norm": 0.21536506712436676,
+      "learning_rate": 2.725446460566821e-05,
+      "loss": 0.1816,
+      "step": 5283
+    },
+    {
+      "epoch": 3.977418140760256,
+      "grad_norm": 0.21669770777225494,
+      "learning_rate": 2.7247024448102148e-05,
+      "loss": 0.225,
+      "step": 5284
+    },
+    {
+      "epoch": 3.9781708694015805,
+      "grad_norm": 0.2555747628211975,
+      "learning_rate": 2.7239584318201115e-05,
+      "loss": 0.2412,
+      "step": 5285
+    },
+    {
+      "epoch": 3.9789235980429054,
+      "grad_norm": 0.23621438443660736,
+      "learning_rate": 2.723214421677875e-05,
+      "loss": 0.2493,
+      "step": 5286
+    },
+    {
+      "epoch": 3.9796763266842303,
+      "grad_norm": 0.18852214515209198,
+      "learning_rate": 2.722470414464869e-05,
+      "loss": 0.2545,
+      "step": 5287
+    },
+    {
+      "epoch": 3.9804290553255552,
+      "grad_norm": 0.23302681744098663,
+      "learning_rate": 2.7217264102624566e-05,
+      "loss": 0.1469,
+      "step": 5288
+    },
+    {
+      "epoch": 3.98118178396688,
+      "grad_norm": 0.21294179558753967,
+      "learning_rate": 2.7209824091520027e-05,
+      "loss": 0.2366,
+      "step": 5289
+    },
+    {
+      "epoch": 3.9819345126082046,
+      "grad_norm": 0.20908093452453613,
+      "learning_rate": 2.7202384112148675e-05,
+      "loss": 0.1465,
+      "step": 5290
+    },
+    {
+      "epoch": 3.9826872412495296,
+      "grad_norm": 0.38081493973731995,
+      "learning_rate": 2.719494416532416e-05,
+      "loss": 0.3068,
+      "step": 5291
+    },
+    {
+      "epoch": 3.9834399698908545,
+      "grad_norm": 0.23938626050949097,
+      "learning_rate": 2.718750425186008e-05,
+      "loss": 0.2505,
+      "step": 5292
+    },
+    {
+      "epoch": 3.984192698532179,
+      "grad_norm": 0.26418042182922363,
+      "learning_rate": 2.7180064372570073e-05,
+      "loss": 0.2581,
+      "step": 5293
+    },
+    {
+      "epoch": 3.984945427173504,
+      "grad_norm": 0.22390827536582947,
+      "learning_rate": 2.7172624528267748e-05,
+      "loss": 0.1769,
+      "step": 5294
+    },
+    {
+      "epoch": 3.985698155814829,
+      "grad_norm": 0.2129838615655899,
+      "learning_rate": 2.7165184719766713e-05,
+      "loss": 0.2037,
+      "step": 5295
+    },
+    {
+      "epoch": 3.9864508844561537,
+      "grad_norm": 0.2639102339744568,
+      "learning_rate": 2.715774494788057e-05,
+      "loss": 0.2607,
+      "step": 5296
+    },
+    {
+      "epoch": 3.9872036130974786,
+      "grad_norm": 0.23060745000839233,
+      "learning_rate": 2.7150305213422928e-05,
+      "loss": 0.2093,
+      "step": 5297
+    },
+    {
+      "epoch": 3.987956341738803,
+      "grad_norm": 0.21141384541988373,
+      "learning_rate": 2.7142865517207372e-05,
+      "loss": 0.2832,
+      "step": 5298
+    },
+    {
+      "epoch": 3.988709070380128,
+      "grad_norm": 0.2359461635351181,
+      "learning_rate": 2.7135425860047513e-05,
+      "loss": 0.1729,
+      "step": 5299
+    },
+    {
+      "epoch": 3.989461799021453,
+      "grad_norm": 0.26199865341186523,
+      "learning_rate": 2.7127986242756936e-05,
+      "loss": 0.251,
+      "step": 5300
+    },
+    {
+      "epoch": 3.9902145276627774,
+      "grad_norm": 0.23214924335479736,
+      "learning_rate": 2.7120546666149225e-05,
+      "loss": 0.1976,
+      "step": 5301
+    },
+    {
+      "epoch": 3.9909672563041023,
+      "grad_norm": 0.2608475387096405,
+      "learning_rate": 2.711310713103796e-05,
+      "loss": 0.2772,
+      "step": 5302
+    },
+    {
+      "epoch": 3.9917199849454272,
+      "grad_norm": 0.22249330580234528,
+      "learning_rate": 2.7105667638236714e-05,
+      "loss": 0.2093,
+      "step": 5303
+    },
+    {
+      "epoch": 3.9924727135867517,
+      "grad_norm": 0.19749614596366882,
+      "learning_rate": 2.7098228188559066e-05,
+      "loss": 0.1509,
+      "step": 5304
+    },
+    {
+      "epoch": 3.9932254422280766,
+      "grad_norm": 0.2490306794643402,
+      "learning_rate": 2.7090788782818565e-05,
+      "loss": 0.2549,
+      "step": 5305
+    },
+    {
+      "epoch": 3.9939781708694015,
+      "grad_norm": 0.1917862594127655,
+      "learning_rate": 2.70833494218288e-05,
+      "loss": 0.1632,
+      "step": 5306
+    },
+    {
+      "epoch": 3.9947308995107265,
+      "grad_norm": 0.2716890573501587,
+      "learning_rate": 2.7075910106403317e-05,
+      "loss": 0.3102,
+      "step": 5307
+    },
+    {
+      "epoch": 3.9954836281520514,
+      "grad_norm": 0.22935888171195984,
+      "learning_rate": 2.706847083735567e-05,
+      "loss": 0.2061,
+      "step": 5308
+    },
+    {
+      "epoch": 3.996236356793376,
+      "grad_norm": 0.1863192468881607,
+      "learning_rate": 2.7061031615499393e-05,
+      "loss": 0.1424,
+      "step": 5309
+    },
+    {
+      "epoch": 3.9969890854347008,
+      "grad_norm": 0.2255145162343979,
+      "learning_rate": 2.7053592441648047e-05,
+      "loss": 0.3229,
+      "step": 5310
+    },
+    {
+      "epoch": 3.9977418140760257,
+      "grad_norm": 0.23559299111366272,
+      "learning_rate": 2.7046153316615153e-05,
+      "loss": 0.211,
+      "step": 5311
+    },
+    {
+      "epoch": 3.99849454271735,
+      "grad_norm": 0.2271050363779068,
+      "learning_rate": 2.703871424121426e-05,
+      "loss": 0.2785,
+      "step": 5312
+    },
+    {
+      "epoch": 3.999247271358675,
+      "grad_norm": 0.19760605692863464,
+      "learning_rate": 2.7031275216258873e-05,
+      "loss": 0.2667,
+      "step": 5313
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.24993963539600372,
+      "learning_rate": 2.7023836242562534e-05,
+      "loss": 0.37,
+      "step": 5314
+    },
+    {
+      "epoch": 4.0007527286413245,
+      "grad_norm": 0.210062637925148,
+      "learning_rate": 2.701639732093874e-05,
+      "loss": 0.2527,
+      "step": 5315
+    },
+    {
+      "epoch": 4.00150545728265,
+      "grad_norm": 0.23028726875782013,
+      "learning_rate": 2.700895845220101e-05,
+      "loss": 0.247,
+      "step": 5316
+    },
+    {
+      "epoch": 4.002258185923974,
+      "grad_norm": 0.24175697565078735,
+      "learning_rate": 2.7001519637162843e-05,
+      "loss": 0.2459,
+      "step": 5317
+    },
+    {
+      "epoch": 4.003010914565299,
+      "grad_norm": 0.23549221456050873,
+      "learning_rate": 2.699408087663774e-05,
+      "loss": 0.2386,
+      "step": 5318
+    },
+    {
+      "epoch": 4.003763643206624,
+      "grad_norm": 0.21199169754981995,
+      "learning_rate": 2.6986642171439196e-05,
+      "loss": 0.2742,
+      "step": 5319
+    },
+    {
+      "epoch": 4.004516371847949,
+      "grad_norm": 0.2206403911113739,
+      "learning_rate": 2.697920352238069e-05,
+      "loss": 0.2476,
+      "step": 5320
+    },
+    {
+      "epoch": 4.005269100489274,
+      "grad_norm": 0.2554951608181,
+      "learning_rate": 2.6971764930275706e-05,
+      "loss": 0.1944,
+      "step": 5321
+    },
+    {
+      "epoch": 4.0060218291305985,
+      "grad_norm": 0.19763551652431488,
+      "learning_rate": 2.6964326395937718e-05,
+      "loss": 0.1997,
+      "step": 5322
+    },
+    {
+      "epoch": 4.006774557771923,
+      "grad_norm": 0.18602444231510162,
+      "learning_rate": 2.695688792018018e-05,
+      "loss": 0.1618,
+      "step": 5323
+    },
+    {
+      "epoch": 4.007527286413248,
+      "grad_norm": 0.21734891831874847,
+      "learning_rate": 2.6949449503816566e-05,
+      "loss": 0.1159,
+      "step": 5324
+    },
+    {
+      "epoch": 4.008280015054573,
+      "grad_norm": 0.2432304322719574,
+      "learning_rate": 2.6942011147660324e-05,
+      "loss": 0.1802,
+      "step": 5325
+    },
+    {
+      "epoch": 4.009032743695897,
+      "grad_norm": 0.21057386696338654,
+      "learning_rate": 2.6934572852524907e-05,
+      "loss": 0.1902,
+      "step": 5326
+    },
+    {
+      "epoch": 4.009785472337223,
+      "grad_norm": 0.20604808628559113,
+      "learning_rate": 2.692713461922376e-05,
+      "loss": 0.2288,
+      "step": 5327
+    },
+    {
+      "epoch": 4.010538200978547,
+      "grad_norm": 0.20397256314754486,
+      "learning_rate": 2.6919696448570304e-05,
+      "loss": 0.1826,
+      "step": 5328
+    },
+    {
+      "epoch": 4.011290929619872,
+      "grad_norm": 0.23055033385753632,
+      "learning_rate": 2.6912258341377973e-05,
+      "loss": 0.2915,
+      "step": 5329
+    },
+    {
+      "epoch": 4.012043658261197,
+      "grad_norm": 0.2285270392894745,
+      "learning_rate": 2.6904820298460175e-05,
+      "loss": 0.2009,
+      "step": 5330
+    },
+    {
+      "epoch": 4.012796386902521,
+      "grad_norm": 0.20441722869873047,
+      "learning_rate": 2.6897382320630345e-05,
+      "loss": 0.1802,
+      "step": 5331
+    },
+    {
+      "epoch": 4.013549115543847,
+      "grad_norm": 0.21460309624671936,
+      "learning_rate": 2.688994440870188e-05,
+      "loss": 0.2755,
+      "step": 5332
+    },
+    {
+      "epoch": 4.014301844185171,
+      "grad_norm": 0.20432282984256744,
+      "learning_rate": 2.6882506563488175e-05,
+      "loss": 0.1439,
+      "step": 5333
+    },
+    {
+      "epoch": 4.015054572826496,
+      "grad_norm": 0.19839370250701904,
+      "learning_rate": 2.687506878580262e-05,
+      "loss": 0.2227,
+      "step": 5334
+    },
+    {
+      "epoch": 4.015807301467821,
+      "grad_norm": 0.21638894081115723,
+      "learning_rate": 2.6867631076458606e-05,
+      "loss": 0.254,
+      "step": 5335
+    },
+    {
+      "epoch": 4.0165600301091455,
+      "grad_norm": 0.22568637132644653,
+      "learning_rate": 2.6860193436269505e-05,
+      "loss": 0.2484,
+      "step": 5336
+    },
+    {
+      "epoch": 4.017312758750471,
+      "grad_norm": 0.22161513566970825,
+      "learning_rate": 2.685275586604869e-05,
+      "loss": 0.2099,
+      "step": 5337
+    },
+    {
+      "epoch": 4.018065487391795,
+      "grad_norm": 0.20889908075332642,
+      "learning_rate": 2.684531836660952e-05,
+      "loss": 0.2576,
+      "step": 5338
+    },
+    {
+      "epoch": 4.01881821603312,
+      "grad_norm": 0.21911045908927917,
+      "learning_rate": 2.6837880938765346e-05,
+      "loss": 0.1337,
+      "step": 5339
+    },
+    {
+      "epoch": 4.019570944674445,
+      "grad_norm": 0.23231647908687592,
+      "learning_rate": 2.6830443583329525e-05,
+      "loss": 0.2896,
+      "step": 5340
+    },
+    {
+      "epoch": 4.02032367331577,
+      "grad_norm": 0.19148406386375427,
+      "learning_rate": 2.682300630111538e-05,
+      "loss": 0.1351,
+      "step": 5341
+    },
+    {
+      "epoch": 4.021076401957094,
+      "grad_norm": 0.19677932560443878,
+      "learning_rate": 2.6815569092936255e-05,
+      "loss": 0.1579,
+      "step": 5342
+    },
+    {
+      "epoch": 4.0218291305984195,
+      "grad_norm": 0.20227360725402832,
+      "learning_rate": 2.680813195960545e-05,
+      "loss": 0.2956,
+      "step": 5343
+    },
+    {
+      "epoch": 4.022581859239744,
+      "grad_norm": 0.19742755591869354,
+      "learning_rate": 2.6800694901936308e-05,
+      "loss": 0.3245,
+      "step": 5344
+    },
+    {
+      "epoch": 4.0233345878810685,
+      "grad_norm": 0.22487394511699677,
+      "learning_rate": 2.6793257920742116e-05,
+      "loss": 0.2036,
+      "step": 5345
+    },
+    {
+      "epoch": 4.024087316522394,
+      "grad_norm": 0.20613858103752136,
+      "learning_rate": 2.6785821016836177e-05,
+      "loss": 0.1829,
+      "step": 5346
+    },
+    {
+      "epoch": 4.024840045163718,
+      "grad_norm": 0.22390300035476685,
+      "learning_rate": 2.6778384191031775e-05,
+      "loss": 0.2427,
+      "step": 5347
+    },
+    {
+      "epoch": 4.025592773805044,
+      "grad_norm": 0.2313970923423767,
+      "learning_rate": 2.6770947444142197e-05,
+      "loss": 0.2259,
+      "step": 5348
+    },
+    {
+      "epoch": 4.026345502446368,
+      "grad_norm": 0.22628282010555267,
+      "learning_rate": 2.67635107769807e-05,
+      "loss": 0.2533,
+      "step": 5349
+    },
+    {
+      "epoch": 4.027098231087693,
+      "grad_norm": 0.195582315325737,
+      "learning_rate": 2.6756074190360563e-05,
+      "loss": 0.2333,
+      "step": 5350
+    },
+    {
+      "epoch": 4.027850959729018,
+      "grad_norm": 0.22996306419372559,
+      "learning_rate": 2.674863768509504e-05,
+      "loss": 0.1737,
+      "step": 5351
+    },
+    {
+      "epoch": 4.028603688370342,
+      "grad_norm": 0.20911015570163727,
+      "learning_rate": 2.6741201261997367e-05,
+      "loss": 0.1886,
+      "step": 5352
+    },
+    {
+      "epoch": 4.029356417011667,
+      "grad_norm": 0.2323939949274063,
+      "learning_rate": 2.6733764921880783e-05,
+      "loss": 0.2554,
+      "step": 5353
+    },
+    {
+      "epoch": 4.030109145652992,
+      "grad_norm": 0.23296067118644714,
+      "learning_rate": 2.6726328665558515e-05,
+      "loss": 0.2191,
+      "step": 5354
+    },
+    {
+      "epoch": 4.030861874294317,
+      "grad_norm": 0.2496226727962494,
+      "learning_rate": 2.6718892493843777e-05,
+      "loss": 0.2208,
+      "step": 5355
+    },
+    {
+      "epoch": 4.031614602935642,
+      "grad_norm": 0.23186449706554413,
+      "learning_rate": 2.6711456407549784e-05,
+      "loss": 0.169,
+      "step": 5356
+    },
+    {
+      "epoch": 4.032367331576967,
+      "grad_norm": 0.22550742328166962,
+      "learning_rate": 2.670402040748973e-05,
+      "loss": 0.2264,
+      "step": 5357
+    },
+    {
+      "epoch": 4.033120060218291,
+      "grad_norm": 0.24459299445152283,
+      "learning_rate": 2.669658449447682e-05,
+      "loss": 0.252,
+      "step": 5358
+    },
+    {
+      "epoch": 4.033872788859616,
+      "grad_norm": 0.224811851978302,
+      "learning_rate": 2.6689148669324205e-05,
+      "loss": 0.1877,
+      "step": 5359
+    },
+    {
+      "epoch": 4.034625517500941,
+      "grad_norm": 0.21583496034145355,
+      "learning_rate": 2.668171293284508e-05,
+      "loss": 0.174,
+      "step": 5360
+    },
+    {
+      "epoch": 4.035378246142265,
+      "grad_norm": 0.21008022129535675,
+      "learning_rate": 2.66742772858526e-05,
+      "loss": 0.1889,
+      "step": 5361
+    },
+    {
+      "epoch": 4.036130974783591,
+      "grad_norm": 0.27265849709510803,
+      "learning_rate": 2.6666841729159912e-05,
+      "loss": 0.2793,
+      "step": 5362
+    },
+    {
+      "epoch": 4.036883703424915,
+      "grad_norm": 0.22634655237197876,
+      "learning_rate": 2.6659406263580157e-05,
+      "loss": 0.3046,
+      "step": 5363
+    },
+    {
+      "epoch": 4.03763643206624,
+      "grad_norm": 0.22578246891498566,
+      "learning_rate": 2.6651970889926475e-05,
+      "loss": 0.1748,
+      "step": 5364
+    },
+    {
+      "epoch": 4.038389160707565,
+      "grad_norm": 0.1883048117160797,
+      "learning_rate": 2.6644535609011977e-05,
+      "loss": 0.2049,
+      "step": 5365
+    },
+    {
+      "epoch": 4.0391418893488895,
+      "grad_norm": 0.25211232900619507,
+      "learning_rate": 2.6637100421649785e-05,
+      "loss": 0.2439,
+      "step": 5366
+    },
+    {
+      "epoch": 4.039894617990215,
+      "grad_norm": 0.1852383017539978,
+      "learning_rate": 2.6629665328652993e-05,
+      "loss": 0.1849,
+      "step": 5367
+    },
+    {
+      "epoch": 4.040647346631539,
+      "grad_norm": 0.2775876224040985,
+      "learning_rate": 2.662223033083468e-05,
+      "loss": 0.2383,
+      "step": 5368
+    },
+    {
+      "epoch": 4.041400075272864,
+      "grad_norm": 0.19809702038764954,
+      "learning_rate": 2.661479542900795e-05,
+      "loss": 0.2358,
+      "step": 5369
+    },
+    {
+      "epoch": 4.042152803914189,
+      "grad_norm": 0.23724907636642456,
+      "learning_rate": 2.6607360623985854e-05,
+      "loss": 0.1803,
+      "step": 5370
+    },
+    {
+      "epoch": 4.042905532555514,
+      "grad_norm": 0.19988511502742767,
+      "learning_rate": 2.6599925916581465e-05,
+      "loss": 0.2078,
+      "step": 5371
+    },
+    {
+      "epoch": 4.043658261196838,
+      "grad_norm": 0.24882113933563232,
+      "learning_rate": 2.6592491307607826e-05,
+      "loss": 0.17,
+      "step": 5372
+    },
+    {
+      "epoch": 4.0444109898381635,
+      "grad_norm": 0.22134336829185486,
+      "learning_rate": 2.6585056797877965e-05,
+      "loss": 0.2378,
+      "step": 5373
+    },
+    {
+      "epoch": 4.045163718479488,
+      "grad_norm": 0.22578708827495575,
+      "learning_rate": 2.6577622388204925e-05,
+      "loss": 0.2439,
+      "step": 5374
+    },
+    {
+      "epoch": 4.045916447120813,
+      "grad_norm": 0.19618375599384308,
+      "learning_rate": 2.6570188079401704e-05,
+      "loss": 0.2411,
+      "step": 5375
+    },
+    {
+      "epoch": 4.046669175762138,
+      "grad_norm": 0.22396893799304962,
+      "learning_rate": 2.656275387228132e-05,
+      "loss": 0.1664,
+      "step": 5376
+    },
+    {
+      "epoch": 4.047421904403462,
+      "grad_norm": 0.2135220170021057,
+      "learning_rate": 2.6555319767656773e-05,
+      "loss": 0.2287,
+      "step": 5377
+    },
+    {
+      "epoch": 4.048174633044788,
+      "grad_norm": 0.23199652135372162,
+      "learning_rate": 2.654788576634103e-05,
+      "loss": 0.2551,
+      "step": 5378
+    },
+    {
+      "epoch": 4.048927361686112,
+      "grad_norm": 0.20517756044864655,
+      "learning_rate": 2.6540451869147064e-05,
+      "loss": 0.1342,
+      "step": 5379
+    },
+    {
+      "epoch": 4.049680090327437,
+      "grad_norm": 0.22914542257785797,
+      "learning_rate": 2.6533018076887838e-05,
+      "loss": 0.2365,
+      "step": 5380
+    },
+    {
+      "epoch": 4.050432818968762,
+      "grad_norm": 0.24591712653636932,
+      "learning_rate": 2.6525584390376308e-05,
+      "loss": 0.2998,
+      "step": 5381
+    },
+    {
+      "epoch": 4.051185547610086,
+      "grad_norm": 0.20129792392253876,
+      "learning_rate": 2.6518150810425396e-05,
+      "loss": 0.2756,
+      "step": 5382
+    },
+    {
+      "epoch": 4.051938276251412,
+      "grad_norm": 0.22989122569561005,
+      "learning_rate": 2.6510717337848046e-05,
+      "loss": 0.2086,
+      "step": 5383
+    },
+    {
+      "epoch": 4.052691004892736,
+      "grad_norm": 0.22700068354606628,
+      "learning_rate": 2.650328397345715e-05,
+      "loss": 0.2555,
+      "step": 5384
+    },
+    {
+      "epoch": 4.053443733534061,
+      "grad_norm": 0.20687909424304962,
+      "learning_rate": 2.6495850718065623e-05,
+      "loss": 0.1804,
+      "step": 5385
+    },
+    {
+      "epoch": 4.054196462175386,
+      "grad_norm": 0.19257673621177673,
+      "learning_rate": 2.6488417572486355e-05,
+      "loss": 0.1913,
+      "step": 5386
+    },
+    {
+      "epoch": 4.054949190816711,
+      "grad_norm": 0.24506345391273499,
+      "learning_rate": 2.6480984537532217e-05,
+      "loss": 0.2215,
+      "step": 5387
+    },
+    {
+      "epoch": 4.055701919458035,
+      "grad_norm": 0.22615478932857513,
+      "learning_rate": 2.647355161401608e-05,
+      "loss": 0.2058,
+      "step": 5388
+    },
+    {
+      "epoch": 4.05645464809936,
+      "grad_norm": 0.21449294686317444,
+      "learning_rate": 2.6466118802750793e-05,
+      "loss": 0.1788,
+      "step": 5389
+    },
+    {
+      "epoch": 4.057207376740685,
+      "grad_norm": 0.20783670246601105,
+      "learning_rate": 2.645868610454921e-05,
+      "loss": 0.2325,
+      "step": 5390
+    },
+    {
+      "epoch": 4.057960105382009,
+      "grad_norm": 0.22869667410850525,
+      "learning_rate": 2.645125352022414e-05,
+      "loss": 0.1985,
+      "step": 5391
+    },
+    {
+      "epoch": 4.058712834023335,
+      "grad_norm": 0.23037059605121613,
+      "learning_rate": 2.644382105058841e-05,
+      "loss": 0.1901,
+      "step": 5392
+    },
+    {
+      "epoch": 4.059465562664659,
+      "grad_norm": 0.20913167297840118,
+      "learning_rate": 2.6436388696454823e-05,
+      "loss": 0.1977,
+      "step": 5393
+    },
+    {
+      "epoch": 4.0602182913059846,
+      "grad_norm": 0.23527897894382477,
+      "learning_rate": 2.6428956458636157e-05,
+      "loss": 0.2557,
+      "step": 5394
+    },
+    {
+      "epoch": 4.060971019947309,
+      "grad_norm": 0.18265099823474884,
+      "learning_rate": 2.6421524337945214e-05,
+      "loss": 0.2162,
+      "step": 5395
+    },
+    {
+      "epoch": 4.0617237485886335,
+      "grad_norm": 0.19819992780685425,
+      "learning_rate": 2.6414092335194745e-05,
+      "loss": 0.2263,
+      "step": 5396
+    },
+    {
+      "epoch": 4.062476477229959,
+      "grad_norm": 0.2196228951215744,
+      "learning_rate": 2.6406660451197513e-05,
+      "loss": 0.1981,
+      "step": 5397
+    },
+    {
+      "epoch": 4.063229205871283,
+      "grad_norm": 0.22567912936210632,
+      "learning_rate": 2.6399228686766243e-05,
+      "loss": 0.2595,
+      "step": 5398
+    },
+    {
+      "epoch": 4.063981934512608,
+      "grad_norm": 0.186574324965477,
+      "learning_rate": 2.639179704271367e-05,
+      "loss": 0.1928,
+      "step": 5399
+    },
+    {
+      "epoch": 4.064734663153933,
+      "grad_norm": 0.22511599957942963,
+      "learning_rate": 2.63843655198525e-05,
+      "loss": 0.2499,
+      "step": 5400
+    },
+    {
+      "epoch": 4.064734663153933,
+      "eval_loss": 0.24411219358444214,
+      "eval_runtime": 452.7779,
+      "eval_samples_per_second": 21.262,
+      "eval_steps_per_second": 0.665,
+      "step": 5400
+    },
+    {
+      "epoch": 4.065487391795258,
+      "grad_norm": 0.2173110395669937,
+      "learning_rate": 2.6376934118995445e-05,
+      "loss": 0.2204,
+      "step": 5401
+    },
+    {
+      "epoch": 4.066240120436583,
+      "grad_norm": 0.19071510434150696,
+      "learning_rate": 2.636950284095519e-05,
+      "loss": 0.1683,
+      "step": 5402
+    },
+    {
+      "epoch": 4.0669928490779075,
+      "grad_norm": 0.2246982753276825,
+      "learning_rate": 2.6362071686544405e-05,
+      "loss": 0.2335,
+      "step": 5403
+    },
+    {
+      "epoch": 4.067745577719232,
+      "grad_norm": 0.16450801491737366,
+      "learning_rate": 2.6354640656575742e-05,
+      "loss": 0.1206,
+      "step": 5404
+    },
+    {
+      "epoch": 4.068498306360557,
+      "grad_norm": 0.227121502161026,
+      "learning_rate": 2.6347209751861855e-05,
+      "loss": 0.231,
+      "step": 5405
+    },
+    {
+      "epoch": 4.069251035001882,
+      "grad_norm": 0.19902415573596954,
+      "learning_rate": 2.633977897321538e-05,
+      "loss": 0.217,
+      "step": 5406
+    },
+    {
+      "epoch": 4.070003763643206,
+      "grad_norm": 0.20334352552890778,
+      "learning_rate": 2.6332348321448936e-05,
+      "loss": 0.2319,
+      "step": 5407
+    },
+    {
+      "epoch": 4.070756492284532,
+      "grad_norm": 0.2252737283706665,
+      "learning_rate": 2.6324917797375115e-05,
+      "loss": 0.263,
+      "step": 5408
+    },
+    {
+      "epoch": 4.071509220925856,
+      "grad_norm": 0.2265598028898239,
+      "learning_rate": 2.6317487401806522e-05,
+      "loss": 0.2538,
+      "step": 5409
+    },
+    {
+      "epoch": 4.0722619495671815,
+      "grad_norm": 0.19219769537448883,
+      "learning_rate": 2.6310057135555734e-05,
+      "loss": 0.2272,
+      "step": 5410
+    },
+    {
+      "epoch": 4.073014678208506,
+      "grad_norm": 0.19124756753444672,
+      "learning_rate": 2.6302626999435305e-05,
+      "loss": 0.2116,
+      "step": 5411
+    },
+    {
+      "epoch": 4.07376740684983,
+      "grad_norm": 0.21937568485736847,
+      "learning_rate": 2.629519699425778e-05,
+      "loss": 0.3042,
+      "step": 5412
+    },
+    {
+      "epoch": 4.074520135491156,
+      "grad_norm": 0.2304798662662506,
+      "learning_rate": 2.62877671208357e-05,
+      "loss": 0.2136,
+      "step": 5413
+    },
+    {
+      "epoch": 4.07527286413248,
+      "grad_norm": 0.1996319591999054,
+      "learning_rate": 2.6280337379981597e-05,
+      "loss": 0.2886,
+      "step": 5414
+    },
+    {
+      "epoch": 4.076025592773805,
+      "grad_norm": 0.2318170666694641,
+      "learning_rate": 2.6272907772507956e-05,
+      "loss": 0.1326,
+      "step": 5415
+    },
+    {
+      "epoch": 4.07677832141513,
+      "grad_norm": 0.20550476014614105,
+      "learning_rate": 2.626547829922728e-05,
+      "loss": 0.1804,
+      "step": 5416
+    },
+    {
+      "epoch": 4.0775310500564546,
+      "grad_norm": 0.2368755042552948,
+      "learning_rate": 2.625804896095204e-05,
+      "loss": 0.2589,
+      "step": 5417
+    },
+    {
+      "epoch": 4.078283778697779,
+      "grad_norm": 0.20753087103366852,
+      "learning_rate": 2.6250619758494705e-05,
+      "loss": 0.119,
+      "step": 5418
+    },
+    {
+      "epoch": 4.079036507339104,
+      "grad_norm": 0.2412748783826828,
+      "learning_rate": 2.6243190692667697e-05,
+      "loss": 0.2247,
+      "step": 5419
+    },
+    {
+      "epoch": 4.079789235980429,
+      "grad_norm": 0.24350044131278992,
+      "learning_rate": 2.623576176428348e-05,
+      "loss": 0.2357,
+      "step": 5420
+    },
+    {
+      "epoch": 4.080541964621754,
+      "grad_norm": 0.23335514962673187,
+      "learning_rate": 2.622833297415445e-05,
+      "loss": 0.2732,
+      "step": 5421
+    },
+    {
+      "epoch": 4.081294693263079,
+      "grad_norm": 0.26887762546539307,
+      "learning_rate": 2.622090432309302e-05,
+      "loss": 0.2022,
+      "step": 5422
+    },
+    {
+      "epoch": 4.082047421904403,
+      "grad_norm": 0.25607192516326904,
+      "learning_rate": 2.6213475811911564e-05,
+      "loss": 0.2321,
+      "step": 5423
+    },
+    {
+      "epoch": 4.0828001505457285,
+      "grad_norm": 0.29186755418777466,
+      "learning_rate": 2.6206047441422467e-05,
+      "loss": 0.2027,
+      "step": 5424
+    },
+    {
+      "epoch": 4.083552879187053,
+      "grad_norm": 0.21473407745361328,
+      "learning_rate": 2.6198619212438064e-05,
+      "loss": 0.1966,
+      "step": 5425
+    },
+    {
+      "epoch": 4.0843056078283775,
+      "grad_norm": 0.23761671781539917,
+      "learning_rate": 2.6191191125770708e-05,
+      "loss": 0.2074,
+      "step": 5426
+    },
+    {
+      "epoch": 4.085058336469703,
+      "grad_norm": 0.23860302567481995,
+      "learning_rate": 2.6183763182232735e-05,
+      "loss": 0.3125,
+      "step": 5427
+    },
+    {
+      "epoch": 4.085811065111027,
+      "grad_norm": 0.20504198968410492,
+      "learning_rate": 2.6176335382636435e-05,
+      "loss": 0.2391,
+      "step": 5428
+    },
+    {
+      "epoch": 4.086563793752353,
+      "grad_norm": 0.25938278436660767,
+      "learning_rate": 2.6168907727794102e-05,
+      "loss": 0.1698,
+      "step": 5429
+    },
+    {
+      "epoch": 4.087316522393677,
+      "grad_norm": 0.21528281271457672,
+      "learning_rate": 2.6161480218518026e-05,
+      "loss": 0.2026,
+      "step": 5430
+    },
+    {
+      "epoch": 4.088069251035002,
+      "grad_norm": 0.2700593173503876,
+      "learning_rate": 2.6154052855620458e-05,
+      "loss": 0.188,
+      "step": 5431
+    },
+    {
+      "epoch": 4.088821979676327,
+      "grad_norm": 0.23982396721839905,
+      "learning_rate": 2.6146625639913645e-05,
+      "loss": 0.2396,
+      "step": 5432
+    },
+    {
+      "epoch": 4.0895747083176515,
+      "grad_norm": 0.2355031967163086,
+      "learning_rate": 2.6139198572209818e-05,
+      "loss": 0.1958,
+      "step": 5433
+    },
+    {
+      "epoch": 4.090327436958976,
+      "grad_norm": 0.22359706461429596,
+      "learning_rate": 2.613177165332119e-05,
+      "loss": 0.1829,
+      "step": 5434
+    },
+    {
+      "epoch": 4.091080165600301,
+      "grad_norm": 0.19706298410892487,
+      "learning_rate": 2.612434488405996e-05,
+      "loss": 0.1946,
+      "step": 5435
+    },
+    {
+      "epoch": 4.091832894241626,
+      "grad_norm": 0.19623975455760956,
+      "learning_rate": 2.6116918265238305e-05,
+      "loss": 0.1782,
+      "step": 5436
+    },
+    {
+      "epoch": 4.09258562288295,
+      "grad_norm": 0.24018622934818268,
+      "learning_rate": 2.6109491797668385e-05,
+      "loss": 0.1932,
+      "step": 5437
+    },
+    {
+      "epoch": 4.093338351524276,
+      "grad_norm": 0.1895081102848053,
+      "learning_rate": 2.6102065482162348e-05,
+      "loss": 0.1982,
+      "step": 5438
+    },
+    {
+      "epoch": 4.0940910801656,
+      "grad_norm": 0.24231421947479248,
+      "learning_rate": 2.6094639319532333e-05,
+      "loss": 0.301,
+      "step": 5439
+    },
+    {
+      "epoch": 4.0948438088069254,
+      "grad_norm": 0.23438267409801483,
+      "learning_rate": 2.6087213310590453e-05,
+      "loss": 0.2755,
+      "step": 5440
+    },
+    {
+      "epoch": 4.09559653744825,
+      "grad_norm": 0.2244761735200882,
+      "learning_rate": 2.6079787456148796e-05,
+      "loss": 0.3053,
+      "step": 5441
+    },
+    {
+      "epoch": 4.096349266089574,
+      "grad_norm": 0.21384525299072266,
+      "learning_rate": 2.607236175701945e-05,
+      "loss": 0.1927,
+      "step": 5442
+    },
+    {
+      "epoch": 4.0971019947309,
+      "grad_norm": 0.20805636048316956,
+      "learning_rate": 2.606493621401448e-05,
+      "loss": 0.3541,
+      "step": 5443
+    },
+    {
+      "epoch": 4.097854723372224,
+      "grad_norm": 0.24034473299980164,
+      "learning_rate": 2.605751082794592e-05,
+      "loss": 0.2086,
+      "step": 5444
+    },
+    {
+      "epoch": 4.098607452013549,
+      "grad_norm": 0.1809123456478119,
+      "learning_rate": 2.60500855996258e-05,
+      "loss": 0.1971,
+      "step": 5445
+    },
+    {
+      "epoch": 4.099360180654874,
+      "grad_norm": 0.24260926246643066,
+      "learning_rate": 2.6042660529866154e-05,
+      "loss": 0.1501,
+      "step": 5446
+    },
+    {
+      "epoch": 4.1001129092961985,
+      "grad_norm": 0.19397364556789398,
+      "learning_rate": 2.603523561947896e-05,
+      "loss": 0.304,
+      "step": 5447
+    },
+    {
+      "epoch": 4.100865637937524,
+      "grad_norm": 0.2390350103378296,
+      "learning_rate": 2.602781086927619e-05,
+      "loss": 0.2883,
+      "step": 5448
+    },
+    {
+      "epoch": 4.101618366578848,
+      "grad_norm": 0.21091680228710175,
+      "learning_rate": 2.6020386280069815e-05,
+      "loss": 0.3577,
+      "step": 5449
+    },
+    {
+      "epoch": 4.102371095220173,
+      "grad_norm": 0.21623577177524567,
+      "learning_rate": 2.6012961852671767e-05,
+      "loss": 0.266,
+      "step": 5450
+    },
+    {
+      "epoch": 4.103123823861498,
+      "grad_norm": 0.21346771717071533,
+      "learning_rate": 2.6005537587893976e-05,
+      "loss": 0.189,
+      "step": 5451
+    },
+    {
+      "epoch": 4.103876552502823,
+      "grad_norm": 0.22665832936763763,
+      "learning_rate": 2.599811348654835e-05,
+      "loss": 0.198,
+      "step": 5452
+    },
+    {
+      "epoch": 4.104629281144147,
+      "grad_norm": 0.21718508005142212,
+      "learning_rate": 2.5990689549446774e-05,
+      "loss": 0.1939,
+      "step": 5453
+    },
+    {
+      "epoch": 4.1053820097854725,
+      "grad_norm": 0.20182381570339203,
+      "learning_rate": 2.5983265777401116e-05,
+      "loss": 0.2301,
+      "step": 5454
+    },
+    {
+      "epoch": 4.106134738426797,
+      "grad_norm": 0.1856013536453247,
+      "learning_rate": 2.5975842171223236e-05,
+      "loss": 0.1891,
+      "step": 5455
+    },
+    {
+      "epoch": 4.106887467068122,
+      "grad_norm": 0.22902844846248627,
+      "learning_rate": 2.5968418731724962e-05,
+      "loss": 0.2489,
+      "step": 5456
+    },
+    {
+      "epoch": 4.107640195709447,
+      "grad_norm": 0.2187919020652771,
+      "learning_rate": 2.5960995459718117e-05,
+      "loss": 0.2198,
+      "step": 5457
+    },
+    {
+      "epoch": 4.108392924350771,
+      "grad_norm": 0.28724345564842224,
+      "learning_rate": 2.595357235601449e-05,
+      "loss": 0.3385,
+      "step": 5458
+    },
+    {
+      "epoch": 4.109145652992097,
+      "grad_norm": 0.20941251516342163,
+      "learning_rate": 2.594614942142587e-05,
+      "loss": 0.2337,
+      "step": 5459
+    },
+    {
+      "epoch": 4.109898381633421,
+      "grad_norm": 0.23638838529586792,
+      "learning_rate": 2.593872665676401e-05,
+      "loss": 0.2813,
+      "step": 5460
+    },
+    {
+      "epoch": 4.110651110274746,
+      "grad_norm": 0.21866358816623688,
+      "learning_rate": 2.593130406284066e-05,
+      "loss": 0.248,
+      "step": 5461
+    },
+    {
+      "epoch": 4.111403838916071,
+      "grad_norm": 0.2639898955821991,
+      "learning_rate": 2.592388164046754e-05,
+      "loss": 0.1903,
+      "step": 5462
+    },
+    {
+      "epoch": 4.1121565675573954,
+      "grad_norm": 0.21715474128723145,
+      "learning_rate": 2.591645939045635e-05,
+      "loss": 0.2243,
+      "step": 5463
+    },
+    {
+      "epoch": 4.112909296198721,
+      "grad_norm": 0.2284434586763382,
+      "learning_rate": 2.5909037313618773e-05,
+      "loss": 0.1527,
+      "step": 5464
+    },
+    {
+      "epoch": 4.113662024840045,
+      "grad_norm": 0.2047313153743744,
+      "learning_rate": 2.590161541076649e-05,
+      "loss": 0.1794,
+      "step": 5465
+    },
+    {
+      "epoch": 4.11441475348137,
+      "grad_norm": 0.2008877694606781,
+      "learning_rate": 2.589419368271115e-05,
+      "loss": 0.1898,
+      "step": 5466
+    },
+    {
+      "epoch": 4.115167482122695,
+      "grad_norm": 0.2126656025648117,
+      "learning_rate": 2.5886772130264364e-05,
+      "loss": 0.2083,
+      "step": 5467
+    },
+    {
+      "epoch": 4.11592021076402,
+      "grad_norm": 0.21376319229602814,
+      "learning_rate": 2.587935075423776e-05,
+      "loss": 0.2033,
+      "step": 5468
+    },
+    {
+      "epoch": 4.116672939405344,
+      "grad_norm": 0.2326332926750183,
+      "learning_rate": 2.587192955544292e-05,
+      "loss": 0.1875,
+      "step": 5469
+    },
+    {
+      "epoch": 4.117425668046669,
+      "grad_norm": 0.24432894587516785,
+      "learning_rate": 2.58645085346914e-05,
+      "loss": 0.283,
+      "step": 5470
+    },
+    {
+      "epoch": 4.118178396687994,
+      "grad_norm": 0.222112238407135,
+      "learning_rate": 2.5857087692794785e-05,
+      "loss": 0.2503,
+      "step": 5471
+    },
+    {
+      "epoch": 4.118931125329318,
+      "grad_norm": 0.2500867247581482,
+      "learning_rate": 2.5849667030564585e-05,
+      "loss": 0.3382,
+      "step": 5472
+    },
+    {
+      "epoch": 4.119683853970644,
+      "grad_norm": 0.2149982899427414,
+      "learning_rate": 2.5842246548812322e-05,
+      "loss": 0.2251,
+      "step": 5473
+    },
+    {
+      "epoch": 4.120436582611968,
+      "grad_norm": 0.21632122993469238,
+      "learning_rate": 2.583482624834948e-05,
+      "loss": 0.1841,
+      "step": 5474
+    },
+    {
+      "epoch": 4.121189311253294,
+      "grad_norm": 0.21565967798233032,
+      "learning_rate": 2.5827406129987535e-05,
+      "loss": 0.1225,
+      "step": 5475
+    },
+    {
+      "epoch": 4.121942039894618,
+      "grad_norm": 0.22779974341392517,
+      "learning_rate": 2.5819986194537944e-05,
+      "loss": 0.2983,
+      "step": 5476
+    },
+    {
+      "epoch": 4.1226947685359425,
+      "grad_norm": 0.19335216283798218,
+      "learning_rate": 2.5812566442812126e-05,
+      "loss": 0.2737,
+      "step": 5477
+    },
+    {
+      "epoch": 4.123447497177268,
+      "grad_norm": 0.2133839726448059,
+      "learning_rate": 2.5805146875621516e-05,
+      "loss": 0.2072,
+      "step": 5478
+    },
+    {
+      "epoch": 4.124200225818592,
+      "grad_norm": 0.20405976474285126,
+      "learning_rate": 2.579772749377749e-05,
+      "loss": 0.1859,
+      "step": 5479
+    },
+    {
+      "epoch": 4.124952954459917,
+      "grad_norm": 0.2253781110048294,
+      "learning_rate": 2.579030829809143e-05,
+      "loss": 0.1964,
+      "step": 5480
+    },
+    {
+      "epoch": 4.125705683101242,
+      "grad_norm": 0.2258410006761551,
+      "learning_rate": 2.578288928937469e-05,
+      "loss": 0.1527,
+      "step": 5481
+    },
+    {
+      "epoch": 4.126458411742567,
+      "grad_norm": 0.22383947670459747,
+      "learning_rate": 2.5775470468438583e-05,
+      "loss": 0.2025,
+      "step": 5482
+    },
+    {
+      "epoch": 4.127211140383892,
+      "grad_norm": 0.18018251657485962,
+      "learning_rate": 2.5768051836094432e-05,
+      "loss": 0.2464,
+      "step": 5483
+    },
+    {
+      "epoch": 4.1279638690252165,
+      "grad_norm": 0.22696569561958313,
+      "learning_rate": 2.576063339315354e-05,
+      "loss": 0.2191,
+      "step": 5484
+    },
+    {
+      "epoch": 4.128716597666541,
+      "grad_norm": 0.2131880223751068,
+      "learning_rate": 2.5753215140427157e-05,
+      "loss": 0.2136,
+      "step": 5485
+    },
+    {
+      "epoch": 4.129469326307866,
+      "grad_norm": 0.21972979605197906,
+      "learning_rate": 2.5745797078726547e-05,
+      "loss": 0.2046,
+      "step": 5486
+    },
+    {
+      "epoch": 4.130222054949191,
+      "grad_norm": 0.24853500723838806,
+      "learning_rate": 2.573837920886293e-05,
+      "loss": 0.3256,
+      "step": 5487
+    },
+    {
+      "epoch": 4.130974783590515,
+      "grad_norm": 0.1999872624874115,
+      "learning_rate": 2.5730961531647512e-05,
+      "loss": 0.1904,
+      "step": 5488
+    },
+    {
+      "epoch": 4.131727512231841,
+      "grad_norm": 0.16997621953487396,
+      "learning_rate": 2.5723544047891472e-05,
+      "loss": 0.1849,
+      "step": 5489
+    },
+    {
+      "epoch": 4.132480240873165,
+      "grad_norm": 0.2042304128408432,
+      "learning_rate": 2.571612675840599e-05,
+      "loss": 0.27,
+      "step": 5490
+    },
+    {
+      "epoch": 4.13323296951449,
+      "grad_norm": 0.19657304883003235,
+      "learning_rate": 2.570870966400221e-05,
+      "loss": 0.191,
+      "step": 5491
+    },
+    {
+      "epoch": 4.133985698155815,
+      "grad_norm": 0.2345583587884903,
+      "learning_rate": 2.5701292765491242e-05,
+      "loss": 0.2746,
+      "step": 5492
+    },
+    {
+      "epoch": 4.134738426797139,
+      "grad_norm": 0.20873737335205078,
+      "learning_rate": 2.56938760636842e-05,
+      "loss": 0.1702,
+      "step": 5493
+    },
+    {
+      "epoch": 4.135491155438465,
+      "grad_norm": 0.22368399798870087,
+      "learning_rate": 2.5686459559392153e-05,
+      "loss": 0.1693,
+      "step": 5494
+    },
+    {
+      "epoch": 4.136243884079789,
+      "grad_norm": 0.20590263605117798,
+      "learning_rate": 2.567904325342616e-05,
+      "loss": 0.2584,
+      "step": 5495
+    },
+    {
+      "epoch": 4.136996612721114,
+      "grad_norm": 0.22690394520759583,
+      "learning_rate": 2.567162714659725e-05,
+      "loss": 0.1756,
+      "step": 5496
+    },
+    {
+      "epoch": 4.137749341362439,
+      "grad_norm": 0.19692832231521606,
+      "learning_rate": 2.5664211239716457e-05,
+      "loss": 0.2054,
+      "step": 5497
+    },
+    {
+      "epoch": 4.138502070003764,
+      "grad_norm": 0.19609478116035461,
+      "learning_rate": 2.5656795533594762e-05,
+      "loss": 0.2215,
+      "step": 5498
+    },
+    {
+      "epoch": 4.139254798645088,
+      "grad_norm": 0.22499965131282806,
+      "learning_rate": 2.5649380029043134e-05,
+      "loss": 0.2261,
+      "step": 5499
+    },
+    {
+      "epoch": 4.140007527286413,
+      "grad_norm": 0.21955087780952454,
+      "learning_rate": 2.5641964726872526e-05,
+      "loss": 0.2221,
+      "step": 5500
+    },
+    {
+      "epoch": 4.140760255927738,
+      "grad_norm": 0.21147918701171875,
+      "learning_rate": 2.5634549627893863e-05,
+      "loss": 0.2422,
+      "step": 5501
+    },
+    {
+      "epoch": 4.141512984569063,
+      "grad_norm": 0.20912273228168488,
+      "learning_rate": 2.562713473291804e-05,
+      "loss": 0.1633,
+      "step": 5502
+    },
+    {
+      "epoch": 4.142265713210388,
+      "grad_norm": 0.20321862399578094,
+      "learning_rate": 2.5619720042755956e-05,
+      "loss": 0.2917,
+      "step": 5503
+    },
+    {
+      "epoch": 4.143018441851712,
+      "grad_norm": 0.2264932543039322,
+      "learning_rate": 2.561230555821846e-05,
+      "loss": 0.1964,
+      "step": 5504
+    },
+    {
+      "epoch": 4.143771170493038,
+      "grad_norm": 0.18295875191688538,
+      "learning_rate": 2.5604891280116382e-05,
+      "loss": 0.2896,
+      "step": 5505
+    },
+    {
+      "epoch": 4.144523899134362,
+      "grad_norm": 0.2289116382598877,
+      "learning_rate": 2.559747720926055e-05,
+      "loss": 0.2873,
+      "step": 5506
+    },
+    {
+      "epoch": 4.1452766277756865,
+      "grad_norm": 0.22545306384563446,
+      "learning_rate": 2.5590063346461757e-05,
+      "loss": 0.2246,
+      "step": 5507
+    },
+    {
+      "epoch": 4.146029356417012,
+      "grad_norm": 0.18258336186408997,
+      "learning_rate": 2.5582649692530748e-05,
+      "loss": 0.2172,
+      "step": 5508
+    },
+    {
+      "epoch": 4.146782085058336,
+      "grad_norm": 0.21743062138557434,
+      "learning_rate": 2.55752362482783e-05,
+      "loss": 0.2201,
+      "step": 5509
+    },
+    {
+      "epoch": 4.147534813699661,
+      "grad_norm": 0.24092546105384827,
+      "learning_rate": 2.5567823014515118e-05,
+      "loss": 0.2595,
+      "step": 5510
+    },
+    {
+      "epoch": 4.148287542340986,
+      "grad_norm": 0.19675639271736145,
+      "learning_rate": 2.5560409992051908e-05,
+      "loss": 0.1577,
+      "step": 5511
+    },
+    {
+      "epoch": 4.149040270982311,
+      "grad_norm": 0.20007038116455078,
+      "learning_rate": 2.5552997181699344e-05,
+      "loss": 0.1987,
+      "step": 5512
+    },
+    {
+      "epoch": 4.149792999623636,
+      "grad_norm": 0.24475352466106415,
+      "learning_rate": 2.5545584584268085e-05,
+      "loss": 0.2061,
+      "step": 5513
+    },
+    {
+      "epoch": 4.1505457282649605,
+      "grad_norm": 0.2025049775838852,
+      "learning_rate": 2.5538172200568754e-05,
+      "loss": 0.2825,
+      "step": 5514
+    },
+    {
+      "epoch": 4.151298456906285,
+      "grad_norm": 0.20557309687137604,
+      "learning_rate": 2.553076003141195e-05,
+      "loss": 0.1775,
+      "step": 5515
+    },
+    {
+      "epoch": 4.15205118554761,
+      "grad_norm": 0.2248924970626831,
+      "learning_rate": 2.5523348077608284e-05,
+      "loss": 0.2076,
+      "step": 5516
+    },
+    {
+      "epoch": 4.152803914188935,
+      "grad_norm": 0.21168211102485657,
+      "learning_rate": 2.5515936339968294e-05,
+      "loss": 0.1841,
+      "step": 5517
+    },
+    {
+      "epoch": 4.153556642830259,
+      "grad_norm": 0.20625008642673492,
+      "learning_rate": 2.5508524819302522e-05,
+      "loss": 0.201,
+      "step": 5518
+    },
+    {
+      "epoch": 4.154309371471585,
+      "grad_norm": 0.2152377814054489,
+      "learning_rate": 2.5501113516421488e-05,
+      "loss": 0.2497,
+      "step": 5519
+    },
+    {
+      "epoch": 4.155062100112909,
+      "grad_norm": 0.20805463194847107,
+      "learning_rate": 2.549370243213567e-05,
+      "loss": 0.1513,
+      "step": 5520
+    },
+    {
+      "epoch": 4.1558148287542345,
+      "grad_norm": 0.2483224868774414,
+      "learning_rate": 2.548629156725553e-05,
+      "loss": 0.2175,
+      "step": 5521
+    },
+    {
+      "epoch": 4.156567557395559,
+      "grad_norm": 0.2198197841644287,
+      "learning_rate": 2.5478880922591525e-05,
+      "loss": 0.2813,
+      "step": 5522
+    },
+    {
+      "epoch": 4.157320286036883,
+      "grad_norm": 0.20657196640968323,
+      "learning_rate": 2.5471470498954064e-05,
+      "loss": 0.163,
+      "step": 5523
+    },
+    {
+      "epoch": 4.158073014678209,
+      "grad_norm": 0.21509341895580292,
+      "learning_rate": 2.546406029715354e-05,
+      "loss": 0.2147,
+      "step": 5524
+    },
+    {
+      "epoch": 4.158825743319533,
+      "grad_norm": 0.16622892022132874,
+      "learning_rate": 2.5456650318000312e-05,
+      "loss": 0.1435,
+      "step": 5525
+    },
+    {
+      "epoch": 4.159578471960858,
+      "grad_norm": 0.220582976937294,
+      "learning_rate": 2.5449240562304743e-05,
+      "loss": 0.1891,
+      "step": 5526
+    },
+    {
+      "epoch": 4.160331200602183,
+      "grad_norm": 0.18690885603427887,
+      "learning_rate": 2.544183103087714e-05,
+      "loss": 0.2163,
+      "step": 5527
+    },
+    {
+      "epoch": 4.161083929243508,
+      "grad_norm": 0.24720481038093567,
+      "learning_rate": 2.54344217245278e-05,
+      "loss": 0.25,
+      "step": 5528
+    },
+    {
+      "epoch": 4.161836657884833,
+      "grad_norm": 0.19398590922355652,
+      "learning_rate": 2.5427012644066996e-05,
+      "loss": 0.1502,
+      "step": 5529
+    },
+    {
+      "epoch": 4.162589386526157,
+      "grad_norm": 0.20858152210712433,
+      "learning_rate": 2.5419603790304973e-05,
+      "loss": 0.2319,
+      "step": 5530
+    },
+    {
+      "epoch": 4.163342115167482,
+      "grad_norm": 0.1896049827337265,
+      "learning_rate": 2.5412195164051956e-05,
+      "loss": 0.2287,
+      "step": 5531
+    },
+    {
+      "epoch": 4.164094843808807,
+      "grad_norm": 0.2145184874534607,
+      "learning_rate": 2.5404786766118132e-05,
+      "loss": 0.1968,
+      "step": 5532
+    },
+    {
+      "epoch": 4.164847572450132,
+      "grad_norm": 0.21139608323574066,
+      "learning_rate": 2.5397378597313677e-05,
+      "loss": 0.1953,
+      "step": 5533
+    },
+    {
+      "epoch": 4.165600301091456,
+      "grad_norm": 0.2107645869255066,
+      "learning_rate": 2.538997065844873e-05,
+      "loss": 0.2322,
+      "step": 5534
+    },
+    {
+      "epoch": 4.1663530297327815,
+      "grad_norm": 0.1941271424293518,
+      "learning_rate": 2.5382562950333426e-05,
+      "loss": 0.1996,
+      "step": 5535
+    },
+    {
+      "epoch": 4.167105758374106,
+      "grad_norm": 0.2163994163274765,
+      "learning_rate": 2.5375155473777858e-05,
+      "loss": 0.2089,
+      "step": 5536
+    },
+    {
+      "epoch": 4.167858487015431,
+      "grad_norm": 0.2041962891817093,
+      "learning_rate": 2.536774822959209e-05,
+      "loss": 0.171,
+      "step": 5537
+    },
+    {
+      "epoch": 4.168611215656756,
+      "grad_norm": 0.2831268310546875,
+      "learning_rate": 2.536034121858617e-05,
+      "loss": 0.2001,
+      "step": 5538
+    },
+    {
+      "epoch": 4.16936394429808,
+      "grad_norm": 0.2011781483888626,
+      "learning_rate": 2.5352934441570126e-05,
+      "loss": 0.2508,
+      "step": 5539
+    },
+    {
+      "epoch": 4.170116672939406,
+      "grad_norm": 0.23279476165771484,
+      "learning_rate": 2.5345527899353922e-05,
+      "loss": 0.184,
+      "step": 5540
+    },
+    {
+      "epoch": 4.17086940158073,
+      "grad_norm": 0.2140761762857437,
+      "learning_rate": 2.533812159274756e-05,
+      "loss": 0.2086,
+      "step": 5541
+    },
+    {
+      "epoch": 4.171622130222055,
+      "grad_norm": 0.24619030952453613,
+      "learning_rate": 2.533071552256097e-05,
+      "loss": 0.2165,
+      "step": 5542
+    },
+    {
+      "epoch": 4.17237485886338,
+      "grad_norm": 0.2419048547744751,
+      "learning_rate": 2.5323309689604074e-05,
+      "loss": 0.2776,
+      "step": 5543
+    },
+    {
+      "epoch": 4.1731275875047045,
+      "grad_norm": 0.2613905370235443,
+      "learning_rate": 2.5315904094686754e-05,
+      "loss": 0.2589,
+      "step": 5544
+    },
+    {
+      "epoch": 4.173880316146029,
+      "grad_norm": 0.16703461110591888,
+      "learning_rate": 2.530849873861888e-05,
+      "loss": 0.1501,
+      "step": 5545
+    },
+    {
+      "epoch": 4.174633044787354,
+      "grad_norm": 0.21907638013362885,
+      "learning_rate": 2.5301093622210283e-05,
+      "loss": 0.2201,
+      "step": 5546
+    },
+    {
+      "epoch": 4.175385773428679,
+      "grad_norm": 0.23806403577327728,
+      "learning_rate": 2.5293688746270784e-05,
+      "loss": 0.1583,
+      "step": 5547
+    },
+    {
+      "epoch": 4.176138502070004,
+      "grad_norm": 0.24112214148044586,
+      "learning_rate": 2.528628411161017e-05,
+      "loss": 0.2017,
+      "step": 5548
+    },
+    {
+      "epoch": 4.176891230711329,
+      "grad_norm": 0.19485877454280853,
+      "learning_rate": 2.5278879719038197e-05,
+      "loss": 0.2182,
+      "step": 5549
+    },
+    {
+      "epoch": 4.177643959352653,
+      "grad_norm": 0.2376643419265747,
+      "learning_rate": 2.527147556936459e-05,
+      "loss": 0.2348,
+      "step": 5550
+    },
+    {
+      "epoch": 4.1783966879939785,
+      "grad_norm": 0.2324780970811844,
+      "learning_rate": 2.5264071663399076e-05,
+      "loss": 0.1876,
+      "step": 5551
+    },
+    {
+      "epoch": 4.179149416635303,
+      "grad_norm": 0.20400792360305786,
+      "learning_rate": 2.5256668001951317e-05,
+      "loss": 0.2228,
+      "step": 5552
+    },
+    {
+      "epoch": 4.179902145276627,
+      "grad_norm": 0.2336265593767166,
+      "learning_rate": 2.5249264585830972e-05,
+      "loss": 0.2318,
+      "step": 5553
+    },
+    {
+      "epoch": 4.180654873917953,
+      "grad_norm": 0.2164028286933899,
+      "learning_rate": 2.524186141584766e-05,
+      "loss": 0.1997,
+      "step": 5554
+    },
+    {
+      "epoch": 4.181407602559277,
+      "grad_norm": 0.23236054182052612,
+      "learning_rate": 2.5234458492811e-05,
+      "loss": 0.1667,
+      "step": 5555
+    },
+    {
+      "epoch": 4.182160331200603,
+      "grad_norm": 0.19358594715595245,
+      "learning_rate": 2.522705581753056e-05,
+      "loss": 0.2025,
+      "step": 5556
+    },
+    {
+      "epoch": 4.182913059841927,
+      "grad_norm": 0.2333727478981018,
+      "learning_rate": 2.521965339081587e-05,
+      "loss": 0.2991,
+      "step": 5557
+    },
+    {
+      "epoch": 4.1836657884832515,
+      "grad_norm": 0.20957507193088531,
+      "learning_rate": 2.5212251213476455e-05,
+      "loss": 0.2313,
+      "step": 5558
+    },
+    {
+      "epoch": 4.184418517124577,
+      "grad_norm": 0.2760397493839264,
+      "learning_rate": 2.5204849286321802e-05,
+      "loss": 0.2098,
+      "step": 5559
+    },
+    {
+      "epoch": 4.185171245765901,
+      "grad_norm": 0.1885673999786377,
+      "learning_rate": 2.519744761016139e-05,
+      "loss": 0.1635,
+      "step": 5560
+    },
+    {
+      "epoch": 4.185923974407226,
+      "grad_norm": 0.22674332559108734,
+      "learning_rate": 2.5190046185804643e-05,
+      "loss": 0.2705,
+      "step": 5561
+    },
+    {
+      "epoch": 4.186676703048551,
+      "grad_norm": 0.20395071804523468,
+      "learning_rate": 2.5182645014060972e-05,
+      "loss": 0.1995,
+      "step": 5562
+    },
+    {
+      "epoch": 4.187429431689876,
+      "grad_norm": 0.2058943510055542,
+      "learning_rate": 2.5175244095739757e-05,
+      "loss": 0.2784,
+      "step": 5563
+    },
+    {
+      "epoch": 4.1881821603312,
+      "grad_norm": 0.21652038395404816,
+      "learning_rate": 2.5167843431650356e-05,
+      "loss": 0.2027,
+      "step": 5564
+    },
+    {
+      "epoch": 4.1889348889725255,
+      "grad_norm": 0.205149844288826,
+      "learning_rate": 2.5160443022602093e-05,
+      "loss": 0.1977,
+      "step": 5565
+    },
+    {
+      "epoch": 4.18968761761385,
+      "grad_norm": 0.24476781487464905,
+      "learning_rate": 2.5153042869404252e-05,
+      "loss": 0.1953,
+      "step": 5566
+    },
+    {
+      "epoch": 4.190440346255175,
+      "grad_norm": 0.22433388233184814,
+      "learning_rate": 2.5145642972866122e-05,
+      "loss": 0.2674,
+      "step": 5567
+    },
+    {
+      "epoch": 4.1911930748965,
+      "grad_norm": 0.18531204760074615,
+      "learning_rate": 2.513824333379694e-05,
+      "loss": 0.164,
+      "step": 5568
+    },
+    {
+      "epoch": 4.191945803537824,
+      "grad_norm": 0.2131650596857071,
+      "learning_rate": 2.513084395300591e-05,
+      "loss": 0.186,
+      "step": 5569
+    },
+    {
+      "epoch": 4.19269853217915,
+      "grad_norm": 0.22872623801231384,
+      "learning_rate": 2.512344483130223e-05,
+      "loss": 0.2273,
+      "step": 5570
+    },
+    {
+      "epoch": 4.193451260820474,
+      "grad_norm": 0.2047608196735382,
+      "learning_rate": 2.5116045969495037e-05,
+      "loss": 0.218,
+      "step": 5571
+    },
+    {
+      "epoch": 4.194203989461799,
+      "grad_norm": 0.19754445552825928,
+      "learning_rate": 2.5108647368393478e-05,
+      "loss": 0.2013,
+      "step": 5572
+    },
+    {
+      "epoch": 4.194956718103124,
+      "grad_norm": 0.25272244215011597,
+      "learning_rate": 2.5101249028806656e-05,
+      "loss": 0.2332,
+      "step": 5573
+    },
+    {
+      "epoch": 4.1957094467444485,
+      "grad_norm": 0.1898188441991806,
+      "learning_rate": 2.509385095154363e-05,
+      "loss": 0.1936,
+      "step": 5574
+    },
+    {
+      "epoch": 4.196462175385774,
+      "grad_norm": 0.21958021819591522,
+      "learning_rate": 2.508645313741344e-05,
+      "loss": 0.2177,
+      "step": 5575
+    },
+    {
+      "epoch": 4.197214904027098,
+      "grad_norm": 0.1836525797843933,
+      "learning_rate": 2.5079055587225115e-05,
+      "loss": 0.2048,
+      "step": 5576
+    },
+    {
+      "epoch": 4.197967632668423,
+      "grad_norm": 0.2986646592617035,
+      "learning_rate": 2.5071658301787625e-05,
+      "loss": 0.2496,
+      "step": 5577
+    },
+    {
+      "epoch": 4.198720361309748,
+      "grad_norm": 0.196857750415802,
+      "learning_rate": 2.5064261281909936e-05,
+      "loss": 0.1494,
+      "step": 5578
+    },
+    {
+      "epoch": 4.199473089951073,
+      "grad_norm": 0.23829561471939087,
+      "learning_rate": 2.5056864528400964e-05,
+      "loss": 0.2365,
+      "step": 5579
+    },
+    {
+      "epoch": 4.200225818592397,
+      "grad_norm": 0.21834972500801086,
+      "learning_rate": 2.504946804206962e-05,
+      "loss": 0.1418,
+      "step": 5580
+    },
+    {
+      "epoch": 4.200978547233722,
+      "grad_norm": 0.22006766498088837,
+      "learning_rate": 2.5042071823724765e-05,
+      "loss": 0.2509,
+      "step": 5581
+    },
+    {
+      "epoch": 4.201731275875047,
+      "grad_norm": 0.20671267807483673,
+      "learning_rate": 2.5034675874175245e-05,
+      "loss": 0.2331,
+      "step": 5582
+    },
+    {
+      "epoch": 4.202484004516372,
+      "grad_norm": 0.20214709639549255,
+      "learning_rate": 2.5027280194229864e-05,
+      "loss": 0.2467,
+      "step": 5583
+    },
+    {
+      "epoch": 4.203236733157697,
+      "grad_norm": 0.21242909133434296,
+      "learning_rate": 2.50198847846974e-05,
+      "loss": 0.2025,
+      "step": 5584
+    },
+    {
+      "epoch": 4.203989461799021,
+      "grad_norm": 0.19811666011810303,
+      "learning_rate": 2.5012489646386598e-05,
+      "loss": 0.2306,
+      "step": 5585
+    },
+    {
+      "epoch": 4.204742190440347,
+      "grad_norm": 0.25037088990211487,
+      "learning_rate": 2.5005094780106197e-05,
+      "loss": 0.2866,
+      "step": 5586
+    },
+    {
+      "epoch": 4.205494919081671,
+      "grad_norm": 0.21689459681510925,
+      "learning_rate": 2.4997700186664885e-05,
+      "loss": 0.2169,
+      "step": 5587
+    },
+    {
+      "epoch": 4.2062476477229955,
+      "grad_norm": 0.22865641117095947,
+      "learning_rate": 2.4990305866871317e-05,
+      "loss": 0.2049,
+      "step": 5588
+    },
+    {
+      "epoch": 4.207000376364321,
+      "grad_norm": 0.2079988270998001,
+      "learning_rate": 2.4982911821534123e-05,
+      "loss": 0.1786,
+      "step": 5589
+    },
+    {
+      "epoch": 4.207753105005645,
+      "grad_norm": 0.17747244238853455,
+      "learning_rate": 2.497551805146191e-05,
+      "loss": 0.1682,
+      "step": 5590
+    },
+    {
+      "epoch": 4.20850583364697,
+      "grad_norm": 0.20964185893535614,
+      "learning_rate": 2.4968124557463236e-05,
+      "loss": 0.1504,
+      "step": 5591
+    },
+    {
+      "epoch": 4.209258562288295,
+      "grad_norm": 0.21024668216705322,
+      "learning_rate": 2.4960731340346666e-05,
+      "loss": 0.164,
+      "step": 5592
+    },
+    {
+      "epoch": 4.21001129092962,
+      "grad_norm": 0.21144850552082062,
+      "learning_rate": 2.4953338400920695e-05,
+      "loss": 0.226,
+      "step": 5593
+    },
+    {
+      "epoch": 4.210764019570945,
+      "grad_norm": 0.2160588949918747,
+      "learning_rate": 2.4945945739993808e-05,
+      "loss": 0.1791,
+      "step": 5594
+    },
+    {
+      "epoch": 4.2115167482122695,
+      "grad_norm": 0.2399984449148178,
+      "learning_rate": 2.493855335837446e-05,
+      "loss": 0.2391,
+      "step": 5595
+    },
+    {
+      "epoch": 4.212269476853594,
+      "grad_norm": 0.22582878172397614,
+      "learning_rate": 2.4931161256871055e-05,
+      "loss": 0.142,
+      "step": 5596
+    },
+    {
+      "epoch": 4.213022205494919,
+      "grad_norm": 0.16870789229869843,
+      "learning_rate": 2.4923769436292e-05,
+      "loss": 0.095,
+      "step": 5597
+    },
+    {
+      "epoch": 4.213774934136244,
+      "grad_norm": 0.2232659012079239,
+      "learning_rate": 2.4916377897445642e-05,
+      "loss": 0.2593,
+      "step": 5598
+    },
+    {
+      "epoch": 4.214527662777568,
+      "grad_norm": 0.20545484125614166,
+      "learning_rate": 2.490898664114032e-05,
+      "loss": 0.1748,
+      "step": 5599
+    },
+    {
+      "epoch": 4.215280391418894,
+      "grad_norm": 0.2433621883392334,
+      "learning_rate": 2.4901595668184314e-05,
+      "loss": 0.256,
+      "step": 5600
+    },
+    {
+      "epoch": 4.215280391418894,
+      "eval_loss": 0.23942534625530243,
+      "eval_runtime": 452.9348,
+      "eval_samples_per_second": 21.255,
+      "eval_steps_per_second": 0.665,
+      "step": 5600
+    },
+    {
+      "epoch": 4.216033120060218,
+      "grad_norm": 0.25421902537345886,
+      "learning_rate": 2.4894204979385905e-05,
+      "loss": 0.338,
+      "step": 5601
+    },
+    {
+      "epoch": 4.2167858487015435,
+      "grad_norm": 0.25086647272109985,
+      "learning_rate": 2.4886814575553323e-05,
+      "loss": 0.2585,
+      "step": 5602
+    },
+    {
+      "epoch": 4.217538577342868,
+      "grad_norm": 0.24403683841228485,
+      "learning_rate": 2.487942445749477e-05,
+      "loss": 0.2961,
+      "step": 5603
+    },
+    {
+      "epoch": 4.218291305984192,
+      "grad_norm": 0.2297692745923996,
+      "learning_rate": 2.487203462601841e-05,
+      "loss": 0.2693,
+      "step": 5604
+    },
+    {
+      "epoch": 4.219044034625518,
+      "grad_norm": 0.2282557487487793,
+      "learning_rate": 2.4864645081932398e-05,
+      "loss": 0.218,
+      "step": 5605
+    },
+    {
+      "epoch": 4.219796763266842,
+      "grad_norm": 0.23693381249904633,
+      "learning_rate": 2.485725582604484e-05,
+      "loss": 0.2537,
+      "step": 5606
+    },
+    {
+      "epoch": 4.220549491908167,
+      "grad_norm": 0.22200238704681396,
+      "learning_rate": 2.4849866859163812e-05,
+      "loss": 0.2404,
+      "step": 5607
+    },
+    {
+      "epoch": 4.221302220549492,
+      "grad_norm": 0.3011147677898407,
+      "learning_rate": 2.4842478182097357e-05,
+      "loss": 0.2501,
+      "step": 5608
+    },
+    {
+      "epoch": 4.222054949190817,
+      "grad_norm": 0.23110699653625488,
+      "learning_rate": 2.4835089795653493e-05,
+      "loss": 0.215,
+      "step": 5609
+    },
+    {
+      "epoch": 4.222807677832142,
+      "grad_norm": 0.22284109890460968,
+      "learning_rate": 2.4827701700640198e-05,
+      "loss": 0.2306,
+      "step": 5610
+    },
+    {
+      "epoch": 4.223560406473466,
+      "grad_norm": 0.2646382749080658,
+      "learning_rate": 2.4820313897865433e-05,
+      "loss": 0.2708,
+      "step": 5611
+    },
+    {
+      "epoch": 4.224313135114791,
+      "grad_norm": 0.2474120557308197,
+      "learning_rate": 2.481292638813711e-05,
+      "loss": 0.255,
+      "step": 5612
+    },
+    {
+      "epoch": 4.225065863756116,
+      "grad_norm": 0.23183676600456238,
+      "learning_rate": 2.480553917226312e-05,
+      "loss": 0.2241,
+      "step": 5613
+    },
+    {
+      "epoch": 4.225818592397441,
+      "grad_norm": 0.21351392567157745,
+      "learning_rate": 2.4798152251051314e-05,
+      "loss": 0.1622,
+      "step": 5614
+    },
+    {
+      "epoch": 4.226571321038765,
+      "grad_norm": 0.21947894990444183,
+      "learning_rate": 2.4790765625309514e-05,
+      "loss": 0.2192,
+      "step": 5615
+    },
+    {
+      "epoch": 4.227324049680091,
+      "grad_norm": 0.21422354876995087,
+      "learning_rate": 2.478337929584551e-05,
+      "loss": 0.1896,
+      "step": 5616
+    },
+    {
+      "epoch": 4.228076778321415,
+      "grad_norm": 0.19885827600955963,
+      "learning_rate": 2.4775993263467056e-05,
+      "loss": 0.2703,
+      "step": 5617
+    },
+    {
+      "epoch": 4.2288295069627395,
+      "grad_norm": 0.1965792030096054,
+      "learning_rate": 2.4768607528981896e-05,
+      "loss": 0.1669,
+      "step": 5618
+    },
+    {
+      "epoch": 4.229582235604065,
+      "grad_norm": 0.1937028020620346,
+      "learning_rate": 2.4761222093197707e-05,
+      "loss": 0.1359,
+      "step": 5619
+    },
+    {
+      "epoch": 4.230334964245389,
+      "grad_norm": 0.2303389310836792,
+      "learning_rate": 2.475383695692215e-05,
+      "loss": 0.3108,
+      "step": 5620
+    },
+    {
+      "epoch": 4.231087692886715,
+      "grad_norm": 0.20600168406963348,
+      "learning_rate": 2.4746452120962847e-05,
+      "loss": 0.2479,
+      "step": 5621
+    },
+    {
+      "epoch": 4.231840421528039,
+      "grad_norm": 0.2180294543504715,
+      "learning_rate": 2.4739067586127414e-05,
+      "loss": 0.174,
+      "step": 5622
+    },
+    {
+      "epoch": 4.232593150169364,
+      "grad_norm": 0.19157640635967255,
+      "learning_rate": 2.473168335322339e-05,
+      "loss": 0.2615,
+      "step": 5623
+    },
+    {
+      "epoch": 4.233345878810689,
+      "grad_norm": 0.2300761193037033,
+      "learning_rate": 2.4724299423058317e-05,
+      "loss": 0.2598,
+      "step": 5624
+    },
+    {
+      "epoch": 4.2340986074520135,
+      "grad_norm": 0.23727518320083618,
+      "learning_rate": 2.4716915796439678e-05,
+      "loss": 0.2693,
+      "step": 5625
+    },
+    {
+      "epoch": 4.234851336093338,
+      "grad_norm": 0.22769983112812042,
+      "learning_rate": 2.470953247417495e-05,
+      "loss": 0.2201,
+      "step": 5626
+    },
+    {
+      "epoch": 4.235604064734663,
+      "grad_norm": 0.19293732941150665,
+      "learning_rate": 2.470214945707156e-05,
+      "loss": 0.1346,
+      "step": 5627
+    },
+    {
+      "epoch": 4.236356793375988,
+      "grad_norm": 0.2132001668214798,
+      "learning_rate": 2.469476674593689e-05,
+      "loss": 0.2153,
+      "step": 5628
+    },
+    {
+      "epoch": 4.237109522017313,
+      "grad_norm": 0.2023821771144867,
+      "learning_rate": 2.4687384341578308e-05,
+      "loss": 0.1413,
+      "step": 5629
+    },
+    {
+      "epoch": 4.237862250658638,
+      "grad_norm": 0.2209831029176712,
+      "learning_rate": 2.4680002244803153e-05,
+      "loss": 0.2158,
+      "step": 5630
+    },
+    {
+      "epoch": 4.238614979299962,
+      "grad_norm": 0.1964791864156723,
+      "learning_rate": 2.4672620456418716e-05,
+      "loss": 0.2264,
+      "step": 5631
+    },
+    {
+      "epoch": 4.2393677079412875,
+      "grad_norm": 0.24588699638843536,
+      "learning_rate": 2.466523897723225e-05,
+      "loss": 0.2106,
+      "step": 5632
+    },
+    {
+      "epoch": 4.240120436582612,
+      "grad_norm": 0.21802546083927155,
+      "learning_rate": 2.465785780805099e-05,
+      "loss": 0.1653,
+      "step": 5633
+    },
+    {
+      "epoch": 4.240873165223936,
+      "grad_norm": 0.21761953830718994,
+      "learning_rate": 2.4650476949682127e-05,
+      "loss": 0.2318,
+      "step": 5634
+    },
+    {
+      "epoch": 4.241625893865262,
+      "grad_norm": 0.19751910865306854,
+      "learning_rate": 2.4643096402932818e-05,
+      "loss": 0.2881,
+      "step": 5635
+    },
+    {
+      "epoch": 4.242378622506586,
+      "grad_norm": 0.23676232993602753,
+      "learning_rate": 2.4635716168610184e-05,
+      "loss": 0.1562,
+      "step": 5636
+    },
+    {
+      "epoch": 4.243131351147911,
+      "grad_norm": 0.21218377351760864,
+      "learning_rate": 2.4628336247521334e-05,
+      "loss": 0.2388,
+      "step": 5637
+    },
+    {
+      "epoch": 4.243884079789236,
+      "grad_norm": 0.21036066114902496,
+      "learning_rate": 2.4620956640473314e-05,
+      "loss": 0.2084,
+      "step": 5638
+    },
+    {
+      "epoch": 4.244636808430561,
+      "grad_norm": 0.25215545296669006,
+      "learning_rate": 2.461357734827315e-05,
+      "loss": 0.2136,
+      "step": 5639
+    },
+    {
+      "epoch": 4.245389537071886,
+      "grad_norm": 0.21295151114463806,
+      "learning_rate": 2.4606198371727818e-05,
+      "loss": 0.244,
+      "step": 5640
+    },
+    {
+      "epoch": 4.24614226571321,
+      "grad_norm": 0.207365021109581,
+      "learning_rate": 2.459881971164429e-05,
+      "loss": 0.2411,
+      "step": 5641
+    },
+    {
+      "epoch": 4.246894994354535,
+      "grad_norm": 0.21446813642978668,
+      "learning_rate": 2.4591441368829467e-05,
+      "loss": 0.1783,
+      "step": 5642
+    },
+    {
+      "epoch": 4.24764772299586,
+      "grad_norm": 0.16578099131584167,
+      "learning_rate": 2.4584063344090248e-05,
+      "loss": 0.1692,
+      "step": 5643
+    },
+    {
+      "epoch": 4.248400451637185,
+      "grad_norm": 0.23017333447933197,
+      "learning_rate": 2.4576685638233478e-05,
+      "loss": 0.203,
+      "step": 5644
+    },
+    {
+      "epoch": 4.249153180278509,
+      "grad_norm": 0.19162234663963318,
+      "learning_rate": 2.456930825206597e-05,
+      "loss": 0.2031,
+      "step": 5645
+    },
+    {
+      "epoch": 4.2499059089198346,
+      "grad_norm": 0.22374150156974792,
+      "learning_rate": 2.4561931186394504e-05,
+      "loss": 0.2114,
+      "step": 5646
+    },
+    {
+      "epoch": 4.250658637561159,
+      "grad_norm": 0.18451650440692902,
+      "learning_rate": 2.4554554442025835e-05,
+      "loss": 0.1767,
+      "step": 5647
+    },
+    {
+      "epoch": 4.251411366202484,
+      "grad_norm": 0.21216611564159393,
+      "learning_rate": 2.4547178019766653e-05,
+      "loss": 0.1661,
+      "step": 5648
+    },
+    {
+      "epoch": 4.252164094843809,
+      "grad_norm": 0.19537487626075745,
+      "learning_rate": 2.4539801920423644e-05,
+      "loss": 0.1993,
+      "step": 5649
+    },
+    {
+      "epoch": 4.252916823485133,
+      "grad_norm": 0.22114747762680054,
+      "learning_rate": 2.453242614480345e-05,
+      "loss": 0.2057,
+      "step": 5650
+    },
+    {
+      "epoch": 4.253669552126459,
+      "grad_norm": 0.1916477233171463,
+      "learning_rate": 2.452505069371267e-05,
+      "loss": 0.2336,
+      "step": 5651
+    },
+    {
+      "epoch": 4.254422280767783,
+      "grad_norm": 0.1968235820531845,
+      "learning_rate": 2.451767556795788e-05,
+      "loss": 0.1562,
+      "step": 5652
+    },
+    {
+      "epoch": 4.255175009409108,
+      "grad_norm": 0.20137332379817963,
+      "learning_rate": 2.45103007683456e-05,
+      "loss": 0.2083,
+      "step": 5653
+    },
+    {
+      "epoch": 4.255927738050433,
+      "grad_norm": 0.1882929652929306,
+      "learning_rate": 2.450292629568234e-05,
+      "loss": 0.225,
+      "step": 5654
+    },
+    {
+      "epoch": 4.2566804666917575,
+      "grad_norm": 0.19396980106830597,
+      "learning_rate": 2.4495552150774536e-05,
+      "loss": 0.2016,
+      "step": 5655
+    },
+    {
+      "epoch": 4.257433195333082,
+      "grad_norm": 0.2349831759929657,
+      "learning_rate": 2.4488178334428645e-05,
+      "loss": 0.2013,
+      "step": 5656
+    },
+    {
+      "epoch": 4.258185923974407,
+      "grad_norm": 0.2173021137714386,
+      "learning_rate": 2.4480804847451045e-05,
+      "loss": 0.2376,
+      "step": 5657
+    },
+    {
+      "epoch": 4.258938652615732,
+      "grad_norm": 0.21482734382152557,
+      "learning_rate": 2.4473431690648087e-05,
+      "loss": 0.1438,
+      "step": 5658
+    },
+    {
+      "epoch": 4.259691381257057,
+      "grad_norm": 0.17118032276630402,
+      "learning_rate": 2.4466058864826085e-05,
+      "loss": 0.175,
+      "step": 5659
+    },
+    {
+      "epoch": 4.260444109898382,
+      "grad_norm": 0.19927719235420227,
+      "learning_rate": 2.4458686370791327e-05,
+      "loss": 0.2841,
+      "step": 5660
+    },
+    {
+      "epoch": 4.261196838539706,
+      "grad_norm": 0.23474466800689697,
+      "learning_rate": 2.4451314209350047e-05,
+      "loss": 0.2367,
+      "step": 5661
+    },
+    {
+      "epoch": 4.2619495671810315,
+      "grad_norm": 0.20811137557029724,
+      "learning_rate": 2.4443942381308466e-05,
+      "loss": 0.1478,
+      "step": 5662
+    },
+    {
+      "epoch": 4.262702295822356,
+      "grad_norm": 0.2290758192539215,
+      "learning_rate": 2.443657088747276e-05,
+      "loss": 0.1757,
+      "step": 5663
+    },
+    {
+      "epoch": 4.263455024463681,
+      "grad_norm": 0.23866912722587585,
+      "learning_rate": 2.4429199728649048e-05,
+      "loss": 0.2398,
+      "step": 5664
+    },
+    {
+      "epoch": 4.264207753105006,
+      "grad_norm": 0.204092338681221,
+      "learning_rate": 2.442182890564344e-05,
+      "loss": 0.1907,
+      "step": 5665
+    },
+    {
+      "epoch": 4.26496048174633,
+      "grad_norm": 0.2503703236579895,
+      "learning_rate": 2.4414458419261993e-05,
+      "loss": 0.2411,
+      "step": 5666
+    },
+    {
+      "epoch": 4.265713210387656,
+      "grad_norm": 0.2233315259218216,
+      "learning_rate": 2.440708827031073e-05,
+      "loss": 0.1919,
+      "step": 5667
+    },
+    {
+      "epoch": 4.26646593902898,
+      "grad_norm": 0.24391578137874603,
+      "learning_rate": 2.439971845959565e-05,
+      "loss": 0.2431,
+      "step": 5668
+    },
+    {
+      "epoch": 4.2672186676703046,
+      "grad_norm": 0.24961601197719574,
+      "learning_rate": 2.4392348987922693e-05,
+      "loss": 0.1762,
+      "step": 5669
+    },
+    {
+      "epoch": 4.26797139631163,
+      "grad_norm": 0.2197190374135971,
+      "learning_rate": 2.4384979856097782e-05,
+      "loss": 0.21,
+      "step": 5670
+    },
+    {
+      "epoch": 4.268724124952954,
+      "grad_norm": 0.20241893827915192,
+      "learning_rate": 2.4377611064926785e-05,
+      "loss": 0.2116,
+      "step": 5671
+    },
+    {
+      "epoch": 4.269476853594279,
+      "grad_norm": 0.19189108908176422,
+      "learning_rate": 2.4370242615215555e-05,
+      "loss": 0.1643,
+      "step": 5672
+    },
+    {
+      "epoch": 4.270229582235604,
+      "grad_norm": 0.2281411588191986,
+      "learning_rate": 2.4362874507769888e-05,
+      "loss": 0.2206,
+      "step": 5673
+    },
+    {
+      "epoch": 4.270982310876929,
+      "grad_norm": 0.23890605568885803,
+      "learning_rate": 2.435550674339554e-05,
+      "loss": 0.1371,
+      "step": 5674
+    },
+    {
+      "epoch": 4.271735039518254,
+      "grad_norm": 0.19139419496059418,
+      "learning_rate": 2.434813932289825e-05,
+      "loss": 0.1931,
+      "step": 5675
+    },
+    {
+      "epoch": 4.2724877681595785,
+      "grad_norm": 0.20883922278881073,
+      "learning_rate": 2.4340772247083704e-05,
+      "loss": 0.1681,
+      "step": 5676
+    },
+    {
+      "epoch": 4.273240496800903,
+      "grad_norm": 0.2198123335838318,
+      "learning_rate": 2.4333405516757562e-05,
+      "loss": 0.2458,
+      "step": 5677
+    },
+    {
+      "epoch": 4.273993225442228,
+      "grad_norm": 0.22387054562568665,
+      "learning_rate": 2.4326039132725425e-05,
+      "loss": 0.1957,
+      "step": 5678
+    },
+    {
+      "epoch": 4.274745954083553,
+      "grad_norm": 0.195397287607193,
+      "learning_rate": 2.431867309579288e-05,
+      "loss": 0.1981,
+      "step": 5679
+    },
+    {
+      "epoch": 4.275498682724877,
+      "grad_norm": 0.24616371095180511,
+      "learning_rate": 2.4311307406765453e-05,
+      "loss": 0.3041,
+      "step": 5680
+    },
+    {
+      "epoch": 4.276251411366203,
+      "grad_norm": 0.19436131417751312,
+      "learning_rate": 2.4303942066448665e-05,
+      "loss": 0.1497,
+      "step": 5681
+    },
+    {
+      "epoch": 4.277004140007527,
+      "grad_norm": 0.21056002378463745,
+      "learning_rate": 2.4296577075647968e-05,
+      "loss": 0.216,
+      "step": 5682
+    },
+    {
+      "epoch": 4.2777568686488525,
+      "grad_norm": 0.1731927990913391,
+      "learning_rate": 2.4289212435168785e-05,
+      "loss": 0.1768,
+      "step": 5683
+    },
+    {
+      "epoch": 4.278509597290177,
+      "grad_norm": 0.19214224815368652,
+      "learning_rate": 2.4281848145816498e-05,
+      "loss": 0.2,
+      "step": 5684
+    },
+    {
+      "epoch": 4.2792623259315015,
+      "grad_norm": 0.1966479867696762,
+      "learning_rate": 2.4274484208396464e-05,
+      "loss": 0.2367,
+      "step": 5685
+    },
+    {
+      "epoch": 4.280015054572827,
+      "grad_norm": 0.23017442226409912,
+      "learning_rate": 2.4267120623713986e-05,
+      "loss": 0.2367,
+      "step": 5686
+    },
+    {
+      "epoch": 4.280767783214151,
+      "grad_norm": 0.21086545288562775,
+      "learning_rate": 2.425975739257433e-05,
+      "loss": 0.1416,
+      "step": 5687
+    },
+    {
+      "epoch": 4.281520511855476,
+      "grad_norm": 0.22463347017765045,
+      "learning_rate": 2.4252394515782734e-05,
+      "loss": 0.1868,
+      "step": 5688
+    },
+    {
+      "epoch": 4.282273240496801,
+      "grad_norm": 0.19235669076442719,
+      "learning_rate": 2.42450319941444e-05,
+      "loss": 0.2008,
+      "step": 5689
+    },
+    {
+      "epoch": 4.283025969138126,
+      "grad_norm": 0.19574737548828125,
+      "learning_rate": 2.4237669828464472e-05,
+      "loss": 0.2836,
+      "step": 5690
+    },
+    {
+      "epoch": 4.28377869777945,
+      "grad_norm": 0.21303342282772064,
+      "learning_rate": 2.4230308019548066e-05,
+      "loss": 0.2075,
+      "step": 5691
+    },
+    {
+      "epoch": 4.2845314264207754,
+      "grad_norm": 0.20643000304698944,
+      "learning_rate": 2.422294656820025e-05,
+      "loss": 0.3214,
+      "step": 5692
+    },
+    {
+      "epoch": 4.2852841550621,
+      "grad_norm": 0.20947808027267456,
+      "learning_rate": 2.4215585475226078e-05,
+      "loss": 0.2337,
+      "step": 5693
+    },
+    {
+      "epoch": 4.286036883703425,
+      "grad_norm": 0.2186335027217865,
+      "learning_rate": 2.420822474143054e-05,
+      "loss": 0.2172,
+      "step": 5694
+    },
+    {
+      "epoch": 4.28678961234475,
+      "grad_norm": 0.21054622530937195,
+      "learning_rate": 2.4200864367618593e-05,
+      "loss": 0.2604,
+      "step": 5695
+    },
+    {
+      "epoch": 4.287542340986074,
+      "grad_norm": 0.23950286209583282,
+      "learning_rate": 2.419350435459516e-05,
+      "loss": 0.2254,
+      "step": 5696
+    },
+    {
+      "epoch": 4.2882950696274,
+      "grad_norm": 0.236442431807518,
+      "learning_rate": 2.418614470316512e-05,
+      "loss": 0.2468,
+      "step": 5697
+    },
+    {
+      "epoch": 4.289047798268724,
+      "grad_norm": 0.19537372887134552,
+      "learning_rate": 2.4178785414133316e-05,
+      "loss": 0.1481,
+      "step": 5698
+    },
+    {
+      "epoch": 4.2898005269100485,
+      "grad_norm": 0.21264798939228058,
+      "learning_rate": 2.417142648830453e-05,
+      "loss": 0.1953,
+      "step": 5699
+    },
+    {
+      "epoch": 4.290553255551374,
+      "grad_norm": 0.22584304213523865,
+      "learning_rate": 2.416406792648355e-05,
+      "loss": 0.2351,
+      "step": 5700
+    },
+    {
+      "epoch": 4.291305984192698,
+      "grad_norm": 0.20269165933132172,
+      "learning_rate": 2.4156709729475098e-05,
+      "loss": 0.2285,
+      "step": 5701
+    },
+    {
+      "epoch": 4.292058712834024,
+      "grad_norm": 0.22047889232635498,
+      "learning_rate": 2.4149351898083832e-05,
+      "loss": 0.2075,
+      "step": 5702
+    },
+    {
+      "epoch": 4.292811441475348,
+      "grad_norm": 0.19918085634708405,
+      "learning_rate": 2.4141994433114408e-05,
+      "loss": 0.2441,
+      "step": 5703
+    },
+    {
+      "epoch": 4.293564170116673,
+      "grad_norm": 0.21852512657642365,
+      "learning_rate": 2.4134637335371425e-05,
+      "loss": 0.2339,
+      "step": 5704
+    },
+    {
+      "epoch": 4.294316898757998,
+      "grad_norm": 0.18654781579971313,
+      "learning_rate": 2.4127280605659437e-05,
+      "loss": 0.1999,
+      "step": 5705
+    },
+    {
+      "epoch": 4.2950696273993225,
+      "grad_norm": 0.18093229830265045,
+      "learning_rate": 2.4119924244782965e-05,
+      "loss": 0.1818,
+      "step": 5706
+    },
+    {
+      "epoch": 4.295822356040647,
+      "grad_norm": 0.2323596328496933,
+      "learning_rate": 2.4112568253546508e-05,
+      "loss": 0.2144,
+      "step": 5707
+    },
+    {
+      "epoch": 4.296575084681972,
+      "grad_norm": 0.21820864081382751,
+      "learning_rate": 2.4105212632754493e-05,
+      "loss": 0.1287,
+      "step": 5708
+    },
+    {
+      "epoch": 4.297327813323297,
+      "grad_norm": 0.196783646941185,
+      "learning_rate": 2.409785738321132e-05,
+      "loss": 0.2507,
+      "step": 5709
+    },
+    {
+      "epoch": 4.298080541964621,
+      "grad_norm": 0.21043553948402405,
+      "learning_rate": 2.4090502505721348e-05,
+      "loss": 0.2309,
+      "step": 5710
+    },
+    {
+      "epoch": 4.298833270605947,
+      "grad_norm": 0.19214586913585663,
+      "learning_rate": 2.4083148001088896e-05,
+      "loss": 0.1067,
+      "step": 5711
+    },
+    {
+      "epoch": 4.299585999247271,
+      "grad_norm": 0.18467874825000763,
+      "learning_rate": 2.407579387011823e-05,
+      "loss": 0.1162,
+      "step": 5712
+    },
+    {
+      "epoch": 4.3003387278885965,
+      "grad_norm": 0.2146999090909958,
+      "learning_rate": 2.406844011361361e-05,
+      "loss": 0.2402,
+      "step": 5713
+    },
+    {
+      "epoch": 4.301091456529921,
+      "grad_norm": 0.17779821157455444,
+      "learning_rate": 2.406108673237922e-05,
+      "loss": 0.2112,
+      "step": 5714
+    },
+    {
+      "epoch": 4.3018441851712454,
+      "grad_norm": 0.20376449823379517,
+      "learning_rate": 2.4053733727219212e-05,
+      "loss": 0.199,
+      "step": 5715
+    },
+    {
+      "epoch": 4.302596913812571,
+      "grad_norm": 0.19216731190681458,
+      "learning_rate": 2.40463810989377e-05,
+      "loss": 0.1548,
+      "step": 5716
+    },
+    {
+      "epoch": 4.303349642453895,
+      "grad_norm": 0.22630444169044495,
+      "learning_rate": 2.4039028848338762e-05,
+      "loss": 0.2289,
+      "step": 5717
+    },
+    {
+      "epoch": 4.30410237109522,
+      "grad_norm": 0.22180557250976562,
+      "learning_rate": 2.4031676976226424e-05,
+      "loss": 0.2079,
+      "step": 5718
+    },
+    {
+      "epoch": 4.304855099736545,
+      "grad_norm": 0.22839143872261047,
+      "learning_rate": 2.402432548340468e-05,
+      "loss": 0.1603,
+      "step": 5719
+    },
+    {
+      "epoch": 4.30560782837787,
+      "grad_norm": 0.21700569987297058,
+      "learning_rate": 2.401697437067747e-05,
+      "loss": 0.1787,
+      "step": 5720
+    },
+    {
+      "epoch": 4.306360557019195,
+      "grad_norm": 0.19783470034599304,
+      "learning_rate": 2.400962363884871e-05,
+      "loss": 0.1059,
+      "step": 5721
+    },
+    {
+      "epoch": 4.307113285660519,
+      "grad_norm": 0.22467941045761108,
+      "learning_rate": 2.4002273288722264e-05,
+      "loss": 0.146,
+      "step": 5722
+    },
+    {
+      "epoch": 4.307866014301844,
+      "grad_norm": 0.21454063057899475,
+      "learning_rate": 2.3994923321101953e-05,
+      "loss": 0.2397,
+      "step": 5723
+    },
+    {
+      "epoch": 4.308618742943169,
+      "grad_norm": 0.2304893136024475,
+      "learning_rate": 2.398757373679156e-05,
+      "loss": 0.2171,
+      "step": 5724
+    },
+    {
+      "epoch": 4.309371471584494,
+      "grad_norm": 0.23051442205905914,
+      "learning_rate": 2.3980224536594803e-05,
+      "loss": 0.2869,
+      "step": 5725
+    },
+    {
+      "epoch": 4.310124200225818,
+      "grad_norm": 0.17078366875648499,
+      "learning_rate": 2.3972875721315413e-05,
+      "loss": 0.1333,
+      "step": 5726
+    },
+    {
+      "epoch": 4.310876928867144,
+      "grad_norm": 0.22981849312782288,
+      "learning_rate": 2.3965527291757033e-05,
+      "loss": 0.2592,
+      "step": 5727
+    },
+    {
+      "epoch": 4.311629657508468,
+      "grad_norm": 0.19877633452415466,
+      "learning_rate": 2.395817924872328e-05,
+      "loss": 0.1802,
+      "step": 5728
+    },
+    {
+      "epoch": 4.312382386149793,
+      "grad_norm": 0.20918776094913483,
+      "learning_rate": 2.395083159301771e-05,
+      "loss": 0.126,
+      "step": 5729
+    },
+    {
+      "epoch": 4.313135114791118,
+      "grad_norm": 0.20783467590808868,
+      "learning_rate": 2.394348432544386e-05,
+      "loss": 0.279,
+      "step": 5730
+    },
+    {
+      "epoch": 4.313887843432442,
+      "grad_norm": 0.20403337478637695,
+      "learning_rate": 2.393613744680521e-05,
+      "loss": 0.176,
+      "step": 5731
+    },
+    {
+      "epoch": 4.314640572073768,
+      "grad_norm": 0.19925591349601746,
+      "learning_rate": 2.3928790957905223e-05,
+      "loss": 0.1927,
+      "step": 5732
+    },
+    {
+      "epoch": 4.315393300715092,
+      "grad_norm": 0.20696809887886047,
+      "learning_rate": 2.3921444859547288e-05,
+      "loss": 0.1897,
+      "step": 5733
+    },
+    {
+      "epoch": 4.316146029356417,
+      "grad_norm": 0.1937345415353775,
+      "learning_rate": 2.391409915253476e-05,
+      "loss": 0.2152,
+      "step": 5734
+    },
+    {
+      "epoch": 4.316898757997742,
+      "grad_norm": 0.1676884889602661,
+      "learning_rate": 2.3906753837670957e-05,
+      "loss": 0.0896,
+      "step": 5735
+    },
+    {
+      "epoch": 4.3176514866390665,
+      "grad_norm": 0.20787999033927917,
+      "learning_rate": 2.389940891575916e-05,
+      "loss": 0.2142,
+      "step": 5736
+    },
+    {
+      "epoch": 4.318404215280392,
+      "grad_norm": 0.22135157883167267,
+      "learning_rate": 2.3892064387602574e-05,
+      "loss": 0.2337,
+      "step": 5737
+    },
+    {
+      "epoch": 4.319156943921716,
+      "grad_norm": 0.23969025909900665,
+      "learning_rate": 2.3884720254004405e-05,
+      "loss": 0.2529,
+      "step": 5738
+    },
+    {
+      "epoch": 4.319909672563041,
+      "grad_norm": 0.21607540547847748,
+      "learning_rate": 2.38773765157678e-05,
+      "loss": 0.1905,
+      "step": 5739
+    },
+    {
+      "epoch": 4.320662401204366,
+      "grad_norm": 0.22193601727485657,
+      "learning_rate": 2.3870033173695856e-05,
+      "loss": 0.2024,
+      "step": 5740
+    },
+    {
+      "epoch": 4.321415129845691,
+      "grad_norm": 0.2155558168888092,
+      "learning_rate": 2.3862690228591618e-05,
+      "loss": 0.1882,
+      "step": 5741
+    },
+    {
+      "epoch": 4.322167858487015,
+      "grad_norm": 0.19312261044979095,
+      "learning_rate": 2.3855347681258117e-05,
+      "loss": 0.1632,
+      "step": 5742
+    },
+    {
+      "epoch": 4.3229205871283405,
+      "grad_norm": 0.24156692624092102,
+      "learning_rate": 2.3848005532498307e-05,
+      "loss": 0.2676,
+      "step": 5743
+    },
+    {
+      "epoch": 4.323673315769665,
+      "grad_norm": 0.1719277948141098,
+      "learning_rate": 2.384066378311512e-05,
+      "loss": 0.1922,
+      "step": 5744
+    },
+    {
+      "epoch": 4.324426044410989,
+      "grad_norm": 0.24538426101207733,
+      "learning_rate": 2.3833322433911443e-05,
+      "loss": 0.2002,
+      "step": 5745
+    },
+    {
+      "epoch": 4.325178773052315,
+      "grad_norm": 0.1970103234052658,
+      "learning_rate": 2.382598148569011e-05,
+      "loss": 0.1756,
+      "step": 5746
+    },
+    {
+      "epoch": 4.325931501693639,
+      "grad_norm": 0.31489014625549316,
+      "learning_rate": 2.3818640939253922e-05,
+      "loss": 0.2739,
+      "step": 5747
+    },
+    {
+      "epoch": 4.326684230334965,
+      "grad_norm": 0.20381474494934082,
+      "learning_rate": 2.381130079540563e-05,
+      "loss": 0.1523,
+      "step": 5748
+    },
+    {
+      "epoch": 4.327436958976289,
+      "grad_norm": 0.3151390552520752,
+      "learning_rate": 2.3803961054947933e-05,
+      "loss": 0.2688,
+      "step": 5749
+    },
+    {
+      "epoch": 4.328189687617614,
+      "grad_norm": 0.17447030544281006,
+      "learning_rate": 2.3796621718683487e-05,
+      "loss": 0.1941,
+      "step": 5750
+    },
+    {
+      "epoch": 4.328942416258939,
+      "grad_norm": 0.3048948049545288,
+      "learning_rate": 2.378928278741493e-05,
+      "loss": 0.1922,
+      "step": 5751
+    },
+    {
+      "epoch": 4.329695144900263,
+      "grad_norm": 0.2352307289838791,
+      "learning_rate": 2.3781944261944834e-05,
+      "loss": 0.1672,
+      "step": 5752
+    },
+    {
+      "epoch": 4.330447873541588,
+      "grad_norm": 0.24347437918186188,
+      "learning_rate": 2.3774606143075717e-05,
+      "loss": 0.2456,
+      "step": 5753
+    },
+    {
+      "epoch": 4.331200602182913,
+      "grad_norm": 0.21192313730716705,
+      "learning_rate": 2.3767268431610078e-05,
+      "loss": 0.184,
+      "step": 5754
+    },
+    {
+      "epoch": 4.331953330824238,
+      "grad_norm": 0.2445749044418335,
+      "learning_rate": 2.3759931128350348e-05,
+      "loss": 0.1895,
+      "step": 5755
+    },
+    {
+      "epoch": 4.332706059465563,
+      "grad_norm": 0.1959114670753479,
+      "learning_rate": 2.3752594234098924e-05,
+      "loss": 0.1626,
+      "step": 5756
+    },
+    {
+      "epoch": 4.333458788106888,
+      "grad_norm": 0.23827750980854034,
+      "learning_rate": 2.3745257749658155e-05,
+      "loss": 0.2042,
+      "step": 5757
+    },
+    {
+      "epoch": 4.334211516748212,
+      "grad_norm": 0.24176941812038422,
+      "learning_rate": 2.373792167583036e-05,
+      "loss": 0.2703,
+      "step": 5758
+    },
+    {
+      "epoch": 4.334964245389537,
+      "grad_norm": 0.2391853779554367,
+      "learning_rate": 2.37305860134178e-05,
+      "loss": 0.2258,
+      "step": 5759
+    },
+    {
+      "epoch": 4.335716974030862,
+      "grad_norm": 0.21886131167411804,
+      "learning_rate": 2.3723250763222683e-05,
+      "loss": 0.2748,
+      "step": 5760
+    },
+    {
+      "epoch": 4.336469702672186,
+      "grad_norm": 0.21191473305225372,
+      "learning_rate": 2.3715915926047178e-05,
+      "loss": 0.1934,
+      "step": 5761
+    },
+    {
+      "epoch": 4.337222431313512,
+      "grad_norm": 0.1818401962518692,
+      "learning_rate": 2.370858150269342e-05,
+      "loss": 0.1468,
+      "step": 5762
+    },
+    {
+      "epoch": 4.337975159954836,
+      "grad_norm": 0.21994835138320923,
+      "learning_rate": 2.370124749396349e-05,
+      "loss": 0.15,
+      "step": 5763
+    },
+    {
+      "epoch": 4.338727888596161,
+      "grad_norm": 0.20545101165771484,
+      "learning_rate": 2.3693913900659426e-05,
+      "loss": 0.2897,
+      "step": 5764
+    },
+    {
+      "epoch": 4.339480617237486,
+      "grad_norm": 0.1938062608242035,
+      "learning_rate": 2.3686580723583214e-05,
+      "loss": 0.2058,
+      "step": 5765
+    },
+    {
+      "epoch": 4.3402333458788105,
+      "grad_norm": 0.213126540184021,
+      "learning_rate": 2.3679247963536793e-05,
+      "loss": 0.2165,
+      "step": 5766
+    },
+    {
+      "epoch": 4.340986074520136,
+      "grad_norm": 0.21864977478981018,
+      "learning_rate": 2.3671915621322076e-05,
+      "loss": 0.1749,
+      "step": 5767
+    },
+    {
+      "epoch": 4.34173880316146,
+      "grad_norm": 0.21347053349018097,
+      "learning_rate": 2.3664583697740918e-05,
+      "loss": 0.1575,
+      "step": 5768
+    },
+    {
+      "epoch": 4.342491531802785,
+      "grad_norm": 0.21198026835918427,
+      "learning_rate": 2.365725219359511e-05,
+      "loss": 0.2505,
+      "step": 5769
+    },
+    {
+      "epoch": 4.34324426044411,
+      "grad_norm": 0.19898301362991333,
+      "learning_rate": 2.3649921109686423e-05,
+      "loss": 0.1723,
+      "step": 5770
+    },
+    {
+      "epoch": 4.343996989085435,
+      "grad_norm": 0.20906135439872742,
+      "learning_rate": 2.364259044681658e-05,
+      "loss": 0.231,
+      "step": 5771
+    },
+    {
+      "epoch": 4.344749717726759,
+      "grad_norm": 0.24128398299217224,
+      "learning_rate": 2.3635260205787252e-05,
+      "loss": 0.2593,
+      "step": 5772
+    },
+    {
+      "epoch": 4.3455024463680845,
+      "grad_norm": 0.21533608436584473,
+      "learning_rate": 2.3627930387400056e-05,
+      "loss": 0.1859,
+      "step": 5773
+    },
+    {
+      "epoch": 4.346255175009409,
+      "grad_norm": 0.21972216665744781,
+      "learning_rate": 2.362060099245657e-05,
+      "loss": 0.2938,
+      "step": 5774
+    },
+    {
+      "epoch": 4.347007903650734,
+      "grad_norm": 0.2228476107120514,
+      "learning_rate": 2.3613272021758325e-05,
+      "loss": 0.2817,
+      "step": 5775
+    },
+    {
+      "epoch": 4.347760632292059,
+      "grad_norm": 0.19130253791809082,
+      "learning_rate": 2.3605943476106803e-05,
+      "loss": 0.2129,
+      "step": 5776
+    },
+    {
+      "epoch": 4.348513360933383,
+      "grad_norm": 0.23922346532344818,
+      "learning_rate": 2.359861535630345e-05,
+      "loss": 0.271,
+      "step": 5777
+    },
+    {
+      "epoch": 4.349266089574709,
+      "grad_norm": 0.24892202019691467,
+      "learning_rate": 2.3591287663149657e-05,
+      "loss": 0.2476,
+      "step": 5778
+    },
+    {
+      "epoch": 4.350018818216033,
+      "grad_norm": 0.22796574234962463,
+      "learning_rate": 2.3583960397446775e-05,
+      "loss": 0.2362,
+      "step": 5779
+    },
+    {
+      "epoch": 4.350771546857358,
+      "grad_norm": 0.21210840344429016,
+      "learning_rate": 2.357663355999609e-05,
+      "loss": 0.1721,
+      "step": 5780
+    },
+    {
+      "epoch": 4.351524275498683,
+      "grad_norm": 0.20786887407302856,
+      "learning_rate": 2.3569307151598868e-05,
+      "loss": 0.2139,
+      "step": 5781
+    },
+    {
+      "epoch": 4.352277004140007,
+      "grad_norm": 0.27187201380729675,
+      "learning_rate": 2.3561981173056292e-05,
+      "loss": 0.2249,
+      "step": 5782
+    },
+    {
+      "epoch": 4.353029732781332,
+      "grad_norm": 0.22791840136051178,
+      "learning_rate": 2.3554655625169543e-05,
+      "loss": 0.2772,
+      "step": 5783
+    },
+    {
+      "epoch": 4.353782461422657,
+      "grad_norm": 0.23380687832832336,
+      "learning_rate": 2.3547330508739724e-05,
+      "loss": 0.1511,
+      "step": 5784
+    },
+    {
+      "epoch": 4.354535190063982,
+      "grad_norm": 0.23833535611629486,
+      "learning_rate": 2.3540005824567894e-05,
+      "loss": 0.2619,
+      "step": 5785
+    },
+    {
+      "epoch": 4.355287918705307,
+      "grad_norm": 0.2501411736011505,
+      "learning_rate": 2.3532681573455072e-05,
+      "loss": 0.2295,
+      "step": 5786
+    },
+    {
+      "epoch": 4.3560406473466315,
+      "grad_norm": 0.2557797133922577,
+      "learning_rate": 2.3525357756202232e-05,
+      "loss": 0.1677,
+      "step": 5787
+    },
+    {
+      "epoch": 4.356793375987956,
+      "grad_norm": 0.19675835967063904,
+      "learning_rate": 2.3518034373610292e-05,
+      "loss": 0.1782,
+      "step": 5788
+    },
+    {
+      "epoch": 4.357546104629281,
+      "grad_norm": 0.20426331460475922,
+      "learning_rate": 2.3510711426480125e-05,
+      "loss": 0.1999,
+      "step": 5789
+    },
+    {
+      "epoch": 4.358298833270606,
+      "grad_norm": 0.23970738053321838,
+      "learning_rate": 2.3503388915612556e-05,
+      "loss": 0.2054,
+      "step": 5790
+    },
+    {
+      "epoch": 4.359051561911931,
+      "grad_norm": 0.20874252915382385,
+      "learning_rate": 2.3496066841808367e-05,
+      "loss": 0.2377,
+      "step": 5791
+    },
+    {
+      "epoch": 4.359804290553256,
+      "grad_norm": 0.23705989122390747,
+      "learning_rate": 2.3488745205868296e-05,
+      "loss": 0.2806,
+      "step": 5792
+    },
+    {
+      "epoch": 4.36055701919458,
+      "grad_norm": 0.20874986052513123,
+      "learning_rate": 2.348142400859301e-05,
+      "loss": 0.169,
+      "step": 5793
+    },
+    {
+      "epoch": 4.3613097478359055,
+      "grad_norm": 0.25301048159599304,
+      "learning_rate": 2.347410325078316e-05,
+      "loss": 0.2513,
+      "step": 5794
+    },
+    {
+      "epoch": 4.36206247647723,
+      "grad_norm": 0.1933937668800354,
+      "learning_rate": 2.3466782933239313e-05,
+      "loss": 0.2317,
+      "step": 5795
+    },
+    {
+      "epoch": 4.3628152051185545,
+      "grad_norm": 0.21530161798000336,
+      "learning_rate": 2.3459463056762033e-05,
+      "loss": 0.2799,
+      "step": 5796
+    },
+    {
+      "epoch": 4.36356793375988,
+      "grad_norm": 0.21551015973091125,
+      "learning_rate": 2.3452143622151795e-05,
+      "loss": 0.1812,
+      "step": 5797
+    },
+    {
+      "epoch": 4.364320662401204,
+      "grad_norm": 0.19449269771575928,
+      "learning_rate": 2.3444824630209045e-05,
+      "loss": 0.2399,
+      "step": 5798
+    },
+    {
+      "epoch": 4.365073391042529,
+      "grad_norm": 0.2459174543619156,
+      "learning_rate": 2.3437506081734173e-05,
+      "loss": 0.2189,
+      "step": 5799
+    },
+    {
+      "epoch": 4.365826119683854,
+      "grad_norm": 0.2422088384628296,
+      "learning_rate": 2.3430187977527533e-05,
+      "loss": 0.2631,
+      "step": 5800
+    },
+    {
+      "epoch": 4.365826119683854,
+      "eval_loss": 0.2346644401550293,
+      "eval_runtime": 452.6848,
+      "eval_samples_per_second": 21.266,
+      "eval_steps_per_second": 0.665,
+      "step": 5800
+    },
+    {
+      "epoch": 4.366578848325179,
+      "grad_norm": 0.22878363728523254,
+      "learning_rate": 2.3422870318389406e-05,
+      "loss": 0.2246,
+      "step": 5801
+    },
+    {
+      "epoch": 4.367331576966504,
+      "grad_norm": 0.2336190789937973,
+      "learning_rate": 2.3415553105120063e-05,
+      "loss": 0.2241,
+      "step": 5802
+    },
+    {
+      "epoch": 4.3680843056078285,
+      "grad_norm": 0.2550087571144104,
+      "learning_rate": 2.340823633851969e-05,
+      "loss": 0.2862,
+      "step": 5803
+    },
+    {
+      "epoch": 4.368837034249153,
+      "grad_norm": 0.21680331230163574,
+      "learning_rate": 2.3400920019388438e-05,
+      "loss": 0.2275,
+      "step": 5804
+    },
+    {
+      "epoch": 4.369589762890478,
+      "grad_norm": 0.20742563903331757,
+      "learning_rate": 2.339360414852641e-05,
+      "loss": 0.1309,
+      "step": 5805
+    },
+    {
+      "epoch": 4.370342491531803,
+      "grad_norm": 0.21924084424972534,
+      "learning_rate": 2.3386288726733663e-05,
+      "loss": 0.1689,
+      "step": 5806
+    },
+    {
+      "epoch": 4.371095220173127,
+      "grad_norm": 0.18349036574363708,
+      "learning_rate": 2.337897375481019e-05,
+      "loss": 0.141,
+      "step": 5807
+    },
+    {
+      "epoch": 4.371847948814453,
+      "grad_norm": 0.2405020147562027,
+      "learning_rate": 2.337165923355594e-05,
+      "loss": 0.2384,
+      "step": 5808
+    },
+    {
+      "epoch": 4.372600677455777,
+      "grad_norm": 0.21972602605819702,
+      "learning_rate": 2.3364345163770845e-05,
+      "loss": 0.2102,
+      "step": 5809
+    },
+    {
+      "epoch": 4.373353406097102,
+      "grad_norm": 0.1804770678281784,
+      "learning_rate": 2.3357031546254744e-05,
+      "loss": 0.2518,
+      "step": 5810
+    },
+    {
+      "epoch": 4.374106134738427,
+      "grad_norm": 0.20790961384773254,
+      "learning_rate": 2.334971838180744e-05,
+      "loss": 0.225,
+      "step": 5811
+    },
+    {
+      "epoch": 4.374858863379751,
+      "grad_norm": 0.2331637740135193,
+      "learning_rate": 2.3342405671228697e-05,
+      "loss": 0.2146,
+      "step": 5812
+    },
+    {
+      "epoch": 4.375611592021077,
+      "grad_norm": 0.22467386722564697,
+      "learning_rate": 2.3335093415318215e-05,
+      "loss": 0.2277,
+      "step": 5813
+    },
+    {
+      "epoch": 4.376364320662401,
+      "grad_norm": 0.20429372787475586,
+      "learning_rate": 2.3327781614875655e-05,
+      "loss": 0.254,
+      "step": 5814
+    },
+    {
+      "epoch": 4.377117049303726,
+      "grad_norm": 0.23733903467655182,
+      "learning_rate": 2.332047027070063e-05,
+      "loss": 0.251,
+      "step": 5815
+    },
+    {
+      "epoch": 4.377869777945051,
+      "grad_norm": 0.23401270806789398,
+      "learning_rate": 2.3313159383592687e-05,
+      "loss": 0.3225,
+      "step": 5816
+    },
+    {
+      "epoch": 4.3786225065863755,
+      "grad_norm": 0.21547015011310577,
+      "learning_rate": 2.3305848954351338e-05,
+      "loss": 0.1887,
+      "step": 5817
+    },
+    {
+      "epoch": 4.3793752352277,
+      "grad_norm": 0.18512047827243805,
+      "learning_rate": 2.3298538983776047e-05,
+      "loss": 0.1241,
+      "step": 5818
+    },
+    {
+      "epoch": 4.380127963869025,
+      "grad_norm": 0.21568818390369415,
+      "learning_rate": 2.3291229472666215e-05,
+      "loss": 0.206,
+      "step": 5819
+    },
+    {
+      "epoch": 4.38088069251035,
+      "grad_norm": 0.19237665832042694,
+      "learning_rate": 2.3283920421821192e-05,
+      "loss": 0.2394,
+      "step": 5820
+    },
+    {
+      "epoch": 4.381633421151675,
+      "grad_norm": 0.21998052299022675,
+      "learning_rate": 2.32766118320403e-05,
+      "loss": 0.2708,
+      "step": 5821
+    },
+    {
+      "epoch": 4.382386149793,
+      "grad_norm": 0.2414284646511078,
+      "learning_rate": 2.326930370412279e-05,
+      "loss": 0.2249,
+      "step": 5822
+    },
+    {
+      "epoch": 4.383138878434324,
+      "grad_norm": 0.22616979479789734,
+      "learning_rate": 2.3261996038867866e-05,
+      "loss": 0.2556,
+      "step": 5823
+    },
+    {
+      "epoch": 4.3838916070756495,
+      "grad_norm": 0.2315456122159958,
+      "learning_rate": 2.3254688837074682e-05,
+      "loss": 0.2531,
+      "step": 5824
+    },
+    {
+      "epoch": 4.384644335716974,
+      "grad_norm": 0.22710219025611877,
+      "learning_rate": 2.3247382099542348e-05,
+      "loss": 0.2085,
+      "step": 5825
+    },
+    {
+      "epoch": 4.3853970643582985,
+      "grad_norm": 0.2613029181957245,
+      "learning_rate": 2.3240075827069912e-05,
+      "loss": 0.3334,
+      "step": 5826
+    },
+    {
+      "epoch": 4.386149792999624,
+      "grad_norm": 0.22995246946811676,
+      "learning_rate": 2.3232770020456368e-05,
+      "loss": 0.2609,
+      "step": 5827
+    },
+    {
+      "epoch": 4.386902521640948,
+      "grad_norm": 0.22000914812088013,
+      "learning_rate": 2.322546468050069e-05,
+      "loss": 0.2533,
+      "step": 5828
+    },
+    {
+      "epoch": 4.387655250282274,
+      "grad_norm": 0.1963583081960678,
+      "learning_rate": 2.3218159808001766e-05,
+      "loss": 0.2663,
+      "step": 5829
+    },
+    {
+      "epoch": 4.388407978923598,
+      "grad_norm": 0.20528173446655273,
+      "learning_rate": 2.3210855403758453e-05,
+      "loss": 0.1681,
+      "step": 5830
+    },
+    {
+      "epoch": 4.389160707564923,
+      "grad_norm": 0.2188722938299179,
+      "learning_rate": 2.320355146856954e-05,
+      "loss": 0.25,
+      "step": 5831
+    },
+    {
+      "epoch": 4.389913436206248,
+      "grad_norm": 0.24888625741004944,
+      "learning_rate": 2.3196248003233777e-05,
+      "loss": 0.3,
+      "step": 5832
+    },
+    {
+      "epoch": 4.390666164847572,
+      "grad_norm": 0.21731002628803253,
+      "learning_rate": 2.3188945008549856e-05,
+      "loss": 0.1841,
+      "step": 5833
+    },
+    {
+      "epoch": 4.391418893488897,
+      "grad_norm": 0.1994468867778778,
+      "learning_rate": 2.3181642485316428e-05,
+      "loss": 0.1841,
+      "step": 5834
+    },
+    {
+      "epoch": 4.392171622130222,
+      "grad_norm": 0.2336721271276474,
+      "learning_rate": 2.3174340434332097e-05,
+      "loss": 0.2338,
+      "step": 5835
+    },
+    {
+      "epoch": 4.392924350771547,
+      "grad_norm": 0.20098024606704712,
+      "learning_rate": 2.3167038856395383e-05,
+      "loss": 0.1451,
+      "step": 5836
+    },
+    {
+      "epoch": 4.393677079412871,
+      "grad_norm": 0.2400442510843277,
+      "learning_rate": 2.315973775230479e-05,
+      "loss": 0.1899,
+      "step": 5837
+    },
+    {
+      "epoch": 4.394429808054197,
+      "grad_norm": 0.24980656802654266,
+      "learning_rate": 2.315243712285875e-05,
+      "loss": 0.2791,
+      "step": 5838
+    },
+    {
+      "epoch": 4.395182536695521,
+      "grad_norm": 0.22405582666397095,
+      "learning_rate": 2.3145136968855646e-05,
+      "loss": 0.1802,
+      "step": 5839
+    },
+    {
+      "epoch": 4.395935265336846,
+      "grad_norm": 0.21176199615001678,
+      "learning_rate": 2.3137837291093818e-05,
+      "loss": 0.1768,
+      "step": 5840
+    },
+    {
+      "epoch": 4.396687993978171,
+      "grad_norm": 0.2027125209569931,
+      "learning_rate": 2.3130538090371547e-05,
+      "loss": 0.1979,
+      "step": 5841
+    },
+    {
+      "epoch": 4.397440722619495,
+      "grad_norm": 0.22128911316394806,
+      "learning_rate": 2.312323936748706e-05,
+      "loss": 0.1441,
+      "step": 5842
+    },
+    {
+      "epoch": 4.398193451260821,
+      "grad_norm": 0.21274779736995697,
+      "learning_rate": 2.311594112323854e-05,
+      "loss": 0.1888,
+      "step": 5843
+    },
+    {
+      "epoch": 4.398946179902145,
+      "grad_norm": 0.21141603589057922,
+      "learning_rate": 2.3108643358424104e-05,
+      "loss": 0.1732,
+      "step": 5844
+    },
+    {
+      "epoch": 4.39969890854347,
+      "grad_norm": 0.21857298910617828,
+      "learning_rate": 2.310134607384183e-05,
+      "loss": 0.1815,
+      "step": 5845
+    },
+    {
+      "epoch": 4.400451637184795,
+      "grad_norm": 0.17543086409568787,
+      "learning_rate": 2.3094049270289718e-05,
+      "loss": 0.1148,
+      "step": 5846
+    },
+    {
+      "epoch": 4.4012043658261195,
+      "grad_norm": 0.20096145570278168,
+      "learning_rate": 2.308675294856577e-05,
+      "loss": 0.1836,
+      "step": 5847
+    },
+    {
+      "epoch": 4.401957094467445,
+      "grad_norm": 0.2211276739835739,
+      "learning_rate": 2.307945710946789e-05,
+      "loss": 0.2829,
+      "step": 5848
+    },
+    {
+      "epoch": 4.402709823108769,
+      "grad_norm": 0.20063939690589905,
+      "learning_rate": 2.3072161753793924e-05,
+      "loss": 0.1725,
+      "step": 5849
+    },
+    {
+      "epoch": 4.403462551750094,
+      "grad_norm": 0.21214713156223297,
+      "learning_rate": 2.3064866882341696e-05,
+      "loss": 0.19,
+      "step": 5850
+    },
+    {
+      "epoch": 4.404215280391419,
+      "grad_norm": 0.2030765861272812,
+      "learning_rate": 2.3057572495908962e-05,
+      "loss": 0.1764,
+      "step": 5851
+    },
+    {
+      "epoch": 4.404968009032744,
+      "grad_norm": 0.21976026892662048,
+      "learning_rate": 2.3050278595293405e-05,
+      "loss": 0.3273,
+      "step": 5852
+    },
+    {
+      "epoch": 4.405720737674068,
+      "grad_norm": 0.20030593872070312,
+      "learning_rate": 2.3042985181292703e-05,
+      "loss": 0.139,
+      "step": 5853
+    },
+    {
+      "epoch": 4.4064734663153935,
+      "grad_norm": 0.21972130239009857,
+      "learning_rate": 2.303569225470444e-05,
+      "loss": 0.2123,
+      "step": 5854
+    },
+    {
+      "epoch": 4.407226194956718,
+      "grad_norm": 0.2778966426849365,
+      "learning_rate": 2.302839981632616e-05,
+      "loss": 0.3396,
+      "step": 5855
+    },
+    {
+      "epoch": 4.407978923598042,
+      "grad_norm": 0.2082005739212036,
+      "learning_rate": 2.3021107866955353e-05,
+      "loss": 0.2399,
+      "step": 5856
+    },
+    {
+      "epoch": 4.408731652239368,
+      "grad_norm": 0.24052490293979645,
+      "learning_rate": 2.3013816407389454e-05,
+      "loss": 0.231,
+      "step": 5857
+    },
+    {
+      "epoch": 4.409484380880692,
+      "grad_norm": 0.21924103796482086,
+      "learning_rate": 2.3006525438425845e-05,
+      "loss": 0.1844,
+      "step": 5858
+    },
+    {
+      "epoch": 4.410237109522018,
+      "grad_norm": 0.2158931940793991,
+      "learning_rate": 2.299923496086186e-05,
+      "loss": 0.2037,
+      "step": 5859
+    },
+    {
+      "epoch": 4.410989838163342,
+      "grad_norm": 0.21137741208076477,
+      "learning_rate": 2.2991944975494777e-05,
+      "loss": 0.263,
+      "step": 5860
+    },
+    {
+      "epoch": 4.411742566804667,
+      "grad_norm": 0.19067691266536713,
+      "learning_rate": 2.298465548312181e-05,
+      "loss": 0.189,
+      "step": 5861
+    },
+    {
+      "epoch": 4.412495295445992,
+      "grad_norm": 0.19958990812301636,
+      "learning_rate": 2.297736648454013e-05,
+      "loss": 0.2472,
+      "step": 5862
+    },
+    {
+      "epoch": 4.413248024087316,
+      "grad_norm": 0.2225978523492813,
+      "learning_rate": 2.2970077980546857e-05,
+      "loss": 0.2134,
+      "step": 5863
+    },
+    {
+      "epoch": 4.414000752728642,
+      "grad_norm": 0.1920253187417984,
+      "learning_rate": 2.296278997193904e-05,
+      "loss": 0.1931,
+      "step": 5864
+    },
+    {
+      "epoch": 4.414753481369966,
+      "grad_norm": 0.22148722410202026,
+      "learning_rate": 2.295550245951369e-05,
+      "loss": 0.2635,
+      "step": 5865
+    },
+    {
+      "epoch": 4.415506210011291,
+      "grad_norm": 0.18176910281181335,
+      "learning_rate": 2.2948215444067755e-05,
+      "loss": 0.1651,
+      "step": 5866
+    },
+    {
+      "epoch": 4.416258938652616,
+      "grad_norm": 0.20582085847854614,
+      "learning_rate": 2.294092892639814e-05,
+      "loss": 0.1822,
+      "step": 5867
+    },
+    {
+      "epoch": 4.417011667293941,
+      "grad_norm": 0.1956900805234909,
+      "learning_rate": 2.2933642907301688e-05,
+      "loss": 0.21,
+      "step": 5868
+    },
+    {
+      "epoch": 4.417764395935265,
+      "grad_norm": 0.22704148292541504,
+      "learning_rate": 2.2926357387575182e-05,
+      "loss": 0.2263,
+      "step": 5869
+    },
+    {
+      "epoch": 4.41851712457659,
+      "grad_norm": 0.20658884942531586,
+      "learning_rate": 2.2919072368015354e-05,
+      "loss": 0.176,
+      "step": 5870
+    },
+    {
+      "epoch": 4.419269853217915,
+      "grad_norm": 0.1976558417081833,
+      "learning_rate": 2.2911787849418874e-05,
+      "loss": 0.1802,
+      "step": 5871
+    },
+    {
+      "epoch": 4.420022581859239,
+      "grad_norm": 0.23012366890907288,
+      "learning_rate": 2.2904503832582384e-05,
+      "loss": 0.162,
+      "step": 5872
+    },
+    {
+      "epoch": 4.420775310500565,
+      "grad_norm": 0.17563797533512115,
+      "learning_rate": 2.289722031830245e-05,
+      "loss": 0.2006,
+      "step": 5873
+    },
+    {
+      "epoch": 4.421528039141889,
+      "grad_norm": 0.19130612909793854,
+      "learning_rate": 2.2889937307375582e-05,
+      "loss": 0.2253,
+      "step": 5874
+    },
+    {
+      "epoch": 4.4222807677832146,
+      "grad_norm": 0.2078777402639389,
+      "learning_rate": 2.288265480059823e-05,
+      "loss": 0.1271,
+      "step": 5875
+    },
+    {
+      "epoch": 4.423033496424539,
+      "grad_norm": 0.23767942190170288,
+      "learning_rate": 2.287537279876682e-05,
+      "loss": 0.2652,
+      "step": 5876
+    },
+    {
+      "epoch": 4.4237862250658635,
+      "grad_norm": 0.18577352166175842,
+      "learning_rate": 2.286809130267768e-05,
+      "loss": 0.2153,
+      "step": 5877
+    },
+    {
+      "epoch": 4.424538953707189,
+      "grad_norm": 0.27499136328697205,
+      "learning_rate": 2.2860810313127102e-05,
+      "loss": 0.3016,
+      "step": 5878
+    },
+    {
+      "epoch": 4.425291682348513,
+      "grad_norm": 0.2277744710445404,
+      "learning_rate": 2.2853529830911342e-05,
+      "loss": 0.2577,
+      "step": 5879
+    },
+    {
+      "epoch": 4.426044410989838,
+      "grad_norm": 0.20763546228408813,
+      "learning_rate": 2.2846249856826578e-05,
+      "loss": 0.1471,
+      "step": 5880
+    },
+    {
+      "epoch": 4.426797139631163,
+      "grad_norm": 0.2440299391746521,
+      "learning_rate": 2.283897039166893e-05,
+      "loss": 0.1958,
+      "step": 5881
+    },
+    {
+      "epoch": 4.427549868272488,
+      "grad_norm": 0.24477456510066986,
+      "learning_rate": 2.283169143623447e-05,
+      "loss": 0.2151,
+      "step": 5882
+    },
+    {
+      "epoch": 4.428302596913813,
+      "grad_norm": 0.18469750881195068,
+      "learning_rate": 2.2824412991319217e-05,
+      "loss": 0.1839,
+      "step": 5883
+    },
+    {
+      "epoch": 4.4290553255551375,
+      "grad_norm": 0.22322650253772736,
+      "learning_rate": 2.2817135057719123e-05,
+      "loss": 0.2374,
+      "step": 5884
+    },
+    {
+      "epoch": 4.429808054196462,
+      "grad_norm": 0.22928397357463837,
+      "learning_rate": 2.280985763623011e-05,
+      "loss": 0.2289,
+      "step": 5885
+    },
+    {
+      "epoch": 4.430560782837787,
+      "grad_norm": 0.19534185528755188,
+      "learning_rate": 2.280258072764801e-05,
+      "loss": 0.2304,
+      "step": 5886
+    },
+    {
+      "epoch": 4.431313511479112,
+      "grad_norm": 0.2178461253643036,
+      "learning_rate": 2.2795304332768612e-05,
+      "loss": 0.201,
+      "step": 5887
+    },
+    {
+      "epoch": 4.432066240120436,
+      "grad_norm": 0.20946615934371948,
+      "learning_rate": 2.2788028452387672e-05,
+      "loss": 0.2177,
+      "step": 5888
+    },
+    {
+      "epoch": 4.432818968761762,
+      "grad_norm": 0.23207977414131165,
+      "learning_rate": 2.278075308730085e-05,
+      "loss": 0.2145,
+      "step": 5889
+    },
+    {
+      "epoch": 4.433571697403086,
+      "grad_norm": 0.21326874196529388,
+      "learning_rate": 2.2773478238303776e-05,
+      "loss": 0.2099,
+      "step": 5890
+    },
+    {
+      "epoch": 4.434324426044411,
+      "grad_norm": 0.2287229597568512,
+      "learning_rate": 2.276620390619201e-05,
+      "loss": 0.2316,
+      "step": 5891
+    },
+    {
+      "epoch": 4.435077154685736,
+      "grad_norm": 0.21407033503055573,
+      "learning_rate": 2.275893009176107e-05,
+      "loss": 0.1736,
+      "step": 5892
+    },
+    {
+      "epoch": 4.43582988332706,
+      "grad_norm": 0.2227305918931961,
+      "learning_rate": 2.2751656795806415e-05,
+      "loss": 0.2205,
+      "step": 5893
+    },
+    {
+      "epoch": 4.436582611968386,
+      "grad_norm": 0.22730867564678192,
+      "learning_rate": 2.2744384019123427e-05,
+      "loss": 0.1968,
+      "step": 5894
+    },
+    {
+      "epoch": 4.43733534060971,
+      "grad_norm": 0.17386667430400848,
+      "learning_rate": 2.273711176250745e-05,
+      "loss": 0.194,
+      "step": 5895
+    },
+    {
+      "epoch": 4.438088069251035,
+      "grad_norm": 0.21332743763923645,
+      "learning_rate": 2.272984002675378e-05,
+      "loss": 0.2465,
+      "step": 5896
+    },
+    {
+      "epoch": 4.43884079789236,
+      "grad_norm": 0.21738651394844055,
+      "learning_rate": 2.2722568812657613e-05,
+      "loss": 0.1972,
+      "step": 5897
+    },
+    {
+      "epoch": 4.4395935265336846,
+      "grad_norm": 0.21935707330703735,
+      "learning_rate": 2.2715298121014154e-05,
+      "loss": 0.2947,
+      "step": 5898
+    },
+    {
+      "epoch": 4.440346255175009,
+      "grad_norm": 0.21586064994335175,
+      "learning_rate": 2.2708027952618498e-05,
+      "loss": 0.241,
+      "step": 5899
+    },
+    {
+      "epoch": 4.441098983816334,
+      "grad_norm": 0.22716259956359863,
+      "learning_rate": 2.27007583082657e-05,
+      "loss": 0.1457,
+      "step": 5900
+    },
+    {
+      "epoch": 4.441851712457659,
+      "grad_norm": 0.19145599007606506,
+      "learning_rate": 2.2693489188750754e-05,
+      "loss": 0.1488,
+      "step": 5901
+    },
+    {
+      "epoch": 4.442604441098984,
+      "grad_norm": 0.22248436510562897,
+      "learning_rate": 2.268622059486861e-05,
+      "loss": 0.2538,
+      "step": 5902
+    },
+    {
+      "epoch": 4.443357169740309,
+      "grad_norm": 0.20645944774150848,
+      "learning_rate": 2.267895252741413e-05,
+      "loss": 0.1652,
+      "step": 5903
+    },
+    {
+      "epoch": 4.444109898381633,
+      "grad_norm": 0.2221689522266388,
+      "learning_rate": 2.2671684987182166e-05,
+      "loss": 0.2321,
+      "step": 5904
+    },
+    {
+      "epoch": 4.4448626270229585,
+      "grad_norm": 0.23507216572761536,
+      "learning_rate": 2.266441797496747e-05,
+      "loss": 0.2518,
+      "step": 5905
+    },
+    {
+      "epoch": 4.445615355664283,
+      "grad_norm": 0.2594968378543854,
+      "learning_rate": 2.265715149156476e-05,
+      "loss": 0.212,
+      "step": 5906
+    },
+    {
+      "epoch": 4.4463680843056075,
+      "grad_norm": 0.18105386197566986,
+      "learning_rate": 2.264988553776868e-05,
+      "loss": 0.192,
+      "step": 5907
+    },
+    {
+      "epoch": 4.447120812946933,
+      "grad_norm": 0.22198201715946198,
+      "learning_rate": 2.2642620114373812e-05,
+      "loss": 0.2493,
+      "step": 5908
+    },
+    {
+      "epoch": 4.447873541588257,
+      "grad_norm": 0.2074604481458664,
+      "learning_rate": 2.263535522217472e-05,
+      "loss": 0.2012,
+      "step": 5909
+    },
+    {
+      "epoch": 4.448626270229582,
+      "grad_norm": 0.21651427447795868,
+      "learning_rate": 2.262809086196586e-05,
+      "loss": 0.0863,
+      "step": 5910
+    },
+    {
+      "epoch": 4.449378998870907,
+      "grad_norm": 0.22907137870788574,
+      "learning_rate": 2.2620827034541664e-05,
+      "loss": 0.258,
+      "step": 5911
+    },
+    {
+      "epoch": 4.450131727512232,
+      "grad_norm": 0.2079429179430008,
+      "learning_rate": 2.261356374069648e-05,
+      "loss": 0.1717,
+      "step": 5912
+    },
+    {
+      "epoch": 4.450884456153557,
+      "grad_norm": 0.23144954442977905,
+      "learning_rate": 2.260630098122462e-05,
+      "loss": 0.1711,
+      "step": 5913
+    },
+    {
+      "epoch": 4.4516371847948815,
+      "grad_norm": 0.22690144181251526,
+      "learning_rate": 2.2599038756920333e-05,
+      "loss": 0.2313,
+      "step": 5914
+    },
+    {
+      "epoch": 4.452389913436206,
+      "grad_norm": 0.20382456481456757,
+      "learning_rate": 2.2591777068577803e-05,
+      "loss": 0.2879,
+      "step": 5915
+    },
+    {
+      "epoch": 4.453142642077531,
+      "grad_norm": 0.19985562562942505,
+      "learning_rate": 2.2584515916991134e-05,
+      "loss": 0.2194,
+      "step": 5916
+    },
+    {
+      "epoch": 4.453895370718856,
+      "grad_norm": 0.22958821058273315,
+      "learning_rate": 2.2577255302954424e-05,
+      "loss": 0.2405,
+      "step": 5917
+    },
+    {
+      "epoch": 4.45464809936018,
+      "grad_norm": 0.2159969061613083,
+      "learning_rate": 2.256999522726168e-05,
+      "loss": 0.1455,
+      "step": 5918
+    },
+    {
+      "epoch": 4.455400828001506,
+      "grad_norm": 0.20597754418849945,
+      "learning_rate": 2.2562735690706843e-05,
+      "loss": 0.2299,
+      "step": 5919
+    },
+    {
+      "epoch": 4.45615355664283,
+      "grad_norm": 0.23310381174087524,
+      "learning_rate": 2.2555476694083804e-05,
+      "loss": 0.2425,
+      "step": 5920
+    },
+    {
+      "epoch": 4.4569062852841554,
+      "grad_norm": 0.2156984806060791,
+      "learning_rate": 2.2548218238186408e-05,
+      "loss": 0.1755,
+      "step": 5921
+    },
+    {
+      "epoch": 4.45765901392548,
+      "grad_norm": 0.22072210907936096,
+      "learning_rate": 2.2540960323808407e-05,
+      "loss": 0.1922,
+      "step": 5922
+    },
+    {
+      "epoch": 4.458411742566804,
+      "grad_norm": 0.22004330158233643,
+      "learning_rate": 2.2533702951743536e-05,
+      "loss": 0.1601,
+      "step": 5923
+    },
+    {
+      "epoch": 4.45916447120813,
+      "grad_norm": 0.1883590668439865,
+      "learning_rate": 2.2526446122785443e-05,
+      "loss": 0.2083,
+      "step": 5924
+    },
+    {
+      "epoch": 4.459917199849454,
+      "grad_norm": 0.25297853350639343,
+      "learning_rate": 2.2519189837727727e-05,
+      "loss": 0.148,
+      "step": 5925
+    },
+    {
+      "epoch": 4.460669928490779,
+      "grad_norm": 0.20887066423892975,
+      "learning_rate": 2.2511934097363923e-05,
+      "loss": 0.2383,
+      "step": 5926
+    },
+    {
+      "epoch": 4.461422657132104,
+      "grad_norm": 0.20831404626369476,
+      "learning_rate": 2.2504678902487508e-05,
+      "loss": 0.1563,
+      "step": 5927
+    },
+    {
+      "epoch": 4.4621753857734285,
+      "grad_norm": 0.24241118133068085,
+      "learning_rate": 2.249742425389189e-05,
+      "loss": 0.236,
+      "step": 5928
+    },
+    {
+      "epoch": 4.462928114414753,
+      "grad_norm": 0.2052774429321289,
+      "learning_rate": 2.249017015237043e-05,
+      "loss": 0.1525,
+      "step": 5929
+    },
+    {
+      "epoch": 4.463680843056078,
+      "grad_norm": 0.21553248167037964,
+      "learning_rate": 2.248291659871644e-05,
+      "loss": 0.1318,
+      "step": 5930
+    },
+    {
+      "epoch": 4.464433571697403,
+      "grad_norm": 0.22765308618545532,
+      "learning_rate": 2.247566359372315e-05,
+      "loss": 0.1792,
+      "step": 5931
+    },
+    {
+      "epoch": 4.465186300338728,
+      "grad_norm": 0.19577588140964508,
+      "learning_rate": 2.2468411138183732e-05,
+      "loss": 0.1986,
+      "step": 5932
+    },
+    {
+      "epoch": 4.465939028980053,
+      "grad_norm": 0.20928321778774261,
+      "learning_rate": 2.2461159232891298e-05,
+      "loss": 0.1536,
+      "step": 5933
+    },
+    {
+      "epoch": 4.466691757621377,
+      "grad_norm": 0.21376071870326996,
+      "learning_rate": 2.245390787863893e-05,
+      "loss": 0.179,
+      "step": 5934
+    },
+    {
+      "epoch": 4.4674444862627025,
+      "grad_norm": 0.1893610805273056,
+      "learning_rate": 2.2446657076219594e-05,
+      "loss": 0.1522,
+      "step": 5935
+    },
+    {
+      "epoch": 4.468197214904027,
+      "grad_norm": 0.21366159617900848,
+      "learning_rate": 2.2439406826426256e-05,
+      "loss": 0.2575,
+      "step": 5936
+    },
+    {
+      "epoch": 4.468949943545352,
+      "grad_norm": 0.22487635910511017,
+      "learning_rate": 2.2432157130051766e-05,
+      "loss": 0.2623,
+      "step": 5937
+    },
+    {
+      "epoch": 4.469702672186677,
+      "grad_norm": 0.2680451273918152,
+      "learning_rate": 2.2424907987888965e-05,
+      "loss": 0.2378,
+      "step": 5938
+    },
+    {
+      "epoch": 4.470455400828001,
+      "grad_norm": 0.19120576977729797,
+      "learning_rate": 2.2417659400730595e-05,
+      "loss": 0.1948,
+      "step": 5939
+    },
+    {
+      "epoch": 4.471208129469327,
+      "grad_norm": 0.2179485410451889,
+      "learning_rate": 2.241041136936935e-05,
+      "loss": 0.198,
+      "step": 5940
+    },
+    {
+      "epoch": 4.471960858110651,
+      "grad_norm": 0.23150652647018433,
+      "learning_rate": 2.2403163894597857e-05,
+      "loss": 0.1418,
+      "step": 5941
+    },
+    {
+      "epoch": 4.472713586751976,
+      "grad_norm": 0.20081833004951477,
+      "learning_rate": 2.2395916977208708e-05,
+      "loss": 0.1827,
+      "step": 5942
+    },
+    {
+      "epoch": 4.473466315393301,
+      "grad_norm": 0.22906944155693054,
+      "learning_rate": 2.23886706179944e-05,
+      "loss": 0.1984,
+      "step": 5943
+    },
+    {
+      "epoch": 4.4742190440346254,
+      "grad_norm": 0.1859503537416458,
+      "learning_rate": 2.2381424817747387e-05,
+      "loss": 0.1827,
+      "step": 5944
+    },
+    {
+      "epoch": 4.47497177267595,
+      "grad_norm": 0.23732125759124756,
+      "learning_rate": 2.2374179577260068e-05,
+      "loss": 0.2305,
+      "step": 5945
+    },
+    {
+      "epoch": 4.475724501317275,
+      "grad_norm": 0.21504786610603333,
+      "learning_rate": 2.236693489732476e-05,
+      "loss": 0.2199,
+      "step": 5946
+    },
+    {
+      "epoch": 4.4764772299586,
+      "grad_norm": 0.19255812466144562,
+      "learning_rate": 2.2359690778733734e-05,
+      "loss": 0.1689,
+      "step": 5947
+    },
+    {
+      "epoch": 4.477229958599925,
+      "grad_norm": 0.1827113926410675,
+      "learning_rate": 2.2352447222279187e-05,
+      "loss": 0.1629,
+      "step": 5948
+    },
+    {
+      "epoch": 4.47798268724125,
+      "grad_norm": 0.2347870022058487,
+      "learning_rate": 2.234520422875328e-05,
+      "loss": 0.1906,
+      "step": 5949
+    },
+    {
+      "epoch": 4.478735415882574,
+      "grad_norm": 0.20953159034252167,
+      "learning_rate": 2.233796179894809e-05,
+      "loss": 0.1549,
+      "step": 5950
+    },
+    {
+      "epoch": 4.479488144523899,
+      "grad_norm": 0.20863528549671173,
+      "learning_rate": 2.233071993365564e-05,
+      "loss": 0.1657,
+      "step": 5951
+    },
+    {
+      "epoch": 4.480240873165224,
+      "grad_norm": 0.2110145539045334,
+      "learning_rate": 2.2323478633667883e-05,
+      "loss": 0.2352,
+      "step": 5952
+    },
+    {
+      "epoch": 4.480993601806548,
+      "grad_norm": 0.2276211977005005,
+      "learning_rate": 2.231623789977672e-05,
+      "loss": 0.251,
+      "step": 5953
+    },
+    {
+      "epoch": 4.481746330447874,
+      "grad_norm": 0.19553864002227783,
+      "learning_rate": 2.2308997732773977e-05,
+      "loss": 0.1769,
+      "step": 5954
+    },
+    {
+      "epoch": 4.482499059089198,
+      "grad_norm": 0.21348801255226135,
+      "learning_rate": 2.2301758133451446e-05,
+      "loss": 0.2251,
+      "step": 5955
+    },
+    {
+      "epoch": 4.483251787730524,
+      "grad_norm": 0.21506120264530182,
+      "learning_rate": 2.2294519102600837e-05,
+      "loss": 0.2811,
+      "step": 5956
+    },
+    {
+      "epoch": 4.484004516371848,
+      "grad_norm": 0.19159533083438873,
+      "learning_rate": 2.2287280641013787e-05,
+      "loss": 0.242,
+      "step": 5957
+    },
+    {
+      "epoch": 4.4847572450131725,
+      "grad_norm": 0.21396741271018982,
+      "learning_rate": 2.228004274948188e-05,
+      "loss": 0.2327,
+      "step": 5958
+    },
+    {
+      "epoch": 4.485509973654498,
+      "grad_norm": 0.22806324064731598,
+      "learning_rate": 2.2272805428796655e-05,
+      "loss": 0.2048,
+      "step": 5959
+    },
+    {
+      "epoch": 4.486262702295822,
+      "grad_norm": 0.20971858501434326,
+      "learning_rate": 2.226556867974957e-05,
+      "loss": 0.236,
+      "step": 5960
+    },
+    {
+      "epoch": 4.487015430937147,
+      "grad_norm": 0.2072150707244873,
+      "learning_rate": 2.225833250313203e-05,
+      "loss": 0.1961,
+      "step": 5961
+    },
+    {
+      "epoch": 4.487768159578472,
+      "grad_norm": 0.19223718345165253,
+      "learning_rate": 2.2251096899735358e-05,
+      "loss": 0.1807,
+      "step": 5962
+    },
+    {
+      "epoch": 4.488520888219797,
+      "grad_norm": 0.217403382062912,
+      "learning_rate": 2.224386187035084e-05,
+      "loss": 0.1787,
+      "step": 5963
+    },
+    {
+      "epoch": 4.489273616861121,
+      "grad_norm": 0.21444962918758392,
+      "learning_rate": 2.223662741576969e-05,
+      "loss": 0.1709,
+      "step": 5964
+    },
+    {
+      "epoch": 4.4900263455024465,
+      "grad_norm": 0.20360417664051056,
+      "learning_rate": 2.2229393536783056e-05,
+      "loss": 0.2344,
+      "step": 5965
+    },
+    {
+      "epoch": 4.490779074143771,
+      "grad_norm": 0.20954005420207977,
+      "learning_rate": 2.222216023418201e-05,
+      "loss": 0.2372,
+      "step": 5966
+    },
+    {
+      "epoch": 4.491531802785096,
+      "grad_norm": 0.2054355889558792,
+      "learning_rate": 2.2214927508757584e-05,
+      "loss": 0.2081,
+      "step": 5967
+    },
+    {
+      "epoch": 4.492284531426421,
+      "grad_norm": 0.20840097963809967,
+      "learning_rate": 2.2207695361300752e-05,
+      "loss": 0.1826,
+      "step": 5968
+    },
+    {
+      "epoch": 4.493037260067745,
+      "grad_norm": 0.21900032460689545,
+      "learning_rate": 2.2200463792602387e-05,
+      "loss": 0.1649,
+      "step": 5969
+    },
+    {
+      "epoch": 4.493789988709071,
+      "grad_norm": 0.21423828601837158,
+      "learning_rate": 2.219323280345334e-05,
+      "loss": 0.2397,
+      "step": 5970
+    },
+    {
+      "epoch": 4.494542717350395,
+      "grad_norm": 0.18508470058441162,
+      "learning_rate": 2.218600239464438e-05,
+      "loss": 0.1021,
+      "step": 5971
+    },
+    {
+      "epoch": 4.49529544599172,
+      "grad_norm": 0.217258021235466,
+      "learning_rate": 2.21787725669662e-05,
+      "loss": 0.1654,
+      "step": 5972
+    },
+    {
+      "epoch": 4.496048174633045,
+      "grad_norm": 0.1994309276342392,
+      "learning_rate": 2.217154332120945e-05,
+      "loss": 0.2501,
+      "step": 5973
+    },
+    {
+      "epoch": 4.496800903274369,
+      "grad_norm": 0.2308073341846466,
+      "learning_rate": 2.2164314658164714e-05,
+      "loss": 0.3274,
+      "step": 5974
+    },
+    {
+      "epoch": 4.497553631915695,
+      "grad_norm": 0.2269303798675537,
+      "learning_rate": 2.215708657862251e-05,
+      "loss": 0.2545,
+      "step": 5975
+    },
+    {
+      "epoch": 4.498306360557019,
+      "grad_norm": 0.17760542035102844,
+      "learning_rate": 2.214985908337328e-05,
+      "loss": 0.1209,
+      "step": 5976
+    },
+    {
+      "epoch": 4.499059089198344,
+      "grad_norm": 0.21166902780532837,
+      "learning_rate": 2.2142632173207424e-05,
+      "loss": 0.2136,
+      "step": 5977
+    },
+    {
+      "epoch": 4.499811817839669,
+      "grad_norm": 0.22115717828273773,
+      "learning_rate": 2.2135405848915253e-05,
+      "loss": 0.167,
+      "step": 5978
+    },
+    {
+      "epoch": 4.500564546480994,
+      "grad_norm": 0.1974421888589859,
+      "learning_rate": 2.212818011128703e-05,
+      "loss": 0.187,
+      "step": 5979
+    },
+    {
+      "epoch": 4.501317275122318,
+      "grad_norm": 0.2029014378786087,
+      "learning_rate": 2.212095496111295e-05,
+      "loss": 0.1725,
+      "step": 5980
+    },
+    {
+      "epoch": 4.502070003763643,
+      "grad_norm": 0.2191023826599121,
+      "learning_rate": 2.211373039918316e-05,
+      "loss": 0.2901,
+      "step": 5981
+    },
+    {
+      "epoch": 4.502822732404968,
+      "grad_norm": 0.186708003282547,
+      "learning_rate": 2.210650642628771e-05,
+      "loss": 0.2826,
+      "step": 5982
+    },
+    {
+      "epoch": 4.503575461046292,
+      "grad_norm": 0.18669098615646362,
+      "learning_rate": 2.2099283043216607e-05,
+      "loss": 0.1848,
+      "step": 5983
+    },
+    {
+      "epoch": 4.504328189687618,
+      "grad_norm": 0.22114834189414978,
+      "learning_rate": 2.209206025075979e-05,
+      "loss": 0.2456,
+      "step": 5984
+    },
+    {
+      "epoch": 4.505080918328942,
+      "grad_norm": 0.20732660591602325,
+      "learning_rate": 2.2084838049707135e-05,
+      "loss": 0.2151,
+      "step": 5985
+    },
+    {
+      "epoch": 4.505833646970268,
+      "grad_norm": 0.211372971534729,
+      "learning_rate": 2.2077616440848442e-05,
+      "loss": 0.1787,
+      "step": 5986
+    },
+    {
+      "epoch": 4.506586375611592,
+      "grad_norm": 0.19332151114940643,
+      "learning_rate": 2.2070395424973473e-05,
+      "loss": 0.2227,
+      "step": 5987
+    },
+    {
+      "epoch": 4.5073391042529165,
+      "grad_norm": 0.1696118712425232,
+      "learning_rate": 2.2063175002871893e-05,
+      "loss": 0.0892,
+      "step": 5988
+    },
+    {
+      "epoch": 4.508091832894242,
+      "grad_norm": 0.18153659999370575,
+      "learning_rate": 2.2055955175333327e-05,
+      "loss": 0.2379,
+      "step": 5989
+    },
+    {
+      "epoch": 4.508844561535566,
+      "grad_norm": 0.2079237401485443,
+      "learning_rate": 2.2048735943147314e-05,
+      "loss": 0.2303,
+      "step": 5990
+    },
+    {
+      "epoch": 4.509597290176892,
+      "grad_norm": 0.18532858788967133,
+      "learning_rate": 2.2041517307103337e-05,
+      "loss": 0.2391,
+      "step": 5991
+    },
+    {
+      "epoch": 4.510350018818216,
+      "grad_norm": 0.2105163037776947,
+      "learning_rate": 2.2034299267990817e-05,
+      "loss": 0.2039,
+      "step": 5992
+    },
+    {
+      "epoch": 4.511102747459541,
+      "grad_norm": 0.19783370196819305,
+      "learning_rate": 2.202708182659912e-05,
+      "loss": 0.1467,
+      "step": 5993
+    },
+    {
+      "epoch": 4.511855476100866,
+      "grad_norm": 0.2038516104221344,
+      "learning_rate": 2.201986498371753e-05,
+      "loss": 0.1899,
+      "step": 5994
+    },
+    {
+      "epoch": 4.5126082047421905,
+      "grad_norm": 0.252783864736557,
+      "learning_rate": 2.201264874013526e-05,
+      "loss": 0.2959,
+      "step": 5995
+    },
+    {
+      "epoch": 4.513360933383515,
+      "grad_norm": 0.19337938725948334,
+      "learning_rate": 2.2005433096641477e-05,
+      "loss": 0.1351,
+      "step": 5996
+    },
+    {
+      "epoch": 4.51411366202484,
+      "grad_norm": 0.2185029685497284,
+      "learning_rate": 2.199821805402527e-05,
+      "loss": 0.2224,
+      "step": 5997
+    },
+    {
+      "epoch": 4.514866390666165,
+      "grad_norm": 0.22388476133346558,
+      "learning_rate": 2.1991003613075662e-05,
+      "loss": 0.1983,
+      "step": 5998
+    },
+    {
+      "epoch": 4.515619119307489,
+      "grad_norm": 0.19236262142658234,
+      "learning_rate": 2.198378977458161e-05,
+      "loss": 0.1987,
+      "step": 5999
+    },
+    {
+      "epoch": 4.516371847948815,
+      "grad_norm": 0.244500070810318,
+      "learning_rate": 2.1976576539332024e-05,
+      "loss": 0.2267,
+      "step": 6000
+    },
+    {
+      "epoch": 4.516371847948815,
+      "eval_loss": 0.23185604810714722,
+      "eval_runtime": 452.7801,
+      "eval_samples_per_second": 21.262,
+      "eval_steps_per_second": 0.665,
+      "step": 6000
+    },
+    {
+      "epoch": 4.517124576590139,
+      "grad_norm": 0.20815491676330566,
+      "learning_rate": 2.196936390811572e-05,
+      "loss": 0.2019,
+      "step": 6001
+    },
+    {
+      "epoch": 4.517877305231464,
+      "grad_norm": 0.234881192445755,
+      "learning_rate": 2.1962151881721464e-05,
+      "loss": 0.2211,
+      "step": 6002
+    },
+    {
+      "epoch": 4.518630033872789,
+      "grad_norm": 0.21030071377754211,
+      "learning_rate": 2.195494046093795e-05,
+      "loss": 0.1821,
+      "step": 6003
+    },
+    {
+      "epoch": 4.519382762514113,
+      "grad_norm": 0.20557573437690735,
+      "learning_rate": 2.1947729646553817e-05,
+      "loss": 0.1869,
+      "step": 6004
+    },
+    {
+      "epoch": 4.520135491155439,
+      "grad_norm": 0.1948746144771576,
+      "learning_rate": 2.194051943935761e-05,
+      "loss": 0.2254,
+      "step": 6005
+    },
+    {
+      "epoch": 4.520888219796763,
+      "grad_norm": 0.2306148260831833,
+      "learning_rate": 2.193330984013785e-05,
+      "loss": 0.1902,
+      "step": 6006
+    },
+    {
+      "epoch": 4.521640948438088,
+      "grad_norm": 0.2330739051103592,
+      "learning_rate": 2.192610084968295e-05,
+      "loss": 0.1361,
+      "step": 6007
+    },
+    {
+      "epoch": 4.522393677079413,
+      "grad_norm": 0.1720714569091797,
+      "learning_rate": 2.1918892468781287e-05,
+      "loss": 0.2009,
+      "step": 6008
+    },
+    {
+      "epoch": 4.523146405720738,
+      "grad_norm": 0.22313304245471954,
+      "learning_rate": 2.1911684698221146e-05,
+      "loss": 0.2416,
+      "step": 6009
+    },
+    {
+      "epoch": 4.523899134362063,
+      "grad_norm": 0.2057969868183136,
+      "learning_rate": 2.1904477538790774e-05,
+      "loss": 0.1939,
+      "step": 6010
+    },
+    {
+      "epoch": 4.524651863003387,
+      "grad_norm": 0.2312842607498169,
+      "learning_rate": 2.1897270991278314e-05,
+      "loss": 0.2665,
+      "step": 6011
+    },
+    {
+      "epoch": 4.525404591644712,
+      "grad_norm": 0.2199803739786148,
+      "learning_rate": 2.1890065056471876e-05,
+      "loss": 0.2273,
+      "step": 6012
+    },
+    {
+      "epoch": 4.526157320286037,
+      "grad_norm": 0.2170543074607849,
+      "learning_rate": 2.1882859735159493e-05,
+      "loss": 0.1894,
+      "step": 6013
+    },
+    {
+      "epoch": 4.526910048927362,
+      "grad_norm": 0.21748404204845428,
+      "learning_rate": 2.187565502812913e-05,
+      "loss": 0.2162,
+      "step": 6014
+    },
+    {
+      "epoch": 4.527662777568686,
+      "grad_norm": 0.2432517111301422,
+      "learning_rate": 2.1868450936168673e-05,
+      "loss": 0.287,
+      "step": 6015
+    },
+    {
+      "epoch": 4.5284155062100115,
+      "grad_norm": 0.2140440195798874,
+      "learning_rate": 2.1861247460065954e-05,
+      "loss": 0.2288,
+      "step": 6016
+    },
+    {
+      "epoch": 4.529168234851336,
+      "grad_norm": 0.22343097627162933,
+      "learning_rate": 2.1854044600608737e-05,
+      "loss": 0.2925,
+      "step": 6017
+    },
+    {
+      "epoch": 4.5299209634926605,
+      "grad_norm": 0.18905530869960785,
+      "learning_rate": 2.18468423585847e-05,
+      "loss": 0.169,
+      "step": 6018
+    },
+    {
+      "epoch": 4.530673692133986,
+      "grad_norm": 0.2087632715702057,
+      "learning_rate": 2.1839640734781498e-05,
+      "loss": 0.2054,
+      "step": 6019
+    },
+    {
+      "epoch": 4.53142642077531,
+      "grad_norm": 0.22000370919704437,
+      "learning_rate": 2.1832439729986676e-05,
+      "loss": 0.2747,
+      "step": 6020
+    },
+    {
+      "epoch": 4.532179149416636,
+      "grad_norm": 0.1979513168334961,
+      "learning_rate": 2.182523934498772e-05,
+      "loss": 0.2012,
+      "step": 6021
+    },
+    {
+      "epoch": 4.53293187805796,
+      "grad_norm": 0.1724737137556076,
+      "learning_rate": 2.1818039580572052e-05,
+      "loss": 0.1322,
+      "step": 6022
+    },
+    {
+      "epoch": 4.533684606699285,
+      "grad_norm": 0.2008916586637497,
+      "learning_rate": 2.1810840437527043e-05,
+      "loss": 0.1963,
+      "step": 6023
+    },
+    {
+      "epoch": 4.53443733534061,
+      "grad_norm": 0.19235526025295258,
+      "learning_rate": 2.1803641916639954e-05,
+      "loss": 0.1975,
+      "step": 6024
+    },
+    {
+      "epoch": 4.5351900639819345,
+      "grad_norm": 0.2197679877281189,
+      "learning_rate": 2.1796444018698036e-05,
+      "loss": 0.1883,
+      "step": 6025
+    },
+    {
+      "epoch": 4.535942792623259,
+      "grad_norm": 0.22618500888347626,
+      "learning_rate": 2.1789246744488422e-05,
+      "loss": 0.1772,
+      "step": 6026
+    },
+    {
+      "epoch": 4.536695521264584,
+      "grad_norm": 0.22709430754184723,
+      "learning_rate": 2.1782050094798195e-05,
+      "loss": 0.2771,
+      "step": 6027
+    },
+    {
+      "epoch": 4.537448249905909,
+      "grad_norm": 0.2188844084739685,
+      "learning_rate": 2.1774854070414375e-05,
+      "loss": 0.3158,
+      "step": 6028
+    },
+    {
+      "epoch": 4.538200978547234,
+      "grad_norm": 0.2299719601869583,
+      "learning_rate": 2.176765867212391e-05,
+      "loss": 0.1876,
+      "step": 6029
+    },
+    {
+      "epoch": 4.538953707188559,
+      "grad_norm": 0.18854394555091858,
+      "learning_rate": 2.1760463900713662e-05,
+      "loss": 0.194,
+      "step": 6030
+    },
+    {
+      "epoch": 4.539706435829883,
+      "grad_norm": 0.19057026505470276,
+      "learning_rate": 2.1753269756970464e-05,
+      "loss": 0.1846,
+      "step": 6031
+    },
+    {
+      "epoch": 4.5404591644712085,
+      "grad_norm": 0.2715662121772766,
+      "learning_rate": 2.1746076241681034e-05,
+      "loss": 0.2194,
+      "step": 6032
+    },
+    {
+      "epoch": 4.541211893112533,
+      "grad_norm": 0.19353188574314117,
+      "learning_rate": 2.1738883355632068e-05,
+      "loss": 0.117,
+      "step": 6033
+    },
+    {
+      "epoch": 4.541964621753857,
+      "grad_norm": 0.2024213969707489,
+      "learning_rate": 2.1731691099610152e-05,
+      "loss": 0.2174,
+      "step": 6034
+    },
+    {
+      "epoch": 4.542717350395183,
+      "grad_norm": 0.19609083235263824,
+      "learning_rate": 2.1724499474401828e-05,
+      "loss": 0.245,
+      "step": 6035
+    },
+    {
+      "epoch": 4.543470079036507,
+      "grad_norm": 0.18351636826992035,
+      "learning_rate": 2.171730848079355e-05,
+      "loss": 0.2156,
+      "step": 6036
+    },
+    {
+      "epoch": 4.544222807677832,
+      "grad_norm": 0.2164335399866104,
+      "learning_rate": 2.171011811957172e-05,
+      "loss": 0.2886,
+      "step": 6037
+    },
+    {
+      "epoch": 4.544975536319157,
+      "grad_norm": 0.1939413845539093,
+      "learning_rate": 2.1702928391522676e-05,
+      "loss": 0.2054,
+      "step": 6038
+    },
+    {
+      "epoch": 4.5457282649604815,
+      "grad_norm": 0.25830018520355225,
+      "learning_rate": 2.169573929743267e-05,
+      "loss": 0.2164,
+      "step": 6039
+    },
+    {
+      "epoch": 4.546480993601807,
+      "grad_norm": 0.19659429788589478,
+      "learning_rate": 2.1688550838087884e-05,
+      "loss": 0.2484,
+      "step": 6040
+    },
+    {
+      "epoch": 4.547233722243131,
+      "grad_norm": 0.20814087986946106,
+      "learning_rate": 2.168136301427444e-05,
+      "loss": 0.1823,
+      "step": 6041
+    },
+    {
+      "epoch": 4.547986450884456,
+      "grad_norm": 0.2081860452890396,
+      "learning_rate": 2.1674175826778392e-05,
+      "loss": 0.1561,
+      "step": 6042
+    },
+    {
+      "epoch": 4.548739179525781,
+      "grad_norm": 0.2187824249267578,
+      "learning_rate": 2.1666989276385703e-05,
+      "loss": 0.2374,
+      "step": 6043
+    },
+    {
+      "epoch": 4.549491908167106,
+      "grad_norm": 0.19272252917289734,
+      "learning_rate": 2.165980336388231e-05,
+      "loss": 0.2068,
+      "step": 6044
+    },
+    {
+      "epoch": 4.550244636808431,
+      "grad_norm": 0.21623139083385468,
+      "learning_rate": 2.165261809005404e-05,
+      "loss": 0.1925,
+      "step": 6045
+    },
+    {
+      "epoch": 4.5509973654497555,
+      "grad_norm": 0.2051752358675003,
+      "learning_rate": 2.1645433455686664e-05,
+      "loss": 0.1487,
+      "step": 6046
+    },
+    {
+      "epoch": 4.55175009409108,
+      "grad_norm": 0.2152714729309082,
+      "learning_rate": 2.1638249461565884e-05,
+      "loss": 0.213,
+      "step": 6047
+    },
+    {
+      "epoch": 4.552502822732405,
+      "grad_norm": 0.2115730494260788,
+      "learning_rate": 2.1631066108477328e-05,
+      "loss": 0.1722,
+      "step": 6048
+    },
+    {
+      "epoch": 4.55325555137373,
+      "grad_norm": 0.19393211603164673,
+      "learning_rate": 2.162388339720655e-05,
+      "loss": 0.2877,
+      "step": 6049
+    },
+    {
+      "epoch": 4.554008280015054,
+      "grad_norm": 0.22504541277885437,
+      "learning_rate": 2.1616701328539057e-05,
+      "loss": 0.266,
+      "step": 6050
+    },
+    {
+      "epoch": 4.55476100865638,
+      "grad_norm": 0.1859370768070221,
+      "learning_rate": 2.160951990326026e-05,
+      "loss": 0.159,
+      "step": 6051
+    },
+    {
+      "epoch": 4.555513737297704,
+      "grad_norm": 0.2162865698337555,
+      "learning_rate": 2.1602339122155512e-05,
+      "loss": 0.1592,
+      "step": 6052
+    },
+    {
+      "epoch": 4.556266465939029,
+      "grad_norm": 0.21819378435611725,
+      "learning_rate": 2.159515898601009e-05,
+      "loss": 0.1424,
+      "step": 6053
+    },
+    {
+      "epoch": 4.557019194580354,
+      "grad_norm": 0.22129791975021362,
+      "learning_rate": 2.1587979495609206e-05,
+      "loss": 0.1868,
+      "step": 6054
+    },
+    {
+      "epoch": 4.5577719232216785,
+      "grad_norm": 0.1941688060760498,
+      "learning_rate": 2.1580800651737994e-05,
+      "loss": 0.1993,
+      "step": 6055
+    },
+    {
+      "epoch": 4.558524651863003,
+      "grad_norm": 0.25217145681381226,
+      "learning_rate": 2.157362245518152e-05,
+      "loss": 0.2454,
+      "step": 6056
+    },
+    {
+      "epoch": 4.559277380504328,
+      "grad_norm": 0.20165367424488068,
+      "learning_rate": 2.156644490672478e-05,
+      "loss": 0.1974,
+      "step": 6057
+    },
+    {
+      "epoch": 4.560030109145653,
+      "grad_norm": 0.24607662856578827,
+      "learning_rate": 2.1559268007152717e-05,
+      "loss": 0.21,
+      "step": 6058
+    },
+    {
+      "epoch": 4.560782837786978,
+      "grad_norm": 0.21407772600650787,
+      "learning_rate": 2.1552091757250164e-05,
+      "loss": 0.2071,
+      "step": 6059
+    },
+    {
+      "epoch": 4.561535566428303,
+      "grad_norm": 0.19100774824619293,
+      "learning_rate": 2.154491615780192e-05,
+      "loss": 0.2065,
+      "step": 6060
+    },
+    {
+      "epoch": 4.562288295069627,
+      "grad_norm": 0.19274815917015076,
+      "learning_rate": 2.1537741209592686e-05,
+      "loss": 0.2386,
+      "step": 6061
+    },
+    {
+      "epoch": 4.563041023710952,
+      "grad_norm": 0.20033487677574158,
+      "learning_rate": 2.1530566913407107e-05,
+      "loss": 0.1911,
+      "step": 6062
+    },
+    {
+      "epoch": 4.563793752352277,
+      "grad_norm": 0.19751892983913422,
+      "learning_rate": 2.1523393270029762e-05,
+      "loss": 0.3266,
+      "step": 6063
+    },
+    {
+      "epoch": 4.564546480993602,
+      "grad_norm": 0.17205218970775604,
+      "learning_rate": 2.151622028024514e-05,
+      "loss": 0.1055,
+      "step": 6064
+    },
+    {
+      "epoch": 4.565299209634927,
+      "grad_norm": 0.21045461297035217,
+      "learning_rate": 2.1509047944837672e-05,
+      "loss": 0.2346,
+      "step": 6065
+    },
+    {
+      "epoch": 4.566051938276251,
+      "grad_norm": 0.21335242688655853,
+      "learning_rate": 2.150187626459172e-05,
+      "loss": 0.1985,
+      "step": 6066
+    },
+    {
+      "epoch": 4.566804666917577,
+      "grad_norm": 0.219119131565094,
+      "learning_rate": 2.1494705240291558e-05,
+      "loss": 0.1477,
+      "step": 6067
+    },
+    {
+      "epoch": 4.567557395558901,
+      "grad_norm": 0.19762489199638367,
+      "learning_rate": 2.1487534872721404e-05,
+      "loss": 0.2101,
+      "step": 6068
+    },
+    {
+      "epoch": 4.5683101242002255,
+      "grad_norm": 0.21504630148410797,
+      "learning_rate": 2.148036516266538e-05,
+      "loss": 0.1959,
+      "step": 6069
+    },
+    {
+      "epoch": 4.569062852841551,
+      "grad_norm": 0.19951185584068298,
+      "learning_rate": 2.1473196110907582e-05,
+      "loss": 0.2518,
+      "step": 6070
+    },
+    {
+      "epoch": 4.569815581482875,
+      "grad_norm": 0.1984003633260727,
+      "learning_rate": 2.1466027718232007e-05,
+      "loss": 0.2115,
+      "step": 6071
+    },
+    {
+      "epoch": 4.5705683101242,
+      "grad_norm": 0.20523416996002197,
+      "learning_rate": 2.145885998542256e-05,
+      "loss": 0.2414,
+      "step": 6072
+    },
+    {
+      "epoch": 4.571321038765525,
+      "grad_norm": 0.21058495342731476,
+      "learning_rate": 2.1451692913263105e-05,
+      "loss": 0.1409,
+      "step": 6073
+    },
+    {
+      "epoch": 4.57207376740685,
+      "grad_norm": 0.1816788762807846,
+      "learning_rate": 2.144452650253741e-05,
+      "loss": 0.2237,
+      "step": 6074
+    },
+    {
+      "epoch": 4.572826496048174,
+      "grad_norm": 0.2158798724412918,
+      "learning_rate": 2.1437360754029197e-05,
+      "loss": 0.2288,
+      "step": 6075
+    },
+    {
+      "epoch": 4.5735792246894995,
+      "grad_norm": 0.22044825553894043,
+      "learning_rate": 2.1430195668522102e-05,
+      "loss": 0.2491,
+      "step": 6076
+    },
+    {
+      "epoch": 4.574331953330824,
+      "grad_norm": 0.20504577457904816,
+      "learning_rate": 2.142303124679968e-05,
+      "loss": 0.2403,
+      "step": 6077
+    },
+    {
+      "epoch": 4.575084681972149,
+      "grad_norm": 0.18421711027622223,
+      "learning_rate": 2.141586748964542e-05,
+      "loss": 0.1613,
+      "step": 6078
+    },
+    {
+      "epoch": 4.575837410613474,
+      "grad_norm": 0.18618954718112946,
+      "learning_rate": 2.140870439784275e-05,
+      "loss": 0.1987,
+      "step": 6079
+    },
+    {
+      "epoch": 4.576590139254798,
+      "grad_norm": 0.21795818209648132,
+      "learning_rate": 2.1401541972175008e-05,
+      "loss": 0.2339,
+      "step": 6080
+    },
+    {
+      "epoch": 4.577342867896124,
+      "grad_norm": 0.21764546632766724,
+      "learning_rate": 2.1394380213425464e-05,
+      "loss": 0.258,
+      "step": 6081
+    },
+    {
+      "epoch": 4.578095596537448,
+      "grad_norm": 0.2141924649477005,
+      "learning_rate": 2.138721912237732e-05,
+      "loss": 0.2324,
+      "step": 6082
+    },
+    {
+      "epoch": 4.5788483251787735,
+      "grad_norm": 0.18989971280097961,
+      "learning_rate": 2.1380058699813705e-05,
+      "loss": 0.1802,
+      "step": 6083
+    },
+    {
+      "epoch": 4.579601053820098,
+      "grad_norm": 0.19518180191516876,
+      "learning_rate": 2.1372898946517673e-05,
+      "loss": 0.2074,
+      "step": 6084
+    },
+    {
+      "epoch": 4.580353782461422,
+      "grad_norm": 0.19328522682189941,
+      "learning_rate": 2.1365739863272193e-05,
+      "loss": 0.1828,
+      "step": 6085
+    },
+    {
+      "epoch": 4.581106511102748,
+      "grad_norm": 0.23378363251686096,
+      "learning_rate": 2.135858145086019e-05,
+      "loss": 0.1902,
+      "step": 6086
+    },
+    {
+      "epoch": 4.581859239744072,
+      "grad_norm": 0.17890989780426025,
+      "learning_rate": 2.1351423710064477e-05,
+      "loss": 0.2746,
+      "step": 6087
+    },
+    {
+      "epoch": 4.582611968385397,
+      "grad_norm": 0.21901629865169525,
+      "learning_rate": 2.1344266641667816e-05,
+      "loss": 0.1884,
+      "step": 6088
+    },
+    {
+      "epoch": 4.583364697026722,
+      "grad_norm": 0.1860281080007553,
+      "learning_rate": 2.1337110246452913e-05,
+      "loss": 0.211,
+      "step": 6089
+    },
+    {
+      "epoch": 4.584117425668047,
+      "grad_norm": 0.21383920311927795,
+      "learning_rate": 2.132995452520236e-05,
+      "loss": 0.1972,
+      "step": 6090
+    },
+    {
+      "epoch": 4.584870154309371,
+      "grad_norm": 0.18785502016544342,
+      "learning_rate": 2.1322799478698708e-05,
+      "loss": 0.2903,
+      "step": 6091
+    },
+    {
+      "epoch": 4.585622882950696,
+      "grad_norm": 0.227357417345047,
+      "learning_rate": 2.131564510772442e-05,
+      "loss": 0.1435,
+      "step": 6092
+    },
+    {
+      "epoch": 4.586375611592021,
+      "grad_norm": 0.1986430436372757,
+      "learning_rate": 2.130849141306188e-05,
+      "loss": 0.1729,
+      "step": 6093
+    },
+    {
+      "epoch": 4.587128340233346,
+      "grad_norm": 0.192689448595047,
+      "learning_rate": 2.1301338395493404e-05,
+      "loss": 0.1703,
+      "step": 6094
+    },
+    {
+      "epoch": 4.587881068874671,
+      "grad_norm": 0.2079346776008606,
+      "learning_rate": 2.1294186055801252e-05,
+      "loss": 0.2574,
+      "step": 6095
+    },
+    {
+      "epoch": 4.588633797515995,
+      "grad_norm": 0.2037295699119568,
+      "learning_rate": 2.128703439476758e-05,
+      "loss": 0.2019,
+      "step": 6096
+    },
+    {
+      "epoch": 4.589386526157321,
+      "grad_norm": 0.21359506249427795,
+      "learning_rate": 2.127988341317449e-05,
+      "loss": 0.2614,
+      "step": 6097
+    },
+    {
+      "epoch": 4.590139254798645,
+      "grad_norm": 0.23332251608371735,
+      "learning_rate": 2.1272733111803993e-05,
+      "loss": 0.2112,
+      "step": 6098
+    },
+    {
+      "epoch": 4.5908919834399695,
+      "grad_norm": 0.17072345316410065,
+      "learning_rate": 2.1265583491438035e-05,
+      "loss": 0.1728,
+      "step": 6099
+    },
+    {
+      "epoch": 4.591644712081295,
+      "grad_norm": 0.20668189227581024,
+      "learning_rate": 2.1258434552858502e-05,
+      "loss": 0.184,
+      "step": 6100
+    },
+    {
+      "epoch": 4.592397440722619,
+      "grad_norm": 0.21163715422153473,
+      "learning_rate": 2.125128629684717e-05,
+      "loss": 0.213,
+      "step": 6101
+    },
+    {
+      "epoch": 4.593150169363945,
+      "grad_norm": 0.21703241765499115,
+      "learning_rate": 2.1244138724185786e-05,
+      "loss": 0.2549,
+      "step": 6102
+    },
+    {
+      "epoch": 4.593902898005269,
+      "grad_norm": 0.21929682791233063,
+      "learning_rate": 2.1236991835655982e-05,
+      "loss": 0.2735,
+      "step": 6103
+    },
+    {
+      "epoch": 4.594655626646594,
+      "grad_norm": 0.1914924681186676,
+      "learning_rate": 2.122984563203934e-05,
+      "loss": 0.2224,
+      "step": 6104
+    },
+    {
+      "epoch": 4.595408355287919,
+      "grad_norm": 0.23531293869018555,
+      "learning_rate": 2.1222700114117343e-05,
+      "loss": 0.2551,
+      "step": 6105
+    },
+    {
+      "epoch": 4.5961610839292435,
+      "grad_norm": 0.22619862854480743,
+      "learning_rate": 2.1215555282671433e-05,
+      "loss": 0.273,
+      "step": 6106
+    },
+    {
+      "epoch": 4.596913812570568,
+      "grad_norm": 0.22178539633750916,
+      "learning_rate": 2.1208411138482936e-05,
+      "loss": 0.2054,
+      "step": 6107
+    },
+    {
+      "epoch": 4.597666541211893,
+      "grad_norm": 0.21013052761554718,
+      "learning_rate": 2.1201267682333147e-05,
+      "loss": 0.212,
+      "step": 6108
+    },
+    {
+      "epoch": 4.598419269853218,
+      "grad_norm": 0.19411849975585938,
+      "learning_rate": 2.1194124915003255e-05,
+      "loss": 0.2211,
+      "step": 6109
+    },
+    {
+      "epoch": 4.599171998494542,
+      "grad_norm": 0.2014913111925125,
+      "learning_rate": 2.1186982837274382e-05,
+      "loss": 0.147,
+      "step": 6110
+    },
+    {
+      "epoch": 4.599924727135868,
+      "grad_norm": 0.2345956414937973,
+      "learning_rate": 2.1179841449927573e-05,
+      "loss": 0.2684,
+      "step": 6111
+    },
+    {
+      "epoch": 4.600677455777192,
+      "grad_norm": 0.21378101408481598,
+      "learning_rate": 2.1172700753743803e-05,
+      "loss": 0.2433,
+      "step": 6112
+    },
+    {
+      "epoch": 4.6014301844185175,
+      "grad_norm": 0.22796876728534698,
+      "learning_rate": 2.1165560749503954e-05,
+      "loss": 0.2046,
+      "step": 6113
+    },
+    {
+      "epoch": 4.602182913059842,
+      "grad_norm": 0.19358544051647186,
+      "learning_rate": 2.1158421437988863e-05,
+      "loss": 0.1689,
+      "step": 6114
+    },
+    {
+      "epoch": 4.602935641701166,
+      "grad_norm": 0.20881634950637817,
+      "learning_rate": 2.1151282819979273e-05,
+      "loss": 0.1809,
+      "step": 6115
+    },
+    {
+      "epoch": 4.603688370342492,
+      "grad_norm": 0.24015715718269348,
+      "learning_rate": 2.114414489625585e-05,
+      "loss": 0.193,
+      "step": 6116
+    },
+    {
+      "epoch": 4.604441098983816,
+      "grad_norm": 0.20933814346790314,
+      "learning_rate": 2.1137007667599185e-05,
+      "loss": 0.1483,
+      "step": 6117
+    },
+    {
+      "epoch": 4.605193827625142,
+      "grad_norm": 0.19272936880588531,
+      "learning_rate": 2.1129871134789796e-05,
+      "loss": 0.125,
+      "step": 6118
+    },
+    {
+      "epoch": 4.605946556266466,
+      "grad_norm": 0.22553600370883942,
+      "learning_rate": 2.112273529860812e-05,
+      "loss": 0.1606,
+      "step": 6119
+    },
+    {
+      "epoch": 4.606699284907791,
+      "grad_norm": 0.24040977656841278,
+      "learning_rate": 2.1115600159834516e-05,
+      "loss": 0.2224,
+      "step": 6120
+    },
+    {
+      "epoch": 4.607452013549116,
+      "grad_norm": 0.2277776151895523,
+      "learning_rate": 2.1108465719249286e-05,
+      "loss": 0.1428,
+      "step": 6121
+    },
+    {
+      "epoch": 4.60820474219044,
+      "grad_norm": 0.2137908786535263,
+      "learning_rate": 2.1101331977632643e-05,
+      "loss": 0.2613,
+      "step": 6122
+    },
+    {
+      "epoch": 4.608957470831765,
+      "grad_norm": 0.22347311675548553,
+      "learning_rate": 2.1094198935764704e-05,
+      "loss": 0.1593,
+      "step": 6123
+    },
+    {
+      "epoch": 4.60971019947309,
+      "grad_norm": 0.19870440661907196,
+      "learning_rate": 2.108706659442554e-05,
+      "loss": 0.1964,
+      "step": 6124
+    },
+    {
+      "epoch": 4.610462928114415,
+      "grad_norm": 0.20303292572498322,
+      "learning_rate": 2.1079934954395137e-05,
+      "loss": 0.1586,
+      "step": 6125
+    },
+    {
+      "epoch": 4.611215656755739,
+      "grad_norm": 0.24607688188552856,
+      "learning_rate": 2.1072804016453385e-05,
+      "loss": 0.3083,
+      "step": 6126
+    },
+    {
+      "epoch": 4.6119683853970646,
+      "grad_norm": 0.2049359828233719,
+      "learning_rate": 2.1065673781380135e-05,
+      "loss": 0.226,
+      "step": 6127
+    },
+    {
+      "epoch": 4.612721114038389,
+      "grad_norm": 0.21754039824008942,
+      "learning_rate": 2.1058544249955115e-05,
+      "loss": 0.2001,
+      "step": 6128
+    },
+    {
+      "epoch": 4.6134738426797135,
+      "grad_norm": 0.20215091109275818,
+      "learning_rate": 2.1051415422958015e-05,
+      "loss": 0.1695,
+      "step": 6129
+    },
+    {
+      "epoch": 4.614226571321039,
+      "grad_norm": 0.2305067479610443,
+      "learning_rate": 2.1044287301168435e-05,
+      "loss": 0.1671,
+      "step": 6130
+    },
+    {
+      "epoch": 4.614979299962363,
+      "grad_norm": 0.203969344496727,
+      "learning_rate": 2.1037159885365888e-05,
+      "loss": 0.1433,
+      "step": 6131
+    },
+    {
+      "epoch": 4.615732028603689,
+      "grad_norm": 0.19940364360809326,
+      "learning_rate": 2.103003317632981e-05,
+      "loss": 0.1757,
+      "step": 6132
+    },
+    {
+      "epoch": 4.616484757245013,
+      "grad_norm": 0.23229335248470306,
+      "learning_rate": 2.1022907174839585e-05,
+      "loss": 0.1649,
+      "step": 6133
+    },
+    {
+      "epoch": 4.617237485886338,
+      "grad_norm": 0.21602126955986023,
+      "learning_rate": 2.101578188167449e-05,
+      "loss": 0.1766,
+      "step": 6134
+    },
+    {
+      "epoch": 4.617990214527663,
+      "grad_norm": 0.2185862809419632,
+      "learning_rate": 2.1008657297613744e-05,
+      "loss": 0.1785,
+      "step": 6135
+    },
+    {
+      "epoch": 4.6187429431689875,
+      "grad_norm": 0.2288241684436798,
+      "learning_rate": 2.1001533423436475e-05,
+      "loss": 0.1731,
+      "step": 6136
+    },
+    {
+      "epoch": 4.619495671810313,
+      "grad_norm": 0.2109624147415161,
+      "learning_rate": 2.0994410259921744e-05,
+      "loss": 0.0924,
+      "step": 6137
+    },
+    {
+      "epoch": 4.620248400451637,
+      "grad_norm": 0.22574475407600403,
+      "learning_rate": 2.0987287807848515e-05,
+      "loss": 0.1739,
+      "step": 6138
+    },
+    {
+      "epoch": 4.621001129092962,
+      "grad_norm": 0.23447218537330627,
+      "learning_rate": 2.09801660679957e-05,
+      "loss": 0.2268,
+      "step": 6139
+    },
+    {
+      "epoch": 4.621753857734287,
+      "grad_norm": 0.21039719879627228,
+      "learning_rate": 2.0973045041142132e-05,
+      "loss": 0.2054,
+      "step": 6140
+    },
+    {
+      "epoch": 4.622506586375612,
+      "grad_norm": 0.23054151237010956,
+      "learning_rate": 2.0965924728066537e-05,
+      "loss": 0.1911,
+      "step": 6141
+    },
+    {
+      "epoch": 4.623259315016936,
+      "grad_norm": 0.20791983604431152,
+      "learning_rate": 2.0958805129547598e-05,
+      "loss": 0.1342,
+      "step": 6142
+    },
+    {
+      "epoch": 4.6240120436582615,
+      "grad_norm": 0.21356280148029327,
+      "learning_rate": 2.095168624636389e-05,
+      "loss": 0.186,
+      "step": 6143
+    },
+    {
+      "epoch": 4.624764772299586,
+      "grad_norm": 0.24296419322490692,
+      "learning_rate": 2.094456807929393e-05,
+      "loss": 0.219,
+      "step": 6144
+    },
+    {
+      "epoch": 4.62551750094091,
+      "grad_norm": 0.21842262148857117,
+      "learning_rate": 2.093745062911614e-05,
+      "loss": 0.2166,
+      "step": 6145
+    },
+    {
+      "epoch": 4.626270229582236,
+      "grad_norm": 0.22337359189987183,
+      "learning_rate": 2.0930333896608893e-05,
+      "loss": 0.2198,
+      "step": 6146
+    },
+    {
+      "epoch": 4.62702295822356,
+      "grad_norm": 0.24409119784832,
+      "learning_rate": 2.092321788255045e-05,
+      "loss": 0.2669,
+      "step": 6147
+    },
+    {
+      "epoch": 4.627775686864885,
+      "grad_norm": 0.21428367495536804,
+      "learning_rate": 2.0916102587719018e-05,
+      "loss": 0.1481,
+      "step": 6148
+    },
+    {
+      "epoch": 4.62852841550621,
+      "grad_norm": 0.20398883521556854,
+      "learning_rate": 2.09089880128927e-05,
+      "loss": 0.2505,
+      "step": 6149
+    },
+    {
+      "epoch": 4.6292811441475346,
+      "grad_norm": 0.1942271739244461,
+      "learning_rate": 2.090187415884955e-05,
+      "loss": 0.2249,
+      "step": 6150
+    },
+    {
+      "epoch": 4.63003387278886,
+      "grad_norm": 0.20544472336769104,
+      "learning_rate": 2.0894761026367514e-05,
+      "loss": 0.1443,
+      "step": 6151
+    },
+    {
+      "epoch": 4.630786601430184,
+      "grad_norm": 0.1662530153989792,
+      "learning_rate": 2.0887648616224486e-05,
+      "loss": 0.1079,
+      "step": 6152
+    },
+    {
+      "epoch": 4.631539330071509,
+      "grad_norm": 0.2223552018404007,
+      "learning_rate": 2.0880536929198263e-05,
+      "loss": 0.2433,
+      "step": 6153
+    },
+    {
+      "epoch": 4.632292058712834,
+      "grad_norm": 0.19650891423225403,
+      "learning_rate": 2.0873425966066566e-05,
+      "loss": 0.2042,
+      "step": 6154
+    },
+    {
+      "epoch": 4.633044787354159,
+      "grad_norm": 0.2044554352760315,
+      "learning_rate": 2.086631572760705e-05,
+      "loss": 0.155,
+      "step": 6155
+    },
+    {
+      "epoch": 4.633797515995484,
+      "grad_norm": 0.23120267689228058,
+      "learning_rate": 2.0859206214597273e-05,
+      "loss": 0.2949,
+      "step": 6156
+    },
+    {
+      "epoch": 4.6345502446368085,
+      "grad_norm": 0.19281834363937378,
+      "learning_rate": 2.0852097427814717e-05,
+      "loss": 0.2103,
+      "step": 6157
+    },
+    {
+      "epoch": 4.635302973278133,
+      "grad_norm": 0.19788607954978943,
+      "learning_rate": 2.0844989368036776e-05,
+      "loss": 0.1477,
+      "step": 6158
+    },
+    {
+      "epoch": 4.636055701919458,
+      "grad_norm": 0.2203785926103592,
+      "learning_rate": 2.083788203604081e-05,
+      "loss": 0.2168,
+      "step": 6159
+    },
+    {
+      "epoch": 4.636808430560783,
+      "grad_norm": 0.1995445042848587,
+      "learning_rate": 2.0830775432604045e-05,
+      "loss": 0.2219,
+      "step": 6160
+    },
+    {
+      "epoch": 4.637561159202107,
+      "grad_norm": 0.2165738046169281,
+      "learning_rate": 2.082366955850365e-05,
+      "loss": 0.2644,
+      "step": 6161
+    },
+    {
+      "epoch": 4.638313887843433,
+      "grad_norm": 0.21867994964122772,
+      "learning_rate": 2.081656441451672e-05,
+      "loss": 0.2429,
+      "step": 6162
+    },
+    {
+      "epoch": 4.639066616484757,
+      "grad_norm": 0.2149602770805359,
+      "learning_rate": 2.080946000142025e-05,
+      "loss": 0.2013,
+      "step": 6163
+    },
+    {
+      "epoch": 4.639819345126082,
+      "grad_norm": 0.21919794380664825,
+      "learning_rate": 2.0802356319991174e-05,
+      "loss": 0.1841,
+      "step": 6164
+    },
+    {
+      "epoch": 4.640572073767407,
+      "grad_norm": 0.22419264912605286,
+      "learning_rate": 2.0795253371006345e-05,
+      "loss": 0.2321,
+      "step": 6165
+    },
+    {
+      "epoch": 4.6413248024087315,
+      "grad_norm": 0.21155957877635956,
+      "learning_rate": 2.078815115524253e-05,
+      "loss": 0.2628,
+      "step": 6166
+    },
+    {
+      "epoch": 4.642077531050057,
+      "grad_norm": 0.2288271188735962,
+      "learning_rate": 2.0781049673476412e-05,
+      "loss": 0.2564,
+      "step": 6167
+    },
+    {
+      "epoch": 4.642830259691381,
+      "grad_norm": 0.1822563111782074,
+      "learning_rate": 2.07739489264846e-05,
+      "loss": 0.2206,
+      "step": 6168
+    },
+    {
+      "epoch": 4.643582988332706,
+      "grad_norm": 0.24044416844844818,
+      "learning_rate": 2.076684891504362e-05,
+      "loss": 0.1632,
+      "step": 6169
+    },
+    {
+      "epoch": 4.644335716974031,
+      "grad_norm": 0.2415785938501358,
+      "learning_rate": 2.0759749639929917e-05,
+      "loss": 0.2083,
+      "step": 6170
+    },
+    {
+      "epoch": 4.645088445615356,
+      "grad_norm": 0.24320095777511597,
+      "learning_rate": 2.075265110191985e-05,
+      "loss": 0.2473,
+      "step": 6171
+    },
+    {
+      "epoch": 4.64584117425668,
+      "grad_norm": 0.1877070963382721,
+      "learning_rate": 2.0745553301789733e-05,
+      "loss": 0.1833,
+      "step": 6172
+    },
+    {
+      "epoch": 4.6465939028980054,
+      "grad_norm": 0.24162320792675018,
+      "learning_rate": 2.0738456240315747e-05,
+      "loss": 0.1992,
+      "step": 6173
+    },
+    {
+      "epoch": 4.64734663153933,
+      "grad_norm": 0.19288446009159088,
+      "learning_rate": 2.0731359918274013e-05,
+      "loss": 0.1443,
+      "step": 6174
+    },
+    {
+      "epoch": 4.648099360180655,
+      "grad_norm": 0.18176555633544922,
+      "learning_rate": 2.0724264336440584e-05,
+      "loss": 0.1287,
+      "step": 6175
+    },
+    {
+      "epoch": 4.64885208882198,
+      "grad_norm": 0.2387997806072235,
+      "learning_rate": 2.0717169495591426e-05,
+      "loss": 0.2089,
+      "step": 6176
+    },
+    {
+      "epoch": 4.649604817463304,
+      "grad_norm": 0.2960723638534546,
+      "learning_rate": 2.071007539650241e-05,
+      "loss": 0.2565,
+      "step": 6177
+    },
+    {
+      "epoch": 4.65035754610463,
+      "grad_norm": 0.2360212802886963,
+      "learning_rate": 2.0702982039949335e-05,
+      "loss": 0.257,
+      "step": 6178
+    },
+    {
+      "epoch": 4.651110274745954,
+      "grad_norm": 0.20497757196426392,
+      "learning_rate": 2.069588942670793e-05,
+      "loss": 0.166,
+      "step": 6179
+    },
+    {
+      "epoch": 4.6518630033872785,
+      "grad_norm": 0.23437169194221497,
+      "learning_rate": 2.068879755755383e-05,
+      "loss": 0.2143,
+      "step": 6180
+    },
+    {
+      "epoch": 4.652615732028604,
+      "grad_norm": 0.2536351978778839,
+      "learning_rate": 2.0681706433262593e-05,
+      "loss": 0.2632,
+      "step": 6181
+    },
+    {
+      "epoch": 4.653368460669928,
+      "grad_norm": 0.23375961184501648,
+      "learning_rate": 2.0674616054609685e-05,
+      "loss": 0.2076,
+      "step": 6182
+    },
+    {
+      "epoch": 4.654121189311253,
+      "grad_norm": 0.23887020349502563,
+      "learning_rate": 2.066752642237049e-05,
+      "loss": 0.2208,
+      "step": 6183
+    },
+    {
+      "epoch": 4.654873917952578,
+      "grad_norm": 0.1764029860496521,
+      "learning_rate": 2.066043753732034e-05,
+      "loss": 0.1761,
+      "step": 6184
+    },
+    {
+      "epoch": 4.655626646593903,
+      "grad_norm": 0.2180500477552414,
+      "learning_rate": 2.0653349400234467e-05,
+      "loss": 0.1313,
+      "step": 6185
+    },
+    {
+      "epoch": 4.656379375235228,
+      "grad_norm": 0.21582752466201782,
+      "learning_rate": 2.064626201188801e-05,
+      "loss": 0.2061,
+      "step": 6186
+    },
+    {
+      "epoch": 4.6571321038765525,
+      "grad_norm": 0.18688149750232697,
+      "learning_rate": 2.063917537305603e-05,
+      "loss": 0.1468,
+      "step": 6187
+    },
+    {
+      "epoch": 4.657884832517877,
+      "grad_norm": 0.206973135471344,
+      "learning_rate": 2.0632089484513517e-05,
+      "loss": 0.1438,
+      "step": 6188
+    },
+    {
+      "epoch": 4.658637561159202,
+      "grad_norm": 0.2280263900756836,
+      "learning_rate": 2.062500434703537e-05,
+      "loss": 0.2281,
+      "step": 6189
+    },
+    {
+      "epoch": 4.659390289800527,
+      "grad_norm": 0.24269115924835205,
+      "learning_rate": 2.061791996139641e-05,
+      "loss": 0.2592,
+      "step": 6190
+    },
+    {
+      "epoch": 4.660143018441852,
+      "grad_norm": 0.2007661759853363,
+      "learning_rate": 2.061083632837138e-05,
+      "loss": 0.2416,
+      "step": 6191
+    },
+    {
+      "epoch": 4.660895747083177,
+      "grad_norm": 0.21249715983867645,
+      "learning_rate": 2.060375344873493e-05,
+      "loss": 0.2276,
+      "step": 6192
+    },
+    {
+      "epoch": 4.661648475724501,
+      "grad_norm": 0.21260793507099152,
+      "learning_rate": 2.0596671323261633e-05,
+      "loss": 0.2003,
+      "step": 6193
+    },
+    {
+      "epoch": 4.6624012043658265,
+      "grad_norm": 0.22672009468078613,
+      "learning_rate": 2.0589589952725986e-05,
+      "loss": 0.2002,
+      "step": 6194
+    },
+    {
+      "epoch": 4.663153933007151,
+      "grad_norm": 0.21304090321063995,
+      "learning_rate": 2.0582509337902384e-05,
+      "loss": 0.202,
+      "step": 6195
+    },
+    {
+      "epoch": 4.6639066616484754,
+      "grad_norm": 0.1997358798980713,
+      "learning_rate": 2.0575429479565168e-05,
+      "loss": 0.3247,
+      "step": 6196
+    },
+    {
+      "epoch": 4.664659390289801,
+      "grad_norm": 0.23548829555511475,
+      "learning_rate": 2.0568350378488572e-05,
+      "loss": 0.1555,
+      "step": 6197
+    },
+    {
+      "epoch": 4.665412118931125,
+      "grad_norm": 0.21644039452075958,
+      "learning_rate": 2.0561272035446756e-05,
+      "loss": 0.1802,
+      "step": 6198
+    },
+    {
+      "epoch": 4.66616484757245,
+      "grad_norm": 0.20642274618148804,
+      "learning_rate": 2.0554194451213794e-05,
+      "loss": 0.2426,
+      "step": 6199
+    },
+    {
+      "epoch": 4.666917576213775,
+      "grad_norm": 0.2157023847103119,
+      "learning_rate": 2.054711762656369e-05,
+      "loss": 0.1981,
+      "step": 6200
+    },
+    {
+      "epoch": 4.666917576213775,
+      "eval_loss": 0.22830942273139954,
+      "eval_runtime": 452.7161,
+      "eval_samples_per_second": 21.265,
+      "eval_steps_per_second": 0.665,
+      "step": 6200
+    },
+    {
+      "epoch": 4.6676703048551,
+      "grad_norm": 0.1817314773797989,
+      "learning_rate": 2.0540041562270345e-05,
+      "loss": 0.1703,
+      "step": 6201
+    },
+    {
+      "epoch": 4.668423033496424,
+      "grad_norm": 0.18745321035385132,
+      "learning_rate": 2.053296625910759e-05,
+      "loss": 0.1866,
+      "step": 6202
+    },
+    {
+      "epoch": 4.669175762137749,
+      "grad_norm": 0.18946295976638794,
+      "learning_rate": 2.0525891717849166e-05,
+      "loss": 0.253,
+      "step": 6203
+    },
+    {
+      "epoch": 4.669928490779074,
+      "grad_norm": 0.19527196884155273,
+      "learning_rate": 2.0518817939268747e-05,
+      "loss": 0.2131,
+      "step": 6204
+    },
+    {
+      "epoch": 4.670681219420399,
+      "grad_norm": 0.20256538689136505,
+      "learning_rate": 2.0511744924139904e-05,
+      "loss": 0.2102,
+      "step": 6205
+    },
+    {
+      "epoch": 4.671433948061724,
+      "grad_norm": 0.18136745691299438,
+      "learning_rate": 2.050467267323612e-05,
+      "loss": 0.1723,
+      "step": 6206
+    },
+    {
+      "epoch": 4.672186676703048,
+      "grad_norm": 0.20673036575317383,
+      "learning_rate": 2.0497601187330824e-05,
+      "loss": 0.253,
+      "step": 6207
+    },
+    {
+      "epoch": 4.672939405344374,
+      "grad_norm": 0.22298765182495117,
+      "learning_rate": 2.0490530467197328e-05,
+      "loss": 0.1415,
+      "step": 6208
+    },
+    {
+      "epoch": 4.673692133985698,
+      "grad_norm": 0.2060922384262085,
+      "learning_rate": 2.0483460513608875e-05,
+      "loss": 0.1123,
+      "step": 6209
+    },
+    {
+      "epoch": 4.674444862627023,
+      "grad_norm": 0.20941658318042755,
+      "learning_rate": 2.0476391327338645e-05,
+      "loss": 0.2243,
+      "step": 6210
+    },
+    {
+      "epoch": 4.675197591268348,
+      "grad_norm": 0.18890735507011414,
+      "learning_rate": 2.0469322909159694e-05,
+      "loss": 0.1977,
+      "step": 6211
+    },
+    {
+      "epoch": 4.675950319909672,
+      "grad_norm": 0.21349135041236877,
+      "learning_rate": 2.0462255259845025e-05,
+      "loss": 0.1692,
+      "step": 6212
+    },
+    {
+      "epoch": 4.676703048550998,
+      "grad_norm": 0.20083454251289368,
+      "learning_rate": 2.0455188380167534e-05,
+      "loss": 0.2406,
+      "step": 6213
+    },
+    {
+      "epoch": 4.677455777192322,
+      "grad_norm": 0.20125456154346466,
+      "learning_rate": 2.044812227090005e-05,
+      "loss": 0.1501,
+      "step": 6214
+    },
+    {
+      "epoch": 4.678208505833647,
+      "grad_norm": 0.18809199333190918,
+      "learning_rate": 2.0441056932815307e-05,
+      "loss": 0.2069,
+      "step": 6215
+    },
+    {
+      "epoch": 4.678961234474972,
+      "grad_norm": 0.21870358288288116,
+      "learning_rate": 2.0433992366685966e-05,
+      "loss": 0.1899,
+      "step": 6216
+    },
+    {
+      "epoch": 4.6797139631162965,
+      "grad_norm": 0.2381441444158554,
+      "learning_rate": 2.0426928573284603e-05,
+      "loss": 0.2184,
+      "step": 6217
+    },
+    {
+      "epoch": 4.680466691757621,
+      "grad_norm": 0.20698240399360657,
+      "learning_rate": 2.04198655533837e-05,
+      "loss": 0.2511,
+      "step": 6218
+    },
+    {
+      "epoch": 4.681219420398946,
+      "grad_norm": 0.19271129369735718,
+      "learning_rate": 2.0412803307755652e-05,
+      "loss": 0.1357,
+      "step": 6219
+    },
+    {
+      "epoch": 4.681972149040271,
+      "grad_norm": 0.1997849941253662,
+      "learning_rate": 2.040574183717277e-05,
+      "loss": 0.2311,
+      "step": 6220
+    },
+    {
+      "epoch": 4.682724877681595,
+      "grad_norm": 0.17277279496192932,
+      "learning_rate": 2.03986811424073e-05,
+      "loss": 0.1548,
+      "step": 6221
+    },
+    {
+      "epoch": 4.683477606322921,
+      "grad_norm": 0.20365314185619354,
+      "learning_rate": 2.0391621224231384e-05,
+      "loss": 0.2426,
+      "step": 6222
+    },
+    {
+      "epoch": 4.684230334964245,
+      "grad_norm": 0.18180964887142181,
+      "learning_rate": 2.0384562083417086e-05,
+      "loss": 0.1883,
+      "step": 6223
+    },
+    {
+      "epoch": 4.6849830636055705,
+      "grad_norm": 0.195704385638237,
+      "learning_rate": 2.0377503720736375e-05,
+      "loss": 0.2479,
+      "step": 6224
+    },
+    {
+      "epoch": 4.685735792246895,
+      "grad_norm": 0.18760356307029724,
+      "learning_rate": 2.037044613696115e-05,
+      "loss": 0.2259,
+      "step": 6225
+    },
+    {
+      "epoch": 4.686488520888219,
+      "grad_norm": 0.19874121248722076,
+      "learning_rate": 2.0363389332863225e-05,
+      "loss": 0.1677,
+      "step": 6226
+    },
+    {
+      "epoch": 4.687241249529545,
+      "grad_norm": 0.2013537883758545,
+      "learning_rate": 2.0356333309214304e-05,
+      "loss": 0.2071,
+      "step": 6227
+    },
+    {
+      "epoch": 4.687993978170869,
+      "grad_norm": 0.17832672595977783,
+      "learning_rate": 2.0349278066786032e-05,
+      "loss": 0.1884,
+      "step": 6228
+    },
+    {
+      "epoch": 4.688746706812195,
+      "grad_norm": 0.18564818799495697,
+      "learning_rate": 2.0342223606349964e-05,
+      "loss": 0.1239,
+      "step": 6229
+    },
+    {
+      "epoch": 4.689499435453519,
+      "grad_norm": 0.18551842868328094,
+      "learning_rate": 2.0335169928677562e-05,
+      "loss": 0.1508,
+      "step": 6230
+    },
+    {
+      "epoch": 4.690252164094844,
+      "grad_norm": 0.17588269710540771,
+      "learning_rate": 2.03281170345402e-05,
+      "loss": 0.1489,
+      "step": 6231
+    },
+    {
+      "epoch": 4.691004892736169,
+      "grad_norm": 0.21282395720481873,
+      "learning_rate": 2.0321064924709178e-05,
+      "loss": 0.2524,
+      "step": 6232
+    },
+    {
+      "epoch": 4.691757621377493,
+      "grad_norm": 0.20604534447193146,
+      "learning_rate": 2.0314013599955704e-05,
+      "loss": 0.197,
+      "step": 6233
+    },
+    {
+      "epoch": 4.692510350018818,
+      "grad_norm": 0.22980262339115143,
+      "learning_rate": 2.0306963061050892e-05,
+      "loss": 0.1582,
+      "step": 6234
+    },
+    {
+      "epoch": 4.693263078660143,
+      "grad_norm": 0.23909801244735718,
+      "learning_rate": 2.029991330876579e-05,
+      "loss": 0.1448,
+      "step": 6235
+    },
+    {
+      "epoch": 4.694015807301468,
+      "grad_norm": 0.2132544368505478,
+      "learning_rate": 2.0292864343871343e-05,
+      "loss": 0.2755,
+      "step": 6236
+    },
+    {
+      "epoch": 4.694768535942792,
+      "grad_norm": 0.22749212384223938,
+      "learning_rate": 2.028581616713842e-05,
+      "loss": 0.2058,
+      "step": 6237
+    },
+    {
+      "epoch": 4.695521264584118,
+      "grad_norm": 0.18576762080192566,
+      "learning_rate": 2.0278768779337792e-05,
+      "loss": 0.1824,
+      "step": 6238
+    },
+    {
+      "epoch": 4.696273993225442,
+      "grad_norm": 0.23119549453258514,
+      "learning_rate": 2.0271722181240156e-05,
+      "loss": 0.2197,
+      "step": 6239
+    },
+    {
+      "epoch": 4.697026721866767,
+      "grad_norm": 0.17707611620426178,
+      "learning_rate": 2.026467637361611e-05,
+      "loss": 0.1713,
+      "step": 6240
+    },
+    {
+      "epoch": 4.697779450508092,
+      "grad_norm": 0.22583214938640594,
+      "learning_rate": 2.0257631357236168e-05,
+      "loss": 0.1968,
+      "step": 6241
+    },
+    {
+      "epoch": 4.698532179149416,
+      "grad_norm": 0.19024279713630676,
+      "learning_rate": 2.0250587132870785e-05,
+      "loss": 0.1896,
+      "step": 6242
+    },
+    {
+      "epoch": 4.699284907790742,
+      "grad_norm": 0.22586597502231598,
+      "learning_rate": 2.0243543701290295e-05,
+      "loss": 0.2204,
+      "step": 6243
+    },
+    {
+      "epoch": 4.700037636432066,
+      "grad_norm": 0.18101699650287628,
+      "learning_rate": 2.023650106326495e-05,
+      "loss": 0.2499,
+      "step": 6244
+    },
+    {
+      "epoch": 4.7007903650733915,
+      "grad_norm": 0.2285294085741043,
+      "learning_rate": 2.0229459219564927e-05,
+      "loss": 0.1651,
+      "step": 6245
+    },
+    {
+      "epoch": 4.701543093714716,
+      "grad_norm": 0.18303729593753815,
+      "learning_rate": 2.0222418170960323e-05,
+      "loss": 0.1719,
+      "step": 6246
+    },
+    {
+      "epoch": 4.7022958223560405,
+      "grad_norm": 0.1856185644865036,
+      "learning_rate": 2.0215377918221116e-05,
+      "loss": 0.2044,
+      "step": 6247
+    },
+    {
+      "epoch": 4.703048550997366,
+      "grad_norm": 0.22072508931159973,
+      "learning_rate": 2.0208338462117234e-05,
+      "loss": 0.1385,
+      "step": 6248
+    },
+    {
+      "epoch": 4.70380127963869,
+      "grad_norm": 0.21946199238300323,
+      "learning_rate": 2.0201299803418482e-05,
+      "loss": 0.2003,
+      "step": 6249
+    },
+    {
+      "epoch": 4.704554008280015,
+      "grad_norm": 0.2111605703830719,
+      "learning_rate": 2.0194261942894628e-05,
+      "loss": 0.1745,
+      "step": 6250
+    },
+    {
+      "epoch": 4.70530673692134,
+      "grad_norm": 0.21167251467704773,
+      "learning_rate": 2.0187224881315292e-05,
+      "loss": 0.1679,
+      "step": 6251
+    },
+    {
+      "epoch": 4.706059465562665,
+      "grad_norm": 0.21721500158309937,
+      "learning_rate": 2.0180188619450052e-05,
+      "loss": 0.1628,
+      "step": 6252
+    },
+    {
+      "epoch": 4.706812194203989,
+      "grad_norm": 0.21123431622982025,
+      "learning_rate": 2.0173153158068376e-05,
+      "loss": 0.2461,
+      "step": 6253
+    },
+    {
+      "epoch": 4.7075649228453145,
+      "grad_norm": 0.19963864982128143,
+      "learning_rate": 2.0166118497939657e-05,
+      "loss": 0.2002,
+      "step": 6254
+    },
+    {
+      "epoch": 4.708317651486639,
+      "grad_norm": 0.16507036983966827,
+      "learning_rate": 2.01590846398332e-05,
+      "loss": 0.0949,
+      "step": 6255
+    },
+    {
+      "epoch": 4.709070380127963,
+      "grad_norm": 0.21352390944957733,
+      "learning_rate": 2.01520515845182e-05,
+      "loss": 0.2104,
+      "step": 6256
+    },
+    {
+      "epoch": 4.709823108769289,
+      "grad_norm": 0.19839628040790558,
+      "learning_rate": 2.0145019332763798e-05,
+      "loss": 0.1999,
+      "step": 6257
+    },
+    {
+      "epoch": 4.710575837410613,
+      "grad_norm": 0.19967679679393768,
+      "learning_rate": 2.013798788533902e-05,
+      "loss": 0.2156,
+      "step": 6258
+    },
+    {
+      "epoch": 4.711328566051939,
+      "grad_norm": 0.19669774174690247,
+      "learning_rate": 2.013095724301282e-05,
+      "loss": 0.1599,
+      "step": 6259
+    },
+    {
+      "epoch": 4.712081294693263,
+      "grad_norm": 0.19361869990825653,
+      "learning_rate": 2.012392740655405e-05,
+      "loss": 0.1663,
+      "step": 6260
+    },
+    {
+      "epoch": 4.712834023334588,
+      "grad_norm": 0.20385539531707764,
+      "learning_rate": 2.01168983767315e-05,
+      "loss": 0.1417,
+      "step": 6261
+    },
+    {
+      "epoch": 4.713586751975913,
+      "grad_norm": 0.19945694506168365,
+      "learning_rate": 2.0109870154313838e-05,
+      "loss": 0.2317,
+      "step": 6262
+    },
+    {
+      "epoch": 4.714339480617237,
+      "grad_norm": 0.20017096400260925,
+      "learning_rate": 2.0102842740069666e-05,
+      "loss": 0.1087,
+      "step": 6263
+    },
+    {
+      "epoch": 4.715092209258563,
+      "grad_norm": 0.22003081440925598,
+      "learning_rate": 2.009581613476749e-05,
+      "loss": 0.1898,
+      "step": 6264
+    },
+    {
+      "epoch": 4.715844937899887,
+      "grad_norm": 0.17702147364616394,
+      "learning_rate": 2.0088790339175735e-05,
+      "loss": 0.2022,
+      "step": 6265
+    },
+    {
+      "epoch": 4.716597666541212,
+      "grad_norm": 0.2777307331562042,
+      "learning_rate": 2.008176535406271e-05,
+      "loss": 0.2564,
+      "step": 6266
+    },
+    {
+      "epoch": 4.717350395182537,
+      "grad_norm": 0.19265879690647125,
+      "learning_rate": 2.0074741180196686e-05,
+      "loss": 0.1971,
+      "step": 6267
+    },
+    {
+      "epoch": 4.7181031238238615,
+      "grad_norm": 0.22501195967197418,
+      "learning_rate": 2.00677178183458e-05,
+      "loss": 0.2614,
+      "step": 6268
+    },
+    {
+      "epoch": 4.718855852465186,
+      "grad_norm": 0.2402229905128479,
+      "learning_rate": 2.006069526927812e-05,
+      "loss": 0.1947,
+      "step": 6269
+    },
+    {
+      "epoch": 4.719608581106511,
+      "grad_norm": 0.19124601781368256,
+      "learning_rate": 2.0053673533761613e-05,
+      "loss": 0.2038,
+      "step": 6270
+    },
+    {
+      "epoch": 4.720361309747836,
+      "grad_norm": 0.23299378156661987,
+      "learning_rate": 2.0046652612564183e-05,
+      "loss": 0.2836,
+      "step": 6271
+    },
+    {
+      "epoch": 4.72111403838916,
+      "grad_norm": 0.22542956471443176,
+      "learning_rate": 2.003963250645361e-05,
+      "loss": 0.1432,
+      "step": 6272
+    },
+    {
+      "epoch": 4.721866767030486,
+      "grad_norm": 0.197622150182724,
+      "learning_rate": 2.0032613216197615e-05,
+      "loss": 0.221,
+      "step": 6273
+    },
+    {
+      "epoch": 4.72261949567181,
+      "grad_norm": 0.19078722596168518,
+      "learning_rate": 2.0025594742563807e-05,
+      "loss": 0.184,
+      "step": 6274
+    },
+    {
+      "epoch": 4.723372224313135,
+      "grad_norm": 0.31359973549842834,
+      "learning_rate": 2.0018577086319724e-05,
+      "loss": 0.2398,
+      "step": 6275
+    },
+    {
+      "epoch": 4.72412495295446,
+      "grad_norm": 0.1763066053390503,
+      "learning_rate": 2.00115602482328e-05,
+      "loss": 0.1593,
+      "step": 6276
+    },
+    {
+      "epoch": 4.7248776815957845,
+      "grad_norm": 0.2769826650619507,
+      "learning_rate": 2.00045442290704e-05,
+      "loss": 0.2236,
+      "step": 6277
+    },
+    {
+      "epoch": 4.72563041023711,
+      "grad_norm": 0.1965869814157486,
+      "learning_rate": 1.999752902959976e-05,
+      "loss": 0.1757,
+      "step": 6278
+    },
+    {
+      "epoch": 4.726383138878434,
+      "grad_norm": 0.2257390320301056,
+      "learning_rate": 1.9990514650588066e-05,
+      "loss": 0.2032,
+      "step": 6279
+    },
+    {
+      "epoch": 4.727135867519759,
+      "grad_norm": 0.2155158519744873,
+      "learning_rate": 1.9983501092802402e-05,
+      "loss": 0.18,
+      "step": 6280
+    },
+    {
+      "epoch": 4.727888596161084,
+      "grad_norm": 0.23705816268920898,
+      "learning_rate": 1.9976488357009766e-05,
+      "loss": 0.2116,
+      "step": 6281
+    },
+    {
+      "epoch": 4.728641324802409,
+      "grad_norm": 0.21993708610534668,
+      "learning_rate": 1.9969476443977043e-05,
+      "loss": 0.223,
+      "step": 6282
+    },
+    {
+      "epoch": 4.729394053443734,
+      "grad_norm": 0.17873473465442657,
+      "learning_rate": 1.9962465354471056e-05,
+      "loss": 0.162,
+      "step": 6283
+    },
+    {
+      "epoch": 4.7301467820850585,
+      "grad_norm": 0.25559428334236145,
+      "learning_rate": 1.9955455089258522e-05,
+      "loss": 0.2186,
+      "step": 6284
+    },
+    {
+      "epoch": 4.730899510726383,
+      "grad_norm": 0.1732226312160492,
+      "learning_rate": 1.9948445649106074e-05,
+      "loss": 0.2654,
+      "step": 6285
+    },
+    {
+      "epoch": 4.731652239367708,
+      "grad_norm": 0.2262609750032425,
+      "learning_rate": 1.9941437034780264e-05,
+      "loss": 0.2174,
+      "step": 6286
+    },
+    {
+      "epoch": 4.732404968009033,
+      "grad_norm": 0.19888128340244293,
+      "learning_rate": 1.9934429247047536e-05,
+      "loss": 0.1091,
+      "step": 6287
+    },
+    {
+      "epoch": 4.733157696650357,
+      "grad_norm": 0.186229407787323,
+      "learning_rate": 1.9927422286674247e-05,
+      "loss": 0.2082,
+      "step": 6288
+    },
+    {
+      "epoch": 4.733910425291683,
+      "grad_norm": 0.2504982650279999,
+      "learning_rate": 1.992041615442668e-05,
+      "loss": 0.2342,
+      "step": 6289
+    },
+    {
+      "epoch": 4.734663153933007,
+      "grad_norm": 0.21356317400932312,
+      "learning_rate": 1.9913410851070996e-05,
+      "loss": 0.2178,
+      "step": 6290
+    },
+    {
+      "epoch": 4.7354158825743315,
+      "grad_norm": 0.21797743439674377,
+      "learning_rate": 1.99064063773733e-05,
+      "loss": 0.2457,
+      "step": 6291
+    },
+    {
+      "epoch": 4.736168611215657,
+      "grad_norm": 0.2591923773288727,
+      "learning_rate": 1.9899402734099576e-05,
+      "loss": 0.2332,
+      "step": 6292
+    },
+    {
+      "epoch": 4.736921339856981,
+      "grad_norm": 0.19835157692432404,
+      "learning_rate": 1.9892399922015753e-05,
+      "loss": 0.2273,
+      "step": 6293
+    },
+    {
+      "epoch": 4.737674068498307,
+      "grad_norm": 0.2039031833410263,
+      "learning_rate": 1.988539794188763e-05,
+      "loss": 0.1976,
+      "step": 6294
+    },
+    {
+      "epoch": 4.738426797139631,
+      "grad_norm": 0.2078014612197876,
+      "learning_rate": 1.9878396794480948e-05,
+      "loss": 0.2267,
+      "step": 6295
+    },
+    {
+      "epoch": 4.739179525780956,
+      "grad_norm": 0.20321592688560486,
+      "learning_rate": 1.9871396480561327e-05,
+      "loss": 0.2574,
+      "step": 6296
+    },
+    {
+      "epoch": 4.739932254422281,
+      "grad_norm": 0.18686257302761078,
+      "learning_rate": 1.9864397000894325e-05,
+      "loss": 0.1907,
+      "step": 6297
+    },
+    {
+      "epoch": 4.7406849830636055,
+      "grad_norm": 0.24771757423877716,
+      "learning_rate": 1.9857398356245373e-05,
+      "loss": 0.2126,
+      "step": 6298
+    },
+    {
+      "epoch": 4.74143771170493,
+      "grad_norm": 0.22620563209056854,
+      "learning_rate": 1.9850400547379855e-05,
+      "loss": 0.2657,
+      "step": 6299
+    },
+    {
+      "epoch": 4.742190440346255,
+      "grad_norm": 0.24123866856098175,
+      "learning_rate": 1.984340357506303e-05,
+      "loss": 0.2177,
+      "step": 6300
+    },
+    {
+      "epoch": 4.74294316898758,
+      "grad_norm": 0.23544977605342865,
+      "learning_rate": 1.9836407440060085e-05,
+      "loss": 0.1519,
+      "step": 6301
+    },
+    {
+      "epoch": 4.743695897628905,
+      "grad_norm": 0.20797693729400635,
+      "learning_rate": 1.9829412143136096e-05,
+      "loss": 0.2474,
+      "step": 6302
+    },
+    {
+      "epoch": 4.74444862627023,
+      "grad_norm": 0.19910722970962524,
+      "learning_rate": 1.9822417685056065e-05,
+      "loss": 0.261,
+      "step": 6303
+    },
+    {
+      "epoch": 4.745201354911554,
+      "grad_norm": 0.21428604423999786,
+      "learning_rate": 1.981542406658488e-05,
+      "loss": 0.1958,
+      "step": 6304
+    },
+    {
+      "epoch": 4.7459540835528795,
+      "grad_norm": 0.21498243510723114,
+      "learning_rate": 1.9808431288487377e-05,
+      "loss": 0.2561,
+      "step": 6305
+    },
+    {
+      "epoch": 4.746706812194204,
+      "grad_norm": 0.19764351844787598,
+      "learning_rate": 1.9801439351528268e-05,
+      "loss": 0.1128,
+      "step": 6306
+    },
+    {
+      "epoch": 4.7474595408355285,
+      "grad_norm": 0.20015442371368408,
+      "learning_rate": 1.979444825647217e-05,
+      "loss": 0.1715,
+      "step": 6307
+    },
+    {
+      "epoch": 4.748212269476854,
+      "grad_norm": 0.2112327218055725,
+      "learning_rate": 1.9787458004083632e-05,
+      "loss": 0.1837,
+      "step": 6308
+    },
+    {
+      "epoch": 4.748964998118178,
+      "grad_norm": 0.188125878572464,
+      "learning_rate": 1.9780468595127087e-05,
+      "loss": 0.1426,
+      "step": 6309
+    },
+    {
+      "epoch": 4.749717726759503,
+      "grad_norm": 0.18421167135238647,
+      "learning_rate": 1.9773480030366892e-05,
+      "loss": 0.1952,
+      "step": 6310
+    },
+    {
+      "epoch": 4.750470455400828,
+      "grad_norm": 0.22413349151611328,
+      "learning_rate": 1.9766492310567293e-05,
+      "loss": 0.1976,
+      "step": 6311
+    },
+    {
+      "epoch": 4.751223184042153,
+      "grad_norm": 0.2567768692970276,
+      "learning_rate": 1.9759505436492476e-05,
+      "loss": 0.2649,
+      "step": 6312
+    },
+    {
+      "epoch": 4.751975912683478,
+      "grad_norm": 0.1873537302017212,
+      "learning_rate": 1.975251940890651e-05,
+      "loss": 0.2383,
+      "step": 6313
+    },
+    {
+      "epoch": 4.752728641324802,
+      "grad_norm": 0.2805364727973938,
+      "learning_rate": 1.974553422857337e-05,
+      "loss": 0.246,
+      "step": 6314
+    },
+    {
+      "epoch": 4.753481369966127,
+      "grad_norm": 0.18771906197071075,
+      "learning_rate": 1.9738549896256947e-05,
+      "loss": 0.1951,
+      "step": 6315
+    },
+    {
+      "epoch": 4.754234098607452,
+      "grad_norm": 0.2009226381778717,
+      "learning_rate": 1.973156641272104e-05,
+      "loss": 0.2629,
+      "step": 6316
+    },
+    {
+      "epoch": 4.754986827248777,
+      "grad_norm": 0.20131240785121918,
+      "learning_rate": 1.972458377872935e-05,
+      "loss": 0.1484,
+      "step": 6317
+    },
+    {
+      "epoch": 4.755739555890102,
+      "grad_norm": 0.2383032888174057,
+      "learning_rate": 1.9717601995045492e-05,
+      "loss": 0.1595,
+      "step": 6318
+    },
+    {
+      "epoch": 4.756492284531427,
+      "grad_norm": 0.20574499666690826,
+      "learning_rate": 1.971062106243297e-05,
+      "loss": 0.1966,
+      "step": 6319
+    },
+    {
+      "epoch": 4.757245013172751,
+      "grad_norm": 0.24149265885353088,
+      "learning_rate": 1.9703640981655226e-05,
+      "loss": 0.215,
+      "step": 6320
+    },
+    {
+      "epoch": 4.757997741814076,
+      "grad_norm": 0.202560693025589,
+      "learning_rate": 1.9696661753475577e-05,
+      "loss": 0.2384,
+      "step": 6321
+    },
+    {
+      "epoch": 4.758750470455401,
+      "grad_norm": 0.19141070544719696,
+      "learning_rate": 1.968968337865727e-05,
+      "loss": 0.1879,
+      "step": 6322
+    },
+    {
+      "epoch": 4.759503199096725,
+      "grad_norm": 0.17021366953849792,
+      "learning_rate": 1.9682705857963434e-05,
+      "loss": 0.141,
+      "step": 6323
+    },
+    {
+      "epoch": 4.760255927738051,
+      "grad_norm": 0.2530558407306671,
+      "learning_rate": 1.967572919215715e-05,
+      "loss": 0.1911,
+      "step": 6324
+    },
+    {
+      "epoch": 4.761008656379375,
+      "grad_norm": 0.22677163779735565,
+      "learning_rate": 1.9668753382001348e-05,
+      "loss": 0.147,
+      "step": 6325
+    },
+    {
+      "epoch": 4.7617613850207,
+      "grad_norm": 0.1792307198047638,
+      "learning_rate": 1.9661778428258908e-05,
+      "loss": 0.2366,
+      "step": 6326
+    },
+    {
+      "epoch": 4.762514113662025,
+      "grad_norm": 0.1827917993068695,
+      "learning_rate": 1.9654804331692594e-05,
+      "loss": 0.2188,
+      "step": 6327
+    },
+    {
+      "epoch": 4.7632668423033495,
+      "grad_norm": 0.2171882838010788,
+      "learning_rate": 1.9647831093065082e-05,
+      "loss": 0.178,
+      "step": 6328
+    },
+    {
+      "epoch": 4.764019570944674,
+      "grad_norm": 0.20616208016872406,
+      "learning_rate": 1.9640858713138958e-05,
+      "loss": 0.2597,
+      "step": 6329
+    },
+    {
+      "epoch": 4.764772299585999,
+      "grad_norm": 0.22851701080799103,
+      "learning_rate": 1.96338871926767e-05,
+      "loss": 0.272,
+      "step": 6330
+    },
+    {
+      "epoch": 4.765525028227324,
+      "grad_norm": 0.19965746998786926,
+      "learning_rate": 1.962691653244072e-05,
+      "loss": 0.2322,
+      "step": 6331
+    },
+    {
+      "epoch": 4.766277756868649,
+      "grad_norm": 0.16608987748622894,
+      "learning_rate": 1.9619946733193306e-05,
+      "loss": 0.0905,
+      "step": 6332
+    },
+    {
+      "epoch": 4.767030485509974,
+      "grad_norm": 0.17506985366344452,
+      "learning_rate": 1.961297779569668e-05,
+      "loss": 0.1523,
+      "step": 6333
+    },
+    {
+      "epoch": 4.767783214151298,
+      "grad_norm": 0.2004774510860443,
+      "learning_rate": 1.9606009720712938e-05,
+      "loss": 0.1501,
+      "step": 6334
+    },
+    {
+      "epoch": 4.7685359427926235,
+      "grad_norm": 0.20598313212394714,
+      "learning_rate": 1.9599042509004102e-05,
+      "loss": 0.2081,
+      "step": 6335
+    },
+    {
+      "epoch": 4.769288671433948,
+      "grad_norm": 0.22605155408382416,
+      "learning_rate": 1.9592076161332095e-05,
+      "loss": 0.1547,
+      "step": 6336
+    },
+    {
+      "epoch": 4.770041400075273,
+      "grad_norm": 0.21639975905418396,
+      "learning_rate": 1.9585110678458758e-05,
+      "loss": 0.25,
+      "step": 6337
+    },
+    {
+      "epoch": 4.770794128716598,
+      "grad_norm": 0.19936494529247284,
+      "learning_rate": 1.9578146061145815e-05,
+      "loss": 0.0983,
+      "step": 6338
+    },
+    {
+      "epoch": 4.771546857357922,
+      "grad_norm": 0.21203257143497467,
+      "learning_rate": 1.9571182310154907e-05,
+      "loss": 0.2398,
+      "step": 6339
+    },
+    {
+      "epoch": 4.772299585999248,
+      "grad_norm": 0.20244212448596954,
+      "learning_rate": 1.9564219426247576e-05,
+      "loss": 0.2056,
+      "step": 6340
+    },
+    {
+      "epoch": 4.773052314640572,
+      "grad_norm": 0.21367907524108887,
+      "learning_rate": 1.9557257410185285e-05,
+      "loss": 0.1825,
+      "step": 6341
+    },
+    {
+      "epoch": 4.773805043281897,
+      "grad_norm": 0.1742090880870819,
+      "learning_rate": 1.955029626272938e-05,
+      "loss": 0.1237,
+      "step": 6342
+    },
+    {
+      "epoch": 4.774557771923222,
+      "grad_norm": 0.19660884141921997,
+      "learning_rate": 1.954333598464112e-05,
+      "loss": 0.2034,
+      "step": 6343
+    },
+    {
+      "epoch": 4.775310500564546,
+      "grad_norm": 0.18406033515930176,
+      "learning_rate": 1.9536376576681676e-05,
+      "loss": 0.1154,
+      "step": 6344
+    },
+    {
+      "epoch": 4.776063229205871,
+      "grad_norm": 0.2051040679216385,
+      "learning_rate": 1.9529418039612118e-05,
+      "loss": 0.2468,
+      "step": 6345
+    },
+    {
+      "epoch": 4.776815957847196,
+      "grad_norm": 0.18971602618694305,
+      "learning_rate": 1.952246037419342e-05,
+      "loss": 0.24,
+      "step": 6346
+    },
+    {
+      "epoch": 4.777568686488521,
+      "grad_norm": 0.1972414255142212,
+      "learning_rate": 1.951550358118646e-05,
+      "loss": 0.1485,
+      "step": 6347
+    },
+    {
+      "epoch": 4.778321415129845,
+      "grad_norm": 0.24690937995910645,
+      "learning_rate": 1.950854766135203e-05,
+      "loss": 0.1815,
+      "step": 6348
+    },
+    {
+      "epoch": 4.779074143771171,
+      "grad_norm": 0.25884345173835754,
+      "learning_rate": 1.9501592615450797e-05,
+      "loss": 0.3127,
+      "step": 6349
+    },
+    {
+      "epoch": 4.779826872412495,
+      "grad_norm": 0.1759379506111145,
+      "learning_rate": 1.949463844424338e-05,
+      "loss": 0.239,
+      "step": 6350
+    },
+    {
+      "epoch": 4.78057960105382,
+      "grad_norm": 0.2113979160785675,
+      "learning_rate": 1.948768514849027e-05,
+      "loss": 0.2021,
+      "step": 6351
+    },
+    {
+      "epoch": 4.781332329695145,
+      "grad_norm": 0.174796000123024,
+      "learning_rate": 1.948073272895186e-05,
+      "loss": 0.1191,
+      "step": 6352
+    },
+    {
+      "epoch": 4.782085058336469,
+      "grad_norm": 0.19949443638324738,
+      "learning_rate": 1.9473781186388466e-05,
+      "loss": 0.1883,
+      "step": 6353
+    },
+    {
+      "epoch": 4.782837786977795,
+      "grad_norm": 0.2064112424850464,
+      "learning_rate": 1.9466830521560297e-05,
+      "loss": 0.1897,
+      "step": 6354
+    },
+    {
+      "epoch": 4.783590515619119,
+      "grad_norm": 0.1923416256904602,
+      "learning_rate": 1.9459880735227453e-05,
+      "loss": 0.1731,
+      "step": 6355
+    },
+    {
+      "epoch": 4.7843432442604445,
+      "grad_norm": 0.18870367109775543,
+      "learning_rate": 1.9452931828149974e-05,
+      "loss": 0.2384,
+      "step": 6356
+    },
+    {
+      "epoch": 4.785095972901769,
+      "grad_norm": 0.22177010774612427,
+      "learning_rate": 1.944598380108777e-05,
+      "loss": 0.1588,
+      "step": 6357
+    },
+    {
+      "epoch": 4.7858487015430935,
+      "grad_norm": 0.20471231639385223,
+      "learning_rate": 1.943903665480067e-05,
+      "loss": 0.2299,
+      "step": 6358
+    },
+    {
+      "epoch": 4.786601430184419,
+      "grad_norm": 0.19244596362113953,
+      "learning_rate": 1.9432090390048395e-05,
+      "loss": 0.119,
+      "step": 6359
+    },
+    {
+      "epoch": 4.787354158825743,
+      "grad_norm": 0.26346883177757263,
+      "learning_rate": 1.9425145007590595e-05,
+      "loss": 0.1887,
+      "step": 6360
+    },
+    {
+      "epoch": 4.788106887467068,
+      "grad_norm": 0.20104439556598663,
+      "learning_rate": 1.9418200508186785e-05,
+      "loss": 0.2151,
+      "step": 6361
+    },
+    {
+      "epoch": 4.788859616108393,
+      "grad_norm": 0.1802431046962738,
+      "learning_rate": 1.9411256892596423e-05,
+      "loss": 0.1789,
+      "step": 6362
+    },
+    {
+      "epoch": 4.789612344749718,
+      "grad_norm": 0.18092338740825653,
+      "learning_rate": 1.9404314161578844e-05,
+      "loss": 0.1655,
+      "step": 6363
+    },
+    {
+      "epoch": 4.790365073391042,
+      "grad_norm": 0.19698718190193176,
+      "learning_rate": 1.9397372315893303e-05,
+      "loss": 0.2162,
+      "step": 6364
+    },
+    {
+      "epoch": 4.7911178020323675,
+      "grad_norm": 0.2008783221244812,
+      "learning_rate": 1.9390431356298932e-05,
+      "loss": 0.2222,
+      "step": 6365
+    },
+    {
+      "epoch": 4.791870530673692,
+      "grad_norm": 0.18303215503692627,
+      "learning_rate": 1.9383491283554806e-05,
+      "loss": 0.1318,
+      "step": 6366
+    },
+    {
+      "epoch": 4.792623259315017,
+      "grad_norm": 0.193119078874588,
+      "learning_rate": 1.9376552098419864e-05,
+      "loss": 0.1736,
+      "step": 6367
+    },
+    {
+      "epoch": 4.793375987956342,
+      "grad_norm": 0.1712648570537567,
+      "learning_rate": 1.9369613801652968e-05,
+      "loss": 0.1184,
+      "step": 6368
+    },
+    {
+      "epoch": 4.794128716597666,
+      "grad_norm": 0.21031148731708527,
+      "learning_rate": 1.936267639401289e-05,
+      "loss": 0.1523,
+      "step": 6369
+    },
+    {
+      "epoch": 4.794881445238992,
+      "grad_norm": 0.2190135419368744,
+      "learning_rate": 1.9355739876258283e-05,
+      "loss": 0.2511,
+      "step": 6370
+    },
+    {
+      "epoch": 4.795634173880316,
+      "grad_norm": 0.21123924851417542,
+      "learning_rate": 1.9348804249147722e-05,
+      "loss": 0.1778,
+      "step": 6371
+    },
+    {
+      "epoch": 4.796386902521641,
+      "grad_norm": 0.1989087015390396,
+      "learning_rate": 1.9341869513439677e-05,
+      "loss": 0.2203,
+      "step": 6372
+    },
+    {
+      "epoch": 4.797139631162966,
+      "grad_norm": 0.19442664086818695,
+      "learning_rate": 1.9334935669892513e-05,
+      "loss": 0.2698,
+      "step": 6373
+    },
+    {
+      "epoch": 4.79789235980429,
+      "grad_norm": 0.20363299548625946,
+      "learning_rate": 1.93280027192645e-05,
+      "loss": 0.2698,
+      "step": 6374
+    },
+    {
+      "epoch": 4.798645088445616,
+      "grad_norm": 0.20693567395210266,
+      "learning_rate": 1.9321070662313827e-05,
+      "loss": 0.2099,
+      "step": 6375
+    },
+    {
+      "epoch": 4.79939781708694,
+      "grad_norm": 0.24073967337608337,
+      "learning_rate": 1.9314139499798572e-05,
+      "loss": 0.2507,
+      "step": 6376
+    },
+    {
+      "epoch": 4.800150545728265,
+      "grad_norm": 0.18473277986049652,
+      "learning_rate": 1.930720923247672e-05,
+      "loss": 0.1328,
+      "step": 6377
+    },
+    {
+      "epoch": 4.80090327436959,
+      "grad_norm": 0.17498765885829926,
+      "learning_rate": 1.930027986110614e-05,
+      "loss": 0.1515,
+      "step": 6378
+    },
+    {
+      "epoch": 4.8016560030109146,
+      "grad_norm": 0.1895669400691986,
+      "learning_rate": 1.9293351386444637e-05,
+      "loss": 0.2572,
+      "step": 6379
+    },
+    {
+      "epoch": 4.802408731652239,
+      "grad_norm": 0.21355648338794708,
+      "learning_rate": 1.9286423809249876e-05,
+      "loss": 0.15,
+      "step": 6380
+    },
+    {
+      "epoch": 4.803161460293564,
+      "grad_norm": 0.21420800685882568,
+      "learning_rate": 1.9279497130279457e-05,
+      "loss": 0.2581,
+      "step": 6381
+    },
+    {
+      "epoch": 4.803914188934889,
+      "grad_norm": 0.20216457545757294,
+      "learning_rate": 1.927257135029088e-05,
+      "loss": 0.2363,
+      "step": 6382
+    },
+    {
+      "epoch": 4.804666917576213,
+      "grad_norm": 0.19096018373966217,
+      "learning_rate": 1.9265646470041527e-05,
+      "loss": 0.1587,
+      "step": 6383
+    },
+    {
+      "epoch": 4.805419646217539,
+      "grad_norm": 0.2159426361322403,
+      "learning_rate": 1.92587224902887e-05,
+      "loss": 0.1291,
+      "step": 6384
+    },
+    {
+      "epoch": 4.806172374858863,
+      "grad_norm": 0.22766925394535065,
+      "learning_rate": 1.925179941178959e-05,
+      "loss": 0.2195,
+      "step": 6385
+    },
+    {
+      "epoch": 4.8069251035001885,
+      "grad_norm": 0.21139325201511383,
+      "learning_rate": 1.924487723530129e-05,
+      "loss": 0.2454,
+      "step": 6386
+    },
+    {
+      "epoch": 4.807677832141513,
+      "grad_norm": 0.21487298607826233,
+      "learning_rate": 1.9237955961580806e-05,
+      "loss": 0.26,
+      "step": 6387
+    },
+    {
+      "epoch": 4.8084305607828375,
+      "grad_norm": 0.1709480881690979,
+      "learning_rate": 1.923103559138504e-05,
+      "loss": 0.1204,
+      "step": 6388
+    },
+    {
+      "epoch": 4.809183289424163,
+      "grad_norm": 0.21548296511173248,
+      "learning_rate": 1.9224116125470785e-05,
+      "loss": 0.1921,
+      "step": 6389
+    },
+    {
+      "epoch": 4.809936018065487,
+      "grad_norm": 0.1929355263710022,
+      "learning_rate": 1.9217197564594746e-05,
+      "loss": 0.1875,
+      "step": 6390
+    },
+    {
+      "epoch": 4.810688746706813,
+      "grad_norm": 0.23249180614948273,
+      "learning_rate": 1.9210279909513534e-05,
+      "loss": 0.252,
+      "step": 6391
+    },
+    {
+      "epoch": 4.811441475348137,
+      "grad_norm": 0.19886629283428192,
+      "learning_rate": 1.9203363160983652e-05,
+      "loss": 0.2289,
+      "step": 6392
+    },
+    {
+      "epoch": 4.812194203989462,
+      "grad_norm": 0.1719769984483719,
+      "learning_rate": 1.919644731976149e-05,
+      "loss": 0.2104,
+      "step": 6393
+    },
+    {
+      "epoch": 4.812946932630787,
+      "grad_norm": 0.19597069919109344,
+      "learning_rate": 1.9189532386603365e-05,
+      "loss": 0.2116,
+      "step": 6394
+    },
+    {
+      "epoch": 4.8136996612721115,
+      "grad_norm": 0.19171658158302307,
+      "learning_rate": 1.9182618362265493e-05,
+      "loss": 0.1854,
+      "step": 6395
+    },
+    {
+      "epoch": 4.814452389913436,
+      "grad_norm": 0.18605545163154602,
+      "learning_rate": 1.9175705247503966e-05,
+      "loss": 0.1975,
+      "step": 6396
+    },
+    {
+      "epoch": 4.815205118554761,
+      "grad_norm": 0.20291879773139954,
+      "learning_rate": 1.91687930430748e-05,
+      "loss": 0.1767,
+      "step": 6397
+    },
+    {
+      "epoch": 4.815957847196086,
+      "grad_norm": 0.21876688301563263,
+      "learning_rate": 1.91618817497339e-05,
+      "loss": 0.2055,
+      "step": 6398
+    },
+    {
+      "epoch": 4.81671057583741,
+      "grad_norm": 0.22159963846206665,
+      "learning_rate": 1.915497136823708e-05,
+      "loss": 0.2317,
+      "step": 6399
+    },
+    {
+      "epoch": 4.817463304478736,
+      "grad_norm": 0.18527288734912872,
+      "learning_rate": 1.914806189934003e-05,
+      "loss": 0.1849,
+      "step": 6400
+    },
+    {
+      "epoch": 4.817463304478736,
+      "eval_loss": 0.22520853579044342,
+      "eval_runtime": 452.6677,
+      "eval_samples_per_second": 21.267,
+      "eval_steps_per_second": 0.665,
+      "step": 6400
+    },
+    {
+      "epoch": 4.81821603312006,
+      "grad_norm": 0.1826200634241104,
+      "learning_rate": 1.9141153343798392e-05,
+      "loss": 0.1898,
+      "step": 6401
+    },
+    {
+      "epoch": 4.818968761761385,
+      "grad_norm": 0.18936248123645782,
+      "learning_rate": 1.913424570236765e-05,
+      "loss": 0.1462,
+      "step": 6402
+    },
+    {
+      "epoch": 4.81972149040271,
+      "grad_norm": 0.2161732167005539,
+      "learning_rate": 1.9127338975803215e-05,
+      "loss": 0.2805,
+      "step": 6403
+    },
+    {
+      "epoch": 4.820474219044034,
+      "grad_norm": 0.23926039040088654,
+      "learning_rate": 1.912043316486041e-05,
+      "loss": 0.198,
+      "step": 6404
+    },
+    {
+      "epoch": 4.82122694768536,
+      "grad_norm": 0.18793343007564545,
+      "learning_rate": 1.9113528270294433e-05,
+      "loss": 0.1555,
+      "step": 6405
+    },
+    {
+      "epoch": 4.821979676326684,
+      "grad_norm": 0.18732650578022003,
+      "learning_rate": 1.9106624292860388e-05,
+      "loss": 0.1088,
+      "step": 6406
+    },
+    {
+      "epoch": 4.822732404968009,
+      "grad_norm": 0.20998580753803253,
+      "learning_rate": 1.90997212333133e-05,
+      "loss": 0.1523,
+      "step": 6407
+    },
+    {
+      "epoch": 4.823485133609334,
+      "grad_norm": 0.2279466837644577,
+      "learning_rate": 1.9092819092408066e-05,
+      "loss": 0.2401,
+      "step": 6408
+    },
+    {
+      "epoch": 4.8242378622506585,
+      "grad_norm": 0.2756803631782532,
+      "learning_rate": 1.9085917870899496e-05,
+      "loss": 0.2148,
+      "step": 6409
+    },
+    {
+      "epoch": 4.824990590891984,
+      "grad_norm": 0.19298389554023743,
+      "learning_rate": 1.9079017569542295e-05,
+      "loss": 0.2333,
+      "step": 6410
+    },
+    {
+      "epoch": 4.825743319533308,
+      "grad_norm": 0.21123313903808594,
+      "learning_rate": 1.9072118189091072e-05,
+      "loss": 0.2338,
+      "step": 6411
+    },
+    {
+      "epoch": 4.826496048174633,
+      "grad_norm": 0.22574925422668457,
+      "learning_rate": 1.906521973030034e-05,
+      "loss": 0.2743,
+      "step": 6412
+    },
+    {
+      "epoch": 4.827248776815958,
+      "grad_norm": 0.23545421659946442,
+      "learning_rate": 1.9058322193924484e-05,
+      "loss": 0.1757,
+      "step": 6413
+    },
+    {
+      "epoch": 4.828001505457283,
+      "grad_norm": 0.26143625378608704,
+      "learning_rate": 1.905142558071783e-05,
+      "loss": 0.2021,
+      "step": 6414
+    },
+    {
+      "epoch": 4.828754234098607,
+      "grad_norm": 0.2400417923927307,
+      "learning_rate": 1.9044529891434566e-05,
+      "loss": 0.2418,
+      "step": 6415
+    },
+    {
+      "epoch": 4.8295069627399325,
+      "grad_norm": 0.1860826015472412,
+      "learning_rate": 1.9037635126828806e-05,
+      "loss": 0.1282,
+      "step": 6416
+    },
+    {
+      "epoch": 4.830259691381257,
+      "grad_norm": 0.18803755939006805,
+      "learning_rate": 1.9030741287654547e-05,
+      "loss": 0.1001,
+      "step": 6417
+    },
+    {
+      "epoch": 4.8310124200225815,
+      "grad_norm": 0.21442361176013947,
+      "learning_rate": 1.902384837466569e-05,
+      "loss": 0.208,
+      "step": 6418
+    },
+    {
+      "epoch": 4.831765148663907,
+      "grad_norm": 0.1601109504699707,
+      "learning_rate": 1.9016956388616013e-05,
+      "loss": 0.1214,
+      "step": 6419
+    },
+    {
+      "epoch": 4.832517877305231,
+      "grad_norm": 0.18865475058555603,
+      "learning_rate": 1.9010065330259248e-05,
+      "loss": 0.2249,
+      "step": 6420
+    },
+    {
+      "epoch": 4.833270605946556,
+      "grad_norm": 0.19167354702949524,
+      "learning_rate": 1.900317520034897e-05,
+      "loss": 0.1352,
+      "step": 6421
+    },
+    {
+      "epoch": 4.834023334587881,
+      "grad_norm": 0.1843806952238083,
+      "learning_rate": 1.899628599963868e-05,
+      "loss": 0.1776,
+      "step": 6422
+    },
+    {
+      "epoch": 4.834776063229206,
+      "grad_norm": 0.18969067931175232,
+      "learning_rate": 1.8989397728881765e-05,
+      "loss": 0.2243,
+      "step": 6423
+    },
+    {
+      "epoch": 4.835528791870531,
+      "grad_norm": 0.2106957733631134,
+      "learning_rate": 1.8982510388831522e-05,
+      "loss": 0.2302,
+      "step": 6424
+    },
+    {
+      "epoch": 4.8362815205118554,
+      "grad_norm": 0.18110667169094086,
+      "learning_rate": 1.8975623980241124e-05,
+      "loss": 0.1606,
+      "step": 6425
+    },
+    {
+      "epoch": 4.83703424915318,
+      "grad_norm": 0.19206273555755615,
+      "learning_rate": 1.8968738503863685e-05,
+      "loss": 0.2079,
+      "step": 6426
+    },
+    {
+      "epoch": 4.837786977794505,
+      "grad_norm": 0.23900943994522095,
+      "learning_rate": 1.8961853960452174e-05,
+      "loss": 0.2418,
+      "step": 6427
+    },
+    {
+      "epoch": 4.83853970643583,
+      "grad_norm": 0.22940793633460999,
+      "learning_rate": 1.8954970350759476e-05,
+      "loss": 0.1908,
+      "step": 6428
+    },
+    {
+      "epoch": 4.839292435077155,
+      "grad_norm": 0.19074560701847076,
+      "learning_rate": 1.8948087675538378e-05,
+      "loss": 0.1581,
+      "step": 6429
+    },
+    {
+      "epoch": 4.84004516371848,
+      "grad_norm": 0.15536880493164062,
+      "learning_rate": 1.894120593554155e-05,
+      "loss": 0.1229,
+      "step": 6430
+    },
+    {
+      "epoch": 4.840797892359804,
+      "grad_norm": 0.23442216217517853,
+      "learning_rate": 1.8934325131521575e-05,
+      "loss": 0.1774,
+      "step": 6431
+    },
+    {
+      "epoch": 4.841550621001129,
+      "grad_norm": 0.19120843708515167,
+      "learning_rate": 1.8927445264230916e-05,
+      "loss": 0.114,
+      "step": 6432
+    },
+    {
+      "epoch": 4.842303349642454,
+      "grad_norm": 0.1794971227645874,
+      "learning_rate": 1.8920566334421964e-05,
+      "loss": 0.2147,
+      "step": 6433
+    },
+    {
+      "epoch": 4.843056078283778,
+      "grad_norm": 0.18433810770511627,
+      "learning_rate": 1.8913688342846975e-05,
+      "loss": 0.1926,
+      "step": 6434
+    },
+    {
+      "epoch": 4.843808806925104,
+      "grad_norm": 0.21728670597076416,
+      "learning_rate": 1.890681129025812e-05,
+      "loss": 0.2414,
+      "step": 6435
+    },
+    {
+      "epoch": 4.844561535566428,
+      "grad_norm": 0.22273562848567963,
+      "learning_rate": 1.8899935177407464e-05,
+      "loss": 0.221,
+      "step": 6436
+    },
+    {
+      "epoch": 4.845314264207753,
+      "grad_norm": 0.20753169059753418,
+      "learning_rate": 1.8893060005046966e-05,
+      "loss": 0.1519,
+      "step": 6437
+    },
+    {
+      "epoch": 4.846066992849078,
+      "grad_norm": 0.18692141771316528,
+      "learning_rate": 1.888618577392848e-05,
+      "loss": 0.1687,
+      "step": 6438
+    },
+    {
+      "epoch": 4.8468197214904025,
+      "grad_norm": 0.2083490490913391,
+      "learning_rate": 1.8879312484803773e-05,
+      "loss": 0.2199,
+      "step": 6439
+    },
+    {
+      "epoch": 4.847572450131728,
+      "grad_norm": 0.21635016798973083,
+      "learning_rate": 1.887244013842449e-05,
+      "loss": 0.271,
+      "step": 6440
+    },
+    {
+      "epoch": 4.848325178773052,
+      "grad_norm": 0.18247105181217194,
+      "learning_rate": 1.886556873554218e-05,
+      "loss": 0.2112,
+      "step": 6441
+    },
+    {
+      "epoch": 4.849077907414377,
+      "grad_norm": 0.22937476634979248,
+      "learning_rate": 1.88586982769083e-05,
+      "loss": 0.1557,
+      "step": 6442
+    },
+    {
+      "epoch": 4.849830636055702,
+      "grad_norm": 0.20462332665920258,
+      "learning_rate": 1.885182876327418e-05,
+      "loss": 0.2228,
+      "step": 6443
+    },
+    {
+      "epoch": 4.850583364697027,
+      "grad_norm": 0.19915828108787537,
+      "learning_rate": 1.884496019539106e-05,
+      "loss": 0.1887,
+      "step": 6444
+    },
+    {
+      "epoch": 4.851336093338351,
+      "grad_norm": 0.23532332479953766,
+      "learning_rate": 1.883809257401008e-05,
+      "loss": 0.1635,
+      "step": 6445
+    },
+    {
+      "epoch": 4.8520888219796765,
+      "grad_norm": 0.21480374038219452,
+      "learning_rate": 1.8831225899882276e-05,
+      "loss": 0.2398,
+      "step": 6446
+    },
+    {
+      "epoch": 4.852841550621001,
+      "grad_norm": 0.24537363648414612,
+      "learning_rate": 1.8824360173758583e-05,
+      "loss": 0.1893,
+      "step": 6447
+    },
+    {
+      "epoch": 4.853594279262326,
+      "grad_norm": 0.16452239453792572,
+      "learning_rate": 1.8817495396389807e-05,
+      "loss": 0.1329,
+      "step": 6448
+    },
+    {
+      "epoch": 4.854347007903651,
+      "grad_norm": 0.2404678910970688,
+      "learning_rate": 1.881063156852669e-05,
+      "loss": 0.2303,
+      "step": 6449
+    },
+    {
+      "epoch": 4.855099736544975,
+      "grad_norm": 0.2360159009695053,
+      "learning_rate": 1.8803768690919832e-05,
+      "loss": 0.214,
+      "step": 6450
+    },
+    {
+      "epoch": 4.855852465186301,
+      "grad_norm": 0.18148982524871826,
+      "learning_rate": 1.879690676431975e-05,
+      "loss": 0.1367,
+      "step": 6451
+    },
+    {
+      "epoch": 4.856605193827625,
+      "grad_norm": 0.26153284311294556,
+      "learning_rate": 1.8790045789476866e-05,
+      "loss": 0.2234,
+      "step": 6452
+    },
+    {
+      "epoch": 4.85735792246895,
+      "grad_norm": 0.181520476937294,
+      "learning_rate": 1.878318576714148e-05,
+      "loss": 0.2107,
+      "step": 6453
+    },
+    {
+      "epoch": 4.858110651110275,
+      "grad_norm": 0.19691632688045502,
+      "learning_rate": 1.8776326698063795e-05,
+      "loss": 0.2122,
+      "step": 6454
+    },
+    {
+      "epoch": 4.858863379751599,
+      "grad_norm": 0.17961828410625458,
+      "learning_rate": 1.8769468582993905e-05,
+      "loss": 0.142,
+      "step": 6455
+    },
+    {
+      "epoch": 4.859616108392924,
+      "grad_norm": 0.2215823084115982,
+      "learning_rate": 1.876261142268181e-05,
+      "loss": 0.2485,
+      "step": 6456
+    },
+    {
+      "epoch": 4.860368837034249,
+      "grad_norm": 0.19310373067855835,
+      "learning_rate": 1.8755755217877373e-05,
+      "loss": 0.1875,
+      "step": 6457
+    },
+    {
+      "epoch": 4.861121565675574,
+      "grad_norm": 0.18033815920352936,
+      "learning_rate": 1.8748899969330414e-05,
+      "loss": 0.2221,
+      "step": 6458
+    },
+    {
+      "epoch": 4.861874294316899,
+      "grad_norm": 0.23099561035633087,
+      "learning_rate": 1.87420456777906e-05,
+      "loss": 0.1688,
+      "step": 6459
+    },
+    {
+      "epoch": 4.862627022958224,
+      "grad_norm": 0.21037335693836212,
+      "learning_rate": 1.87351923440075e-05,
+      "loss": 0.2671,
+      "step": 6460
+    },
+    {
+      "epoch": 4.863379751599548,
+      "grad_norm": 0.21069248020648956,
+      "learning_rate": 1.8728339968730584e-05,
+      "loss": 0.1403,
+      "step": 6461
+    },
+    {
+      "epoch": 4.864132480240873,
+      "grad_norm": 0.21305491030216217,
+      "learning_rate": 1.8721488552709226e-05,
+      "loss": 0.2344,
+      "step": 6462
+    },
+    {
+      "epoch": 4.864885208882198,
+      "grad_norm": 0.21105559170246124,
+      "learning_rate": 1.8714638096692676e-05,
+      "loss": 0.2163,
+      "step": 6463
+    },
+    {
+      "epoch": 4.865637937523523,
+      "grad_norm": 0.21163570880889893,
+      "learning_rate": 1.8707788601430097e-05,
+      "loss": 0.1811,
+      "step": 6464
+    },
+    {
+      "epoch": 4.866390666164848,
+      "grad_norm": 0.21653003990650177,
+      "learning_rate": 1.8700940067670535e-05,
+      "loss": 0.2148,
+      "step": 6465
+    },
+    {
+      "epoch": 4.867143394806172,
+      "grad_norm": 0.2129582166671753,
+      "learning_rate": 1.8694092496162945e-05,
+      "loss": 0.2358,
+      "step": 6466
+    },
+    {
+      "epoch": 4.867896123447498,
+      "grad_norm": 0.17892242968082428,
+      "learning_rate": 1.8687245887656162e-05,
+      "loss": 0.1791,
+      "step": 6467
+    },
+    {
+      "epoch": 4.868648852088822,
+      "grad_norm": 0.20060348510742188,
+      "learning_rate": 1.8680400242898917e-05,
+      "loss": 0.1967,
+      "step": 6468
+    },
+    {
+      "epoch": 4.8694015807301465,
+      "grad_norm": 0.22473527491092682,
+      "learning_rate": 1.867355556263984e-05,
+      "loss": 0.141,
+      "step": 6469
+    },
+    {
+      "epoch": 4.870154309371472,
+      "grad_norm": 0.18464122712612152,
+      "learning_rate": 1.8666711847627445e-05,
+      "loss": 0.1824,
+      "step": 6470
+    },
+    {
+      "epoch": 4.870907038012796,
+      "grad_norm": 0.22038592398166656,
+      "learning_rate": 1.8659869098610172e-05,
+      "loss": 0.1504,
+      "step": 6471
+    },
+    {
+      "epoch": 4.871659766654121,
+      "grad_norm": 0.192650705575943,
+      "learning_rate": 1.8653027316336326e-05,
+      "loss": 0.1981,
+      "step": 6472
+    },
+    {
+      "epoch": 4.872412495295446,
+      "grad_norm": 0.21936899423599243,
+      "learning_rate": 1.8646186501554113e-05,
+      "loss": 0.1421,
+      "step": 6473
+    },
+    {
+      "epoch": 4.873165223936771,
+      "grad_norm": 0.21089564263820648,
+      "learning_rate": 1.8639346655011624e-05,
+      "loss": 0.1142,
+      "step": 6474
+    },
+    {
+      "epoch": 4.873917952578095,
+      "grad_norm": 0.20083515346050262,
+      "learning_rate": 1.863250777745687e-05,
+      "loss": 0.2204,
+      "step": 6475
+    },
+    {
+      "epoch": 4.8746706812194205,
+      "grad_norm": 0.23093242943286896,
+      "learning_rate": 1.8625669869637724e-05,
+      "loss": 0.2196,
+      "step": 6476
+    },
+    {
+      "epoch": 4.875423409860745,
+      "grad_norm": 0.20463231205940247,
+      "learning_rate": 1.861883293230199e-05,
+      "loss": 0.1674,
+      "step": 6477
+    },
+    {
+      "epoch": 4.87617613850207,
+      "grad_norm": 0.19608108699321747,
+      "learning_rate": 1.8611996966197327e-05,
+      "loss": 0.1851,
+      "step": 6478
+    },
+    {
+      "epoch": 4.876928867143395,
+      "grad_norm": 0.18011820316314697,
+      "learning_rate": 1.8605161972071317e-05,
+      "loss": 0.1276,
+      "step": 6479
+    },
+    {
+      "epoch": 4.877681595784719,
+      "grad_norm": 0.19670534133911133,
+      "learning_rate": 1.8598327950671418e-05,
+      "loss": 0.2005,
+      "step": 6480
+    },
+    {
+      "epoch": 4.878434324426045,
+      "grad_norm": 0.19809724390506744,
+      "learning_rate": 1.8591494902744994e-05,
+      "loss": 0.1855,
+      "step": 6481
+    },
+    {
+      "epoch": 4.879187053067369,
+      "grad_norm": 0.23431070148944855,
+      "learning_rate": 1.858466282903929e-05,
+      "loss": 0.2309,
+      "step": 6482
+    },
+    {
+      "epoch": 4.8799397817086945,
+      "grad_norm": 0.22356165945529938,
+      "learning_rate": 1.8577831730301448e-05,
+      "loss": 0.193,
+      "step": 6483
+    },
+    {
+      "epoch": 4.880692510350019,
+      "grad_norm": 0.2079782485961914,
+      "learning_rate": 1.8571001607278525e-05,
+      "loss": 0.2053,
+      "step": 6484
+    },
+    {
+      "epoch": 4.881445238991343,
+      "grad_norm": 0.188282772898674,
+      "learning_rate": 1.856417246071744e-05,
+      "loss": 0.1805,
+      "step": 6485
+    },
+    {
+      "epoch": 4.882197967632669,
+      "grad_norm": 0.22337304055690765,
+      "learning_rate": 1.8557344291365015e-05,
+      "loss": 0.2007,
+      "step": 6486
+    },
+    {
+      "epoch": 4.882950696273993,
+      "grad_norm": 0.18206654489040375,
+      "learning_rate": 1.8550517099967978e-05,
+      "loss": 0.22,
+      "step": 6487
+    },
+    {
+      "epoch": 4.883703424915318,
+      "grad_norm": 0.1946128010749817,
+      "learning_rate": 1.8543690887272934e-05,
+      "loss": 0.2201,
+      "step": 6488
+    },
+    {
+      "epoch": 4.884456153556643,
+      "grad_norm": 0.17499446868896484,
+      "learning_rate": 1.8536865654026393e-05,
+      "loss": 0.2157,
+      "step": 6489
+    },
+    {
+      "epoch": 4.885208882197968,
+      "grad_norm": 0.1923496425151825,
+      "learning_rate": 1.853004140097474e-05,
+      "loss": 0.2294,
+      "step": 6490
+    },
+    {
+      "epoch": 4.885961610839292,
+      "grad_norm": 0.18841944634914398,
+      "learning_rate": 1.8523218128864277e-05,
+      "loss": 0.1847,
+      "step": 6491
+    },
+    {
+      "epoch": 4.886714339480617,
+      "grad_norm": 0.21714362502098083,
+      "learning_rate": 1.8516395838441185e-05,
+      "loss": 0.2111,
+      "step": 6492
+    },
+    {
+      "epoch": 4.887467068121942,
+      "grad_norm": 0.22760000824928284,
+      "learning_rate": 1.8509574530451536e-05,
+      "loss": 0.2316,
+      "step": 6493
+    },
+    {
+      "epoch": 4.888219796763266,
+      "grad_norm": 0.19565503299236298,
+      "learning_rate": 1.8502754205641303e-05,
+      "loss": 0.2174,
+      "step": 6494
+    },
+    {
+      "epoch": 4.888972525404592,
+      "grad_norm": 0.20667383074760437,
+      "learning_rate": 1.8495934864756332e-05,
+      "loss": 0.2146,
+      "step": 6495
+    },
+    {
+      "epoch": 4.889725254045916,
+      "grad_norm": 0.1690371334552765,
+      "learning_rate": 1.84891165085424e-05,
+      "loss": 0.2005,
+      "step": 6496
+    },
+    {
+      "epoch": 4.8904779826872415,
+      "grad_norm": 0.20448842644691467,
+      "learning_rate": 1.848229913774513e-05,
+      "loss": 0.1716,
+      "step": 6497
+    },
+    {
+      "epoch": 4.891230711328566,
+      "grad_norm": 0.20748160779476166,
+      "learning_rate": 1.8475482753110075e-05,
+      "loss": 0.2419,
+      "step": 6498
+    },
+    {
+      "epoch": 4.8919834399698905,
+      "grad_norm": 0.24446333944797516,
+      "learning_rate": 1.8468667355382652e-05,
+      "loss": 0.1756,
+      "step": 6499
+    },
+    {
+      "epoch": 4.892736168611216,
+      "grad_norm": 0.2857748568058014,
+      "learning_rate": 1.8461852945308196e-05,
+      "loss": 0.2393,
+      "step": 6500
+    },
+    {
+      "epoch": 4.89348889725254,
+      "grad_norm": 0.2661631107330322,
+      "learning_rate": 1.8455039523631908e-05,
+      "loss": 0.2093,
+      "step": 6501
+    },
+    {
+      "epoch": 4.894241625893866,
+      "grad_norm": 0.22580987215042114,
+      "learning_rate": 1.844822709109889e-05,
+      "loss": 0.1322,
+      "step": 6502
+    },
+    {
+      "epoch": 4.89499435453519,
+      "grad_norm": 0.22201409935951233,
+      "learning_rate": 1.8441415648454157e-05,
+      "loss": 0.2294,
+      "step": 6503
+    },
+    {
+      "epoch": 4.895747083176515,
+      "grad_norm": 0.20747095346450806,
+      "learning_rate": 1.8434605196442595e-05,
+      "loss": 0.1451,
+      "step": 6504
+    },
+    {
+      "epoch": 4.89649981181784,
+      "grad_norm": 0.18907959759235382,
+      "learning_rate": 1.8427795735808973e-05,
+      "loss": 0.1629,
+      "step": 6505
+    },
+    {
+      "epoch": 4.8972525404591645,
+      "grad_norm": 0.19591231644153595,
+      "learning_rate": 1.8420987267297973e-05,
+      "loss": 0.2009,
+      "step": 6506
+    },
+    {
+      "epoch": 4.898005269100489,
+      "grad_norm": 0.2007550299167633,
+      "learning_rate": 1.841417979165415e-05,
+      "loss": 0.2578,
+      "step": 6507
+    },
+    {
+      "epoch": 4.898757997741814,
+      "grad_norm": 0.21527442336082458,
+      "learning_rate": 1.8407373309621963e-05,
+      "loss": 0.1584,
+      "step": 6508
+    },
+    {
+      "epoch": 4.899510726383139,
+      "grad_norm": 0.1833324432373047,
+      "learning_rate": 1.8400567821945764e-05,
+      "loss": 0.1693,
+      "step": 6509
+    },
+    {
+      "epoch": 4.900263455024463,
+      "grad_norm": 0.22072671353816986,
+      "learning_rate": 1.839376332936979e-05,
+      "loss": 0.2206,
+      "step": 6510
+    },
+    {
+      "epoch": 4.901016183665789,
+      "grad_norm": 0.22317183017730713,
+      "learning_rate": 1.838695983263816e-05,
+      "loss": 0.2068,
+      "step": 6511
+    },
+    {
+      "epoch": 4.901768912307113,
+      "grad_norm": 0.19224902987480164,
+      "learning_rate": 1.8380157332494912e-05,
+      "loss": 0.2829,
+      "step": 6512
+    },
+    {
+      "epoch": 4.9025216409484385,
+      "grad_norm": 0.18729344010353088,
+      "learning_rate": 1.837335582968394e-05,
+      "loss": 0.2255,
+      "step": 6513
+    },
+    {
+      "epoch": 4.903274369589763,
+      "grad_norm": 0.1888328641653061,
+      "learning_rate": 1.8366555324949054e-05,
+      "loss": 0.1816,
+      "step": 6514
+    },
+    {
+      "epoch": 4.904027098231087,
+      "grad_norm": 0.19789700210094452,
+      "learning_rate": 1.835975581903394e-05,
+      "loss": 0.2231,
+      "step": 6515
+    },
+    {
+      "epoch": 4.904779826872413,
+      "grad_norm": 0.22588826715946198,
+      "learning_rate": 1.8352957312682197e-05,
+      "loss": 0.2491,
+      "step": 6516
+    },
+    {
+      "epoch": 4.905532555513737,
+      "grad_norm": 0.1720208078622818,
+      "learning_rate": 1.8346159806637287e-05,
+      "loss": 0.1642,
+      "step": 6517
+    },
+    {
+      "epoch": 4.906285284155063,
+      "grad_norm": 0.1962442696094513,
+      "learning_rate": 1.8339363301642576e-05,
+      "loss": 0.1819,
+      "step": 6518
+    },
+    {
+      "epoch": 4.907038012796387,
+      "grad_norm": 0.2112351655960083,
+      "learning_rate": 1.8332567798441324e-05,
+      "loss": 0.2119,
+      "step": 6519
+    },
+    {
+      "epoch": 4.9077907414377115,
+      "grad_norm": 0.1951994150876999,
+      "learning_rate": 1.832577329777668e-05,
+      "loss": 0.2151,
+      "step": 6520
+    },
+    {
+      "epoch": 4.908543470079037,
+      "grad_norm": 0.18095554411411285,
+      "learning_rate": 1.8318979800391655e-05,
+      "loss": 0.15,
+      "step": 6521
+    },
+    {
+      "epoch": 4.909296198720361,
+      "grad_norm": 0.2097228318452835,
+      "learning_rate": 1.831218730702921e-05,
+      "loss": 0.2936,
+      "step": 6522
+    },
+    {
+      "epoch": 4.910048927361686,
+      "grad_norm": 0.22125661373138428,
+      "learning_rate": 1.8305395818432154e-05,
+      "loss": 0.1952,
+      "step": 6523
+    },
+    {
+      "epoch": 4.910801656003011,
+      "grad_norm": 0.17821963131427765,
+      "learning_rate": 1.8298605335343178e-05,
+      "loss": 0.2206,
+      "step": 6524
+    },
+    {
+      "epoch": 4.911554384644336,
+      "grad_norm": 0.20298565924167633,
+      "learning_rate": 1.829181585850489e-05,
+      "loss": 0.0882,
+      "step": 6525
+    },
+    {
+      "epoch": 4.91230711328566,
+      "grad_norm": 0.21636945009231567,
+      "learning_rate": 1.8285027388659776e-05,
+      "loss": 0.2241,
+      "step": 6526
+    },
+    {
+      "epoch": 4.9130598419269855,
+      "grad_norm": 0.19619950652122498,
+      "learning_rate": 1.8278239926550205e-05,
+      "loss": 0.277,
+      "step": 6527
+    },
+    {
+      "epoch": 4.91381257056831,
+      "grad_norm": 0.1749694049358368,
+      "learning_rate": 1.8271453472918458e-05,
+      "loss": 0.2327,
+      "step": 6528
+    },
+    {
+      "epoch": 4.9145652992096345,
+      "grad_norm": 0.21237006783485413,
+      "learning_rate": 1.826466802850668e-05,
+      "loss": 0.1795,
+      "step": 6529
+    },
+    {
+      "epoch": 4.91531802785096,
+      "grad_norm": 0.21213600039482117,
+      "learning_rate": 1.8257883594056936e-05,
+      "loss": 0.2354,
+      "step": 6530
+    },
+    {
+      "epoch": 4.916070756492284,
+      "grad_norm": 0.19375693798065186,
+      "learning_rate": 1.825110017031113e-05,
+      "loss": 0.2426,
+      "step": 6531
+    },
+    {
+      "epoch": 4.91682348513361,
+      "grad_norm": 0.22004692256450653,
+      "learning_rate": 1.824431775801111e-05,
+      "loss": 0.2039,
+      "step": 6532
+    },
+    {
+      "epoch": 4.917576213774934,
+      "grad_norm": 0.18662749230861664,
+      "learning_rate": 1.8237536357898584e-05,
+      "loss": 0.2605,
+      "step": 6533
+    },
+    {
+      "epoch": 4.918328942416259,
+      "grad_norm": 0.18646499514579773,
+      "learning_rate": 1.8230755970715152e-05,
+      "loss": 0.1748,
+      "step": 6534
+    },
+    {
+      "epoch": 4.919081671057584,
+      "grad_norm": 0.16515617072582245,
+      "learning_rate": 1.822397659720232e-05,
+      "loss": 0.1133,
+      "step": 6535
+    },
+    {
+      "epoch": 4.9198343996989085,
+      "grad_norm": 0.18442481756210327,
+      "learning_rate": 1.8217198238101453e-05,
+      "loss": 0.1821,
+      "step": 6536
+    },
+    {
+      "epoch": 4.920587128340234,
+      "grad_norm": 0.1764274686574936,
+      "learning_rate": 1.8210420894153833e-05,
+      "loss": 0.106,
+      "step": 6537
+    },
+    {
+      "epoch": 4.921339856981558,
+      "grad_norm": 0.23018525540828705,
+      "learning_rate": 1.820364456610062e-05,
+      "loss": 0.1937,
+      "step": 6538
+    },
+    {
+      "epoch": 4.922092585622883,
+      "grad_norm": 0.1999259889125824,
+      "learning_rate": 1.819686925468286e-05,
+      "loss": 0.1493,
+      "step": 6539
+    },
+    {
+      "epoch": 4.922845314264208,
+      "grad_norm": 0.18804073333740234,
+      "learning_rate": 1.819009496064148e-05,
+      "loss": 0.2042,
+      "step": 6540
+    },
+    {
+      "epoch": 4.923598042905533,
+      "grad_norm": 0.2014806866645813,
+      "learning_rate": 1.818332168471733e-05,
+      "loss": 0.1962,
+      "step": 6541
+    },
+    {
+      "epoch": 4.924350771546857,
+      "grad_norm": 0.20380628108978271,
+      "learning_rate": 1.8176549427651108e-05,
+      "loss": 0.2325,
+      "step": 6542
+    },
+    {
+      "epoch": 4.925103500188182,
+      "grad_norm": 0.20260968804359436,
+      "learning_rate": 1.816977819018343e-05,
+      "loss": 0.1972,
+      "step": 6543
+    },
+    {
+      "epoch": 4.925856228829507,
+      "grad_norm": 0.22906745970249176,
+      "learning_rate": 1.8163007973054775e-05,
+      "loss": 0.2329,
+      "step": 6544
+    },
+    {
+      "epoch": 4.926608957470831,
+      "grad_norm": 0.2078077644109726,
+      "learning_rate": 1.8156238777005536e-05,
+      "loss": 0.2585,
+      "step": 6545
+    },
+    {
+      "epoch": 4.927361686112157,
+      "grad_norm": 0.21100033819675446,
+      "learning_rate": 1.8149470602775958e-05,
+      "loss": 0.1993,
+      "step": 6546
+    },
+    {
+      "epoch": 4.928114414753481,
+      "grad_norm": 0.20546284317970276,
+      "learning_rate": 1.8142703451106234e-05,
+      "loss": 0.1676,
+      "step": 6547
+    },
+    {
+      "epoch": 4.928867143394806,
+      "grad_norm": 0.23626482486724854,
+      "learning_rate": 1.813593732273639e-05,
+      "loss": 0.2337,
+      "step": 6548
+    },
+    {
+      "epoch": 4.929619872036131,
+      "grad_norm": 0.2095954567193985,
+      "learning_rate": 1.812917221840636e-05,
+      "loss": 0.259,
+      "step": 6549
+    },
+    {
+      "epoch": 4.9303726006774555,
+      "grad_norm": 0.22191841900348663,
+      "learning_rate": 1.8122408138855974e-05,
+      "loss": 0.2083,
+      "step": 6550
+    },
+    {
+      "epoch": 4.931125329318781,
+      "grad_norm": 0.19991865754127502,
+      "learning_rate": 1.811564508482492e-05,
+      "loss": 0.1957,
+      "step": 6551
+    },
+    {
+      "epoch": 4.931878057960105,
+      "grad_norm": 0.18539373576641083,
+      "learning_rate": 1.8108883057052828e-05,
+      "loss": 0.1064,
+      "step": 6552
+    },
+    {
+      "epoch": 4.93263078660143,
+      "grad_norm": 0.1993197798728943,
+      "learning_rate": 1.810212205627914e-05,
+      "loss": 0.1333,
+      "step": 6553
+    },
+    {
+      "epoch": 4.933383515242755,
+      "grad_norm": 0.2058860957622528,
+      "learning_rate": 1.8095362083243283e-05,
+      "loss": 0.225,
+      "step": 6554
+    },
+    {
+      "epoch": 4.93413624388408,
+      "grad_norm": 0.182353213429451,
+      "learning_rate": 1.8088603138684474e-05,
+      "loss": 0.2068,
+      "step": 6555
+    },
+    {
+      "epoch": 4.934888972525405,
+      "grad_norm": 0.20860905945301056,
+      "learning_rate": 1.808184522334188e-05,
+      "loss": 0.2533,
+      "step": 6556
+    },
+    {
+      "epoch": 4.9356417011667295,
+      "grad_norm": 0.22957533597946167,
+      "learning_rate": 1.8075088337954528e-05,
+      "loss": 0.2141,
+      "step": 6557
+    },
+    {
+      "epoch": 4.936394429808054,
+      "grad_norm": 0.22387227416038513,
+      "learning_rate": 1.806833248326136e-05,
+      "loss": 0.1965,
+      "step": 6558
+    },
+    {
+      "epoch": 4.937147158449379,
+      "grad_norm": 0.18591105937957764,
+      "learning_rate": 1.8061577660001144e-05,
+      "loss": 0.1351,
+      "step": 6559
+    },
+    {
+      "epoch": 4.937899887090704,
+      "grad_norm": 0.2008380889892578,
+      "learning_rate": 1.8054823868912623e-05,
+      "loss": 0.1746,
+      "step": 6560
+    },
+    {
+      "epoch": 4.938652615732028,
+      "grad_norm": 0.19452694058418274,
+      "learning_rate": 1.804807111073436e-05,
+      "loss": 0.2685,
+      "step": 6561
+    },
+    {
+      "epoch": 4.939405344373354,
+      "grad_norm": 0.21529272198677063,
+      "learning_rate": 1.8041319386204836e-05,
+      "loss": 0.1559,
+      "step": 6562
+    },
+    {
+      "epoch": 4.940158073014678,
+      "grad_norm": 0.2017052322626114,
+      "learning_rate": 1.8034568696062393e-05,
+      "loss": 0.1549,
+      "step": 6563
+    },
+    {
+      "epoch": 4.940910801656003,
+      "grad_norm": 0.19438646733760834,
+      "learning_rate": 1.8027819041045295e-05,
+      "loss": 0.1961,
+      "step": 6564
+    },
+    {
+      "epoch": 4.941663530297328,
+      "grad_norm": 0.20900483429431915,
+      "learning_rate": 1.802107042189165e-05,
+      "loss": 0.2182,
+      "step": 6565
+    },
+    {
+      "epoch": 4.942416258938652,
+      "grad_norm": 0.1925707906484604,
+      "learning_rate": 1.8014322839339508e-05,
+      "loss": 0.1769,
+      "step": 6566
+    },
+    {
+      "epoch": 4.943168987579978,
+      "grad_norm": 0.1710643470287323,
+      "learning_rate": 1.8007576294126747e-05,
+      "loss": 0.1776,
+      "step": 6567
+    },
+    {
+      "epoch": 4.943921716221302,
+      "grad_norm": 0.19990918040275574,
+      "learning_rate": 1.8000830786991184e-05,
+      "loss": 0.1592,
+      "step": 6568
+    },
+    {
+      "epoch": 4.944674444862627,
+      "grad_norm": 0.2171078622341156,
+      "learning_rate": 1.7994086318670477e-05,
+      "loss": 0.1853,
+      "step": 6569
+    },
+    {
+      "epoch": 4.945427173503952,
+      "grad_norm": 0.23853832483291626,
+      "learning_rate": 1.7987342889902197e-05,
+      "loss": 0.2397,
+      "step": 6570
+    },
+    {
+      "epoch": 4.946179902145277,
+      "grad_norm": 0.19218212366104126,
+      "learning_rate": 1.7980600501423805e-05,
+      "loss": 0.2612,
+      "step": 6571
+    },
+    {
+      "epoch": 4.946932630786601,
+      "grad_norm": 0.19949185848236084,
+      "learning_rate": 1.797385915397262e-05,
+      "loss": 0.1745,
+      "step": 6572
+    },
+    {
+      "epoch": 4.947685359427926,
+      "grad_norm": 0.2086709588766098,
+      "learning_rate": 1.7967118848285874e-05,
+      "loss": 0.1992,
+      "step": 6573
+    },
+    {
+      "epoch": 4.948438088069251,
+      "grad_norm": 0.20717747509479523,
+      "learning_rate": 1.7960379585100684e-05,
+      "loss": 0.1954,
+      "step": 6574
+    },
+    {
+      "epoch": 4.949190816710576,
+      "grad_norm": 0.24254077672958374,
+      "learning_rate": 1.7953641365154046e-05,
+      "loss": 0.2444,
+      "step": 6575
+    },
+    {
+      "epoch": 4.949943545351901,
+      "grad_norm": 0.19865496456623077,
+      "learning_rate": 1.7946904189182824e-05,
+      "loss": 0.1865,
+      "step": 6576
+    },
+    {
+      "epoch": 4.950696273993225,
+      "grad_norm": 0.1913125067949295,
+      "learning_rate": 1.7940168057923813e-05,
+      "loss": 0.1616,
+      "step": 6577
+    },
+    {
+      "epoch": 4.951449002634551,
+      "grad_norm": 0.17987968027591705,
+      "learning_rate": 1.7933432972113623e-05,
+      "loss": 0.1812,
+      "step": 6578
+    },
+    {
+      "epoch": 4.952201731275875,
+      "grad_norm": 0.18706554174423218,
+      "learning_rate": 1.7926698932488844e-05,
+      "loss": 0.1203,
+      "step": 6579
+    },
+    {
+      "epoch": 4.9529544599171995,
+      "grad_norm": 0.1818874031305313,
+      "learning_rate": 1.7919965939785866e-05,
+      "loss": 0.2689,
+      "step": 6580
+    },
+    {
+      "epoch": 4.953707188558525,
+      "grad_norm": 0.20072788000106812,
+      "learning_rate": 1.791323399474102e-05,
+      "loss": 0.2261,
+      "step": 6581
+    },
+    {
+      "epoch": 4.954459917199849,
+      "grad_norm": 0.1929055154323578,
+      "learning_rate": 1.7906503098090478e-05,
+      "loss": 0.1308,
+      "step": 6582
+    },
+    {
+      "epoch": 4.955212645841174,
+      "grad_norm": 0.20149315893650055,
+      "learning_rate": 1.789977325057034e-05,
+      "loss": 0.1677,
+      "step": 6583
+    },
+    {
+      "epoch": 4.955965374482499,
+      "grad_norm": 0.21338623762130737,
+      "learning_rate": 1.7893044452916557e-05,
+      "loss": 0.2917,
+      "step": 6584
+    },
+    {
+      "epoch": 4.956718103123824,
+      "grad_norm": 0.18467172980308533,
+      "learning_rate": 1.7886316705865002e-05,
+      "loss": 0.1348,
+      "step": 6585
+    },
+    {
+      "epoch": 4.957470831765149,
+      "grad_norm": 0.19360263645648956,
+      "learning_rate": 1.787959001015139e-05,
+      "loss": 0.1977,
+      "step": 6586
+    },
+    {
+      "epoch": 4.9582235604064735,
+      "grad_norm": 0.19502173364162445,
+      "learning_rate": 1.7872864366511354e-05,
+      "loss": 0.2271,
+      "step": 6587
+    },
+    {
+      "epoch": 4.958976289047798,
+      "grad_norm": 0.1940426528453827,
+      "learning_rate": 1.7866139775680398e-05,
+      "loss": 0.2102,
+      "step": 6588
+    },
+    {
+      "epoch": 4.959729017689123,
+      "grad_norm": 0.22058908641338348,
+      "learning_rate": 1.785941623839391e-05,
+      "loss": 0.1653,
+      "step": 6589
+    },
+    {
+      "epoch": 4.960481746330448,
+      "grad_norm": 0.194154754281044,
+      "learning_rate": 1.7852693755387167e-05,
+      "loss": 0.2424,
+      "step": 6590
+    },
+    {
+      "epoch": 4.961234474971773,
+      "grad_norm": 0.2007720172405243,
+      "learning_rate": 1.7845972327395326e-05,
+      "loss": 0.1279,
+      "step": 6591
+    },
+    {
+      "epoch": 4.961987203613098,
+      "grad_norm": 0.2276049554347992,
+      "learning_rate": 1.7839251955153437e-05,
+      "loss": 0.2274,
+      "step": 6592
+    },
+    {
+      "epoch": 4.962739932254422,
+      "grad_norm": 0.2017461061477661,
+      "learning_rate": 1.7832532639396437e-05,
+      "loss": 0.1658,
+      "step": 6593
+    },
+    {
+      "epoch": 4.9634926608957475,
+      "grad_norm": 0.2809276580810547,
+      "learning_rate": 1.7825814380859128e-05,
+      "loss": 0.184,
+      "step": 6594
+    },
+    {
+      "epoch": 4.964245389537072,
+      "grad_norm": 0.1871432065963745,
+      "learning_rate": 1.7819097180276213e-05,
+      "loss": 0.2029,
+      "step": 6595
+    },
+    {
+      "epoch": 4.964998118178396,
+      "grad_norm": 0.21483632922172546,
+      "learning_rate": 1.7812381038382285e-05,
+      "loss": 0.1398,
+      "step": 6596
+    },
+    {
+      "epoch": 4.965750846819722,
+      "grad_norm": 0.1841498613357544,
+      "learning_rate": 1.7805665955911788e-05,
+      "loss": 0.2351,
+      "step": 6597
+    },
+    {
+      "epoch": 4.966503575461046,
+      "grad_norm": 0.21394822001457214,
+      "learning_rate": 1.779895193359909e-05,
+      "loss": 0.1752,
+      "step": 6598
+    },
+    {
+      "epoch": 4.967256304102371,
+      "grad_norm": 0.1914028376340866,
+      "learning_rate": 1.7792238972178422e-05,
+      "loss": 0.2229,
+      "step": 6599
+    },
+    {
+      "epoch": 4.968009032743696,
+      "grad_norm": 0.2055130898952484,
+      "learning_rate": 1.778552707238391e-05,
+      "loss": 0.1805,
+      "step": 6600
+    },
+    {
+      "epoch": 4.968009032743696,
+      "eval_loss": 0.2229611724615097,
+      "eval_runtime": 452.6643,
+      "eval_samples_per_second": 21.267,
+      "eval_steps_per_second": 0.665,
+      "step": 6600
+    },
+    {
+      "epoch": 4.968761761385021,
+      "grad_norm": 0.19394853711128235,
+      "learning_rate": 1.7778816234949552e-05,
+      "loss": 0.237,
+      "step": 6601
+    },
+    {
+      "epoch": 4.969514490026345,
+      "grad_norm": 0.2187439203262329,
+      "learning_rate": 1.7772106460609238e-05,
+      "loss": 0.2856,
+      "step": 6602
+    },
+    {
+      "epoch": 4.97026721866767,
+      "grad_norm": 0.21517948806285858,
+      "learning_rate": 1.776539775009673e-05,
+      "loss": 0.1768,
+      "step": 6603
+    },
+    {
+      "epoch": 4.971019947308995,
+      "grad_norm": 0.2293052077293396,
+      "learning_rate": 1.775869010414568e-05,
+      "loss": 0.2141,
+      "step": 6604
+    },
+    {
+      "epoch": 4.97177267595032,
+      "grad_norm": 0.21352097392082214,
+      "learning_rate": 1.7751983523489647e-05,
+      "loss": 0.1811,
+      "step": 6605
+    },
+    {
+      "epoch": 4.972525404591645,
+      "grad_norm": 0.21981588006019592,
+      "learning_rate": 1.774527800886204e-05,
+      "loss": 0.2234,
+      "step": 6606
+    },
+    {
+      "epoch": 4.973278133232969,
+      "grad_norm": 0.202365905046463,
+      "learning_rate": 1.7738573560996158e-05,
+      "loss": 0.1941,
+      "step": 6607
+    },
+    {
+      "epoch": 4.9740308618742946,
+      "grad_norm": 0.2048071175813675,
+      "learning_rate": 1.77318701806252e-05,
+      "loss": 0.1162,
+      "step": 6608
+    },
+    {
+      "epoch": 4.974783590515619,
+      "grad_norm": 0.21259494125843048,
+      "learning_rate": 1.7725167868482228e-05,
+      "loss": 0.1582,
+      "step": 6609
+    },
+    {
+      "epoch": 4.975536319156944,
+      "grad_norm": 0.21921031177043915,
+      "learning_rate": 1.77184666253002e-05,
+      "loss": 0.1913,
+      "step": 6610
+    },
+    {
+      "epoch": 4.976289047798269,
+      "grad_norm": 0.20274853706359863,
+      "learning_rate": 1.7711766451811957e-05,
+      "loss": 0.2141,
+      "step": 6611
+    },
+    {
+      "epoch": 4.977041776439593,
+      "grad_norm": 0.20155437290668488,
+      "learning_rate": 1.7705067348750225e-05,
+      "loss": 0.2702,
+      "step": 6612
+    },
+    {
+      "epoch": 4.977794505080919,
+      "grad_norm": 0.205840066075325,
+      "learning_rate": 1.76983693168476e-05,
+      "loss": 0.1208,
+      "step": 6613
+    },
+    {
+      "epoch": 4.978547233722243,
+      "grad_norm": 0.1871134340763092,
+      "learning_rate": 1.769167235683657e-05,
+      "loss": 0.1468,
+      "step": 6614
+    },
+    {
+      "epoch": 4.979299962363568,
+      "grad_norm": 0.19755512475967407,
+      "learning_rate": 1.76849764694495e-05,
+      "loss": 0.2649,
+      "step": 6615
+    },
+    {
+      "epoch": 4.980052691004893,
+      "grad_norm": 0.20438143610954285,
+      "learning_rate": 1.7678281655418642e-05,
+      "loss": 0.1949,
+      "step": 6616
+    },
+    {
+      "epoch": 4.9808054196462175,
+      "grad_norm": 0.18388792872428894,
+      "learning_rate": 1.767158791547614e-05,
+      "loss": 0.1566,
+      "step": 6617
+    },
+    {
+      "epoch": 4.981558148287542,
+      "grad_norm": 0.18030817806720734,
+      "learning_rate": 1.766489525035402e-05,
+      "loss": 0.1691,
+      "step": 6618
+    },
+    {
+      "epoch": 4.982310876928867,
+      "grad_norm": 0.19440826773643494,
+      "learning_rate": 1.765820366078415e-05,
+      "loss": 0.1942,
+      "step": 6619
+    },
+    {
+      "epoch": 4.983063605570192,
+      "grad_norm": 0.2007468342781067,
+      "learning_rate": 1.7651513147498334e-05,
+      "loss": 0.184,
+      "step": 6620
+    },
+    {
+      "epoch": 4.983816334211516,
+      "grad_norm": 0.18418702483177185,
+      "learning_rate": 1.7644823711228243e-05,
+      "loss": 0.2795,
+      "step": 6621
+    },
+    {
+      "epoch": 4.984569062852842,
+      "grad_norm": 0.2033672034740448,
+      "learning_rate": 1.7638135352705402e-05,
+      "loss": 0.1563,
+      "step": 6622
+    },
+    {
+      "epoch": 4.985321791494166,
+      "grad_norm": 0.1785466969013214,
+      "learning_rate": 1.763144807266126e-05,
+      "loss": 0.1138,
+      "step": 6623
+    },
+    {
+      "epoch": 4.9860745201354915,
+      "grad_norm": 0.1995280534029007,
+      "learning_rate": 1.762476187182711e-05,
+      "loss": 0.231,
+      "step": 6624
+    },
+    {
+      "epoch": 4.986827248776816,
+      "grad_norm": 0.17116282880306244,
+      "learning_rate": 1.761807675093417e-05,
+      "loss": 0.1519,
+      "step": 6625
+    },
+    {
+      "epoch": 4.98757997741814,
+      "grad_norm": 0.1865827888250351,
+      "learning_rate": 1.7611392710713488e-05,
+      "loss": 0.1806,
+      "step": 6626
+    },
+    {
+      "epoch": 4.988332706059466,
+      "grad_norm": 0.19049888849258423,
+      "learning_rate": 1.7604709751896042e-05,
+      "loss": 0.139,
+      "step": 6627
+    },
+    {
+      "epoch": 4.98908543470079,
+      "grad_norm": 0.1768139898777008,
+      "learning_rate": 1.7598027875212653e-05,
+      "loss": 0.1532,
+      "step": 6628
+    },
+    {
+      "epoch": 4.989838163342116,
+      "grad_norm": 0.1891268491744995,
+      "learning_rate": 1.7591347081394043e-05,
+      "loss": 0.1711,
+      "step": 6629
+    },
+    {
+      "epoch": 4.99059089198344,
+      "grad_norm": 0.17284829914569855,
+      "learning_rate": 1.758466737117082e-05,
+      "loss": 0.2187,
+      "step": 6630
+    },
+    {
+      "epoch": 4.9913436206247646,
+      "grad_norm": 0.19628120958805084,
+      "learning_rate": 1.7577988745273476e-05,
+      "loss": 0.1614,
+      "step": 6631
+    },
+    {
+      "epoch": 4.99209634926609,
+      "grad_norm": 0.20196731388568878,
+      "learning_rate": 1.7571311204432356e-05,
+      "loss": 0.2375,
+      "step": 6632
+    },
+    {
+      "epoch": 4.992849077907414,
+      "grad_norm": 0.22954857349395752,
+      "learning_rate": 1.756463474937773e-05,
+      "loss": 0.2399,
+      "step": 6633
+    },
+    {
+      "epoch": 4.993601806548739,
+      "grad_norm": 0.18136467039585114,
+      "learning_rate": 1.755795938083969e-05,
+      "loss": 0.1863,
+      "step": 6634
+    },
+    {
+      "epoch": 4.994354535190064,
+      "grad_norm": 0.21668162941932678,
+      "learning_rate": 1.755128509954827e-05,
+      "loss": 0.2883,
+      "step": 6635
+    },
+    {
+      "epoch": 4.995107263831389,
+      "grad_norm": 0.18467579782009125,
+      "learning_rate": 1.7544611906233355e-05,
+      "loss": 0.2245,
+      "step": 6636
+    },
+    {
+      "epoch": 4.995859992472713,
+      "grad_norm": 0.19667762517929077,
+      "learning_rate": 1.753793980162472e-05,
+      "loss": 0.1333,
+      "step": 6637
+    },
+    {
+      "epoch": 4.9966127211140385,
+      "grad_norm": 0.1970939189195633,
+      "learning_rate": 1.7531268786452004e-05,
+      "loss": 0.162,
+      "step": 6638
+    },
+    {
+      "epoch": 4.997365449755363,
+      "grad_norm": 0.23568932712078094,
+      "learning_rate": 1.752459886144475e-05,
+      "loss": 0.2324,
+      "step": 6639
+    },
+    {
+      "epoch": 4.998118178396688,
+      "grad_norm": 0.18335792422294617,
+      "learning_rate": 1.751793002733237e-05,
+      "loss": 0.2114,
+      "step": 6640
+    },
+    {
+      "epoch": 4.998870907038013,
+      "grad_norm": 0.2503625750541687,
+      "learning_rate": 1.751126228484414e-05,
+      "loss": 0.2454,
+      "step": 6641
+    },
+    {
+      "epoch": 4.999623635679337,
+      "grad_norm": 0.1999194473028183,
+      "learning_rate": 1.750459563470926e-05,
+      "loss": 0.1642,
+      "step": 6642
+    },
+    {
+      "epoch": 5.000376364320663,
+      "grad_norm": 0.271030068397522,
+      "learning_rate": 1.7497930077656776e-05,
+      "loss": 0.2347,
+      "step": 6643
+    },
+    {
+      "epoch": 5.001129092961987,
+      "grad_norm": 0.2037947177886963,
+      "learning_rate": 1.7491265614415613e-05,
+      "loss": 0.2258,
+      "step": 6644
+    },
+    {
+      "epoch": 5.001881821603312,
+      "grad_norm": 0.23524735867977142,
+      "learning_rate": 1.7484602245714596e-05,
+      "loss": 0.2774,
+      "step": 6645
+    },
+    {
+      "epoch": 5.002634550244637,
+      "grad_norm": 0.21713675558567047,
+      "learning_rate": 1.7477939972282426e-05,
+      "loss": 0.1554,
+      "step": 6646
+    },
+    {
+      "epoch": 5.0033872788859615,
+      "grad_norm": 0.196938619017601,
+      "learning_rate": 1.7471278794847664e-05,
+      "loss": 0.138,
+      "step": 6647
+    },
+    {
+      "epoch": 5.004140007527287,
+      "grad_norm": 0.22003184258937836,
+      "learning_rate": 1.7464618714138778e-05,
+      "loss": 0.1708,
+      "step": 6648
+    },
+    {
+      "epoch": 5.004892736168611,
+      "grad_norm": 0.19674384593963623,
+      "learning_rate": 1.74579597308841e-05,
+      "loss": 0.1716,
+      "step": 6649
+    },
+    {
+      "epoch": 5.005645464809936,
+      "grad_norm": 0.22571425139904022,
+      "learning_rate": 1.7451301845811857e-05,
+      "loss": 0.1424,
+      "step": 6650
+    },
+    {
+      "epoch": 5.006398193451261,
+      "grad_norm": 0.21999472379684448,
+      "learning_rate": 1.7444645059650126e-05,
+      "loss": 0.1495,
+      "step": 6651
+    },
+    {
+      "epoch": 5.007150922092586,
+      "grad_norm": 0.1901761144399643,
+      "learning_rate": 1.7437989373126905e-05,
+      "loss": 0.1903,
+      "step": 6652
+    },
+    {
+      "epoch": 5.00790365073391,
+      "grad_norm": 0.22287186980247498,
+      "learning_rate": 1.7431334786970027e-05,
+      "loss": 0.1642,
+      "step": 6653
+    },
+    {
+      "epoch": 5.0086563793752354,
+      "grad_norm": 0.2124577909708023,
+      "learning_rate": 1.7424681301907236e-05,
+      "loss": 0.2359,
+      "step": 6654
+    },
+    {
+      "epoch": 5.00940910801656,
+      "grad_norm": 0.17077240347862244,
+      "learning_rate": 1.7418028918666152e-05,
+      "loss": 0.1443,
+      "step": 6655
+    },
+    {
+      "epoch": 5.010161836657884,
+      "grad_norm": 0.20455190539360046,
+      "learning_rate": 1.741137763797428e-05,
+      "loss": 0.2093,
+      "step": 6656
+    },
+    {
+      "epoch": 5.01091456529921,
+      "grad_norm": 0.1917990893125534,
+      "learning_rate": 1.7404727460558966e-05,
+      "loss": 0.1388,
+      "step": 6657
+    },
+    {
+      "epoch": 5.011667293940534,
+      "grad_norm": 0.17138344049453735,
+      "learning_rate": 1.7398078387147493e-05,
+      "loss": 0.1454,
+      "step": 6658
+    },
+    {
+      "epoch": 5.01242002258186,
+      "grad_norm": 0.17969642579555511,
+      "learning_rate": 1.739143041846697e-05,
+      "loss": 0.2533,
+      "step": 6659
+    },
+    {
+      "epoch": 5.013172751223184,
+      "grad_norm": 0.20435991883277893,
+      "learning_rate": 1.7384783555244425e-05,
+      "loss": 0.1895,
+      "step": 6660
+    },
+    {
+      "epoch": 5.0139254798645085,
+      "grad_norm": 0.23332732915878296,
+      "learning_rate": 1.737813779820673e-05,
+      "loss": 0.2362,
+      "step": 6661
+    },
+    {
+      "epoch": 5.014678208505834,
+      "grad_norm": 0.2060462385416031,
+      "learning_rate": 1.7371493148080677e-05,
+      "loss": 0.1741,
+      "step": 6662
+    },
+    {
+      "epoch": 5.015430937147158,
+      "grad_norm": 0.1734015941619873,
+      "learning_rate": 1.7364849605592905e-05,
+      "loss": 0.0906,
+      "step": 6663
+    },
+    {
+      "epoch": 5.016183665788483,
+      "grad_norm": 0.20328454673290253,
+      "learning_rate": 1.7358207171469943e-05,
+      "loss": 0.1732,
+      "step": 6664
+    },
+    {
+      "epoch": 5.016936394429808,
+      "grad_norm": 0.2010139524936676,
+      "learning_rate": 1.735156584643819e-05,
+      "loss": 0.1867,
+      "step": 6665
+    },
+    {
+      "epoch": 5.017689123071133,
+      "grad_norm": 0.1902051568031311,
+      "learning_rate": 1.734492563122394e-05,
+      "loss": 0.2311,
+      "step": 6666
+    },
+    {
+      "epoch": 5.018441851712458,
+      "grad_norm": 0.1879478543996811,
+      "learning_rate": 1.7338286526553354e-05,
+      "loss": 0.1857,
+      "step": 6667
+    },
+    {
+      "epoch": 5.0191945803537825,
+      "grad_norm": 0.22484146058559418,
+      "learning_rate": 1.7331648533152485e-05,
+      "loss": 0.1855,
+      "step": 6668
+    },
+    {
+      "epoch": 5.019947308995107,
+      "grad_norm": 0.20277363061904907,
+      "learning_rate": 1.7325011651747236e-05,
+      "loss": 0.1939,
+      "step": 6669
+    },
+    {
+      "epoch": 5.020700037636432,
+      "grad_norm": 0.1787460744380951,
+      "learning_rate": 1.7318375883063413e-05,
+      "loss": 0.1528,
+      "step": 6670
+    },
+    {
+      "epoch": 5.021452766277757,
+      "grad_norm": 0.21354426443576813,
+      "learning_rate": 1.731174122782671e-05,
+      "loss": 0.1815,
+      "step": 6671
+    },
+    {
+      "epoch": 5.022205494919081,
+      "grad_norm": 0.2000601589679718,
+      "learning_rate": 1.7305107686762656e-05,
+      "loss": 0.162,
+      "step": 6672
+    },
+    {
+      "epoch": 5.022958223560407,
+      "grad_norm": 0.2024368941783905,
+      "learning_rate": 1.7298475260596705e-05,
+      "loss": 0.1522,
+      "step": 6673
+    },
+    {
+      "epoch": 5.023710952201731,
+      "grad_norm": 0.1798451691865921,
+      "learning_rate": 1.7291843950054143e-05,
+      "loss": 0.2417,
+      "step": 6674
+    },
+    {
+      "epoch": 5.0244636808430565,
+      "grad_norm": 0.18418526649475098,
+      "learning_rate": 1.72852137558602e-05,
+      "loss": 0.1438,
+      "step": 6675
+    },
+    {
+      "epoch": 5.025216409484381,
+      "grad_norm": 0.21067869663238525,
+      "learning_rate": 1.7278584678739915e-05,
+      "loss": 0.1977,
+      "step": 6676
+    },
+    {
+      "epoch": 5.0259691381257054,
+      "grad_norm": 0.1697929948568344,
+      "learning_rate": 1.7271956719418248e-05,
+      "loss": 0.1407,
+      "step": 6677
+    },
+    {
+      "epoch": 5.026721866767031,
+      "grad_norm": 0.21399490535259247,
+      "learning_rate": 1.7265329878620007e-05,
+      "loss": 0.2352,
+      "step": 6678
+    },
+    {
+      "epoch": 5.027474595408355,
+      "grad_norm": 0.20930607616901398,
+      "learning_rate": 1.7258704157069913e-05,
+      "loss": 0.1224,
+      "step": 6679
+    },
+    {
+      "epoch": 5.02822732404968,
+      "grad_norm": 0.16824766993522644,
+      "learning_rate": 1.7252079555492517e-05,
+      "loss": 0.1711,
+      "step": 6680
+    },
+    {
+      "epoch": 5.028980052691005,
+      "grad_norm": 0.18907375633716583,
+      "learning_rate": 1.724545607461231e-05,
+      "loss": 0.1608,
+      "step": 6681
+    },
+    {
+      "epoch": 5.02973278133233,
+      "grad_norm": 0.17429661750793457,
+      "learning_rate": 1.7238833715153602e-05,
+      "loss": 0.1582,
+      "step": 6682
+    },
+    {
+      "epoch": 5.030485509973654,
+      "grad_norm": 0.21927517652511597,
+      "learning_rate": 1.723221247784062e-05,
+      "loss": 0.1899,
+      "step": 6683
+    },
+    {
+      "epoch": 5.031238238614979,
+      "grad_norm": 0.19815535843372345,
+      "learning_rate": 1.7225592363397437e-05,
+      "loss": 0.1957,
+      "step": 6684
+    },
+    {
+      "epoch": 5.031990967256304,
+      "grad_norm": 0.18643689155578613,
+      "learning_rate": 1.721897337254803e-05,
+      "loss": 0.231,
+      "step": 6685
+    },
+    {
+      "epoch": 5.032743695897629,
+      "grad_norm": 0.22198010981082916,
+      "learning_rate": 1.7212355506016224e-05,
+      "loss": 0.1995,
+      "step": 6686
+    },
+    {
+      "epoch": 5.033496424538954,
+      "grad_norm": 0.2229168862104416,
+      "learning_rate": 1.7205738764525766e-05,
+      "loss": 0.2321,
+      "step": 6687
+    },
+    {
+      "epoch": 5.034249153180278,
+      "grad_norm": 0.1659086048603058,
+      "learning_rate": 1.7199123148800235e-05,
+      "loss": 0.1343,
+      "step": 6688
+    },
+    {
+      "epoch": 5.035001881821604,
+      "grad_norm": 0.1906573325395584,
+      "learning_rate": 1.7192508659563116e-05,
+      "loss": 0.2071,
+      "step": 6689
+    },
+    {
+      "epoch": 5.035754610462928,
+      "grad_norm": 0.18024636805057526,
+      "learning_rate": 1.7185895297537745e-05,
+      "loss": 0.1356,
+      "step": 6690
+    },
+    {
+      "epoch": 5.0365073391042525,
+      "grad_norm": 0.22673757374286652,
+      "learning_rate": 1.717928306344736e-05,
+      "loss": 0.1635,
+      "step": 6691
+    },
+    {
+      "epoch": 5.037260067745578,
+      "grad_norm": 0.21093705296516418,
+      "learning_rate": 1.717267195801507e-05,
+      "loss": 0.2194,
+      "step": 6692
+    },
+    {
+      "epoch": 5.038012796386902,
+      "grad_norm": 0.17070768773555756,
+      "learning_rate": 1.7166061981963834e-05,
+      "loss": 0.1255,
+      "step": 6693
+    },
+    {
+      "epoch": 5.038765525028228,
+      "grad_norm": 0.23323853313922882,
+      "learning_rate": 1.7159453136016526e-05,
+      "loss": 0.2835,
+      "step": 6694
+    },
+    {
+      "epoch": 5.039518253669552,
+      "grad_norm": 0.19931234419345856,
+      "learning_rate": 1.7152845420895875e-05,
+      "loss": 0.2178,
+      "step": 6695
+    },
+    {
+      "epoch": 5.040270982310877,
+      "grad_norm": 0.17618152499198914,
+      "learning_rate": 1.7146238837324502e-05,
+      "loss": 0.1601,
+      "step": 6696
+    },
+    {
+      "epoch": 5.041023710952202,
+      "grad_norm": 0.18513914942741394,
+      "learning_rate": 1.7139633386024872e-05,
+      "loss": 0.1499,
+      "step": 6697
+    },
+    {
+      "epoch": 5.0417764395935265,
+      "grad_norm": 0.19447875022888184,
+      "learning_rate": 1.7133029067719372e-05,
+      "loss": 0.1986,
+      "step": 6698
+    },
+    {
+      "epoch": 5.042529168234851,
+      "grad_norm": 0.18965108692646027,
+      "learning_rate": 1.7126425883130206e-05,
+      "loss": 0.1291,
+      "step": 6699
+    },
+    {
+      "epoch": 5.043281896876176,
+      "grad_norm": 0.18983067572116852,
+      "learning_rate": 1.711982383297953e-05,
+      "loss": 0.1444,
+      "step": 6700
+    },
+    {
+      "epoch": 5.044034625517501,
+      "grad_norm": 0.20210714638233185,
+      "learning_rate": 1.7113222917989298e-05,
+      "loss": 0.24,
+      "step": 6701
+    },
+    {
+      "epoch": 5.044787354158826,
+      "grad_norm": 0.17344211041927338,
+      "learning_rate": 1.7106623138881402e-05,
+      "loss": 0.1724,
+      "step": 6702
+    },
+    {
+      "epoch": 5.045540082800151,
+      "grad_norm": 0.20511959493160248,
+      "learning_rate": 1.710002449637757e-05,
+      "loss": 0.1565,
+      "step": 6703
+    },
+    {
+      "epoch": 5.046292811441475,
+      "grad_norm": 0.19388031959533691,
+      "learning_rate": 1.7093426991199428e-05,
+      "loss": 0.1518,
+      "step": 6704
+    },
+    {
+      "epoch": 5.0470455400828005,
+      "grad_norm": 0.22723297774791718,
+      "learning_rate": 1.7086830624068443e-05,
+      "loss": 0.2734,
+      "step": 6705
+    },
+    {
+      "epoch": 5.047798268724125,
+      "grad_norm": 0.16583117842674255,
+      "learning_rate": 1.708023539570603e-05,
+      "loss": 0.2481,
+      "step": 6706
+    },
+    {
+      "epoch": 5.048550997365449,
+      "grad_norm": 0.236783966422081,
+      "learning_rate": 1.7073641306833393e-05,
+      "loss": 0.24,
+      "step": 6707
+    },
+    {
+      "epoch": 5.049303726006775,
+      "grad_norm": 0.18922841548919678,
+      "learning_rate": 1.7067048358171682e-05,
+      "loss": 0.2846,
+      "step": 6708
+    },
+    {
+      "epoch": 5.050056454648099,
+      "grad_norm": 0.2007347047328949,
+      "learning_rate": 1.706045655044186e-05,
+      "loss": 0.1437,
+      "step": 6709
+    },
+    {
+      "epoch": 5.050809183289424,
+      "grad_norm": 0.1983328014612198,
+      "learning_rate": 1.7053865884364827e-05,
+      "loss": 0.2136,
+      "step": 6710
+    },
+    {
+      "epoch": 5.051561911930749,
+      "grad_norm": 0.19802196323871613,
+      "learning_rate": 1.7047276360661306e-05,
+      "loss": 0.1826,
+      "step": 6711
+    },
+    {
+      "epoch": 5.052314640572074,
+      "grad_norm": 0.2142874002456665,
+      "learning_rate": 1.704068798005193e-05,
+      "loss": 0.2147,
+      "step": 6712
+    },
+    {
+      "epoch": 5.053067369213399,
+      "grad_norm": 0.18503819406032562,
+      "learning_rate": 1.7034100743257188e-05,
+      "loss": 0.2423,
+      "step": 6713
+    },
+    {
+      "epoch": 5.053820097854723,
+      "grad_norm": 0.20841075479984283,
+      "learning_rate": 1.7027514650997466e-05,
+      "loss": 0.1712,
+      "step": 6714
+    },
+    {
+      "epoch": 5.054572826496048,
+      "grad_norm": 0.1709815263748169,
+      "learning_rate": 1.7020929703992983e-05,
+      "loss": 0.143,
+      "step": 6715
+    },
+    {
+      "epoch": 5.055325555137373,
+      "grad_norm": 0.1898762285709381,
+      "learning_rate": 1.7014345902963875e-05,
+      "loss": 0.1582,
+      "step": 6716
+    },
+    {
+      "epoch": 5.056078283778698,
+      "grad_norm": 0.21825256943702698,
+      "learning_rate": 1.700776324863014e-05,
+      "loss": 0.29,
+      "step": 6717
+    },
+    {
+      "epoch": 5.056831012420022,
+      "grad_norm": 0.18703009188175201,
+      "learning_rate": 1.7001181741711624e-05,
+      "loss": 0.2236,
+      "step": 6718
+    },
+    {
+      "epoch": 5.057583741061348,
+      "grad_norm": 0.2101692110300064,
+      "learning_rate": 1.699460138292811e-05,
+      "loss": 0.2314,
+      "step": 6719
+    },
+    {
+      "epoch": 5.058336469702672,
+      "grad_norm": 0.1984216719865799,
+      "learning_rate": 1.6988022172999182e-05,
+      "loss": 0.1906,
+      "step": 6720
+    },
+    {
+      "epoch": 5.059089198343997,
+      "grad_norm": 0.21347782015800476,
+      "learning_rate": 1.6981444112644354e-05,
+      "loss": 0.2391,
+      "step": 6721
+    },
+    {
+      "epoch": 5.059841926985322,
+      "grad_norm": 0.18705488741397858,
+      "learning_rate": 1.697486720258297e-05,
+      "loss": 0.2166,
+      "step": 6722
+    },
+    {
+      "epoch": 5.060594655626646,
+      "grad_norm": 0.18967269361019135,
+      "learning_rate": 1.6968291443534302e-05,
+      "loss": 0.1917,
+      "step": 6723
+    },
+    {
+      "epoch": 5.061347384267972,
+      "grad_norm": 0.16188590228557587,
+      "learning_rate": 1.6961716836217435e-05,
+      "loss": 0.2272,
+      "step": 6724
+    },
+    {
+      "epoch": 5.062100112909296,
+      "grad_norm": 0.20403195917606354,
+      "learning_rate": 1.695514338135137e-05,
+      "loss": 0.1615,
+      "step": 6725
+    },
+    {
+      "epoch": 5.062852841550621,
+      "grad_norm": 0.24464626610279083,
+      "learning_rate": 1.6948571079654975e-05,
+      "loss": 0.1652,
+      "step": 6726
+    },
+    {
+      "epoch": 5.063605570191946,
+      "grad_norm": 0.20671530067920685,
+      "learning_rate": 1.6941999931846985e-05,
+      "loss": 0.2412,
+      "step": 6727
+    },
+    {
+      "epoch": 5.0643582988332705,
+      "grad_norm": 0.1934547871351242,
+      "learning_rate": 1.6935429938646e-05,
+      "loss": 0.2161,
+      "step": 6728
+    },
+    {
+      "epoch": 5.065111027474595,
+      "grad_norm": 0.20002008974552155,
+      "learning_rate": 1.692886110077053e-05,
+      "loss": 0.1057,
+      "step": 6729
+    },
+    {
+      "epoch": 5.06586375611592,
+      "grad_norm": 0.2022731900215149,
+      "learning_rate": 1.69222934189389e-05,
+      "loss": 0.2375,
+      "step": 6730
+    },
+    {
+      "epoch": 5.066616484757245,
+      "grad_norm": 0.22117839753627777,
+      "learning_rate": 1.6915726893869362e-05,
+      "loss": 0.179,
+      "step": 6731
+    },
+    {
+      "epoch": 5.06736921339857,
+      "grad_norm": 0.21579240262508392,
+      "learning_rate": 1.690916152628002e-05,
+      "loss": 0.1939,
+      "step": 6732
+    },
+    {
+      "epoch": 5.068121942039895,
+      "grad_norm": 0.17947258055210114,
+      "learning_rate": 1.690259731688885e-05,
+      "loss": 0.1432,
+      "step": 6733
+    },
+    {
+      "epoch": 5.068874670681219,
+      "grad_norm": 0.2079716920852661,
+      "learning_rate": 1.68960342664137e-05,
+      "loss": 0.151,
+      "step": 6734
+    },
+    {
+      "epoch": 5.0696273993225445,
+      "grad_norm": 0.22548240423202515,
+      "learning_rate": 1.6889472375572313e-05,
+      "loss": 0.1736,
+      "step": 6735
+    },
+    {
+      "epoch": 5.070380127963869,
+      "grad_norm": 0.2280488908290863,
+      "learning_rate": 1.6882911645082265e-05,
+      "loss": 0.16,
+      "step": 6736
+    },
+    {
+      "epoch": 5.071132856605193,
+      "grad_norm": 0.20302584767341614,
+      "learning_rate": 1.6876352075661037e-05,
+      "loss": 0.2463,
+      "step": 6737
+    },
+    {
+      "epoch": 5.071885585246519,
+      "grad_norm": 0.2034115493297577,
+      "learning_rate": 1.6869793668025968e-05,
+      "loss": 0.2325,
+      "step": 6738
+    },
+    {
+      "epoch": 5.072638313887843,
+      "grad_norm": 0.19083893299102783,
+      "learning_rate": 1.6863236422894287e-05,
+      "loss": 0.2199,
+      "step": 6739
+    },
+    {
+      "epoch": 5.073391042529169,
+      "grad_norm": 0.21028605103492737,
+      "learning_rate": 1.6856680340983084e-05,
+      "loss": 0.17,
+      "step": 6740
+    },
+    {
+      "epoch": 5.074143771170493,
+      "grad_norm": 0.212730273604393,
+      "learning_rate": 1.685012542300931e-05,
+      "loss": 0.2025,
+      "step": 6741
+    },
+    {
+      "epoch": 5.074896499811818,
+      "grad_norm": 0.21810244023799896,
+      "learning_rate": 1.6843571669689818e-05,
+      "loss": 0.1826,
+      "step": 6742
+    },
+    {
+      "epoch": 5.075649228453143,
+      "grad_norm": 0.20811109244823456,
+      "learning_rate": 1.6837019081741292e-05,
+      "loss": 0.2317,
+      "step": 6743
+    },
+    {
+      "epoch": 5.076401957094467,
+      "grad_norm": 0.20984645187854767,
+      "learning_rate": 1.683046765988033e-05,
+      "loss": 0.2272,
+      "step": 6744
+    },
+    {
+      "epoch": 5.077154685735792,
+      "grad_norm": 0.1786101907491684,
+      "learning_rate": 1.6823917404823385e-05,
+      "loss": 0.159,
+      "step": 6745
+    },
+    {
+      "epoch": 5.077907414377117,
+      "grad_norm": 0.19986025989055634,
+      "learning_rate": 1.681736831728679e-05,
+      "loss": 0.1607,
+      "step": 6746
+    },
+    {
+      "epoch": 5.078660143018442,
+      "grad_norm": 0.19716396927833557,
+      "learning_rate": 1.6810820397986728e-05,
+      "loss": 0.1843,
+      "step": 6747
+    },
+    {
+      "epoch": 5.079412871659767,
+      "grad_norm": 0.1917249858379364,
+      "learning_rate": 1.680427364763928e-05,
+      "loss": 0.1509,
+      "step": 6748
+    },
+    {
+      "epoch": 5.0801656003010915,
+      "grad_norm": 0.18186752498149872,
+      "learning_rate": 1.6797728066960377e-05,
+      "loss": 0.1953,
+      "step": 6749
+    },
+    {
+      "epoch": 5.080918328942416,
+      "grad_norm": 0.19396059215068817,
+      "learning_rate": 1.6791183656665846e-05,
+      "loss": 0.1203,
+      "step": 6750
+    },
+    {
+      "epoch": 5.081671057583741,
+      "grad_norm": 0.18450096249580383,
+      "learning_rate": 1.678464041747137e-05,
+      "loss": 0.2603,
+      "step": 6751
+    },
+    {
+      "epoch": 5.082423786225066,
+      "grad_norm": 0.19342775642871857,
+      "learning_rate": 1.6778098350092513e-05,
+      "loss": 0.1824,
+      "step": 6752
+    },
+    {
+      "epoch": 5.08317651486639,
+      "grad_norm": 0.19598476588726044,
+      "learning_rate": 1.677155745524469e-05,
+      "loss": 0.2302,
+      "step": 6753
+    },
+    {
+      "epoch": 5.083929243507716,
+      "grad_norm": 0.2129865288734436,
+      "learning_rate": 1.6765017733643228e-05,
+      "loss": 0.1694,
+      "step": 6754
+    },
+    {
+      "epoch": 5.08468197214904,
+      "grad_norm": 0.20610225200653076,
+      "learning_rate": 1.6758479186003273e-05,
+      "loss": 0.0972,
+      "step": 6755
+    },
+    {
+      "epoch": 5.085434700790365,
+      "grad_norm": 0.2017820179462433,
+      "learning_rate": 1.6751941813039886e-05,
+      "loss": 0.1378,
+      "step": 6756
+    },
+    {
+      "epoch": 5.08618742943169,
+      "grad_norm": 0.15169672667980194,
+      "learning_rate": 1.6745405615467988e-05,
+      "loss": 0.1379,
+      "step": 6757
+    },
+    {
+      "epoch": 5.0869401580730145,
+      "grad_norm": 0.23045602440834045,
+      "learning_rate": 1.6738870594002364e-05,
+      "loss": 0.1547,
+      "step": 6758
+    },
+    {
+      "epoch": 5.08769288671434,
+      "grad_norm": 0.17851261794567108,
+      "learning_rate": 1.673233674935767e-05,
+      "loss": 0.2187,
+      "step": 6759
+    },
+    {
+      "epoch": 5.088445615355664,
+      "grad_norm": 0.21582184731960297,
+      "learning_rate": 1.6725804082248437e-05,
+      "loss": 0.1376,
+      "step": 6760
+    },
+    {
+      "epoch": 5.089198343996989,
+      "grad_norm": 0.19784756004810333,
+      "learning_rate": 1.6719272593389078e-05,
+      "loss": 0.164,
+      "step": 6761
+    },
+    {
+      "epoch": 5.089951072638314,
+      "grad_norm": 0.2301049381494522,
+      "learning_rate": 1.6712742283493855e-05,
+      "loss": 0.2341,
+      "step": 6762
+    },
+    {
+      "epoch": 5.090703801279639,
+      "grad_norm": 0.20008450746536255,
+      "learning_rate": 1.6706213153276915e-05,
+      "loss": 0.2118,
+      "step": 6763
+    },
+    {
+      "epoch": 5.091456529920963,
+      "grad_norm": 0.22852583229541779,
+      "learning_rate": 1.6699685203452277e-05,
+      "loss": 0.2227,
+      "step": 6764
+    },
+    {
+      "epoch": 5.0922092585622885,
+      "grad_norm": 0.18467943370342255,
+      "learning_rate": 1.6693158434733832e-05,
+      "loss": 0.1242,
+      "step": 6765
+    },
+    {
+      "epoch": 5.092961987203613,
+      "grad_norm": 0.21075940132141113,
+      "learning_rate": 1.6686632847835333e-05,
+      "loss": 0.1891,
+      "step": 6766
+    },
+    {
+      "epoch": 5.093714715844938,
+      "grad_norm": 0.1920529305934906,
+      "learning_rate": 1.668010844347041e-05,
+      "loss": 0.1264,
+      "step": 6767
+    },
+    {
+      "epoch": 5.094467444486263,
+      "grad_norm": 0.21770749986171722,
+      "learning_rate": 1.667358522235255e-05,
+      "loss": 0.1923,
+      "step": 6768
+    },
+    {
+      "epoch": 5.095220173127587,
+      "grad_norm": 0.17573513090610504,
+      "learning_rate": 1.6667063185195135e-05,
+      "loss": 0.1763,
+      "step": 6769
+    },
+    {
+      "epoch": 5.095972901768913,
+      "grad_norm": 0.20477767288684845,
+      "learning_rate": 1.6660542332711403e-05,
+      "loss": 0.1453,
+      "step": 6770
+    },
+    {
+      "epoch": 5.096725630410237,
+      "grad_norm": 0.20830091834068298,
+      "learning_rate": 1.6654022665614475e-05,
+      "loss": 0.1713,
+      "step": 6771
+    },
+    {
+      "epoch": 5.0974783590515615,
+      "grad_norm": 0.19644097983837128,
+      "learning_rate": 1.664750418461731e-05,
+      "loss": 0.163,
+      "step": 6772
+    },
+    {
+      "epoch": 5.098231087692887,
+      "grad_norm": 0.18275082111358643,
+      "learning_rate": 1.6640986890432785e-05,
+      "loss": 0.1434,
+      "step": 6773
+    },
+    {
+      "epoch": 5.098983816334211,
+      "grad_norm": 0.21035659313201904,
+      "learning_rate": 1.6634470783773592e-05,
+      "loss": 0.1769,
+      "step": 6774
+    },
+    {
+      "epoch": 5.099736544975537,
+      "grad_norm": 0.20432990789413452,
+      "learning_rate": 1.6627955865352343e-05,
+      "loss": 0.1973,
+      "step": 6775
+    },
+    {
+      "epoch": 5.100489273616861,
+      "grad_norm": 0.20386987924575806,
+      "learning_rate": 1.662144213588149e-05,
+      "loss": 0.1898,
+      "step": 6776
+    },
+    {
+      "epoch": 5.101242002258186,
+      "grad_norm": 0.2034950703382492,
+      "learning_rate": 1.661492959607338e-05,
+      "loss": 0.197,
+      "step": 6777
+    },
+    {
+      "epoch": 5.101994730899511,
+      "grad_norm": 0.20700545608997345,
+      "learning_rate": 1.6608418246640196e-05,
+      "loss": 0.2932,
+      "step": 6778
+    },
+    {
+      "epoch": 5.1027474595408355,
+      "grad_norm": 0.1885414868593216,
+      "learning_rate": 1.6601908088294026e-05,
+      "loss": 0.1721,
+      "step": 6779
+    },
+    {
+      "epoch": 5.10350018818216,
+      "grad_norm": 0.17383354902267456,
+      "learning_rate": 1.6595399121746798e-05,
+      "loss": 0.1687,
+      "step": 6780
+    },
+    {
+      "epoch": 5.104252916823485,
+      "grad_norm": 0.21866241097450256,
+      "learning_rate": 1.6588891347710316e-05,
+      "loss": 0.187,
+      "step": 6781
+    },
+    {
+      "epoch": 5.10500564546481,
+      "grad_norm": 0.2106594443321228,
+      "learning_rate": 1.658238476689628e-05,
+      "loss": 0.2574,
+      "step": 6782
+    },
+    {
+      "epoch": 5.105758374106134,
+      "grad_norm": 0.20750829577445984,
+      "learning_rate": 1.6575879380016244e-05,
+      "loss": 0.1888,
+      "step": 6783
+    },
+    {
+      "epoch": 5.10651110274746,
+      "grad_norm": 0.22377942502498627,
+      "learning_rate": 1.65693751877816e-05,
+      "loss": 0.1966,
+      "step": 6784
+    },
+    {
+      "epoch": 5.107263831388784,
+      "grad_norm": 0.1987222135066986,
+      "learning_rate": 1.656287219090365e-05,
+      "loss": 0.2171,
+      "step": 6785
+    },
+    {
+      "epoch": 5.1080165600301095,
+      "grad_norm": 0.20425757765769958,
+      "learning_rate": 1.655637039009356e-05,
+      "loss": 0.1855,
+      "step": 6786
+    },
+    {
+      "epoch": 5.108769288671434,
+      "grad_norm": 0.21153679490089417,
+      "learning_rate": 1.654986978606235e-05,
+      "loss": 0.2359,
+      "step": 6787
+    },
+    {
+      "epoch": 5.1095220173127585,
+      "grad_norm": 0.19331839680671692,
+      "learning_rate": 1.6543370379520918e-05,
+      "loss": 0.1917,
+      "step": 6788
+    },
+    {
+      "epoch": 5.110274745954084,
+      "grad_norm": 0.20438174903392792,
+      "learning_rate": 1.6536872171180016e-05,
+      "loss": 0.199,
+      "step": 6789
+    },
+    {
+      "epoch": 5.111027474595408,
+      "grad_norm": 0.21667011082172394,
+      "learning_rate": 1.653037516175031e-05,
+      "loss": 0.1574,
+      "step": 6790
+    },
+    {
+      "epoch": 5.111780203236733,
+      "grad_norm": 0.1851176917552948,
+      "learning_rate": 1.6523879351942273e-05,
+      "loss": 0.2356,
+      "step": 6791
+    },
+    {
+      "epoch": 5.112532931878058,
+      "grad_norm": 0.17955902218818665,
+      "learning_rate": 1.6517384742466295e-05,
+      "loss": 0.177,
+      "step": 6792
+    },
+    {
+      "epoch": 5.113285660519383,
+      "grad_norm": 0.23778477311134338,
+      "learning_rate": 1.6510891334032603e-05,
+      "loss": 0.235,
+      "step": 6793
+    },
+    {
+      "epoch": 5.114038389160708,
+      "grad_norm": 0.18448469042778015,
+      "learning_rate": 1.6504399127351317e-05,
+      "loss": 0.1129,
+      "step": 6794
+    },
+    {
+      "epoch": 5.114791117802032,
+      "grad_norm": 0.23781445622444153,
+      "learning_rate": 1.6497908123132395e-05,
+      "loss": 0.1866,
+      "step": 6795
+    },
+    {
+      "epoch": 5.115543846443357,
+      "grad_norm": 0.18397897481918335,
+      "learning_rate": 1.6491418322085717e-05,
+      "loss": 0.1443,
+      "step": 6796
+    },
+    {
+      "epoch": 5.116296575084682,
+      "grad_norm": 0.2546220123767853,
+      "learning_rate": 1.6484929724920972e-05,
+      "loss": 0.2251,
+      "step": 6797
+    },
+    {
+      "epoch": 5.117049303726007,
+      "grad_norm": 0.20017346739768982,
+      "learning_rate": 1.647844233234776e-05,
+      "loss": 0.2092,
+      "step": 6798
+    },
+    {
+      "epoch": 5.117802032367331,
+      "grad_norm": 0.20341482758522034,
+      "learning_rate": 1.6471956145075514e-05,
+      "loss": 0.2275,
+      "step": 6799
+    },
+    {
+      "epoch": 5.118554761008657,
+      "grad_norm": 0.24738158285617828,
+      "learning_rate": 1.6465471163813574e-05,
+      "loss": 0.2323,
+      "step": 6800
+    },
+    {
+      "epoch": 5.118554761008657,
+      "eval_loss": 0.2203199863433838,
+      "eval_runtime": 452.5381,
+      "eval_samples_per_second": 21.273,
+      "eval_steps_per_second": 0.665,
+      "step": 6800
+    },
+    {
+      "epoch": 5.119307489649981,
+      "grad_norm": 0.20729447901248932,
+      "learning_rate": 1.64589873892711e-05,
+      "loss": 0.1952,
+      "step": 6801
+    },
+    {
+      "epoch": 5.120060218291306,
+      "grad_norm": 0.2369091659784317,
+      "learning_rate": 1.6452504822157182e-05,
+      "loss": 0.173,
+      "step": 6802
+    },
+    {
+      "epoch": 5.120812946932631,
+      "grad_norm": 0.2020484358072281,
+      "learning_rate": 1.644602346318072e-05,
+      "loss": 0.1394,
+      "step": 6803
+    },
+    {
+      "epoch": 5.121565675573955,
+      "grad_norm": 0.18813063204288483,
+      "learning_rate": 1.643954331305052e-05,
+      "loss": 0.1621,
+      "step": 6804
+    },
+    {
+      "epoch": 5.122318404215281,
+      "grad_norm": 0.1833622306585312,
+      "learning_rate": 1.6433064372475225e-05,
+      "loss": 0.0919,
+      "step": 6805
+    },
+    {
+      "epoch": 5.123071132856605,
+      "grad_norm": 0.2082682102918625,
+      "learning_rate": 1.6426586642163373e-05,
+      "loss": 0.2001,
+      "step": 6806
+    },
+    {
+      "epoch": 5.12382386149793,
+      "grad_norm": 0.21093176305294037,
+      "learning_rate": 1.6420110122823355e-05,
+      "loss": 0.1696,
+      "step": 6807
+    },
+    {
+      "epoch": 5.124576590139255,
+      "grad_norm": 0.19668875634670258,
+      "learning_rate": 1.6413634815163442e-05,
+      "loss": 0.2854,
+      "step": 6808
+    },
+    {
+      "epoch": 5.1253293187805795,
+      "grad_norm": 0.21460959315299988,
+      "learning_rate": 1.6407160719891755e-05,
+      "loss": 0.1462,
+      "step": 6809
+    },
+    {
+      "epoch": 5.126082047421904,
+      "grad_norm": 0.22007520496845245,
+      "learning_rate": 1.6400687837716288e-05,
+      "loss": 0.2243,
+      "step": 6810
+    },
+    {
+      "epoch": 5.126834776063229,
+      "grad_norm": 0.16412435472011566,
+      "learning_rate": 1.6394216169344926e-05,
+      "loss": 0.1732,
+      "step": 6811
+    },
+    {
+      "epoch": 5.127587504704554,
+      "grad_norm": 0.23854254186153412,
+      "learning_rate": 1.6387745715485368e-05,
+      "loss": 0.1645,
+      "step": 6812
+    },
+    {
+      "epoch": 5.128340233345879,
+      "grad_norm": 0.2239697128534317,
+      "learning_rate": 1.6381276476845254e-05,
+      "loss": 0.1456,
+      "step": 6813
+    },
+    {
+      "epoch": 5.129092961987204,
+      "grad_norm": 0.18524764478206635,
+      "learning_rate": 1.6374808454131998e-05,
+      "loss": 0.159,
+      "step": 6814
+    },
+    {
+      "epoch": 5.129845690628528,
+      "grad_norm": 0.2247796356678009,
+      "learning_rate": 1.6368341648052988e-05,
+      "loss": 0.2077,
+      "step": 6815
+    },
+    {
+      "epoch": 5.1305984192698535,
+      "grad_norm": 0.22062797844409943,
+      "learning_rate": 1.636187605931539e-05,
+      "loss": 0.1958,
+      "step": 6816
+    },
+    {
+      "epoch": 5.131351147911178,
+      "grad_norm": 0.19209334254264832,
+      "learning_rate": 1.6355411688626286e-05,
+      "loss": 0.2551,
+      "step": 6817
+    },
+    {
+      "epoch": 5.132103876552502,
+      "grad_norm": 0.1919633001089096,
+      "learning_rate": 1.6348948536692594e-05,
+      "loss": 0.131,
+      "step": 6818
+    },
+    {
+      "epoch": 5.132856605193828,
+      "grad_norm": 0.196451336145401,
+      "learning_rate": 1.6342486604221134e-05,
+      "loss": 0.1478,
+      "step": 6819
+    },
+    {
+      "epoch": 5.133609333835152,
+      "grad_norm": 0.21808210015296936,
+      "learning_rate": 1.633602589191855e-05,
+      "loss": 0.2461,
+      "step": 6820
+    },
+    {
+      "epoch": 5.134362062476478,
+      "grad_norm": 0.19797325134277344,
+      "learning_rate": 1.6329566400491408e-05,
+      "loss": 0.172,
+      "step": 6821
+    },
+    {
+      "epoch": 5.135114791117802,
+      "grad_norm": 0.22200214862823486,
+      "learning_rate": 1.6323108130646075e-05,
+      "loss": 0.2912,
+      "step": 6822
+    },
+    {
+      "epoch": 5.135867519759127,
+      "grad_norm": 0.20549823343753815,
+      "learning_rate": 1.6316651083088845e-05,
+      "loss": 0.15,
+      "step": 6823
+    },
+    {
+      "epoch": 5.136620248400452,
+      "grad_norm": 0.2101885825395584,
+      "learning_rate": 1.6310195258525834e-05,
+      "loss": 0.2525,
+      "step": 6824
+    },
+    {
+      "epoch": 5.137372977041776,
+      "grad_norm": 0.18714836239814758,
+      "learning_rate": 1.6303740657663045e-05,
+      "loss": 0.1881,
+      "step": 6825
+    },
+    {
+      "epoch": 5.138125705683101,
+      "grad_norm": 0.20977291464805603,
+      "learning_rate": 1.6297287281206334e-05,
+      "loss": 0.1977,
+      "step": 6826
+    },
+    {
+      "epoch": 5.138878434324426,
+      "grad_norm": 0.19961963593959808,
+      "learning_rate": 1.629083512986146e-05,
+      "loss": 0.2243,
+      "step": 6827
+    },
+    {
+      "epoch": 5.139631162965751,
+      "grad_norm": 0.1969051957130432,
+      "learning_rate": 1.6284384204333992e-05,
+      "loss": 0.2043,
+      "step": 6828
+    },
+    {
+      "epoch": 5.140383891607076,
+      "grad_norm": 0.18385367095470428,
+      "learning_rate": 1.6277934505329416e-05,
+      "loss": 0.1738,
+      "step": 6829
+    },
+    {
+      "epoch": 5.141136620248401,
+      "grad_norm": 0.193377286195755,
+      "learning_rate": 1.6271486033553042e-05,
+      "loss": 0.1984,
+      "step": 6830
+    },
+    {
+      "epoch": 5.141889348889725,
+      "grad_norm": 0.2063473016023636,
+      "learning_rate": 1.6265038789710072e-05,
+      "loss": 0.1871,
+      "step": 6831
+    },
+    {
+      "epoch": 5.14264207753105,
+      "grad_norm": 0.18302269279956818,
+      "learning_rate": 1.6258592774505577e-05,
+      "loss": 0.127,
+      "step": 6832
+    },
+    {
+      "epoch": 5.143394806172375,
+      "grad_norm": 0.2074163854122162,
+      "learning_rate": 1.6252147988644467e-05,
+      "loss": 0.2786,
+      "step": 6833
+    },
+    {
+      "epoch": 5.144147534813699,
+      "grad_norm": 0.198365718126297,
+      "learning_rate": 1.624570443283154e-05,
+      "loss": 0.1296,
+      "step": 6834
+    },
+    {
+      "epoch": 5.144900263455025,
+      "grad_norm": 0.20197372138500214,
+      "learning_rate": 1.623926210777146e-05,
+      "loss": 0.1313,
+      "step": 6835
+    },
+    {
+      "epoch": 5.145652992096349,
+      "grad_norm": 0.21470943093299866,
+      "learning_rate": 1.6232821014168746e-05,
+      "loss": 0.2462,
+      "step": 6836
+    },
+    {
+      "epoch": 5.146405720737674,
+      "grad_norm": 0.17650410532951355,
+      "learning_rate": 1.6226381152727788e-05,
+      "loss": 0.0837,
+      "step": 6837
+    },
+    {
+      "epoch": 5.147158449378999,
+      "grad_norm": 0.22055457532405853,
+      "learning_rate": 1.6219942524152843e-05,
+      "loss": 0.2194,
+      "step": 6838
+    },
+    {
+      "epoch": 5.1479111780203235,
+      "grad_norm": 0.20177045464515686,
+      "learning_rate": 1.6213505129147998e-05,
+      "loss": 0.2259,
+      "step": 6839
+    },
+    {
+      "epoch": 5.148663906661649,
+      "grad_norm": 0.19700433313846588,
+      "learning_rate": 1.620706896841729e-05,
+      "loss": 0.1678,
+      "step": 6840
+    },
+    {
+      "epoch": 5.149416635302973,
+      "grad_norm": 0.22045516967773438,
+      "learning_rate": 1.620063404266453e-05,
+      "loss": 0.2427,
+      "step": 6841
+    },
+    {
+      "epoch": 5.150169363944298,
+      "grad_norm": 0.19632041454315186,
+      "learning_rate": 1.619420035259345e-05,
+      "loss": 0.2071,
+      "step": 6842
+    },
+    {
+      "epoch": 5.150922092585623,
+      "grad_norm": 0.21999743580818176,
+      "learning_rate": 1.6187767898907617e-05,
+      "loss": 0.1665,
+      "step": 6843
+    },
+    {
+      "epoch": 5.151674821226948,
+      "grad_norm": 0.2331784963607788,
+      "learning_rate": 1.6181336682310483e-05,
+      "loss": 0.2508,
+      "step": 6844
+    },
+    {
+      "epoch": 5.152427549868272,
+      "grad_norm": 0.1954835206270218,
+      "learning_rate": 1.6174906703505345e-05,
+      "loss": 0.2052,
+      "step": 6845
+    },
+    {
+      "epoch": 5.1531802785095975,
+      "grad_norm": 0.16344404220581055,
+      "learning_rate": 1.6168477963195384e-05,
+      "loss": 0.1204,
+      "step": 6846
+    },
+    {
+      "epoch": 5.153933007150922,
+      "grad_norm": 0.2163626104593277,
+      "learning_rate": 1.6162050462083634e-05,
+      "loss": 0.1668,
+      "step": 6847
+    },
+    {
+      "epoch": 5.154685735792247,
+      "grad_norm": 0.23404721915721893,
+      "learning_rate": 1.615562420087301e-05,
+      "loss": 0.2541,
+      "step": 6848
+    },
+    {
+      "epoch": 5.155438464433572,
+      "grad_norm": 0.18276235461235046,
+      "learning_rate": 1.6149199180266255e-05,
+      "loss": 0.127,
+      "step": 6849
+    },
+    {
+      "epoch": 5.156191193074896,
+      "grad_norm": 0.18330501019954681,
+      "learning_rate": 1.6142775400966023e-05,
+      "loss": 0.1387,
+      "step": 6850
+    },
+    {
+      "epoch": 5.156943921716222,
+      "grad_norm": 0.19854386150836945,
+      "learning_rate": 1.6136352863674792e-05,
+      "loss": 0.2486,
+      "step": 6851
+    },
+    {
+      "epoch": 5.157696650357546,
+      "grad_norm": 0.19110967218875885,
+      "learning_rate": 1.6129931569094923e-05,
+      "loss": 0.1902,
+      "step": 6852
+    },
+    {
+      "epoch": 5.158449378998871,
+      "grad_norm": 0.1900116354227066,
+      "learning_rate": 1.6123511517928645e-05,
+      "loss": 0.1248,
+      "step": 6853
+    },
+    {
+      "epoch": 5.159202107640196,
+      "grad_norm": 0.19917260110378265,
+      "learning_rate": 1.6117092710878052e-05,
+      "loss": 0.206,
+      "step": 6854
+    },
+    {
+      "epoch": 5.15995483628152,
+      "grad_norm": 0.2132454663515091,
+      "learning_rate": 1.6110675148645073e-05,
+      "loss": 0.1985,
+      "step": 6855
+    },
+    {
+      "epoch": 5.160707564922845,
+      "grad_norm": 0.21629862487316132,
+      "learning_rate": 1.6104258831931542e-05,
+      "loss": 0.2058,
+      "step": 6856
+    },
+    {
+      "epoch": 5.16146029356417,
+      "grad_norm": 0.18417908251285553,
+      "learning_rate": 1.6097843761439138e-05,
+      "loss": 0.167,
+      "step": 6857
+    },
+    {
+      "epoch": 5.162213022205495,
+      "grad_norm": 0.19585399329662323,
+      "learning_rate": 1.6091429937869386e-05,
+      "loss": 0.111,
+      "step": 6858
+    },
+    {
+      "epoch": 5.16296575084682,
+      "grad_norm": 0.1921507865190506,
+      "learning_rate": 1.608501736192371e-05,
+      "loss": 0.1796,
+      "step": 6859
+    },
+    {
+      "epoch": 5.1637184794881446,
+      "grad_norm": 0.1931551992893219,
+      "learning_rate": 1.6078606034303366e-05,
+      "loss": 0.1831,
+      "step": 6860
+    },
+    {
+      "epoch": 5.164471208129469,
+      "grad_norm": 0.19943192601203918,
+      "learning_rate": 1.607219595570951e-05,
+      "loss": 0.1813,
+      "step": 6861
+    },
+    {
+      "epoch": 5.165223936770794,
+      "grad_norm": 0.18825066089630127,
+      "learning_rate": 1.6065787126843108e-05,
+      "loss": 0.1828,
+      "step": 6862
+    },
+    {
+      "epoch": 5.165976665412119,
+      "grad_norm": 0.2192976027727127,
+      "learning_rate": 1.6059379548405044e-05,
+      "loss": 0.1952,
+      "step": 6863
+    },
+    {
+      "epoch": 5.166729394053443,
+      "grad_norm": 0.1907716542482376,
+      "learning_rate": 1.605297322109603e-05,
+      "loss": 0.1844,
+      "step": 6864
+    },
+    {
+      "epoch": 5.167482122694769,
+      "grad_norm": 0.20915472507476807,
+      "learning_rate": 1.6046568145616648e-05,
+      "loss": 0.3585,
+      "step": 6865
+    },
+    {
+      "epoch": 5.168234851336093,
+      "grad_norm": 0.1756918728351593,
+      "learning_rate": 1.6040164322667356e-05,
+      "loss": 0.1775,
+      "step": 6866
+    },
+    {
+      "epoch": 5.1689875799774185,
+      "grad_norm": 0.19970831274986267,
+      "learning_rate": 1.603376175294847e-05,
+      "loss": 0.1636,
+      "step": 6867
+    },
+    {
+      "epoch": 5.169740308618743,
+      "grad_norm": 0.205888032913208,
+      "learning_rate": 1.602736043716015e-05,
+      "loss": 0.2924,
+      "step": 6868
+    },
+    {
+      "epoch": 5.1704930372600675,
+      "grad_norm": 0.1897847056388855,
+      "learning_rate": 1.6020960376002457e-05,
+      "loss": 0.1258,
+      "step": 6869
+    },
+    {
+      "epoch": 5.171245765901393,
+      "grad_norm": 0.20911246538162231,
+      "learning_rate": 1.601456157017527e-05,
+      "loss": 0.1869,
+      "step": 6870
+    },
+    {
+      "epoch": 5.171998494542717,
+      "grad_norm": 0.20309455692768097,
+      "learning_rate": 1.6008164020378358e-05,
+      "loss": 0.1423,
+      "step": 6871
+    },
+    {
+      "epoch": 5.172751223184042,
+      "grad_norm": 0.20543228089809418,
+      "learning_rate": 1.6001767727311352e-05,
+      "loss": 0.3048,
+      "step": 6872
+    },
+    {
+      "epoch": 5.173503951825367,
+      "grad_norm": 0.19321087002754211,
+      "learning_rate": 1.5995372691673748e-05,
+      "loss": 0.1891,
+      "step": 6873
+    },
+    {
+      "epoch": 5.174256680466692,
+      "grad_norm": 0.1787692755460739,
+      "learning_rate": 1.598897891416488e-05,
+      "loss": 0.1026,
+      "step": 6874
+    },
+    {
+      "epoch": 5.175009409108017,
+      "grad_norm": 0.1817975789308548,
+      "learning_rate": 1.5982586395483983e-05,
+      "loss": 0.2356,
+      "step": 6875
+    },
+    {
+      "epoch": 5.1757621377493415,
+      "grad_norm": 0.2044084370136261,
+      "learning_rate": 1.5976195136330113e-05,
+      "loss": 0.2516,
+      "step": 6876
+    },
+    {
+      "epoch": 5.176514866390666,
+      "grad_norm": 0.17483024299144745,
+      "learning_rate": 1.5969805137402216e-05,
+      "loss": 0.1817,
+      "step": 6877
+    },
+    {
+      "epoch": 5.177267595031991,
+      "grad_norm": 0.18486732244491577,
+      "learning_rate": 1.5963416399399095e-05,
+      "loss": 0.1452,
+      "step": 6878
+    },
+    {
+      "epoch": 5.178020323673316,
+      "grad_norm": 0.1848524808883667,
+      "learning_rate": 1.595702892301942e-05,
+      "loss": 0.1362,
+      "step": 6879
+    },
+    {
+      "epoch": 5.17877305231464,
+      "grad_norm": 0.21424198150634766,
+      "learning_rate": 1.59506427089617e-05,
+      "loss": 0.1935,
+      "step": 6880
+    },
+    {
+      "epoch": 5.179525780955966,
+      "grad_norm": 0.16397716104984283,
+      "learning_rate": 1.5944257757924324e-05,
+      "loss": 0.1322,
+      "step": 6881
+    },
+    {
+      "epoch": 5.18027850959729,
+      "grad_norm": 0.19430327415466309,
+      "learning_rate": 1.5937874070605557e-05,
+      "loss": 0.0846,
+      "step": 6882
+    },
+    {
+      "epoch": 5.1810312382386146,
+      "grad_norm": 0.22000248730182648,
+      "learning_rate": 1.593149164770349e-05,
+      "loss": 0.2061,
+      "step": 6883
+    },
+    {
+      "epoch": 5.18178396687994,
+      "grad_norm": 0.20032785832881927,
+      "learning_rate": 1.5925110489916106e-05,
+      "loss": 0.1625,
+      "step": 6884
+    },
+    {
+      "epoch": 5.182536695521264,
+      "grad_norm": 0.1934167891740799,
+      "learning_rate": 1.5918730597941235e-05,
+      "loss": 0.1639,
+      "step": 6885
+    },
+    {
+      "epoch": 5.18328942416259,
+      "grad_norm": 0.20601677894592285,
+      "learning_rate": 1.591235197247658e-05,
+      "loss": 0.3023,
+      "step": 6886
+    },
+    {
+      "epoch": 5.184042152803914,
+      "grad_norm": 0.18875761330127716,
+      "learning_rate": 1.5905974614219683e-05,
+      "loss": 0.2718,
+      "step": 6887
+    },
+    {
+      "epoch": 5.184794881445239,
+      "grad_norm": 0.20733177661895752,
+      "learning_rate": 1.589959852386798e-05,
+      "loss": 0.2047,
+      "step": 6888
+    },
+    {
+      "epoch": 5.185547610086564,
+      "grad_norm": 0.2102918177843094,
+      "learning_rate": 1.589322370211873e-05,
+      "loss": 0.1633,
+      "step": 6889
+    },
+    {
+      "epoch": 5.1863003387278885,
+      "grad_norm": 0.18436190485954285,
+      "learning_rate": 1.588685014966909e-05,
+      "loss": 0.196,
+      "step": 6890
+    },
+    {
+      "epoch": 5.187053067369213,
+      "grad_norm": 0.2234569638967514,
+      "learning_rate": 1.588047786721606e-05,
+      "loss": 0.2419,
+      "step": 6891
+    },
+    {
+      "epoch": 5.187805796010538,
+      "grad_norm": 0.1973486691713333,
+      "learning_rate": 1.5874106855456504e-05,
+      "loss": 0.1497,
+      "step": 6892
+    },
+    {
+      "epoch": 5.188558524651863,
+      "grad_norm": 0.18115998804569244,
+      "learning_rate": 1.5867737115087137e-05,
+      "loss": 0.1772,
+      "step": 6893
+    },
+    {
+      "epoch": 5.189311253293188,
+      "grad_norm": 0.1828973889350891,
+      "learning_rate": 1.5861368646804556e-05,
+      "loss": 0.1921,
+      "step": 6894
+    },
+    {
+      "epoch": 5.190063981934513,
+      "grad_norm": 0.1975078284740448,
+      "learning_rate": 1.5855001451305195e-05,
+      "loss": 0.1672,
+      "step": 6895
+    },
+    {
+      "epoch": 5.190816710575837,
+      "grad_norm": 0.23067273199558258,
+      "learning_rate": 1.5848635529285368e-05,
+      "loss": 0.1843,
+      "step": 6896
+    },
+    {
+      "epoch": 5.1915694392171625,
+      "grad_norm": 0.21770846843719482,
+      "learning_rate": 1.584227088144124e-05,
+      "loss": 0.227,
+      "step": 6897
+    },
+    {
+      "epoch": 5.192322167858487,
+      "grad_norm": 0.18627290427684784,
+      "learning_rate": 1.583590750846885e-05,
+      "loss": 0.2389,
+      "step": 6898
+    },
+    {
+      "epoch": 5.1930748964998115,
+      "grad_norm": 0.16783219575881958,
+      "learning_rate": 1.5829545411064074e-05,
+      "loss": 0.162,
+      "step": 6899
+    },
+    {
+      "epoch": 5.193827625141137,
+      "grad_norm": 0.245481476187706,
+      "learning_rate": 1.582318458992267e-05,
+      "loss": 0.1679,
+      "step": 6900
+    },
+    {
+      "epoch": 5.194580353782461,
+      "grad_norm": 0.17710351943969727,
+      "learning_rate": 1.5816825045740232e-05,
+      "loss": 0.1875,
+      "step": 6901
+    },
+    {
+      "epoch": 5.195333082423787,
+      "grad_norm": 0.20323695242404938,
+      "learning_rate": 1.5810466779212243e-05,
+      "loss": 0.2162,
+      "step": 6902
+    },
+    {
+      "epoch": 5.196085811065111,
+      "grad_norm": 0.21337567269802094,
+      "learning_rate": 1.5804109791034035e-05,
+      "loss": 0.2017,
+      "step": 6903
+    },
+    {
+      "epoch": 5.196838539706436,
+      "grad_norm": 0.19210685789585114,
+      "learning_rate": 1.5797754081900802e-05,
+      "loss": 0.1531,
+      "step": 6904
+    },
+    {
+      "epoch": 5.197591268347761,
+      "grad_norm": 0.2531861960887909,
+      "learning_rate": 1.5791399652507582e-05,
+      "loss": 0.2651,
+      "step": 6905
+    },
+    {
+      "epoch": 5.1983439969890854,
+      "grad_norm": 0.17132413387298584,
+      "learning_rate": 1.5785046503549294e-05,
+      "loss": 0.1625,
+      "step": 6906
+    },
+    {
+      "epoch": 5.19909672563041,
+      "grad_norm": 0.19676269590854645,
+      "learning_rate": 1.5778694635720713e-05,
+      "loss": 0.2377,
+      "step": 6907
+    },
+    {
+      "epoch": 5.199849454271735,
+      "grad_norm": 0.1818702667951584,
+      "learning_rate": 1.577234404971645e-05,
+      "loss": 0.1107,
+      "step": 6908
+    },
+    {
+      "epoch": 5.20060218291306,
+      "grad_norm": 0.18674997985363007,
+      "learning_rate": 1.5765994746231024e-05,
+      "loss": 0.1704,
+      "step": 6909
+    },
+    {
+      "epoch": 5.201354911554384,
+      "grad_norm": 0.1912098228931427,
+      "learning_rate": 1.575964672595876e-05,
+      "loss": 0.1387,
+      "step": 6910
+    },
+    {
+      "epoch": 5.20210764019571,
+      "grad_norm": 0.20121639966964722,
+      "learning_rate": 1.5753299989593886e-05,
+      "loss": 0.1855,
+      "step": 6911
+    },
+    {
+      "epoch": 5.202860368837034,
+      "grad_norm": 0.22186066210269928,
+      "learning_rate": 1.574695453783046e-05,
+      "loss": 0.1342,
+      "step": 6912
+    },
+    {
+      "epoch": 5.203613097478359,
+      "grad_norm": 0.19472672045230865,
+      "learning_rate": 1.5740610371362423e-05,
+      "loss": 0.2072,
+      "step": 6913
+    },
+    {
+      "epoch": 5.204365826119684,
+      "grad_norm": 0.21769261360168457,
+      "learning_rate": 1.5734267490883548e-05,
+      "loss": 0.179,
+      "step": 6914
+    },
+    {
+      "epoch": 5.205118554761008,
+      "grad_norm": 0.20200887322425842,
+      "learning_rate": 1.5727925897087496e-05,
+      "loss": 0.2212,
+      "step": 6915
+    },
+    {
+      "epoch": 5.205871283402334,
+      "grad_norm": 0.19761723279953003,
+      "learning_rate": 1.5721585590667754e-05,
+      "loss": 0.1175,
+      "step": 6916
+    },
+    {
+      "epoch": 5.206624012043658,
+      "grad_norm": 0.20394407212734222,
+      "learning_rate": 1.571524657231772e-05,
+      "loss": 0.2047,
+      "step": 6917
+    },
+    {
+      "epoch": 5.207376740684983,
+      "grad_norm": 0.20493577420711517,
+      "learning_rate": 1.5708908842730596e-05,
+      "loss": 0.1992,
+      "step": 6918
+    },
+    {
+      "epoch": 5.208129469326308,
+      "grad_norm": 0.22024038434028625,
+      "learning_rate": 1.570257240259948e-05,
+      "loss": 0.1822,
+      "step": 6919
+    },
+    {
+      "epoch": 5.2088821979676325,
+      "grad_norm": 0.20663362741470337,
+      "learning_rate": 1.5696237252617302e-05,
+      "loss": 0.1584,
+      "step": 6920
+    },
+    {
+      "epoch": 5.209634926608958,
+      "grad_norm": 0.19821880757808685,
+      "learning_rate": 1.5689903393476876e-05,
+      "loss": 0.1983,
+      "step": 6921
+    },
+    {
+      "epoch": 5.210387655250282,
+      "grad_norm": 0.18092191219329834,
+      "learning_rate": 1.5683570825870847e-05,
+      "loss": 0.1995,
+      "step": 6922
+    },
+    {
+      "epoch": 5.211140383891607,
+      "grad_norm": 0.19114235043525696,
+      "learning_rate": 1.567723955049176e-05,
+      "loss": 0.1481,
+      "step": 6923
+    },
+    {
+      "epoch": 5.211893112532932,
+      "grad_norm": 0.18940679728984833,
+      "learning_rate": 1.567090956803197e-05,
+      "loss": 0.1887,
+      "step": 6924
+    },
+    {
+      "epoch": 5.212645841174257,
+      "grad_norm": 0.19685164093971252,
+      "learning_rate": 1.5664580879183737e-05,
+      "loss": 0.1436,
+      "step": 6925
+    },
+    {
+      "epoch": 5.213398569815581,
+      "grad_norm": 0.19910749793052673,
+      "learning_rate": 1.5658253484639134e-05,
+      "loss": 0.2214,
+      "step": 6926
+    },
+    {
+      "epoch": 5.2141512984569065,
+      "grad_norm": 0.20601874589920044,
+      "learning_rate": 1.565192738509013e-05,
+      "loss": 0.1593,
+      "step": 6927
+    },
+    {
+      "epoch": 5.214904027098231,
+      "grad_norm": 0.19432809948921204,
+      "learning_rate": 1.564560258122853e-05,
+      "loss": 0.1813,
+      "step": 6928
+    },
+    {
+      "epoch": 5.2156567557395555,
+      "grad_norm": 0.21208904683589935,
+      "learning_rate": 1.5639279073746023e-05,
+      "loss": 0.2168,
+      "step": 6929
+    },
+    {
+      "epoch": 5.216409484380881,
+      "grad_norm": 0.2353551685810089,
+      "learning_rate": 1.5632956863334107e-05,
+      "loss": 0.2044,
+      "step": 6930
+    },
+    {
+      "epoch": 5.217162213022205,
+      "grad_norm": 0.17367485165596008,
+      "learning_rate": 1.5626635950684192e-05,
+      "loss": 0.1588,
+      "step": 6931
+    },
+    {
+      "epoch": 5.217914941663531,
+      "grad_norm": 0.23626679182052612,
+      "learning_rate": 1.562031633648753e-05,
+      "loss": 0.1586,
+      "step": 6932
+    },
+    {
+      "epoch": 5.218667670304855,
+      "grad_norm": 0.18960323929786682,
+      "learning_rate": 1.5613998021435196e-05,
+      "loss": 0.2053,
+      "step": 6933
+    },
+    {
+      "epoch": 5.21942039894618,
+      "grad_norm": 0.21240799129009247,
+      "learning_rate": 1.5607681006218182e-05,
+      "loss": 0.3007,
+      "step": 6934
+    },
+    {
+      "epoch": 5.220173127587505,
+      "grad_norm": 0.1748628318309784,
+      "learning_rate": 1.5601365291527277e-05,
+      "loss": 0.1698,
+      "step": 6935
+    },
+    {
+      "epoch": 5.220925856228829,
+      "grad_norm": 0.17131106555461884,
+      "learning_rate": 1.5595050878053187e-05,
+      "loss": 0.1819,
+      "step": 6936
+    },
+    {
+      "epoch": 5.221678584870154,
+      "grad_norm": 0.18226127326488495,
+      "learning_rate": 1.5588737766486426e-05,
+      "loss": 0.1591,
+      "step": 6937
+    },
+    {
+      "epoch": 5.222431313511479,
+      "grad_norm": 0.17117829620838165,
+      "learning_rate": 1.5582425957517404e-05,
+      "loss": 0.0807,
+      "step": 6938
+    },
+    {
+      "epoch": 5.223184042152804,
+      "grad_norm": 0.22042544186115265,
+      "learning_rate": 1.5576115451836355e-05,
+      "loss": 0.2307,
+      "step": 6939
+    },
+    {
+      "epoch": 5.223936770794129,
+      "grad_norm": 0.21648643910884857,
+      "learning_rate": 1.55698062501334e-05,
+      "loss": 0.2522,
+      "step": 6940
+    },
+    {
+      "epoch": 5.224689499435454,
+      "grad_norm": 0.18348824977874756,
+      "learning_rate": 1.556349835309848e-05,
+      "loss": 0.2324,
+      "step": 6941
+    },
+    {
+      "epoch": 5.225442228076778,
+      "grad_norm": 0.21329446136951447,
+      "learning_rate": 1.5557191761421453e-05,
+      "loss": 0.149,
+      "step": 6942
+    },
+    {
+      "epoch": 5.226194956718103,
+      "grad_norm": 0.17483407258987427,
+      "learning_rate": 1.5550886475791965e-05,
+      "loss": 0.1561,
+      "step": 6943
+    },
+    {
+      "epoch": 5.226947685359428,
+      "grad_norm": 0.19208358228206635,
+      "learning_rate": 1.554458249689958e-05,
+      "loss": 0.2609,
+      "step": 6944
+    },
+    {
+      "epoch": 5.227700414000752,
+      "grad_norm": 0.17742113769054413,
+      "learning_rate": 1.553827982543367e-05,
+      "loss": 0.1388,
+      "step": 6945
+    },
+    {
+      "epoch": 5.228453142642078,
+      "grad_norm": 0.19866347312927246,
+      "learning_rate": 1.5531978462083503e-05,
+      "loss": 0.169,
+      "step": 6946
+    },
+    {
+      "epoch": 5.229205871283402,
+      "grad_norm": 0.18249931931495667,
+      "learning_rate": 1.552567840753816e-05,
+      "loss": 0.1062,
+      "step": 6947
+    },
+    {
+      "epoch": 5.229958599924728,
+      "grad_norm": 0.2004917562007904,
+      "learning_rate": 1.5519379662486635e-05,
+      "loss": 0.2293,
+      "step": 6948
+    },
+    {
+      "epoch": 5.230711328566052,
+      "grad_norm": 0.1950550079345703,
+      "learning_rate": 1.5513082227617732e-05,
+      "loss": 0.2034,
+      "step": 6949
+    },
+    {
+      "epoch": 5.2314640572073765,
+      "grad_norm": 0.17744772136211395,
+      "learning_rate": 1.5506786103620146e-05,
+      "loss": 0.1811,
+      "step": 6950
+    },
+    {
+      "epoch": 5.232216785848702,
+      "grad_norm": 0.1748194545507431,
+      "learning_rate": 1.550049129118239e-05,
+      "loss": 0.1721,
+      "step": 6951
+    },
+    {
+      "epoch": 5.232969514490026,
+      "grad_norm": 0.2075432986021042,
+      "learning_rate": 1.5494197790992865e-05,
+      "loss": 0.166,
+      "step": 6952
+    },
+    {
+      "epoch": 5.233722243131351,
+      "grad_norm": 0.20287556946277618,
+      "learning_rate": 1.5487905603739828e-05,
+      "loss": 0.1727,
+      "step": 6953
+    },
+    {
+      "epoch": 5.234474971772676,
+      "grad_norm": 0.1920081377029419,
+      "learning_rate": 1.5481614730111365e-05,
+      "loss": 0.1908,
+      "step": 6954
+    },
+    {
+      "epoch": 5.235227700414001,
+      "grad_norm": 0.18870557844638824,
+      "learning_rate": 1.5475325170795445e-05,
+      "loss": 0.2606,
+      "step": 6955
+    },
+    {
+      "epoch": 5.235980429055326,
+      "grad_norm": 0.19162002205848694,
+      "learning_rate": 1.546903692647989e-05,
+      "loss": 0.197,
+      "step": 6956
+    },
+    {
+      "epoch": 5.2367331576966505,
+      "grad_norm": 0.21661458909511566,
+      "learning_rate": 1.546274999785238e-05,
+      "loss": 0.1841,
+      "step": 6957
+    },
+    {
+      "epoch": 5.237485886337975,
+      "grad_norm": 0.21450932323932648,
+      "learning_rate": 1.545646438560041e-05,
+      "loss": 0.2122,
+      "step": 6958
+    },
+    {
+      "epoch": 5.2382386149793,
+      "grad_norm": 0.1633850336074829,
+      "learning_rate": 1.5450180090411408e-05,
+      "loss": 0.161,
+      "step": 6959
+    },
+    {
+      "epoch": 5.238991343620625,
+      "grad_norm": 0.2282143235206604,
+      "learning_rate": 1.544389711297257e-05,
+      "loss": 0.1613,
+      "step": 6960
+    },
+    {
+      "epoch": 5.239744072261949,
+      "grad_norm": 0.1911134272813797,
+      "learning_rate": 1.543761545397104e-05,
+      "loss": 0.198,
+      "step": 6961
+    },
+    {
+      "epoch": 5.240496800903275,
+      "grad_norm": 0.20168398320674896,
+      "learning_rate": 1.5431335114093737e-05,
+      "loss": 0.2621,
+      "step": 6962
+    },
+    {
+      "epoch": 5.241249529544599,
+      "grad_norm": 0.17330129444599152,
+      "learning_rate": 1.5425056094027493e-05,
+      "loss": 0.1596,
+      "step": 6963
+    },
+    {
+      "epoch": 5.242002258185924,
+      "grad_norm": 0.1901233196258545,
+      "learning_rate": 1.541877839445895e-05,
+      "loss": 0.1678,
+      "step": 6964
+    },
+    {
+      "epoch": 5.242754986827249,
+      "grad_norm": 0.16977640986442566,
+      "learning_rate": 1.5412502016074646e-05,
+      "loss": 0.2031,
+      "step": 6965
+    },
+    {
+      "epoch": 5.243507715468573,
+      "grad_norm": 0.19384002685546875,
+      "learning_rate": 1.540622695956093e-05,
+      "loss": 0.2201,
+      "step": 6966
+    },
+    {
+      "epoch": 5.244260444109899,
+      "grad_norm": 0.18800145387649536,
+      "learning_rate": 1.5399953225604067e-05,
+      "loss": 0.2038,
+      "step": 6967
+    },
+    {
+      "epoch": 5.245013172751223,
+      "grad_norm": 0.21602359414100647,
+      "learning_rate": 1.5393680814890124e-05,
+      "loss": 0.1727,
+      "step": 6968
+    },
+    {
+      "epoch": 5.245765901392548,
+      "grad_norm": 0.20622768998146057,
+      "learning_rate": 1.538740972810505e-05,
+      "loss": 0.1727,
+      "step": 6969
+    },
+    {
+      "epoch": 5.246518630033873,
+      "grad_norm": 0.20608791708946228,
+      "learning_rate": 1.5381139965934625e-05,
+      "loss": 0.1992,
+      "step": 6970
+    },
+    {
+      "epoch": 5.247271358675198,
+      "grad_norm": 0.22737164795398712,
+      "learning_rate": 1.537487152906453e-05,
+      "loss": 0.2187,
+      "step": 6971
+    },
+    {
+      "epoch": 5.248024087316522,
+      "grad_norm": 0.23582184314727783,
+      "learning_rate": 1.5368604418180242e-05,
+      "loss": 0.1677,
+      "step": 6972
+    },
+    {
+      "epoch": 5.248776815957847,
+      "grad_norm": 0.20266638696193695,
+      "learning_rate": 1.5362338633967133e-05,
+      "loss": 0.182,
+      "step": 6973
+    },
+    {
+      "epoch": 5.249529544599172,
+      "grad_norm": 0.1712830513715744,
+      "learning_rate": 1.5356074177110424e-05,
+      "loss": 0.1225,
+      "step": 6974
+    },
+    {
+      "epoch": 5.250282273240497,
+      "grad_norm": 0.2270311415195465,
+      "learning_rate": 1.5349811048295196e-05,
+      "loss": 0.2403,
+      "step": 6975
+    },
+    {
+      "epoch": 5.251035001881822,
+      "grad_norm": 0.2362806797027588,
+      "learning_rate": 1.5343549248206353e-05,
+      "loss": 0.2705,
+      "step": 6976
+    },
+    {
+      "epoch": 5.251787730523146,
+      "grad_norm": 0.2061147540807724,
+      "learning_rate": 1.533728877752869e-05,
+      "loss": 0.1987,
+      "step": 6977
+    },
+    {
+      "epoch": 5.2525404591644715,
+      "grad_norm": 0.20129837095737457,
+      "learning_rate": 1.5331029636946844e-05,
+      "loss": 0.197,
+      "step": 6978
+    },
+    {
+      "epoch": 5.253293187805796,
+      "grad_norm": 0.15853962302207947,
+      "learning_rate": 1.5324771827145295e-05,
+      "loss": 0.116,
+      "step": 6979
+    },
+    {
+      "epoch": 5.2540459164471205,
+      "grad_norm": 0.1845935881137848,
+      "learning_rate": 1.5318515348808395e-05,
+      "loss": 0.2078,
+      "step": 6980
+    },
+    {
+      "epoch": 5.254798645088446,
+      "grad_norm": 0.2012375295162201,
+      "learning_rate": 1.5312260202620343e-05,
+      "loss": 0.2191,
+      "step": 6981
+    },
+    {
+      "epoch": 5.25555137372977,
+      "grad_norm": 0.17796450853347778,
+      "learning_rate": 1.5306006389265197e-05,
+      "loss": 0.2312,
+      "step": 6982
+    },
+    {
+      "epoch": 5.256304102371095,
+      "grad_norm": 0.18921703100204468,
+      "learning_rate": 1.529975390942685e-05,
+      "loss": 0.1986,
+      "step": 6983
+    },
+    {
+      "epoch": 5.25705683101242,
+      "grad_norm": 0.18370066583156586,
+      "learning_rate": 1.5293502763789084e-05,
+      "loss": 0.1323,
+      "step": 6984
+    },
+    {
+      "epoch": 5.257809559653745,
+      "grad_norm": 0.21719405055046082,
+      "learning_rate": 1.5287252953035492e-05,
+      "loss": 0.1761,
+      "step": 6985
+    },
+    {
+      "epoch": 5.25856228829507,
+      "grad_norm": 0.1983148753643036,
+      "learning_rate": 1.5281004477849557e-05,
+      "loss": 0.2382,
+      "step": 6986
+    },
+    {
+      "epoch": 5.2593150169363945,
+      "grad_norm": 0.21538472175598145,
+      "learning_rate": 1.5274757338914603e-05,
+      "loss": 0.26,
+      "step": 6987
+    },
+    {
+      "epoch": 5.260067745577719,
+      "grad_norm": 0.18505749106407166,
+      "learning_rate": 1.526851153691381e-05,
+      "loss": 0.1307,
+      "step": 6988
+    },
+    {
+      "epoch": 5.260820474219044,
+      "grad_norm": 0.2014084756374359,
+      "learning_rate": 1.5262267072530202e-05,
+      "loss": 0.1877,
+      "step": 6989
+    },
+    {
+      "epoch": 5.261573202860369,
+      "grad_norm": 0.20545974373817444,
+      "learning_rate": 1.5256023946446674e-05,
+      "loss": 0.2582,
+      "step": 6990
+    },
+    {
+      "epoch": 5.262325931501693,
+      "grad_norm": 0.1855407953262329,
+      "learning_rate": 1.5249782159345949e-05,
+      "loss": 0.2062,
+      "step": 6991
+    },
+    {
+      "epoch": 5.263078660143019,
+      "grad_norm": 0.19321133196353912,
+      "learning_rate": 1.5243541711910627e-05,
+      "loss": 0.2773,
+      "step": 6992
+    },
+    {
+      "epoch": 5.263831388784343,
+      "grad_norm": 0.17432326078414917,
+      "learning_rate": 1.5237302604823162e-05,
+      "loss": 0.1572,
+      "step": 6993
+    },
+    {
+      "epoch": 5.2645841174256685,
+      "grad_norm": 0.18890932202339172,
+      "learning_rate": 1.5231064838765852e-05,
+      "loss": 0.1336,
+      "step": 6994
+    },
+    {
+      "epoch": 5.265336846066993,
+      "grad_norm": 0.18392370641231537,
+      "learning_rate": 1.522482841442084e-05,
+      "loss": 0.1388,
+      "step": 6995
+    },
+    {
+      "epoch": 5.266089574708317,
+      "grad_norm": 0.18065111339092255,
+      "learning_rate": 1.521859333247014e-05,
+      "loss": 0.1875,
+      "step": 6996
+    },
+    {
+      "epoch": 5.266842303349643,
+      "grad_norm": 0.18969303369522095,
+      "learning_rate": 1.5212359593595604e-05,
+      "loss": 0.1732,
+      "step": 6997
+    },
+    {
+      "epoch": 5.267595031990967,
+      "grad_norm": 0.19741708040237427,
+      "learning_rate": 1.5206127198478948e-05,
+      "loss": 0.1549,
+      "step": 6998
+    },
+    {
+      "epoch": 5.268347760632292,
+      "grad_norm": 0.194239541888237,
+      "learning_rate": 1.519989614780174e-05,
+      "loss": 0.2111,
+      "step": 6999
+    },
+    {
+      "epoch": 5.269100489273617,
+      "grad_norm": 0.16288648545742035,
+      "learning_rate": 1.5193666442245402e-05,
+      "loss": 0.1543,
+      "step": 7000
+    },
+    {
+      "epoch": 5.269100489273617,
+      "eval_loss": 0.21799756586551666,
+      "eval_runtime": 452.7794,
+      "eval_samples_per_second": 21.262,
+      "eval_steps_per_second": 0.665,
+      "step": 7000
+    },
+    {
+      "epoch": 5.2698532179149415,
+      "grad_norm": 0.1766362339258194,
+      "learning_rate": 1.518743808249119e-05,
+      "loss": 0.1556,
+      "step": 7001
+    },
+    {
+      "epoch": 5.270605946556266,
+      "grad_norm": 0.20997858047485352,
+      "learning_rate": 1.5181211069220239e-05,
+      "loss": 0.1684,
+      "step": 7002
+    },
+    {
+      "epoch": 5.271358675197591,
+      "grad_norm": 0.1842949092388153,
+      "learning_rate": 1.5174985403113534e-05,
+      "loss": 0.1832,
+      "step": 7003
+    },
+    {
+      "epoch": 5.272111403838916,
+      "grad_norm": 0.18876467645168304,
+      "learning_rate": 1.5168761084851884e-05,
+      "loss": 0.216,
+      "step": 7004
+    },
+    {
+      "epoch": 5.272864132480241,
+      "grad_norm": 0.1734398603439331,
+      "learning_rate": 1.5162538115115987e-05,
+      "loss": 0.1494,
+      "step": 7005
+    },
+    {
+      "epoch": 5.273616861121566,
+      "grad_norm": 0.180531844496727,
+      "learning_rate": 1.515631649458637e-05,
+      "loss": 0.2135,
+      "step": 7006
+    },
+    {
+      "epoch": 5.27436958976289,
+      "grad_norm": 0.20493638515472412,
+      "learning_rate": 1.515009622394343e-05,
+      "loss": 0.2094,
+      "step": 7007
+    },
+    {
+      "epoch": 5.2751223184042155,
+      "grad_norm": 0.1673840433359146,
+      "learning_rate": 1.5143877303867391e-05,
+      "loss": 0.1076,
+      "step": 7008
+    },
+    {
+      "epoch": 5.27587504704554,
+      "grad_norm": 0.17632687091827393,
+      "learning_rate": 1.5137659735038361e-05,
+      "loss": 0.18,
+      "step": 7009
+    },
+    {
+      "epoch": 5.2766277756868645,
+      "grad_norm": 0.1771775335073471,
+      "learning_rate": 1.5131443518136267e-05,
+      "loss": 0.1049,
+      "step": 7010
+    },
+    {
+      "epoch": 5.27738050432819,
+      "grad_norm": 0.1760403960943222,
+      "learning_rate": 1.5125228653840917e-05,
+      "loss": 0.2655,
+      "step": 7011
+    },
+    {
+      "epoch": 5.278133232969514,
+      "grad_norm": 0.2011805921792984,
+      "learning_rate": 1.5119015142831955e-05,
+      "loss": 0.1356,
+      "step": 7012
+    },
+    {
+      "epoch": 5.27888596161084,
+      "grad_norm": 0.20846842229366302,
+      "learning_rate": 1.5112802985788893e-05,
+      "loss": 0.223,
+      "step": 7013
+    },
+    {
+      "epoch": 5.279638690252164,
+      "grad_norm": 0.2028125375509262,
+      "learning_rate": 1.5106592183391061e-05,
+      "loss": 0.1409,
+      "step": 7014
+    },
+    {
+      "epoch": 5.280391418893489,
+      "grad_norm": 0.18192408978939056,
+      "learning_rate": 1.510038273631768e-05,
+      "loss": 0.2285,
+      "step": 7015
+    },
+    {
+      "epoch": 5.281144147534814,
+      "grad_norm": 0.22021111845970154,
+      "learning_rate": 1.5094174645247794e-05,
+      "loss": 0.1789,
+      "step": 7016
+    },
+    {
+      "epoch": 5.2818968761761385,
+      "grad_norm": 0.2303897887468338,
+      "learning_rate": 1.5087967910860324e-05,
+      "loss": 0.1614,
+      "step": 7017
+    },
+    {
+      "epoch": 5.282649604817463,
+      "grad_norm": 0.19712571799755096,
+      "learning_rate": 1.5081762533834015e-05,
+      "loss": 0.2299,
+      "step": 7018
+    },
+    {
+      "epoch": 5.283402333458788,
+      "grad_norm": 0.16131165623664856,
+      "learning_rate": 1.5075558514847493e-05,
+      "loss": 0.1532,
+      "step": 7019
+    },
+    {
+      "epoch": 5.284155062100113,
+      "grad_norm": 0.18678946793079376,
+      "learning_rate": 1.5069355854579203e-05,
+      "loss": 0.1354,
+      "step": 7020
+    },
+    {
+      "epoch": 5.284907790741438,
+      "grad_norm": 0.18208858370780945,
+      "learning_rate": 1.5063154553707482e-05,
+      "loss": 0.1788,
+      "step": 7021
+    },
+    {
+      "epoch": 5.285660519382763,
+      "grad_norm": 0.19182419776916504,
+      "learning_rate": 1.5056954612910468e-05,
+      "loss": 0.1431,
+      "step": 7022
+    },
+    {
+      "epoch": 5.286413248024087,
+      "grad_norm": 0.1801423281431198,
+      "learning_rate": 1.5050756032866187e-05,
+      "loss": 0.1479,
+      "step": 7023
+    },
+    {
+      "epoch": 5.287165976665412,
+      "grad_norm": 0.20503823459148407,
+      "learning_rate": 1.5044558814252513e-05,
+      "loss": 0.1921,
+      "step": 7024
+    },
+    {
+      "epoch": 5.287918705306737,
+      "grad_norm": 0.1832212507724762,
+      "learning_rate": 1.5038362957747171e-05,
+      "loss": 0.1504,
+      "step": 7025
+    },
+    {
+      "epoch": 5.288671433948061,
+      "grad_norm": 0.17600169777870178,
+      "learning_rate": 1.5032168464027707e-05,
+      "loss": 0.149,
+      "step": 7026
+    },
+    {
+      "epoch": 5.289424162589387,
+      "grad_norm": 0.22866705060005188,
+      "learning_rate": 1.5025975333771559e-05,
+      "loss": 0.1432,
+      "step": 7027
+    },
+    {
+      "epoch": 5.290176891230711,
+      "grad_norm": 0.24993203580379486,
+      "learning_rate": 1.5019783567655999e-05,
+      "loss": 0.2236,
+      "step": 7028
+    },
+    {
+      "epoch": 5.290929619872037,
+      "grad_norm": 0.18942596018314362,
+      "learning_rate": 1.5013593166358137e-05,
+      "loss": 0.203,
+      "step": 7029
+    },
+    {
+      "epoch": 5.291682348513361,
+      "grad_norm": 0.2293238639831543,
+      "learning_rate": 1.500740413055495e-05,
+      "loss": 0.1989,
+      "step": 7030
+    },
+    {
+      "epoch": 5.2924350771546855,
+      "grad_norm": 0.2121075540781021,
+      "learning_rate": 1.5001216460923268e-05,
+      "loss": 0.19,
+      "step": 7031
+    },
+    {
+      "epoch": 5.293187805796011,
+      "grad_norm": 0.23734773695468903,
+      "learning_rate": 1.4995030158139772e-05,
+      "loss": 0.2688,
+      "step": 7032
+    },
+    {
+      "epoch": 5.293940534437335,
+      "grad_norm": 0.18602204322814941,
+      "learning_rate": 1.4988845222880968e-05,
+      "loss": 0.1376,
+      "step": 7033
+    },
+    {
+      "epoch": 5.29469326307866,
+      "grad_norm": 0.21575391292572021,
+      "learning_rate": 1.4982661655823247e-05,
+      "loss": 0.2102,
+      "step": 7034
+    },
+    {
+      "epoch": 5.295445991719985,
+      "grad_norm": 0.19750544428825378,
+      "learning_rate": 1.4976479457642817e-05,
+      "loss": 0.1682,
+      "step": 7035
+    },
+    {
+      "epoch": 5.29619872036131,
+      "grad_norm": 0.20648150146007538,
+      "learning_rate": 1.497029862901578e-05,
+      "loss": 0.2085,
+      "step": 7036
+    },
+    {
+      "epoch": 5.296951449002634,
+      "grad_norm": 0.23581582307815552,
+      "learning_rate": 1.4964119170618024e-05,
+      "loss": 0.2381,
+      "step": 7037
+    },
+    {
+      "epoch": 5.2977041776439595,
+      "grad_norm": 0.26071226596832275,
+      "learning_rate": 1.495794108312537e-05,
+      "loss": 0.223,
+      "step": 7038
+    },
+    {
+      "epoch": 5.298456906285284,
+      "grad_norm": 0.17283326387405396,
+      "learning_rate": 1.4951764367213409e-05,
+      "loss": 0.2125,
+      "step": 7039
+    },
+    {
+      "epoch": 5.299209634926609,
+      "grad_norm": 0.23084644973278046,
+      "learning_rate": 1.4945589023557643e-05,
+      "loss": 0.2353,
+      "step": 7040
+    },
+    {
+      "epoch": 5.299962363567934,
+      "grad_norm": 0.19100946187973022,
+      "learning_rate": 1.4939415052833378e-05,
+      "loss": 0.1939,
+      "step": 7041
+    },
+    {
+      "epoch": 5.300715092209258,
+      "grad_norm": 0.19910500943660736,
+      "learning_rate": 1.4933242455715802e-05,
+      "loss": 0.1824,
+      "step": 7042
+    },
+    {
+      "epoch": 5.301467820850584,
+      "grad_norm": 0.1866569221019745,
+      "learning_rate": 1.4927071232879922e-05,
+      "loss": 0.159,
+      "step": 7043
+    },
+    {
+      "epoch": 5.302220549491908,
+      "grad_norm": 0.16818678379058838,
+      "learning_rate": 1.492090138500065e-05,
+      "loss": 0.1501,
+      "step": 7044
+    },
+    {
+      "epoch": 5.302973278133233,
+      "grad_norm": 0.2305917888879776,
+      "learning_rate": 1.4914732912752677e-05,
+      "loss": 0.1695,
+      "step": 7045
+    },
+    {
+      "epoch": 5.303726006774558,
+      "grad_norm": 0.20075209438800812,
+      "learning_rate": 1.4908565816810599e-05,
+      "loss": 0.1955,
+      "step": 7046
+    },
+    {
+      "epoch": 5.304478735415882,
+      "grad_norm": 0.18637825548648834,
+      "learning_rate": 1.4902400097848829e-05,
+      "loss": 0.2421,
+      "step": 7047
+    },
+    {
+      "epoch": 5.305231464057208,
+      "grad_norm": 0.22888267040252686,
+      "learning_rate": 1.489623575654164e-05,
+      "loss": 0.1583,
+      "step": 7048
+    },
+    {
+      "epoch": 5.305984192698532,
+      "grad_norm": 0.16641362011432648,
+      "learning_rate": 1.4890072793563159e-05,
+      "loss": 0.2072,
+      "step": 7049
+    },
+    {
+      "epoch": 5.306736921339857,
+      "grad_norm": 0.19683551788330078,
+      "learning_rate": 1.4883911209587368e-05,
+      "loss": 0.1932,
+      "step": 7050
+    },
+    {
+      "epoch": 5.307489649981182,
+      "grad_norm": 0.18836228549480438,
+      "learning_rate": 1.487775100528807e-05,
+      "loss": 0.1345,
+      "step": 7051
+    },
+    {
+      "epoch": 5.308242378622507,
+      "grad_norm": 0.1765003800392151,
+      "learning_rate": 1.4871592181338947e-05,
+      "loss": 0.2163,
+      "step": 7052
+    },
+    {
+      "epoch": 5.308995107263831,
+      "grad_norm": 0.1940252035856247,
+      "learning_rate": 1.4865434738413524e-05,
+      "loss": 0.217,
+      "step": 7053
+    },
+    {
+      "epoch": 5.309747835905156,
+      "grad_norm": 0.1754763126373291,
+      "learning_rate": 1.4859278677185152e-05,
+      "loss": 0.2201,
+      "step": 7054
+    },
+    {
+      "epoch": 5.310500564546481,
+      "grad_norm": 0.18737073242664337,
+      "learning_rate": 1.4853123998327068e-05,
+      "loss": 0.2713,
+      "step": 7055
+    },
+    {
+      "epoch": 5.311253293187805,
+      "grad_norm": 0.2172868400812149,
+      "learning_rate": 1.4846970702512314e-05,
+      "loss": 0.1543,
+      "step": 7056
+    },
+    {
+      "epoch": 5.312006021829131,
+      "grad_norm": 0.2035790979862213,
+      "learning_rate": 1.4840818790413836e-05,
+      "loss": 0.1923,
+      "step": 7057
+    },
+    {
+      "epoch": 5.312758750470455,
+      "grad_norm": 0.17060494422912598,
+      "learning_rate": 1.4834668262704376e-05,
+      "loss": 0.2074,
+      "step": 7058
+    },
+    {
+      "epoch": 5.313511479111781,
+      "grad_norm": 0.1946296989917755,
+      "learning_rate": 1.4828519120056562e-05,
+      "loss": 0.1545,
+      "step": 7059
+    },
+    {
+      "epoch": 5.314264207753105,
+      "grad_norm": 0.19552679359912872,
+      "learning_rate": 1.4822371363142831e-05,
+      "loss": 0.1775,
+      "step": 7060
+    },
+    {
+      "epoch": 5.3150169363944295,
+      "grad_norm": 0.20634551346302032,
+      "learning_rate": 1.4816224992635518e-05,
+      "loss": 0.2223,
+      "step": 7061
+    },
+    {
+      "epoch": 5.315769665035755,
+      "grad_norm": 0.2062556892633438,
+      "learning_rate": 1.4810080009206756e-05,
+      "loss": 0.1616,
+      "step": 7062
+    },
+    {
+      "epoch": 5.316522393677079,
+      "grad_norm": 0.1937459260225296,
+      "learning_rate": 1.480393641352858e-05,
+      "loss": 0.1662,
+      "step": 7063
+    },
+    {
+      "epoch": 5.317275122318404,
+      "grad_norm": 0.1703866869211197,
+      "learning_rate": 1.4797794206272822e-05,
+      "loss": 0.1837,
+      "step": 7064
+    },
+    {
+      "epoch": 5.318027850959729,
+      "grad_norm": 0.2517925202846527,
+      "learning_rate": 1.4791653388111199e-05,
+      "loss": 0.2498,
+      "step": 7065
+    },
+    {
+      "epoch": 5.318780579601054,
+      "grad_norm": 0.2042030543088913,
+      "learning_rate": 1.4785513959715248e-05,
+      "loss": 0.178,
+      "step": 7066
+    },
+    {
+      "epoch": 5.319533308242379,
+      "grad_norm": 0.17216317355632782,
+      "learning_rate": 1.4779375921756384e-05,
+      "loss": 0.1225,
+      "step": 7067
+    },
+    {
+      "epoch": 5.3202860368837035,
+      "grad_norm": 0.20578257739543915,
+      "learning_rate": 1.4773239274905825e-05,
+      "loss": 0.2479,
+      "step": 7068
+    },
+    {
+      "epoch": 5.321038765525028,
+      "grad_norm": 0.16270186007022858,
+      "learning_rate": 1.4767104019834702e-05,
+      "loss": 0.1208,
+      "step": 7069
+    },
+    {
+      "epoch": 5.321791494166353,
+      "grad_norm": 0.22050556540489197,
+      "learning_rate": 1.4760970157213932e-05,
+      "loss": 0.195,
+      "step": 7070
+    },
+    {
+      "epoch": 5.322544222807678,
+      "grad_norm": 0.2171730250120163,
+      "learning_rate": 1.4754837687714318e-05,
+      "loss": 0.1827,
+      "step": 7071
+    },
+    {
+      "epoch": 5.323296951449002,
+      "grad_norm": 0.22486311197280884,
+      "learning_rate": 1.4748706612006488e-05,
+      "loss": 0.159,
+      "step": 7072
+    },
+    {
+      "epoch": 5.324049680090328,
+      "grad_norm": 0.20925194025039673,
+      "learning_rate": 1.4742576930760929e-05,
+      "loss": 0.1819,
+      "step": 7073
+    },
+    {
+      "epoch": 5.324802408731652,
+      "grad_norm": 0.2047904133796692,
+      "learning_rate": 1.4736448644647985e-05,
+      "loss": 0.2392,
+      "step": 7074
+    },
+    {
+      "epoch": 5.325555137372977,
+      "grad_norm": 0.2003067582845688,
+      "learning_rate": 1.4730321754337823e-05,
+      "loss": 0.1981,
+      "step": 7075
+    },
+    {
+      "epoch": 5.326307866014302,
+      "grad_norm": 0.1623067706823349,
+      "learning_rate": 1.472419626050047e-05,
+      "loss": 0.0845,
+      "step": 7076
+    },
+    {
+      "epoch": 5.327060594655626,
+      "grad_norm": 0.21141119301319122,
+      "learning_rate": 1.4718072163805807e-05,
+      "loss": 0.1347,
+      "step": 7077
+    },
+    {
+      "epoch": 5.327813323296952,
+      "grad_norm": 0.22544710338115692,
+      "learning_rate": 1.4711949464923564e-05,
+      "loss": 0.1634,
+      "step": 7078
+    },
+    {
+      "epoch": 5.328566051938276,
+      "grad_norm": 0.17989657819271088,
+      "learning_rate": 1.4705828164523294e-05,
+      "loss": 0.1505,
+      "step": 7079
+    },
+    {
+      "epoch": 5.329318780579601,
+      "grad_norm": 0.2130405157804489,
+      "learning_rate": 1.4699708263274426e-05,
+      "loss": 0.2395,
+      "step": 7080
+    },
+    {
+      "epoch": 5.330071509220926,
+      "grad_norm": 0.2624771296977997,
+      "learning_rate": 1.4693589761846203e-05,
+      "loss": 0.2513,
+      "step": 7081
+    },
+    {
+      "epoch": 5.330824237862251,
+      "grad_norm": 0.169358491897583,
+      "learning_rate": 1.4687472660907763e-05,
+      "loss": 0.1124,
+      "step": 7082
+    },
+    {
+      "epoch": 5.331576966503576,
+      "grad_norm": 0.215037003159523,
+      "learning_rate": 1.4681356961128042e-05,
+      "loss": 0.1985,
+      "step": 7083
+    },
+    {
+      "epoch": 5.3323296951449,
+      "grad_norm": 0.19925138354301453,
+      "learning_rate": 1.4675242663175858e-05,
+      "loss": 0.1545,
+      "step": 7084
+    },
+    {
+      "epoch": 5.333082423786225,
+      "grad_norm": 0.20614320039749146,
+      "learning_rate": 1.4669129767719853e-05,
+      "loss": 0.1255,
+      "step": 7085
+    },
+    {
+      "epoch": 5.33383515242755,
+      "grad_norm": 0.18944895267486572,
+      "learning_rate": 1.466301827542853e-05,
+      "loss": 0.1557,
+      "step": 7086
+    },
+    {
+      "epoch": 5.334587881068875,
+      "grad_norm": 0.17986108362674713,
+      "learning_rate": 1.4656908186970208e-05,
+      "loss": 0.1255,
+      "step": 7087
+    },
+    {
+      "epoch": 5.335340609710199,
+      "grad_norm": 0.20689785480499268,
+      "learning_rate": 1.4650799503013119e-05,
+      "loss": 0.1717,
+      "step": 7088
+    },
+    {
+      "epoch": 5.3360933383515246,
+      "grad_norm": 0.16772966086864471,
+      "learning_rate": 1.464469222422527e-05,
+      "loss": 0.1183,
+      "step": 7089
+    },
+    {
+      "epoch": 5.336846066992849,
+      "grad_norm": 0.24465054273605347,
+      "learning_rate": 1.4638586351274557e-05,
+      "loss": 0.2186,
+      "step": 7090
+    },
+    {
+      "epoch": 5.3375987956341735,
+      "grad_norm": 0.1923346221446991,
+      "learning_rate": 1.4632481884828696e-05,
+      "loss": 0.1431,
+      "step": 7091
+    },
+    {
+      "epoch": 5.338351524275499,
+      "grad_norm": 0.19449593126773834,
+      "learning_rate": 1.4626378825555281e-05,
+      "loss": 0.2277,
+      "step": 7092
+    },
+    {
+      "epoch": 5.339104252916823,
+      "grad_norm": 0.18906262516975403,
+      "learning_rate": 1.462027717412171e-05,
+      "loss": 0.1267,
+      "step": 7093
+    },
+    {
+      "epoch": 5.339856981558149,
+      "grad_norm": 0.2071405053138733,
+      "learning_rate": 1.4614176931195265e-05,
+      "loss": 0.2488,
+      "step": 7094
+    },
+    {
+      "epoch": 5.340609710199473,
+      "grad_norm": 0.1942216157913208,
+      "learning_rate": 1.4608078097443056e-05,
+      "loss": 0.2051,
+      "step": 7095
+    },
+    {
+      "epoch": 5.341362438840798,
+      "grad_norm": 0.21784649789333344,
+      "learning_rate": 1.4601980673532057e-05,
+      "loss": 0.1241,
+      "step": 7096
+    },
+    {
+      "epoch": 5.342115167482123,
+      "grad_norm": 0.22926223278045654,
+      "learning_rate": 1.4595884660129044e-05,
+      "loss": 0.2025,
+      "step": 7097
+    },
+    {
+      "epoch": 5.3428678961234475,
+      "grad_norm": 0.18237854540348053,
+      "learning_rate": 1.4589790057900687e-05,
+      "loss": 0.1461,
+      "step": 7098
+    },
+    {
+      "epoch": 5.343620624764772,
+      "grad_norm": 0.1853754222393036,
+      "learning_rate": 1.4583696867513489e-05,
+      "loss": 0.1888,
+      "step": 7099
+    },
+    {
+      "epoch": 5.344373353406097,
+      "grad_norm": 0.21313534677028656,
+      "learning_rate": 1.4577605089633773e-05,
+      "loss": 0.1906,
+      "step": 7100
+    },
+    {
+      "epoch": 5.345126082047422,
+      "grad_norm": 0.20669235289096832,
+      "learning_rate": 1.4571514724927731e-05,
+      "loss": 0.2401,
+      "step": 7101
+    },
+    {
+      "epoch": 5.345878810688747,
+      "grad_norm": 0.19437529146671295,
+      "learning_rate": 1.4565425774061406e-05,
+      "loss": 0.1159,
+      "step": 7102
+    },
+    {
+      "epoch": 5.346631539330072,
+      "grad_norm": 0.22650660574436188,
+      "learning_rate": 1.4559338237700681e-05,
+      "loss": 0.2228,
+      "step": 7103
+    },
+    {
+      "epoch": 5.347384267971396,
+      "grad_norm": 0.2044593244791031,
+      "learning_rate": 1.4553252116511256e-05,
+      "loss": 0.1967,
+      "step": 7104
+    },
+    {
+      "epoch": 5.3481369966127215,
+      "grad_norm": 0.2439054548740387,
+      "learning_rate": 1.454716741115873e-05,
+      "loss": 0.182,
+      "step": 7105
+    },
+    {
+      "epoch": 5.348889725254046,
+      "grad_norm": 0.18622389435768127,
+      "learning_rate": 1.4541084122308488e-05,
+      "loss": 0.1485,
+      "step": 7106
+    },
+    {
+      "epoch": 5.34964245389537,
+      "grad_norm": 0.19826997816562653,
+      "learning_rate": 1.4535002250625807e-05,
+      "loss": 0.1509,
+      "step": 7107
+    },
+    {
+      "epoch": 5.350395182536696,
+      "grad_norm": 0.18373285233974457,
+      "learning_rate": 1.4528921796775785e-05,
+      "loss": 0.1443,
+      "step": 7108
+    },
+    {
+      "epoch": 5.35114791117802,
+      "grad_norm": 0.22265581786632538,
+      "learning_rate": 1.4522842761423382e-05,
+      "loss": 0.2922,
+      "step": 7109
+    },
+    {
+      "epoch": 5.351900639819345,
+      "grad_norm": 0.200959250330925,
+      "learning_rate": 1.4516765145233374e-05,
+      "loss": 0.1649,
+      "step": 7110
+    },
+    {
+      "epoch": 5.35265336846067,
+      "grad_norm": 0.22381454706192017,
+      "learning_rate": 1.4510688948870418e-05,
+      "loss": 0.1979,
+      "step": 7111
+    },
+    {
+      "epoch": 5.3534060971019946,
+      "grad_norm": 0.20971792936325073,
+      "learning_rate": 1.4504614172998985e-05,
+      "loss": 0.2786,
+      "step": 7112
+    },
+    {
+      "epoch": 5.35415882574332,
+      "grad_norm": 0.17972469329833984,
+      "learning_rate": 1.4498540818283402e-05,
+      "loss": 0.1675,
+      "step": 7113
+    },
+    {
+      "epoch": 5.354911554384644,
+      "grad_norm": 0.2039380818605423,
+      "learning_rate": 1.4492468885387856e-05,
+      "loss": 0.1635,
+      "step": 7114
+    },
+    {
+      "epoch": 5.355664283025969,
+      "grad_norm": 0.18662422895431519,
+      "learning_rate": 1.4486398374976356e-05,
+      "loss": 0.2682,
+      "step": 7115
+    },
+    {
+      "epoch": 5.356417011667294,
+      "grad_norm": 0.1997787356376648,
+      "learning_rate": 1.4480329287712762e-05,
+      "loss": 0.182,
+      "step": 7116
+    },
+    {
+      "epoch": 5.357169740308619,
+      "grad_norm": 0.21743649244308472,
+      "learning_rate": 1.4474261624260785e-05,
+      "loss": 0.2537,
+      "step": 7117
+    },
+    {
+      "epoch": 5.357922468949943,
+      "grad_norm": 0.20365720987319946,
+      "learning_rate": 1.4468195385283967e-05,
+      "loss": 0.1679,
+      "step": 7118
+    },
+    {
+      "epoch": 5.3586751975912685,
+      "grad_norm": 0.18457716703414917,
+      "learning_rate": 1.446213057144571e-05,
+      "loss": 0.186,
+      "step": 7119
+    },
+    {
+      "epoch": 5.359427926232593,
+      "grad_norm": 0.2053317278623581,
+      "learning_rate": 1.445606718340925e-05,
+      "loss": 0.1797,
+      "step": 7120
+    },
+    {
+      "epoch": 5.360180654873918,
+      "grad_norm": 0.20178556442260742,
+      "learning_rate": 1.445000522183768e-05,
+      "loss": 0.2073,
+      "step": 7121
+    },
+    {
+      "epoch": 5.360933383515243,
+      "grad_norm": 0.16644586622714996,
+      "learning_rate": 1.4443944687393913e-05,
+      "loss": 0.1935,
+      "step": 7122
+    },
+    {
+      "epoch": 5.361686112156567,
+      "grad_norm": 0.18892385065555573,
+      "learning_rate": 1.4437885580740724e-05,
+      "loss": 0.2571,
+      "step": 7123
+    },
+    {
+      "epoch": 5.362438840797893,
+      "grad_norm": 0.17493543028831482,
+      "learning_rate": 1.4431827902540734e-05,
+      "loss": 0.1929,
+      "step": 7124
+    },
+    {
+      "epoch": 5.363191569439217,
+      "grad_norm": 0.19739550352096558,
+      "learning_rate": 1.442577165345639e-05,
+      "loss": 0.1268,
+      "step": 7125
+    },
+    {
+      "epoch": 5.363944298080542,
+      "grad_norm": 0.17616845667362213,
+      "learning_rate": 1.4419716834150004e-05,
+      "loss": 0.1244,
+      "step": 7126
+    },
+    {
+      "epoch": 5.364697026721867,
+      "grad_norm": 0.2277706265449524,
+      "learning_rate": 1.4413663445283717e-05,
+      "loss": 0.2139,
+      "step": 7127
+    },
+    {
+      "epoch": 5.3654497553631915,
+      "grad_norm": 0.19659298658370972,
+      "learning_rate": 1.4407611487519529e-05,
+      "loss": 0.2496,
+      "step": 7128
+    },
+    {
+      "epoch": 5.366202484004516,
+      "grad_norm": 0.17712216079235077,
+      "learning_rate": 1.4401560961519259e-05,
+      "loss": 0.1368,
+      "step": 7129
+    },
+    {
+      "epoch": 5.366955212645841,
+      "grad_norm": 0.24205520749092102,
+      "learning_rate": 1.4395511867944595e-05,
+      "loss": 0.1981,
+      "step": 7130
+    },
+    {
+      "epoch": 5.367707941287166,
+      "grad_norm": 0.20129051804542542,
+      "learning_rate": 1.4389464207457042e-05,
+      "loss": 0.2143,
+      "step": 7131
+    },
+    {
+      "epoch": 5.368460669928491,
+      "grad_norm": 0.19684669375419617,
+      "learning_rate": 1.4383417980717972e-05,
+      "loss": 0.1975,
+      "step": 7132
+    },
+    {
+      "epoch": 5.369213398569816,
+      "grad_norm": 0.21456775069236755,
+      "learning_rate": 1.4377373188388593e-05,
+      "loss": 0.1492,
+      "step": 7133
+    },
+    {
+      "epoch": 5.36996612721114,
+      "grad_norm": 0.1896369606256485,
+      "learning_rate": 1.437132983112996e-05,
+      "loss": 0.2363,
+      "step": 7134
+    },
+    {
+      "epoch": 5.3707188558524654,
+      "grad_norm": 0.1755903959274292,
+      "learning_rate": 1.4365287909602948e-05,
+      "loss": 0.1996,
+      "step": 7135
+    },
+    {
+      "epoch": 5.37147158449379,
+      "grad_norm": 0.1816374808549881,
+      "learning_rate": 1.4359247424468312e-05,
+      "loss": 0.18,
+      "step": 7136
+    },
+    {
+      "epoch": 5.372224313135114,
+      "grad_norm": 0.18020883202552795,
+      "learning_rate": 1.4353208376386612e-05,
+      "loss": 0.1843,
+      "step": 7137
+    },
+    {
+      "epoch": 5.37297704177644,
+      "grad_norm": 0.19209825992584229,
+      "learning_rate": 1.4347170766018281e-05,
+      "loss": 0.1767,
+      "step": 7138
+    },
+    {
+      "epoch": 5.373729770417764,
+      "grad_norm": 0.22499124705791473,
+      "learning_rate": 1.4341134594023576e-05,
+      "loss": 0.1343,
+      "step": 7139
+    },
+    {
+      "epoch": 5.37448249905909,
+      "grad_norm": 0.2528038024902344,
+      "learning_rate": 1.4335099861062618e-05,
+      "loss": 0.2537,
+      "step": 7140
+    },
+    {
+      "epoch": 5.375235227700414,
+      "grad_norm": 0.1637861281633377,
+      "learning_rate": 1.4329066567795338e-05,
+      "loss": 0.2095,
+      "step": 7141
+    },
+    {
+      "epoch": 5.3759879563417385,
+      "grad_norm": 0.205913245677948,
+      "learning_rate": 1.4323034714881545e-05,
+      "loss": 0.2161,
+      "step": 7142
+    },
+    {
+      "epoch": 5.376740684983064,
+      "grad_norm": 0.19953928887844086,
+      "learning_rate": 1.4317004302980852e-05,
+      "loss": 0.1593,
+      "step": 7143
+    },
+    {
+      "epoch": 5.377493413624388,
+      "grad_norm": 0.2024744302034378,
+      "learning_rate": 1.4310975332752752e-05,
+      "loss": 0.2178,
+      "step": 7144
+    },
+    {
+      "epoch": 5.378246142265713,
+      "grad_norm": 0.19638891518115997,
+      "learning_rate": 1.430494780485656e-05,
+      "loss": 0.1439,
+      "step": 7145
+    },
+    {
+      "epoch": 5.378998870907038,
+      "grad_norm": 0.19512617588043213,
+      "learning_rate": 1.4298921719951444e-05,
+      "loss": 0.1634,
+      "step": 7146
+    },
+    {
+      "epoch": 5.379751599548363,
+      "grad_norm": 0.19930823147296906,
+      "learning_rate": 1.4292897078696393e-05,
+      "loss": 0.1678,
+      "step": 7147
+    },
+    {
+      "epoch": 5.380504328189687,
+      "grad_norm": 0.18959708511829376,
+      "learning_rate": 1.4286873881750263e-05,
+      "loss": 0.1829,
+      "step": 7148
+    },
+    {
+      "epoch": 5.3812570568310125,
+      "grad_norm": 0.20268577337265015,
+      "learning_rate": 1.4280852129771744e-05,
+      "loss": 0.2164,
+      "step": 7149
+    },
+    {
+      "epoch": 5.382009785472337,
+      "grad_norm": 0.19090169668197632,
+      "learning_rate": 1.427483182341936e-05,
+      "loss": 0.1474,
+      "step": 7150
+    },
+    {
+      "epoch": 5.382762514113662,
+      "grad_norm": 0.17255139350891113,
+      "learning_rate": 1.4268812963351475e-05,
+      "loss": 0.2353,
+      "step": 7151
+    },
+    {
+      "epoch": 5.383515242754987,
+      "grad_norm": 0.17187456786632538,
+      "learning_rate": 1.4262795550226319e-05,
+      "loss": 0.1716,
+      "step": 7152
+    },
+    {
+      "epoch": 5.384267971396311,
+      "grad_norm": 0.17826908826828003,
+      "learning_rate": 1.4256779584701946e-05,
+      "loss": 0.1814,
+      "step": 7153
+    },
+    {
+      "epoch": 5.385020700037637,
+      "grad_norm": 0.18596187233924866,
+      "learning_rate": 1.4250765067436237e-05,
+      "loss": 0.1441,
+      "step": 7154
+    },
+    {
+      "epoch": 5.385773428678961,
+      "grad_norm": 0.18735076487064362,
+      "learning_rate": 1.424475199908695e-05,
+      "loss": 0.1002,
+      "step": 7155
+    },
+    {
+      "epoch": 5.3865261573202865,
+      "grad_norm": 0.19681280851364136,
+      "learning_rate": 1.4238740380311646e-05,
+      "loss": 0.1157,
+      "step": 7156
+    },
+    {
+      "epoch": 5.387278885961611,
+      "grad_norm": 0.18197348713874817,
+      "learning_rate": 1.4232730211767765e-05,
+      "loss": 0.1812,
+      "step": 7157
+    },
+    {
+      "epoch": 5.3880316146029354,
+      "grad_norm": 0.20749150216579437,
+      "learning_rate": 1.4226721494112544e-05,
+      "loss": 0.1696,
+      "step": 7158
+    },
+    {
+      "epoch": 5.388784343244261,
+      "grad_norm": 0.1863340586423874,
+      "learning_rate": 1.4220714228003123e-05,
+      "loss": 0.1444,
+      "step": 7159
+    },
+    {
+      "epoch": 5.389537071885585,
+      "grad_norm": 0.1931474357843399,
+      "learning_rate": 1.4214708414096417e-05,
+      "loss": 0.1998,
+      "step": 7160
+    },
+    {
+      "epoch": 5.39028980052691,
+      "grad_norm": 0.19004607200622559,
+      "learning_rate": 1.4208704053049232e-05,
+      "loss": 0.1456,
+      "step": 7161
+    },
+    {
+      "epoch": 5.391042529168235,
+      "grad_norm": 0.1884436011314392,
+      "learning_rate": 1.4202701145518188e-05,
+      "loss": 0.2143,
+      "step": 7162
+    },
+    {
+      "epoch": 5.39179525780956,
+      "grad_norm": 0.1702367067337036,
+      "learning_rate": 1.4196699692159754e-05,
+      "loss": 0.201,
+      "step": 7163
+    },
+    {
+      "epoch": 5.392547986450884,
+      "grad_norm": 0.2011524885892868,
+      "learning_rate": 1.419069969363023e-05,
+      "loss": 0.3021,
+      "step": 7164
+    },
+    {
+      "epoch": 5.393300715092209,
+      "grad_norm": 0.18272513151168823,
+      "learning_rate": 1.4184701150585788e-05,
+      "loss": 0.1586,
+      "step": 7165
+    },
+    {
+      "epoch": 5.394053443733534,
+      "grad_norm": 0.19126811623573303,
+      "learning_rate": 1.4178704063682408e-05,
+      "loss": 0.1357,
+      "step": 7166
+    },
+    {
+      "epoch": 5.394806172374859,
+      "grad_norm": 0.22560903429985046,
+      "learning_rate": 1.4172708433575926e-05,
+      "loss": 0.2383,
+      "step": 7167
+    },
+    {
+      "epoch": 5.395558901016184,
+      "grad_norm": 0.21093648672103882,
+      "learning_rate": 1.4166714260922004e-05,
+      "loss": 0.2404,
+      "step": 7168
+    },
+    {
+      "epoch": 5.396311629657508,
+      "grad_norm": 0.18835775554180145,
+      "learning_rate": 1.4160721546376169e-05,
+      "loss": 0.1515,
+      "step": 7169
+    },
+    {
+      "epoch": 5.397064358298834,
+      "grad_norm": 0.20655880868434906,
+      "learning_rate": 1.4154730290593766e-05,
+      "loss": 0.1488,
+      "step": 7170
+    },
+    {
+      "epoch": 5.397817086940158,
+      "grad_norm": 0.18501189351081848,
+      "learning_rate": 1.4148740494230006e-05,
+      "loss": 0.1273,
+      "step": 7171
+    },
+    {
+      "epoch": 5.3985698155814825,
+      "grad_norm": 0.1916174441576004,
+      "learning_rate": 1.4142752157939903e-05,
+      "loss": 0.1934,
+      "step": 7172
+    },
+    {
+      "epoch": 5.399322544222808,
+      "grad_norm": 0.18981176614761353,
+      "learning_rate": 1.4136765282378339e-05,
+      "loss": 0.1747,
+      "step": 7173
+    },
+    {
+      "epoch": 5.400075272864132,
+      "grad_norm": 0.16862978041172028,
+      "learning_rate": 1.4130779868200043e-05,
+      "loss": 0.1679,
+      "step": 7174
+    },
+    {
+      "epoch": 5.400828001505458,
+      "grad_norm": 0.22623518109321594,
+      "learning_rate": 1.412479591605955e-05,
+      "loss": 0.2156,
+      "step": 7175
+    },
+    {
+      "epoch": 5.401580730146782,
+      "grad_norm": 0.1807098090648651,
+      "learning_rate": 1.4118813426611278e-05,
+      "loss": 0.1722,
+      "step": 7176
+    },
+    {
+      "epoch": 5.402333458788107,
+      "grad_norm": 0.20130494236946106,
+      "learning_rate": 1.411283240050943e-05,
+      "loss": 0.2339,
+      "step": 7177
+    },
+    {
+      "epoch": 5.403086187429432,
+      "grad_norm": 0.19098949432373047,
+      "learning_rate": 1.4106852838408125e-05,
+      "loss": 0.1969,
+      "step": 7178
+    },
+    {
+      "epoch": 5.4038389160707565,
+      "grad_norm": 0.2118140310049057,
+      "learning_rate": 1.4100874740961245e-05,
+      "loss": 0.2421,
+      "step": 7179
+    },
+    {
+      "epoch": 5.404591644712081,
+      "grad_norm": 0.1864185482263565,
+      "learning_rate": 1.4094898108822569e-05,
+      "loss": 0.1266,
+      "step": 7180
+    },
+    {
+      "epoch": 5.405344373353406,
+      "grad_norm": 0.19771797955036163,
+      "learning_rate": 1.408892294264567e-05,
+      "loss": 0.1563,
+      "step": 7181
+    },
+    {
+      "epoch": 5.406097101994731,
+      "grad_norm": 0.20131993293762207,
+      "learning_rate": 1.4082949243084003e-05,
+      "loss": 0.1699,
+      "step": 7182
+    },
+    {
+      "epoch": 5.406849830636055,
+      "grad_norm": 0.17132991552352905,
+      "learning_rate": 1.4076977010790823e-05,
+      "loss": 0.1622,
+      "step": 7183
+    },
+    {
+      "epoch": 5.407602559277381,
+      "grad_norm": 0.195194274187088,
+      "learning_rate": 1.4071006246419268e-05,
+      "loss": 0.1847,
+      "step": 7184
+    },
+    {
+      "epoch": 5.408355287918705,
+      "grad_norm": 0.16813500225543976,
+      "learning_rate": 1.4065036950622273e-05,
+      "loss": 0.1735,
+      "step": 7185
+    },
+    {
+      "epoch": 5.4091080165600305,
+      "grad_norm": 0.2044905424118042,
+      "learning_rate": 1.4059069124052648e-05,
+      "loss": 0.2206,
+      "step": 7186
+    },
+    {
+      "epoch": 5.409860745201355,
+      "grad_norm": 0.16498751938343048,
+      "learning_rate": 1.4053102767363002e-05,
+      "loss": 0.196,
+      "step": 7187
+    },
+    {
+      "epoch": 5.410613473842679,
+      "grad_norm": 0.1768319457769394,
+      "learning_rate": 1.4047137881205835e-05,
+      "loss": 0.1104,
+      "step": 7188
+    },
+    {
+      "epoch": 5.411366202484005,
+      "grad_norm": 0.17732295393943787,
+      "learning_rate": 1.4041174466233423e-05,
+      "loss": 0.1901,
+      "step": 7189
+    },
+    {
+      "epoch": 5.412118931125329,
+      "grad_norm": 0.1669139415025711,
+      "learning_rate": 1.4035212523097955e-05,
+      "loss": 0.127,
+      "step": 7190
+    },
+    {
+      "epoch": 5.412871659766654,
+      "grad_norm": 0.20156601071357727,
+      "learning_rate": 1.4029252052451392e-05,
+      "loss": 0.2016,
+      "step": 7191
+    },
+    {
+      "epoch": 5.413624388407979,
+      "grad_norm": 0.17579135298728943,
+      "learning_rate": 1.402329305494558e-05,
+      "loss": 0.2549,
+      "step": 7192
+    },
+    {
+      "epoch": 5.414377117049304,
+      "grad_norm": 0.19776380062103271,
+      "learning_rate": 1.4017335531232167e-05,
+      "loss": 0.2293,
+      "step": 7193
+    },
+    {
+      "epoch": 5.415129845690629,
+      "grad_norm": 0.17277899384498596,
+      "learning_rate": 1.4011379481962669e-05,
+      "loss": 0.1199,
+      "step": 7194
+    },
+    {
+      "epoch": 5.415882574331953,
+      "grad_norm": 0.17900627851486206,
+      "learning_rate": 1.4005424907788444e-05,
+      "loss": 0.2085,
+      "step": 7195
+    },
+    {
+      "epoch": 5.416635302973278,
+      "grad_norm": 0.19168348610401154,
+      "learning_rate": 1.3999471809360651e-05,
+      "loss": 0.1675,
+      "step": 7196
+    },
+    {
+      "epoch": 5.417388031614603,
+      "grad_norm": 0.17970621585845947,
+      "learning_rate": 1.3993520187330322e-05,
+      "loss": 0.2138,
+      "step": 7197
+    },
+    {
+      "epoch": 5.418140760255928,
+      "grad_norm": 0.17799890041351318,
+      "learning_rate": 1.398757004234832e-05,
+      "loss": 0.1972,
+      "step": 7198
+    },
+    {
+      "epoch": 5.418893488897252,
+      "grad_norm": 0.18119210004806519,
+      "learning_rate": 1.3981621375065355e-05,
+      "loss": 0.2284,
+      "step": 7199
+    },
+    {
+      "epoch": 5.419646217538578,
+      "grad_norm": 0.1965169459581375,
+      "learning_rate": 1.3975674186131937e-05,
+      "loss": 0.2504,
+      "step": 7200
+    },
+    {
+      "epoch": 5.419646217538578,
+      "eval_loss": 0.21595826745033264,
+      "eval_runtime": 452.7794,
+      "eval_samples_per_second": 21.262,
+      "eval_steps_per_second": 0.665,
+      "step": 7200
+    },
+    {
+      "epoch": 5.420398946179902,
+      "grad_norm": 0.21930043399333954,
+      "learning_rate": 1.3969728476198468e-05,
+      "loss": 0.1585,
+      "step": 7201
+    },
+    {
+      "epoch": 5.4211516748212265,
+      "grad_norm": 0.21137551963329315,
+      "learning_rate": 1.3963784245915135e-05,
+      "loss": 0.2235,
+      "step": 7202
+    },
+    {
+      "epoch": 5.421904403462552,
+      "grad_norm": 0.17491061985492706,
+      "learning_rate": 1.3957841495932016e-05,
+      "loss": 0.2177,
+      "step": 7203
+    },
+    {
+      "epoch": 5.422657132103876,
+      "grad_norm": 0.18801142275333405,
+      "learning_rate": 1.3951900226898989e-05,
+      "loss": 0.1534,
+      "step": 7204
+    },
+    {
+      "epoch": 5.423409860745202,
+      "grad_norm": 0.1703701764345169,
+      "learning_rate": 1.3945960439465788e-05,
+      "loss": 0.1374,
+      "step": 7205
+    },
+    {
+      "epoch": 5.424162589386526,
+      "grad_norm": 0.18632866442203522,
+      "learning_rate": 1.3940022134281972e-05,
+      "loss": 0.1741,
+      "step": 7206
+    },
+    {
+      "epoch": 5.424915318027851,
+      "grad_norm": 0.19951243698596954,
+      "learning_rate": 1.3934085311996953e-05,
+      "loss": 0.1384,
+      "step": 7207
+    },
+    {
+      "epoch": 5.425668046669176,
+      "grad_norm": 0.22417795658111572,
+      "learning_rate": 1.3928149973259952e-05,
+      "loss": 0.2407,
+      "step": 7208
+    },
+    {
+      "epoch": 5.4264207753105005,
+      "grad_norm": 0.187375009059906,
+      "learning_rate": 1.392221611872008e-05,
+      "loss": 0.1618,
+      "step": 7209
+    },
+    {
+      "epoch": 5.427173503951825,
+      "grad_norm": 0.19695481657981873,
+      "learning_rate": 1.3916283749026232e-05,
+      "loss": 0.2048,
+      "step": 7210
+    },
+    {
+      "epoch": 5.42792623259315,
+      "grad_norm": 0.19154901802539825,
+      "learning_rate": 1.391035286482718e-05,
+      "loss": 0.1894,
+      "step": 7211
+    },
+    {
+      "epoch": 5.428678961234475,
+      "grad_norm": 0.21060197055339813,
+      "learning_rate": 1.3904423466771498e-05,
+      "loss": 0.2032,
+      "step": 7212
+    },
+    {
+      "epoch": 5.4294316898758,
+      "grad_norm": 0.20022937655448914,
+      "learning_rate": 1.3898495555507635e-05,
+      "loss": 0.2327,
+      "step": 7213
+    },
+    {
+      "epoch": 5.430184418517125,
+      "grad_norm": 0.22886285185813904,
+      "learning_rate": 1.3892569131683837e-05,
+      "loss": 0.2547,
+      "step": 7214
+    },
+    {
+      "epoch": 5.430937147158449,
+      "grad_norm": 0.24600598216056824,
+      "learning_rate": 1.3886644195948223e-05,
+      "loss": 0.2342,
+      "step": 7215
+    },
+    {
+      "epoch": 5.4316898757997745,
+      "grad_norm": 0.18932849168777466,
+      "learning_rate": 1.3880720748948731e-05,
+      "loss": 0.1087,
+      "step": 7216
+    },
+    {
+      "epoch": 5.432442604441099,
+      "grad_norm": 0.18632282316684723,
+      "learning_rate": 1.3874798791333149e-05,
+      "loss": 0.2768,
+      "step": 7217
+    },
+    {
+      "epoch": 5.433195333082423,
+      "grad_norm": 0.19194495677947998,
+      "learning_rate": 1.386887832374908e-05,
+      "loss": 0.1594,
+      "step": 7218
+    },
+    {
+      "epoch": 5.433948061723749,
+      "grad_norm": 0.19552579522132874,
+      "learning_rate": 1.3862959346843982e-05,
+      "loss": 0.2602,
+      "step": 7219
+    },
+    {
+      "epoch": 5.434700790365073,
+      "grad_norm": 0.1939883679151535,
+      "learning_rate": 1.3857041861265152e-05,
+      "loss": 0.1675,
+      "step": 7220
+    },
+    {
+      "epoch": 5.435453519006398,
+      "grad_norm": 0.19377519190311432,
+      "learning_rate": 1.385112586765971e-05,
+      "loss": 0.1943,
+      "step": 7221
+    },
+    {
+      "epoch": 5.436206247647723,
+      "grad_norm": 0.1887626051902771,
+      "learning_rate": 1.3845211366674618e-05,
+      "loss": 0.1473,
+      "step": 7222
+    },
+    {
+      "epoch": 5.436958976289048,
+      "grad_norm": 0.18266171216964722,
+      "learning_rate": 1.3839298358956684e-05,
+      "loss": 0.2417,
+      "step": 7223
+    },
+    {
+      "epoch": 5.437711704930373,
+      "grad_norm": 0.19056160748004913,
+      "learning_rate": 1.3833386845152551e-05,
+      "loss": 0.1729,
+      "step": 7224
+    },
+    {
+      "epoch": 5.438464433571697,
+      "grad_norm": 0.1773747205734253,
+      "learning_rate": 1.3827476825908674e-05,
+      "loss": 0.1649,
+      "step": 7225
+    },
+    {
+      "epoch": 5.439217162213022,
+      "grad_norm": 0.1634763926267624,
+      "learning_rate": 1.3821568301871384e-05,
+      "loss": 0.1066,
+      "step": 7226
+    },
+    {
+      "epoch": 5.439969890854347,
+      "grad_norm": 0.18013648688793182,
+      "learning_rate": 1.3815661273686813e-05,
+      "loss": 0.1311,
+      "step": 7227
+    },
+    {
+      "epoch": 5.440722619495672,
+      "grad_norm": 0.1845812201499939,
+      "learning_rate": 1.3809755742000951e-05,
+      "loss": 0.1576,
+      "step": 7228
+    },
+    {
+      "epoch": 5.441475348136997,
+      "grad_norm": 0.1954963654279709,
+      "learning_rate": 1.3803851707459617e-05,
+      "loss": 0.1915,
+      "step": 7229
+    },
+    {
+      "epoch": 5.4422280767783215,
+      "grad_norm": 0.20964182913303375,
+      "learning_rate": 1.3797949170708477e-05,
+      "loss": 0.149,
+      "step": 7230
+    },
+    {
+      "epoch": 5.442980805419646,
+      "grad_norm": 0.18887494504451752,
+      "learning_rate": 1.3792048132393007e-05,
+      "loss": 0.2076,
+      "step": 7231
+    },
+    {
+      "epoch": 5.443733534060971,
+      "grad_norm": 0.2363666445016861,
+      "learning_rate": 1.3786148593158551e-05,
+      "loss": 0.182,
+      "step": 7232
+    },
+    {
+      "epoch": 5.444486262702296,
+      "grad_norm": 0.17761817574501038,
+      "learning_rate": 1.378025055365026e-05,
+      "loss": 0.134,
+      "step": 7233
+    },
+    {
+      "epoch": 5.44523899134362,
+      "grad_norm": 0.19007883965969086,
+      "learning_rate": 1.3774354014513141e-05,
+      "loss": 0.2257,
+      "step": 7234
+    },
+    {
+      "epoch": 5.445991719984946,
+      "grad_norm": 0.17882327735424042,
+      "learning_rate": 1.376845897639203e-05,
+      "loss": 0.1476,
+      "step": 7235
+    },
+    {
+      "epoch": 5.44674444862627,
+      "grad_norm": 0.18957453966140747,
+      "learning_rate": 1.3762565439931606e-05,
+      "loss": 0.2144,
+      "step": 7236
+    },
+    {
+      "epoch": 5.447497177267595,
+      "grad_norm": 0.20013917982578278,
+      "learning_rate": 1.3756673405776366e-05,
+      "loss": 0.2012,
+      "step": 7237
+    },
+    {
+      "epoch": 5.44824990590892,
+      "grad_norm": 0.17915308475494385,
+      "learning_rate": 1.3750782874570667e-05,
+      "loss": 0.1684,
+      "step": 7238
+    },
+    {
+      "epoch": 5.4490026345502445,
+      "grad_norm": 0.21598650515079498,
+      "learning_rate": 1.3744893846958673e-05,
+      "loss": 0.2396,
+      "step": 7239
+    },
+    {
+      "epoch": 5.44975536319157,
+      "grad_norm": 0.19694387912750244,
+      "learning_rate": 1.3739006323584407e-05,
+      "loss": 0.1916,
+      "step": 7240
+    },
+    {
+      "epoch": 5.450508091832894,
+      "grad_norm": 0.2145547866821289,
+      "learning_rate": 1.3733120305091719e-05,
+      "loss": 0.3078,
+      "step": 7241
+    },
+    {
+      "epoch": 5.451260820474219,
+      "grad_norm": 0.20230130851268768,
+      "learning_rate": 1.3727235792124308e-05,
+      "loss": 0.1452,
+      "step": 7242
+    },
+    {
+      "epoch": 5.452013549115544,
+      "grad_norm": 0.16987331211566925,
+      "learning_rate": 1.3721352785325673e-05,
+      "loss": 0.1325,
+      "step": 7243
+    },
+    {
+      "epoch": 5.452766277756869,
+      "grad_norm": 0.17687532305717468,
+      "learning_rate": 1.3715471285339183e-05,
+      "loss": 0.2283,
+      "step": 7244
+    },
+    {
+      "epoch": 5.453519006398193,
+      "grad_norm": 0.20404726266860962,
+      "learning_rate": 1.3709591292808035e-05,
+      "loss": 0.2584,
+      "step": 7245
+    },
+    {
+      "epoch": 5.4542717350395185,
+      "grad_norm": 0.1943594366312027,
+      "learning_rate": 1.3703712808375241e-05,
+      "loss": 0.1844,
+      "step": 7246
+    },
+    {
+      "epoch": 5.455024463680843,
+      "grad_norm": 0.18738201260566711,
+      "learning_rate": 1.3697835832683676e-05,
+      "loss": 0.2702,
+      "step": 7247
+    },
+    {
+      "epoch": 5.455777192322168,
+      "grad_norm": 0.1585284173488617,
+      "learning_rate": 1.3691960366376028e-05,
+      "loss": 0.1772,
+      "step": 7248
+    },
+    {
+      "epoch": 5.456529920963493,
+      "grad_norm": 0.21412749588489532,
+      "learning_rate": 1.3686086410094845e-05,
+      "loss": 0.1902,
+      "step": 7249
+    },
+    {
+      "epoch": 5.457282649604817,
+      "grad_norm": 0.178921177983284,
+      "learning_rate": 1.3680213964482475e-05,
+      "loss": 0.2647,
+      "step": 7250
+    },
+    {
+      "epoch": 5.458035378246143,
+      "grad_norm": 0.2162439376115799,
+      "learning_rate": 1.3674343030181136e-05,
+      "loss": 0.2082,
+      "step": 7251
+    },
+    {
+      "epoch": 5.458788106887467,
+      "grad_norm": 0.20377077162265778,
+      "learning_rate": 1.366847360783285e-05,
+      "loss": 0.1262,
+      "step": 7252
+    },
+    {
+      "epoch": 5.4595408355287915,
+      "grad_norm": 0.21603113412857056,
+      "learning_rate": 1.3662605698079495e-05,
+      "loss": 0.2056,
+      "step": 7253
+    },
+    {
+      "epoch": 5.460293564170117,
+      "grad_norm": 0.19674977660179138,
+      "learning_rate": 1.3656739301562771e-05,
+      "loss": 0.0892,
+      "step": 7254
+    },
+    {
+      "epoch": 5.461046292811441,
+      "grad_norm": 0.18155518174171448,
+      "learning_rate": 1.3650874418924236e-05,
+      "loss": 0.2465,
+      "step": 7255
+    },
+    {
+      "epoch": 5.461799021452766,
+      "grad_norm": 0.18641817569732666,
+      "learning_rate": 1.3645011050805245e-05,
+      "loss": 0.2633,
+      "step": 7256
+    },
+    {
+      "epoch": 5.462551750094091,
+      "grad_norm": 0.17374937236309052,
+      "learning_rate": 1.3639149197847024e-05,
+      "loss": 0.1627,
+      "step": 7257
+    },
+    {
+      "epoch": 5.463304478735416,
+      "grad_norm": 0.19682958722114563,
+      "learning_rate": 1.3633288860690596e-05,
+      "loss": 0.1487,
+      "step": 7258
+    },
+    {
+      "epoch": 5.464057207376741,
+      "grad_norm": 0.18251357972621918,
+      "learning_rate": 1.3627430039976851e-05,
+      "loss": 0.2159,
+      "step": 7259
+    },
+    {
+      "epoch": 5.4648099360180655,
+      "grad_norm": 0.2404390424489975,
+      "learning_rate": 1.3621572736346505e-05,
+      "loss": 0.2134,
+      "step": 7260
+    },
+    {
+      "epoch": 5.46556266465939,
+      "grad_norm": 0.18374669551849365,
+      "learning_rate": 1.36157169504401e-05,
+      "loss": 0.1738,
+      "step": 7261
+    },
+    {
+      "epoch": 5.466315393300715,
+      "grad_norm": 0.1700722873210907,
+      "learning_rate": 1.3609862682898009e-05,
+      "loss": 0.1574,
+      "step": 7262
+    },
+    {
+      "epoch": 5.46706812194204,
+      "grad_norm": 0.22002525627613068,
+      "learning_rate": 1.3604009934360459e-05,
+      "loss": 0.1722,
+      "step": 7263
+    },
+    {
+      "epoch": 5.467820850583364,
+      "grad_norm": 0.1992993801832199,
+      "learning_rate": 1.3598158705467482e-05,
+      "loss": 0.2132,
+      "step": 7264
+    },
+    {
+      "epoch": 5.46857357922469,
+      "grad_norm": 0.2126915454864502,
+      "learning_rate": 1.3592308996858968e-05,
+      "loss": 0.1513,
+      "step": 7265
+    },
+    {
+      "epoch": 5.469326307866014,
+      "grad_norm": 0.18559753894805908,
+      "learning_rate": 1.3586460809174634e-05,
+      "loss": 0.2286,
+      "step": 7266
+    },
+    {
+      "epoch": 5.4700790365073395,
+      "grad_norm": 0.1928398162126541,
+      "learning_rate": 1.3580614143054032e-05,
+      "loss": 0.154,
+      "step": 7267
+    },
+    {
+      "epoch": 5.470831765148664,
+      "grad_norm": 0.19126413762569427,
+      "learning_rate": 1.3574768999136534e-05,
+      "loss": 0.2049,
+      "step": 7268
+    },
+    {
+      "epoch": 5.4715844937899885,
+      "grad_norm": 0.24722442030906677,
+      "learning_rate": 1.3568925378061365e-05,
+      "loss": 0.2316,
+      "step": 7269
+    },
+    {
+      "epoch": 5.472337222431314,
+      "grad_norm": 0.18367895483970642,
+      "learning_rate": 1.3563083280467575e-05,
+      "loss": 0.1571,
+      "step": 7270
+    },
+    {
+      "epoch": 5.473089951072638,
+      "grad_norm": 0.19855237007141113,
+      "learning_rate": 1.3557242706994033e-05,
+      "loss": 0.1489,
+      "step": 7271
+    },
+    {
+      "epoch": 5.473842679713963,
+      "grad_norm": 0.19684399664402008,
+      "learning_rate": 1.3551403658279472e-05,
+      "loss": 0.1709,
+      "step": 7272
+    },
+    {
+      "epoch": 5.474595408355288,
+      "grad_norm": 0.20004379749298096,
+      "learning_rate": 1.3545566134962433e-05,
+      "loss": 0.1999,
+      "step": 7273
+    },
+    {
+      "epoch": 5.475348136996613,
+      "grad_norm": 0.23555617034435272,
+      "learning_rate": 1.3539730137681311e-05,
+      "loss": 0.1869,
+      "step": 7274
+    },
+    {
+      "epoch": 5.476100865637937,
+      "grad_norm": 0.1878862828016281,
+      "learning_rate": 1.35338956670743e-05,
+      "loss": 0.1693,
+      "step": 7275
+    },
+    {
+      "epoch": 5.476853594279262,
+      "grad_norm": 0.2666642367839813,
+      "learning_rate": 1.3528062723779472e-05,
+      "loss": 0.2299,
+      "step": 7276
+    },
+    {
+      "epoch": 5.477606322920587,
+      "grad_norm": 0.2429923117160797,
+      "learning_rate": 1.352223130843469e-05,
+      "loss": 0.2478,
+      "step": 7277
+    },
+    {
+      "epoch": 5.478359051561912,
+      "grad_norm": 0.1909763365983963,
+      "learning_rate": 1.3516401421677676e-05,
+      "loss": 0.1246,
+      "step": 7278
+    },
+    {
+      "epoch": 5.479111780203237,
+      "grad_norm": 0.17455001175403595,
+      "learning_rate": 1.3510573064145981e-05,
+      "loss": 0.1019,
+      "step": 7279
+    },
+    {
+      "epoch": 5.479864508844561,
+      "grad_norm": 0.20228199660778046,
+      "learning_rate": 1.350474623647699e-05,
+      "loss": 0.1604,
+      "step": 7280
+    },
+    {
+      "epoch": 5.480617237485887,
+      "grad_norm": 0.18427889049053192,
+      "learning_rate": 1.3498920939307898e-05,
+      "loss": 0.2055,
+      "step": 7281
+    },
+    {
+      "epoch": 5.481369966127211,
+      "grad_norm": 0.1854594349861145,
+      "learning_rate": 1.3493097173275773e-05,
+      "loss": 0.1924,
+      "step": 7282
+    },
+    {
+      "epoch": 5.4821226947685355,
+      "grad_norm": 0.2095312923192978,
+      "learning_rate": 1.3487274939017475e-05,
+      "loss": 0.1792,
+      "step": 7283
+    },
+    {
+      "epoch": 5.482875423409861,
+      "grad_norm": 0.2121540755033493,
+      "learning_rate": 1.3481454237169724e-05,
+      "loss": 0.1736,
+      "step": 7284
+    },
+    {
+      "epoch": 5.483628152051185,
+      "grad_norm": 0.2243456095457077,
+      "learning_rate": 1.347563506836905e-05,
+      "loss": 0.2082,
+      "step": 7285
+    },
+    {
+      "epoch": 5.484380880692511,
+      "grad_norm": 0.1616576910018921,
+      "learning_rate": 1.3469817433251854e-05,
+      "loss": 0.1903,
+      "step": 7286
+    },
+    {
+      "epoch": 5.485133609333835,
+      "grad_norm": 0.2041795253753662,
+      "learning_rate": 1.3464001332454324e-05,
+      "loss": 0.1826,
+      "step": 7287
+    },
+    {
+      "epoch": 5.48588633797516,
+      "grad_norm": 0.21037228405475616,
+      "learning_rate": 1.3458186766612513e-05,
+      "loss": 0.2168,
+      "step": 7288
+    },
+    {
+      "epoch": 5.486639066616485,
+      "grad_norm": 0.19006240367889404,
+      "learning_rate": 1.3452373736362273e-05,
+      "loss": 0.1206,
+      "step": 7289
+    },
+    {
+      "epoch": 5.4873917952578095,
+      "grad_norm": 0.1786503940820694,
+      "learning_rate": 1.3446562242339325e-05,
+      "loss": 0.1745,
+      "step": 7290
+    },
+    {
+      "epoch": 5.488144523899134,
+      "grad_norm": 0.19398826360702515,
+      "learning_rate": 1.3440752285179197e-05,
+      "loss": 0.1363,
+      "step": 7291
+    },
+    {
+      "epoch": 5.488897252540459,
+      "grad_norm": 0.24177995324134827,
+      "learning_rate": 1.3434943865517275e-05,
+      "loss": 0.1542,
+      "step": 7292
+    },
+    {
+      "epoch": 5.489649981181784,
+      "grad_norm": 0.15108422935009003,
+      "learning_rate": 1.3429136983988736e-05,
+      "loss": 0.1319,
+      "step": 7293
+    },
+    {
+      "epoch": 5.490402709823109,
+      "grad_norm": 0.1941325068473816,
+      "learning_rate": 1.3423331641228617e-05,
+      "loss": 0.2152,
+      "step": 7294
+    },
+    {
+      "epoch": 5.491155438464434,
+      "grad_norm": 0.171546071767807,
+      "learning_rate": 1.3417527837871797e-05,
+      "loss": 0.1706,
+      "step": 7295
+    },
+    {
+      "epoch": 5.491908167105758,
+      "grad_norm": 0.18488675355911255,
+      "learning_rate": 1.3411725574552952e-05,
+      "loss": 0.1644,
+      "step": 7296
+    },
+    {
+      "epoch": 5.4926608957470835,
+      "grad_norm": 0.19173121452331543,
+      "learning_rate": 1.3405924851906626e-05,
+      "loss": 0.2399,
+      "step": 7297
+    },
+    {
+      "epoch": 5.493413624388408,
+      "grad_norm": 0.25105020403862,
+      "learning_rate": 1.3400125670567149e-05,
+      "loss": 0.129,
+      "step": 7298
+    },
+    {
+      "epoch": 5.494166353029732,
+      "grad_norm": 0.19497929513454437,
+      "learning_rate": 1.3394328031168749e-05,
+      "loss": 0.1834,
+      "step": 7299
+    },
+    {
+      "epoch": 5.494919081671058,
+      "grad_norm": 0.1850377768278122,
+      "learning_rate": 1.3388531934345416e-05,
+      "loss": 0.1425,
+      "step": 7300
+    },
+    {
+      "epoch": 5.495671810312382,
+      "grad_norm": 0.22694072127342224,
+      "learning_rate": 1.3382737380731028e-05,
+      "loss": 0.1948,
+      "step": 7301
+    },
+    {
+      "epoch": 5.496424538953708,
+      "grad_norm": 0.21046659350395203,
+      "learning_rate": 1.3376944370959244e-05,
+      "loss": 0.3094,
+      "step": 7302
+    },
+    {
+      "epoch": 5.497177267595032,
+      "grad_norm": 0.1923946887254715,
+      "learning_rate": 1.3371152905663598e-05,
+      "loss": 0.1524,
+      "step": 7303
+    },
+    {
+      "epoch": 5.497929996236357,
+      "grad_norm": 0.23392683267593384,
+      "learning_rate": 1.3365362985477408e-05,
+      "loss": 0.2008,
+      "step": 7304
+    },
+    {
+      "epoch": 5.498682724877682,
+      "grad_norm": 0.20400820672512054,
+      "learning_rate": 1.3359574611033893e-05,
+      "loss": 0.2125,
+      "step": 7305
+    },
+    {
+      "epoch": 5.499435453519006,
+      "grad_norm": 0.23428164422512054,
+      "learning_rate": 1.3353787782966027e-05,
+      "loss": 0.1885,
+      "step": 7306
+    },
+    {
+      "epoch": 5.500188182160331,
+      "grad_norm": 0.17140504717826843,
+      "learning_rate": 1.334800250190667e-05,
+      "loss": 0.2291,
+      "step": 7307
+    },
+    {
+      "epoch": 5.500940910801656,
+      "grad_norm": 0.18854740262031555,
+      "learning_rate": 1.3342218768488473e-05,
+      "loss": 0.1434,
+      "step": 7308
+    },
+    {
+      "epoch": 5.501693639442981,
+      "grad_norm": 0.19387058913707733,
+      "learning_rate": 1.3336436583343952e-05,
+      "loss": 0.1492,
+      "step": 7309
+    },
+    {
+      "epoch": 5.502446368084305,
+      "grad_norm": 0.17738422751426697,
+      "learning_rate": 1.3330655947105419e-05,
+      "loss": 0.1577,
+      "step": 7310
+    },
+    {
+      "epoch": 5.503199096725631,
+      "grad_norm": 0.19897134602069855,
+      "learning_rate": 1.3324876860405066e-05,
+      "loss": 0.2128,
+      "step": 7311
+    },
+    {
+      "epoch": 5.503951825366955,
+      "grad_norm": 0.20817190408706665,
+      "learning_rate": 1.3319099323874857e-05,
+      "loss": 0.1696,
+      "step": 7312
+    },
+    {
+      "epoch": 5.5047045540082795,
+      "grad_norm": 0.19864696264266968,
+      "learning_rate": 1.3313323338146633e-05,
+      "loss": 0.24,
+      "step": 7313
+    },
+    {
+      "epoch": 5.505457282649605,
+      "grad_norm": 0.19964192807674408,
+      "learning_rate": 1.330754890385203e-05,
+      "loss": 0.265,
+      "step": 7314
+    },
+    {
+      "epoch": 5.506210011290929,
+      "grad_norm": 0.14874468743801117,
+      "learning_rate": 1.3301776021622542e-05,
+      "loss": 0.1608,
+      "step": 7315
+    },
+    {
+      "epoch": 5.506962739932255,
+      "grad_norm": 0.1802331954240799,
+      "learning_rate": 1.3296004692089491e-05,
+      "loss": 0.2517,
+      "step": 7316
+    },
+    {
+      "epoch": 5.507715468573579,
+      "grad_norm": 0.18412497639656067,
+      "learning_rate": 1.3290234915883995e-05,
+      "loss": 0.1865,
+      "step": 7317
+    },
+    {
+      "epoch": 5.508468197214904,
+      "grad_norm": 0.21239297091960907,
+      "learning_rate": 1.328446669363706e-05,
+      "loss": 0.2342,
+      "step": 7318
+    },
+    {
+      "epoch": 5.509220925856229,
+      "grad_norm": 0.1978355348110199,
+      "learning_rate": 1.3278700025979468e-05,
+      "loss": 0.2589,
+      "step": 7319
+    },
+    {
+      "epoch": 5.5099736544975535,
+      "grad_norm": 0.1599096655845642,
+      "learning_rate": 1.3272934913541865e-05,
+      "loss": 0.1379,
+      "step": 7320
+    },
+    {
+      "epoch": 5.510726383138879,
+      "grad_norm": 0.19019776582717896,
+      "learning_rate": 1.32671713569547e-05,
+      "loss": 0.1505,
+      "step": 7321
+    },
+    {
+      "epoch": 5.511479111780203,
+      "grad_norm": 0.17627525329589844,
+      "learning_rate": 1.3261409356848292e-05,
+      "loss": 0.1647,
+      "step": 7322
+    },
+    {
+      "epoch": 5.512231840421528,
+      "grad_norm": 0.1823684573173523,
+      "learning_rate": 1.3255648913852725e-05,
+      "loss": 0.1782,
+      "step": 7323
+    },
+    {
+      "epoch": 5.512984569062853,
+      "grad_norm": 0.1932966411113739,
+      "learning_rate": 1.3249890028598e-05,
+      "loss": 0.1547,
+      "step": 7324
+    },
+    {
+      "epoch": 5.513737297704178,
+      "grad_norm": 0.18808379769325256,
+      "learning_rate": 1.3244132701713866e-05,
+      "loss": 0.2192,
+      "step": 7325
+    },
+    {
+      "epoch": 5.514490026345502,
+      "grad_norm": 0.18756070733070374,
+      "learning_rate": 1.3238376933829951e-05,
+      "loss": 0.182,
+      "step": 7326
+    },
+    {
+      "epoch": 5.5152427549868275,
+      "grad_norm": 0.19263732433319092,
+      "learning_rate": 1.3232622725575688e-05,
+      "loss": 0.2026,
+      "step": 7327
+    },
+    {
+      "epoch": 5.515995483628152,
+      "grad_norm": 0.192935049533844,
+      "learning_rate": 1.3226870077580361e-05,
+      "loss": 0.1659,
+      "step": 7328
+    },
+    {
+      "epoch": 5.516748212269476,
+      "grad_norm": 0.1999576985836029,
+      "learning_rate": 1.3221118990473048e-05,
+      "loss": 0.265,
+      "step": 7329
+    },
+    {
+      "epoch": 5.517500940910802,
+      "grad_norm": 0.19178614020347595,
+      "learning_rate": 1.3215369464882709e-05,
+      "loss": 0.1821,
+      "step": 7330
+    },
+    {
+      "epoch": 5.518253669552126,
+      "grad_norm": 0.196236714720726,
+      "learning_rate": 1.320962150143808e-05,
+      "loss": 0.1815,
+      "step": 7331
+    },
+    {
+      "epoch": 5.519006398193452,
+      "grad_norm": 0.20691655576229095,
+      "learning_rate": 1.3203875100767765e-05,
+      "loss": 0.2,
+      "step": 7332
+    },
+    {
+      "epoch": 5.519759126834776,
+      "grad_norm": 0.1571476012468338,
+      "learning_rate": 1.319813026350017e-05,
+      "loss": 0.1639,
+      "step": 7333
+    },
+    {
+      "epoch": 5.520511855476101,
+      "grad_norm": 0.1981617957353592,
+      "learning_rate": 1.3192386990263555e-05,
+      "loss": 0.2425,
+      "step": 7334
+    },
+    {
+      "epoch": 5.521264584117426,
+      "grad_norm": 0.2594397962093353,
+      "learning_rate": 1.3186645281685977e-05,
+      "loss": 0.2098,
+      "step": 7335
+    },
+    {
+      "epoch": 5.52201731275875,
+      "grad_norm": 0.19083410501480103,
+      "learning_rate": 1.318090513839535e-05,
+      "loss": 0.1196,
+      "step": 7336
+    },
+    {
+      "epoch": 5.522770041400076,
+      "grad_norm": 0.1865665167570114,
+      "learning_rate": 1.3175166561019412e-05,
+      "loss": 0.2583,
+      "step": 7337
+    },
+    {
+      "epoch": 5.5235227700414,
+      "grad_norm": 0.21611998975276947,
+      "learning_rate": 1.3169429550185722e-05,
+      "loss": 0.1692,
+      "step": 7338
+    },
+    {
+      "epoch": 5.524275498682725,
+      "grad_norm": 0.17772242426872253,
+      "learning_rate": 1.3163694106521677e-05,
+      "loss": 0.141,
+      "step": 7339
+    },
+    {
+      "epoch": 5.52502822732405,
+      "grad_norm": 0.1813821643590927,
+      "learning_rate": 1.3157960230654482e-05,
+      "loss": 0.2111,
+      "step": 7340
+    },
+    {
+      "epoch": 5.5257809559653746,
+      "grad_norm": 0.176394984126091,
+      "learning_rate": 1.3152227923211202e-05,
+      "loss": 0.1798,
+      "step": 7341
+    },
+    {
+      "epoch": 5.526533684606699,
+      "grad_norm": 0.19328799843788147,
+      "learning_rate": 1.314649718481869e-05,
+      "loss": 0.1717,
+      "step": 7342
+    },
+    {
+      "epoch": 5.527286413248024,
+      "grad_norm": 0.17985840141773224,
+      "learning_rate": 1.314076801610368e-05,
+      "loss": 0.1915,
+      "step": 7343
+    },
+    {
+      "epoch": 5.528039141889349,
+      "grad_norm": 0.17577001452445984,
+      "learning_rate": 1.3135040417692685e-05,
+      "loss": 0.1139,
+      "step": 7344
+    },
+    {
+      "epoch": 5.528791870530673,
+      "grad_norm": 0.19330216944217682,
+      "learning_rate": 1.3129314390212078e-05,
+      "loss": 0.2254,
+      "step": 7345
+    },
+    {
+      "epoch": 5.529544599171999,
+      "grad_norm": 0.18526896834373474,
+      "learning_rate": 1.3123589934288038e-05,
+      "loss": 0.1862,
+      "step": 7346
+    },
+    {
+      "epoch": 5.530297327813323,
+      "grad_norm": 0.18089312314987183,
+      "learning_rate": 1.3117867050546593e-05,
+      "loss": 0.194,
+      "step": 7347
+    },
+    {
+      "epoch": 5.531050056454648,
+      "grad_norm": 0.18138659000396729,
+      "learning_rate": 1.3112145739613576e-05,
+      "loss": 0.1591,
+      "step": 7348
+    },
+    {
+      "epoch": 5.531802785095973,
+      "grad_norm": 0.18556421995162964,
+      "learning_rate": 1.3106426002114672e-05,
+      "loss": 0.1137,
+      "step": 7349
+    },
+    {
+      "epoch": 5.5325555137372975,
+      "grad_norm": 0.22655165195465088,
+      "learning_rate": 1.3100707838675377e-05,
+      "loss": 0.186,
+      "step": 7350
+    },
+    {
+      "epoch": 5.533308242378623,
+      "grad_norm": 0.18901591002941132,
+      "learning_rate": 1.3094991249921035e-05,
+      "loss": 0.1748,
+      "step": 7351
+    },
+    {
+      "epoch": 5.534060971019947,
+      "grad_norm": 0.16947966814041138,
+      "learning_rate": 1.3089276236476782e-05,
+      "loss": 0.1545,
+      "step": 7352
+    },
+    {
+      "epoch": 5.534813699661272,
+      "grad_norm": 0.17557726800441742,
+      "learning_rate": 1.3083562798967621e-05,
+      "loss": 0.0868,
+      "step": 7353
+    },
+    {
+      "epoch": 5.535566428302597,
+      "grad_norm": 0.21784856915473938,
+      "learning_rate": 1.3077850938018352e-05,
+      "loss": 0.2225,
+      "step": 7354
+    },
+    {
+      "epoch": 5.536319156943922,
+      "grad_norm": 0.1635206639766693,
+      "learning_rate": 1.3072140654253618e-05,
+      "loss": 0.1516,
+      "step": 7355
+    },
+    {
+      "epoch": 5.537071885585247,
+      "grad_norm": 0.17751802504062653,
+      "learning_rate": 1.3066431948297892e-05,
+      "loss": 0.191,
+      "step": 7356
+    },
+    {
+      "epoch": 5.5378246142265715,
+      "grad_norm": 0.18115238845348358,
+      "learning_rate": 1.3060724820775477e-05,
+      "loss": 0.1177,
+      "step": 7357
+    },
+    {
+      "epoch": 5.538577342867896,
+      "grad_norm": 0.19745564460754395,
+      "learning_rate": 1.3055019272310476e-05,
+      "loss": 0.1382,
+      "step": 7358
+    },
+    {
+      "epoch": 5.539330071509221,
+      "grad_norm": 0.17493580281734467,
+      "learning_rate": 1.3049315303526852e-05,
+      "loss": 0.2449,
+      "step": 7359
+    },
+    {
+      "epoch": 5.540082800150546,
+      "grad_norm": 0.20749270915985107,
+      "learning_rate": 1.3043612915048386e-05,
+      "loss": 0.1875,
+      "step": 7360
+    },
+    {
+      "epoch": 5.54083552879187,
+      "grad_norm": 0.19132252037525177,
+      "learning_rate": 1.303791210749867e-05,
+      "loss": 0.1992,
+      "step": 7361
+    },
+    {
+      "epoch": 5.541588257433196,
+      "grad_norm": 0.19324536621570587,
+      "learning_rate": 1.3032212881501144e-05,
+      "loss": 0.1514,
+      "step": 7362
+    },
+    {
+      "epoch": 5.54234098607452,
+      "grad_norm": 0.19102966785430908,
+      "learning_rate": 1.3026515237679064e-05,
+      "loss": 0.1505,
+      "step": 7363
+    },
+    {
+      "epoch": 5.5430937147158446,
+      "grad_norm": 0.1829076111316681,
+      "learning_rate": 1.3020819176655524e-05,
+      "loss": 0.1327,
+      "step": 7364
+    },
+    {
+      "epoch": 5.54384644335717,
+      "grad_norm": 0.1511651873588562,
+      "learning_rate": 1.3015124699053419e-05,
+      "loss": 0.126,
+      "step": 7365
+    },
+    {
+      "epoch": 5.544599171998494,
+      "grad_norm": 0.22364069521427155,
+      "learning_rate": 1.3009431805495511e-05,
+      "loss": 0.1952,
+      "step": 7366
+    },
+    {
+      "epoch": 5.545351900639819,
+      "grad_norm": 0.18462219834327698,
+      "learning_rate": 1.3003740496604346e-05,
+      "loss": 0.2395,
+      "step": 7367
+    },
+    {
+      "epoch": 5.546104629281144,
+      "grad_norm": 0.20127324759960175,
+      "learning_rate": 1.2998050773002324e-05,
+      "loss": 0.173,
+      "step": 7368
+    },
+    {
+      "epoch": 5.546857357922469,
+      "grad_norm": 0.1855308711528778,
+      "learning_rate": 1.2992362635311669e-05,
+      "loss": 0.1655,
+      "step": 7369
+    },
+    {
+      "epoch": 5.547610086563794,
+      "grad_norm": 0.1822502464056015,
+      "learning_rate": 1.298667608415443e-05,
+      "loss": 0.1801,
+      "step": 7370
+    },
+    {
+      "epoch": 5.5483628152051185,
+      "grad_norm": 0.1858977973461151,
+      "learning_rate": 1.2980991120152466e-05,
+      "loss": 0.1909,
+      "step": 7371
+    },
+    {
+      "epoch": 5.549115543846443,
+      "grad_norm": 0.1899484097957611,
+      "learning_rate": 1.2975307743927493e-05,
+      "loss": 0.2155,
+      "step": 7372
+    },
+    {
+      "epoch": 5.549868272487768,
+      "grad_norm": 0.1908283680677414,
+      "learning_rate": 1.296962595610102e-05,
+      "loss": 0.1958,
+      "step": 7373
+    },
+    {
+      "epoch": 5.550621001129093,
+      "grad_norm": 0.18502889573574066,
+      "learning_rate": 1.2963945757294407e-05,
+      "loss": 0.1354,
+      "step": 7374
+    },
+    {
+      "epoch": 5.551373729770418,
+      "grad_norm": 0.1808599978685379,
+      "learning_rate": 1.2958267148128828e-05,
+      "loss": 0.1869,
+      "step": 7375
+    },
+    {
+      "epoch": 5.552126458411743,
+      "grad_norm": 0.18829120695590973,
+      "learning_rate": 1.2952590129225301e-05,
+      "loss": 0.1551,
+      "step": 7376
+    },
+    {
+      "epoch": 5.552879187053067,
+      "grad_norm": 0.21524029970169067,
+      "learning_rate": 1.2946914701204638e-05,
+      "loss": 0.1944,
+      "step": 7377
+    },
+    {
+      "epoch": 5.5536319156943925,
+      "grad_norm": 0.18873324990272522,
+      "learning_rate": 1.294124086468751e-05,
+      "loss": 0.1555,
+      "step": 7378
+    },
+    {
+      "epoch": 5.554384644335717,
+      "grad_norm": 0.16157987713813782,
+      "learning_rate": 1.2935568620294387e-05,
+      "loss": 0.1483,
+      "step": 7379
+    },
+    {
+      "epoch": 5.5551373729770415,
+      "grad_norm": 0.19039282202720642,
+      "learning_rate": 1.2929897968645577e-05,
+      "loss": 0.2765,
+      "step": 7380
+    },
+    {
+      "epoch": 5.555890101618367,
+      "grad_norm": 0.19746285676956177,
+      "learning_rate": 1.2924228910361224e-05,
+      "loss": 0.1837,
+      "step": 7381
+    },
+    {
+      "epoch": 5.556642830259691,
+      "grad_norm": 0.1647823452949524,
+      "learning_rate": 1.291856144606129e-05,
+      "loss": 0.1519,
+      "step": 7382
+    },
+    {
+      "epoch": 5.557395558901016,
+      "grad_norm": 0.1991378664970398,
+      "learning_rate": 1.2912895576365542e-05,
+      "loss": 0.1871,
+      "step": 7383
+    },
+    {
+      "epoch": 5.558148287542341,
+      "grad_norm": 0.18572144210338593,
+      "learning_rate": 1.2907231301893607e-05,
+      "loss": 0.2081,
+      "step": 7384
+    },
+    {
+      "epoch": 5.558901016183666,
+      "grad_norm": 0.19477732479572296,
+      "learning_rate": 1.2901568623264923e-05,
+      "loss": 0.12,
+      "step": 7385
+    },
+    {
+      "epoch": 5.559653744824991,
+      "grad_norm": 0.18076103925704956,
+      "learning_rate": 1.289590754109874e-05,
+      "loss": 0.1202,
+      "step": 7386
+    },
+    {
+      "epoch": 5.5604064734663154,
+      "grad_norm": 0.1979312151670456,
+      "learning_rate": 1.2890248056014145e-05,
+      "loss": 0.1829,
+      "step": 7387
+    },
+    {
+      "epoch": 5.56115920210764,
+      "grad_norm": 0.18525288999080658,
+      "learning_rate": 1.2884590168630061e-05,
+      "loss": 0.1491,
+      "step": 7388
+    },
+    {
+      "epoch": 5.561911930748965,
+      "grad_norm": 0.23742859065532684,
+      "learning_rate": 1.287893387956523e-05,
+      "loss": 0.2534,
+      "step": 7389
+    },
+    {
+      "epoch": 5.56266465939029,
+      "grad_norm": 0.22799602150917053,
+      "learning_rate": 1.2873279189438204e-05,
+      "loss": 0.1928,
+      "step": 7390
+    },
+    {
+      "epoch": 5.563417388031614,
+      "grad_norm": 0.18864475190639496,
+      "learning_rate": 1.2867626098867381e-05,
+      "loss": 0.2571,
+      "step": 7391
+    },
+    {
+      "epoch": 5.56417011667294,
+      "grad_norm": 0.22405734658241272,
+      "learning_rate": 1.2861974608470956e-05,
+      "loss": 0.2561,
+      "step": 7392
+    },
+    {
+      "epoch": 5.564922845314264,
+      "grad_norm": 0.20086652040481567,
+      "learning_rate": 1.285632471886698e-05,
+      "loss": 0.21,
+      "step": 7393
+    },
+    {
+      "epoch": 5.565675573955589,
+      "grad_norm": 0.18259483575820923,
+      "learning_rate": 1.2850676430673319e-05,
+      "loss": 0.1739,
+      "step": 7394
+    },
+    {
+      "epoch": 5.566428302596914,
+      "grad_norm": 0.19021472334861755,
+      "learning_rate": 1.284502974450767e-05,
+      "loss": 0.1744,
+      "step": 7395
+    },
+    {
+      "epoch": 5.567181031238238,
+      "grad_norm": 0.16126257181167603,
+      "learning_rate": 1.2839384660987521e-05,
+      "loss": 0.1829,
+      "step": 7396
+    },
+    {
+      "epoch": 5.567933759879564,
+      "grad_norm": 0.21083423495292664,
+      "learning_rate": 1.2833741180730233e-05,
+      "loss": 0.2003,
+      "step": 7397
+    },
+    {
+      "epoch": 5.568686488520888,
+      "grad_norm": 0.18226587772369385,
+      "learning_rate": 1.2828099304352954e-05,
+      "loss": 0.0859,
+      "step": 7398
+    },
+    {
+      "epoch": 5.569439217162213,
+      "grad_norm": 0.19376616179943085,
+      "learning_rate": 1.2822459032472672e-05,
+      "loss": 0.2094,
+      "step": 7399
+    },
+    {
+      "epoch": 5.570191945803538,
+      "grad_norm": 0.20046333968639374,
+      "learning_rate": 1.2816820365706206e-05,
+      "loss": 0.1794,
+      "step": 7400
+    },
+    {
+      "epoch": 5.570191945803538,
+      "eval_loss": 0.21361961960792542,
+      "eval_runtime": 452.8545,
+      "eval_samples_per_second": 21.258,
+      "eval_steps_per_second": 0.665,
+      "step": 7400
+    },
+    {
+      "epoch": 5.5709446744448625,
+      "grad_norm": 0.1778251975774765,
+      "learning_rate": 1.2811183304670196e-05,
+      "loss": 0.2087,
+      "step": 7401
+    },
+    {
+      "epoch": 5.571697403086187,
+      "grad_norm": 0.18530873954296112,
+      "learning_rate": 1.2805547849981086e-05,
+      "loss": 0.1847,
+      "step": 7402
+    },
+    {
+      "epoch": 5.572450131727512,
+      "grad_norm": 0.1835397332906723,
+      "learning_rate": 1.2799914002255179e-05,
+      "loss": 0.24,
+      "step": 7403
+    },
+    {
+      "epoch": 5.573202860368837,
+      "grad_norm": 0.2141299843788147,
+      "learning_rate": 1.2794281762108567e-05,
+      "loss": 0.2881,
+      "step": 7404
+    },
+    {
+      "epoch": 5.573955589010162,
+      "grad_norm": 0.1729857474565506,
+      "learning_rate": 1.2788651130157195e-05,
+      "loss": 0.2238,
+      "step": 7405
+    },
+    {
+      "epoch": 5.574708317651487,
+      "grad_norm": 0.1947253942489624,
+      "learning_rate": 1.2783022107016815e-05,
+      "loss": 0.1092,
+      "step": 7406
+    },
+    {
+      "epoch": 5.575461046292811,
+      "grad_norm": 0.1865909844636917,
+      "learning_rate": 1.277739469330302e-05,
+      "loss": 0.1044,
+      "step": 7407
+    },
+    {
+      "epoch": 5.5762137749341365,
+      "grad_norm": 0.2003934681415558,
+      "learning_rate": 1.2771768889631197e-05,
+      "loss": 0.2135,
+      "step": 7408
+    },
+    {
+      "epoch": 5.576966503575461,
+      "grad_norm": 0.1721356362104416,
+      "learning_rate": 1.2766144696616582e-05,
+      "loss": 0.1262,
+      "step": 7409
+    },
+    {
+      "epoch": 5.577719232216786,
+      "grad_norm": 0.21023748815059662,
+      "learning_rate": 1.276052211487424e-05,
+      "loss": 0.1296,
+      "step": 7410
+    },
+    {
+      "epoch": 5.578471960858111,
+      "grad_norm": 0.2152605652809143,
+      "learning_rate": 1.2754901145019035e-05,
+      "loss": 0.215,
+      "step": 7411
+    },
+    {
+      "epoch": 5.579224689499435,
+      "grad_norm": 0.17313726246356964,
+      "learning_rate": 1.274928178766567e-05,
+      "loss": 0.2073,
+      "step": 7412
+    },
+    {
+      "epoch": 5.579977418140761,
+      "grad_norm": 0.17078571021556854,
+      "learning_rate": 1.2743664043428671e-05,
+      "loss": 0.1034,
+      "step": 7413
+    },
+    {
+      "epoch": 5.580730146782085,
+      "grad_norm": 0.18193495273590088,
+      "learning_rate": 1.2738047912922396e-05,
+      "loss": 0.1779,
+      "step": 7414
+    },
+    {
+      "epoch": 5.58148287542341,
+      "grad_norm": 0.17942291498184204,
+      "learning_rate": 1.2732433396760995e-05,
+      "loss": 0.1727,
+      "step": 7415
+    },
+    {
+      "epoch": 5.582235604064735,
+      "grad_norm": 0.17590658366680145,
+      "learning_rate": 1.2726820495558484e-05,
+      "loss": 0.1812,
+      "step": 7416
+    },
+    {
+      "epoch": 5.582988332706059,
+      "grad_norm": 0.21957315504550934,
+      "learning_rate": 1.2721209209928667e-05,
+      "loss": 0.2096,
+      "step": 7417
+    },
+    {
+      "epoch": 5.583741061347384,
+      "grad_norm": 0.23197171092033386,
+      "learning_rate": 1.2715599540485196e-05,
+      "loss": 0.2468,
+      "step": 7418
+    },
+    {
+      "epoch": 5.584493789988709,
+      "grad_norm": 0.17242611944675446,
+      "learning_rate": 1.2709991487841518e-05,
+      "loss": 0.1607,
+      "step": 7419
+    },
+    {
+      "epoch": 5.585246518630034,
+      "grad_norm": 0.19498707354068756,
+      "learning_rate": 1.2704385052610952e-05,
+      "loss": 0.1223,
+      "step": 7420
+    },
+    {
+      "epoch": 5.585999247271358,
+      "grad_norm": 0.18118752539157867,
+      "learning_rate": 1.2698780235406582e-05,
+      "loss": 0.114,
+      "step": 7421
+    },
+    {
+      "epoch": 5.586751975912684,
+      "grad_norm": 0.17230193316936493,
+      "learning_rate": 1.2693177036841359e-05,
+      "loss": 0.1149,
+      "step": 7422
+    },
+    {
+      "epoch": 5.587504704554008,
+      "grad_norm": 0.20460085570812225,
+      "learning_rate": 1.2687575457528025e-05,
+      "loss": 0.1831,
+      "step": 7423
+    },
+    {
+      "epoch": 5.588257433195333,
+      "grad_norm": 0.20540638267993927,
+      "learning_rate": 1.268197549807918e-05,
+      "loss": 0.2224,
+      "step": 7424
+    },
+    {
+      "epoch": 5.589010161836658,
+      "grad_norm": 0.1865558922290802,
+      "learning_rate": 1.2676377159107194e-05,
+      "loss": 0.1502,
+      "step": 7425
+    },
+    {
+      "epoch": 5.589762890477982,
+      "grad_norm": 0.18675769865512848,
+      "learning_rate": 1.2670780441224334e-05,
+      "loss": 0.1203,
+      "step": 7426
+    },
+    {
+      "epoch": 5.590515619119308,
+      "grad_norm": 0.19951973855495453,
+      "learning_rate": 1.2665185345042624e-05,
+      "loss": 0.1943,
+      "step": 7427
+    },
+    {
+      "epoch": 5.591268347760632,
+      "grad_norm": 0.18778182566165924,
+      "learning_rate": 1.2659591871173947e-05,
+      "loss": 0.1018,
+      "step": 7428
+    },
+    {
+      "epoch": 5.592021076401958,
+      "grad_norm": 0.19287659227848053,
+      "learning_rate": 1.2654000020229984e-05,
+      "loss": 0.2442,
+      "step": 7429
+    },
+    {
+      "epoch": 5.592773805043282,
+      "grad_norm": 0.19104355573654175,
+      "learning_rate": 1.2648409792822255e-05,
+      "loss": 0.1414,
+      "step": 7430
+    },
+    {
+      "epoch": 5.5935265336846065,
+      "grad_norm": 0.17942045629024506,
+      "learning_rate": 1.2642821189562104e-05,
+      "loss": 0.1869,
+      "step": 7431
+    },
+    {
+      "epoch": 5.594279262325932,
+      "grad_norm": 0.19221758842468262,
+      "learning_rate": 1.26372342110607e-05,
+      "loss": 0.2202,
+      "step": 7432
+    },
+    {
+      "epoch": 5.595031990967256,
+      "grad_norm": 0.23682250082492828,
+      "learning_rate": 1.2631648857929007e-05,
+      "loss": 0.2731,
+      "step": 7433
+    },
+    {
+      "epoch": 5.595784719608581,
+      "grad_norm": 0.19556429982185364,
+      "learning_rate": 1.2626065130777842e-05,
+      "loss": 0.1576,
+      "step": 7434
+    },
+    {
+      "epoch": 5.596537448249906,
+      "grad_norm": 0.17187829315662384,
+      "learning_rate": 1.2620483030217838e-05,
+      "loss": 0.1737,
+      "step": 7435
+    },
+    {
+      "epoch": 5.597290176891231,
+      "grad_norm": 0.21789886057376862,
+      "learning_rate": 1.2614902556859431e-05,
+      "loss": 0.2329,
+      "step": 7436
+    },
+    {
+      "epoch": 5.598042905532555,
+      "grad_norm": 0.1836283951997757,
+      "learning_rate": 1.260932371131291e-05,
+      "loss": 0.1986,
+      "step": 7437
+    },
+    {
+      "epoch": 5.5987956341738805,
+      "grad_norm": 0.1860847920179367,
+      "learning_rate": 1.2603746494188339e-05,
+      "loss": 0.1394,
+      "step": 7438
+    },
+    {
+      "epoch": 5.599548362815205,
+      "grad_norm": 0.2005406767129898,
+      "learning_rate": 1.2598170906095678e-05,
+      "loss": 0.1952,
+      "step": 7439
+    },
+    {
+      "epoch": 5.600301091456529,
+      "grad_norm": 0.24868132174015045,
+      "learning_rate": 1.2592596947644633e-05,
+      "loss": 0.2004,
+      "step": 7440
+    },
+    {
+      "epoch": 5.601053820097855,
+      "grad_norm": 0.18140706419944763,
+      "learning_rate": 1.2587024619444777e-05,
+      "loss": 0.1822,
+      "step": 7441
+    },
+    {
+      "epoch": 5.601806548739179,
+      "grad_norm": 0.21997103095054626,
+      "learning_rate": 1.2581453922105485e-05,
+      "loss": 0.2339,
+      "step": 7442
+    },
+    {
+      "epoch": 5.602559277380505,
+      "grad_norm": 0.20099250972270966,
+      "learning_rate": 1.2575884856235968e-05,
+      "loss": 0.2143,
+      "step": 7443
+    },
+    {
+      "epoch": 5.603312006021829,
+      "grad_norm": 0.19935132563114166,
+      "learning_rate": 1.2570317422445226e-05,
+      "loss": 0.1988,
+      "step": 7444
+    },
+    {
+      "epoch": 5.604064734663154,
+      "grad_norm": 0.21528422832489014,
+      "learning_rate": 1.2564751621342141e-05,
+      "loss": 0.2007,
+      "step": 7445
+    },
+    {
+      "epoch": 5.604817463304479,
+      "grad_norm": 0.1880209594964981,
+      "learning_rate": 1.255918745353536e-05,
+      "loss": 0.2095,
+      "step": 7446
+    },
+    {
+      "epoch": 5.605570191945803,
+      "grad_norm": 0.2505228519439697,
+      "learning_rate": 1.255362491963338e-05,
+      "loss": 0.2312,
+      "step": 7447
+    },
+    {
+      "epoch": 5.606322920587129,
+      "grad_norm": 0.1517082154750824,
+      "learning_rate": 1.2548064020244505e-05,
+      "loss": 0.0933,
+      "step": 7448
+    },
+    {
+      "epoch": 5.607075649228453,
+      "grad_norm": 0.19462089240550995,
+      "learning_rate": 1.254250475597687e-05,
+      "loss": 0.1472,
+      "step": 7449
+    },
+    {
+      "epoch": 5.607828377869778,
+      "grad_norm": 0.21755358576774597,
+      "learning_rate": 1.2536947127438415e-05,
+      "loss": 0.147,
+      "step": 7450
+    },
+    {
+      "epoch": 5.608581106511103,
+      "grad_norm": 0.20430198311805725,
+      "learning_rate": 1.2531391135236937e-05,
+      "loss": 0.2279,
+      "step": 7451
+    },
+    {
+      "epoch": 5.609333835152428,
+      "grad_norm": 0.17820362746715546,
+      "learning_rate": 1.2525836779980013e-05,
+      "loss": 0.2062,
+      "step": 7452
+    },
+    {
+      "epoch": 5.610086563793752,
+      "grad_norm": 0.17442584037780762,
+      "learning_rate": 1.2520284062275076e-05,
+      "loss": 0.2461,
+      "step": 7453
+    },
+    {
+      "epoch": 5.610839292435077,
+      "grad_norm": 0.1845102161169052,
+      "learning_rate": 1.2514732982729341e-05,
+      "loss": 0.1634,
+      "step": 7454
+    },
+    {
+      "epoch": 5.611592021076402,
+      "grad_norm": 0.1893865168094635,
+      "learning_rate": 1.2509183541949881e-05,
+      "loss": 0.1402,
+      "step": 7455
+    },
+    {
+      "epoch": 5.612344749717726,
+      "grad_norm": 0.19211721420288086,
+      "learning_rate": 1.2503635740543573e-05,
+      "loss": 0.1365,
+      "step": 7456
+    },
+    {
+      "epoch": 5.613097478359052,
+      "grad_norm": 0.1904422491788864,
+      "learning_rate": 1.249808957911711e-05,
+      "loss": 0.1816,
+      "step": 7457
+    },
+    {
+      "epoch": 5.613850207000376,
+      "grad_norm": 0.17760400474071503,
+      "learning_rate": 1.2492545058277015e-05,
+      "loss": 0.1666,
+      "step": 7458
+    },
+    {
+      "epoch": 5.6146029356417015,
+      "grad_norm": 0.18795613944530487,
+      "learning_rate": 1.2487002178629628e-05,
+      "loss": 0.1809,
+      "step": 7459
+    },
+    {
+      "epoch": 5.615355664283026,
+      "grad_norm": 0.18067511916160583,
+      "learning_rate": 1.2481460940781122e-05,
+      "loss": 0.1202,
+      "step": 7460
+    },
+    {
+      "epoch": 5.6161083929243505,
+      "grad_norm": 0.18427909910678864,
+      "learning_rate": 1.2475921345337456e-05,
+      "loss": 0.1767,
+      "step": 7461
+    },
+    {
+      "epoch": 5.616861121565676,
+      "grad_norm": 0.196591317653656,
+      "learning_rate": 1.247038339290445e-05,
+      "loss": 0.168,
+      "step": 7462
+    },
+    {
+      "epoch": 5.617613850207,
+      "grad_norm": 0.1734241545200348,
+      "learning_rate": 1.246484708408771e-05,
+      "loss": 0.091,
+      "step": 7463
+    },
+    {
+      "epoch": 5.618366578848325,
+      "grad_norm": 0.19523678719997406,
+      "learning_rate": 1.2459312419492697e-05,
+      "loss": 0.201,
+      "step": 7464
+    },
+    {
+      "epoch": 5.61911930748965,
+      "grad_norm": 0.2181016355752945,
+      "learning_rate": 1.245377939972466e-05,
+      "loss": 0.1989,
+      "step": 7465
+    },
+    {
+      "epoch": 5.619872036130975,
+      "grad_norm": 0.201655313372612,
+      "learning_rate": 1.2448248025388693e-05,
+      "loss": 0.1652,
+      "step": 7466
+    },
+    {
+      "epoch": 5.6206247647723,
+      "grad_norm": 0.19085335731506348,
+      "learning_rate": 1.2442718297089686e-05,
+      "loss": 0.2513,
+      "step": 7467
+    },
+    {
+      "epoch": 5.6213774934136245,
+      "grad_norm": 0.1799253225326538,
+      "learning_rate": 1.2437190215432375e-05,
+      "loss": 0.1831,
+      "step": 7468
+    },
+    {
+      "epoch": 5.622130222054949,
+      "grad_norm": 0.19805099070072174,
+      "learning_rate": 1.2431663781021285e-05,
+      "loss": 0.1626,
+      "step": 7469
+    },
+    {
+      "epoch": 5.622882950696274,
+      "grad_norm": 0.17743965983390808,
+      "learning_rate": 1.242613899446079e-05,
+      "loss": 0.1377,
+      "step": 7470
+    },
+    {
+      "epoch": 5.623635679337599,
+      "grad_norm": 0.22164106369018555,
+      "learning_rate": 1.2420615856355072e-05,
+      "loss": 0.1674,
+      "step": 7471
+    },
+    {
+      "epoch": 5.624388407978923,
+      "grad_norm": 0.19644179940223694,
+      "learning_rate": 1.241509436730814e-05,
+      "loss": 0.306,
+      "step": 7472
+    },
+    {
+      "epoch": 5.625141136620249,
+      "grad_norm": 0.16608178615570068,
+      "learning_rate": 1.24095745279238e-05,
+      "loss": 0.1452,
+      "step": 7473
+    },
+    {
+      "epoch": 5.625893865261573,
+      "grad_norm": 0.18921247124671936,
+      "learning_rate": 1.2404056338805711e-05,
+      "loss": 0.1752,
+      "step": 7474
+    },
+    {
+      "epoch": 5.626646593902898,
+      "grad_norm": 0.19001884758472443,
+      "learning_rate": 1.2398539800557317e-05,
+      "loss": 0.2004,
+      "step": 7475
+    },
+    {
+      "epoch": 5.627399322544223,
+      "grad_norm": 0.16306453943252563,
+      "learning_rate": 1.2393024913781904e-05,
+      "loss": 0.1789,
+      "step": 7476
+    },
+    {
+      "epoch": 5.628152051185547,
+      "grad_norm": 0.19826847314834595,
+      "learning_rate": 1.2387511679082573e-05,
+      "loss": 0.2588,
+      "step": 7477
+    },
+    {
+      "epoch": 5.628904779826873,
+      "grad_norm": 0.18575677275657654,
+      "learning_rate": 1.2382000097062252e-05,
+      "loss": 0.194,
+      "step": 7478
+    },
+    {
+      "epoch": 5.629657508468197,
+      "grad_norm": 0.21184907853603363,
+      "learning_rate": 1.2376490168323663e-05,
+      "loss": 0.1947,
+      "step": 7479
+    },
+    {
+      "epoch": 5.630410237109522,
+      "grad_norm": 0.18341460824012756,
+      "learning_rate": 1.2370981893469372e-05,
+      "loss": 0.127,
+      "step": 7480
+    },
+    {
+      "epoch": 5.631162965750847,
+      "grad_norm": 0.18283693492412567,
+      "learning_rate": 1.2365475273101761e-05,
+      "loss": 0.2243,
+      "step": 7481
+    },
+    {
+      "epoch": 5.6319156943921715,
+      "grad_norm": 0.1772419512271881,
+      "learning_rate": 1.2359970307823011e-05,
+      "loss": 0.2157,
+      "step": 7482
+    },
+    {
+      "epoch": 5.632668423033497,
+      "grad_norm": 0.17591464519500732,
+      "learning_rate": 1.2354466998235149e-05,
+      "loss": 0.1257,
+      "step": 7483
+    },
+    {
+      "epoch": 5.633421151674821,
+      "grad_norm": 0.20040366053581238,
+      "learning_rate": 1.2348965344940006e-05,
+      "loss": 0.1159,
+      "step": 7484
+    },
+    {
+      "epoch": 5.634173880316146,
+      "grad_norm": 0.18799585103988647,
+      "learning_rate": 1.2343465348539237e-05,
+      "loss": 0.2137,
+      "step": 7485
+    },
+    {
+      "epoch": 5.634926608957471,
+      "grad_norm": 0.18342214822769165,
+      "learning_rate": 1.2337967009634303e-05,
+      "loss": 0.1958,
+      "step": 7486
+    },
+    {
+      "epoch": 5.635679337598796,
+      "grad_norm": 0.18644456565380096,
+      "learning_rate": 1.233247032882651e-05,
+      "loss": 0.1385,
+      "step": 7487
+    },
+    {
+      "epoch": 5.63643206624012,
+      "grad_norm": 0.21033735573291779,
+      "learning_rate": 1.2326975306716951e-05,
+      "loss": 0.1848,
+      "step": 7488
+    },
+    {
+      "epoch": 5.6371847948814455,
+      "grad_norm": 0.2006317377090454,
+      "learning_rate": 1.2321481943906559e-05,
+      "loss": 0.2402,
+      "step": 7489
+    },
+    {
+      "epoch": 5.63793752352277,
+      "grad_norm": 0.2028709203004837,
+      "learning_rate": 1.231599024099608e-05,
+      "loss": 0.2394,
+      "step": 7490
+    },
+    {
+      "epoch": 5.6386902521640945,
+      "grad_norm": 0.22384865581989288,
+      "learning_rate": 1.2310500198586089e-05,
+      "loss": 0.1272,
+      "step": 7491
+    },
+    {
+      "epoch": 5.63944298080542,
+      "grad_norm": 0.21459689736366272,
+      "learning_rate": 1.230501181727695e-05,
+      "loss": 0.2197,
+      "step": 7492
+    },
+    {
+      "epoch": 5.640195709446744,
+      "grad_norm": 0.18394756317138672,
+      "learning_rate": 1.2299525097668884e-05,
+      "loss": 0.2497,
+      "step": 7493
+    },
+    {
+      "epoch": 5.640948438088069,
+      "grad_norm": 0.1833568811416626,
+      "learning_rate": 1.229404004036189e-05,
+      "loss": 0.1438,
+      "step": 7494
+    },
+    {
+      "epoch": 5.641701166729394,
+      "grad_norm": 0.17294694483280182,
+      "learning_rate": 1.2288556645955814e-05,
+      "loss": 0.1872,
+      "step": 7495
+    },
+    {
+      "epoch": 5.642453895370719,
+      "grad_norm": 0.2186087816953659,
+      "learning_rate": 1.2283074915050315e-05,
+      "loss": 0.1594,
+      "step": 7496
+    },
+    {
+      "epoch": 5.643206624012044,
+      "grad_norm": 0.16765424609184265,
+      "learning_rate": 1.2277594848244871e-05,
+      "loss": 0.2139,
+      "step": 7497
+    },
+    {
+      "epoch": 5.6439593526533685,
+      "grad_norm": 0.18590080738067627,
+      "learning_rate": 1.2272116446138763e-05,
+      "loss": 0.1204,
+      "step": 7498
+    },
+    {
+      "epoch": 5.644712081294693,
+      "grad_norm": 0.1914040446281433,
+      "learning_rate": 1.226663970933111e-05,
+      "loss": 0.1329,
+      "step": 7499
+    },
+    {
+      "epoch": 5.645464809936018,
+      "grad_norm": 0.16038067638874054,
+      "learning_rate": 1.2261164638420832e-05,
+      "loss": 0.1966,
+      "step": 7500
+    },
+    {
+      "epoch": 5.646217538577343,
+      "grad_norm": 0.19198077917099,
+      "learning_rate": 1.2255691234006674e-05,
+      "loss": 0.1529,
+      "step": 7501
+    },
+    {
+      "epoch": 5.646970267218668,
+      "grad_norm": 0.18284891545772552,
+      "learning_rate": 1.2250219496687203e-05,
+      "loss": 0.1647,
+      "step": 7502
+    },
+    {
+      "epoch": 5.647722995859993,
+      "grad_norm": 0.16093626618385315,
+      "learning_rate": 1.224474942706081e-05,
+      "loss": 0.1306,
+      "step": 7503
+    },
+    {
+      "epoch": 5.648475724501317,
+      "grad_norm": 0.2038043588399887,
+      "learning_rate": 1.2239281025725676e-05,
+      "loss": 0.2076,
+      "step": 7504
+    },
+    {
+      "epoch": 5.649228453142642,
+      "grad_norm": 0.1974611133337021,
+      "learning_rate": 1.2233814293279825e-05,
+      "loss": 0.1627,
+      "step": 7505
+    },
+    {
+      "epoch": 5.649981181783967,
+      "grad_norm": 0.1764024943113327,
+      "learning_rate": 1.2228349230321096e-05,
+      "loss": 0.1717,
+      "step": 7506
+    },
+    {
+      "epoch": 5.650733910425291,
+      "grad_norm": 0.20837870240211487,
+      "learning_rate": 1.2222885837447128e-05,
+      "loss": 0.1423,
+      "step": 7507
+    },
+    {
+      "epoch": 5.651486639066617,
+      "grad_norm": 0.2020895928144455,
+      "learning_rate": 1.2217424115255397e-05,
+      "loss": 0.2217,
+      "step": 7508
+    },
+    {
+      "epoch": 5.652239367707941,
+      "grad_norm": 0.1731901466846466,
+      "learning_rate": 1.2211964064343185e-05,
+      "loss": 0.2458,
+      "step": 7509
+    },
+    {
+      "epoch": 5.652992096349266,
+      "grad_norm": 0.20546075701713562,
+      "learning_rate": 1.220650568530761e-05,
+      "loss": 0.1904,
+      "step": 7510
+    },
+    {
+      "epoch": 5.653744824990591,
+      "grad_norm": 0.20876875519752502,
+      "learning_rate": 1.220104897874557e-05,
+      "loss": 0.1534,
+      "step": 7511
+    },
+    {
+      "epoch": 5.6544975536319155,
+      "grad_norm": 0.22137387096881866,
+      "learning_rate": 1.2195593945253817e-05,
+      "loss": 0.1914,
+      "step": 7512
+    },
+    {
+      "epoch": 5.65525028227324,
+      "grad_norm": 0.1769166886806488,
+      "learning_rate": 1.2190140585428895e-05,
+      "loss": 0.2133,
+      "step": 7513
+    },
+    {
+      "epoch": 5.656003010914565,
+      "grad_norm": 0.2602977752685547,
+      "learning_rate": 1.2184688899867182e-05,
+      "loss": 0.2786,
+      "step": 7514
+    },
+    {
+      "epoch": 5.65675573955589,
+      "grad_norm": 0.1684648096561432,
+      "learning_rate": 1.2179238889164862e-05,
+      "loss": 0.1313,
+      "step": 7515
+    },
+    {
+      "epoch": 5.657508468197215,
+      "grad_norm": 0.2150682806968689,
+      "learning_rate": 1.2173790553917953e-05,
+      "loss": 0.2242,
+      "step": 7516
+    },
+    {
+      "epoch": 5.65826119683854,
+      "grad_norm": 0.2258751094341278,
+      "learning_rate": 1.2168343894722257e-05,
+      "loss": 0.2375,
+      "step": 7517
+    },
+    {
+      "epoch": 5.659013925479864,
+      "grad_norm": 0.1853095144033432,
+      "learning_rate": 1.2162898912173428e-05,
+      "loss": 0.1728,
+      "step": 7518
+    },
+    {
+      "epoch": 5.6597666541211895,
+      "grad_norm": 0.19899700582027435,
+      "learning_rate": 1.215745560686691e-05,
+      "loss": 0.1418,
+      "step": 7519
+    },
+    {
+      "epoch": 5.660519382762514,
+      "grad_norm": 0.1652509868144989,
+      "learning_rate": 1.2152013979397978e-05,
+      "loss": 0.1152,
+      "step": 7520
+    },
+    {
+      "epoch": 5.661272111403839,
+      "grad_norm": 0.1694941520690918,
+      "learning_rate": 1.2146574030361722e-05,
+      "loss": 0.1201,
+      "step": 7521
+    },
+    {
+      "epoch": 5.662024840045164,
+      "grad_norm": 0.19950641691684723,
+      "learning_rate": 1.2141135760353054e-05,
+      "loss": 0.2417,
+      "step": 7522
+    },
+    {
+      "epoch": 5.662777568686488,
+      "grad_norm": 0.21450386941432953,
+      "learning_rate": 1.2135699169966682e-05,
+      "loss": 0.1286,
+      "step": 7523
+    },
+    {
+      "epoch": 5.663530297327814,
+      "grad_norm": 0.19170936942100525,
+      "learning_rate": 1.2130264259797156e-05,
+      "loss": 0.2361,
+      "step": 7524
+    },
+    {
+      "epoch": 5.664283025969138,
+      "grad_norm": 0.18996207416057587,
+      "learning_rate": 1.2124831030438813e-05,
+      "loss": 0.1797,
+      "step": 7525
+    },
+    {
+      "epoch": 5.665035754610463,
+      "grad_norm": 0.20719854533672333,
+      "learning_rate": 1.2119399482485835e-05,
+      "loss": 0.1565,
+      "step": 7526
+    },
+    {
+      "epoch": 5.665788483251788,
+      "grad_norm": 0.19092898070812225,
+      "learning_rate": 1.2113969616532202e-05,
+      "loss": 0.2043,
+      "step": 7527
+    },
+    {
+      "epoch": 5.666541211893112,
+      "grad_norm": 0.21577371656894684,
+      "learning_rate": 1.210854143317173e-05,
+      "loss": 0.2495,
+      "step": 7528
+    },
+    {
+      "epoch": 5.667293940534437,
+      "grad_norm": 0.20681579411029816,
+      "learning_rate": 1.2103114932998015e-05,
+      "loss": 0.1836,
+      "step": 7529
+    },
+    {
+      "epoch": 5.668046669175762,
+      "grad_norm": 0.16374611854553223,
+      "learning_rate": 1.2097690116604504e-05,
+      "loss": 0.1314,
+      "step": 7530
+    },
+    {
+      "epoch": 5.668799397817087,
+      "grad_norm": 0.1987484097480774,
+      "learning_rate": 1.2092266984584452e-05,
+      "loss": 0.18,
+      "step": 7531
+    },
+    {
+      "epoch": 5.669552126458412,
+      "grad_norm": 0.19378876686096191,
+      "learning_rate": 1.208684553753091e-05,
+      "loss": 0.2527,
+      "step": 7532
+    },
+    {
+      "epoch": 5.670304855099737,
+      "grad_norm": 0.16842283308506012,
+      "learning_rate": 1.2081425776036768e-05,
+      "loss": 0.1432,
+      "step": 7533
+    },
+    {
+      "epoch": 5.671057583741061,
+      "grad_norm": 0.19734147191047668,
+      "learning_rate": 1.2076007700694723e-05,
+      "loss": 0.196,
+      "step": 7534
+    },
+    {
+      "epoch": 5.671810312382386,
+      "grad_norm": 0.19442321360111237,
+      "learning_rate": 1.2070591312097288e-05,
+      "loss": 0.143,
+      "step": 7535
+    },
+    {
+      "epoch": 5.672563041023711,
+      "grad_norm": 0.21794532239437103,
+      "learning_rate": 1.2065176610836785e-05,
+      "loss": 0.198,
+      "step": 7536
+    },
+    {
+      "epoch": 5.673315769665035,
+      "grad_norm": 0.19322721660137177,
+      "learning_rate": 1.2059763597505369e-05,
+      "loss": 0.1566,
+      "step": 7537
+    },
+    {
+      "epoch": 5.674068498306361,
+      "grad_norm": 0.18064266443252563,
+      "learning_rate": 1.2054352272694986e-05,
+      "loss": 0.2199,
+      "step": 7538
+    },
+    {
+      "epoch": 5.674821226947685,
+      "grad_norm": 0.19132502377033234,
+      "learning_rate": 1.2048942636997422e-05,
+      "loss": 0.2387,
+      "step": 7539
+    },
+    {
+      "epoch": 5.675573955589011,
+      "grad_norm": 0.18325674533843994,
+      "learning_rate": 1.2043534691004248e-05,
+      "loss": 0.1901,
+      "step": 7540
+    },
+    {
+      "epoch": 5.676326684230335,
+      "grad_norm": 0.2154097557067871,
+      "learning_rate": 1.2038128435306895e-05,
+      "loss": 0.1958,
+      "step": 7541
+    },
+    {
+      "epoch": 5.6770794128716595,
+      "grad_norm": 0.18168281018733978,
+      "learning_rate": 1.2032723870496562e-05,
+      "loss": 0.1486,
+      "step": 7542
+    },
+    {
+      "epoch": 5.677832141512985,
+      "grad_norm": 0.19649367034435272,
+      "learning_rate": 1.20273209971643e-05,
+      "loss": 0.1723,
+      "step": 7543
+    },
+    {
+      "epoch": 5.678584870154309,
+      "grad_norm": 0.1845475286245346,
+      "learning_rate": 1.2021919815900942e-05,
+      "loss": 0.1907,
+      "step": 7544
+    },
+    {
+      "epoch": 5.679337598795634,
+      "grad_norm": 0.17867013812065125,
+      "learning_rate": 1.2016520327297172e-05,
+      "loss": 0.1341,
+      "step": 7545
+    },
+    {
+      "epoch": 5.680090327436959,
+      "grad_norm": 0.1972634494304657,
+      "learning_rate": 1.2011122531943445e-05,
+      "loss": 0.2231,
+      "step": 7546
+    },
+    {
+      "epoch": 5.680843056078284,
+      "grad_norm": 0.1870119720697403,
+      "learning_rate": 1.2005726430430084e-05,
+      "loss": 0.1489,
+      "step": 7547
+    },
+    {
+      "epoch": 5.681595784719608,
+      "grad_norm": 0.1839354783296585,
+      "learning_rate": 1.2000332023347178e-05,
+      "loss": 0.2058,
+      "step": 7548
+    },
+    {
+      "epoch": 5.6823485133609335,
+      "grad_norm": 0.21493926644325256,
+      "learning_rate": 1.1994939311284665e-05,
+      "loss": 0.2334,
+      "step": 7549
+    },
+    {
+      "epoch": 5.683101242002258,
+      "grad_norm": 0.15573187172412872,
+      "learning_rate": 1.198954829483227e-05,
+      "loss": 0.1367,
+      "step": 7550
+    },
+    {
+      "epoch": 5.683853970643583,
+      "grad_norm": 0.1900930553674698,
+      "learning_rate": 1.1984158974579556e-05,
+      "loss": 0.1822,
+      "step": 7551
+    },
+    {
+      "epoch": 5.684606699284908,
+      "grad_norm": 0.1940416544675827,
+      "learning_rate": 1.197877135111589e-05,
+      "loss": 0.2214,
+      "step": 7552
+    },
+    {
+      "epoch": 5.685359427926232,
+      "grad_norm": 0.17402926087379456,
+      "learning_rate": 1.1973385425030457e-05,
+      "loss": 0.1658,
+      "step": 7553
+    },
+    {
+      "epoch": 5.686112156567558,
+      "grad_norm": 0.1950986087322235,
+      "learning_rate": 1.1968001196912244e-05,
+      "loss": 0.2189,
+      "step": 7554
+    },
+    {
+      "epoch": 5.686864885208882,
+      "grad_norm": 0.1853010654449463,
+      "learning_rate": 1.1962618667350068e-05,
+      "loss": 0.1634,
+      "step": 7555
+    },
+    {
+      "epoch": 5.6876176138502075,
+      "grad_norm": 0.19950851798057556,
+      "learning_rate": 1.1957237836932563e-05,
+      "loss": 0.2327,
+      "step": 7556
+    },
+    {
+      "epoch": 5.688370342491532,
+      "grad_norm": 0.21153561770915985,
+      "learning_rate": 1.195185870624815e-05,
+      "loss": 0.2406,
+      "step": 7557
+    },
+    {
+      "epoch": 5.689123071132856,
+      "grad_norm": 0.18555641174316406,
+      "learning_rate": 1.1946481275885102e-05,
+      "loss": 0.1495,
+      "step": 7558
+    },
+    {
+      "epoch": 5.689875799774182,
+      "grad_norm": 0.18537940084934235,
+      "learning_rate": 1.1941105546431463e-05,
+      "loss": 0.1691,
+      "step": 7559
+    },
+    {
+      "epoch": 5.690628528415506,
+      "grad_norm": 0.18060486018657684,
+      "learning_rate": 1.1935731518475142e-05,
+      "loss": 0.2252,
+      "step": 7560
+    },
+    {
+      "epoch": 5.691381257056831,
+      "grad_norm": 0.18175913393497467,
+      "learning_rate": 1.1930359192603816e-05,
+      "loss": 0.0788,
+      "step": 7561
+    },
+    {
+      "epoch": 5.692133985698156,
+      "grad_norm": 0.17025530338287354,
+      "learning_rate": 1.1924988569405006e-05,
+      "loss": 0.1908,
+      "step": 7562
+    },
+    {
+      "epoch": 5.692886714339481,
+      "grad_norm": 0.20035682618618011,
+      "learning_rate": 1.1919619649466025e-05,
+      "loss": 0.1379,
+      "step": 7563
+    },
+    {
+      "epoch": 5.693639442980805,
+      "grad_norm": 0.197805255651474,
+      "learning_rate": 1.1914252433374017e-05,
+      "loss": 0.207,
+      "step": 7564
+    },
+    {
+      "epoch": 5.69439217162213,
+      "grad_norm": 0.20719057321548462,
+      "learning_rate": 1.1908886921715919e-05,
+      "loss": 0.2314,
+      "step": 7565
+    },
+    {
+      "epoch": 5.695144900263455,
+      "grad_norm": 0.17758090794086456,
+      "learning_rate": 1.1903523115078518e-05,
+      "loss": 0.1272,
+      "step": 7566
+    },
+    {
+      "epoch": 5.695897628904779,
+      "grad_norm": 0.201105535030365,
+      "learning_rate": 1.1898161014048374e-05,
+      "loss": 0.1876,
+      "step": 7567
+    },
+    {
+      "epoch": 5.696650357546105,
+      "grad_norm": 0.18590597808361053,
+      "learning_rate": 1.1892800619211893e-05,
+      "loss": 0.1704,
+      "step": 7568
+    },
+    {
+      "epoch": 5.697403086187429,
+      "grad_norm": 0.18639060854911804,
+      "learning_rate": 1.1887441931155263e-05,
+      "loss": 0.1527,
+      "step": 7569
+    },
+    {
+      "epoch": 5.6981558148287546,
+      "grad_norm": 0.1941802203655243,
+      "learning_rate": 1.1882084950464519e-05,
+      "loss": 0.2345,
+      "step": 7570
+    },
+    {
+      "epoch": 5.698908543470079,
+      "grad_norm": 0.17813636362552643,
+      "learning_rate": 1.1876729677725468e-05,
+      "loss": 0.1218,
+      "step": 7571
+    },
+    {
+      "epoch": 5.6996612721114035,
+      "grad_norm": 0.19557997584342957,
+      "learning_rate": 1.1871376113523784e-05,
+      "loss": 0.1134,
+      "step": 7572
+    },
+    {
+      "epoch": 5.700414000752729,
+      "grad_norm": 0.19762642681598663,
+      "learning_rate": 1.1866024258444904e-05,
+      "loss": 0.1699,
+      "step": 7573
+    },
+    {
+      "epoch": 5.701166729394053,
+      "grad_norm": 0.16351351141929626,
+      "learning_rate": 1.1860674113074111e-05,
+      "loss": 0.1533,
+      "step": 7574
+    },
+    {
+      "epoch": 5.701919458035379,
+      "grad_norm": 0.22710146009922028,
+      "learning_rate": 1.1855325677996482e-05,
+      "loss": 0.1874,
+      "step": 7575
+    },
+    {
+      "epoch": 5.702672186676703,
+      "grad_norm": 0.18142420053482056,
+      "learning_rate": 1.1849978953796913e-05,
+      "loss": 0.1994,
+      "step": 7576
+    },
+    {
+      "epoch": 5.703424915318028,
+      "grad_norm": 0.17361295223236084,
+      "learning_rate": 1.1844633941060122e-05,
+      "loss": 0.1832,
+      "step": 7577
+    },
+    {
+      "epoch": 5.704177643959353,
+      "grad_norm": 0.19753926992416382,
+      "learning_rate": 1.183929064037062e-05,
+      "loss": 0.164,
+      "step": 7578
+    },
+    {
+      "epoch": 5.7049303726006775,
+      "grad_norm": 0.17107060551643372,
+      "learning_rate": 1.1833949052312747e-05,
+      "loss": 0.1802,
+      "step": 7579
+    },
+    {
+      "epoch": 5.705683101242002,
+      "grad_norm": 0.19981087744235992,
+      "learning_rate": 1.1828609177470652e-05,
+      "loss": 0.1517,
+      "step": 7580
+    },
+    {
+      "epoch": 5.706435829883327,
+      "grad_norm": 0.20839528739452362,
+      "learning_rate": 1.1823271016428302e-05,
+      "loss": 0.1926,
+      "step": 7581
+    },
+    {
+      "epoch": 5.707188558524652,
+      "grad_norm": 0.19802048802375793,
+      "learning_rate": 1.1817934569769457e-05,
+      "loss": 0.215,
+      "step": 7582
+    },
+    {
+      "epoch": 5.707941287165976,
+      "grad_norm": 0.1893249899148941,
+      "learning_rate": 1.1812599838077718e-05,
+      "loss": 0.1598,
+      "step": 7583
+    },
+    {
+      "epoch": 5.708694015807302,
+      "grad_norm": 0.20647013187408447,
+      "learning_rate": 1.180726682193646e-05,
+      "loss": 0.2218,
+      "step": 7584
+    },
+    {
+      "epoch": 5.709446744448626,
+      "grad_norm": 0.17659667134284973,
+      "learning_rate": 1.1801935521928926e-05,
+      "loss": 0.2127,
+      "step": 7585
+    },
+    {
+      "epoch": 5.7101994730899515,
+      "grad_norm": 0.19433267414569855,
+      "learning_rate": 1.179660593863811e-05,
+      "loss": 0.2092,
+      "step": 7586
+    },
+    {
+      "epoch": 5.710952201731276,
+      "grad_norm": 0.17829161882400513,
+      "learning_rate": 1.179127807264687e-05,
+      "loss": 0.1325,
+      "step": 7587
+    },
+    {
+      "epoch": 5.7117049303726,
+      "grad_norm": 0.194311261177063,
+      "learning_rate": 1.1785951924537832e-05,
+      "loss": 0.1716,
+      "step": 7588
+    },
+    {
+      "epoch": 5.712457659013926,
+      "grad_norm": 0.15538868308067322,
+      "learning_rate": 1.1780627494893474e-05,
+      "loss": 0.1635,
+      "step": 7589
+    },
+    {
+      "epoch": 5.71321038765525,
+      "grad_norm": 0.21472221612930298,
+      "learning_rate": 1.1775304784296044e-05,
+      "loss": 0.1961,
+      "step": 7590
+    },
+    {
+      "epoch": 5.713963116296575,
+      "grad_norm": 0.19291001558303833,
+      "learning_rate": 1.176998379332766e-05,
+      "loss": 0.1947,
+      "step": 7591
+    },
+    {
+      "epoch": 5.7147158449379,
+      "grad_norm": 0.17220252752304077,
+      "learning_rate": 1.1764664522570188e-05,
+      "loss": 0.1723,
+      "step": 7592
+    },
+    {
+      "epoch": 5.7154685735792246,
+      "grad_norm": 0.18070586025714874,
+      "learning_rate": 1.1759346972605353e-05,
+      "loss": 0.1728,
+      "step": 7593
+    },
+    {
+      "epoch": 5.71622130222055,
+      "grad_norm": 0.20313668251037598,
+      "learning_rate": 1.175403114401466e-05,
+      "loss": 0.1543,
+      "step": 7594
+    },
+    {
+      "epoch": 5.716974030861874,
+      "grad_norm": 0.20430909097194672,
+      "learning_rate": 1.1748717037379455e-05,
+      "loss": 0.1496,
+      "step": 7595
+    },
+    {
+      "epoch": 5.717726759503199,
+      "grad_norm": 0.17616018652915955,
+      "learning_rate": 1.1743404653280867e-05,
+      "loss": 0.2034,
+      "step": 7596
+    },
+    {
+      "epoch": 5.718479488144524,
+      "grad_norm": 0.1962384730577469,
+      "learning_rate": 1.1738093992299854e-05,
+      "loss": 0.1625,
+      "step": 7597
+    },
+    {
+      "epoch": 5.719232216785849,
+      "grad_norm": 0.19532446563243866,
+      "learning_rate": 1.1732785055017185e-05,
+      "loss": 0.2141,
+      "step": 7598
+    },
+    {
+      "epoch": 5.719984945427173,
+      "grad_norm": 0.1922546625137329,
+      "learning_rate": 1.1727477842013446e-05,
+      "loss": 0.2279,
+      "step": 7599
+    },
+    {
+      "epoch": 5.7207376740684985,
+      "grad_norm": 0.1820259988307953,
+      "learning_rate": 1.1722172353869008e-05,
+      "loss": 0.1234,
+      "step": 7600
+    },
+    {
+      "epoch": 5.7207376740684985,
+      "eval_loss": 0.21171638369560242,
+      "eval_runtime": 452.8651,
+      "eval_samples_per_second": 21.258,
+      "eval_steps_per_second": 0.665,
+      "step": 7600
+    },
+    {
+      "epoch": 5.721490402709823,
+      "grad_norm": 0.18755339086055756,
+      "learning_rate": 1.1716868591164081e-05,
+      "loss": 0.1594,
+      "step": 7601
+    },
+    {
+      "epoch": 5.7222431313511475,
+      "grad_norm": 0.17747458815574646,
+      "learning_rate": 1.171156655447868e-05,
+      "loss": 0.1329,
+      "step": 7602
+    },
+    {
+      "epoch": 5.722995859992473,
+      "grad_norm": 0.18851405382156372,
+      "learning_rate": 1.170626624439262e-05,
+      "loss": 0.1367,
+      "step": 7603
+    },
+    {
+      "epoch": 5.723748588633797,
+      "grad_norm": 0.20160679519176483,
+      "learning_rate": 1.1700967661485535e-05,
+      "loss": 0.217,
+      "step": 7604
+    },
+    {
+      "epoch": 5.724501317275123,
+      "grad_norm": 0.18556436896324158,
+      "learning_rate": 1.1695670806336872e-05,
+      "loss": 0.1862,
+      "step": 7605
+    },
+    {
+      "epoch": 5.725254045916447,
+      "grad_norm": 0.17542214691638947,
+      "learning_rate": 1.1690375679525898e-05,
+      "loss": 0.1479,
+      "step": 7606
+    },
+    {
+      "epoch": 5.726006774557772,
+      "grad_norm": 0.2125927060842514,
+      "learning_rate": 1.168508228163166e-05,
+      "loss": 0.2135,
+      "step": 7607
+    },
+    {
+      "epoch": 5.726759503199097,
+      "grad_norm": 0.20779214799404144,
+      "learning_rate": 1.1679790613233057e-05,
+      "loss": 0.1901,
+      "step": 7608
+    },
+    {
+      "epoch": 5.7275122318404215,
+      "grad_norm": 0.20908123254776,
+      "learning_rate": 1.1674500674908756e-05,
+      "loss": 0.1795,
+      "step": 7609
+    },
+    {
+      "epoch": 5.728264960481747,
+      "grad_norm": 0.17409025132656097,
+      "learning_rate": 1.166921246723727e-05,
+      "loss": 0.127,
+      "step": 7610
+    },
+    {
+      "epoch": 5.729017689123071,
+      "grad_norm": 0.2068019062280655,
+      "learning_rate": 1.1663925990796907e-05,
+      "loss": 0.1489,
+      "step": 7611
+    },
+    {
+      "epoch": 5.729770417764396,
+      "grad_norm": 0.17424145340919495,
+      "learning_rate": 1.1658641246165799e-05,
+      "loss": 0.1624,
+      "step": 7612
+    },
+    {
+      "epoch": 5.730523146405721,
+      "grad_norm": 0.22006668150424957,
+      "learning_rate": 1.1653358233921856e-05,
+      "loss": 0.2422,
+      "step": 7613
+    },
+    {
+      "epoch": 5.731275875047046,
+      "grad_norm": 0.18173035979270935,
+      "learning_rate": 1.1648076954642839e-05,
+      "loss": 0.1517,
+      "step": 7614
+    },
+    {
+      "epoch": 5.73202860368837,
+      "grad_norm": 0.21640938520431519,
+      "learning_rate": 1.1642797408906288e-05,
+      "loss": 0.1764,
+      "step": 7615
+    },
+    {
+      "epoch": 5.7327813323296954,
+      "grad_norm": 0.20455317199230194,
+      "learning_rate": 1.163751959728957e-05,
+      "loss": 0.1542,
+      "step": 7616
+    },
+    {
+      "epoch": 5.73353406097102,
+      "grad_norm": 0.2372986227273941,
+      "learning_rate": 1.163224352036986e-05,
+      "loss": 0.1928,
+      "step": 7617
+    },
+    {
+      "epoch": 5.734286789612344,
+      "grad_norm": 0.16143466532230377,
+      "learning_rate": 1.162696917872415e-05,
+      "loss": 0.0854,
+      "step": 7618
+    },
+    {
+      "epoch": 5.73503951825367,
+      "grad_norm": 0.2184748351573944,
+      "learning_rate": 1.1621696572929216e-05,
+      "loss": 0.2343,
+      "step": 7619
+    },
+    {
+      "epoch": 5.735792246894994,
+      "grad_norm": 0.18151994049549103,
+      "learning_rate": 1.1616425703561683e-05,
+      "loss": 0.1915,
+      "step": 7620
+    },
+    {
+      "epoch": 5.736544975536319,
+      "grad_norm": 0.20874713361263275,
+      "learning_rate": 1.1611156571197945e-05,
+      "loss": 0.1675,
+      "step": 7621
+    },
+    {
+      "epoch": 5.737297704177644,
+      "grad_norm": 0.18178118765354156,
+      "learning_rate": 1.1605889176414235e-05,
+      "loss": 0.1734,
+      "step": 7622
+    },
+    {
+      "epoch": 5.7380504328189685,
+      "grad_norm": 0.16175031661987305,
+      "learning_rate": 1.1600623519786591e-05,
+      "loss": 0.1756,
+      "step": 7623
+    },
+    {
+      "epoch": 5.738803161460294,
+      "grad_norm": 0.18329979479312897,
+      "learning_rate": 1.159535960189086e-05,
+      "loss": 0.2208,
+      "step": 7624
+    },
+    {
+      "epoch": 5.739555890101618,
+      "grad_norm": 0.1844458132982254,
+      "learning_rate": 1.1590097423302684e-05,
+      "loss": 0.1274,
+      "step": 7625
+    },
+    {
+      "epoch": 5.740308618742943,
+      "grad_norm": 0.2001466602087021,
+      "learning_rate": 1.1584836984597533e-05,
+      "loss": 0.182,
+      "step": 7626
+    },
+    {
+      "epoch": 5.741061347384268,
+      "grad_norm": 0.20597636699676514,
+      "learning_rate": 1.1579578286350689e-05,
+      "loss": 0.2005,
+      "step": 7627
+    },
+    {
+      "epoch": 5.741814076025593,
+      "grad_norm": 0.18696831166744232,
+      "learning_rate": 1.157432132913722e-05,
+      "loss": 0.1688,
+      "step": 7628
+    },
+    {
+      "epoch": 5.742566804666918,
+      "grad_norm": 0.18790322542190552,
+      "learning_rate": 1.1569066113532028e-05,
+      "loss": 0.1107,
+      "step": 7629
+    },
+    {
+      "epoch": 5.7433195333082425,
+      "grad_norm": 0.17746171355247498,
+      "learning_rate": 1.156381264010981e-05,
+      "loss": 0.2215,
+      "step": 7630
+    },
+    {
+      "epoch": 5.744072261949567,
+      "grad_norm": 0.1606818288564682,
+      "learning_rate": 1.1558560909445095e-05,
+      "loss": 0.1677,
+      "step": 7631
+    },
+    {
+      "epoch": 5.744824990590892,
+      "grad_norm": 0.1723446100950241,
+      "learning_rate": 1.155331092211218e-05,
+      "loss": 0.1078,
+      "step": 7632
+    },
+    {
+      "epoch": 5.745577719232217,
+      "grad_norm": 0.17558971047401428,
+      "learning_rate": 1.1548062678685215e-05,
+      "loss": 0.1113,
+      "step": 7633
+    },
+    {
+      "epoch": 5.746330447873541,
+      "grad_norm": 0.19719049334526062,
+      "learning_rate": 1.1542816179738126e-05,
+      "loss": 0.1783,
+      "step": 7634
+    },
+    {
+      "epoch": 5.747083176514867,
+      "grad_norm": 0.1630217432975769,
+      "learning_rate": 1.1537571425844664e-05,
+      "loss": 0.1632,
+      "step": 7635
+    },
+    {
+      "epoch": 5.747835905156191,
+      "grad_norm": 0.16601254045963287,
+      "learning_rate": 1.1532328417578398e-05,
+      "loss": 0.1246,
+      "step": 7636
+    },
+    {
+      "epoch": 5.748588633797516,
+      "grad_norm": 0.19097372889518738,
+      "learning_rate": 1.1527087155512694e-05,
+      "loss": 0.2308,
+      "step": 7637
+    },
+    {
+      "epoch": 5.749341362438841,
+      "grad_norm": 0.19964085519313812,
+      "learning_rate": 1.1521847640220716e-05,
+      "loss": 0.22,
+      "step": 7638
+    },
+    {
+      "epoch": 5.7500940910801654,
+      "grad_norm": 0.18599830567836761,
+      "learning_rate": 1.1516609872275463e-05,
+      "loss": 0.1374,
+      "step": 7639
+    },
+    {
+      "epoch": 5.75084681972149,
+      "grad_norm": 0.15749257802963257,
+      "learning_rate": 1.1511373852249719e-05,
+      "loss": 0.1894,
+      "step": 7640
+    },
+    {
+      "epoch": 5.751599548362815,
+      "grad_norm": 0.1637045443058014,
+      "learning_rate": 1.1506139580716092e-05,
+      "loss": 0.1458,
+      "step": 7641
+    },
+    {
+      "epoch": 5.75235227700414,
+      "grad_norm": 0.1881384700536728,
+      "learning_rate": 1.1500907058246993e-05,
+      "loss": 0.2375,
+      "step": 7642
+    },
+    {
+      "epoch": 5.753105005645465,
+      "grad_norm": 0.16812175512313843,
+      "learning_rate": 1.1495676285414652e-05,
+      "loss": 0.1718,
+      "step": 7643
+    },
+    {
+      "epoch": 5.75385773428679,
+      "grad_norm": 0.20235727727413177,
+      "learning_rate": 1.149044726279108e-05,
+      "loss": 0.1812,
+      "step": 7644
+    },
+    {
+      "epoch": 5.754610462928114,
+      "grad_norm": 0.2047838419675827,
+      "learning_rate": 1.1485219990948136e-05,
+      "loss": 0.2521,
+      "step": 7645
+    },
+    {
+      "epoch": 5.755363191569439,
+      "grad_norm": 0.17614597082138062,
+      "learning_rate": 1.1479994470457448e-05,
+      "loss": 0.1402,
+      "step": 7646
+    },
+    {
+      "epoch": 5.756115920210764,
+      "grad_norm": 0.21370024979114532,
+      "learning_rate": 1.1474770701890479e-05,
+      "loss": 0.1818,
+      "step": 7647
+    },
+    {
+      "epoch": 5.756868648852089,
+      "grad_norm": 0.17192402482032776,
+      "learning_rate": 1.1469548685818492e-05,
+      "loss": 0.1793,
+      "step": 7648
+    },
+    {
+      "epoch": 5.757621377493414,
+      "grad_norm": 0.17930012941360474,
+      "learning_rate": 1.1464328422812565e-05,
+      "loss": 0.1518,
+      "step": 7649
+    },
+    {
+      "epoch": 5.758374106134738,
+      "grad_norm": 0.1932365745306015,
+      "learning_rate": 1.1459109913443567e-05,
+      "loss": 0.1677,
+      "step": 7650
+    },
+    {
+      "epoch": 5.759126834776064,
+      "grad_norm": 0.19831304252147675,
+      "learning_rate": 1.1453893158282194e-05,
+      "loss": 0.1625,
+      "step": 7651
+    },
+    {
+      "epoch": 5.759879563417388,
+      "grad_norm": 0.16569335758686066,
+      "learning_rate": 1.1448678157898945e-05,
+      "loss": 0.1737,
+      "step": 7652
+    },
+    {
+      "epoch": 5.7606322920587125,
+      "grad_norm": 0.19212369620800018,
+      "learning_rate": 1.144346491286411e-05,
+      "loss": 0.1812,
+      "step": 7653
+    },
+    {
+      "epoch": 5.761385020700038,
+      "grad_norm": 0.2231551557779312,
+      "learning_rate": 1.1438253423747817e-05,
+      "loss": 0.2138,
+      "step": 7654
+    },
+    {
+      "epoch": 5.762137749341362,
+      "grad_norm": 0.18324019014835358,
+      "learning_rate": 1.1433043691119975e-05,
+      "loss": 0.1448,
+      "step": 7655
+    },
+    {
+      "epoch": 5.762890477982687,
+      "grad_norm": 0.18198010325431824,
+      "learning_rate": 1.142783571555033e-05,
+      "loss": 0.1385,
+      "step": 7656
+    },
+    {
+      "epoch": 5.763643206624012,
+      "grad_norm": 0.1893947869539261,
+      "learning_rate": 1.14226294976084e-05,
+      "loss": 0.1167,
+      "step": 7657
+    },
+    {
+      "epoch": 5.764395935265337,
+      "grad_norm": 0.18526524305343628,
+      "learning_rate": 1.1417425037863542e-05,
+      "loss": 0.1868,
+      "step": 7658
+    },
+    {
+      "epoch": 5.765148663906662,
+      "grad_norm": 0.2029908448457718,
+      "learning_rate": 1.1412222336884894e-05,
+      "loss": 0.1989,
+      "step": 7659
+    },
+    {
+      "epoch": 5.7659013925479865,
+      "grad_norm": 0.22949492931365967,
+      "learning_rate": 1.1407021395241432e-05,
+      "loss": 0.2219,
+      "step": 7660
+    },
+    {
+      "epoch": 5.766654121189311,
+      "grad_norm": 0.2067500650882721,
+      "learning_rate": 1.1401822213501902e-05,
+      "loss": 0.1617,
+      "step": 7661
+    },
+    {
+      "epoch": 5.767406849830636,
+      "grad_norm": 0.18170759081840515,
+      "learning_rate": 1.1396624792234902e-05,
+      "loss": 0.2257,
+      "step": 7662
+    },
+    {
+      "epoch": 5.768159578471961,
+      "grad_norm": 0.15493462979793549,
+      "learning_rate": 1.1391429132008799e-05,
+      "loss": 0.1074,
+      "step": 7663
+    },
+    {
+      "epoch": 5.768912307113285,
+      "grad_norm": 0.18269655108451843,
+      "learning_rate": 1.1386235233391794e-05,
+      "loss": 0.1723,
+      "step": 7664
+    },
+    {
+      "epoch": 5.769665035754611,
+      "grad_norm": 0.19604036211967468,
+      "learning_rate": 1.138104309695187e-05,
+      "loss": 0.2114,
+      "step": 7665
+    },
+    {
+      "epoch": 5.770417764395935,
+      "grad_norm": 0.1826332062482834,
+      "learning_rate": 1.1375852723256847e-05,
+      "loss": 0.232,
+      "step": 7666
+    },
+    {
+      "epoch": 5.7711704930372605,
+      "grad_norm": 0.2215484231710434,
+      "learning_rate": 1.1370664112874312e-05,
+      "loss": 0.2192,
+      "step": 7667
+    },
+    {
+      "epoch": 5.771923221678585,
+      "grad_norm": 0.19128505885601044,
+      "learning_rate": 1.1365477266371715e-05,
+      "loss": 0.2504,
+      "step": 7668
+    },
+    {
+      "epoch": 5.772675950319909,
+      "grad_norm": 0.20941972732543945,
+      "learning_rate": 1.1360292184316257e-05,
+      "loss": 0.1759,
+      "step": 7669
+    },
+    {
+      "epoch": 5.773428678961235,
+      "grad_norm": 0.17569777369499207,
+      "learning_rate": 1.1355108867274988e-05,
+      "loss": 0.2206,
+      "step": 7670
+    },
+    {
+      "epoch": 5.774181407602559,
+      "grad_norm": 0.18950392305850983,
+      "learning_rate": 1.1349927315814735e-05,
+      "loss": 0.1612,
+      "step": 7671
+    },
+    {
+      "epoch": 5.774934136243884,
+      "grad_norm": 0.18528421223163605,
+      "learning_rate": 1.1344747530502146e-05,
+      "loss": 0.1556,
+      "step": 7672
+    },
+    {
+      "epoch": 5.775686864885209,
+      "grad_norm": 0.19401177763938904,
+      "learning_rate": 1.133956951190368e-05,
+      "loss": 0.1187,
+      "step": 7673
+    },
+    {
+      "epoch": 5.776439593526534,
+      "grad_norm": 0.1518394649028778,
+      "learning_rate": 1.13343932605856e-05,
+      "loss": 0.0978,
+      "step": 7674
+    },
+    {
+      "epoch": 5.777192322167858,
+      "grad_norm": 0.18485309183597565,
+      "learning_rate": 1.1329218777113965e-05,
+      "loss": 0.1301,
+      "step": 7675
+    },
+    {
+      "epoch": 5.777945050809183,
+      "grad_norm": 0.18460242450237274,
+      "learning_rate": 1.132404606205465e-05,
+      "loss": 0.1201,
+      "step": 7676
+    },
+    {
+      "epoch": 5.778697779450508,
+      "grad_norm": 0.2112095057964325,
+      "learning_rate": 1.1318875115973345e-05,
+      "loss": 0.2139,
+      "step": 7677
+    },
+    {
+      "epoch": 5.779450508091833,
+      "grad_norm": 0.17469823360443115,
+      "learning_rate": 1.1313705939435523e-05,
+      "loss": 0.1439,
+      "step": 7678
+    },
+    {
+      "epoch": 5.780203236733158,
+      "grad_norm": 0.18990403413772583,
+      "learning_rate": 1.1308538533006488e-05,
+      "loss": 0.2385,
+      "step": 7679
+    },
+    {
+      "epoch": 5.780955965374482,
+      "grad_norm": 0.19677701592445374,
+      "learning_rate": 1.1303372897251324e-05,
+      "loss": 0.1656,
+      "step": 7680
+    },
+    {
+      "epoch": 5.781708694015808,
+      "grad_norm": 0.15629392862319946,
+      "learning_rate": 1.1298209032734961e-05,
+      "loss": 0.1143,
+      "step": 7681
+    },
+    {
+      "epoch": 5.782461422657132,
+      "grad_norm": 0.18351444602012634,
+      "learning_rate": 1.1293046940022095e-05,
+      "loss": 0.1795,
+      "step": 7682
+    },
+    {
+      "epoch": 5.783214151298457,
+      "grad_norm": 0.20505140721797943,
+      "learning_rate": 1.1287886619677255e-05,
+      "loss": 0.2435,
+      "step": 7683
+    },
+    {
+      "epoch": 5.783966879939782,
+      "grad_norm": 0.1812034547328949,
+      "learning_rate": 1.1282728072264751e-05,
+      "loss": 0.1784,
+      "step": 7684
+    },
+    {
+      "epoch": 5.784719608581106,
+      "grad_norm": 0.20074252784252167,
+      "learning_rate": 1.127757129834873e-05,
+      "loss": 0.115,
+      "step": 7685
+    },
+    {
+      "epoch": 5.785472337222432,
+      "grad_norm": 0.18022100627422333,
+      "learning_rate": 1.1272416298493109e-05,
+      "loss": 0.1991,
+      "step": 7686
+    },
+    {
+      "epoch": 5.786225065863756,
+      "grad_norm": 0.18985703587532043,
+      "learning_rate": 1.1267263073261659e-05,
+      "loss": 0.1515,
+      "step": 7687
+    },
+    {
+      "epoch": 5.786977794505081,
+      "grad_norm": 0.17596174776554108,
+      "learning_rate": 1.1262111623217903e-05,
+      "loss": 0.1741,
+      "step": 7688
+    },
+    {
+      "epoch": 5.787730523146406,
+      "grad_norm": 0.18399928510189056,
+      "learning_rate": 1.1256961948925216e-05,
+      "loss": 0.1473,
+      "step": 7689
+    },
+    {
+      "epoch": 5.7884832517877305,
+      "grad_norm": 0.18601563572883606,
+      "learning_rate": 1.1251814050946743e-05,
+      "loss": 0.1986,
+      "step": 7690
+    },
+    {
+      "epoch": 5.789235980429055,
+      "grad_norm": 0.1561710685491562,
+      "learning_rate": 1.124666792984546e-05,
+      "loss": 0.177,
+      "step": 7691
+    },
+    {
+      "epoch": 5.78998870907038,
+      "grad_norm": 0.16879057884216309,
+      "learning_rate": 1.1241523586184125e-05,
+      "loss": 0.1922,
+      "step": 7692
+    },
+    {
+      "epoch": 5.790741437711705,
+      "grad_norm": 0.2068219929933548,
+      "learning_rate": 1.1236381020525339e-05,
+      "loss": 0.1642,
+      "step": 7693
+    },
+    {
+      "epoch": 5.791494166353029,
+      "grad_norm": 0.16467836499214172,
+      "learning_rate": 1.1231240233431467e-05,
+      "loss": 0.1164,
+      "step": 7694
+    },
+    {
+      "epoch": 5.792246894994355,
+      "grad_norm": 0.1673501580953598,
+      "learning_rate": 1.1226101225464707e-05,
+      "loss": 0.062,
+      "step": 7695
+    },
+    {
+      "epoch": 5.792999623635679,
+      "grad_norm": 0.19563189148902893,
+      "learning_rate": 1.1220963997187046e-05,
+      "loss": 0.1898,
+      "step": 7696
+    },
+    {
+      "epoch": 5.7937523522770045,
+      "grad_norm": 0.16967357695102692,
+      "learning_rate": 1.1215828549160284e-05,
+      "loss": 0.2405,
+      "step": 7697
+    },
+    {
+      "epoch": 5.794505080918329,
+      "grad_norm": 0.19012218713760376,
+      "learning_rate": 1.121069488194604e-05,
+      "loss": 0.1399,
+      "step": 7698
+    },
+    {
+      "epoch": 5.795257809559653,
+      "grad_norm": 0.1697230041027069,
+      "learning_rate": 1.1205562996105704e-05,
+      "loss": 0.1269,
+      "step": 7699
+    },
+    {
+      "epoch": 5.796010538200979,
+      "grad_norm": 0.20059847831726074,
+      "learning_rate": 1.12004328922005e-05,
+      "loss": 0.1782,
+      "step": 7700
+    },
+    {
+      "epoch": 5.796763266842303,
+      "grad_norm": 0.16896232962608337,
+      "learning_rate": 1.1195304570791451e-05,
+      "loss": 0.1666,
+      "step": 7701
+    },
+    {
+      "epoch": 5.797515995483629,
+      "grad_norm": 0.20742973685264587,
+      "learning_rate": 1.1190178032439383e-05,
+      "loss": 0.2289,
+      "step": 7702
+    },
+    {
+      "epoch": 5.798268724124953,
+      "grad_norm": 0.18096236884593964,
+      "learning_rate": 1.118505327770492e-05,
+      "loss": 0.1929,
+      "step": 7703
+    },
+    {
+      "epoch": 5.799021452766278,
+      "grad_norm": 0.16409897804260254,
+      "learning_rate": 1.1179930307148509e-05,
+      "loss": 0.1447,
+      "step": 7704
+    },
+    {
+      "epoch": 5.799774181407603,
+      "grad_norm": 0.1960889846086502,
+      "learning_rate": 1.1174809121330366e-05,
+      "loss": 0.194,
+      "step": 7705
+    },
+    {
+      "epoch": 5.800526910048927,
+      "grad_norm": 0.16229066252708435,
+      "learning_rate": 1.116968972081057e-05,
+      "loss": 0.0821,
+      "step": 7706
+    },
+    {
+      "epoch": 5.801279638690252,
+      "grad_norm": 0.21893024444580078,
+      "learning_rate": 1.1164572106148944e-05,
+      "loss": 0.1898,
+      "step": 7707
+    },
+    {
+      "epoch": 5.802032367331577,
+      "grad_norm": 0.17689503729343414,
+      "learning_rate": 1.1159456277905164e-05,
+      "loss": 0.198,
+      "step": 7708
+    },
+    {
+      "epoch": 5.802785095972902,
+      "grad_norm": 0.186757430434227,
+      "learning_rate": 1.1154342236638668e-05,
+      "loss": 0.2172,
+      "step": 7709
+    },
+    {
+      "epoch": 5.803537824614226,
+      "grad_norm": 0.17651569843292236,
+      "learning_rate": 1.1149229982908736e-05,
+      "loss": 0.1735,
+      "step": 7710
+    },
+    {
+      "epoch": 5.8042905532555515,
+      "grad_norm": 0.21231521666049957,
+      "learning_rate": 1.1144119517274414e-05,
+      "loss": 0.2249,
+      "step": 7711
+    },
+    {
+      "epoch": 5.805043281896876,
+      "grad_norm": 0.1768522411584854,
+      "learning_rate": 1.113901084029461e-05,
+      "loss": 0.1833,
+      "step": 7712
+    },
+    {
+      "epoch": 5.8057960105382005,
+      "grad_norm": 0.19356156885623932,
+      "learning_rate": 1.1133903952527972e-05,
+      "loss": 0.0877,
+      "step": 7713
+    },
+    {
+      "epoch": 5.806548739179526,
+      "grad_norm": 0.20096057653427124,
+      "learning_rate": 1.1128798854533001e-05,
+      "loss": 0.2004,
+      "step": 7714
+    },
+    {
+      "epoch": 5.80730146782085,
+      "grad_norm": 0.1634809821844101,
+      "learning_rate": 1.1123695546867962e-05,
+      "loss": 0.1187,
+      "step": 7715
+    },
+    {
+      "epoch": 5.808054196462176,
+      "grad_norm": 0.21843981742858887,
+      "learning_rate": 1.1118594030090968e-05,
+      "loss": 0.2094,
+      "step": 7716
+    },
+    {
+      "epoch": 5.8088069251035,
+      "grad_norm": 0.16877593100070953,
+      "learning_rate": 1.1113494304759895e-05,
+      "loss": 0.1776,
+      "step": 7717
+    },
+    {
+      "epoch": 5.809559653744825,
+      "grad_norm": 0.1990184336900711,
+      "learning_rate": 1.1108396371432444e-05,
+      "loss": 0.1738,
+      "step": 7718
+    },
+    {
+      "epoch": 5.81031238238615,
+      "grad_norm": 0.2006627768278122,
+      "learning_rate": 1.1103300230666126e-05,
+      "loss": 0.1789,
+      "step": 7719
+    },
+    {
+      "epoch": 5.8110651110274745,
+      "grad_norm": 0.18542663753032684,
+      "learning_rate": 1.1098205883018244e-05,
+      "loss": 0.1728,
+      "step": 7720
+    },
+    {
+      "epoch": 5.8118178396688,
+      "grad_norm": 0.19814032316207886,
+      "learning_rate": 1.1093113329045904e-05,
+      "loss": 0.2648,
+      "step": 7721
+    },
+    {
+      "epoch": 5.812570568310124,
+      "grad_norm": 0.19607706367969513,
+      "learning_rate": 1.1088022569306022e-05,
+      "loss": 0.1848,
+      "step": 7722
+    },
+    {
+      "epoch": 5.813323296951449,
+      "grad_norm": 0.17646963894367218,
+      "learning_rate": 1.1082933604355327e-05,
+      "loss": 0.1581,
+      "step": 7723
+    },
+    {
+      "epoch": 5.814076025592774,
+      "grad_norm": 0.18790745735168457,
+      "learning_rate": 1.107784643475032e-05,
+      "loss": 0.209,
+      "step": 7724
+    },
+    {
+      "epoch": 5.814828754234099,
+      "grad_norm": 0.1778683066368103,
+      "learning_rate": 1.1072761061047338e-05,
+      "loss": 0.2677,
+      "step": 7725
+    },
+    {
+      "epoch": 5.815581482875423,
+      "grad_norm": 0.19713176786899567,
+      "learning_rate": 1.1067677483802505e-05,
+      "loss": 0.1491,
+      "step": 7726
+    },
+    {
+      "epoch": 5.8163342115167485,
+      "grad_norm": 0.1880626231431961,
+      "learning_rate": 1.1062595703571768e-05,
+      "loss": 0.1997,
+      "step": 7727
+    },
+    {
+      "epoch": 5.817086940158073,
+      "grad_norm": 0.20212562382221222,
+      "learning_rate": 1.1057515720910842e-05,
+      "loss": 0.2329,
+      "step": 7728
+    },
+    {
+      "epoch": 5.817839668799397,
+      "grad_norm": 0.20239591598510742,
+      "learning_rate": 1.1052437536375285e-05,
+      "loss": 0.2105,
+      "step": 7729
+    },
+    {
+      "epoch": 5.818592397440723,
+      "grad_norm": 0.1781044900417328,
+      "learning_rate": 1.1047361150520418e-05,
+      "loss": 0.1504,
+      "step": 7730
+    },
+    {
+      "epoch": 5.819345126082047,
+      "grad_norm": 0.19302469491958618,
+      "learning_rate": 1.1042286563901405e-05,
+      "loss": 0.2372,
+      "step": 7731
+    },
+    {
+      "epoch": 5.820097854723373,
+      "grad_norm": 0.18924282491207123,
+      "learning_rate": 1.1037213777073188e-05,
+      "loss": 0.2136,
+      "step": 7732
+    },
+    {
+      "epoch": 5.820850583364697,
+      "grad_norm": 0.1971602588891983,
+      "learning_rate": 1.1032142790590528e-05,
+      "loss": 0.1637,
+      "step": 7733
+    },
+    {
+      "epoch": 5.8216033120060215,
+      "grad_norm": 0.19025187194347382,
+      "learning_rate": 1.1027073605007968e-05,
+      "loss": 0.1719,
+      "step": 7734
+    },
+    {
+      "epoch": 5.822356040647347,
+      "grad_norm": 0.18661677837371826,
+      "learning_rate": 1.1022006220879874e-05,
+      "loss": 0.1372,
+      "step": 7735
+    },
+    {
+      "epoch": 5.823108769288671,
+      "grad_norm": 0.17214977741241455,
+      "learning_rate": 1.1016940638760402e-05,
+      "loss": 0.1601,
+      "step": 7736
+    },
+    {
+      "epoch": 5.823861497929996,
+      "grad_norm": 0.17890390753746033,
+      "learning_rate": 1.1011876859203521e-05,
+      "loss": 0.1468,
+      "step": 7737
+    },
+    {
+      "epoch": 5.824614226571321,
+      "grad_norm": 0.16620418429374695,
+      "learning_rate": 1.1006814882762995e-05,
+      "loss": 0.178,
+      "step": 7738
+    },
+    {
+      "epoch": 5.825366955212646,
+      "grad_norm": 0.1980728656053543,
+      "learning_rate": 1.1001754709992404e-05,
+      "loss": 0.1265,
+      "step": 7739
+    },
+    {
+      "epoch": 5.826119683853971,
+      "grad_norm": 0.22055767476558685,
+      "learning_rate": 1.0996696341445107e-05,
+      "loss": 0.2234,
+      "step": 7740
+    },
+    {
+      "epoch": 5.8268724124952955,
+      "grad_norm": 0.2014564424753189,
+      "learning_rate": 1.0991639777674293e-05,
+      "loss": 0.1803,
+      "step": 7741
+    },
+    {
+      "epoch": 5.82762514113662,
+      "grad_norm": 0.1726893037557602,
+      "learning_rate": 1.0986585019232923e-05,
+      "loss": 0.1501,
+      "step": 7742
+    },
+    {
+      "epoch": 5.828377869777945,
+      "grad_norm": 0.1675926297903061,
+      "learning_rate": 1.098153206667379e-05,
+      "loss": 0.1164,
+      "step": 7743
+    },
+    {
+      "epoch": 5.82913059841927,
+      "grad_norm": 0.18545199930667877,
+      "learning_rate": 1.0976480920549472e-05,
+      "loss": 0.1344,
+      "step": 7744
+    },
+    {
+      "epoch": 5.829883327060594,
+      "grad_norm": 0.18924787640571594,
+      "learning_rate": 1.0971431581412368e-05,
+      "loss": 0.1289,
+      "step": 7745
+    },
+    {
+      "epoch": 5.83063605570192,
+      "grad_norm": 0.17331954836845398,
+      "learning_rate": 1.0966384049814648e-05,
+      "loss": 0.1747,
+      "step": 7746
+    },
+    {
+      "epoch": 5.831388784343244,
+      "grad_norm": 0.20274607837200165,
+      "learning_rate": 1.096133832630831e-05,
+      "loss": 0.2639,
+      "step": 7747
+    },
+    {
+      "epoch": 5.832141512984569,
+      "grad_norm": 0.19224531948566437,
+      "learning_rate": 1.0956294411445155e-05,
+      "loss": 0.183,
+      "step": 7748
+    },
+    {
+      "epoch": 5.832894241625894,
+      "grad_norm": 0.2023853212594986,
+      "learning_rate": 1.0951252305776765e-05,
+      "loss": 0.1878,
+      "step": 7749
+    },
+    {
+      "epoch": 5.8336469702672185,
+      "grad_norm": 0.17003753781318665,
+      "learning_rate": 1.094621200985454e-05,
+      "loss": 0.1484,
+      "step": 7750
+    },
+    {
+      "epoch": 5.834399698908544,
+      "grad_norm": 0.17372611165046692,
+      "learning_rate": 1.0941173524229681e-05,
+      "loss": 0.1174,
+      "step": 7751
+    },
+    {
+      "epoch": 5.835152427549868,
+      "grad_norm": 0.1607314646244049,
+      "learning_rate": 1.0936136849453197e-05,
+      "loss": 0.1104,
+      "step": 7752
+    },
+    {
+      "epoch": 5.835905156191193,
+      "grad_norm": 0.1808215230703354,
+      "learning_rate": 1.0931101986075876e-05,
+      "loss": 0.1455,
+      "step": 7753
+    },
+    {
+      "epoch": 5.836657884832518,
+      "grad_norm": 0.22739823162555695,
+      "learning_rate": 1.0926068934648339e-05,
+      "loss": 0.1743,
+      "step": 7754
+    },
+    {
+      "epoch": 5.837410613473843,
+      "grad_norm": 0.20626068115234375,
+      "learning_rate": 1.0921037695720976e-05,
+      "loss": 0.1669,
+      "step": 7755
+    },
+    {
+      "epoch": 5.838163342115168,
+      "grad_norm": 0.2164037972688675,
+      "learning_rate": 1.0916008269844006e-05,
+      "loss": 0.2182,
+      "step": 7756
+    },
+    {
+      "epoch": 5.838916070756492,
+      "grad_norm": 0.2063000202178955,
+      "learning_rate": 1.0910980657567438e-05,
+      "loss": 0.1641,
+      "step": 7757
+    },
+    {
+      "epoch": 5.839668799397817,
+      "grad_norm": 0.18803177773952484,
+      "learning_rate": 1.090595485944109e-05,
+      "loss": 0.1807,
+      "step": 7758
+    },
+    {
+      "epoch": 5.840421528039142,
+      "grad_norm": 0.1756078451871872,
+      "learning_rate": 1.0900930876014564e-05,
+      "loss": 0.1832,
+      "step": 7759
+    },
+    {
+      "epoch": 5.841174256680467,
+      "grad_norm": 0.19426926970481873,
+      "learning_rate": 1.0895908707837287e-05,
+      "loss": 0.1822,
+      "step": 7760
+    },
+    {
+      "epoch": 5.841926985321791,
+      "grad_norm": 0.18737143278121948,
+      "learning_rate": 1.0890888355458464e-05,
+      "loss": 0.2323,
+      "step": 7761
+    },
+    {
+      "epoch": 5.842679713963117,
+      "grad_norm": 0.18296940624713898,
+      "learning_rate": 1.0885869819427119e-05,
+      "loss": 0.2454,
+      "step": 7762
+    },
+    {
+      "epoch": 5.843432442604441,
+      "grad_norm": 0.19341035187244415,
+      "learning_rate": 1.0880853100292071e-05,
+      "loss": 0.1556,
+      "step": 7763
+    },
+    {
+      "epoch": 5.8441851712457655,
+      "grad_norm": 0.1909821480512619,
+      "learning_rate": 1.0875838198601949e-05,
+      "loss": 0.1468,
+      "step": 7764
+    },
+    {
+      "epoch": 5.844937899887091,
+      "grad_norm": 0.17142337560653687,
+      "learning_rate": 1.0870825114905162e-05,
+      "loss": 0.1613,
+      "step": 7765
+    },
+    {
+      "epoch": 5.845690628528415,
+      "grad_norm": 0.18369950354099274,
+      "learning_rate": 1.0865813849749942e-05,
+      "loss": 0.1513,
+      "step": 7766
+    },
+    {
+      "epoch": 5.84644335716974,
+      "grad_norm": 0.17675215005874634,
+      "learning_rate": 1.0860804403684305e-05,
+      "loss": 0.1563,
+      "step": 7767
+    },
+    {
+      "epoch": 5.847196085811065,
+      "grad_norm": 0.1713147610425949,
+      "learning_rate": 1.0855796777256083e-05,
+      "loss": 0.1724,
+      "step": 7768
+    },
+    {
+      "epoch": 5.84794881445239,
+      "grad_norm": 0.2183070331811905,
+      "learning_rate": 1.0850790971012898e-05,
+      "loss": 0.2973,
+      "step": 7769
+    },
+    {
+      "epoch": 5.848701543093715,
+      "grad_norm": 0.18634453415870667,
+      "learning_rate": 1.0845786985502189e-05,
+      "loss": 0.1532,
+      "step": 7770
+    },
+    {
+      "epoch": 5.8494542717350395,
+      "grad_norm": 0.19868610799312592,
+      "learning_rate": 1.0840784821271171e-05,
+      "loss": 0.2019,
+      "step": 7771
+    },
+    {
+      "epoch": 5.850207000376364,
+      "grad_norm": 0.16666795313358307,
+      "learning_rate": 1.0835784478866877e-05,
+      "loss": 0.2357,
+      "step": 7772
+    },
+    {
+      "epoch": 5.850959729017689,
+      "grad_norm": 0.21193411946296692,
+      "learning_rate": 1.0830785958836145e-05,
+      "loss": 0.1725,
+      "step": 7773
+    },
+    {
+      "epoch": 5.851712457659014,
+      "grad_norm": 0.19169166684150696,
+      "learning_rate": 1.0825789261725593e-05,
+      "loss": 0.1872,
+      "step": 7774
+    },
+    {
+      "epoch": 5.852465186300339,
+      "grad_norm": 0.1699906587600708,
+      "learning_rate": 1.0820794388081657e-05,
+      "loss": 0.1965,
+      "step": 7775
+    },
+    {
+      "epoch": 5.853217914941664,
+      "grad_norm": 0.1644848883152008,
+      "learning_rate": 1.0815801338450571e-05,
+      "loss": 0.1671,
+      "step": 7776
+    },
+    {
+      "epoch": 5.853970643582988,
+      "grad_norm": 0.1779901087284088,
+      "learning_rate": 1.0810810113378373e-05,
+      "loss": 0.129,
+      "step": 7777
+    },
+    {
+      "epoch": 5.8547233722243135,
+      "grad_norm": 0.18988293409347534,
+      "learning_rate": 1.0805820713410878e-05,
+      "loss": 0.1636,
+      "step": 7778
+    },
+    {
+      "epoch": 5.855476100865638,
+      "grad_norm": 0.18649263679981232,
+      "learning_rate": 1.0800833139093742e-05,
+      "loss": 0.1129,
+      "step": 7779
+    },
+    {
+      "epoch": 5.856228829506962,
+      "grad_norm": 0.22453434765338898,
+      "learning_rate": 1.079584739097238e-05,
+      "loss": 0.2012,
+      "step": 7780
+    },
+    {
+      "epoch": 5.856981558148288,
+      "grad_norm": 0.20388460159301758,
+      "learning_rate": 1.0790863469592031e-05,
+      "loss": 0.1551,
+      "step": 7781
+    },
+    {
+      "epoch": 5.857734286789612,
+      "grad_norm": 0.17854540050029755,
+      "learning_rate": 1.0785881375497731e-05,
+      "loss": 0.2214,
+      "step": 7782
+    },
+    {
+      "epoch": 5.858487015430937,
+      "grad_norm": 0.1897020936012268,
+      "learning_rate": 1.0780901109234322e-05,
+      "loss": 0.1265,
+      "step": 7783
+    },
+    {
+      "epoch": 5.859239744072262,
+      "grad_norm": 0.17794173955917358,
+      "learning_rate": 1.0775922671346425e-05,
+      "loss": 0.1705,
+      "step": 7784
+    },
+    {
+      "epoch": 5.859992472713587,
+      "grad_norm": 0.18251720070838928,
+      "learning_rate": 1.0770946062378487e-05,
+      "loss": 0.1808,
+      "step": 7785
+    },
+    {
+      "epoch": 5.860745201354911,
+      "grad_norm": 0.1712121218442917,
+      "learning_rate": 1.076597128287473e-05,
+      "loss": 0.1621,
+      "step": 7786
+    },
+    {
+      "epoch": 5.861497929996236,
+      "grad_norm": 0.19697599112987518,
+      "learning_rate": 1.0760998333379199e-05,
+      "loss": 0.1717,
+      "step": 7787
+    },
+    {
+      "epoch": 5.862250658637561,
+      "grad_norm": 0.18092623353004456,
+      "learning_rate": 1.075602721443571e-05,
+      "loss": 0.2107,
+      "step": 7788
+    },
+    {
+      "epoch": 5.863003387278886,
+      "grad_norm": 0.21754738688468933,
+      "learning_rate": 1.0751057926587927e-05,
+      "loss": 0.1886,
+      "step": 7789
+    },
+    {
+      "epoch": 5.863756115920211,
+      "grad_norm": 0.16474993526935577,
+      "learning_rate": 1.0746090470379259e-05,
+      "loss": 0.127,
+      "step": 7790
+    },
+    {
+      "epoch": 5.864508844561535,
+      "grad_norm": 0.1855790913105011,
+      "learning_rate": 1.0741124846352951e-05,
+      "loss": 0.1584,
+      "step": 7791
+    },
+    {
+      "epoch": 5.865261573202861,
+      "grad_norm": 0.1879911720752716,
+      "learning_rate": 1.0736161055052031e-05,
+      "loss": 0.2076,
+      "step": 7792
+    },
+    {
+      "epoch": 5.866014301844185,
+      "grad_norm": 0.16984882950782776,
+      "learning_rate": 1.0731199097019334e-05,
+      "loss": 0.1808,
+      "step": 7793
+    },
+    {
+      "epoch": 5.86676703048551,
+      "grad_norm": 0.18410493433475494,
+      "learning_rate": 1.0726238972797492e-05,
+      "loss": 0.1906,
+      "step": 7794
+    },
+    {
+      "epoch": 5.867519759126835,
+      "grad_norm": 0.16657724976539612,
+      "learning_rate": 1.0721280682928942e-05,
+      "loss": 0.2578,
+      "step": 7795
+    },
+    {
+      "epoch": 5.868272487768159,
+      "grad_norm": 0.17646929621696472,
+      "learning_rate": 1.0716324227955904e-05,
+      "loss": 0.1275,
+      "step": 7796
+    },
+    {
+      "epoch": 5.869025216409485,
+      "grad_norm": 0.19322815537452698,
+      "learning_rate": 1.0711369608420414e-05,
+      "loss": 0.1807,
+      "step": 7797
+    },
+    {
+      "epoch": 5.869777945050809,
+      "grad_norm": 0.1778169423341751,
+      "learning_rate": 1.0706416824864311e-05,
+      "loss": 0.1648,
+      "step": 7798
+    },
+    {
+      "epoch": 5.870530673692134,
+      "grad_norm": 0.1833476573228836,
+      "learning_rate": 1.0701465877829205e-05,
+      "loss": 0.1806,
+      "step": 7799
+    },
+    {
+      "epoch": 5.871283402333459,
+      "grad_norm": 0.166673481464386,
+      "learning_rate": 1.0696516767856546e-05,
+      "loss": 0.1477,
+      "step": 7800
+    },
+    {
+      "epoch": 5.871283402333459,
+      "eval_loss": 0.20970036089420319,
+      "eval_runtime": 453.2426,
+      "eval_samples_per_second": 21.24,
+      "eval_steps_per_second": 0.664,
+      "step": 7800
+    },
+    {
+      "epoch": 5.8720361309747835,
+      "grad_norm": 0.16951106488704681,
+      "learning_rate": 1.0691569495487531e-05,
+      "loss": 0.1206,
+      "step": 7801
+    },
+    {
+      "epoch": 5.872788859616108,
+      "grad_norm": 0.16745300590991974,
+      "learning_rate": 1.068662406126322e-05,
+      "loss": 0.1711,
+      "step": 7802
+    },
+    {
+      "epoch": 5.873541588257433,
+      "grad_norm": 0.21482063829898834,
+      "learning_rate": 1.068168046572442e-05,
+      "loss": 0.2,
+      "step": 7803
+    },
+    {
+      "epoch": 5.874294316898758,
+      "grad_norm": 0.2368444800376892,
+      "learning_rate": 1.067673870941176e-05,
+      "loss": 0.1963,
+      "step": 7804
+    },
+    {
+      "epoch": 5.875047045540083,
+      "grad_norm": 0.19649024307727814,
+      "learning_rate": 1.0671798792865659e-05,
+      "loss": 0.1662,
+      "step": 7805
+    },
+    {
+      "epoch": 5.875799774181408,
+      "grad_norm": 0.170277401804924,
+      "learning_rate": 1.0666860716626347e-05,
+      "loss": 0.1202,
+      "step": 7806
+    },
+    {
+      "epoch": 5.876552502822732,
+      "grad_norm": 0.21207663416862488,
+      "learning_rate": 1.0661924481233823e-05,
+      "loss": 0.205,
+      "step": 7807
+    },
+    {
+      "epoch": 5.8773052314640575,
+      "grad_norm": 0.17658007144927979,
+      "learning_rate": 1.0656990087227938e-05,
+      "loss": 0.2329,
+      "step": 7808
+    },
+    {
+      "epoch": 5.878057960105382,
+      "grad_norm": 0.178333580493927,
+      "learning_rate": 1.0652057535148282e-05,
+      "loss": 0.2266,
+      "step": 7809
+    },
+    {
+      "epoch": 5.878810688746706,
+      "grad_norm": 0.20729613304138184,
+      "learning_rate": 1.0647126825534299e-05,
+      "loss": 0.1516,
+      "step": 7810
+    },
+    {
+      "epoch": 5.879563417388032,
+      "grad_norm": 0.18374569714069366,
+      "learning_rate": 1.0642197958925174e-05,
+      "loss": 0.2144,
+      "step": 7811
+    },
+    {
+      "epoch": 5.880316146029356,
+      "grad_norm": 0.19207534193992615,
+      "learning_rate": 1.0637270935859944e-05,
+      "loss": 0.1897,
+      "step": 7812
+    },
+    {
+      "epoch": 5.881068874670682,
+      "grad_norm": 0.16876280307769775,
+      "learning_rate": 1.0632345756877401e-05,
+      "loss": 0.1871,
+      "step": 7813
+    },
+    {
+      "epoch": 5.881821603312006,
+      "grad_norm": 0.17834530770778656,
+      "learning_rate": 1.0627422422516176e-05,
+      "loss": 0.1148,
+      "step": 7814
+    },
+    {
+      "epoch": 5.882574331953331,
+      "grad_norm": 0.1822824478149414,
+      "learning_rate": 1.0622500933314661e-05,
+      "loss": 0.189,
+      "step": 7815
+    },
+    {
+      "epoch": 5.883327060594656,
+      "grad_norm": 0.19745855033397675,
+      "learning_rate": 1.0617581289811077e-05,
+      "loss": 0.1415,
+      "step": 7816
+    },
+    {
+      "epoch": 5.88407978923598,
+      "grad_norm": 0.15290215611457825,
+      "learning_rate": 1.0612663492543412e-05,
+      "loss": 0.0842,
+      "step": 7817
+    },
+    {
+      "epoch": 5.884832517877305,
+      "grad_norm": 0.24663034081459045,
+      "learning_rate": 1.0607747542049479e-05,
+      "loss": 0.2027,
+      "step": 7818
+    },
+    {
+      "epoch": 5.88558524651863,
+      "grad_norm": 0.21029239892959595,
+      "learning_rate": 1.0602833438866886e-05,
+      "loss": 0.2403,
+      "step": 7819
+    },
+    {
+      "epoch": 5.886337975159955,
+      "grad_norm": 0.1718427836894989,
+      "learning_rate": 1.0597921183533017e-05,
+      "loss": 0.1066,
+      "step": 7820
+    },
+    {
+      "epoch": 5.887090703801279,
+      "grad_norm": 0.16433054208755493,
+      "learning_rate": 1.0593010776585074e-05,
+      "loss": 0.1443,
+      "step": 7821
+    },
+    {
+      "epoch": 5.8878434324426046,
+      "grad_norm": 0.19389720261096954,
+      "learning_rate": 1.0588102218560054e-05,
+      "loss": 0.1876,
+      "step": 7822
+    },
+    {
+      "epoch": 5.888596161083929,
+      "grad_norm": 0.1851845681667328,
+      "learning_rate": 1.0583195509994752e-05,
+      "loss": 0.2028,
+      "step": 7823
+    },
+    {
+      "epoch": 5.889348889725254,
+      "grad_norm": 0.17869111895561218,
+      "learning_rate": 1.057829065142575e-05,
+      "loss": 0.2132,
+      "step": 7824
+    },
+    {
+      "epoch": 5.890101618366579,
+      "grad_norm": 0.17694218456745148,
+      "learning_rate": 1.0573387643389451e-05,
+      "loss": 0.2541,
+      "step": 7825
+    },
+    {
+      "epoch": 5.890854347007903,
+      "grad_norm": 0.1907970905303955,
+      "learning_rate": 1.0568486486422013e-05,
+      "loss": 0.118,
+      "step": 7826
+    },
+    {
+      "epoch": 5.891607075649229,
+      "grad_norm": 0.1678181141614914,
+      "learning_rate": 1.0563587181059451e-05,
+      "loss": 0.1126,
+      "step": 7827
+    },
+    {
+      "epoch": 5.892359804290553,
+      "grad_norm": 0.16499756276607513,
+      "learning_rate": 1.0558689727837525e-05,
+      "loss": 0.1621,
+      "step": 7828
+    },
+    {
+      "epoch": 5.8931125329318785,
+      "grad_norm": 0.17301584780216217,
+      "learning_rate": 1.0553794127291827e-05,
+      "loss": 0.0852,
+      "step": 7829
+    },
+    {
+      "epoch": 5.893865261573203,
+      "grad_norm": 0.17573325335979462,
+      "learning_rate": 1.0548900379957718e-05,
+      "loss": 0.1542,
+      "step": 7830
+    },
+    {
+      "epoch": 5.8946179902145275,
+      "grad_norm": 0.1443299949169159,
+      "learning_rate": 1.0544008486370384e-05,
+      "loss": 0.1292,
+      "step": 7831
+    },
+    {
+      "epoch": 5.895370718855853,
+      "grad_norm": 0.16085167229175568,
+      "learning_rate": 1.0539118447064775e-05,
+      "loss": 0.1585,
+      "step": 7832
+    },
+    {
+      "epoch": 5.896123447497177,
+      "grad_norm": 0.21097628772258759,
+      "learning_rate": 1.0534230262575689e-05,
+      "loss": 0.1742,
+      "step": 7833
+    },
+    {
+      "epoch": 5.896876176138502,
+      "grad_norm": 0.1785847395658493,
+      "learning_rate": 1.0529343933437664e-05,
+      "loss": 0.1103,
+      "step": 7834
+    },
+    {
+      "epoch": 5.897628904779827,
+      "grad_norm": 0.2301696091890335,
+      "learning_rate": 1.0524459460185079e-05,
+      "loss": 0.2705,
+      "step": 7835
+    },
+    {
+      "epoch": 5.898381633421152,
+      "grad_norm": 0.18194139003753662,
+      "learning_rate": 1.0519576843352077e-05,
+      "loss": 0.1856,
+      "step": 7836
+    },
+    {
+      "epoch": 5.899134362062476,
+      "grad_norm": 0.1797322928905487,
+      "learning_rate": 1.051469608347263e-05,
+      "loss": 0.1298,
+      "step": 7837
+    },
+    {
+      "epoch": 5.8998870907038015,
+      "grad_norm": 0.18591181933879852,
+      "learning_rate": 1.0509817181080476e-05,
+      "loss": 0.2008,
+      "step": 7838
+    },
+    {
+      "epoch": 5.900639819345126,
+      "grad_norm": 0.22504335641860962,
+      "learning_rate": 1.0504940136709172e-05,
+      "loss": 0.2371,
+      "step": 7839
+    },
+    {
+      "epoch": 5.90139254798645,
+      "grad_norm": 0.18718086183071136,
+      "learning_rate": 1.0500064950892061e-05,
+      "loss": 0.141,
+      "step": 7840
+    },
+    {
+      "epoch": 5.902145276627776,
+      "grad_norm": 0.18510393798351288,
+      "learning_rate": 1.0495191624162295e-05,
+      "loss": 0.1841,
+      "step": 7841
+    },
+    {
+      "epoch": 5.9028980052691,
+      "grad_norm": 0.19794760644435883,
+      "learning_rate": 1.0490320157052798e-05,
+      "loss": 0.1564,
+      "step": 7842
+    },
+    {
+      "epoch": 5.903650733910426,
+      "grad_norm": 0.26055872440338135,
+      "learning_rate": 1.0485450550096318e-05,
+      "loss": 0.1734,
+      "step": 7843
+    },
+    {
+      "epoch": 5.90440346255175,
+      "grad_norm": 0.18747378885746002,
+      "learning_rate": 1.048058280382539e-05,
+      "loss": 0.2096,
+      "step": 7844
+    },
+    {
+      "epoch": 5.9051561911930746,
+      "grad_norm": 0.1708677113056183,
+      "learning_rate": 1.0475716918772329e-05,
+      "loss": 0.1821,
+      "step": 7845
+    },
+    {
+      "epoch": 5.9059089198344,
+      "grad_norm": 0.18697457015514374,
+      "learning_rate": 1.0470852895469275e-05,
+      "loss": 0.1669,
+      "step": 7846
+    },
+    {
+      "epoch": 5.906661648475724,
+      "grad_norm": 0.20160262286663055,
+      "learning_rate": 1.0465990734448142e-05,
+      "loss": 0.1736,
+      "step": 7847
+    },
+    {
+      "epoch": 5.90741437711705,
+      "grad_norm": 0.15979957580566406,
+      "learning_rate": 1.0461130436240663e-05,
+      "loss": 0.1076,
+      "step": 7848
+    },
+    {
+      "epoch": 5.908167105758374,
+      "grad_norm": 0.1732211709022522,
+      "learning_rate": 1.0456272001378331e-05,
+      "loss": 0.1645,
+      "step": 7849
+    },
+    {
+      "epoch": 5.908919834399699,
+      "grad_norm": 0.20390279591083527,
+      "learning_rate": 1.0451415430392474e-05,
+      "loss": 0.1296,
+      "step": 7850
+    },
+    {
+      "epoch": 5.909672563041024,
+      "grad_norm": 0.17691653966903687,
+      "learning_rate": 1.044656072381419e-05,
+      "loss": 0.1361,
+      "step": 7851
+    },
+    {
+      "epoch": 5.9104252916823485,
+      "grad_norm": 0.1884659379720688,
+      "learning_rate": 1.0441707882174384e-05,
+      "loss": 0.157,
+      "step": 7852
+    },
+    {
+      "epoch": 5.911178020323673,
+      "grad_norm": 0.17874178290367126,
+      "learning_rate": 1.0436856906003758e-05,
+      "loss": 0.1187,
+      "step": 7853
+    },
+    {
+      "epoch": 5.911930748964998,
+      "grad_norm": 0.17289064824581146,
+      "learning_rate": 1.0432007795832813e-05,
+      "loss": 0.1216,
+      "step": 7854
+    },
+    {
+      "epoch": 5.912683477606323,
+      "grad_norm": 0.20610016584396362,
+      "learning_rate": 1.0427160552191827e-05,
+      "loss": 0.24,
+      "step": 7855
+    },
+    {
+      "epoch": 5.913436206247647,
+      "grad_norm": 0.19455622136592865,
+      "learning_rate": 1.0422315175610902e-05,
+      "loss": 0.1197,
+      "step": 7856
+    },
+    {
+      "epoch": 5.914188934888973,
+      "grad_norm": 0.19501039385795593,
+      "learning_rate": 1.0417471666619905e-05,
+      "loss": 0.2201,
+      "step": 7857
+    },
+    {
+      "epoch": 5.914941663530297,
+      "grad_norm": 0.18510055541992188,
+      "learning_rate": 1.0412630025748527e-05,
+      "loss": 0.1589,
+      "step": 7858
+    },
+    {
+      "epoch": 5.9156943921716225,
+      "grad_norm": 0.18614967167377472,
+      "learning_rate": 1.0407790253526236e-05,
+      "loss": 0.166,
+      "step": 7859
+    },
+    {
+      "epoch": 5.916447120812947,
+      "grad_norm": 0.20401036739349365,
+      "learning_rate": 1.0402952350482312e-05,
+      "loss": 0.1932,
+      "step": 7860
+    },
+    {
+      "epoch": 5.9171998494542715,
+      "grad_norm": 0.17207644879817963,
+      "learning_rate": 1.0398116317145807e-05,
+      "loss": 0.1573,
+      "step": 7861
+    },
+    {
+      "epoch": 5.917952578095597,
+      "grad_norm": 0.1763281226158142,
+      "learning_rate": 1.0393282154045598e-05,
+      "loss": 0.0878,
+      "step": 7862
+    },
+    {
+      "epoch": 5.918705306736921,
+      "grad_norm": 0.19537658989429474,
+      "learning_rate": 1.0388449861710323e-05,
+      "loss": 0.1258,
+      "step": 7863
+    },
+    {
+      "epoch": 5.919458035378246,
+      "grad_norm": 0.19857269525527954,
+      "learning_rate": 1.0383619440668449e-05,
+      "loss": 0.2217,
+      "step": 7864
+    },
+    {
+      "epoch": 5.920210764019571,
+      "grad_norm": 0.17926359176635742,
+      "learning_rate": 1.0378790891448215e-05,
+      "loss": 0.1899,
+      "step": 7865
+    },
+    {
+      "epoch": 5.920963492660896,
+      "grad_norm": 0.192916139960289,
+      "learning_rate": 1.0373964214577675e-05,
+      "loss": 0.224,
+      "step": 7866
+    },
+    {
+      "epoch": 5.921716221302221,
+      "grad_norm": 0.16569092869758606,
+      "learning_rate": 1.0369139410584654e-05,
+      "loss": 0.1553,
+      "step": 7867
+    },
+    {
+      "epoch": 5.9224689499435454,
+      "grad_norm": 0.1653866320848465,
+      "learning_rate": 1.0364316479996792e-05,
+      "loss": 0.1392,
+      "step": 7868
+    },
+    {
+      "epoch": 5.92322167858487,
+      "grad_norm": 0.18908195197582245,
+      "learning_rate": 1.035949542334152e-05,
+      "loss": 0.1654,
+      "step": 7869
+    },
+    {
+      "epoch": 5.923974407226195,
+      "grad_norm": 0.19136519730091095,
+      "learning_rate": 1.0354676241146055e-05,
+      "loss": 0.2146,
+      "step": 7870
+    },
+    {
+      "epoch": 5.92472713586752,
+      "grad_norm": 0.17178978025913239,
+      "learning_rate": 1.0349858933937417e-05,
+      "loss": 0.1429,
+      "step": 7871
+    },
+    {
+      "epoch": 5.925479864508844,
+      "grad_norm": 0.19063693284988403,
+      "learning_rate": 1.0345043502242422e-05,
+      "loss": 0.2203,
+      "step": 7872
+    },
+    {
+      "epoch": 5.92623259315017,
+      "grad_norm": 0.17472364008426666,
+      "learning_rate": 1.0340229946587683e-05,
+      "loss": 0.168,
+      "step": 7873
+    },
+    {
+      "epoch": 5.926985321791494,
+      "grad_norm": 0.20320269465446472,
+      "learning_rate": 1.0335418267499593e-05,
+      "loss": 0.2012,
+      "step": 7874
+    },
+    {
+      "epoch": 5.9277380504328185,
+      "grad_norm": 0.17805790901184082,
+      "learning_rate": 1.0330608465504357e-05,
+      "loss": 0.2135,
+      "step": 7875
+    },
+    {
+      "epoch": 5.928490779074144,
+      "grad_norm": 0.18371546268463135,
+      "learning_rate": 1.0325800541127964e-05,
+      "loss": 0.1246,
+      "step": 7876
+    },
+    {
+      "epoch": 5.929243507715468,
+      "grad_norm": 0.19280579686164856,
+      "learning_rate": 1.03209944948962e-05,
+      "loss": 0.1662,
+      "step": 7877
+    },
+    {
+      "epoch": 5.929996236356794,
+      "grad_norm": 0.16670307517051697,
+      "learning_rate": 1.031619032733465e-05,
+      "loss": 0.0869,
+      "step": 7878
+    },
+    {
+      "epoch": 5.930748964998118,
+      "grad_norm": 0.20357339084148407,
+      "learning_rate": 1.0311388038968701e-05,
+      "loss": 0.1299,
+      "step": 7879
+    },
+    {
+      "epoch": 5.931501693639443,
+      "grad_norm": 0.16486714780330658,
+      "learning_rate": 1.03065876303235e-05,
+      "loss": 0.1384,
+      "step": 7880
+    },
+    {
+      "epoch": 5.932254422280768,
+      "grad_norm": 0.18535013496875763,
+      "learning_rate": 1.0301789101924035e-05,
+      "loss": 0.1985,
+      "step": 7881
+    },
+    {
+      "epoch": 5.9330071509220925,
+      "grad_norm": 0.19591061770915985,
+      "learning_rate": 1.0296992454295052e-05,
+      "loss": 0.2187,
+      "step": 7882
+    },
+    {
+      "epoch": 5.933759879563418,
+      "grad_norm": 0.162959024310112,
+      "learning_rate": 1.0292197687961108e-05,
+      "loss": 0.1556,
+      "step": 7883
+    },
+    {
+      "epoch": 5.934512608204742,
+      "grad_norm": 0.18426793813705444,
+      "learning_rate": 1.0287404803446556e-05,
+      "loss": 0.1667,
+      "step": 7884
+    },
+    {
+      "epoch": 5.935265336846067,
+      "grad_norm": 0.1855568140745163,
+      "learning_rate": 1.0282613801275539e-05,
+      "loss": 0.1557,
+      "step": 7885
+    },
+    {
+      "epoch": 5.936018065487392,
+      "grad_norm": 0.1766507625579834,
+      "learning_rate": 1.0277824681971985e-05,
+      "loss": 0.123,
+      "step": 7886
+    },
+    {
+      "epoch": 5.936770794128717,
+      "grad_norm": 0.1938115954399109,
+      "learning_rate": 1.0273037446059638e-05,
+      "loss": 0.1836,
+      "step": 7887
+    },
+    {
+      "epoch": 5.937523522770041,
+      "grad_norm": 0.18180306255817413,
+      "learning_rate": 1.026825209406201e-05,
+      "loss": 0.2056,
+      "step": 7888
+    },
+    {
+      "epoch": 5.9382762514113665,
+      "grad_norm": 0.2217472344636917,
+      "learning_rate": 1.0263468626502424e-05,
+      "loss": 0.1844,
+      "step": 7889
+    },
+    {
+      "epoch": 5.939028980052691,
+      "grad_norm": 0.17871296405792236,
+      "learning_rate": 1.0258687043903993e-05,
+      "loss": 0.1017,
+      "step": 7890
+    },
+    {
+      "epoch": 5.9397817086940154,
+      "grad_norm": 0.2000368982553482,
+      "learning_rate": 1.0253907346789632e-05,
+      "loss": 0.2176,
+      "step": 7891
+    },
+    {
+      "epoch": 5.940534437335341,
+      "grad_norm": 0.1860024482011795,
+      "learning_rate": 1.0249129535682027e-05,
+      "loss": 0.2543,
+      "step": 7892
+    },
+    {
+      "epoch": 5.941287165976665,
+      "grad_norm": 0.18554732203483582,
+      "learning_rate": 1.0244353611103685e-05,
+      "loss": 0.2412,
+      "step": 7893
+    },
+    {
+      "epoch": 5.94203989461799,
+      "grad_norm": 0.16688817739486694,
+      "learning_rate": 1.023957957357689e-05,
+      "loss": 0.1538,
+      "step": 7894
+    },
+    {
+      "epoch": 5.942792623259315,
+      "grad_norm": 0.22165672481060028,
+      "learning_rate": 1.0234807423623718e-05,
+      "loss": 0.2317,
+      "step": 7895
+    },
+    {
+      "epoch": 5.94354535190064,
+      "grad_norm": 0.1881219744682312,
+      "learning_rate": 1.0230037161766049e-05,
+      "loss": 0.3106,
+      "step": 7896
+    },
+    {
+      "epoch": 5.944298080541965,
+      "grad_norm": 0.23244896531105042,
+      "learning_rate": 1.0225268788525548e-05,
+      "loss": 0.2134,
+      "step": 7897
+    },
+    {
+      "epoch": 5.945050809183289,
+      "grad_norm": 0.21213982999324799,
+      "learning_rate": 1.0220502304423687e-05,
+      "loss": 0.1659,
+      "step": 7898
+    },
+    {
+      "epoch": 5.945803537824614,
+      "grad_norm": 0.19288165867328644,
+      "learning_rate": 1.0215737709981713e-05,
+      "loss": 0.2072,
+      "step": 7899
+    },
+    {
+      "epoch": 5.946556266465939,
+      "grad_norm": 0.1883694976568222,
+      "learning_rate": 1.0210975005720677e-05,
+      "loss": 0.159,
+      "step": 7900
+    },
+    {
+      "epoch": 5.947308995107264,
+      "grad_norm": 0.2634642422199249,
+      "learning_rate": 1.0206214192161415e-05,
+      "loss": 0.2158,
+      "step": 7901
+    },
+    {
+      "epoch": 5.948061723748589,
+      "grad_norm": 0.1863466054201126,
+      "learning_rate": 1.020145526982457e-05,
+      "loss": 0.2122,
+      "step": 7902
+    },
+    {
+      "epoch": 5.948814452389914,
+      "grad_norm": 0.1522994339466095,
+      "learning_rate": 1.0196698239230571e-05,
+      "loss": 0.1287,
+      "step": 7903
+    },
+    {
+      "epoch": 5.949567181031238,
+      "grad_norm": 0.19483941793441772,
+      "learning_rate": 1.019194310089964e-05,
+      "loss": 0.1586,
+      "step": 7904
+    },
+    {
+      "epoch": 5.950319909672563,
+      "grad_norm": 0.1860089749097824,
+      "learning_rate": 1.0187189855351784e-05,
+      "loss": 0.1734,
+      "step": 7905
+    },
+    {
+      "epoch": 5.951072638313888,
+      "grad_norm": 0.178179532289505,
+      "learning_rate": 1.0182438503106825e-05,
+      "loss": 0.155,
+      "step": 7906
+    },
+    {
+      "epoch": 5.951825366955212,
+      "grad_norm": 0.19346116483211517,
+      "learning_rate": 1.0177689044684343e-05,
+      "loss": 0.1956,
+      "step": 7907
+    },
+    {
+      "epoch": 5.952578095596538,
+      "grad_norm": 0.21126465499401093,
+      "learning_rate": 1.0172941480603754e-05,
+      "loss": 0.1752,
+      "step": 7908
+    },
+    {
+      "epoch": 5.953330824237862,
+      "grad_norm": 0.17846469581127167,
+      "learning_rate": 1.0168195811384219e-05,
+      "loss": 0.152,
+      "step": 7909
+    },
+    {
+      "epoch": 5.954083552879187,
+      "grad_norm": 0.1627771109342575,
+      "learning_rate": 1.0163452037544746e-05,
+      "loss": 0.2242,
+      "step": 7910
+    },
+    {
+      "epoch": 5.954836281520512,
+      "grad_norm": 0.21515361964702606,
+      "learning_rate": 1.0158710159604088e-05,
+      "loss": 0.2019,
+      "step": 7911
+    },
+    {
+      "epoch": 5.9555890101618365,
+      "grad_norm": 0.16429246962070465,
+      "learning_rate": 1.0153970178080816e-05,
+      "loss": 0.1444,
+      "step": 7912
+    },
+    {
+      "epoch": 5.956341738803161,
+      "grad_norm": 0.21398895978927612,
+      "learning_rate": 1.0149232093493284e-05,
+      "loss": 0.2028,
+      "step": 7913
+    },
+    {
+      "epoch": 5.957094467444486,
+      "grad_norm": 0.18086136877536774,
+      "learning_rate": 1.0144495906359644e-05,
+      "loss": 0.2247,
+      "step": 7914
+    },
+    {
+      "epoch": 5.957847196085811,
+      "grad_norm": 0.159880593419075,
+      "learning_rate": 1.0139761617197835e-05,
+      "loss": 0.1563,
+      "step": 7915
+    },
+    {
+      "epoch": 5.958599924727136,
+      "grad_norm": 0.1840042769908905,
+      "learning_rate": 1.0135029226525597e-05,
+      "loss": 0.1401,
+      "step": 7916
+    },
+    {
+      "epoch": 5.959352653368461,
+      "grad_norm": 0.15724746882915497,
+      "learning_rate": 1.0130298734860462e-05,
+      "loss": 0.1131,
+      "step": 7917
+    },
+    {
+      "epoch": 5.960105382009785,
+      "grad_norm": 0.1991526484489441,
+      "learning_rate": 1.0125570142719734e-05,
+      "loss": 0.1994,
+      "step": 7918
+    },
+    {
+      "epoch": 5.9608581106511105,
+      "grad_norm": 0.18658261001110077,
+      "learning_rate": 1.0120843450620541e-05,
+      "loss": 0.1425,
+      "step": 7919
+    },
+    {
+      "epoch": 5.961610839292435,
+      "grad_norm": 0.1808723658323288,
+      "learning_rate": 1.011611865907977e-05,
+      "loss": 0.1825,
+      "step": 7920
+    },
+    {
+      "epoch": 5.96236356793376,
+      "grad_norm": 0.17133557796478271,
+      "learning_rate": 1.0111395768614138e-05,
+      "loss": 0.1283,
+      "step": 7921
+    },
+    {
+      "epoch": 5.963116296575085,
+      "grad_norm": 0.17711491882801056,
+      "learning_rate": 1.0106674779740103e-05,
+      "loss": 0.1938,
+      "step": 7922
+    },
+    {
+      "epoch": 5.963869025216409,
+      "grad_norm": 0.17953535914421082,
+      "learning_rate": 1.010195569297398e-05,
+      "loss": 0.2208,
+      "step": 7923
+    },
+    {
+      "epoch": 5.964621753857735,
+      "grad_norm": 0.1949918568134308,
+      "learning_rate": 1.0097238508831817e-05,
+      "loss": 0.2207,
+      "step": 7924
+    },
+    {
+      "epoch": 5.965374482499059,
+      "grad_norm": 0.18875020742416382,
+      "learning_rate": 1.0092523227829493e-05,
+      "loss": 0.2366,
+      "step": 7925
+    },
+    {
+      "epoch": 5.966127211140384,
+      "grad_norm": 0.180150106549263,
+      "learning_rate": 1.008780985048265e-05,
+      "loss": 0.1691,
+      "step": 7926
+    },
+    {
+      "epoch": 5.966879939781709,
+      "grad_norm": 0.16865652799606323,
+      "learning_rate": 1.0083098377306748e-05,
+      "loss": 0.1513,
+      "step": 7927
+    },
+    {
+      "epoch": 5.967632668423033,
+      "grad_norm": 0.1703604757785797,
+      "learning_rate": 1.007838880881701e-05,
+      "loss": 0.1318,
+      "step": 7928
+    },
+    {
+      "epoch": 5.968385397064358,
+      "grad_norm": 0.18322737514972687,
+      "learning_rate": 1.0073681145528491e-05,
+      "loss": 0.1361,
+      "step": 7929
+    },
+    {
+      "epoch": 5.969138125705683,
+      "grad_norm": 0.18802011013031006,
+      "learning_rate": 1.0068975387955997e-05,
+      "loss": 0.2236,
+      "step": 7930
+    },
+    {
+      "epoch": 5.969890854347008,
+      "grad_norm": 0.16607120633125305,
+      "learning_rate": 1.0064271536614153e-05,
+      "loss": 0.1294,
+      "step": 7931
+    },
+    {
+      "epoch": 5.970643582988333,
+      "grad_norm": 0.17561207711696625,
+      "learning_rate": 1.0059569592017354e-05,
+      "loss": 0.1609,
+      "step": 7932
+    },
+    {
+      "epoch": 5.971396311629658,
+      "grad_norm": 0.17743906378746033,
+      "learning_rate": 1.0054869554679808e-05,
+      "loss": 0.1864,
+      "step": 7933
+    },
+    {
+      "epoch": 5.972149040270982,
+      "grad_norm": 0.16574592888355255,
+      "learning_rate": 1.0050171425115487e-05,
+      "loss": 0.1355,
+      "step": 7934
+    },
+    {
+      "epoch": 5.972901768912307,
+      "grad_norm": 0.18215009570121765,
+      "learning_rate": 1.0045475203838198e-05,
+      "loss": 0.14,
+      "step": 7935
+    },
+    {
+      "epoch": 5.973654497553632,
+      "grad_norm": 0.17594365775585175,
+      "learning_rate": 1.0040780891361489e-05,
+      "loss": 0.2006,
+      "step": 7936
+    },
+    {
+      "epoch": 5.974407226194956,
+      "grad_norm": 0.16771769523620605,
+      "learning_rate": 1.0036088488198733e-05,
+      "loss": 0.174,
+      "step": 7937
+    },
+    {
+      "epoch": 5.975159954836282,
+      "grad_norm": 0.17592336237430573,
+      "learning_rate": 1.003139799486309e-05,
+      "loss": 0.1508,
+      "step": 7938
+    },
+    {
+      "epoch": 5.975912683477606,
+      "grad_norm": 0.16094255447387695,
+      "learning_rate": 1.0026709411867493e-05,
+      "loss": 0.1918,
+      "step": 7939
+    },
+    {
+      "epoch": 5.9766654121189315,
+      "grad_norm": 0.21374765038490295,
+      "learning_rate": 1.0022022739724688e-05,
+      "loss": 0.2296,
+      "step": 7940
+    },
+    {
+      "epoch": 5.977418140760256,
+      "grad_norm": 0.18074874579906464,
+      "learning_rate": 1.001733797894719e-05,
+      "loss": 0.2167,
+      "step": 7941
+    },
+    {
+      "epoch": 5.9781708694015805,
+      "grad_norm": 0.1863674819469452,
+      "learning_rate": 1.0012655130047335e-05,
+      "loss": 0.1457,
+      "step": 7942
+    },
+    {
+      "epoch": 5.978923598042906,
+      "grad_norm": 0.1714567244052887,
+      "learning_rate": 1.0007974193537218e-05,
+      "loss": 0.2006,
+      "step": 7943
+    },
+    {
+      "epoch": 5.97967632668423,
+      "grad_norm": 0.16577856242656708,
+      "learning_rate": 1.000329516992875e-05,
+      "loss": 0.1094,
+      "step": 7944
+    },
+    {
+      "epoch": 5.980429055325555,
+      "grad_norm": 0.15562570095062256,
+      "learning_rate": 9.99861805973361e-06,
+      "loss": 0.1224,
+      "step": 7945
+    },
+    {
+      "epoch": 5.98118178396688,
+      "grad_norm": 0.17697042226791382,
+      "learning_rate": 9.993942863463295e-06,
+      "loss": 0.2836,
+      "step": 7946
+    },
+    {
+      "epoch": 5.981934512608205,
+      "grad_norm": 0.19623316824436188,
+      "learning_rate": 9.989269581629052e-06,
+      "loss": 0.1917,
+      "step": 7947
+    },
+    {
+      "epoch": 5.982687241249529,
+      "grad_norm": 0.16522882878780365,
+      "learning_rate": 9.984598214741975e-06,
+      "loss": 0.1562,
+      "step": 7948
+    },
+    {
+      "epoch": 5.9834399698908545,
+      "grad_norm": 0.19221879541873932,
+      "learning_rate": 9.9799287633129e-06,
+      "loss": 0.1711,
+      "step": 7949
+    },
+    {
+      "epoch": 5.984192698532179,
+      "grad_norm": 0.16842840611934662,
+      "learning_rate": 9.975261227852472e-06,
+      "loss": 0.1243,
+      "step": 7950
+    },
+    {
+      "epoch": 5.984945427173504,
+      "grad_norm": 0.18317899107933044,
+      "learning_rate": 9.97059560887113e-06,
+      "loss": 0.1735,
+      "step": 7951
+    },
+    {
+      "epoch": 5.985698155814829,
+      "grad_norm": 0.1743563711643219,
+      "learning_rate": 9.9659319068791e-06,
+      "loss": 0.1272,
+      "step": 7952
+    },
+    {
+      "epoch": 5.986450884456153,
+      "grad_norm": 0.22248855233192444,
+      "learning_rate": 9.961270122386385e-06,
+      "loss": 0.24,
+      "step": 7953
+    },
+    {
+      "epoch": 5.987203613097479,
+      "grad_norm": 0.1956307590007782,
+      "learning_rate": 9.956610255902809e-06,
+      "loss": 0.1537,
+      "step": 7954
+    },
+    {
+      "epoch": 5.987956341738803,
+      "grad_norm": 0.17051708698272705,
+      "learning_rate": 9.951952307937953e-06,
+      "loss": 0.2122,
+      "step": 7955
+    },
+    {
+      "epoch": 5.9887090703801285,
+      "grad_norm": 0.20562376081943512,
+      "learning_rate": 9.947296279001219e-06,
+      "loss": 0.2091,
+      "step": 7956
+    },
+    {
+      "epoch": 5.989461799021453,
+      "grad_norm": 0.1698097139596939,
+      "learning_rate": 9.942642169601765e-06,
+      "loss": 0.2144,
+      "step": 7957
+    },
+    {
+      "epoch": 5.990214527662777,
+      "grad_norm": 0.1670960634946823,
+      "learning_rate": 9.937989980248569e-06,
+      "loss": 0.2436,
+      "step": 7958
+    },
+    {
+      "epoch": 5.990967256304103,
+      "grad_norm": 0.16151118278503418,
+      "learning_rate": 9.93333971145039e-06,
+      "loss": 0.107,
+      "step": 7959
+    },
+    {
+      "epoch": 5.991719984945427,
+      "grad_norm": 0.14876770973205566,
+      "learning_rate": 9.928691363715762e-06,
+      "loss": 0.1266,
+      "step": 7960
+    },
+    {
+      "epoch": 5.992472713586752,
+      "grad_norm": 0.17624466121196747,
+      "learning_rate": 9.924044937553026e-06,
+      "loss": 0.1904,
+      "step": 7961
+    },
+    {
+      "epoch": 5.993225442228077,
+      "grad_norm": 0.17551475763320923,
+      "learning_rate": 9.919400433470314e-06,
+      "loss": 0.1677,
+      "step": 7962
+    },
+    {
+      "epoch": 5.9939781708694015,
+      "grad_norm": 0.17941442131996155,
+      "learning_rate": 9.914757851975543e-06,
+      "loss": 0.2402,
+      "step": 7963
+    },
+    {
+      "epoch": 5.994730899510726,
+      "grad_norm": 0.18911607563495636,
+      "learning_rate": 9.91011719357641e-06,
+      "loss": 0.1901,
+      "step": 7964
+    },
+    {
+      "epoch": 5.995483628152051,
+      "grad_norm": 0.1940196007490158,
+      "learning_rate": 9.905478458780417e-06,
+      "loss": 0.1842,
+      "step": 7965
+    },
+    {
+      "epoch": 5.996236356793376,
+      "grad_norm": 0.15967459976673126,
+      "learning_rate": 9.90084164809484e-06,
+      "loss": 0.1668,
+      "step": 7966
+    },
+    {
+      "epoch": 5.9969890854347,
+      "grad_norm": 0.18763099610805511,
+      "learning_rate": 9.896206762026769e-06,
+      "loss": 0.1642,
+      "step": 7967
+    },
+    {
+      "epoch": 5.997741814076026,
+      "grad_norm": 0.21454210579395294,
+      "learning_rate": 9.891573801083055e-06,
+      "loss": 0.1989,
+      "step": 7968
+    },
+    {
+      "epoch": 5.99849454271735,
+      "grad_norm": 0.20020495355129242,
+      "learning_rate": 9.886942765770361e-06,
+      "loss": 0.2025,
+      "step": 7969
+    },
+    {
+      "epoch": 5.9992472713586755,
+      "grad_norm": 0.1932109296321869,
+      "learning_rate": 9.882313656595118e-06,
+      "loss": 0.1431,
+      "step": 7970
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 0.17852620780467987,
+      "learning_rate": 9.877686474063575e-06,
+      "loss": 0.2089,
+      "step": 7971
+    },
+    {
+      "epoch": 6.0007527286413245,
+      "grad_norm": 0.17070230841636658,
+      "learning_rate": 9.873061218681737e-06,
+      "loss": 0.1556,
+      "step": 7972
+    },
+    {
+      "epoch": 6.00150545728265,
+      "grad_norm": 0.2165829986333847,
+      "learning_rate": 9.868437890955424e-06,
+      "loss": 0.2236,
+      "step": 7973
+    },
+    {
+      "epoch": 6.002258185923974,
+      "grad_norm": 0.16307607293128967,
+      "learning_rate": 9.863816491390232e-06,
+      "loss": 0.1636,
+      "step": 7974
+    },
+    {
+      "epoch": 6.003010914565299,
+      "grad_norm": 0.19232909381389618,
+      "learning_rate": 9.859197020491559e-06,
+      "loss": 0.1744,
+      "step": 7975
+    },
+    {
+      "epoch": 6.003763643206624,
+      "grad_norm": 0.18801036477088928,
+      "learning_rate": 9.854579478764573e-06,
+      "loss": 0.1456,
+      "step": 7976
+    },
+    {
+      "epoch": 6.004516371847949,
+      "grad_norm": 0.1729387491941452,
+      "learning_rate": 9.849963866714252e-06,
+      "loss": 0.1301,
+      "step": 7977
+    },
+    {
+      "epoch": 6.005269100489274,
+      "grad_norm": 0.2237861007452011,
+      "learning_rate": 9.845350184845338e-06,
+      "loss": 0.1916,
+      "step": 7978
+    },
+    {
+      "epoch": 6.0060218291305985,
+      "grad_norm": 0.1847735047340393,
+      "learning_rate": 9.840738433662386e-06,
+      "loss": 0.2382,
+      "step": 7979
+    },
+    {
+      "epoch": 6.006774557771923,
+      "grad_norm": 0.16050522029399872,
+      "learning_rate": 9.836128613669726e-06,
+      "loss": 0.2301,
+      "step": 7980
+    },
+    {
+      "epoch": 6.007527286413248,
+      "grad_norm": 0.1908998042345047,
+      "learning_rate": 9.831520725371493e-06,
+      "loss": 0.1783,
+      "step": 7981
+    },
+    {
+      "epoch": 6.008280015054573,
+      "grad_norm": 0.1917969286441803,
+      "learning_rate": 9.826914769271581e-06,
+      "loss": 0.1779,
+      "step": 7982
+    },
+    {
+      "epoch": 6.009032743695897,
+      "grad_norm": 0.16535179316997528,
+      "learning_rate": 9.822310745873701e-06,
+      "loss": 0.1269,
+      "step": 7983
+    },
+    {
+      "epoch": 6.009785472337223,
+      "grad_norm": 0.18845775723457336,
+      "learning_rate": 9.817708655681345e-06,
+      "loss": 0.1276,
+      "step": 7984
+    },
+    {
+      "epoch": 6.010538200978547,
+      "grad_norm": 0.17131680250167847,
+      "learning_rate": 9.81310849919778e-06,
+      "loss": 0.1801,
+      "step": 7985
+    },
+    {
+      "epoch": 6.011290929619872,
+      "grad_norm": 0.19359616935253143,
+      "learning_rate": 9.808510276926075e-06,
+      "loss": 0.1899,
+      "step": 7986
+    },
+    {
+      "epoch": 6.012043658261197,
+      "grad_norm": 0.18386000394821167,
+      "learning_rate": 9.80391398936909e-06,
+      "loss": 0.2175,
+      "step": 7987
+    },
+    {
+      "epoch": 6.012796386902521,
+      "grad_norm": 0.20390352606773376,
+      "learning_rate": 9.79931963702947e-06,
+      "loss": 0.163,
+      "step": 7988
+    },
+    {
+      "epoch": 6.013549115543847,
+      "grad_norm": 0.180230975151062,
+      "learning_rate": 9.794727220409635e-06,
+      "loss": 0.1717,
+      "step": 7989
+    },
+    {
+      "epoch": 6.014301844185171,
+      "grad_norm": 0.17502504587173462,
+      "learning_rate": 9.79013674001182e-06,
+      "loss": 0.2089,
+      "step": 7990
+    },
+    {
+      "epoch": 6.015054572826496,
+      "grad_norm": 0.2125883251428604,
+      "learning_rate": 9.785548196338018e-06,
+      "loss": 0.1643,
+      "step": 7991
+    },
+    {
+      "epoch": 6.015807301467821,
+      "grad_norm": 0.17926263809204102,
+      "learning_rate": 9.780961589890033e-06,
+      "loss": 0.1761,
+      "step": 7992
+    },
+    {
+      "epoch": 6.0165600301091455,
+      "grad_norm": 0.18224923312664032,
+      "learning_rate": 9.776376921169444e-06,
+      "loss": 0.1475,
+      "step": 7993
+    },
+    {
+      "epoch": 6.017312758750471,
+      "grad_norm": 0.19006119668483734,
+      "learning_rate": 9.771794190677639e-06,
+      "loss": 0.1687,
+      "step": 7994
+    },
+    {
+      "epoch": 6.018065487391795,
+      "grad_norm": 0.19016380608081818,
+      "learning_rate": 9.76721339891576e-06,
+      "loss": 0.1857,
+      "step": 7995
+    },
+    {
+      "epoch": 6.01881821603312,
+      "grad_norm": 0.20830382406711578,
+      "learning_rate": 9.762634546384766e-06,
+      "loss": 0.2255,
+      "step": 7996
+    },
+    {
+      "epoch": 6.019570944674445,
+      "grad_norm": 0.19657564163208008,
+      "learning_rate": 9.758057633585388e-06,
+      "loss": 0.1614,
+      "step": 7997
+    },
+    {
+      "epoch": 6.02032367331577,
+      "grad_norm": 0.19113287329673767,
+      "learning_rate": 9.75348266101815e-06,
+      "loss": 0.1723,
+      "step": 7998
+    },
+    {
+      "epoch": 6.021076401957094,
+      "grad_norm": 0.16473917663097382,
+      "learning_rate": 9.74890962918337e-06,
+      "loss": 0.1903,
+      "step": 7999
+    },
+    {
+      "epoch": 6.0218291305984195,
+      "grad_norm": 0.1748979240655899,
+      "learning_rate": 9.744338538581147e-06,
+      "loss": 0.1416,
+      "step": 8000
+    },
+    {
+      "epoch": 6.0218291305984195,
+      "eval_loss": 0.20822034776210785,
+      "eval_runtime": 452.6448,
+      "eval_samples_per_second": 21.268,
+      "eval_steps_per_second": 0.665,
+      "step": 8000
+    },
+    {
+      "epoch": 6.022581859239744,
+      "grad_norm": 0.19145846366882324,
+      "learning_rate": 9.739769389711361e-06,
+      "loss": 0.2122,
+      "step": 8001
+    },
+    {
+      "epoch": 6.0233345878810685,
+      "grad_norm": 0.16384214162826538,
+      "learning_rate": 9.735202183073702e-06,
+      "loss": 0.1474,
+      "step": 8002
+    },
+    {
+      "epoch": 6.024087316522394,
+      "grad_norm": 0.19704356789588928,
+      "learning_rate": 9.730636919167613e-06,
+      "loss": 0.2655,
+      "step": 8003
+    },
+    {
+      "epoch": 6.024840045163718,
+      "grad_norm": 0.16561545431613922,
+      "learning_rate": 9.726073598492355e-06,
+      "loss": 0.1215,
+      "step": 8004
+    },
+    {
+      "epoch": 6.025592773805044,
+      "grad_norm": 0.17520521581172943,
+      "learning_rate": 9.721512221546966e-06,
+      "loss": 0.1626,
+      "step": 8005
+    },
+    {
+      "epoch": 6.026345502446368,
+      "grad_norm": 0.1945396214723587,
+      "learning_rate": 9.716952788830278e-06,
+      "loss": 0.1931,
+      "step": 8006
+    },
+    {
+      "epoch": 6.027098231087693,
+      "grad_norm": 0.18432922661304474,
+      "learning_rate": 9.71239530084089e-06,
+      "loss": 0.2102,
+      "step": 8007
+    },
+    {
+      "epoch": 6.027850959729018,
+      "grad_norm": 0.16351264715194702,
+      "learning_rate": 9.707839758077209e-06,
+      "loss": 0.1555,
+      "step": 8008
+    },
+    {
+      "epoch": 6.028603688370342,
+      "grad_norm": 0.19681642949581146,
+      "learning_rate": 9.703286161037427e-06,
+      "loss": 0.1029,
+      "step": 8009
+    },
+    {
+      "epoch": 6.029356417011667,
+      "grad_norm": 0.17093248665332794,
+      "learning_rate": 9.698734510219507e-06,
+      "loss": 0.141,
+      "step": 8010
+    },
+    {
+      "epoch": 6.030109145652992,
+      "grad_norm": 0.17641031742095947,
+      "learning_rate": 9.694184806121221e-06,
+      "loss": 0.13,
+      "step": 8011
+    },
+    {
+      "epoch": 6.030861874294317,
+      "grad_norm": 0.18621030449867249,
+      "learning_rate": 9.689637049240108e-06,
+      "loss": 0.1587,
+      "step": 8012
+    },
+    {
+      "epoch": 6.031614602935642,
+      "grad_norm": 0.16768483817577362,
+      "learning_rate": 9.685091240073518e-06,
+      "loss": 0.1638,
+      "step": 8013
+    },
+    {
+      "epoch": 6.032367331576967,
+      "grad_norm": 0.1894766241312027,
+      "learning_rate": 9.680547379118563e-06,
+      "loss": 0.1371,
+      "step": 8014
+    },
+    {
+      "epoch": 6.033120060218291,
+      "grad_norm": 0.2853589951992035,
+      "learning_rate": 9.676005466872159e-06,
+      "loss": 0.227,
+      "step": 8015
+    },
+    {
+      "epoch": 6.033872788859616,
+      "grad_norm": 0.1886426955461502,
+      "learning_rate": 9.671465503830992e-06,
+      "loss": 0.1385,
+      "step": 8016
+    },
+    {
+      "epoch": 6.034625517500941,
+      "grad_norm": 0.19343046844005585,
+      "learning_rate": 9.666927490491555e-06,
+      "loss": 0.1723,
+      "step": 8017
+    },
+    {
+      "epoch": 6.035378246142265,
+      "grad_norm": 0.2177373170852661,
+      "learning_rate": 9.662391427350116e-06,
+      "loss": 0.3057,
+      "step": 8018
+    },
+    {
+      "epoch": 6.036130974783591,
+      "grad_norm": 0.20888741314411163,
+      "learning_rate": 9.657857314902735e-06,
+      "loss": 0.1902,
+      "step": 8019
+    },
+    {
+      "epoch": 6.036883703424915,
+      "grad_norm": 0.1624508947134018,
+      "learning_rate": 9.653325153645249e-06,
+      "loss": 0.1312,
+      "step": 8020
+    },
+    {
+      "epoch": 6.03763643206624,
+      "grad_norm": 0.16994522511959076,
+      "learning_rate": 9.648794944073297e-06,
+      "loss": 0.1262,
+      "step": 8021
+    },
+    {
+      "epoch": 6.038389160707565,
+      "grad_norm": 0.22016307711601257,
+      "learning_rate": 9.644266686682289e-06,
+      "loss": 0.2258,
+      "step": 8022
+    },
+    {
+      "epoch": 6.0391418893488895,
+      "grad_norm": 0.18793299794197083,
+      "learning_rate": 9.639740381967425e-06,
+      "loss": 0.2393,
+      "step": 8023
+    },
+    {
+      "epoch": 6.039894617990215,
+      "grad_norm": 0.1736435890197754,
+      "learning_rate": 9.635216030423705e-06,
+      "loss": 0.187,
+      "step": 8024
+    },
+    {
+      "epoch": 6.040647346631539,
+      "grad_norm": 0.21770799160003662,
+      "learning_rate": 9.630693632545904e-06,
+      "loss": 0.1794,
+      "step": 8025
+    },
+    {
+      "epoch": 6.041400075272864,
+      "grad_norm": 0.19674880802631378,
+      "learning_rate": 9.626173188828574e-06,
+      "loss": 0.2109,
+      "step": 8026
+    },
+    {
+      "epoch": 6.042152803914189,
+      "grad_norm": 0.15429067611694336,
+      "learning_rate": 9.621654699766078e-06,
+      "loss": 0.1098,
+      "step": 8027
+    },
+    {
+      "epoch": 6.042905532555514,
+      "grad_norm": 0.17694710195064545,
+      "learning_rate": 9.617138165852541e-06,
+      "loss": 0.1882,
+      "step": 8028
+    },
+    {
+      "epoch": 6.043658261196838,
+      "grad_norm": 0.17477528750896454,
+      "learning_rate": 9.612623587581886e-06,
+      "loss": 0.1199,
+      "step": 8029
+    },
+    {
+      "epoch": 6.0444109898381635,
+      "grad_norm": 0.20671483874320984,
+      "learning_rate": 9.60811096544782e-06,
+      "loss": 0.3086,
+      "step": 8030
+    },
+    {
+      "epoch": 6.045163718479488,
+      "grad_norm": 0.14626653492450714,
+      "learning_rate": 9.603600299943849e-06,
+      "loss": 0.0745,
+      "step": 8031
+    },
+    {
+      "epoch": 6.045916447120813,
+      "grad_norm": 0.18177969753742218,
+      "learning_rate": 9.599091591563233e-06,
+      "loss": 0.1817,
+      "step": 8032
+    },
+    {
+      "epoch": 6.046669175762138,
+      "grad_norm": 0.14785057306289673,
+      "learning_rate": 9.594584840799048e-06,
+      "loss": 0.1168,
+      "step": 8033
+    },
+    {
+      "epoch": 6.047421904403462,
+      "grad_norm": 0.18687471747398376,
+      "learning_rate": 9.590080048144152e-06,
+      "loss": 0.1405,
+      "step": 8034
+    },
+    {
+      "epoch": 6.048174633044788,
+      "grad_norm": 0.1652068793773651,
+      "learning_rate": 9.585577214091166e-06,
+      "loss": 0.1494,
+      "step": 8035
+    },
+    {
+      "epoch": 6.048927361686112,
+      "grad_norm": 0.16085393726825714,
+      "learning_rate": 9.581076339132525e-06,
+      "loss": 0.1564,
+      "step": 8036
+    },
+    {
+      "epoch": 6.049680090327437,
+      "grad_norm": 0.1966649889945984,
+      "learning_rate": 9.576577423760432e-06,
+      "loss": 0.2524,
+      "step": 8037
+    },
+    {
+      "epoch": 6.050432818968762,
+      "grad_norm": 0.16176362335681915,
+      "learning_rate": 9.572080468466892e-06,
+      "loss": 0.1699,
+      "step": 8038
+    },
+    {
+      "epoch": 6.051185547610086,
+      "grad_norm": 0.1726902425289154,
+      "learning_rate": 9.567585473743671e-06,
+      "loss": 0.1714,
+      "step": 8039
+    },
+    {
+      "epoch": 6.051938276251412,
+      "grad_norm": 0.18412046134471893,
+      "learning_rate": 9.563092440082347e-06,
+      "loss": 0.1889,
+      "step": 8040
+    },
+    {
+      "epoch": 6.052691004892736,
+      "grad_norm": 0.1943395435810089,
+      "learning_rate": 9.55860136797426e-06,
+      "loss": 0.1452,
+      "step": 8041
+    },
+    {
+      "epoch": 6.053443733534061,
+      "grad_norm": 0.19566690921783447,
+      "learning_rate": 9.55411225791056e-06,
+      "loss": 0.2139,
+      "step": 8042
+    },
+    {
+      "epoch": 6.054196462175386,
+      "grad_norm": 0.190326526761055,
+      "learning_rate": 9.549625110382151e-06,
+      "loss": 0.1613,
+      "step": 8043
+    },
+    {
+      "epoch": 6.054949190816711,
+      "grad_norm": 0.18464258313179016,
+      "learning_rate": 9.545139925879762e-06,
+      "loss": 0.1331,
+      "step": 8044
+    },
+    {
+      "epoch": 6.055701919458035,
+      "grad_norm": 0.21453577280044556,
+      "learning_rate": 9.540656704893871e-06,
+      "loss": 0.1819,
+      "step": 8045
+    },
+    {
+      "epoch": 6.05645464809936,
+      "grad_norm": 0.1830378919839859,
+      "learning_rate": 9.536175447914767e-06,
+      "loss": 0.1575,
+      "step": 8046
+    },
+    {
+      "epoch": 6.057207376740685,
+      "grad_norm": 0.16455501317977905,
+      "learning_rate": 9.531696155432504e-06,
+      "loss": 0.2178,
+      "step": 8047
+    },
+    {
+      "epoch": 6.057960105382009,
+      "grad_norm": 0.20252548158168793,
+      "learning_rate": 9.527218827936937e-06,
+      "loss": 0.2629,
+      "step": 8048
+    },
+    {
+      "epoch": 6.058712834023335,
+      "grad_norm": 0.1910725086927414,
+      "learning_rate": 9.522743465917686e-06,
+      "loss": 0.1132,
+      "step": 8049
+    },
+    {
+      "epoch": 6.059465562664659,
+      "grad_norm": 0.18206043541431427,
+      "learning_rate": 9.518270069864195e-06,
+      "loss": 0.1924,
+      "step": 8050
+    },
+    {
+      "epoch": 6.0602182913059846,
+      "grad_norm": 0.18976543843746185,
+      "learning_rate": 9.513798640265647e-06,
+      "loss": 0.1457,
+      "step": 8051
+    },
+    {
+      "epoch": 6.060971019947309,
+      "grad_norm": 0.1776922345161438,
+      "learning_rate": 9.509329177611042e-06,
+      "loss": 0.1897,
+      "step": 8052
+    },
+    {
+      "epoch": 6.0617237485886335,
+      "grad_norm": 0.20805560052394867,
+      "learning_rate": 9.504861682389147e-06,
+      "loss": 0.1574,
+      "step": 8053
+    },
+    {
+      "epoch": 6.062476477229959,
+      "grad_norm": 0.18185588717460632,
+      "learning_rate": 9.500396155088523e-06,
+      "loss": 0.1182,
+      "step": 8054
+    },
+    {
+      "epoch": 6.063229205871283,
+      "grad_norm": 0.18178899586200714,
+      "learning_rate": 9.495932596197512e-06,
+      "loss": 0.2119,
+      "step": 8055
+    },
+    {
+      "epoch": 6.063981934512608,
+      "grad_norm": 0.1947861760854721,
+      "learning_rate": 9.49147100620425e-06,
+      "loss": 0.1852,
+      "step": 8056
+    },
+    {
+      "epoch": 6.064734663153933,
+      "grad_norm": 0.1842043697834015,
+      "learning_rate": 9.487011385596636e-06,
+      "loss": 0.1361,
+      "step": 8057
+    },
+    {
+      "epoch": 6.065487391795258,
+      "grad_norm": 0.1607372760772705,
+      "learning_rate": 9.482553734862374e-06,
+      "loss": 0.181,
+      "step": 8058
+    },
+    {
+      "epoch": 6.066240120436583,
+      "grad_norm": 0.18826182186603546,
+      "learning_rate": 9.478098054488956e-06,
+      "loss": 0.2014,
+      "step": 8059
+    },
+    {
+      "epoch": 6.0669928490779075,
+      "grad_norm": 0.17673251032829285,
+      "learning_rate": 9.473644344963631e-06,
+      "loss": 0.1831,
+      "step": 8060
+    },
+    {
+      "epoch": 6.067745577719232,
+      "grad_norm": 0.18977104127407074,
+      "learning_rate": 9.469192606773463e-06,
+      "loss": 0.1892,
+      "step": 8061
+    },
+    {
+      "epoch": 6.068498306360557,
+      "grad_norm": 0.17403729259967804,
+      "learning_rate": 9.464742840405275e-06,
+      "loss": 0.1502,
+      "step": 8062
+    },
+    {
+      "epoch": 6.069251035001882,
+      "grad_norm": 0.18036967515945435,
+      "learning_rate": 9.460295046345702e-06,
+      "loss": 0.1679,
+      "step": 8063
+    },
+    {
+      "epoch": 6.070003763643206,
+      "grad_norm": 0.17038851976394653,
+      "learning_rate": 9.45584922508114e-06,
+      "loss": 0.1237,
+      "step": 8064
+    },
+    {
+      "epoch": 6.070756492284532,
+      "grad_norm": 0.1819281280040741,
+      "learning_rate": 9.451405377097777e-06,
+      "loss": 0.1551,
+      "step": 8065
+    },
+    {
+      "epoch": 6.071509220925856,
+      "grad_norm": 0.19779758155345917,
+      "learning_rate": 9.446963502881584e-06,
+      "loss": 0.1796,
+      "step": 8066
+    },
+    {
+      "epoch": 6.0722619495671815,
+      "grad_norm": 0.19070486724376678,
+      "learning_rate": 9.442523602918326e-06,
+      "loss": 0.2321,
+      "step": 8067
+    },
+    {
+      "epoch": 6.073014678208506,
+      "grad_norm": 0.18509170413017273,
+      "learning_rate": 9.438085677693525e-06,
+      "loss": 0.1502,
+      "step": 8068
+    },
+    {
+      "epoch": 6.07376740684983,
+      "grad_norm": 0.191454216837883,
+      "learning_rate": 9.433649727692532e-06,
+      "loss": 0.2425,
+      "step": 8069
+    },
+    {
+      "epoch": 6.074520135491156,
+      "grad_norm": 0.17004330456256866,
+      "learning_rate": 9.429215753400435e-06,
+      "loss": 0.1201,
+      "step": 8070
+    },
+    {
+      "epoch": 6.07527286413248,
+      "grad_norm": 0.18114200234413147,
+      "learning_rate": 9.42478375530214e-06,
+      "loss": 0.1897,
+      "step": 8071
+    },
+    {
+      "epoch": 6.076025592773805,
+      "grad_norm": 0.17657962441444397,
+      "learning_rate": 9.420353733882312e-06,
+      "loss": 0.1953,
+      "step": 8072
+    },
+    {
+      "epoch": 6.07677832141513,
+      "grad_norm": 0.1691080927848816,
+      "learning_rate": 9.415925689625424e-06,
+      "loss": 0.1775,
+      "step": 8073
+    },
+    {
+      "epoch": 6.0775310500564546,
+      "grad_norm": 0.23190230131149292,
+      "learning_rate": 9.411499623015701e-06,
+      "loss": 0.1945,
+      "step": 8074
+    },
+    {
+      "epoch": 6.078283778697779,
+      "grad_norm": 0.17465533316135406,
+      "learning_rate": 9.407075534537196e-06,
+      "loss": 0.1732,
+      "step": 8075
+    },
+    {
+      "epoch": 6.079036507339104,
+      "grad_norm": 0.18154393136501312,
+      "learning_rate": 9.402653424673702e-06,
+      "loss": 0.1081,
+      "step": 8076
+    },
+    {
+      "epoch": 6.079789235980429,
+      "grad_norm": 0.17904694378376007,
+      "learning_rate": 9.398233293908826e-06,
+      "loss": 0.252,
+      "step": 8077
+    },
+    {
+      "epoch": 6.080541964621754,
+      "grad_norm": 0.19032540917396545,
+      "learning_rate": 9.393815142725933e-06,
+      "loss": 0.1548,
+      "step": 8078
+    },
+    {
+      "epoch": 6.081294693263079,
+      "grad_norm": 0.1697811633348465,
+      "learning_rate": 9.389398971608198e-06,
+      "loss": 0.1625,
+      "step": 8079
+    },
+    {
+      "epoch": 6.082047421904403,
+      "grad_norm": 0.1698836088180542,
+      "learning_rate": 9.384984781038566e-06,
+      "loss": 0.1737,
+      "step": 8080
+    },
+    {
+      "epoch": 6.0828001505457285,
+      "grad_norm": 0.18664512038230896,
+      "learning_rate": 9.380572571499758e-06,
+      "loss": 0.2186,
+      "step": 8081
+    },
+    {
+      "epoch": 6.083552879187053,
+      "grad_norm": 0.20953628420829773,
+      "learning_rate": 9.376162343474292e-06,
+      "loss": 0.237,
+      "step": 8082
+    },
+    {
+      "epoch": 6.0843056078283775,
+      "grad_norm": 0.17748279869556427,
+      "learning_rate": 9.371754097444462e-06,
+      "loss": 0.179,
+      "step": 8083
+    },
+    {
+      "epoch": 6.085058336469703,
+      "grad_norm": 0.17628929018974304,
+      "learning_rate": 9.367347833892357e-06,
+      "loss": 0.1522,
+      "step": 8084
+    },
+    {
+      "epoch": 6.085811065111027,
+      "grad_norm": 0.1868543177843094,
+      "learning_rate": 9.362943553299821e-06,
+      "loss": 0.1976,
+      "step": 8085
+    },
+    {
+      "epoch": 6.086563793752353,
+      "grad_norm": 0.15578536689281464,
+      "learning_rate": 9.35854125614852e-06,
+      "loss": 0.1771,
+      "step": 8086
+    },
+    {
+      "epoch": 6.087316522393677,
+      "grad_norm": 0.18666386604309082,
+      "learning_rate": 9.354140942919857e-06,
+      "loss": 0.1923,
+      "step": 8087
+    },
+    {
+      "epoch": 6.088069251035002,
+      "grad_norm": 0.1516411453485489,
+      "learning_rate": 9.34974261409507e-06,
+      "loss": 0.1617,
+      "step": 8088
+    },
+    {
+      "epoch": 6.088821979676327,
+      "grad_norm": 0.17541943490505219,
+      "learning_rate": 9.345346270155137e-06,
+      "loss": 0.1461,
+      "step": 8089
+    },
+    {
+      "epoch": 6.0895747083176515,
+      "grad_norm": 0.17478257417678833,
+      "learning_rate": 9.340951911580848e-06,
+      "loss": 0.1582,
+      "step": 8090
+    },
+    {
+      "epoch": 6.090327436958976,
+      "grad_norm": 0.17196664214134216,
+      "learning_rate": 9.33655953885275e-06,
+      "loss": 0.1549,
+      "step": 8091
+    },
+    {
+      "epoch": 6.091080165600301,
+      "grad_norm": 0.1894526332616806,
+      "learning_rate": 9.332169152451198e-06,
+      "loss": 0.212,
+      "step": 8092
+    },
+    {
+      "epoch": 6.091832894241626,
+      "grad_norm": 0.1775461882352829,
+      "learning_rate": 9.327780752856296e-06,
+      "loss": 0.1122,
+      "step": 8093
+    },
+    {
+      "epoch": 6.09258562288295,
+      "grad_norm": 0.20372147858142853,
+      "learning_rate": 9.323394340547987e-06,
+      "loss": 0.1755,
+      "step": 8094
+    },
+    {
+      "epoch": 6.093338351524276,
+      "grad_norm": 0.16729189455509186,
+      "learning_rate": 9.319009916005937e-06,
+      "loss": 0.1627,
+      "step": 8095
+    },
+    {
+      "epoch": 6.0940910801656,
+      "grad_norm": 0.18789279460906982,
+      "learning_rate": 9.314627479709633e-06,
+      "loss": 0.1346,
+      "step": 8096
+    },
+    {
+      "epoch": 6.0948438088069254,
+      "grad_norm": 0.18417994678020477,
+      "learning_rate": 9.310247032138325e-06,
+      "loss": 0.2297,
+      "step": 8097
+    },
+    {
+      "epoch": 6.09559653744825,
+      "grad_norm": 0.21273861825466156,
+      "learning_rate": 9.305868573771056e-06,
+      "loss": 0.1995,
+      "step": 8098
+    },
+    {
+      "epoch": 6.096349266089574,
+      "grad_norm": 0.1681799441576004,
+      "learning_rate": 9.301492105086636e-06,
+      "loss": 0.1757,
+      "step": 8099
+    },
+    {
+      "epoch": 6.0971019947309,
+      "grad_norm": 0.20802557468414307,
+      "learning_rate": 9.297117626563687e-06,
+      "loss": 0.1635,
+      "step": 8100
+    },
+    {
+      "epoch": 6.097854723372224,
+      "grad_norm": 0.16632549464702606,
+      "learning_rate": 9.292745138680577e-06,
+      "loss": 0.1815,
+      "step": 8101
+    },
+    {
+      "epoch": 6.098607452013549,
+      "grad_norm": 0.17857760190963745,
+      "learning_rate": 9.288374641915497e-06,
+      "loss": 0.1326,
+      "step": 8102
+    },
+    {
+      "epoch": 6.099360180654874,
+      "grad_norm": 0.16376669704914093,
+      "learning_rate": 9.284006136746379e-06,
+      "loss": 0.1455,
+      "step": 8103
+    },
+    {
+      "epoch": 6.1001129092961985,
+      "grad_norm": 0.21437641978263855,
+      "learning_rate": 9.27963962365096e-06,
+      "loss": 0.1573,
+      "step": 8104
+    },
+    {
+      "epoch": 6.100865637937524,
+      "grad_norm": 0.18423956632614136,
+      "learning_rate": 9.275275103106767e-06,
+      "loss": 0.141,
+      "step": 8105
+    },
+    {
+      "epoch": 6.101618366578848,
+      "grad_norm": 0.16899853944778442,
+      "learning_rate": 9.270912575591079e-06,
+      "loss": 0.201,
+      "step": 8106
+    },
+    {
+      "epoch": 6.102371095220173,
+      "grad_norm": 0.18983173370361328,
+      "learning_rate": 9.266552041580987e-06,
+      "loss": 0.1584,
+      "step": 8107
+    },
+    {
+      "epoch": 6.103123823861498,
+      "grad_norm": 0.17493240535259247,
+      "learning_rate": 9.262193501553348e-06,
+      "loss": 0.0922,
+      "step": 8108
+    },
+    {
+      "epoch": 6.103876552502823,
+      "grad_norm": 0.19684597849845886,
+      "learning_rate": 9.257836955984813e-06,
+      "loss": 0.2352,
+      "step": 8109
+    },
+    {
+      "epoch": 6.104629281144147,
+      "grad_norm": 0.18750151991844177,
+      "learning_rate": 9.2534824053518e-06,
+      "loss": 0.225,
+      "step": 8110
+    },
+    {
+      "epoch": 6.1053820097854725,
+      "grad_norm": 0.2001110315322876,
+      "learning_rate": 9.24912985013052e-06,
+      "loss": 0.1788,
+      "step": 8111
+    },
+    {
+      "epoch": 6.106134738426797,
+      "grad_norm": 0.18402306735515594,
+      "learning_rate": 9.244779290796957e-06,
+      "loss": 0.205,
+      "step": 8112
+    },
+    {
+      "epoch": 6.106887467068122,
+      "grad_norm": 0.18215444684028625,
+      "learning_rate": 9.240430727826881e-06,
+      "loss": 0.1499,
+      "step": 8113
+    },
+    {
+      "epoch": 6.107640195709447,
+      "grad_norm": 0.19409684836864471,
+      "learning_rate": 9.236084161695851e-06,
+      "loss": 0.2129,
+      "step": 8114
+    },
+    {
+      "epoch": 6.108392924350771,
+      "grad_norm": 0.16692250967025757,
+      "learning_rate": 9.231739592879202e-06,
+      "loss": 0.2219,
+      "step": 8115
+    },
+    {
+      "epoch": 6.109145652992097,
+      "grad_norm": 0.17731386423110962,
+      "learning_rate": 9.227397021852038e-06,
+      "loss": 0.1306,
+      "step": 8116
+    },
+    {
+      "epoch": 6.109898381633421,
+      "grad_norm": 0.1751507967710495,
+      "learning_rate": 9.22305644908927e-06,
+      "loss": 0.2049,
+      "step": 8117
+    },
+    {
+      "epoch": 6.110651110274746,
+      "grad_norm": 0.18417900800704956,
+      "learning_rate": 9.218717875065566e-06,
+      "loss": 0.1886,
+      "step": 8118
+    },
+    {
+      "epoch": 6.111403838916071,
+      "grad_norm": 0.21956154704093933,
+      "learning_rate": 9.214381300255386e-06,
+      "loss": 0.1529,
+      "step": 8119
+    },
+    {
+      "epoch": 6.1121565675573954,
+      "grad_norm": 0.16585150361061096,
+      "learning_rate": 9.210046725132979e-06,
+      "loss": 0.1112,
+      "step": 8120
+    },
+    {
+      "epoch": 6.112909296198721,
+      "grad_norm": 0.2083546221256256,
+      "learning_rate": 9.205714150172367e-06,
+      "loss": 0.1894,
+      "step": 8121
+    },
+    {
+      "epoch": 6.113662024840045,
+      "grad_norm": 0.17399096488952637,
+      "learning_rate": 9.201383575847345e-06,
+      "loss": 0.0838,
+      "step": 8122
+    },
+    {
+      "epoch": 6.11441475348137,
+      "grad_norm": 0.18066996335983276,
+      "learning_rate": 9.19705500263151e-06,
+      "loss": 0.2299,
+      "step": 8123
+    },
+    {
+      "epoch": 6.115167482122695,
+      "grad_norm": 0.1857077181339264,
+      "learning_rate": 9.192728430998215e-06,
+      "loss": 0.2033,
+      "step": 8124
+    },
+    {
+      "epoch": 6.11592021076402,
+      "grad_norm": 0.17373903095722198,
+      "learning_rate": 9.188403861420615e-06,
+      "loss": 0.1746,
+      "step": 8125
+    },
+    {
+      "epoch": 6.116672939405344,
+      "grad_norm": 0.1920178234577179,
+      "learning_rate": 9.184081294371637e-06,
+      "loss": 0.2869,
+      "step": 8126
+    },
+    {
+      "epoch": 6.117425668046669,
+      "grad_norm": 0.17097516357898712,
+      "learning_rate": 9.179760730323996e-06,
+      "loss": 0.1103,
+      "step": 8127
+    },
+    {
+      "epoch": 6.118178396687994,
+      "grad_norm": 0.18318307399749756,
+      "learning_rate": 9.175442169750175e-06,
+      "loss": 0.1892,
+      "step": 8128
+    },
+    {
+      "epoch": 6.118931125329318,
+      "grad_norm": 0.17875845730304718,
+      "learning_rate": 9.171125613122447e-06,
+      "loss": 0.1667,
+      "step": 8129
+    },
+    {
+      "epoch": 6.119683853970644,
+      "grad_norm": 0.14785873889923096,
+      "learning_rate": 9.166811060912869e-06,
+      "loss": 0.1339,
+      "step": 8130
+    },
+    {
+      "epoch": 6.120436582611968,
+      "grad_norm": 0.1887831836938858,
+      "learning_rate": 9.162498513593264e-06,
+      "loss": 0.1281,
+      "step": 8131
+    },
+    {
+      "epoch": 6.121189311253294,
+      "grad_norm": 0.18038348853588104,
+      "learning_rate": 9.158187971635252e-06,
+      "loss": 0.199,
+      "step": 8132
+    },
+    {
+      "epoch": 6.121942039894618,
+      "grad_norm": 0.18237677216529846,
+      "learning_rate": 9.153879435510226e-06,
+      "loss": 0.2449,
+      "step": 8133
+    },
+    {
+      "epoch": 6.1226947685359425,
+      "grad_norm": 0.18255986273288727,
+      "learning_rate": 9.14957290568937e-06,
+      "loss": 0.176,
+      "step": 8134
+    },
+    {
+      "epoch": 6.123447497177268,
+      "grad_norm": 0.18499545753002167,
+      "learning_rate": 9.145268382643623e-06,
+      "loss": 0.2086,
+      "step": 8135
+    },
+    {
+      "epoch": 6.124200225818592,
+      "grad_norm": 0.18507340550422668,
+      "learning_rate": 9.140965866843737e-06,
+      "loss": 0.1706,
+      "step": 8136
+    },
+    {
+      "epoch": 6.124952954459917,
+      "grad_norm": 0.19107083976268768,
+      "learning_rate": 9.136665358760216e-06,
+      "loss": 0.1895,
+      "step": 8137
+    },
+    {
+      "epoch": 6.125705683101242,
+      "grad_norm": 0.188368558883667,
+      "learning_rate": 9.132366858863361e-06,
+      "loss": 0.1703,
+      "step": 8138
+    },
+    {
+      "epoch": 6.126458411742567,
+      "grad_norm": 0.1685337871313095,
+      "learning_rate": 9.12807036762325e-06,
+      "loss": 0.2392,
+      "step": 8139
+    },
+    {
+      "epoch": 6.127211140383892,
+      "grad_norm": 0.16087697446346283,
+      "learning_rate": 9.123775885509753e-06,
+      "loss": 0.2113,
+      "step": 8140
+    },
+    {
+      "epoch": 6.1279638690252165,
+      "grad_norm": 0.19019921123981476,
+      "learning_rate": 9.119483412992489e-06,
+      "loss": 0.1591,
+      "step": 8141
+    },
+    {
+      "epoch": 6.128716597666541,
+      "grad_norm": 0.19705329835414886,
+      "learning_rate": 9.115192950540889e-06,
+      "loss": 0.1368,
+      "step": 8142
+    },
+    {
+      "epoch": 6.129469326307866,
+      "grad_norm": 0.1524621993303299,
+      "learning_rate": 9.110904498624139e-06,
+      "loss": 0.166,
+      "step": 8143
+    },
+    {
+      "epoch": 6.130222054949191,
+      "grad_norm": 0.19126318395137787,
+      "learning_rate": 9.106618057711227e-06,
+      "loss": 0.1932,
+      "step": 8144
+    },
+    {
+      "epoch": 6.130974783590515,
+      "grad_norm": 0.14446784555912018,
+      "learning_rate": 9.10233362827091e-06,
+      "loss": 0.1367,
+      "step": 8145
+    },
+    {
+      "epoch": 6.131727512231841,
+      "grad_norm": 0.20231714844703674,
+      "learning_rate": 9.098051210771736e-06,
+      "loss": 0.247,
+      "step": 8146
+    },
+    {
+      "epoch": 6.132480240873165,
+      "grad_norm": 0.1665763258934021,
+      "learning_rate": 9.093770805682004e-06,
+      "loss": 0.1858,
+      "step": 8147
+    },
+    {
+      "epoch": 6.13323296951449,
+      "grad_norm": 0.17053113877773285,
+      "learning_rate": 9.089492413469833e-06,
+      "loss": 0.1488,
+      "step": 8148
+    },
+    {
+      "epoch": 6.133985698155815,
+      "grad_norm": 0.19238099455833435,
+      "learning_rate": 9.085216034603084e-06,
+      "loss": 0.1248,
+      "step": 8149
+    },
+    {
+      "epoch": 6.134738426797139,
+      "grad_norm": 0.18364746868610382,
+      "learning_rate": 9.080941669549423e-06,
+      "loss": 0.2019,
+      "step": 8150
+    },
+    {
+      "epoch": 6.135491155438465,
+      "grad_norm": 0.15396836400032043,
+      "learning_rate": 9.076669318776287e-06,
+      "loss": 0.1522,
+      "step": 8151
+    },
+    {
+      "epoch": 6.136243884079789,
+      "grad_norm": 0.1627126783132553,
+      "learning_rate": 9.072398982750905e-06,
+      "loss": 0.0887,
+      "step": 8152
+    },
+    {
+      "epoch": 6.136996612721114,
+      "grad_norm": 0.17626017332077026,
+      "learning_rate": 9.068130661940255e-06,
+      "loss": 0.095,
+      "step": 8153
+    },
+    {
+      "epoch": 6.137749341362439,
+      "grad_norm": 0.21410880982875824,
+      "learning_rate": 9.063864356811124e-06,
+      "loss": 0.2228,
+      "step": 8154
+    },
+    {
+      "epoch": 6.138502070003764,
+      "grad_norm": 0.19214312732219696,
+      "learning_rate": 9.059600067830079e-06,
+      "loss": 0.1769,
+      "step": 8155
+    },
+    {
+      "epoch": 6.139254798645088,
+      "grad_norm": 0.15994037687778473,
+      "learning_rate": 9.055337795463436e-06,
+      "loss": 0.1228,
+      "step": 8156
+    },
+    {
+      "epoch": 6.140007527286413,
+      "grad_norm": 0.1904904544353485,
+      "learning_rate": 9.051077540177321e-06,
+      "loss": 0.188,
+      "step": 8157
+    },
+    {
+      "epoch": 6.140760255927738,
+      "grad_norm": 0.16106820106506348,
+      "learning_rate": 9.046819302437632e-06,
+      "loss": 0.0873,
+      "step": 8158
+    },
+    {
+      "epoch": 6.141512984569063,
+      "grad_norm": 0.2007487267255783,
+      "learning_rate": 9.042563082710042e-06,
+      "loss": 0.2573,
+      "step": 8159
+    },
+    {
+      "epoch": 6.142265713210388,
+      "grad_norm": 0.16535326838493347,
+      "learning_rate": 9.03830888146e-06,
+      "loss": 0.1466,
+      "step": 8160
+    },
+    {
+      "epoch": 6.143018441851712,
+      "grad_norm": 0.16428524255752563,
+      "learning_rate": 9.034056699152746e-06,
+      "loss": 0.1708,
+      "step": 8161
+    },
+    {
+      "epoch": 6.143771170493038,
+      "grad_norm": 0.17880640923976898,
+      "learning_rate": 9.029806536253282e-06,
+      "loss": 0.1831,
+      "step": 8162
+    },
+    {
+      "epoch": 6.144523899134362,
+      "grad_norm": 0.1752961277961731,
+      "learning_rate": 9.025558393226413e-06,
+      "loss": 0.1092,
+      "step": 8163
+    },
+    {
+      "epoch": 6.1452766277756865,
+      "grad_norm": 0.19280926883220673,
+      "learning_rate": 9.021312270536691e-06,
+      "loss": 0.1501,
+      "step": 8164
+    },
+    {
+      "epoch": 6.146029356417012,
+      "grad_norm": 0.19474811851978302,
+      "learning_rate": 9.017068168648488e-06,
+      "loss": 0.1656,
+      "step": 8165
+    },
+    {
+      "epoch": 6.146782085058336,
+      "grad_norm": 0.13894616067409515,
+      "learning_rate": 9.012826088025914e-06,
+      "loss": 0.1247,
+      "step": 8166
+    },
+    {
+      "epoch": 6.147534813699661,
+      "grad_norm": 0.17279168963432312,
+      "learning_rate": 9.00858602913289e-06,
+      "loss": 0.2277,
+      "step": 8167
+    },
+    {
+      "epoch": 6.148287542340986,
+      "grad_norm": 0.16572198271751404,
+      "learning_rate": 9.00434799243309e-06,
+      "loss": 0.1748,
+      "step": 8168
+    },
+    {
+      "epoch": 6.149040270982311,
+      "grad_norm": 0.18340839445590973,
+      "learning_rate": 9.00011197838999e-06,
+      "loss": 0.1518,
+      "step": 8169
+    },
+    {
+      "epoch": 6.149792999623636,
+      "grad_norm": 0.1876882165670395,
+      "learning_rate": 8.995877987466824e-06,
+      "loss": 0.1663,
+      "step": 8170
+    },
+    {
+      "epoch": 6.1505457282649605,
+      "grad_norm": 0.1713455617427826,
+      "learning_rate": 8.991646020126629e-06,
+      "loss": 0.2111,
+      "step": 8171
+    },
+    {
+      "epoch": 6.151298456906285,
+      "grad_norm": 0.16948948800563812,
+      "learning_rate": 8.987416076832195e-06,
+      "loss": 0.1125,
+      "step": 8172
+    },
+    {
+      "epoch": 6.15205118554761,
+      "grad_norm": 0.16769585013389587,
+      "learning_rate": 8.983188158046109e-06,
+      "loss": 0.1243,
+      "step": 8173
+    },
+    {
+      "epoch": 6.152803914188935,
+      "grad_norm": 0.18409746885299683,
+      "learning_rate": 8.97896226423072e-06,
+      "loss": 0.2097,
+      "step": 8174
+    },
+    {
+      "epoch": 6.153556642830259,
+      "grad_norm": 0.16856448352336884,
+      "learning_rate": 8.97473839584817e-06,
+      "loss": 0.1898,
+      "step": 8175
+    },
+    {
+      "epoch": 6.154309371471585,
+      "grad_norm": 0.1882065087556839,
+      "learning_rate": 8.970516553360384e-06,
+      "loss": 0.1802,
+      "step": 8176
+    },
+    {
+      "epoch": 6.155062100112909,
+      "grad_norm": 0.1830425262451172,
+      "learning_rate": 8.966296737229049e-06,
+      "loss": 0.1549,
+      "step": 8177
+    },
+    {
+      "epoch": 6.1558148287542345,
+      "grad_norm": 0.1687668263912201,
+      "learning_rate": 8.962078947915632e-06,
+      "loss": 0.1708,
+      "step": 8178
+    },
+    {
+      "epoch": 6.156567557395559,
+      "grad_norm": 0.17190349102020264,
+      "learning_rate": 8.957863185881392e-06,
+      "loss": 0.1254,
+      "step": 8179
+    },
+    {
+      "epoch": 6.157320286036883,
+      "grad_norm": 0.16737666726112366,
+      "learning_rate": 8.953649451587361e-06,
+      "loss": 0.1108,
+      "step": 8180
+    },
+    {
+      "epoch": 6.158073014678209,
+      "grad_norm": 0.194136843085289,
+      "learning_rate": 8.949437745494338e-06,
+      "loss": 0.169,
+      "step": 8181
+    },
+    {
+      "epoch": 6.158825743319533,
+      "grad_norm": 0.17167401313781738,
+      "learning_rate": 8.945228068062916e-06,
+      "loss": 0.1675,
+      "step": 8182
+    },
+    {
+      "epoch": 6.159578471960858,
+      "grad_norm": 0.1600593477487564,
+      "learning_rate": 8.941020419753446e-06,
+      "loss": 0.1261,
+      "step": 8183
+    },
+    {
+      "epoch": 6.160331200602183,
+      "grad_norm": 0.1698751449584961,
+      "learning_rate": 8.936814801026089e-06,
+      "loss": 0.1156,
+      "step": 8184
+    },
+    {
+      "epoch": 6.161083929243508,
+      "grad_norm": 0.1979142129421234,
+      "learning_rate": 8.932611212340753e-06,
+      "loss": 0.1546,
+      "step": 8185
+    },
+    {
+      "epoch": 6.161836657884833,
+      "grad_norm": 0.20740261673927307,
+      "learning_rate": 8.928409654157144e-06,
+      "loss": 0.1968,
+      "step": 8186
+    },
+    {
+      "epoch": 6.162589386526157,
+      "grad_norm": 0.1951761692762375,
+      "learning_rate": 8.924210126934729e-06,
+      "loss": 0.1415,
+      "step": 8187
+    },
+    {
+      "epoch": 6.163342115167482,
+      "grad_norm": 0.1609565019607544,
+      "learning_rate": 8.920012631132769e-06,
+      "loss": 0.1039,
+      "step": 8188
+    },
+    {
+      "epoch": 6.164094843808807,
+      "grad_norm": 0.17775088548660278,
+      "learning_rate": 8.915817167210282e-06,
+      "loss": 0.1639,
+      "step": 8189
+    },
+    {
+      "epoch": 6.164847572450132,
+      "grad_norm": 0.1737736016511917,
+      "learning_rate": 8.911623735626101e-06,
+      "loss": 0.1895,
+      "step": 8190
+    },
+    {
+      "epoch": 6.165600301091456,
+      "grad_norm": 0.17534637451171875,
+      "learning_rate": 8.907432336838796e-06,
+      "loss": 0.1937,
+      "step": 8191
+    },
+    {
+      "epoch": 6.1663530297327815,
+      "grad_norm": 0.16657082736492157,
+      "learning_rate": 8.903242971306744e-06,
+      "loss": 0.1701,
+      "step": 8192
+    },
+    {
+      "epoch": 6.167105758374106,
+      "grad_norm": 0.16405561566352844,
+      "learning_rate": 8.899055639488074e-06,
+      "loss": 0.1752,
+      "step": 8193
+    },
+    {
+      "epoch": 6.167858487015431,
+      "grad_norm": 0.15652546286582947,
+      "learning_rate": 8.894870341840717e-06,
+      "loss": 0.1455,
+      "step": 8194
+    },
+    {
+      "epoch": 6.168611215656756,
+      "grad_norm": 0.16826196014881134,
+      "learning_rate": 8.890687078822357e-06,
+      "loss": 0.1315,
+      "step": 8195
+    },
+    {
+      "epoch": 6.16936394429808,
+      "grad_norm": 0.1823587715625763,
+      "learning_rate": 8.886505850890492e-06,
+      "loss": 0.2273,
+      "step": 8196
+    },
+    {
+      "epoch": 6.170116672939406,
+      "grad_norm": 0.17534200847148895,
+      "learning_rate": 8.882326658502355e-06,
+      "loss": 0.1454,
+      "step": 8197
+    },
+    {
+      "epoch": 6.17086940158073,
+      "grad_norm": 0.17352332174777985,
+      "learning_rate": 8.87814950211499e-06,
+      "loss": 0.0935,
+      "step": 8198
+    },
+    {
+      "epoch": 6.171622130222055,
+      "grad_norm": 0.18694019317626953,
+      "learning_rate": 8.87397438218519e-06,
+      "loss": 0.1833,
+      "step": 8199
+    },
+    {
+      "epoch": 6.17237485886338,
+      "grad_norm": 0.1916595697402954,
+      "learning_rate": 8.86980129916955e-06,
+      "loss": 0.1813,
+      "step": 8200
+    },
+    {
+      "epoch": 6.17237485886338,
+      "eval_loss": 0.20622830092906952,
+      "eval_runtime": 452.8626,
+      "eval_samples_per_second": 21.258,
+      "eval_steps_per_second": 0.665,
+      "step": 8200
+    },
+    {
+      "epoch": 6.1731275875047045,
+      "grad_norm": 0.16931019723415375,
+      "learning_rate": 8.865630253524434e-06,
+      "loss": 0.1424,
+      "step": 8201
+    },
+    {
+      "epoch": 6.173880316146029,
+      "grad_norm": 0.1999819427728653,
+      "learning_rate": 8.861461245705972e-06,
+      "loss": 0.2062,
+      "step": 8202
+    },
+    {
+      "epoch": 6.174633044787354,
+      "grad_norm": 0.1702113151550293,
+      "learning_rate": 8.857294276170086e-06,
+      "loss": 0.1293,
+      "step": 8203
+    },
+    {
+      "epoch": 6.175385773428679,
+      "grad_norm": 0.1680496782064438,
+      "learning_rate": 8.853129345372471e-06,
+      "loss": 0.102,
+      "step": 8204
+    },
+    {
+      "epoch": 6.176138502070004,
+      "grad_norm": 0.17681917548179626,
+      "learning_rate": 8.848966453768597e-06,
+      "loss": 0.1453,
+      "step": 8205
+    },
+    {
+      "epoch": 6.176891230711329,
+      "grad_norm": 0.1751679629087448,
+      "learning_rate": 8.844805601813708e-06,
+      "loss": 0.1296,
+      "step": 8206
+    },
+    {
+      "epoch": 6.177643959352653,
+      "grad_norm": 0.17817988991737366,
+      "learning_rate": 8.840646789962836e-06,
+      "loss": 0.1775,
+      "step": 8207
+    },
+    {
+      "epoch": 6.1783966879939785,
+      "grad_norm": 0.2135886698961258,
+      "learning_rate": 8.836490018670765e-06,
+      "loss": 0.1894,
+      "step": 8208
+    },
+    {
+      "epoch": 6.179149416635303,
+      "grad_norm": 0.1861853003501892,
+      "learning_rate": 8.832335288392097e-06,
+      "loss": 0.1167,
+      "step": 8209
+    },
+    {
+      "epoch": 6.179902145276627,
+      "grad_norm": 0.18908721208572388,
+      "learning_rate": 8.828182599581166e-06,
+      "loss": 0.1632,
+      "step": 8210
+    },
+    {
+      "epoch": 6.180654873917953,
+      "grad_norm": 0.1608656942844391,
+      "learning_rate": 8.824031952692124e-06,
+      "loss": 0.1201,
+      "step": 8211
+    },
+    {
+      "epoch": 6.181407602559277,
+      "grad_norm": 0.18054483830928802,
+      "learning_rate": 8.819883348178861e-06,
+      "loss": 0.2085,
+      "step": 8212
+    },
+    {
+      "epoch": 6.182160331200603,
+      "grad_norm": 0.19761915504932404,
+      "learning_rate": 8.815736786495077e-06,
+      "loss": 0.1596,
+      "step": 8213
+    },
+    {
+      "epoch": 6.182913059841927,
+      "grad_norm": 0.21195538341999054,
+      "learning_rate": 8.811592268094212e-06,
+      "loss": 0.1282,
+      "step": 8214
+    },
+    {
+      "epoch": 6.1836657884832515,
+      "grad_norm": 0.2024705559015274,
+      "learning_rate": 8.807449793429535e-06,
+      "loss": 0.1397,
+      "step": 8215
+    },
+    {
+      "epoch": 6.184418517124577,
+      "grad_norm": 0.16609954833984375,
+      "learning_rate": 8.803309362954037e-06,
+      "loss": 0.181,
+      "step": 8216
+    },
+    {
+      "epoch": 6.185171245765901,
+      "grad_norm": 0.19882358610630035,
+      "learning_rate": 8.79917097712052e-06,
+      "loss": 0.1256,
+      "step": 8217
+    },
+    {
+      "epoch": 6.185923974407226,
+      "grad_norm": 0.1682499498128891,
+      "learning_rate": 8.795034636381545e-06,
+      "loss": 0.2073,
+      "step": 8218
+    },
+    {
+      "epoch": 6.186676703048551,
+      "grad_norm": 0.16802291572093964,
+      "learning_rate": 8.790900341189468e-06,
+      "loss": 0.1751,
+      "step": 8219
+    },
+    {
+      "epoch": 6.187429431689876,
+      "grad_norm": 0.1808565855026245,
+      "learning_rate": 8.786768091996392e-06,
+      "loss": 0.1069,
+      "step": 8220
+    },
+    {
+      "epoch": 6.1881821603312,
+      "grad_norm": 0.15788234770298004,
+      "learning_rate": 8.782637889254222e-06,
+      "loss": 0.1188,
+      "step": 8221
+    },
+    {
+      "epoch": 6.1889348889725255,
+      "grad_norm": 0.19225667417049408,
+      "learning_rate": 8.778509733414633e-06,
+      "loss": 0.1644,
+      "step": 8222
+    },
+    {
+      "epoch": 6.18968761761385,
+      "grad_norm": 0.1764637976884842,
+      "learning_rate": 8.774383624929078e-06,
+      "loss": 0.1856,
+      "step": 8223
+    },
+    {
+      "epoch": 6.190440346255175,
+      "grad_norm": 0.18168048560619354,
+      "learning_rate": 8.77025956424877e-06,
+      "loss": 0.173,
+      "step": 8224
+    },
+    {
+      "epoch": 6.1911930748965,
+      "grad_norm": 0.1637091338634491,
+      "learning_rate": 8.766137551824718e-06,
+      "loss": 0.202,
+      "step": 8225
+    },
+    {
+      "epoch": 6.191945803537824,
+      "grad_norm": 0.1635071337223053,
+      "learning_rate": 8.762017588107701e-06,
+      "loss": 0.115,
+      "step": 8226
+    },
+    {
+      "epoch": 6.19269853217915,
+      "grad_norm": 0.17936953902244568,
+      "learning_rate": 8.757899673548264e-06,
+      "loss": 0.1486,
+      "step": 8227
+    },
+    {
+      "epoch": 6.193451260820474,
+      "grad_norm": 0.20122657716274261,
+      "learning_rate": 8.753783808596738e-06,
+      "loss": 0.2099,
+      "step": 8228
+    },
+    {
+      "epoch": 6.194203989461799,
+      "grad_norm": 0.15658749639987946,
+      "learning_rate": 8.749669993703233e-06,
+      "loss": 0.0934,
+      "step": 8229
+    },
+    {
+      "epoch": 6.194956718103124,
+      "grad_norm": 0.17835558950901031,
+      "learning_rate": 8.745558229317633e-06,
+      "loss": 0.2077,
+      "step": 8230
+    },
+    {
+      "epoch": 6.1957094467444485,
+      "grad_norm": 0.17351368069648743,
+      "learning_rate": 8.741448515889581e-06,
+      "loss": 0.2113,
+      "step": 8231
+    },
+    {
+      "epoch": 6.196462175385774,
+      "grad_norm": 0.17999562621116638,
+      "learning_rate": 8.737340853868523e-06,
+      "loss": 0.204,
+      "step": 8232
+    },
+    {
+      "epoch": 6.197214904027098,
+      "grad_norm": 0.19909431040287018,
+      "learning_rate": 8.73323524370365e-06,
+      "loss": 0.1149,
+      "step": 8233
+    },
+    {
+      "epoch": 6.197967632668423,
+      "grad_norm": 0.2104235738515854,
+      "learning_rate": 8.729131685843959e-06,
+      "loss": 0.2181,
+      "step": 8234
+    },
+    {
+      "epoch": 6.198720361309748,
+      "grad_norm": 0.16277152299880981,
+      "learning_rate": 8.725030180738205e-06,
+      "loss": 0.1813,
+      "step": 8235
+    },
+    {
+      "epoch": 6.199473089951073,
+      "grad_norm": 0.18786026537418365,
+      "learning_rate": 8.720930728834927e-06,
+      "loss": 0.1758,
+      "step": 8236
+    },
+    {
+      "epoch": 6.200225818592397,
+      "grad_norm": 0.17235133051872253,
+      "learning_rate": 8.716833330582427e-06,
+      "loss": 0.1574,
+      "step": 8237
+    },
+    {
+      "epoch": 6.200978547233722,
+      "grad_norm": 0.18282479047775269,
+      "learning_rate": 8.712737986428799e-06,
+      "loss": 0.1876,
+      "step": 8238
+    },
+    {
+      "epoch": 6.201731275875047,
+      "grad_norm": 0.2149437516927719,
+      "learning_rate": 8.708644696821892e-06,
+      "loss": 0.231,
+      "step": 8239
+    },
+    {
+      "epoch": 6.202484004516372,
+      "grad_norm": 0.1706322282552719,
+      "learning_rate": 8.704553462209348e-06,
+      "loss": 0.1827,
+      "step": 8240
+    },
+    {
+      "epoch": 6.203236733157697,
+      "grad_norm": 0.20701424777507782,
+      "learning_rate": 8.70046428303858e-06,
+      "loss": 0.137,
+      "step": 8241
+    },
+    {
+      "epoch": 6.203989461799021,
+      "grad_norm": 0.16502174735069275,
+      "learning_rate": 8.696377159756781e-06,
+      "loss": 0.1177,
+      "step": 8242
+    },
+    {
+      "epoch": 6.204742190440347,
+      "grad_norm": 0.1855711191892624,
+      "learning_rate": 8.692292092810895e-06,
+      "loss": 0.1428,
+      "step": 8243
+    },
+    {
+      "epoch": 6.205494919081671,
+      "grad_norm": 0.16833782196044922,
+      "learning_rate": 8.688209082647678e-06,
+      "loss": 0.1488,
+      "step": 8244
+    },
+    {
+      "epoch": 6.2062476477229955,
+      "grad_norm": 0.19679318368434906,
+      "learning_rate": 8.684128129713626e-06,
+      "loss": 0.1904,
+      "step": 8245
+    },
+    {
+      "epoch": 6.207000376364321,
+      "grad_norm": 0.18364745378494263,
+      "learning_rate": 8.680049234455032e-06,
+      "loss": 0.1801,
+      "step": 8246
+    },
+    {
+      "epoch": 6.207753105005645,
+      "grad_norm": 0.20243950188159943,
+      "learning_rate": 8.675972397317958e-06,
+      "loss": 0.1797,
+      "step": 8247
+    },
+    {
+      "epoch": 6.20850583364697,
+      "grad_norm": 0.18364572525024414,
+      "learning_rate": 8.671897618748248e-06,
+      "loss": 0.1342,
+      "step": 8248
+    },
+    {
+      "epoch": 6.209258562288295,
+      "grad_norm": 0.1885259449481964,
+      "learning_rate": 8.6678248991915e-06,
+      "loss": 0.1464,
+      "step": 8249
+    },
+    {
+      "epoch": 6.21001129092962,
+      "grad_norm": 0.19598069787025452,
+      "learning_rate": 8.663754239093109e-06,
+      "loss": 0.1493,
+      "step": 8250
+    },
+    {
+      "epoch": 6.210764019570945,
+      "grad_norm": 0.1831820011138916,
+      "learning_rate": 8.659685638898238e-06,
+      "loss": 0.1483,
+      "step": 8251
+    },
+    {
+      "epoch": 6.2115167482122695,
+      "grad_norm": 0.1665833592414856,
+      "learning_rate": 8.655619099051814e-06,
+      "loss": 0.2511,
+      "step": 8252
+    },
+    {
+      "epoch": 6.212269476853594,
+      "grad_norm": 0.19293437898159027,
+      "learning_rate": 8.651554619998554e-06,
+      "loss": 0.2456,
+      "step": 8253
+    },
+    {
+      "epoch": 6.213022205494919,
+      "grad_norm": 0.1728106439113617,
+      "learning_rate": 8.647492202182944e-06,
+      "loss": 0.2288,
+      "step": 8254
+    },
+    {
+      "epoch": 6.213774934136244,
+      "grad_norm": 0.1879809945821762,
+      "learning_rate": 8.643431846049243e-06,
+      "loss": 0.1718,
+      "step": 8255
+    },
+    {
+      "epoch": 6.214527662777568,
+      "grad_norm": 0.19561293721199036,
+      "learning_rate": 8.639373552041484e-06,
+      "loss": 0.1726,
+      "step": 8256
+    },
+    {
+      "epoch": 6.215280391418894,
+      "grad_norm": 0.16849465668201447,
+      "learning_rate": 8.635317320603482e-06,
+      "loss": 0.1027,
+      "step": 8257
+    },
+    {
+      "epoch": 6.216033120060218,
+      "grad_norm": 0.1860169768333435,
+      "learning_rate": 8.631263152178807e-06,
+      "loss": 0.198,
+      "step": 8258
+    },
+    {
+      "epoch": 6.2167858487015435,
+      "grad_norm": 0.2049635499715805,
+      "learning_rate": 8.62721104721083e-06,
+      "loss": 0.1975,
+      "step": 8259
+    },
+    {
+      "epoch": 6.217538577342868,
+      "grad_norm": 0.18473117053508759,
+      "learning_rate": 8.623161006142673e-06,
+      "loss": 0.1219,
+      "step": 8260
+    },
+    {
+      "epoch": 6.218291305984192,
+      "grad_norm": 0.18477237224578857,
+      "learning_rate": 8.619113029417253e-06,
+      "loss": 0.1711,
+      "step": 8261
+    },
+    {
+      "epoch": 6.219044034625518,
+      "grad_norm": 0.19090253114700317,
+      "learning_rate": 8.615067117477244e-06,
+      "loss": 0.2684,
+      "step": 8262
+    },
+    {
+      "epoch": 6.219796763266842,
+      "grad_norm": 0.17323997616767883,
+      "learning_rate": 8.611023270765104e-06,
+      "loss": 0.1833,
+      "step": 8263
+    },
+    {
+      "epoch": 6.220549491908167,
+      "grad_norm": 0.17057974636554718,
+      "learning_rate": 8.606981489723056e-06,
+      "loss": 0.2435,
+      "step": 8264
+    },
+    {
+      "epoch": 6.221302220549492,
+      "grad_norm": 0.200887069106102,
+      "learning_rate": 8.602941774793104e-06,
+      "loss": 0.152,
+      "step": 8265
+    },
+    {
+      "epoch": 6.222054949190817,
+      "grad_norm": 0.16619016230106354,
+      "learning_rate": 8.598904126417032e-06,
+      "loss": 0.1279,
+      "step": 8266
+    },
+    {
+      "epoch": 6.222807677832142,
+      "grad_norm": 0.18003292381763458,
+      "learning_rate": 8.594868545036391e-06,
+      "loss": 0.1594,
+      "step": 8267
+    },
+    {
+      "epoch": 6.223560406473466,
+      "grad_norm": 0.1819816529750824,
+      "learning_rate": 8.590835031092497e-06,
+      "loss": 0.1725,
+      "step": 8268
+    },
+    {
+      "epoch": 6.224313135114791,
+      "grad_norm": 0.18542467057704926,
+      "learning_rate": 8.586803585026462e-06,
+      "loss": 0.1707,
+      "step": 8269
+    },
+    {
+      "epoch": 6.225065863756116,
+      "grad_norm": 0.18174245953559875,
+      "learning_rate": 8.582774207279146e-06,
+      "loss": 0.1063,
+      "step": 8270
+    },
+    {
+      "epoch": 6.225818592397441,
+      "grad_norm": 0.18153725564479828,
+      "learning_rate": 8.578746898291199e-06,
+      "loss": 0.2424,
+      "step": 8271
+    },
+    {
+      "epoch": 6.226571321038765,
+      "grad_norm": 0.17021295428276062,
+      "learning_rate": 8.574721658503043e-06,
+      "loss": 0.15,
+      "step": 8272
+    },
+    {
+      "epoch": 6.227324049680091,
+      "grad_norm": 0.19427622854709625,
+      "learning_rate": 8.570698488354878e-06,
+      "loss": 0.1616,
+      "step": 8273
+    },
+    {
+      "epoch": 6.228076778321415,
+      "grad_norm": 0.1562936007976532,
+      "learning_rate": 8.566677388286662e-06,
+      "loss": 0.1619,
+      "step": 8274
+    },
+    {
+      "epoch": 6.2288295069627395,
+      "grad_norm": 0.14487159252166748,
+      "learning_rate": 8.56265835873814e-06,
+      "loss": 0.1,
+      "step": 8275
+    },
+    {
+      "epoch": 6.229582235604065,
+      "grad_norm": 0.18048569560050964,
+      "learning_rate": 8.558641400148837e-06,
+      "loss": 0.1184,
+      "step": 8276
+    },
+    {
+      "epoch": 6.230334964245389,
+      "grad_norm": 0.174120232462883,
+      "learning_rate": 8.554626512958022e-06,
+      "loss": 0.2086,
+      "step": 8277
+    },
+    {
+      "epoch": 6.231087692886715,
+      "grad_norm": 0.19222186505794525,
+      "learning_rate": 8.550613697604773e-06,
+      "loss": 0.1242,
+      "step": 8278
+    },
+    {
+      "epoch": 6.231840421528039,
+      "grad_norm": 0.19073256850242615,
+      "learning_rate": 8.546602954527916e-06,
+      "loss": 0.1894,
+      "step": 8279
+    },
+    {
+      "epoch": 6.232593150169364,
+      "grad_norm": 0.16831351816654205,
+      "learning_rate": 8.542594284166073e-06,
+      "loss": 0.1802,
+      "step": 8280
+    },
+    {
+      "epoch": 6.233345878810689,
+      "grad_norm": 0.18355382978916168,
+      "learning_rate": 8.538587686957607e-06,
+      "loss": 0.2144,
+      "step": 8281
+    },
+    {
+      "epoch": 6.2340986074520135,
+      "grad_norm": 0.18520641326904297,
+      "learning_rate": 8.534583163340691e-06,
+      "loss": 0.1619,
+      "step": 8282
+    },
+    {
+      "epoch": 6.234851336093338,
+      "grad_norm": 0.19489094614982605,
+      "learning_rate": 8.530580713753241e-06,
+      "loss": 0.1319,
+      "step": 8283
+    },
+    {
+      "epoch": 6.235604064734663,
+      "grad_norm": 0.17416292428970337,
+      "learning_rate": 8.52658033863297e-06,
+      "loss": 0.1364,
+      "step": 8284
+    },
+    {
+      "epoch": 6.236356793375988,
+      "grad_norm": 0.16726407408714294,
+      "learning_rate": 8.522582038417335e-06,
+      "loss": 0.1855,
+      "step": 8285
+    },
+    {
+      "epoch": 6.237109522017313,
+      "grad_norm": 0.17440450191497803,
+      "learning_rate": 8.518585813543608e-06,
+      "loss": 0.1774,
+      "step": 8286
+    },
+    {
+      "epoch": 6.237862250658638,
+      "grad_norm": 0.17890550196170807,
+      "learning_rate": 8.514591664448792e-06,
+      "loss": 0.1716,
+      "step": 8287
+    },
+    {
+      "epoch": 6.238614979299962,
+      "grad_norm": 0.17547354102134705,
+      "learning_rate": 8.510599591569693e-06,
+      "loss": 0.1529,
+      "step": 8288
+    },
+    {
+      "epoch": 6.2393677079412875,
+      "grad_norm": 0.16584394872188568,
+      "learning_rate": 8.506609595342869e-06,
+      "loss": 0.1136,
+      "step": 8289
+    },
+    {
+      "epoch": 6.240120436582612,
+      "grad_norm": 0.17216435074806213,
+      "learning_rate": 8.502621676204667e-06,
+      "loss": 0.13,
+      "step": 8290
+    },
+    {
+      "epoch": 6.240873165223936,
+      "grad_norm": 0.19211052358150482,
+      "learning_rate": 8.498635834591186e-06,
+      "loss": 0.1961,
+      "step": 8291
+    },
+    {
+      "epoch": 6.241625893865262,
+      "grad_norm": 0.18279273808002472,
+      "learning_rate": 8.494652070938332e-06,
+      "loss": 0.2605,
+      "step": 8292
+    },
+    {
+      "epoch": 6.242378622506586,
+      "grad_norm": 0.20758919417858124,
+      "learning_rate": 8.49067038568175e-06,
+      "loss": 0.2529,
+      "step": 8293
+    },
+    {
+      "epoch": 6.243131351147911,
+      "grad_norm": 0.17579753696918488,
+      "learning_rate": 8.486690779256876e-06,
+      "loss": 0.1184,
+      "step": 8294
+    },
+    {
+      "epoch": 6.243884079789236,
+      "grad_norm": 0.17390353977680206,
+      "learning_rate": 8.482713252098911e-06,
+      "loss": 0.1176,
+      "step": 8295
+    },
+    {
+      "epoch": 6.244636808430561,
+      "grad_norm": 0.18450134992599487,
+      "learning_rate": 8.478737804642833e-06,
+      "loss": 0.2684,
+      "step": 8296
+    },
+    {
+      "epoch": 6.245389537071886,
+      "grad_norm": 0.16701096296310425,
+      "learning_rate": 8.474764437323387e-06,
+      "loss": 0.1694,
+      "step": 8297
+    },
+    {
+      "epoch": 6.24614226571321,
+      "grad_norm": 0.1752966195344925,
+      "learning_rate": 8.470793150575107e-06,
+      "loss": 0.1479,
+      "step": 8298
+    },
+    {
+      "epoch": 6.246894994354535,
+      "grad_norm": 0.15781909227371216,
+      "learning_rate": 8.466823944832269e-06,
+      "loss": 0.0898,
+      "step": 8299
+    },
+    {
+      "epoch": 6.24764772299586,
+      "grad_norm": 0.1777845025062561,
+      "learning_rate": 8.462856820528952e-06,
+      "loss": 0.1543,
+      "step": 8300
+    },
+    {
+      "epoch": 6.248400451637185,
+      "grad_norm": 0.15982989966869354,
+      "learning_rate": 8.458891778098995e-06,
+      "loss": 0.1432,
+      "step": 8301
+    },
+    {
+      "epoch": 6.249153180278509,
+      "grad_norm": 0.19306838512420654,
+      "learning_rate": 8.454928817976001e-06,
+      "loss": 0.1766,
+      "step": 8302
+    },
+    {
+      "epoch": 6.2499059089198346,
+      "grad_norm": 0.18523460626602173,
+      "learning_rate": 8.45096794059336e-06,
+      "loss": 0.2062,
+      "step": 8303
+    },
+    {
+      "epoch": 6.250658637561159,
+      "grad_norm": 0.17531496286392212,
+      "learning_rate": 8.447009146384219e-06,
+      "loss": 0.1314,
+      "step": 8304
+    },
+    {
+      "epoch": 6.251411366202484,
+      "grad_norm": 0.21394570171833038,
+      "learning_rate": 8.443052435781518e-06,
+      "loss": 0.15,
+      "step": 8305
+    },
+    {
+      "epoch": 6.252164094843809,
+      "grad_norm": 0.18119440972805023,
+      "learning_rate": 8.439097809217948e-06,
+      "loss": 0.1865,
+      "step": 8306
+    },
+    {
+      "epoch": 6.252916823485133,
+      "grad_norm": 0.18342623114585876,
+      "learning_rate": 8.43514526712599e-06,
+      "loss": 0.1681,
+      "step": 8307
+    },
+    {
+      "epoch": 6.253669552126459,
+      "grad_norm": 0.17653155326843262,
+      "learning_rate": 8.431194809937876e-06,
+      "loss": 0.1837,
+      "step": 8308
+    },
+    {
+      "epoch": 6.254422280767783,
+      "grad_norm": 0.18157656490802765,
+      "learning_rate": 8.427246438085632e-06,
+      "loss": 0.1932,
+      "step": 8309
+    },
+    {
+      "epoch": 6.255175009409108,
+      "grad_norm": 0.16765938699245453,
+      "learning_rate": 8.423300152001034e-06,
+      "loss": 0.1478,
+      "step": 8310
+    },
+    {
+      "epoch": 6.255927738050433,
+      "grad_norm": 0.18541841208934784,
+      "learning_rate": 8.41935595211566e-06,
+      "loss": 0.2076,
+      "step": 8311
+    },
+    {
+      "epoch": 6.2566804666917575,
+      "grad_norm": 0.17122448980808258,
+      "learning_rate": 8.415413838860824e-06,
+      "loss": 0.1282,
+      "step": 8312
+    },
+    {
+      "epoch": 6.257433195333082,
+      "grad_norm": 0.17053304612636566,
+      "learning_rate": 8.411473812667645e-06,
+      "loss": 0.1329,
+      "step": 8313
+    },
+    {
+      "epoch": 6.258185923974407,
+      "grad_norm": 0.1836603581905365,
+      "learning_rate": 8.407535873966985e-06,
+      "loss": 0.196,
+      "step": 8314
+    },
+    {
+      "epoch": 6.258938652615732,
+      "grad_norm": 0.17495186626911163,
+      "learning_rate": 8.4036000231895e-06,
+      "loss": 0.1923,
+      "step": 8315
+    },
+    {
+      "epoch": 6.259691381257057,
+      "grad_norm": 0.21120966970920563,
+      "learning_rate": 8.399666260765597e-06,
+      "loss": 0.2007,
+      "step": 8316
+    },
+    {
+      "epoch": 6.260444109898382,
+      "grad_norm": 0.19297459721565247,
+      "learning_rate": 8.395734587125486e-06,
+      "loss": 0.1556,
+      "step": 8317
+    },
+    {
+      "epoch": 6.261196838539706,
+      "grad_norm": 0.1779804676771164,
+      "learning_rate": 8.391805002699111e-06,
+      "loss": 0.1455,
+      "step": 8318
+    },
+    {
+      "epoch": 6.2619495671810315,
+      "grad_norm": 0.17874805629253387,
+      "learning_rate": 8.387877507916217e-06,
+      "loss": 0.1031,
+      "step": 8319
+    },
+    {
+      "epoch": 6.262702295822356,
+      "grad_norm": 0.24663633108139038,
+      "learning_rate": 8.3839521032063e-06,
+      "loss": 0.1775,
+      "step": 8320
+    },
+    {
+      "epoch": 6.263455024463681,
+      "grad_norm": 0.18513420224189758,
+      "learning_rate": 8.38002878899864e-06,
+      "loss": 0.1848,
+      "step": 8321
+    },
+    {
+      "epoch": 6.264207753105006,
+      "grad_norm": 0.18967792391777039,
+      "learning_rate": 8.376107565722289e-06,
+      "loss": 0.1586,
+      "step": 8322
+    },
+    {
+      "epoch": 6.26496048174633,
+      "grad_norm": 0.19964370131492615,
+      "learning_rate": 8.37218843380606e-06,
+      "loss": 0.1483,
+      "step": 8323
+    },
+    {
+      "epoch": 6.265713210387656,
+      "grad_norm": 0.18157264590263367,
+      "learning_rate": 8.368271393678545e-06,
+      "loss": 0.268,
+      "step": 8324
+    },
+    {
+      "epoch": 6.26646593902898,
+      "grad_norm": 0.20310643315315247,
+      "learning_rate": 8.364356445768107e-06,
+      "loss": 0.1331,
+      "step": 8325
+    },
+    {
+      "epoch": 6.2672186676703046,
+      "grad_norm": 0.19900761544704437,
+      "learning_rate": 8.360443590502881e-06,
+      "loss": 0.1893,
+      "step": 8326
+    },
+    {
+      "epoch": 6.26797139631163,
+      "grad_norm": 0.1743140071630478,
+      "learning_rate": 8.356532828310762e-06,
+      "loss": 0.1566,
+      "step": 8327
+    },
+    {
+      "epoch": 6.268724124952954,
+      "grad_norm": 0.20371128618717194,
+      "learning_rate": 8.352624159619438e-06,
+      "loss": 0.1691,
+      "step": 8328
+    },
+    {
+      "epoch": 6.269476853594279,
+      "grad_norm": 0.1840018332004547,
+      "learning_rate": 8.348717584856338e-06,
+      "loss": 0.1379,
+      "step": 8329
+    },
+    {
+      "epoch": 6.270229582235604,
+      "grad_norm": 0.15606488287448883,
+      "learning_rate": 8.344813104448699e-06,
+      "loss": 0.1349,
+      "step": 8330
+    },
+    {
+      "epoch": 6.270982310876929,
+      "grad_norm": 0.20933394134044647,
+      "learning_rate": 8.340910718823495e-06,
+      "loss": 0.1448,
+      "step": 8331
+    },
+    {
+      "epoch": 6.271735039518254,
+      "grad_norm": 0.1656845211982727,
+      "learning_rate": 8.337010428407495e-06,
+      "loss": 0.1759,
+      "step": 8332
+    },
+    {
+      "epoch": 6.2724877681595785,
+      "grad_norm": 0.18392090499401093,
+      "learning_rate": 8.333112233627219e-06,
+      "loss": 0.1579,
+      "step": 8333
+    },
+    {
+      "epoch": 6.273240496800903,
+      "grad_norm": 0.18964503705501556,
+      "learning_rate": 8.329216134908976e-06,
+      "loss": 0.19,
+      "step": 8334
+    },
+    {
+      "epoch": 6.273993225442228,
+      "grad_norm": 0.17418460547924042,
+      "learning_rate": 8.325322132678822e-06,
+      "loss": 0.1265,
+      "step": 8335
+    },
+    {
+      "epoch": 6.274745954083553,
+      "grad_norm": 0.17685584723949432,
+      "learning_rate": 8.321430227362623e-06,
+      "loss": 0.1833,
+      "step": 8336
+    },
+    {
+      "epoch": 6.275498682724877,
+      "grad_norm": 0.15614213049411774,
+      "learning_rate": 8.317540419385974e-06,
+      "loss": 0.1105,
+      "step": 8337
+    },
+    {
+      "epoch": 6.276251411366203,
+      "grad_norm": 0.1617499440908432,
+      "learning_rate": 8.31365270917427e-06,
+      "loss": 0.1705,
+      "step": 8338
+    },
+    {
+      "epoch": 6.277004140007527,
+      "grad_norm": 0.15588928759098053,
+      "learning_rate": 8.309767097152654e-06,
+      "loss": 0.1511,
+      "step": 8339
+    },
+    {
+      "epoch": 6.2777568686488525,
+      "grad_norm": 0.17297203838825226,
+      "learning_rate": 8.305883583746062e-06,
+      "loss": 0.2284,
+      "step": 8340
+    },
+    {
+      "epoch": 6.278509597290177,
+      "grad_norm": 0.16162562370300293,
+      "learning_rate": 8.302002169379176e-06,
+      "loss": 0.1655,
+      "step": 8341
+    },
+    {
+      "epoch": 6.2792623259315015,
+      "grad_norm": 0.20813219249248505,
+      "learning_rate": 8.29812285447647e-06,
+      "loss": 0.2137,
+      "step": 8342
+    },
+    {
+      "epoch": 6.280015054572827,
+      "grad_norm": 0.15603242814540863,
+      "learning_rate": 8.294245639462181e-06,
+      "loss": 0.1391,
+      "step": 8343
+    },
+    {
+      "epoch": 6.280767783214151,
+      "grad_norm": 0.17748527228832245,
+      "learning_rate": 8.290370524760318e-06,
+      "loss": 0.1721,
+      "step": 8344
+    },
+    {
+      "epoch": 6.281520511855476,
+      "grad_norm": 0.17343859374523163,
+      "learning_rate": 8.286497510794646e-06,
+      "loss": 0.1761,
+      "step": 8345
+    },
+    {
+      "epoch": 6.282273240496801,
+      "grad_norm": 0.1699882596731186,
+      "learning_rate": 8.282626597988722e-06,
+      "loss": 0.1827,
+      "step": 8346
+    },
+    {
+      "epoch": 6.283025969138126,
+      "grad_norm": 0.18912270665168762,
+      "learning_rate": 8.278757786765862e-06,
+      "loss": 0.194,
+      "step": 8347
+    },
+    {
+      "epoch": 6.28377869777945,
+      "grad_norm": 0.1784655600786209,
+      "learning_rate": 8.274891077549149e-06,
+      "loss": 0.1238,
+      "step": 8348
+    },
+    {
+      "epoch": 6.2845314264207754,
+      "grad_norm": 0.17393985390663147,
+      "learning_rate": 8.271026470761443e-06,
+      "loss": 0.1368,
+      "step": 8349
+    },
+    {
+      "epoch": 6.2852841550621,
+      "grad_norm": 0.15448884665966034,
+      "learning_rate": 8.26716396682537e-06,
+      "loss": 0.0937,
+      "step": 8350
+    },
+    {
+      "epoch": 6.286036883703425,
+      "grad_norm": 0.17277225852012634,
+      "learning_rate": 8.263303566163338e-06,
+      "loss": 0.1634,
+      "step": 8351
+    },
+    {
+      "epoch": 6.28678961234475,
+      "grad_norm": 0.1831797957420349,
+      "learning_rate": 8.2594452691975e-06,
+      "loss": 0.1274,
+      "step": 8352
+    },
+    {
+      "epoch": 6.287542340986074,
+      "grad_norm": 0.18674252927303314,
+      "learning_rate": 8.255589076349801e-06,
+      "loss": 0.1419,
+      "step": 8353
+    },
+    {
+      "epoch": 6.2882950696274,
+      "grad_norm": 0.18252016603946686,
+      "learning_rate": 8.251734988041946e-06,
+      "loss": 0.2087,
+      "step": 8354
+    },
+    {
+      "epoch": 6.289047798268724,
+      "grad_norm": 0.1812998503446579,
+      "learning_rate": 8.247883004695412e-06,
+      "loss": 0.1436,
+      "step": 8355
+    },
+    {
+      "epoch": 6.2898005269100485,
+      "grad_norm": 0.17751604318618774,
+      "learning_rate": 8.244033126731448e-06,
+      "loss": 0.1527,
+      "step": 8356
+    },
+    {
+      "epoch": 6.290553255551374,
+      "grad_norm": 0.16428625583648682,
+      "learning_rate": 8.240185354571075e-06,
+      "loss": 0.1346,
+      "step": 8357
+    },
+    {
+      "epoch": 6.291305984192698,
+      "grad_norm": 0.16908839344978333,
+      "learning_rate": 8.23633968863507e-06,
+      "loss": 0.2068,
+      "step": 8358
+    },
+    {
+      "epoch": 6.292058712834024,
+      "grad_norm": 0.17517396807670593,
+      "learning_rate": 8.232496129343998e-06,
+      "loss": 0.147,
+      "step": 8359
+    },
+    {
+      "epoch": 6.292811441475348,
+      "grad_norm": 0.15948808193206787,
+      "learning_rate": 8.228654677118181e-06,
+      "loss": 0.1459,
+      "step": 8360
+    },
+    {
+      "epoch": 6.293564170116673,
+      "grad_norm": 0.1817011833190918,
+      "learning_rate": 8.224815332377708e-06,
+      "loss": 0.1964,
+      "step": 8361
+    },
+    {
+      "epoch": 6.294316898757998,
+      "grad_norm": 0.16078144311904907,
+      "learning_rate": 8.220978095542453e-06,
+      "loss": 0.0938,
+      "step": 8362
+    },
+    {
+      "epoch": 6.2950696273993225,
+      "grad_norm": 0.19102366268634796,
+      "learning_rate": 8.217142967032054e-06,
+      "loss": 0.1498,
+      "step": 8363
+    },
+    {
+      "epoch": 6.295822356040647,
+      "grad_norm": 0.1692999005317688,
+      "learning_rate": 8.213309947265901e-06,
+      "loss": 0.0879,
+      "step": 8364
+    },
+    {
+      "epoch": 6.296575084681972,
+      "grad_norm": 0.17854507267475128,
+      "learning_rate": 8.209479036663182e-06,
+      "loss": 0.1675,
+      "step": 8365
+    },
+    {
+      "epoch": 6.297327813323297,
+      "grad_norm": 0.19736282527446747,
+      "learning_rate": 8.205650235642828e-06,
+      "loss": 0.168,
+      "step": 8366
+    },
+    {
+      "epoch": 6.298080541964621,
+      "grad_norm": 0.15524345636367798,
+      "learning_rate": 8.201823544623555e-06,
+      "loss": 0.1881,
+      "step": 8367
+    },
+    {
+      "epoch": 6.298833270605947,
+      "grad_norm": 0.19123029708862305,
+      "learning_rate": 8.197998964023843e-06,
+      "loss": 0.2665,
+      "step": 8368
+    },
+    {
+      "epoch": 6.299585999247271,
+      "grad_norm": 0.15626052021980286,
+      "learning_rate": 8.19417649426195e-06,
+      "loss": 0.1416,
+      "step": 8369
+    },
+    {
+      "epoch": 6.3003387278885965,
+      "grad_norm": 0.1818406581878662,
+      "learning_rate": 8.190356135755888e-06,
+      "loss": 0.1535,
+      "step": 8370
+    },
+    {
+      "epoch": 6.301091456529921,
+      "grad_norm": 0.1713232398033142,
+      "learning_rate": 8.186537888923446e-06,
+      "loss": 0.1687,
+      "step": 8371
+    },
+    {
+      "epoch": 6.3018441851712454,
+      "grad_norm": 0.19696590304374695,
+      "learning_rate": 8.182721754182186e-06,
+      "loss": 0.2494,
+      "step": 8372
+    },
+    {
+      "epoch": 6.302596913812571,
+      "grad_norm": 0.16366411745548248,
+      "learning_rate": 8.178907731949428e-06,
+      "loss": 0.1065,
+      "step": 8373
+    },
+    {
+      "epoch": 6.303349642453895,
+      "grad_norm": 0.177986741065979,
+      "learning_rate": 8.175095822642273e-06,
+      "loss": 0.1309,
+      "step": 8374
+    },
+    {
+      "epoch": 6.30410237109522,
+      "grad_norm": 0.14510932564735413,
+      "learning_rate": 8.171286026677583e-06,
+      "loss": 0.1037,
+      "step": 8375
+    },
+    {
+      "epoch": 6.304855099736545,
+      "grad_norm": 0.17907193303108215,
+      "learning_rate": 8.167478344471997e-06,
+      "loss": 0.1481,
+      "step": 8376
+    },
+    {
+      "epoch": 6.30560782837787,
+      "grad_norm": 0.18078473210334778,
+      "learning_rate": 8.163672776441909e-06,
+      "loss": 0.1866,
+      "step": 8377
+    },
+    {
+      "epoch": 6.306360557019195,
+      "grad_norm": 0.15999643504619598,
+      "learning_rate": 8.1598693230035e-06,
+      "loss": 0.2105,
+      "step": 8378
+    },
+    {
+      "epoch": 6.307113285660519,
+      "grad_norm": 0.1681225597858429,
+      "learning_rate": 8.1560679845727e-06,
+      "loss": 0.1243,
+      "step": 8379
+    },
+    {
+      "epoch": 6.307866014301844,
+      "grad_norm": 0.15121707320213318,
+      "learning_rate": 8.15226876156522e-06,
+      "loss": 0.1873,
+      "step": 8380
+    },
+    {
+      "epoch": 6.308618742943169,
+      "grad_norm": 0.18031731247901917,
+      "learning_rate": 8.14847165439654e-06,
+      "loss": 0.1332,
+      "step": 8381
+    },
+    {
+      "epoch": 6.309371471584494,
+      "grad_norm": 0.16747494041919708,
+      "learning_rate": 8.14467666348191e-06,
+      "loss": 0.1237,
+      "step": 8382
+    },
+    {
+      "epoch": 6.310124200225818,
+      "grad_norm": 0.19191157817840576,
+      "learning_rate": 8.140883789236336e-06,
+      "loss": 0.2559,
+      "step": 8383
+    },
+    {
+      "epoch": 6.310876928867144,
+      "grad_norm": 0.1869773119688034,
+      "learning_rate": 8.137093032074611e-06,
+      "loss": 0.2495,
+      "step": 8384
+    },
+    {
+      "epoch": 6.311629657508468,
+      "grad_norm": 0.19098249077796936,
+      "learning_rate": 8.133304392411272e-06,
+      "loss": 0.2013,
+      "step": 8385
+    },
+    {
+      "epoch": 6.312382386149793,
+      "grad_norm": 0.1750202625989914,
+      "learning_rate": 8.129517870660649e-06,
+      "loss": 0.1379,
+      "step": 8386
+    },
+    {
+      "epoch": 6.313135114791118,
+      "grad_norm": 0.17088843882083893,
+      "learning_rate": 8.125733467236826e-06,
+      "loss": 0.1488,
+      "step": 8387
+    },
+    {
+      "epoch": 6.313887843432442,
+      "grad_norm": 0.1737898737192154,
+      "learning_rate": 8.12195118255367e-06,
+      "loss": 0.1291,
+      "step": 8388
+    },
+    {
+      "epoch": 6.314640572073768,
+      "grad_norm": 0.16908079385757446,
+      "learning_rate": 8.118171017024792e-06,
+      "loss": 0.1511,
+      "step": 8389
+    },
+    {
+      "epoch": 6.315393300715092,
+      "grad_norm": 0.15231353044509888,
+      "learning_rate": 8.1143929710636e-06,
+      "loss": 0.1007,
+      "step": 8390
+    },
+    {
+      "epoch": 6.316146029356417,
+      "grad_norm": 0.15911628305912018,
+      "learning_rate": 8.110617045083239e-06,
+      "loss": 0.1584,
+      "step": 8391
+    },
+    {
+      "epoch": 6.316898757997742,
+      "grad_norm": 0.20299968123435974,
+      "learning_rate": 8.106843239496646e-06,
+      "loss": 0.2927,
+      "step": 8392
+    },
+    {
+      "epoch": 6.3176514866390665,
+      "grad_norm": 0.18170106410980225,
+      "learning_rate": 8.103071554716518e-06,
+      "loss": 0.1598,
+      "step": 8393
+    },
+    {
+      "epoch": 6.318404215280392,
+      "grad_norm": 0.139839768409729,
+      "learning_rate": 8.099301991155329e-06,
+      "loss": 0.1328,
+      "step": 8394
+    },
+    {
+      "epoch": 6.319156943921716,
+      "grad_norm": 0.2111634910106659,
+      "learning_rate": 8.0955345492253e-06,
+      "loss": 0.1566,
+      "step": 8395
+    },
+    {
+      "epoch": 6.319909672563041,
+      "grad_norm": 0.1732516884803772,
+      "learning_rate": 8.091769229338439e-06,
+      "loss": 0.1395,
+      "step": 8396
+    },
+    {
+      "epoch": 6.320662401204366,
+      "grad_norm": 0.19130656123161316,
+      "learning_rate": 8.08800603190652e-06,
+      "loss": 0.1122,
+      "step": 8397
+    },
+    {
+      "epoch": 6.321415129845691,
+      "grad_norm": 0.16873537003993988,
+      "learning_rate": 8.084244957341072e-06,
+      "loss": 0.0981,
+      "step": 8398
+    },
+    {
+      "epoch": 6.322167858487015,
+      "grad_norm": 0.21891151368618011,
+      "learning_rate": 8.080486006053402e-06,
+      "loss": 0.2468,
+      "step": 8399
+    },
+    {
+      "epoch": 6.3229205871283405,
+      "grad_norm": 0.16989995539188385,
+      "learning_rate": 8.076729178454588e-06,
+      "loss": 0.1836,
+      "step": 8400
+    },
+    {
+      "epoch": 6.3229205871283405,
+      "eval_loss": 0.20531049370765686,
+      "eval_runtime": 452.6234,
+      "eval_samples_per_second": 21.269,
+      "eval_steps_per_second": 0.665,
+      "step": 8400
+    },
+    {
+      "epoch": 6.323673315769665,
+      "grad_norm": 0.17933228611946106,
+      "learning_rate": 8.072974474955471e-06,
+      "loss": 0.2886,
+      "step": 8401
+    },
+    {
+      "epoch": 6.324426044410989,
+      "grad_norm": 0.18163010478019714,
+      "learning_rate": 8.069221895966654e-06,
+      "loss": 0.1767,
+      "step": 8402
+    },
+    {
+      "epoch": 6.325178773052315,
+      "grad_norm": 0.15721170604228973,
+      "learning_rate": 8.065471441898522e-06,
+      "loss": 0.1225,
+      "step": 8403
+    },
+    {
+      "epoch": 6.325931501693639,
+      "grad_norm": 0.20781125128269196,
+      "learning_rate": 8.061723113161207e-06,
+      "loss": 0.1839,
+      "step": 8404
+    },
+    {
+      "epoch": 6.326684230334965,
+      "grad_norm": 0.17221632599830627,
+      "learning_rate": 8.05797691016463e-06,
+      "loss": 0.1771,
+      "step": 8405
+    },
+    {
+      "epoch": 6.327436958976289,
+      "grad_norm": 0.1962944120168686,
+      "learning_rate": 8.054232833318465e-06,
+      "loss": 0.2059,
+      "step": 8406
+    },
+    {
+      "epoch": 6.328189687617614,
+      "grad_norm": 0.1888764500617981,
+      "learning_rate": 8.050490883032167e-06,
+      "loss": 0.2187,
+      "step": 8407
+    },
+    {
+      "epoch": 6.328942416258939,
+      "grad_norm": 0.18382887542247772,
+      "learning_rate": 8.04675105971494e-06,
+      "loss": 0.1212,
+      "step": 8408
+    },
+    {
+      "epoch": 6.329695144900263,
+      "grad_norm": 0.15681298077106476,
+      "learning_rate": 8.043013363775773e-06,
+      "loss": 0.1828,
+      "step": 8409
+    },
+    {
+      "epoch": 6.330447873541588,
+      "grad_norm": 0.16491185128688812,
+      "learning_rate": 8.039277795623407e-06,
+      "loss": 0.0876,
+      "step": 8410
+    },
+    {
+      "epoch": 6.331200602182913,
+      "grad_norm": 0.1725318729877472,
+      "learning_rate": 8.035544355666368e-06,
+      "loss": 0.1018,
+      "step": 8411
+    },
+    {
+      "epoch": 6.331953330824238,
+      "grad_norm": 0.17123712599277496,
+      "learning_rate": 8.031813044312924e-06,
+      "loss": 0.1367,
+      "step": 8412
+    },
+    {
+      "epoch": 6.332706059465563,
+      "grad_norm": 0.18252457678318024,
+      "learning_rate": 8.028083861971143e-06,
+      "loss": 0.1845,
+      "step": 8413
+    },
+    {
+      "epoch": 6.333458788106888,
+      "grad_norm": 0.16270312666893005,
+      "learning_rate": 8.024356809048834e-06,
+      "loss": 0.1368,
+      "step": 8414
+    },
+    {
+      "epoch": 6.334211516748212,
+      "grad_norm": 0.19126848876476288,
+      "learning_rate": 8.020631885953585e-06,
+      "loss": 0.1446,
+      "step": 8415
+    },
+    {
+      "epoch": 6.334964245389537,
+      "grad_norm": 0.18350987136363983,
+      "learning_rate": 8.016909093092742e-06,
+      "loss": 0.1071,
+      "step": 8416
+    },
+    {
+      "epoch": 6.335716974030862,
+      "grad_norm": 0.1887713521718979,
+      "learning_rate": 8.013188430873431e-06,
+      "loss": 0.1904,
+      "step": 8417
+    },
+    {
+      "epoch": 6.336469702672186,
+      "grad_norm": 0.18009096384048462,
+      "learning_rate": 8.009469899702534e-06,
+      "loss": 0.1315,
+      "step": 8418
+    },
+    {
+      "epoch": 6.337222431313512,
+      "grad_norm": 0.18358375132083893,
+      "learning_rate": 8.005753499986708e-06,
+      "loss": 0.2207,
+      "step": 8419
+    },
+    {
+      "epoch": 6.337975159954836,
+      "grad_norm": 0.1610904335975647,
+      "learning_rate": 8.002039232132366e-06,
+      "loss": 0.1039,
+      "step": 8420
+    },
+    {
+      "epoch": 6.338727888596161,
+      "grad_norm": 0.19457438588142395,
+      "learning_rate": 7.9983270965457e-06,
+      "loss": 0.1941,
+      "step": 8421
+    },
+    {
+      "epoch": 6.339480617237486,
+      "grad_norm": 0.18529623746871948,
+      "learning_rate": 7.994617093632665e-06,
+      "loss": 0.2079,
+      "step": 8422
+    },
+    {
+      "epoch": 6.3402333458788105,
+      "grad_norm": 0.16555973887443542,
+      "learning_rate": 7.990909223798976e-06,
+      "loss": 0.1285,
+      "step": 8423
+    },
+    {
+      "epoch": 6.340986074520136,
+      "grad_norm": 0.16674676537513733,
+      "learning_rate": 7.987203487450128e-06,
+      "loss": 0.1513,
+      "step": 8424
+    },
+    {
+      "epoch": 6.34173880316146,
+      "grad_norm": 0.15555663406848907,
+      "learning_rate": 7.983499884991356e-06,
+      "loss": 0.1339,
+      "step": 8425
+    },
+    {
+      "epoch": 6.342491531802785,
+      "grad_norm": 0.18561524152755737,
+      "learning_rate": 7.979798416827706e-06,
+      "loss": 0.134,
+      "step": 8426
+    },
+    {
+      "epoch": 6.34324426044411,
+      "grad_norm": 0.19297271966934204,
+      "learning_rate": 7.976099083363951e-06,
+      "loss": 0.1074,
+      "step": 8427
+    },
+    {
+      "epoch": 6.343996989085435,
+      "grad_norm": 0.18590715527534485,
+      "learning_rate": 7.97240188500465e-06,
+      "loss": 0.1707,
+      "step": 8428
+    },
+    {
+      "epoch": 6.344749717726759,
+      "grad_norm": 0.17817316949367523,
+      "learning_rate": 7.968706822154113e-06,
+      "loss": 0.1822,
+      "step": 8429
+    },
+    {
+      "epoch": 6.3455024463680845,
+      "grad_norm": 0.19365261495113373,
+      "learning_rate": 7.965013895216438e-06,
+      "loss": 0.2765,
+      "step": 8430
+    },
+    {
+      "epoch": 6.346255175009409,
+      "grad_norm": 0.17401786148548126,
+      "learning_rate": 7.961323104595464e-06,
+      "loss": 0.1797,
+      "step": 8431
+    },
+    {
+      "epoch": 6.347007903650734,
+      "grad_norm": 0.1500379741191864,
+      "learning_rate": 7.957634450694826e-06,
+      "loss": 0.1078,
+      "step": 8432
+    },
+    {
+      "epoch": 6.347760632292059,
+      "grad_norm": 0.17923873662948608,
+      "learning_rate": 7.953947933917903e-06,
+      "loss": 0.1466,
+      "step": 8433
+    },
+    {
+      "epoch": 6.348513360933383,
+      "grad_norm": 0.17712581157684326,
+      "learning_rate": 7.95026355466785e-06,
+      "loss": 0.1363,
+      "step": 8434
+    },
+    {
+      "epoch": 6.349266089574709,
+      "grad_norm": 0.16465461254119873,
+      "learning_rate": 7.946581313347578e-06,
+      "loss": 0.1909,
+      "step": 8435
+    },
+    {
+      "epoch": 6.350018818216033,
+      "grad_norm": 0.16587485373020172,
+      "learning_rate": 7.942901210359778e-06,
+      "loss": 0.1414,
+      "step": 8436
+    },
+    {
+      "epoch": 6.350771546857358,
+      "grad_norm": 0.18029062449932098,
+      "learning_rate": 7.93922324610689e-06,
+      "loss": 0.1551,
+      "step": 8437
+    },
+    {
+      "epoch": 6.351524275498683,
+      "grad_norm": 0.16934549808502197,
+      "learning_rate": 7.935547420991146e-06,
+      "loss": 0.1497,
+      "step": 8438
+    },
+    {
+      "epoch": 6.352277004140007,
+      "grad_norm": 0.1675984114408493,
+      "learning_rate": 7.931873735414519e-06,
+      "loss": 0.198,
+      "step": 8439
+    },
+    {
+      "epoch": 6.353029732781332,
+      "grad_norm": 0.19445966184139252,
+      "learning_rate": 7.928202189778765e-06,
+      "loss": 0.1748,
+      "step": 8440
+    },
+    {
+      "epoch": 6.353782461422657,
+      "grad_norm": 0.184829980134964,
+      "learning_rate": 7.924532784485386e-06,
+      "loss": 0.1047,
+      "step": 8441
+    },
+    {
+      "epoch": 6.354535190063982,
+      "grad_norm": 0.14518684148788452,
+      "learning_rate": 7.920865519935674e-06,
+      "loss": 0.0647,
+      "step": 8442
+    },
+    {
+      "epoch": 6.355287918705307,
+      "grad_norm": 0.17017053067684174,
+      "learning_rate": 7.917200396530676e-06,
+      "loss": 0.1178,
+      "step": 8443
+    },
+    {
+      "epoch": 6.3560406473466315,
+      "grad_norm": 0.17254269123077393,
+      "learning_rate": 7.913537414671194e-06,
+      "loss": 0.1399,
+      "step": 8444
+    },
+    {
+      "epoch": 6.356793375987956,
+      "grad_norm": 0.1540118306875229,
+      "learning_rate": 7.909876574757813e-06,
+      "loss": 0.1511,
+      "step": 8445
+    },
+    {
+      "epoch": 6.357546104629281,
+      "grad_norm": 0.19120727479457855,
+      "learning_rate": 7.906217877190878e-06,
+      "loss": 0.1627,
+      "step": 8446
+    },
+    {
+      "epoch": 6.358298833270606,
+      "grad_norm": 0.1815195232629776,
+      "learning_rate": 7.9025613223705e-06,
+      "loss": 0.1849,
+      "step": 8447
+    },
+    {
+      "epoch": 6.359051561911931,
+      "grad_norm": 0.1839710772037506,
+      "learning_rate": 7.898906910696548e-06,
+      "loss": 0.2086,
+      "step": 8448
+    },
+    {
+      "epoch": 6.359804290553256,
+      "grad_norm": 0.20567484200000763,
+      "learning_rate": 7.89525464256867e-06,
+      "loss": 0.2152,
+      "step": 8449
+    },
+    {
+      "epoch": 6.36055701919458,
+      "grad_norm": 0.17452479898929596,
+      "learning_rate": 7.89160451838626e-06,
+      "loss": 0.1675,
+      "step": 8450
+    },
+    {
+      "epoch": 6.3613097478359055,
+      "grad_norm": 0.1758532077074051,
+      "learning_rate": 7.887956538548507e-06,
+      "loss": 0.1745,
+      "step": 8451
+    },
+    {
+      "epoch": 6.36206247647723,
+      "grad_norm": 0.18853746354579926,
+      "learning_rate": 7.88431070345434e-06,
+      "loss": 0.2177,
+      "step": 8452
+    },
+    {
+      "epoch": 6.3628152051185545,
+      "grad_norm": 0.19956322014331818,
+      "learning_rate": 7.880667013502464e-06,
+      "loss": 0.18,
+      "step": 8453
+    },
+    {
+      "epoch": 6.36356793375988,
+      "grad_norm": 0.2117021232843399,
+      "learning_rate": 7.877025469091343e-06,
+      "loss": 0.2178,
+      "step": 8454
+    },
+    {
+      "epoch": 6.364320662401204,
+      "grad_norm": 0.20129582285881042,
+      "learning_rate": 7.873386070619219e-06,
+      "loss": 0.161,
+      "step": 8455
+    },
+    {
+      "epoch": 6.365073391042529,
+      "grad_norm": 0.188773512840271,
+      "learning_rate": 7.869748818484074e-06,
+      "loss": 0.1724,
+      "step": 8456
+    },
+    {
+      "epoch": 6.365826119683854,
+      "grad_norm": 0.2006678581237793,
+      "learning_rate": 7.866113713083695e-06,
+      "loss": 0.1896,
+      "step": 8457
+    },
+    {
+      "epoch": 6.366578848325179,
+      "grad_norm": 0.18287383019924164,
+      "learning_rate": 7.8624807548156e-06,
+      "loss": 0.1267,
+      "step": 8458
+    },
+    {
+      "epoch": 6.367331576966504,
+      "grad_norm": 0.1715794950723648,
+      "learning_rate": 7.858849944077086e-06,
+      "loss": 0.1675,
+      "step": 8459
+    },
+    {
+      "epoch": 6.3680843056078285,
+      "grad_norm": 0.17195767164230347,
+      "learning_rate": 7.855221281265208e-06,
+      "loss": 0.1238,
+      "step": 8460
+    },
+    {
+      "epoch": 6.368837034249153,
+      "grad_norm": 0.17715343832969666,
+      "learning_rate": 7.851594766776802e-06,
+      "loss": 0.2857,
+      "step": 8461
+    },
+    {
+      "epoch": 6.369589762890478,
+      "grad_norm": 0.2008422464132309,
+      "learning_rate": 7.847970401008446e-06,
+      "loss": 0.1836,
+      "step": 8462
+    },
+    {
+      "epoch": 6.370342491531803,
+      "grad_norm": 0.2122603952884674,
+      "learning_rate": 7.844348184356499e-06,
+      "loss": 0.2409,
+      "step": 8463
+    },
+    {
+      "epoch": 6.371095220173127,
+      "grad_norm": 0.18951067328453064,
+      "learning_rate": 7.840728117217084e-06,
+      "loss": 0.2545,
+      "step": 8464
+    },
+    {
+      "epoch": 6.371847948814453,
+      "grad_norm": 0.16863837838172913,
+      "learning_rate": 7.837110199986087e-06,
+      "loss": 0.1959,
+      "step": 8465
+    },
+    {
+      "epoch": 6.372600677455777,
+      "grad_norm": 0.16581740975379944,
+      "learning_rate": 7.833494433059154e-06,
+      "loss": 0.0843,
+      "step": 8466
+    },
+    {
+      "epoch": 6.373353406097102,
+      "grad_norm": 0.19206686317920685,
+      "learning_rate": 7.829880816831701e-06,
+      "loss": 0.2176,
+      "step": 8467
+    },
+    {
+      "epoch": 6.374106134738427,
+      "grad_norm": 0.16138455271720886,
+      "learning_rate": 7.826269351698912e-06,
+      "loss": 0.1718,
+      "step": 8468
+    },
+    {
+      "epoch": 6.374858863379751,
+      "grad_norm": 0.18906599283218384,
+      "learning_rate": 7.822660038055724e-06,
+      "loss": 0.1978,
+      "step": 8469
+    },
+    {
+      "epoch": 6.375611592021077,
+      "grad_norm": 0.16451495885849,
+      "learning_rate": 7.81905287629685e-06,
+      "loss": 0.0959,
+      "step": 8470
+    },
+    {
+      "epoch": 6.376364320662401,
+      "grad_norm": 0.17178013920783997,
+      "learning_rate": 7.815447866816762e-06,
+      "loss": 0.1046,
+      "step": 8471
+    },
+    {
+      "epoch": 6.377117049303726,
+      "grad_norm": 0.16750003397464752,
+      "learning_rate": 7.811845010009709e-06,
+      "loss": 0.1579,
+      "step": 8472
+    },
+    {
+      "epoch": 6.377869777945051,
+      "grad_norm": 0.18685053288936615,
+      "learning_rate": 7.808244306269676e-06,
+      "loss": 0.1719,
+      "step": 8473
+    },
+    {
+      "epoch": 6.3786225065863755,
+      "grad_norm": 0.18535222113132477,
+      "learning_rate": 7.804645755990446e-06,
+      "loss": 0.1693,
+      "step": 8474
+    },
+    {
+      "epoch": 6.3793752352277,
+      "grad_norm": 0.17474466562271118,
+      "learning_rate": 7.801049359565538e-06,
+      "loss": 0.1999,
+      "step": 8475
+    },
+    {
+      "epoch": 6.380127963869025,
+      "grad_norm": 0.16867133975028992,
+      "learning_rate": 7.797455117388256e-06,
+      "loss": 0.157,
+      "step": 8476
+    },
+    {
+      "epoch": 6.38088069251035,
+      "grad_norm": 0.16346858441829681,
+      "learning_rate": 7.79386302985166e-06,
+      "loss": 0.1296,
+      "step": 8477
+    },
+    {
+      "epoch": 6.381633421151675,
+      "grad_norm": 0.17278681695461273,
+      "learning_rate": 7.790273097348582e-06,
+      "loss": 0.1696,
+      "step": 8478
+    },
+    {
+      "epoch": 6.382386149793,
+      "grad_norm": 0.16706210374832153,
+      "learning_rate": 7.786685320271599e-06,
+      "loss": 0.1497,
+      "step": 8479
+    },
+    {
+      "epoch": 6.383138878434324,
+      "grad_norm": 0.15728768706321716,
+      "learning_rate": 7.783099699013075e-06,
+      "loss": 0.1409,
+      "step": 8480
+    },
+    {
+      "epoch": 6.3838916070756495,
+      "grad_norm": 0.19820483028888702,
+      "learning_rate": 7.77951623396512e-06,
+      "loss": 0.2151,
+      "step": 8481
+    },
+    {
+      "epoch": 6.384644335716974,
+      "grad_norm": 0.1778547465801239,
+      "learning_rate": 7.775934925519617e-06,
+      "loss": 0.176,
+      "step": 8482
+    },
+    {
+      "epoch": 6.3853970643582985,
+      "grad_norm": 0.18959541618824005,
+      "learning_rate": 7.772355774068222e-06,
+      "loss": 0.2173,
+      "step": 8483
+    },
+    {
+      "epoch": 6.386149792999624,
+      "grad_norm": 0.1910848319530487,
+      "learning_rate": 7.76877878000234e-06,
+      "loss": 0.1399,
+      "step": 8484
+    },
+    {
+      "epoch": 6.386902521640948,
+      "grad_norm": 0.1543068289756775,
+      "learning_rate": 7.76520394371314e-06,
+      "loss": 0.1704,
+      "step": 8485
+    },
+    {
+      "epoch": 6.387655250282274,
+      "grad_norm": 0.16970010101795197,
+      "learning_rate": 7.761631265591573e-06,
+      "loss": 0.2063,
+      "step": 8486
+    },
+    {
+      "epoch": 6.388407978923598,
+      "grad_norm": 0.15592484176158905,
+      "learning_rate": 7.758060746028329e-06,
+      "loss": 0.214,
+      "step": 8487
+    },
+    {
+      "epoch": 6.389160707564923,
+      "grad_norm": 0.17869749665260315,
+      "learning_rate": 7.754492385413879e-06,
+      "loss": 0.1992,
+      "step": 8488
+    },
+    {
+      "epoch": 6.389913436206248,
+      "grad_norm": 0.18303516507148743,
+      "learning_rate": 7.750926184138458e-06,
+      "loss": 0.1338,
+      "step": 8489
+    },
+    {
+      "epoch": 6.390666164847572,
+      "grad_norm": 0.17952662706375122,
+      "learning_rate": 7.747362142592058e-06,
+      "loss": 0.2463,
+      "step": 8490
+    },
+    {
+      "epoch": 6.391418893488897,
+      "grad_norm": 0.17207615077495575,
+      "learning_rate": 7.743800261164432e-06,
+      "loss": 0.1659,
+      "step": 8491
+    },
+    {
+      "epoch": 6.392171622130222,
+      "grad_norm": 0.17922644317150116,
+      "learning_rate": 7.74024054024511e-06,
+      "loss": 0.2044,
+      "step": 8492
+    },
+    {
+      "epoch": 6.392924350771547,
+      "grad_norm": 0.18443042039871216,
+      "learning_rate": 7.736682980223375e-06,
+      "loss": 0.1313,
+      "step": 8493
+    },
+    {
+      "epoch": 6.393677079412871,
+      "grad_norm": 0.1695532351732254,
+      "learning_rate": 7.733127581488274e-06,
+      "loss": 0.1351,
+      "step": 8494
+    },
+    {
+      "epoch": 6.394429808054197,
+      "grad_norm": 0.19422130286693573,
+      "learning_rate": 7.72957434442862e-06,
+      "loss": 0.2022,
+      "step": 8495
+    },
+    {
+      "epoch": 6.395182536695521,
+      "grad_norm": 0.16666589677333832,
+      "learning_rate": 7.726023269432991e-06,
+      "loss": 0.195,
+      "step": 8496
+    },
+    {
+      "epoch": 6.395935265336846,
+      "grad_norm": 0.2075410932302475,
+      "learning_rate": 7.722474356889736e-06,
+      "loss": 0.1836,
+      "step": 8497
+    },
+    {
+      "epoch": 6.396687993978171,
+      "grad_norm": 0.17491722106933594,
+      "learning_rate": 7.71892760718694e-06,
+      "loss": 0.1944,
+      "step": 8498
+    },
+    {
+      "epoch": 6.397440722619495,
+      "grad_norm": 0.1781875193119049,
+      "learning_rate": 7.715383020712489e-06,
+      "loss": 0.1511,
+      "step": 8499
+    },
+    {
+      "epoch": 6.398193451260821,
+      "grad_norm": 0.20657582581043243,
+      "learning_rate": 7.711840597853998e-06,
+      "loss": 0.1575,
+      "step": 8500
+    },
+    {
+      "epoch": 6.398946179902145,
+      "grad_norm": 0.2227921187877655,
+      "learning_rate": 7.70830033899887e-06,
+      "loss": 0.2243,
+      "step": 8501
+    },
+    {
+      "epoch": 6.39969890854347,
+      "grad_norm": 0.21205194294452667,
+      "learning_rate": 7.704762244534259e-06,
+      "loss": 0.1499,
+      "step": 8502
+    },
+    {
+      "epoch": 6.400451637184795,
+      "grad_norm": 0.1767154186964035,
+      "learning_rate": 7.701226314847093e-06,
+      "loss": 0.1779,
+      "step": 8503
+    },
+    {
+      "epoch": 6.4012043658261195,
+      "grad_norm": 0.19003282487392426,
+      "learning_rate": 7.697692550324044e-06,
+      "loss": 0.1704,
+      "step": 8504
+    },
+    {
+      "epoch": 6.401957094467445,
+      "grad_norm": 0.16436825692653656,
+      "learning_rate": 7.694160951351572e-06,
+      "loss": 0.1341,
+      "step": 8505
+    },
+    {
+      "epoch": 6.402709823108769,
+      "grad_norm": 0.17818313837051392,
+      "learning_rate": 7.690631518315876e-06,
+      "loss": 0.1265,
+      "step": 8506
+    },
+    {
+      "epoch": 6.403462551750094,
+      "grad_norm": 0.1888289600610733,
+      "learning_rate": 7.687104251602933e-06,
+      "loss": 0.18,
+      "step": 8507
+    },
+    {
+      "epoch": 6.404215280391419,
+      "grad_norm": 0.17720681428909302,
+      "learning_rate": 7.683579151598482e-06,
+      "loss": 0.1545,
+      "step": 8508
+    },
+    {
+      "epoch": 6.404968009032744,
+      "grad_norm": 0.17122051119804382,
+      "learning_rate": 7.680056218688024e-06,
+      "loss": 0.1338,
+      "step": 8509
+    },
+    {
+      "epoch": 6.405720737674068,
+      "grad_norm": 0.15313777327537537,
+      "learning_rate": 7.676535453256816e-06,
+      "loss": 0.1327,
+      "step": 8510
+    },
+    {
+      "epoch": 6.4064734663153935,
+      "grad_norm": 0.16884514689445496,
+      "learning_rate": 7.673016855689891e-06,
+      "loss": 0.2165,
+      "step": 8511
+    },
+    {
+      "epoch": 6.407226194956718,
+      "grad_norm": 0.1923987716436386,
+      "learning_rate": 7.669500426372032e-06,
+      "loss": 0.2004,
+      "step": 8512
+    },
+    {
+      "epoch": 6.407978923598042,
+      "grad_norm": 0.1861022263765335,
+      "learning_rate": 7.665986165687787e-06,
+      "loss": 0.1957,
+      "step": 8513
+    },
+    {
+      "epoch": 6.408731652239368,
+      "grad_norm": 0.1723240315914154,
+      "learning_rate": 7.662474074021474e-06,
+      "loss": 0.2225,
+      "step": 8514
+    },
+    {
+      "epoch": 6.409484380880692,
+      "grad_norm": 0.15584611892700195,
+      "learning_rate": 7.658964151757175e-06,
+      "loss": 0.1035,
+      "step": 8515
+    },
+    {
+      "epoch": 6.410237109522018,
+      "grad_norm": 0.17078381776809692,
+      "learning_rate": 7.65545639927873e-06,
+      "loss": 0.1693,
+      "step": 8516
+    },
+    {
+      "epoch": 6.410989838163342,
+      "grad_norm": 0.16031695902347565,
+      "learning_rate": 7.65195081696973e-06,
+      "loss": 0.2196,
+      "step": 8517
+    },
+    {
+      "epoch": 6.411742566804667,
+      "grad_norm": 0.18744519352912903,
+      "learning_rate": 7.648447405213555e-06,
+      "loss": 0.2059,
+      "step": 8518
+    },
+    {
+      "epoch": 6.412495295445992,
+      "grad_norm": 0.17903700470924377,
+      "learning_rate": 7.64494616439332e-06,
+      "loss": 0.1463,
+      "step": 8519
+    },
+    {
+      "epoch": 6.413248024087316,
+      "grad_norm": 0.16786512732505798,
+      "learning_rate": 7.641447094891922e-06,
+      "loss": 0.1599,
+      "step": 8520
+    },
+    {
+      "epoch": 6.414000752728642,
+      "grad_norm": 0.15818893909454346,
+      "learning_rate": 7.637950197092009e-06,
+      "loss": 0.1568,
+      "step": 8521
+    },
+    {
+      "epoch": 6.414753481369966,
+      "grad_norm": 0.17586135864257812,
+      "learning_rate": 7.634455471376007e-06,
+      "loss": 0.1553,
+      "step": 8522
+    },
+    {
+      "epoch": 6.415506210011291,
+      "grad_norm": 0.1680774837732315,
+      "learning_rate": 7.630962918126082e-06,
+      "loss": 0.195,
+      "step": 8523
+    },
+    {
+      "epoch": 6.416258938652616,
+      "grad_norm": 0.16971904039382935,
+      "learning_rate": 7.627472537724184e-06,
+      "loss": 0.1742,
+      "step": 8524
+    },
+    {
+      "epoch": 6.417011667293941,
+      "grad_norm": 0.17650258541107178,
+      "learning_rate": 7.623984330552009e-06,
+      "loss": 0.13,
+      "step": 8525
+    },
+    {
+      "epoch": 6.417764395935265,
+      "grad_norm": 0.1731211096048355,
+      "learning_rate": 7.620498296991023e-06,
+      "loss": 0.2336,
+      "step": 8526
+    },
+    {
+      "epoch": 6.41851712457659,
+      "grad_norm": 0.18388831615447998,
+      "learning_rate": 7.617014437422457e-06,
+      "loss": 0.1591,
+      "step": 8527
+    },
+    {
+      "epoch": 6.419269853217915,
+      "grad_norm": 0.1625320315361023,
+      "learning_rate": 7.613532752227298e-06,
+      "loss": 0.1312,
+      "step": 8528
+    },
+    {
+      "epoch": 6.420022581859239,
+      "grad_norm": 0.18004532158374786,
+      "learning_rate": 7.6100532417862985e-06,
+      "loss": 0.2097,
+      "step": 8529
+    },
+    {
+      "epoch": 6.420775310500565,
+      "grad_norm": 0.18996833264827728,
+      "learning_rate": 7.6065759064799735e-06,
+      "loss": 0.1945,
+      "step": 8530
+    },
+    {
+      "epoch": 6.421528039141889,
+      "grad_norm": 0.16389264166355133,
+      "learning_rate": 7.603100746688596e-06,
+      "loss": 0.1465,
+      "step": 8531
+    },
+    {
+      "epoch": 6.4222807677832146,
+      "grad_norm": 0.1651359498500824,
+      "learning_rate": 7.599627762792211e-06,
+      "loss": 0.0817,
+      "step": 8532
+    },
+    {
+      "epoch": 6.423033496424539,
+      "grad_norm": 0.17174862325191498,
+      "learning_rate": 7.5961569551706035e-06,
+      "loss": 0.1489,
+      "step": 8533
+    },
+    {
+      "epoch": 6.4237862250658635,
+      "grad_norm": 0.16846422851085663,
+      "learning_rate": 7.592688324203358e-06,
+      "loss": 0.1906,
+      "step": 8534
+    },
+    {
+      "epoch": 6.424538953707189,
+      "grad_norm": 0.22804875671863556,
+      "learning_rate": 7.589221870269779e-06,
+      "loss": 0.1713,
+      "step": 8535
+    },
+    {
+      "epoch": 6.425291682348513,
+      "grad_norm": 0.15660731494426727,
+      "learning_rate": 7.585757593748964e-06,
+      "loss": 0.1027,
+      "step": 8536
+    },
+    {
+      "epoch": 6.426044410989838,
+      "grad_norm": 0.16001489758491516,
+      "learning_rate": 7.58229549501976e-06,
+      "loss": 0.14,
+      "step": 8537
+    },
+    {
+      "epoch": 6.426797139631163,
+      "grad_norm": 0.17761625349521637,
+      "learning_rate": 7.578835574460773e-06,
+      "loss": 0.2385,
+      "step": 8538
+    },
+    {
+      "epoch": 6.427549868272488,
+      "grad_norm": 0.16351602971553802,
+      "learning_rate": 7.575377832450373e-06,
+      "loss": 0.2334,
+      "step": 8539
+    },
+    {
+      "epoch": 6.428302596913813,
+      "grad_norm": 0.1958485245704651,
+      "learning_rate": 7.571922269366699e-06,
+      "loss": 0.2217,
+      "step": 8540
+    },
+    {
+      "epoch": 6.4290553255551375,
+      "grad_norm": 0.17729803919792175,
+      "learning_rate": 7.56846888558765e-06,
+      "loss": 0.1679,
+      "step": 8541
+    },
+    {
+      "epoch": 6.429808054196462,
+      "grad_norm": 0.13873827457427979,
+      "learning_rate": 7.565017681490871e-06,
+      "loss": 0.1645,
+      "step": 8542
+    },
+    {
+      "epoch": 6.430560782837787,
+      "grad_norm": 0.1647200882434845,
+      "learning_rate": 7.561568657453792e-06,
+      "loss": 0.1792,
+      "step": 8543
+    },
+    {
+      "epoch": 6.431313511479112,
+      "grad_norm": 0.17166350781917572,
+      "learning_rate": 7.558121813853581e-06,
+      "loss": 0.1931,
+      "step": 8544
+    },
+    {
+      "epoch": 6.432066240120436,
+      "grad_norm": 0.1808650940656662,
+      "learning_rate": 7.554677151067192e-06,
+      "loss": 0.223,
+      "step": 8545
+    },
+    {
+      "epoch": 6.432818968761762,
+      "grad_norm": 0.1860407441854477,
+      "learning_rate": 7.551234669471313e-06,
+      "loss": 0.2302,
+      "step": 8546
+    },
+    {
+      "epoch": 6.433571697403086,
+      "grad_norm": 0.15443682670593262,
+      "learning_rate": 7.547794369442431e-06,
+      "loss": 0.1536,
+      "step": 8547
+    },
+    {
+      "epoch": 6.434324426044411,
+      "grad_norm": 0.18417046964168549,
+      "learning_rate": 7.544356251356751e-06,
+      "loss": 0.1398,
+      "step": 8548
+    },
+    {
+      "epoch": 6.435077154685736,
+      "grad_norm": 0.18717405200004578,
+      "learning_rate": 7.540920315590272e-06,
+      "loss": 0.1522,
+      "step": 8549
+    },
+    {
+      "epoch": 6.43582988332706,
+      "grad_norm": 0.15754063427448273,
+      "learning_rate": 7.537486562518735e-06,
+      "loss": 0.0988,
+      "step": 8550
+    },
+    {
+      "epoch": 6.436582611968386,
+      "grad_norm": 0.18257050216197968,
+      "learning_rate": 7.5340549925176614e-06,
+      "loss": 0.1496,
+      "step": 8551
+    },
+    {
+      "epoch": 6.43733534060971,
+      "grad_norm": 0.19083210825920105,
+      "learning_rate": 7.530625605962303e-06,
+      "loss": 0.1698,
+      "step": 8552
+    },
+    {
+      "epoch": 6.438088069251035,
+      "grad_norm": 0.1873539686203003,
+      "learning_rate": 7.5271984032277154e-06,
+      "loss": 0.1426,
+      "step": 8553
+    },
+    {
+      "epoch": 6.43884079789236,
+      "grad_norm": 0.1754458099603653,
+      "learning_rate": 7.523773384688679e-06,
+      "loss": 0.1609,
+      "step": 8554
+    },
+    {
+      "epoch": 6.4395935265336846,
+      "grad_norm": 0.1502501517534256,
+      "learning_rate": 7.520350550719756e-06,
+      "loss": 0.2071,
+      "step": 8555
+    },
+    {
+      "epoch": 6.440346255175009,
+      "grad_norm": 0.18195807933807373,
+      "learning_rate": 7.51692990169525e-06,
+      "loss": 0.1592,
+      "step": 8556
+    },
+    {
+      "epoch": 6.441098983816334,
+      "grad_norm": 0.21334931254386902,
+      "learning_rate": 7.513511437989248e-06,
+      "loss": 0.1722,
+      "step": 8557
+    },
+    {
+      "epoch": 6.441851712457659,
+      "grad_norm": 0.17398233711719513,
+      "learning_rate": 7.510095159975587e-06,
+      "loss": 0.205,
+      "step": 8558
+    },
+    {
+      "epoch": 6.442604441098984,
+      "grad_norm": 0.1897376924753189,
+      "learning_rate": 7.506681068027866e-06,
+      "loss": 0.2101,
+      "step": 8559
+    },
+    {
+      "epoch": 6.443357169740309,
+      "grad_norm": 0.19991818070411682,
+      "learning_rate": 7.503269162519442e-06,
+      "loss": 0.1771,
+      "step": 8560
+    },
+    {
+      "epoch": 6.444109898381633,
+      "grad_norm": 0.16688884794712067,
+      "learning_rate": 7.4998594438234375e-06,
+      "loss": 0.148,
+      "step": 8561
+    },
+    {
+      "epoch": 6.4448626270229585,
+      "grad_norm": 0.1799275279045105,
+      "learning_rate": 7.496451912312739e-06,
+      "loss": 0.1711,
+      "step": 8562
+    },
+    {
+      "epoch": 6.445615355664283,
+      "grad_norm": 0.19448840618133545,
+      "learning_rate": 7.493046568359979e-06,
+      "loss": 0.1597,
+      "step": 8563
+    },
+    {
+      "epoch": 6.4463680843056075,
+      "grad_norm": 0.16848328709602356,
+      "learning_rate": 7.489643412337571e-06,
+      "loss": 0.1799,
+      "step": 8564
+    },
+    {
+      "epoch": 6.447120812946933,
+      "grad_norm": 0.18787899613380432,
+      "learning_rate": 7.486242444617666e-06,
+      "loss": 0.1608,
+      "step": 8565
+    },
+    {
+      "epoch": 6.447873541588257,
+      "grad_norm": 0.1563655138015747,
+      "learning_rate": 7.482843665572207e-06,
+      "loss": 0.1632,
+      "step": 8566
+    },
+    {
+      "epoch": 6.448626270229582,
+      "grad_norm": 0.16346606612205505,
+      "learning_rate": 7.4794470755728625e-06,
+      "loss": 0.1395,
+      "step": 8567
+    },
+    {
+      "epoch": 6.449378998870907,
+      "grad_norm": 0.18497918546199799,
+      "learning_rate": 7.476052674991091e-06,
+      "loss": 0.1888,
+      "step": 8568
+    },
+    {
+      "epoch": 6.450131727512232,
+      "grad_norm": 0.16865411400794983,
+      "learning_rate": 7.472660464198087e-06,
+      "loss": 0.1728,
+      "step": 8569
+    },
+    {
+      "epoch": 6.450884456153557,
+      "grad_norm": 0.1900576800107956,
+      "learning_rate": 7.4692704435648315e-06,
+      "loss": 0.1391,
+      "step": 8570
+    },
+    {
+      "epoch": 6.4516371847948815,
+      "grad_norm": 0.16701185703277588,
+      "learning_rate": 7.465882613462035e-06,
+      "loss": 0.2132,
+      "step": 8571
+    },
+    {
+      "epoch": 6.452389913436206,
+      "grad_norm": 0.18810217082500458,
+      "learning_rate": 7.462496974260202e-06,
+      "loss": 0.1545,
+      "step": 8572
+    },
+    {
+      "epoch": 6.453142642077531,
+      "grad_norm": 0.17913001775741577,
+      "learning_rate": 7.4591135263295725e-06,
+      "loss": 0.2008,
+      "step": 8573
+    },
+    {
+      "epoch": 6.453895370718856,
+      "grad_norm": 0.20452788472175598,
+      "learning_rate": 7.455732270040159e-06,
+      "loss": 0.1646,
+      "step": 8574
+    },
+    {
+      "epoch": 6.45464809936018,
+      "grad_norm": 0.1993197649717331,
+      "learning_rate": 7.452353205761725e-06,
+      "loss": 0.1699,
+      "step": 8575
+    },
+    {
+      "epoch": 6.455400828001506,
+      "grad_norm": 0.19163912534713745,
+      "learning_rate": 7.4489763338638065e-06,
+      "loss": 0.2082,
+      "step": 8576
+    },
+    {
+      "epoch": 6.45615355664283,
+      "grad_norm": 0.19300583004951477,
+      "learning_rate": 7.4456016547156825e-06,
+      "loss": 0.1541,
+      "step": 8577
+    },
+    {
+      "epoch": 6.4569062852841554,
+      "grad_norm": 0.1712518185377121,
+      "learning_rate": 7.44222916868642e-06,
+      "loss": 0.1866,
+      "step": 8578
+    },
+    {
+      "epoch": 6.45765901392548,
+      "grad_norm": 0.19058220088481903,
+      "learning_rate": 7.438858876144811e-06,
+      "loss": 0.1907,
+      "step": 8579
+    },
+    {
+      "epoch": 6.458411742566804,
+      "grad_norm": 0.18842831254005432,
+      "learning_rate": 7.43549077745944e-06,
+      "loss": 0.2236,
+      "step": 8580
+    },
+    {
+      "epoch": 6.45916447120813,
+      "grad_norm": 0.19358602166175842,
+      "learning_rate": 7.432124872998627e-06,
+      "loss": 0.1591,
+      "step": 8581
+    },
+    {
+      "epoch": 6.459917199849454,
+      "grad_norm": 0.1887989491224289,
+      "learning_rate": 7.4287611631304665e-06,
+      "loss": 0.2451,
+      "step": 8582
+    },
+    {
+      "epoch": 6.460669928490779,
+      "grad_norm": 0.17678554356098175,
+      "learning_rate": 7.4253996482228125e-06,
+      "loss": 0.1412,
+      "step": 8583
+    },
+    {
+      "epoch": 6.461422657132104,
+      "grad_norm": 0.18059377372264862,
+      "learning_rate": 7.422040328643267e-06,
+      "loss": 0.1608,
+      "step": 8584
+    },
+    {
+      "epoch": 6.4621753857734285,
+      "grad_norm": 0.16441799700260162,
+      "learning_rate": 7.418683204759203e-06,
+      "loss": 0.1536,
+      "step": 8585
+    },
+    {
+      "epoch": 6.462928114414753,
+      "grad_norm": 0.16479045152664185,
+      "learning_rate": 7.415328276937755e-06,
+      "loss": 0.1502,
+      "step": 8586
+    },
+    {
+      "epoch": 6.463680843056078,
+      "grad_norm": 0.16248928010463715,
+      "learning_rate": 7.41197554554581e-06,
+      "loss": 0.1447,
+      "step": 8587
+    },
+    {
+      "epoch": 6.464433571697403,
+      "grad_norm": 0.16410884261131287,
+      "learning_rate": 7.408625010950016e-06,
+      "loss": 0.198,
+      "step": 8588
+    },
+    {
+      "epoch": 6.465186300338728,
+      "grad_norm": 0.1857629120349884,
+      "learning_rate": 7.405276673516788e-06,
+      "loss": 0.1831,
+      "step": 8589
+    },
+    {
+      "epoch": 6.465939028980053,
+      "grad_norm": 0.1712464541196823,
+      "learning_rate": 7.401930533612282e-06,
+      "loss": 0.2207,
+      "step": 8590
+    },
+    {
+      "epoch": 6.466691757621377,
+      "grad_norm": 0.18267428874969482,
+      "learning_rate": 7.398586591602445e-06,
+      "loss": 0.2608,
+      "step": 8591
+    },
+    {
+      "epoch": 6.4674444862627025,
+      "grad_norm": 0.19957943260669708,
+      "learning_rate": 7.395244847852952e-06,
+      "loss": 0.1394,
+      "step": 8592
+    },
+    {
+      "epoch": 6.468197214904027,
+      "grad_norm": 0.19329193234443665,
+      "learning_rate": 7.391905302729258e-06,
+      "loss": 0.2733,
+      "step": 8593
+    },
+    {
+      "epoch": 6.468949943545352,
+      "grad_norm": 0.18820683658123016,
+      "learning_rate": 7.388567956596566e-06,
+      "loss": 0.1502,
+      "step": 8594
+    },
+    {
+      "epoch": 6.469702672186677,
+      "grad_norm": 0.18694505095481873,
+      "learning_rate": 7.385232809819847e-06,
+      "loss": 0.1833,
+      "step": 8595
+    },
+    {
+      "epoch": 6.470455400828001,
+      "grad_norm": 0.20208749175071716,
+      "learning_rate": 7.381899862763825e-06,
+      "loss": 0.2013,
+      "step": 8596
+    },
+    {
+      "epoch": 6.471208129469327,
+      "grad_norm": 0.19781187176704407,
+      "learning_rate": 7.378569115792984e-06,
+      "loss": 0.149,
+      "step": 8597
+    },
+    {
+      "epoch": 6.471960858110651,
+      "grad_norm": 0.1557220071554184,
+      "learning_rate": 7.3752405692715725e-06,
+      "loss": 0.1603,
+      "step": 8598
+    },
+    {
+      "epoch": 6.472713586751976,
+      "grad_norm": 0.1554795205593109,
+      "learning_rate": 7.3719142235636014e-06,
+      "loss": 0.1279,
+      "step": 8599
+    },
+    {
+      "epoch": 6.473466315393301,
+      "grad_norm": 0.19753091037273407,
+      "learning_rate": 7.368590079032822e-06,
+      "loss": 0.1927,
+      "step": 8600
+    },
+    {
+      "epoch": 6.473466315393301,
+      "eval_loss": 0.20481540262699127,
+      "eval_runtime": 452.6339,
+      "eval_samples_per_second": 21.269,
+      "eval_steps_per_second": 0.665,
+      "step": 8600
+    },
+    {
+      "epoch": 6.4742190440346254,
+      "grad_norm": 0.21264825761318207,
+      "learning_rate": 7.365268136042772e-06,
+      "loss": 0.1589,
+      "step": 8601
+    },
+    {
+      "epoch": 6.47497177267595,
+      "grad_norm": 0.17639386653900146,
+      "learning_rate": 7.361948394956719e-06,
+      "loss": 0.0749,
+      "step": 8602
+    },
+    {
+      "epoch": 6.475724501317275,
+      "grad_norm": 0.16508576273918152,
+      "learning_rate": 7.358630856137714e-06,
+      "loss": 0.101,
+      "step": 8603
+    },
+    {
+      "epoch": 6.4764772299586,
+      "grad_norm": 0.1819247454404831,
+      "learning_rate": 7.355315519948556e-06,
+      "loss": 0.2558,
+      "step": 8604
+    },
+    {
+      "epoch": 6.477229958599925,
+      "grad_norm": 0.20395106077194214,
+      "learning_rate": 7.35200238675181e-06,
+      "loss": 0.1687,
+      "step": 8605
+    },
+    {
+      "epoch": 6.47798268724125,
+      "grad_norm": 0.18058380484580994,
+      "learning_rate": 7.348691456909787e-06,
+      "loss": 0.1866,
+      "step": 8606
+    },
+    {
+      "epoch": 6.478735415882574,
+      "grad_norm": 0.15431274473667145,
+      "learning_rate": 7.345382730784568e-06,
+      "loss": 0.15,
+      "step": 8607
+    },
+    {
+      "epoch": 6.479488144523899,
+      "grad_norm": 0.17147494852542877,
+      "learning_rate": 7.342076208737998e-06,
+      "loss": 0.1342,
+      "step": 8608
+    },
+    {
+      "epoch": 6.480240873165224,
+      "grad_norm": 0.1925792545080185,
+      "learning_rate": 7.3387718911316625e-06,
+      "loss": 0.1224,
+      "step": 8609
+    },
+    {
+      "epoch": 6.480993601806548,
+      "grad_norm": 0.1541198194026947,
+      "learning_rate": 7.33546977832692e-06,
+      "loss": 0.1621,
+      "step": 8610
+    },
+    {
+      "epoch": 6.481746330447874,
+      "grad_norm": 0.17691759765148163,
+      "learning_rate": 7.332169870684884e-06,
+      "loss": 0.151,
+      "step": 8611
+    },
+    {
+      "epoch": 6.482499059089198,
+      "grad_norm": 0.18294695019721985,
+      "learning_rate": 7.328872168566435e-06,
+      "loss": 0.1695,
+      "step": 8612
+    },
+    {
+      "epoch": 6.483251787730524,
+      "grad_norm": 0.18157127499580383,
+      "learning_rate": 7.3255766723321935e-06,
+      "loss": 0.2417,
+      "step": 8613
+    },
+    {
+      "epoch": 6.484004516371848,
+      "grad_norm": 0.18243664503097534,
+      "learning_rate": 7.322283382342562e-06,
+      "loss": 0.1595,
+      "step": 8614
+    },
+    {
+      "epoch": 6.4847572450131725,
+      "grad_norm": 0.15974771976470947,
+      "learning_rate": 7.318992298957676e-06,
+      "loss": 0.1105,
+      "step": 8615
+    },
+    {
+      "epoch": 6.485509973654498,
+      "grad_norm": 0.1846655309200287,
+      "learning_rate": 7.315703422537453e-06,
+      "loss": 0.2235,
+      "step": 8616
+    },
+    {
+      "epoch": 6.486262702295822,
+      "grad_norm": 0.17093998193740845,
+      "learning_rate": 7.312416753441553e-06,
+      "loss": 0.1256,
+      "step": 8617
+    },
+    {
+      "epoch": 6.487015430937147,
+      "grad_norm": 0.21246632933616638,
+      "learning_rate": 7.309132292029414e-06,
+      "loss": 0.2503,
+      "step": 8618
+    },
+    {
+      "epoch": 6.487768159578472,
+      "grad_norm": 0.16112278401851654,
+      "learning_rate": 7.305850038660205e-06,
+      "loss": 0.1334,
+      "step": 8619
+    },
+    {
+      "epoch": 6.488520888219797,
+      "grad_norm": 0.1983078569173813,
+      "learning_rate": 7.302569993692879e-06,
+      "loss": 0.1648,
+      "step": 8620
+    },
+    {
+      "epoch": 6.489273616861121,
+      "grad_norm": 0.20987948775291443,
+      "learning_rate": 7.299292157486127e-06,
+      "loss": 0.1881,
+      "step": 8621
+    },
+    {
+      "epoch": 6.4900263455024465,
+      "grad_norm": 0.18228231370449066,
+      "learning_rate": 7.296016530398414e-06,
+      "loss": 0.147,
+      "step": 8622
+    },
+    {
+      "epoch": 6.490779074143771,
+      "grad_norm": 0.16797059774398804,
+      "learning_rate": 7.292743112787958e-06,
+      "loss": 0.1385,
+      "step": 8623
+    },
+    {
+      "epoch": 6.491531802785096,
+      "grad_norm": 0.1693168729543686,
+      "learning_rate": 7.289471905012735e-06,
+      "loss": 0.1463,
+      "step": 8624
+    },
+    {
+      "epoch": 6.492284531426421,
+      "grad_norm": 0.14109045267105103,
+      "learning_rate": 7.286202907430477e-06,
+      "loss": 0.1654,
+      "step": 8625
+    },
+    {
+      "epoch": 6.493037260067745,
+      "grad_norm": 0.17250825464725494,
+      "learning_rate": 7.282936120398681e-06,
+      "loss": 0.1598,
+      "step": 8626
+    },
+    {
+      "epoch": 6.493789988709071,
+      "grad_norm": 0.18086999654769897,
+      "learning_rate": 7.279671544274591e-06,
+      "loss": 0.226,
+      "step": 8627
+    },
+    {
+      "epoch": 6.494542717350395,
+      "grad_norm": 0.14344222843647003,
+      "learning_rate": 7.2764091794152205e-06,
+      "loss": 0.1084,
+      "step": 8628
+    },
+    {
+      "epoch": 6.49529544599172,
+      "grad_norm": 0.17385877668857574,
+      "learning_rate": 7.273149026177337e-06,
+      "loss": 0.1735,
+      "step": 8629
+    },
+    {
+      "epoch": 6.496048174633045,
+      "grad_norm": 0.16827036440372467,
+      "learning_rate": 7.269891084917469e-06,
+      "loss": 0.1581,
+      "step": 8630
+    },
+    {
+      "epoch": 6.496800903274369,
+      "grad_norm": 0.16069886088371277,
+      "learning_rate": 7.266635355991891e-06,
+      "loss": 0.174,
+      "step": 8631
+    },
+    {
+      "epoch": 6.497553631915695,
+      "grad_norm": 0.18221400678157806,
+      "learning_rate": 7.263381839756652e-06,
+      "loss": 0.1451,
+      "step": 8632
+    },
+    {
+      "epoch": 6.498306360557019,
+      "grad_norm": 0.18784518539905548,
+      "learning_rate": 7.260130536567554e-06,
+      "loss": 0.1907,
+      "step": 8633
+    },
+    {
+      "epoch": 6.499059089198344,
+      "grad_norm": 0.1773412525653839,
+      "learning_rate": 7.256881446780145e-06,
+      "loss": 0.2174,
+      "step": 8634
+    },
+    {
+      "epoch": 6.499811817839669,
+      "grad_norm": 0.1760575771331787,
+      "learning_rate": 7.253634570749744e-06,
+      "loss": 0.1507,
+      "step": 8635
+    },
+    {
+      "epoch": 6.500564546480994,
+      "grad_norm": 0.15016792714595795,
+      "learning_rate": 7.25038990883143e-06,
+      "loss": 0.1491,
+      "step": 8636
+    },
+    {
+      "epoch": 6.501317275122318,
+      "grad_norm": 0.17008240520954132,
+      "learning_rate": 7.247147461380033e-06,
+      "loss": 0.1659,
+      "step": 8637
+    },
+    {
+      "epoch": 6.502070003763643,
+      "grad_norm": 0.15186147391796112,
+      "learning_rate": 7.243907228750136e-06,
+      "loss": 0.1274,
+      "step": 8638
+    },
+    {
+      "epoch": 6.502822732404968,
+      "grad_norm": 0.16892164945602417,
+      "learning_rate": 7.240669211296096e-06,
+      "loss": 0.1512,
+      "step": 8639
+    },
+    {
+      "epoch": 6.503575461046292,
+      "grad_norm": 0.19768182933330536,
+      "learning_rate": 7.237433409372005e-06,
+      "loss": 0.1869,
+      "step": 8640
+    },
+    {
+      "epoch": 6.504328189687618,
+      "grad_norm": 0.15862199664115906,
+      "learning_rate": 7.234199823331732e-06,
+      "loss": 0.1736,
+      "step": 8641
+    },
+    {
+      "epoch": 6.505080918328942,
+      "grad_norm": 0.154807910323143,
+      "learning_rate": 7.230968453528898e-06,
+      "loss": 0.1292,
+      "step": 8642
+    },
+    {
+      "epoch": 6.505833646970268,
+      "grad_norm": 0.17775674164295197,
+      "learning_rate": 7.227739300316885e-06,
+      "loss": 0.1782,
+      "step": 8643
+    },
+    {
+      "epoch": 6.506586375611592,
+      "grad_norm": 0.1810070425271988,
+      "learning_rate": 7.2245123640488175e-06,
+      "loss": 0.172,
+      "step": 8644
+    },
+    {
+      "epoch": 6.5073391042529165,
+      "grad_norm": 0.1971685290336609,
+      "learning_rate": 7.221287645077597e-06,
+      "loss": 0.1479,
+      "step": 8645
+    },
+    {
+      "epoch": 6.508091832894242,
+      "grad_norm": 0.16063302755355835,
+      "learning_rate": 7.21806514375587e-06,
+      "loss": 0.1144,
+      "step": 8646
+    },
+    {
+      "epoch": 6.508844561535566,
+      "grad_norm": 0.18919748067855835,
+      "learning_rate": 7.214844860436043e-06,
+      "loss": 0.1596,
+      "step": 8647
+    },
+    {
+      "epoch": 6.509597290176892,
+      "grad_norm": 0.1790386438369751,
+      "learning_rate": 7.211626795470286e-06,
+      "loss": 0.2463,
+      "step": 8648
+    },
+    {
+      "epoch": 6.510350018818216,
+      "grad_norm": 0.1820467710494995,
+      "learning_rate": 7.20841094921052e-06,
+      "loss": 0.2011,
+      "step": 8649
+    },
+    {
+      "epoch": 6.511102747459541,
+      "grad_norm": 0.16146540641784668,
+      "learning_rate": 7.205197322008425e-06,
+      "loss": 0.1567,
+      "step": 8650
+    },
+    {
+      "epoch": 6.511855476100866,
+      "grad_norm": 0.17264056205749512,
+      "learning_rate": 7.20198591421544e-06,
+      "loss": 0.1185,
+      "step": 8651
+    },
+    {
+      "epoch": 6.5126082047421905,
+      "grad_norm": 0.16781729459762573,
+      "learning_rate": 7.198776726182754e-06,
+      "loss": 0.1176,
+      "step": 8652
+    },
+    {
+      "epoch": 6.513360933383515,
+      "grad_norm": 0.16635200381278992,
+      "learning_rate": 7.195569758261322e-06,
+      "loss": 0.2207,
+      "step": 8653
+    },
+    {
+      "epoch": 6.51411366202484,
+      "grad_norm": 0.18358290195465088,
+      "learning_rate": 7.192365010801856e-06,
+      "loss": 0.2341,
+      "step": 8654
+    },
+    {
+      "epoch": 6.514866390666165,
+      "grad_norm": 0.17002849280834198,
+      "learning_rate": 7.189162484154824e-06,
+      "loss": 0.1261,
+      "step": 8655
+    },
+    {
+      "epoch": 6.515619119307489,
+      "grad_norm": 0.17498169839382172,
+      "learning_rate": 7.185962178670439e-06,
+      "loss": 0.2823,
+      "step": 8656
+    },
+    {
+      "epoch": 6.516371847948815,
+      "grad_norm": 0.17021633684635162,
+      "learning_rate": 7.182764094698691e-06,
+      "loss": 0.1973,
+      "step": 8657
+    },
+    {
+      "epoch": 6.517124576590139,
+      "grad_norm": 0.17028658092021942,
+      "learning_rate": 7.179568232589319e-06,
+      "loss": 0.1976,
+      "step": 8658
+    },
+    {
+      "epoch": 6.517877305231464,
+      "grad_norm": 0.17492006719112396,
+      "learning_rate": 7.176374592691808e-06,
+      "loss": 0.1201,
+      "step": 8659
+    },
+    {
+      "epoch": 6.518630033872789,
+      "grad_norm": 0.17490562796592712,
+      "learning_rate": 7.1731831753554166e-06,
+      "loss": 0.1157,
+      "step": 8660
+    },
+    {
+      "epoch": 6.519382762514113,
+      "grad_norm": 0.14255507290363312,
+      "learning_rate": 7.169993980929154e-06,
+      "loss": 0.1148,
+      "step": 8661
+    },
+    {
+      "epoch": 6.520135491155439,
+      "grad_norm": 0.17967331409454346,
+      "learning_rate": 7.166807009761786e-06,
+      "loss": 0.206,
+      "step": 8662
+    },
+    {
+      "epoch": 6.520888219796763,
+      "grad_norm": 0.18436068296432495,
+      "learning_rate": 7.1636222622018284e-06,
+      "loss": 0.1435,
+      "step": 8663
+    },
+    {
+      "epoch": 6.521640948438088,
+      "grad_norm": 0.19171303510665894,
+      "learning_rate": 7.16043973859757e-06,
+      "loss": 0.1157,
+      "step": 8664
+    },
+    {
+      "epoch": 6.522393677079413,
+      "grad_norm": 0.1549486219882965,
+      "learning_rate": 7.157259439297036e-06,
+      "loss": 0.1595,
+      "step": 8665
+    },
+    {
+      "epoch": 6.523146405720738,
+      "grad_norm": 0.17186720669269562,
+      "learning_rate": 7.154081364648032e-06,
+      "loss": 0.242,
+      "step": 8666
+    },
+    {
+      "epoch": 6.523899134362063,
+      "grad_norm": 0.18136602640151978,
+      "learning_rate": 7.150905514998088e-06,
+      "loss": 0.1763,
+      "step": 8667
+    },
+    {
+      "epoch": 6.524651863003387,
+      "grad_norm": 0.1762436479330063,
+      "learning_rate": 7.147731890694534e-06,
+      "loss": 0.1868,
+      "step": 8668
+    },
+    {
+      "epoch": 6.525404591644712,
+      "grad_norm": 0.16839082539081573,
+      "learning_rate": 7.1445604920844174e-06,
+      "loss": 0.1714,
+      "step": 8669
+    },
+    {
+      "epoch": 6.526157320286037,
+      "grad_norm": 0.1651378720998764,
+      "learning_rate": 7.141391319514565e-06,
+      "loss": 0.1574,
+      "step": 8670
+    },
+    {
+      "epoch": 6.526910048927362,
+      "grad_norm": 0.1841336041688919,
+      "learning_rate": 7.138224373331542e-06,
+      "loss": 0.2344,
+      "step": 8671
+    },
+    {
+      "epoch": 6.527662777568686,
+      "grad_norm": 0.1784990280866623,
+      "learning_rate": 7.135059653881696e-06,
+      "loss": 0.107,
+      "step": 8672
+    },
+    {
+      "epoch": 6.5284155062100115,
+      "grad_norm": 0.1893773227930069,
+      "learning_rate": 7.1318971615111e-06,
+      "loss": 0.1321,
+      "step": 8673
+    },
+    {
+      "epoch": 6.529168234851336,
+      "grad_norm": 0.1963060349225998,
+      "learning_rate": 7.128736896565613e-06,
+      "loss": 0.2202,
+      "step": 8674
+    },
+    {
+      "epoch": 6.5299209634926605,
+      "grad_norm": 0.14271585643291473,
+      "learning_rate": 7.125578859390829e-06,
+      "loss": 0.1164,
+      "step": 8675
+    },
+    {
+      "epoch": 6.530673692133986,
+      "grad_norm": 0.16065406799316406,
+      "learning_rate": 7.122423050332111e-06,
+      "loss": 0.1717,
+      "step": 8676
+    },
+    {
+      "epoch": 6.53142642077531,
+      "grad_norm": 0.20744173228740692,
+      "learning_rate": 7.119269469734567e-06,
+      "loss": 0.1596,
+      "step": 8677
+    },
+    {
+      "epoch": 6.532179149416636,
+      "grad_norm": 0.17039306461811066,
+      "learning_rate": 7.116118117943073e-06,
+      "loss": 0.2044,
+      "step": 8678
+    },
+    {
+      "epoch": 6.53293187805796,
+      "grad_norm": 0.18708369135856628,
+      "learning_rate": 7.112968995302255e-06,
+      "loss": 0.2095,
+      "step": 8679
+    },
+    {
+      "epoch": 6.533684606699285,
+      "grad_norm": 0.2110554426908493,
+      "learning_rate": 7.1098221021564995e-06,
+      "loss": 0.1901,
+      "step": 8680
+    },
+    {
+      "epoch": 6.53443733534061,
+      "grad_norm": 0.1907154619693756,
+      "learning_rate": 7.106677438849937e-06,
+      "loss": 0.1471,
+      "step": 8681
+    },
+    {
+      "epoch": 6.5351900639819345,
+      "grad_norm": 0.21015319228172302,
+      "learning_rate": 7.103535005726472e-06,
+      "loss": 0.2108,
+      "step": 8682
+    },
+    {
+      "epoch": 6.535942792623259,
+      "grad_norm": 0.17477384209632874,
+      "learning_rate": 7.100394803129755e-06,
+      "loss": 0.1032,
+      "step": 8683
+    },
+    {
+      "epoch": 6.536695521264584,
+      "grad_norm": 0.188886821269989,
+      "learning_rate": 7.09725683140319e-06,
+      "loss": 0.1845,
+      "step": 8684
+    },
+    {
+      "epoch": 6.537448249905909,
+      "grad_norm": 0.1801908314228058,
+      "learning_rate": 7.094121090889945e-06,
+      "loss": 0.1023,
+      "step": 8685
+    },
+    {
+      "epoch": 6.538200978547234,
+      "grad_norm": 0.18543006479740143,
+      "learning_rate": 7.0909875819329315e-06,
+      "loss": 0.1374,
+      "step": 8686
+    },
+    {
+      "epoch": 6.538953707188559,
+      "grad_norm": 0.1716979295015335,
+      "learning_rate": 7.087856304874839e-06,
+      "loss": 0.1289,
+      "step": 8687
+    },
+    {
+      "epoch": 6.539706435829883,
+      "grad_norm": 0.16346833109855652,
+      "learning_rate": 7.084727260058087e-06,
+      "loss": 0.1494,
+      "step": 8688
+    },
+    {
+      "epoch": 6.5404591644712085,
+      "grad_norm": 0.17400692403316498,
+      "learning_rate": 7.081600447824871e-06,
+      "loss": 0.1386,
+      "step": 8689
+    },
+    {
+      "epoch": 6.541211893112533,
+      "grad_norm": 0.18214616179466248,
+      "learning_rate": 7.078475868517129e-06,
+      "loss": 0.0598,
+      "step": 8690
+    },
+    {
+      "epoch": 6.541964621753857,
+      "grad_norm": 0.19266574084758759,
+      "learning_rate": 7.0753535224765685e-06,
+      "loss": 0.1582,
+      "step": 8691
+    },
+    {
+      "epoch": 6.542717350395183,
+      "grad_norm": 0.20527704060077667,
+      "learning_rate": 7.072233410044628e-06,
+      "loss": 0.1881,
+      "step": 8692
+    },
+    {
+      "epoch": 6.543470079036507,
+      "grad_norm": 0.16181211173534393,
+      "learning_rate": 7.069115531562538e-06,
+      "loss": 0.0989,
+      "step": 8693
+    },
+    {
+      "epoch": 6.544222807677832,
+      "grad_norm": 0.1856871098279953,
+      "learning_rate": 7.065999887371249e-06,
+      "loss": 0.1677,
+      "step": 8694
+    },
+    {
+      "epoch": 6.544975536319157,
+      "grad_norm": 0.15506573021411896,
+      "learning_rate": 7.062886477811499e-06,
+      "loss": 0.1253,
+      "step": 8695
+    },
+    {
+      "epoch": 6.5457282649604815,
+      "grad_norm": 0.16163823008537292,
+      "learning_rate": 7.059775303223752e-06,
+      "loss": 0.1456,
+      "step": 8696
+    },
+    {
+      "epoch": 6.546480993601807,
+      "grad_norm": 0.17330923676490784,
+      "learning_rate": 7.056666363948249e-06,
+      "loss": 0.1481,
+      "step": 8697
+    },
+    {
+      "epoch": 6.547233722243131,
+      "grad_norm": 0.17828772962093353,
+      "learning_rate": 7.053559660324967e-06,
+      "loss": 0.1575,
+      "step": 8698
+    },
+    {
+      "epoch": 6.547986450884456,
+      "grad_norm": 0.14509451389312744,
+      "learning_rate": 7.050455192693672e-06,
+      "loss": 0.0932,
+      "step": 8699
+    },
+    {
+      "epoch": 6.548739179525781,
+      "grad_norm": 0.19216197729110718,
+      "learning_rate": 7.047352961393844e-06,
+      "loss": 0.1125,
+      "step": 8700
+    },
+    {
+      "epoch": 6.549491908167106,
+      "grad_norm": 0.18833604454994202,
+      "learning_rate": 7.0442529667647525e-06,
+      "loss": 0.1529,
+      "step": 8701
+    },
+    {
+      "epoch": 6.550244636808431,
+      "grad_norm": 0.17198337614536285,
+      "learning_rate": 7.041155209145396e-06,
+      "loss": 0.1172,
+      "step": 8702
+    },
+    {
+      "epoch": 6.5509973654497555,
+      "grad_norm": 0.2130417376756668,
+      "learning_rate": 7.038059688874549e-06,
+      "loss": 0.2379,
+      "step": 8703
+    },
+    {
+      "epoch": 6.55175009409108,
+      "grad_norm": 0.18716904520988464,
+      "learning_rate": 7.034966406290731e-06,
+      "loss": 0.1946,
+      "step": 8704
+    },
+    {
+      "epoch": 6.552502822732405,
+      "grad_norm": 0.17529110610485077,
+      "learning_rate": 7.031875361732214e-06,
+      "loss": 0.1096,
+      "step": 8705
+    },
+    {
+      "epoch": 6.55325555137373,
+      "grad_norm": 0.20972490310668945,
+      "learning_rate": 7.028786555537037e-06,
+      "loss": 0.2783,
+      "step": 8706
+    },
+    {
+      "epoch": 6.554008280015054,
+      "grad_norm": 0.16695645451545715,
+      "learning_rate": 7.025699988042982e-06,
+      "loss": 0.1626,
+      "step": 8707
+    },
+    {
+      "epoch": 6.55476100865638,
+      "grad_norm": 0.19295114278793335,
+      "learning_rate": 7.022615659587599e-06,
+      "loss": 0.0913,
+      "step": 8708
+    },
+    {
+      "epoch": 6.555513737297704,
+      "grad_norm": 0.17290306091308594,
+      "learning_rate": 7.019533570508177e-06,
+      "loss": 0.1292,
+      "step": 8709
+    },
+    {
+      "epoch": 6.556266465939029,
+      "grad_norm": 0.1823475956916809,
+      "learning_rate": 7.016453721141774e-06,
+      "loss": 0.1687,
+      "step": 8710
+    },
+    {
+      "epoch": 6.557019194580354,
+      "grad_norm": 0.1874026656150818,
+      "learning_rate": 7.013376111825189e-06,
+      "loss": 0.1765,
+      "step": 8711
+    },
+    {
+      "epoch": 6.5577719232216785,
+      "grad_norm": 0.20174232125282288,
+      "learning_rate": 7.010300742894997e-06,
+      "loss": 0.2004,
+      "step": 8712
+    },
+    {
+      "epoch": 6.558524651863003,
+      "grad_norm": 0.17043083906173706,
+      "learning_rate": 7.007227614687507e-06,
+      "loss": 0.1929,
+      "step": 8713
+    },
+    {
+      "epoch": 6.559277380504328,
+      "grad_norm": 0.16441985964775085,
+      "learning_rate": 7.004156727538799e-06,
+      "loss": 0.1785,
+      "step": 8714
+    },
+    {
+      "epoch": 6.560030109145653,
+      "grad_norm": 0.16362488269805908,
+      "learning_rate": 7.001088081784693e-06,
+      "loss": 0.1516,
+      "step": 8715
+    },
+    {
+      "epoch": 6.560782837786978,
+      "grad_norm": 0.15362992882728577,
+      "learning_rate": 6.998021677760777e-06,
+      "loss": 0.143,
+      "step": 8716
+    },
+    {
+      "epoch": 6.561535566428303,
+      "grad_norm": 0.21505522727966309,
+      "learning_rate": 6.9949575158023775e-06,
+      "loss": 0.1309,
+      "step": 8717
+    },
+    {
+      "epoch": 6.562288295069627,
+      "grad_norm": 0.18075744807720184,
+      "learning_rate": 6.991895596244605e-06,
+      "loss": 0.1725,
+      "step": 8718
+    },
+    {
+      "epoch": 6.563041023710952,
+      "grad_norm": 0.1866886168718338,
+      "learning_rate": 6.9888359194222945e-06,
+      "loss": 0.2166,
+      "step": 8719
+    },
+    {
+      "epoch": 6.563793752352277,
+      "grad_norm": 0.18078143894672394,
+      "learning_rate": 6.9857784856700495e-06,
+      "loss": 0.1274,
+      "step": 8720
+    },
+    {
+      "epoch": 6.564546480993602,
+      "grad_norm": 0.16110828518867493,
+      "learning_rate": 6.982723295322227e-06,
+      "loss": 0.1167,
+      "step": 8721
+    },
+    {
+      "epoch": 6.565299209634927,
+      "grad_norm": 0.175758495926857,
+      "learning_rate": 6.979670348712937e-06,
+      "loss": 0.1988,
+      "step": 8722
+    },
+    {
+      "epoch": 6.566051938276251,
+      "grad_norm": 0.19445061683654785,
+      "learning_rate": 6.976619646176047e-06,
+      "loss": 0.122,
+      "step": 8723
+    },
+    {
+      "epoch": 6.566804666917577,
+      "grad_norm": 0.16651630401611328,
+      "learning_rate": 6.97357118804517e-06,
+      "loss": 0.1321,
+      "step": 8724
+    },
+    {
+      "epoch": 6.567557395558901,
+      "grad_norm": 0.16336560249328613,
+      "learning_rate": 6.97052497465369e-06,
+      "loss": 0.1671,
+      "step": 8725
+    },
+    {
+      "epoch": 6.5683101242002255,
+      "grad_norm": 0.20148351788520813,
+      "learning_rate": 6.967481006334738e-06,
+      "loss": 0.1615,
+      "step": 8726
+    },
+    {
+      "epoch": 6.569062852841551,
+      "grad_norm": 0.17917798459529877,
+      "learning_rate": 6.964439283421188e-06,
+      "loss": 0.1647,
+      "step": 8727
+    },
+    {
+      "epoch": 6.569815581482875,
+      "grad_norm": 0.1619548201560974,
+      "learning_rate": 6.961399806245683e-06,
+      "loss": 0.178,
+      "step": 8728
+    },
+    {
+      "epoch": 6.5705683101242,
+      "grad_norm": 0.16736602783203125,
+      "learning_rate": 6.9583625751406215e-06,
+      "loss": 0.163,
+      "step": 8729
+    },
+    {
+      "epoch": 6.571321038765525,
+      "grad_norm": 0.17504936456680298,
+      "learning_rate": 6.95532759043814e-06,
+      "loss": 0.1958,
+      "step": 8730
+    },
+    {
+      "epoch": 6.57207376740685,
+      "grad_norm": 0.16598738729953766,
+      "learning_rate": 6.952294852470144e-06,
+      "loss": 0.1168,
+      "step": 8731
+    },
+    {
+      "epoch": 6.572826496048174,
+      "grad_norm": 0.18185509741306305,
+      "learning_rate": 6.94926436156829e-06,
+      "loss": 0.1813,
+      "step": 8732
+    },
+    {
+      "epoch": 6.5735792246894995,
+      "grad_norm": 0.1841517686843872,
+      "learning_rate": 6.946236118063992e-06,
+      "loss": 0.1609,
+      "step": 8733
+    },
+    {
+      "epoch": 6.574331953330824,
+      "grad_norm": 0.1882031261920929,
+      "learning_rate": 6.943210122288406e-06,
+      "loss": 0.1878,
+      "step": 8734
+    },
+    {
+      "epoch": 6.575084681972149,
+      "grad_norm": 0.1840013712644577,
+      "learning_rate": 6.940186374572458e-06,
+      "loss": 0.1583,
+      "step": 8735
+    },
+    {
+      "epoch": 6.575837410613474,
+      "grad_norm": 0.19778448343276978,
+      "learning_rate": 6.937164875246813e-06,
+      "loss": 0.175,
+      "step": 8736
+    },
+    {
+      "epoch": 6.576590139254798,
+      "grad_norm": 0.182045578956604,
+      "learning_rate": 6.934145624641903e-06,
+      "loss": 0.2622,
+      "step": 8737
+    },
+    {
+      "epoch": 6.577342867896124,
+      "grad_norm": 0.18719130754470825,
+      "learning_rate": 6.9311286230879065e-06,
+      "loss": 0.1156,
+      "step": 8738
+    },
+    {
+      "epoch": 6.578095596537448,
+      "grad_norm": 0.18541914224624634,
+      "learning_rate": 6.928113870914761e-06,
+      "loss": 0.2019,
+      "step": 8739
+    },
+    {
+      "epoch": 6.5788483251787735,
+      "grad_norm": 0.17916598916053772,
+      "learning_rate": 6.925101368452151e-06,
+      "loss": 0.1813,
+      "step": 8740
+    },
+    {
+      "epoch": 6.579601053820098,
+      "grad_norm": 0.19070203602313995,
+      "learning_rate": 6.9220911160295264e-06,
+      "loss": 0.2651,
+      "step": 8741
+    },
+    {
+      "epoch": 6.580353782461422,
+      "grad_norm": 0.17931179702281952,
+      "learning_rate": 6.919083113976071e-06,
+      "loss": 0.1433,
+      "step": 8742
+    },
+    {
+      "epoch": 6.581106511102748,
+      "grad_norm": 0.15463873744010925,
+      "learning_rate": 6.916077362620748e-06,
+      "loss": 0.0795,
+      "step": 8743
+    },
+    {
+      "epoch": 6.581859239744072,
+      "grad_norm": 0.18152925372123718,
+      "learning_rate": 6.913073862292256e-06,
+      "loss": 0.1918,
+      "step": 8744
+    },
+    {
+      "epoch": 6.582611968385397,
+      "grad_norm": 0.17126090824604034,
+      "learning_rate": 6.910072613319056e-06,
+      "loss": 0.2407,
+      "step": 8745
+    },
+    {
+      "epoch": 6.583364697026722,
+      "grad_norm": 0.19360865652561188,
+      "learning_rate": 6.907073616029357e-06,
+      "loss": 0.179,
+      "step": 8746
+    },
+    {
+      "epoch": 6.584117425668047,
+      "grad_norm": 0.1651044636964798,
+      "learning_rate": 6.904076870751129e-06,
+      "loss": 0.1722,
+      "step": 8747
+    },
+    {
+      "epoch": 6.584870154309371,
+      "grad_norm": 0.19431647658348083,
+      "learning_rate": 6.901082377812083e-06,
+      "loss": 0.1385,
+      "step": 8748
+    },
+    {
+      "epoch": 6.585622882950696,
+      "grad_norm": 0.16501113772392273,
+      "learning_rate": 6.898090137539702e-06,
+      "loss": 0.1898,
+      "step": 8749
+    },
+    {
+      "epoch": 6.586375611592021,
+      "grad_norm": 0.17647790908813477,
+      "learning_rate": 6.8951001502612065e-06,
+      "loss": 0.1645,
+      "step": 8750
+    },
+    {
+      "epoch": 6.587128340233346,
+      "grad_norm": 0.17191019654273987,
+      "learning_rate": 6.8921124163035865e-06,
+      "loss": 0.1523,
+      "step": 8751
+    },
+    {
+      "epoch": 6.587881068874671,
+      "grad_norm": 0.1746109127998352,
+      "learning_rate": 6.889126935993562e-06,
+      "loss": 0.1141,
+      "step": 8752
+    },
+    {
+      "epoch": 6.588633797515995,
+      "grad_norm": 0.1517232358455658,
+      "learning_rate": 6.8861437096576305e-06,
+      "loss": 0.1006,
+      "step": 8753
+    },
+    {
+      "epoch": 6.589386526157321,
+      "grad_norm": 0.21362264454364777,
+      "learning_rate": 6.883162737622032e-06,
+      "loss": 0.107,
+      "step": 8754
+    },
+    {
+      "epoch": 6.590139254798645,
+      "grad_norm": 0.16137628257274628,
+      "learning_rate": 6.880184020212757e-06,
+      "loss": 0.1264,
+      "step": 8755
+    },
+    {
+      "epoch": 6.5908919834399695,
+      "grad_norm": 0.16213738918304443,
+      "learning_rate": 6.877207557755555e-06,
+      "loss": 0.1126,
+      "step": 8756
+    },
+    {
+      "epoch": 6.591644712081295,
+      "grad_norm": 0.1766202300786972,
+      "learning_rate": 6.87423335057593e-06,
+      "loss": 0.1412,
+      "step": 8757
+    },
+    {
+      "epoch": 6.592397440722619,
+      "grad_norm": 0.1546422690153122,
+      "learning_rate": 6.871261398999139e-06,
+      "loss": 0.0873,
+      "step": 8758
+    },
+    {
+      "epoch": 6.593150169363945,
+      "grad_norm": 0.1712450236082077,
+      "learning_rate": 6.8682917033501795e-06,
+      "loss": 0.116,
+      "step": 8759
+    },
+    {
+      "epoch": 6.593902898005269,
+      "grad_norm": 0.1727568805217743,
+      "learning_rate": 6.865324263953829e-06,
+      "loss": 0.1465,
+      "step": 8760
+    },
+    {
+      "epoch": 6.594655626646594,
+      "grad_norm": 0.20834539830684662,
+      "learning_rate": 6.862359081134585e-06,
+      "loss": 0.1823,
+      "step": 8761
+    },
+    {
+      "epoch": 6.595408355287919,
+      "grad_norm": 0.1812629997730255,
+      "learning_rate": 6.8593961552167266e-06,
+      "loss": 0.1835,
+      "step": 8762
+    },
+    {
+      "epoch": 6.5961610839292435,
+      "grad_norm": 0.15491341054439545,
+      "learning_rate": 6.8564354865242705e-06,
+      "loss": 0.1501,
+      "step": 8763
+    },
+    {
+      "epoch": 6.596913812570568,
+      "grad_norm": 0.18418744206428528,
+      "learning_rate": 6.853477075380997e-06,
+      "loss": 0.1555,
+      "step": 8764
+    },
+    {
+      "epoch": 6.597666541211893,
+      "grad_norm": 0.16252389550209045,
+      "learning_rate": 6.8505209221104245e-06,
+      "loss": 0.1767,
+      "step": 8765
+    },
+    {
+      "epoch": 6.598419269853218,
+      "grad_norm": 0.18282605707645416,
+      "learning_rate": 6.8475670270358424e-06,
+      "loss": 0.0894,
+      "step": 8766
+    },
+    {
+      "epoch": 6.599171998494542,
+      "grad_norm": 0.19777922332286835,
+      "learning_rate": 6.844615390480274e-06,
+      "loss": 0.1885,
+      "step": 8767
+    },
+    {
+      "epoch": 6.599924727135868,
+      "grad_norm": 0.16545997560024261,
+      "learning_rate": 6.841666012766515e-06,
+      "loss": 0.1839,
+      "step": 8768
+    },
+    {
+      "epoch": 6.600677455777192,
+      "grad_norm": 0.20650655031204224,
+      "learning_rate": 6.8387188942171e-06,
+      "loss": 0.244,
+      "step": 8769
+    },
+    {
+      "epoch": 6.6014301844185175,
+      "grad_norm": 0.1640968769788742,
+      "learning_rate": 6.835774035154326e-06,
+      "loss": 0.1387,
+      "step": 8770
+    },
+    {
+      "epoch": 6.602182913059842,
+      "grad_norm": 0.38798296451568604,
+      "learning_rate": 6.832831435900233e-06,
+      "loss": 0.258,
+      "step": 8771
+    },
+    {
+      "epoch": 6.602935641701166,
+      "grad_norm": 0.16226357221603394,
+      "learning_rate": 6.829891096776625e-06,
+      "loss": 0.2202,
+      "step": 8772
+    },
+    {
+      "epoch": 6.603688370342492,
+      "grad_norm": 0.1825396567583084,
+      "learning_rate": 6.826953018105045e-06,
+      "loss": 0.1907,
+      "step": 8773
+    },
+    {
+      "epoch": 6.604441098983816,
+      "grad_norm": 0.2055538296699524,
+      "learning_rate": 6.8240172002068005e-06,
+      "loss": 0.1831,
+      "step": 8774
+    },
+    {
+      "epoch": 6.605193827625142,
+      "grad_norm": 0.18315626680850983,
+      "learning_rate": 6.82108364340295e-06,
+      "loss": 0.2182,
+      "step": 8775
+    },
+    {
+      "epoch": 6.605946556266466,
+      "grad_norm": 0.16431283950805664,
+      "learning_rate": 6.818152348014306e-06,
+      "loss": 0.1337,
+      "step": 8776
+    },
+    {
+      "epoch": 6.606699284907791,
+      "grad_norm": 0.20885993540287018,
+      "learning_rate": 6.815223314361424e-06,
+      "loss": 0.1151,
+      "step": 8777
+    },
+    {
+      "epoch": 6.607452013549116,
+      "grad_norm": 0.1657983958721161,
+      "learning_rate": 6.81229654276462e-06,
+      "loss": 0.1786,
+      "step": 8778
+    },
+    {
+      "epoch": 6.60820474219044,
+      "grad_norm": 0.16966092586517334,
+      "learning_rate": 6.809372033543965e-06,
+      "loss": 0.0801,
+      "step": 8779
+    },
+    {
+      "epoch": 6.608957470831765,
+      "grad_norm": 0.17843039333820343,
+      "learning_rate": 6.806449787019274e-06,
+      "loss": 0.2506,
+      "step": 8780
+    },
+    {
+      "epoch": 6.60971019947309,
+      "grad_norm": 0.16214944422245026,
+      "learning_rate": 6.803529803510122e-06,
+      "loss": 0.1053,
+      "step": 8781
+    },
+    {
+      "epoch": 6.610462928114415,
+      "grad_norm": 0.189442440867424,
+      "learning_rate": 6.800612083335836e-06,
+      "loss": 0.2072,
+      "step": 8782
+    },
+    {
+      "epoch": 6.611215656755739,
+      "grad_norm": 0.18750214576721191,
+      "learning_rate": 6.797696626815493e-06,
+      "loss": 0.2339,
+      "step": 8783
+    },
+    {
+      "epoch": 6.6119683853970646,
+      "grad_norm": 0.15877769887447357,
+      "learning_rate": 6.794783434267918e-06,
+      "loss": 0.1296,
+      "step": 8784
+    },
+    {
+      "epoch": 6.612721114038389,
+      "grad_norm": 0.18937177956104279,
+      "learning_rate": 6.791872506011701e-06,
+      "loss": 0.2146,
+      "step": 8785
+    },
+    {
+      "epoch": 6.6134738426797135,
+      "grad_norm": 0.15070746839046478,
+      "learning_rate": 6.788963842365169e-06,
+      "loss": 0.1064,
+      "step": 8786
+    },
+    {
+      "epoch": 6.614226571321039,
+      "grad_norm": 0.1627299040555954,
+      "learning_rate": 6.786057443646415e-06,
+      "loss": 0.1484,
+      "step": 8787
+    },
+    {
+      "epoch": 6.614979299962363,
+      "grad_norm": 0.16055108606815338,
+      "learning_rate": 6.78315331017327e-06,
+      "loss": 0.1343,
+      "step": 8788
+    },
+    {
+      "epoch": 6.615732028603689,
+      "grad_norm": 0.17508849501609802,
+      "learning_rate": 6.780251442263337e-06,
+      "loss": 0.121,
+      "step": 8789
+    },
+    {
+      "epoch": 6.616484757245013,
+      "grad_norm": 0.17959697544574738,
+      "learning_rate": 6.777351840233952e-06,
+      "loss": 0.1552,
+      "step": 8790
+    },
+    {
+      "epoch": 6.617237485886338,
+      "grad_norm": 0.17748263478279114,
+      "learning_rate": 6.774454504402218e-06,
+      "loss": 0.1953,
+      "step": 8791
+    },
+    {
+      "epoch": 6.617990214527663,
+      "grad_norm": 0.19089967012405396,
+      "learning_rate": 6.771559435084976e-06,
+      "loss": 0.1787,
+      "step": 8792
+    },
+    {
+      "epoch": 6.6187429431689875,
+      "grad_norm": 0.18505772948265076,
+      "learning_rate": 6.768666632598833e-06,
+      "loss": 0.2496,
+      "step": 8793
+    },
+    {
+      "epoch": 6.619495671810313,
+      "grad_norm": 0.20344126224517822,
+      "learning_rate": 6.765776097260128e-06,
+      "loss": 0.2145,
+      "step": 8794
+    },
+    {
+      "epoch": 6.620248400451637,
+      "grad_norm": 0.17080731689929962,
+      "learning_rate": 6.762887829384986e-06,
+      "loss": 0.1457,
+      "step": 8795
+    },
+    {
+      "epoch": 6.621001129092962,
+      "grad_norm": 0.1856684386730194,
+      "learning_rate": 6.760001829289248e-06,
+      "loss": 0.2214,
+      "step": 8796
+    },
+    {
+      "epoch": 6.621753857734287,
+      "grad_norm": 0.16184449195861816,
+      "learning_rate": 6.7571180972885316e-06,
+      "loss": 0.1494,
+      "step": 8797
+    },
+    {
+      "epoch": 6.622506586375612,
+      "grad_norm": 0.17844931781291962,
+      "learning_rate": 6.75423663369819e-06,
+      "loss": 0.1156,
+      "step": 8798
+    },
+    {
+      "epoch": 6.623259315016936,
+      "grad_norm": 0.19564004242420197,
+      "learning_rate": 6.75135743883334e-06,
+      "loss": 0.1728,
+      "step": 8799
+    },
+    {
+      "epoch": 6.6240120436582615,
+      "grad_norm": 0.17380884289741516,
+      "learning_rate": 6.748480513008844e-06,
+      "loss": 0.1806,
+      "step": 8800
+    },
+    {
+      "epoch": 6.6240120436582615,
+      "eval_loss": 0.2037574052810669,
+      "eval_runtime": 452.9187,
+      "eval_samples_per_second": 21.255,
+      "eval_steps_per_second": 0.665,
+      "step": 8800
+    },
+    {
+      "epoch": 6.624764772299586,
+      "grad_norm": 0.17840361595153809,
+      "learning_rate": 6.745605856539325e-06,
+      "loss": 0.1165,
+      "step": 8801
+    },
+    {
+      "epoch": 6.62551750094091,
+      "grad_norm": 0.2075352966785431,
+      "learning_rate": 6.742733469739139e-06,
+      "loss": 0.2589,
+      "step": 8802
+    },
+    {
+      "epoch": 6.626270229582236,
+      "grad_norm": 0.16186392307281494,
+      "learning_rate": 6.739863352922415e-06,
+      "loss": 0.1129,
+      "step": 8803
+    },
+    {
+      "epoch": 6.62702295822356,
+      "grad_norm": 0.17682614922523499,
+      "learning_rate": 6.736995506403026e-06,
+      "loss": 0.1391,
+      "step": 8804
+    },
+    {
+      "epoch": 6.627775686864885,
+      "grad_norm": 0.1814216673374176,
+      "learning_rate": 6.734129930494589e-06,
+      "loss": 0.1634,
+      "step": 8805
+    },
+    {
+      "epoch": 6.62852841550621,
+      "grad_norm": 0.16512463986873627,
+      "learning_rate": 6.731266625510485e-06,
+      "loss": 0.1852,
+      "step": 8806
+    },
+    {
+      "epoch": 6.6292811441475346,
+      "grad_norm": 0.18459227681159973,
+      "learning_rate": 6.72840559176383e-06,
+      "loss": 0.1779,
+      "step": 8807
+    },
+    {
+      "epoch": 6.63003387278886,
+      "grad_norm": 0.19095271825790405,
+      "learning_rate": 6.725546829567515e-06,
+      "loss": 0.165,
+      "step": 8808
+    },
+    {
+      "epoch": 6.630786601430184,
+      "grad_norm": 0.17216555774211884,
+      "learning_rate": 6.722690339234164e-06,
+      "loss": 0.1871,
+      "step": 8809
+    },
+    {
+      "epoch": 6.631539330071509,
+      "grad_norm": 0.16516803205013275,
+      "learning_rate": 6.7198361210761615e-06,
+      "loss": 0.1102,
+      "step": 8810
+    },
+    {
+      "epoch": 6.632292058712834,
+      "grad_norm": 0.15854287147521973,
+      "learning_rate": 6.716984175405636e-06,
+      "loss": 0.1208,
+      "step": 8811
+    },
+    {
+      "epoch": 6.633044787354159,
+      "grad_norm": 0.18389087915420532,
+      "learning_rate": 6.714134502534476e-06,
+      "loss": 0.093,
+      "step": 8812
+    },
+    {
+      "epoch": 6.633797515995484,
+      "grad_norm": 0.19035851955413818,
+      "learning_rate": 6.711287102774308e-06,
+      "loss": 0.2308,
+      "step": 8813
+    },
+    {
+      "epoch": 6.6345502446368085,
+      "grad_norm": 0.18646378815174103,
+      "learning_rate": 6.708441976436536e-06,
+      "loss": 0.1965,
+      "step": 8814
+    },
+    {
+      "epoch": 6.635302973278133,
+      "grad_norm": 0.17890432476997375,
+      "learning_rate": 6.7055991238322845e-06,
+      "loss": 0.1414,
+      "step": 8815
+    },
+    {
+      "epoch": 6.636055701919458,
+      "grad_norm": 0.17183028161525726,
+      "learning_rate": 6.702758545272457e-06,
+      "loss": 0.2198,
+      "step": 8816
+    },
+    {
+      "epoch": 6.636808430560783,
+      "grad_norm": 0.19253118336200714,
+      "learning_rate": 6.699920241067678e-06,
+      "loss": 0.149,
+      "step": 8817
+    },
+    {
+      "epoch": 6.637561159202107,
+      "grad_norm": 0.17340059578418732,
+      "learning_rate": 6.697084211528356e-06,
+      "loss": 0.1701,
+      "step": 8818
+    },
+    {
+      "epoch": 6.638313887843433,
+      "grad_norm": 0.1784570962190628,
+      "learning_rate": 6.694250456964618e-06,
+      "loss": 0.2309,
+      "step": 8819
+    },
+    {
+      "epoch": 6.639066616484757,
+      "grad_norm": 0.18754905462265015,
+      "learning_rate": 6.6914189776863785e-06,
+      "loss": 0.1972,
+      "step": 8820
+    },
+    {
+      "epoch": 6.639819345126082,
+      "grad_norm": 0.15689398348331451,
+      "learning_rate": 6.6885897740032706e-06,
+      "loss": 0.1099,
+      "step": 8821
+    },
+    {
+      "epoch": 6.640572073767407,
+      "grad_norm": 0.17418351769447327,
+      "learning_rate": 6.685762846224697e-06,
+      "loss": 0.2151,
+      "step": 8822
+    },
+    {
+      "epoch": 6.6413248024087315,
+      "grad_norm": 0.18522609770298004,
+      "learning_rate": 6.682938194659802e-06,
+      "loss": 0.2557,
+      "step": 8823
+    },
+    {
+      "epoch": 6.642077531050057,
+      "grad_norm": 0.16826683282852173,
+      "learning_rate": 6.68011581961749e-06,
+      "loss": 0.193,
+      "step": 8824
+    },
+    {
+      "epoch": 6.642830259691381,
+      "grad_norm": 0.16684070229530334,
+      "learning_rate": 6.677295721406411e-06,
+      "loss": 0.1753,
+      "step": 8825
+    },
+    {
+      "epoch": 6.643582988332706,
+      "grad_norm": 0.17024734616279602,
+      "learning_rate": 6.6744779003349635e-06,
+      "loss": 0.1677,
+      "step": 8826
+    },
+    {
+      "epoch": 6.644335716974031,
+      "grad_norm": 0.17440123856067657,
+      "learning_rate": 6.6716623567113e-06,
+      "loss": 0.1568,
+      "step": 8827
+    },
+    {
+      "epoch": 6.645088445615356,
+      "grad_norm": 0.17933225631713867,
+      "learning_rate": 6.668849090843328e-06,
+      "loss": 0.1688,
+      "step": 8828
+    },
+    {
+      "epoch": 6.64584117425668,
+      "grad_norm": 0.1797940731048584,
+      "learning_rate": 6.6660381030387035e-06,
+      "loss": 0.1925,
+      "step": 8829
+    },
+    {
+      "epoch": 6.6465939028980054,
+      "grad_norm": 0.18359167873859406,
+      "learning_rate": 6.663229393604825e-06,
+      "loss": 0.1012,
+      "step": 8830
+    },
+    {
+      "epoch": 6.64734663153933,
+      "grad_norm": 0.1684533953666687,
+      "learning_rate": 6.660422962848857e-06,
+      "loss": 0.1596,
+      "step": 8831
+    },
+    {
+      "epoch": 6.648099360180655,
+      "grad_norm": 0.18051832914352417,
+      "learning_rate": 6.657618811077694e-06,
+      "loss": 0.2082,
+      "step": 8832
+    },
+    {
+      "epoch": 6.64885208882198,
+      "grad_norm": 0.17496368288993835,
+      "learning_rate": 6.654816938598007e-06,
+      "loss": 0.1528,
+      "step": 8833
+    },
+    {
+      "epoch": 6.649604817463304,
+      "grad_norm": 0.17888997495174408,
+      "learning_rate": 6.652017345716198e-06,
+      "loss": 0.1707,
+      "step": 8834
+    },
+    {
+      "epoch": 6.65035754610463,
+      "grad_norm": 0.17364706099033356,
+      "learning_rate": 6.649220032738432e-06,
+      "loss": 0.1447,
+      "step": 8835
+    },
+    {
+      "epoch": 6.651110274745954,
+      "grad_norm": 0.15908293426036835,
+      "learning_rate": 6.6464249999706084e-06,
+      "loss": 0.1707,
+      "step": 8836
+    },
+    {
+      "epoch": 6.6518630033872785,
+      "grad_norm": 0.1818983405828476,
+      "learning_rate": 6.6436322477184e-06,
+      "loss": 0.2162,
+      "step": 8837
+    },
+    {
+      "epoch": 6.652615732028604,
+      "grad_norm": 0.15095414221286774,
+      "learning_rate": 6.640841776287203e-06,
+      "loss": 0.12,
+      "step": 8838
+    },
+    {
+      "epoch": 6.653368460669928,
+      "grad_norm": 0.157008096575737,
+      "learning_rate": 6.6380535859821945e-06,
+      "loss": 0.1216,
+      "step": 8839
+    },
+    {
+      "epoch": 6.654121189311253,
+      "grad_norm": 0.1731209009885788,
+      "learning_rate": 6.63526767710828e-06,
+      "loss": 0.1741,
+      "step": 8840
+    },
+    {
+      "epoch": 6.654873917952578,
+      "grad_norm": 0.15735168755054474,
+      "learning_rate": 6.632484049970123e-06,
+      "loss": 0.1599,
+      "step": 8841
+    },
+    {
+      "epoch": 6.655626646593903,
+      "grad_norm": 0.17047099769115448,
+      "learning_rate": 6.629702704872135e-06,
+      "loss": 0.151,
+      "step": 8842
+    },
+    {
+      "epoch": 6.656379375235228,
+      "grad_norm": 0.18132604658603668,
+      "learning_rate": 6.626923642118482e-06,
+      "loss": 0.2941,
+      "step": 8843
+    },
+    {
+      "epoch": 6.6571321038765525,
+      "grad_norm": 0.18477217853069305,
+      "learning_rate": 6.624146862013077e-06,
+      "loss": 0.1558,
+      "step": 8844
+    },
+    {
+      "epoch": 6.657884832517877,
+      "grad_norm": 0.2114408314228058,
+      "learning_rate": 6.6213723648595825e-06,
+      "loss": 0.1899,
+      "step": 8845
+    },
+    {
+      "epoch": 6.658637561159202,
+      "grad_norm": 0.15264002978801727,
+      "learning_rate": 6.618600150961417e-06,
+      "loss": 0.14,
+      "step": 8846
+    },
+    {
+      "epoch": 6.659390289800527,
+      "grad_norm": 0.17309625446796417,
+      "learning_rate": 6.615830220621747e-06,
+      "loss": 0.1794,
+      "step": 8847
+    },
+    {
+      "epoch": 6.660143018441852,
+      "grad_norm": 0.19279877841472626,
+      "learning_rate": 6.613062574143483e-06,
+      "loss": 0.1391,
+      "step": 8848
+    },
+    {
+      "epoch": 6.660895747083177,
+      "grad_norm": 0.16293732821941376,
+      "learning_rate": 6.610297211829292e-06,
+      "loss": 0.1736,
+      "step": 8849
+    },
+    {
+      "epoch": 6.661648475724501,
+      "grad_norm": 0.15724821388721466,
+      "learning_rate": 6.607534133981594e-06,
+      "loss": 0.1491,
+      "step": 8850
+    },
+    {
+      "epoch": 6.6624012043658265,
+      "grad_norm": 0.14891518652439117,
+      "learning_rate": 6.6047733409025485e-06,
+      "loss": 0.0901,
+      "step": 8851
+    },
+    {
+      "epoch": 6.663153933007151,
+      "grad_norm": 0.17724917829036713,
+      "learning_rate": 6.602014832894074e-06,
+      "loss": 0.2095,
+      "step": 8852
+    },
+    {
+      "epoch": 6.6639066616484754,
+      "grad_norm": 0.1716257631778717,
+      "learning_rate": 6.599258610257839e-06,
+      "loss": 0.1551,
+      "step": 8853
+    },
+    {
+      "epoch": 6.664659390289801,
+      "grad_norm": 0.17563995718955994,
+      "learning_rate": 6.596504673295262e-06,
+      "loss": 0.164,
+      "step": 8854
+    },
+    {
+      "epoch": 6.665412118931125,
+      "grad_norm": 0.1851726770401001,
+      "learning_rate": 6.593753022307501e-06,
+      "loss": 0.2114,
+      "step": 8855
+    },
+    {
+      "epoch": 6.66616484757245,
+      "grad_norm": 0.17496605217456818,
+      "learning_rate": 6.59100365759548e-06,
+      "loss": 0.1102,
+      "step": 8856
+    },
+    {
+      "epoch": 6.666917576213775,
+      "grad_norm": 0.15413746237754822,
+      "learning_rate": 6.588256579459863e-06,
+      "loss": 0.1796,
+      "step": 8857
+    },
+    {
+      "epoch": 6.6676703048551,
+      "grad_norm": 0.1726401150226593,
+      "learning_rate": 6.585511788201064e-06,
+      "loss": 0.0972,
+      "step": 8858
+    },
+    {
+      "epoch": 6.668423033496424,
+      "grad_norm": 0.17310211062431335,
+      "learning_rate": 6.582769284119251e-06,
+      "loss": 0.2646,
+      "step": 8859
+    },
+    {
+      "epoch": 6.669175762137749,
+      "grad_norm": 0.15535874664783478,
+      "learning_rate": 6.580029067514346e-06,
+      "loss": 0.1914,
+      "step": 8860
+    },
+    {
+      "epoch": 6.669928490779074,
+      "grad_norm": 0.21656808257102966,
+      "learning_rate": 6.577291138686005e-06,
+      "loss": 0.2148,
+      "step": 8861
+    },
+    {
+      "epoch": 6.670681219420399,
+      "grad_norm": 0.1750870794057846,
+      "learning_rate": 6.57455549793365e-06,
+      "loss": 0.1824,
+      "step": 8862
+    },
+    {
+      "epoch": 6.671433948061724,
+      "grad_norm": 0.15777452290058136,
+      "learning_rate": 6.571822145556444e-06,
+      "loss": 0.184,
+      "step": 8863
+    },
+    {
+      "epoch": 6.672186676703048,
+      "grad_norm": 0.18802887201309204,
+      "learning_rate": 6.569091081853302e-06,
+      "loss": 0.1919,
+      "step": 8864
+    },
+    {
+      "epoch": 6.672939405344374,
+      "grad_norm": 0.1871045082807541,
+      "learning_rate": 6.566362307122889e-06,
+      "loss": 0.1457,
+      "step": 8865
+    },
+    {
+      "epoch": 6.673692133985698,
+      "grad_norm": 0.15714317560195923,
+      "learning_rate": 6.5636358216636244e-06,
+      "loss": 0.1339,
+      "step": 8866
+    },
+    {
+      "epoch": 6.674444862627023,
+      "grad_norm": 0.16461952030658722,
+      "learning_rate": 6.5609116257736675e-06,
+      "loss": 0.2153,
+      "step": 8867
+    },
+    {
+      "epoch": 6.675197591268348,
+      "grad_norm": 0.18650086224079132,
+      "learning_rate": 6.558189719750938e-06,
+      "loss": 0.1358,
+      "step": 8868
+    },
+    {
+      "epoch": 6.675950319909672,
+      "grad_norm": 0.15508130192756653,
+      "learning_rate": 6.555470103893089e-06,
+      "loss": 0.1421,
+      "step": 8869
+    },
+    {
+      "epoch": 6.676703048550998,
+      "grad_norm": 0.1880018413066864,
+      "learning_rate": 6.552752778497541e-06,
+      "loss": 0.1048,
+      "step": 8870
+    },
+    {
+      "epoch": 6.677455777192322,
+      "grad_norm": 0.14912304282188416,
+      "learning_rate": 6.550037743861455e-06,
+      "loss": 0.1561,
+      "step": 8871
+    },
+    {
+      "epoch": 6.678208505833647,
+      "grad_norm": 0.1854422241449356,
+      "learning_rate": 6.547325000281749e-06,
+      "loss": 0.1933,
+      "step": 8872
+    },
+    {
+      "epoch": 6.678961234474972,
+      "grad_norm": 0.19706428050994873,
+      "learning_rate": 6.544614548055071e-06,
+      "loss": 0.2092,
+      "step": 8873
+    },
+    {
+      "epoch": 6.6797139631162965,
+      "grad_norm": 0.1808384209871292,
+      "learning_rate": 6.541906387477845e-06,
+      "loss": 0.1518,
+      "step": 8874
+    },
+    {
+      "epoch": 6.680466691757621,
+      "grad_norm": 0.15561263263225555,
+      "learning_rate": 6.539200518846225e-06,
+      "loss": 0.1869,
+      "step": 8875
+    },
+    {
+      "epoch": 6.681219420398946,
+      "grad_norm": 0.14604097604751587,
+      "learning_rate": 6.5364969424561215e-06,
+      "loss": 0.0974,
+      "step": 8876
+    },
+    {
+      "epoch": 6.681972149040271,
+      "grad_norm": 0.15590263903141022,
+      "learning_rate": 6.533795658603195e-06,
+      "loss": 0.1542,
+      "step": 8877
+    },
+    {
+      "epoch": 6.682724877681595,
+      "grad_norm": 0.17231369018554688,
+      "learning_rate": 6.531096667582852e-06,
+      "loss": 0.1459,
+      "step": 8878
+    },
+    {
+      "epoch": 6.683477606322921,
+      "grad_norm": 0.1637333780527115,
+      "learning_rate": 6.528399969690255e-06,
+      "loss": 0.1516,
+      "step": 8879
+    },
+    {
+      "epoch": 6.684230334964245,
+      "grad_norm": 0.17837470769882202,
+      "learning_rate": 6.525705565220304e-06,
+      "loss": 0.115,
+      "step": 8880
+    },
+    {
+      "epoch": 6.6849830636055705,
+      "grad_norm": 0.1681775450706482,
+      "learning_rate": 6.52301345446766e-06,
+      "loss": 0.1609,
+      "step": 8881
+    },
+    {
+      "epoch": 6.685735792246895,
+      "grad_norm": 0.15581563115119934,
+      "learning_rate": 6.520323637726725e-06,
+      "loss": 0.1271,
+      "step": 8882
+    },
+    {
+      "epoch": 6.686488520888219,
+      "grad_norm": 0.1359677016735077,
+      "learning_rate": 6.517636115291653e-06,
+      "loss": 0.0731,
+      "step": 8883
+    },
+    {
+      "epoch": 6.687241249529545,
+      "grad_norm": 0.20562252402305603,
+      "learning_rate": 6.514950887456346e-06,
+      "loss": 0.1732,
+      "step": 8884
+    },
+    {
+      "epoch": 6.687993978170869,
+      "grad_norm": 0.17797361314296722,
+      "learning_rate": 6.512267954514468e-06,
+      "loss": 0.1409,
+      "step": 8885
+    },
+    {
+      "epoch": 6.688746706812195,
+      "grad_norm": 0.14667481184005737,
+      "learning_rate": 6.509587316759405e-06,
+      "loss": 0.1083,
+      "step": 8886
+    },
+    {
+      "epoch": 6.689499435453519,
+      "grad_norm": 0.1931498795747757,
+      "learning_rate": 6.506908974484322e-06,
+      "loss": 0.205,
+      "step": 8887
+    },
+    {
+      "epoch": 6.690252164094844,
+      "grad_norm": 0.18147379159927368,
+      "learning_rate": 6.504232927982104e-06,
+      "loss": 0.1561,
+      "step": 8888
+    },
+    {
+      "epoch": 6.691004892736169,
+      "grad_norm": 0.1820165514945984,
+      "learning_rate": 6.501559177545406e-06,
+      "loss": 0.1116,
+      "step": 8889
+    },
+    {
+      "epoch": 6.691757621377493,
+      "grad_norm": 0.1931135207414627,
+      "learning_rate": 6.498887723466626e-06,
+      "loss": 0.1381,
+      "step": 8890
+    },
+    {
+      "epoch": 6.692510350018818,
+      "grad_norm": 0.15802064538002014,
+      "learning_rate": 6.496218566037915e-06,
+      "loss": 0.1646,
+      "step": 8891
+    },
+    {
+      "epoch": 6.693263078660143,
+      "grad_norm": 0.17901824414730072,
+      "learning_rate": 6.493551705551158e-06,
+      "loss": 0.1693,
+      "step": 8892
+    },
+    {
+      "epoch": 6.694015807301468,
+      "grad_norm": 0.16911356151103973,
+      "learning_rate": 6.4908871422980075e-06,
+      "loss": 0.1041,
+      "step": 8893
+    },
+    {
+      "epoch": 6.694768535942792,
+      "grad_norm": 0.17957356572151184,
+      "learning_rate": 6.48822487656985e-06,
+      "loss": 0.2104,
+      "step": 8894
+    },
+    {
+      "epoch": 6.695521264584118,
+      "grad_norm": 0.19025671482086182,
+      "learning_rate": 6.4855649086578265e-06,
+      "loss": 0.2007,
+      "step": 8895
+    },
+    {
+      "epoch": 6.696273993225442,
+      "grad_norm": 0.16648079454898834,
+      "learning_rate": 6.482907238852832e-06,
+      "loss": 0.1303,
+      "step": 8896
+    },
+    {
+      "epoch": 6.697026721866767,
+      "grad_norm": 0.1692754626274109,
+      "learning_rate": 6.480251867445507e-06,
+      "loss": 0.15,
+      "step": 8897
+    },
+    {
+      "epoch": 6.697779450508092,
+      "grad_norm": 0.17757202684879303,
+      "learning_rate": 6.477598794726231e-06,
+      "loss": 0.1703,
+      "step": 8898
+    },
+    {
+      "epoch": 6.698532179149416,
+      "grad_norm": 0.16882288455963135,
+      "learning_rate": 6.474948020985144e-06,
+      "loss": 0.1451,
+      "step": 8899
+    },
+    {
+      "epoch": 6.699284907790742,
+      "grad_norm": 0.16692180931568146,
+      "learning_rate": 6.472299546512134e-06,
+      "loss": 0.1046,
+      "step": 8900
+    },
+    {
+      "epoch": 6.700037636432066,
+      "grad_norm": 0.14751961827278137,
+      "learning_rate": 6.4696533715968275e-06,
+      "loss": 0.1387,
+      "step": 8901
+    },
+    {
+      "epoch": 6.7007903650733915,
+      "grad_norm": 0.2038920372724533,
+      "learning_rate": 6.46700949652861e-06,
+      "loss": 0.2565,
+      "step": 8902
+    },
+    {
+      "epoch": 6.701543093714716,
+      "grad_norm": 0.18232207000255585,
+      "learning_rate": 6.4643679215966105e-06,
+      "loss": 0.1311,
+      "step": 8903
+    },
+    {
+      "epoch": 6.7022958223560405,
+      "grad_norm": 0.16707181930541992,
+      "learning_rate": 6.461728647089712e-06,
+      "loss": 0.1566,
+      "step": 8904
+    },
+    {
+      "epoch": 6.703048550997366,
+      "grad_norm": 0.193992018699646,
+      "learning_rate": 6.4590916732965355e-06,
+      "loss": 0.1936,
+      "step": 8905
+    },
+    {
+      "epoch": 6.70380127963869,
+      "grad_norm": 0.18169742822647095,
+      "learning_rate": 6.4564570005054625e-06,
+      "loss": 0.1625,
+      "step": 8906
+    },
+    {
+      "epoch": 6.704554008280015,
+      "grad_norm": 0.16747339069843292,
+      "learning_rate": 6.453824629004608e-06,
+      "loss": 0.1923,
+      "step": 8907
+    },
+    {
+      "epoch": 6.70530673692134,
+      "grad_norm": 0.19847923517227173,
+      "learning_rate": 6.451194559081857e-06,
+      "loss": 0.1817,
+      "step": 8908
+    },
+    {
+      "epoch": 6.706059465562665,
+      "grad_norm": 0.15201063454151154,
+      "learning_rate": 6.448566791024811e-06,
+      "loss": 0.1469,
+      "step": 8909
+    },
+    {
+      "epoch": 6.706812194203989,
+      "grad_norm": 0.1782679706811905,
+      "learning_rate": 6.445941325120861e-06,
+      "loss": 0.1922,
+      "step": 8910
+    },
+    {
+      "epoch": 6.7075649228453145,
+      "grad_norm": 0.18173189461231232,
+      "learning_rate": 6.4433181616571094e-06,
+      "loss": 0.1099,
+      "step": 8911
+    },
+    {
+      "epoch": 6.708317651486639,
+      "grad_norm": 0.16770732402801514,
+      "learning_rate": 6.4406973009204275e-06,
+      "loss": 0.1472,
+      "step": 8912
+    },
+    {
+      "epoch": 6.709070380127963,
+      "grad_norm": 0.18340982496738434,
+      "learning_rate": 6.438078743197423e-06,
+      "loss": 0.2041,
+      "step": 8913
+    },
+    {
+      "epoch": 6.709823108769289,
+      "grad_norm": 0.18370947241783142,
+      "learning_rate": 6.435462488774464e-06,
+      "loss": 0.1315,
+      "step": 8914
+    },
+    {
+      "epoch": 6.710575837410613,
+      "grad_norm": 0.17949846386909485,
+      "learning_rate": 6.432848537937651e-06,
+      "loss": 0.1495,
+      "step": 8915
+    },
+    {
+      "epoch": 6.711328566051939,
+      "grad_norm": 0.15173503756523132,
+      "learning_rate": 6.430236890972857e-06,
+      "loss": 0.0733,
+      "step": 8916
+    },
+    {
+      "epoch": 6.712081294693263,
+      "grad_norm": 0.17629142105579376,
+      "learning_rate": 6.427627548165674e-06,
+      "loss": 0.1787,
+      "step": 8917
+    },
+    {
+      "epoch": 6.712834023334588,
+      "grad_norm": 0.17092068493366241,
+      "learning_rate": 6.4250205098014635e-06,
+      "loss": 0.1211,
+      "step": 8918
+    },
+    {
+      "epoch": 6.713586751975913,
+      "grad_norm": 0.15409792959690094,
+      "learning_rate": 6.422415776165321e-06,
+      "loss": 0.1366,
+      "step": 8919
+    },
+    {
+      "epoch": 6.714339480617237,
+      "grad_norm": 0.16228975355625153,
+      "learning_rate": 6.419813347542099e-06,
+      "loss": 0.0991,
+      "step": 8920
+    },
+    {
+      "epoch": 6.715092209258563,
+      "grad_norm": 0.17863374948501587,
+      "learning_rate": 6.417213224216394e-06,
+      "loss": 0.1382,
+      "step": 8921
+    },
+    {
+      "epoch": 6.715844937899887,
+      "grad_norm": 0.1755044311285019,
+      "learning_rate": 6.41461540647256e-06,
+      "loss": 0.1674,
+      "step": 8922
+    },
+    {
+      "epoch": 6.716597666541212,
+      "grad_norm": 0.16387133300304413,
+      "learning_rate": 6.41201989459468e-06,
+      "loss": 0.1685,
+      "step": 8923
+    },
+    {
+      "epoch": 6.717350395182537,
+      "grad_norm": 0.1846618950366974,
+      "learning_rate": 6.409426688866597e-06,
+      "loss": 0.1737,
+      "step": 8924
+    },
+    {
+      "epoch": 6.7181031238238615,
+      "grad_norm": 0.17331255972385406,
+      "learning_rate": 6.406835789571905e-06,
+      "loss": 0.1327,
+      "step": 8925
+    },
+    {
+      "epoch": 6.718855852465186,
+      "grad_norm": 0.1929626166820526,
+      "learning_rate": 6.404247196993937e-06,
+      "loss": 0.2247,
+      "step": 8926
+    },
+    {
+      "epoch": 6.719608581106511,
+      "grad_norm": 0.1806914508342743,
+      "learning_rate": 6.401660911415782e-06,
+      "loss": 0.1154,
+      "step": 8927
+    },
+    {
+      "epoch": 6.720361309747836,
+      "grad_norm": 0.1577715426683426,
+      "learning_rate": 6.3990769331202595e-06,
+      "loss": 0.1869,
+      "step": 8928
+    },
+    {
+      "epoch": 6.72111403838916,
+      "grad_norm": 0.16789941489696503,
+      "learning_rate": 6.396495262389968e-06,
+      "loss": 0.0669,
+      "step": 8929
+    },
+    {
+      "epoch": 6.721866767030486,
+      "grad_norm": 0.1669304072856903,
+      "learning_rate": 6.393915899507223e-06,
+      "loss": 0.184,
+      "step": 8930
+    },
+    {
+      "epoch": 6.72261949567181,
+      "grad_norm": 0.17814335227012634,
+      "learning_rate": 6.391338844754103e-06,
+      "loss": 0.2209,
+      "step": 8931
+    },
+    {
+      "epoch": 6.723372224313135,
+      "grad_norm": 0.16131211817264557,
+      "learning_rate": 6.388764098412431e-06,
+      "loss": 0.1572,
+      "step": 8932
+    },
+    {
+      "epoch": 6.72412495295446,
+      "grad_norm": 0.196302592754364,
+      "learning_rate": 6.386191660763778e-06,
+      "loss": 0.1174,
+      "step": 8933
+    },
+    {
+      "epoch": 6.7248776815957845,
+      "grad_norm": 0.17246825993061066,
+      "learning_rate": 6.383621532089453e-06,
+      "loss": 0.183,
+      "step": 8934
+    },
+    {
+      "epoch": 6.72563041023711,
+      "grad_norm": 0.1584896594285965,
+      "learning_rate": 6.381053712670539e-06,
+      "loss": 0.1751,
+      "step": 8935
+    },
+    {
+      "epoch": 6.726383138878434,
+      "grad_norm": 0.18267378211021423,
+      "learning_rate": 6.378488202787834e-06,
+      "loss": 0.1758,
+      "step": 8936
+    },
+    {
+      "epoch": 6.727135867519759,
+      "grad_norm": 0.16658000648021698,
+      "learning_rate": 6.375925002721907e-06,
+      "loss": 0.1564,
+      "step": 8937
+    },
+    {
+      "epoch": 6.727888596161084,
+      "grad_norm": 0.18890772759914398,
+      "learning_rate": 6.373364112753058e-06,
+      "loss": 0.1538,
+      "step": 8938
+    },
+    {
+      "epoch": 6.728641324802409,
+      "grad_norm": 0.18067742884159088,
+      "learning_rate": 6.37080553316135e-06,
+      "loss": 0.134,
+      "step": 8939
+    },
+    {
+      "epoch": 6.729394053443734,
+      "grad_norm": 0.16189102828502655,
+      "learning_rate": 6.3682492642265765e-06,
+      "loss": 0.195,
+      "step": 8940
+    },
+    {
+      "epoch": 6.7301467820850585,
+      "grad_norm": 0.1745390146970749,
+      "learning_rate": 6.365695306228296e-06,
+      "loss": 0.2079,
+      "step": 8941
+    },
+    {
+      "epoch": 6.730899510726383,
+      "grad_norm": 0.17288683354854584,
+      "learning_rate": 6.363143659445801e-06,
+      "loss": 0.1821,
+      "step": 8942
+    },
+    {
+      "epoch": 6.731652239367708,
+      "grad_norm": 0.1607411652803421,
+      "learning_rate": 6.360594324158139e-06,
+      "loss": 0.1017,
+      "step": 8943
+    },
+    {
+      "epoch": 6.732404968009033,
+      "grad_norm": 0.1780715137720108,
+      "learning_rate": 6.358047300644096e-06,
+      "loss": 0.1793,
+      "step": 8944
+    },
+    {
+      "epoch": 6.733157696650357,
+      "grad_norm": 0.167287215590477,
+      "learning_rate": 6.355502589182213e-06,
+      "loss": 0.1507,
+      "step": 8945
+    },
+    {
+      "epoch": 6.733910425291683,
+      "grad_norm": 0.18762217462062836,
+      "learning_rate": 6.352960190050781e-06,
+      "loss": 0.1979,
+      "step": 8946
+    },
+    {
+      "epoch": 6.734663153933007,
+      "grad_norm": 0.1876154989004135,
+      "learning_rate": 6.350420103527827e-06,
+      "loss": 0.1618,
+      "step": 8947
+    },
+    {
+      "epoch": 6.7354158825743315,
+      "grad_norm": 0.1690579354763031,
+      "learning_rate": 6.347882329891129e-06,
+      "loss": 0.1563,
+      "step": 8948
+    },
+    {
+      "epoch": 6.736168611215657,
+      "grad_norm": 0.1763400286436081,
+      "learning_rate": 6.34534686941822e-06,
+      "loss": 0.2339,
+      "step": 8949
+    },
+    {
+      "epoch": 6.736921339856981,
+      "grad_norm": 0.19069531559944153,
+      "learning_rate": 6.342813722386374e-06,
+      "loss": 0.1679,
+      "step": 8950
+    },
+    {
+      "epoch": 6.737674068498307,
+      "grad_norm": 0.1782960742712021,
+      "learning_rate": 6.340282889072606e-06,
+      "loss": 0.0839,
+      "step": 8951
+    },
+    {
+      "epoch": 6.738426797139631,
+      "grad_norm": 0.17068339884281158,
+      "learning_rate": 6.337754369753693e-06,
+      "loss": 0.2044,
+      "step": 8952
+    },
+    {
+      "epoch": 6.739179525780956,
+      "grad_norm": 0.1905069202184677,
+      "learning_rate": 6.3352281647061375e-06,
+      "loss": 0.2374,
+      "step": 8953
+    },
+    {
+      "epoch": 6.739932254422281,
+      "grad_norm": 0.17546379566192627,
+      "learning_rate": 6.332704274206215e-06,
+      "loss": 0.2785,
+      "step": 8954
+    },
+    {
+      "epoch": 6.7406849830636055,
+      "grad_norm": 0.18623045086860657,
+      "learning_rate": 6.330182698529928e-06,
+      "loss": 0.2133,
+      "step": 8955
+    },
+    {
+      "epoch": 6.74143771170493,
+      "grad_norm": 0.18739080429077148,
+      "learning_rate": 6.327663437953032e-06,
+      "loss": 0.1871,
+      "step": 8956
+    },
+    {
+      "epoch": 6.742190440346255,
+      "grad_norm": 0.1539396047592163,
+      "learning_rate": 6.32514649275103e-06,
+      "loss": 0.1626,
+      "step": 8957
+    },
+    {
+      "epoch": 6.74294316898758,
+      "grad_norm": 0.1487175077199936,
+      "learning_rate": 6.322631863199174e-06,
+      "loss": 0.1854,
+      "step": 8958
+    },
+    {
+      "epoch": 6.743695897628905,
+      "grad_norm": 0.17263482511043549,
+      "learning_rate": 6.320119549572452e-06,
+      "loss": 0.1711,
+      "step": 8959
+    },
+    {
+      "epoch": 6.74444862627023,
+      "grad_norm": 0.20374462008476257,
+      "learning_rate": 6.317609552145619e-06,
+      "loss": 0.1898,
+      "step": 8960
+    },
+    {
+      "epoch": 6.745201354911554,
+      "grad_norm": 0.1674741804599762,
+      "learning_rate": 6.315101871193154e-06,
+      "loss": 0.2129,
+      "step": 8961
+    },
+    {
+      "epoch": 6.7459540835528795,
+      "grad_norm": 0.17240725457668304,
+      "learning_rate": 6.3125965069893e-06,
+      "loss": 0.229,
+      "step": 8962
+    },
+    {
+      "epoch": 6.746706812194204,
+      "grad_norm": 0.17466269433498383,
+      "learning_rate": 6.310093459808038e-06,
+      "loss": 0.1417,
+      "step": 8963
+    },
+    {
+      "epoch": 6.7474595408355285,
+      "grad_norm": 0.17193035781383514,
+      "learning_rate": 6.307592729923095e-06,
+      "loss": 0.2009,
+      "step": 8964
+    },
+    {
+      "epoch": 6.748212269476854,
+      "grad_norm": 0.1777704507112503,
+      "learning_rate": 6.305094317607948e-06,
+      "loss": 0.1616,
+      "step": 8965
+    },
+    {
+      "epoch": 6.748964998118178,
+      "grad_norm": 0.1912297010421753,
+      "learning_rate": 6.302598223135818e-06,
+      "loss": 0.2483,
+      "step": 8966
+    },
+    {
+      "epoch": 6.749717726759503,
+      "grad_norm": 0.18605847656726837,
+      "learning_rate": 6.30010444677968e-06,
+      "loss": 0.1521,
+      "step": 8967
+    },
+    {
+      "epoch": 6.750470455400828,
+      "grad_norm": 0.1554461270570755,
+      "learning_rate": 6.297612988812246e-06,
+      "loss": 0.1754,
+      "step": 8968
+    },
+    {
+      "epoch": 6.751223184042153,
+      "grad_norm": 0.18369971215724945,
+      "learning_rate": 6.2951238495059765e-06,
+      "loss": 0.181,
+      "step": 8969
+    },
+    {
+      "epoch": 6.751975912683478,
+      "grad_norm": 0.19398736953735352,
+      "learning_rate": 6.292637029133083e-06,
+      "loss": 0.2106,
+      "step": 8970
+    },
+    {
+      "epoch": 6.752728641324802,
+      "grad_norm": 0.1626664102077484,
+      "learning_rate": 6.290152527965522e-06,
+      "loss": 0.1212,
+      "step": 8971
+    },
+    {
+      "epoch": 6.753481369966127,
+      "grad_norm": 0.1716885268688202,
+      "learning_rate": 6.287670346274987e-06,
+      "loss": 0.1849,
+      "step": 8972
+    },
+    {
+      "epoch": 6.754234098607452,
+      "grad_norm": 0.1788913458585739,
+      "learning_rate": 6.285190484332932e-06,
+      "loss": 0.1367,
+      "step": 8973
+    },
+    {
+      "epoch": 6.754986827248777,
+      "grad_norm": 0.18723240494728088,
+      "learning_rate": 6.282712942410548e-06,
+      "loss": 0.1405,
+      "step": 8974
+    },
+    {
+      "epoch": 6.755739555890102,
+      "grad_norm": 0.1731182187795639,
+      "learning_rate": 6.2802377207787825e-06,
+      "loss": 0.1474,
+      "step": 8975
+    },
+    {
+      "epoch": 6.756492284531427,
+      "grad_norm": 0.17391161620616913,
+      "learning_rate": 6.277764819708313e-06,
+      "loss": 0.2017,
+      "step": 8976
+    },
+    {
+      "epoch": 6.757245013172751,
+      "grad_norm": 0.18209438025951385,
+      "learning_rate": 6.275294239469577e-06,
+      "loss": 0.1483,
+      "step": 8977
+    },
+    {
+      "epoch": 6.757997741814076,
+      "grad_norm": 0.1594463735818863,
+      "learning_rate": 6.272825980332751e-06,
+      "loss": 0.1866,
+      "step": 8978
+    },
+    {
+      "epoch": 6.758750470455401,
+      "grad_norm": 0.18643637001514435,
+      "learning_rate": 6.270360042567759e-06,
+      "loss": 0.142,
+      "step": 8979
+    },
+    {
+      "epoch": 6.759503199096725,
+      "grad_norm": 0.18887215852737427,
+      "learning_rate": 6.2678964264442765e-06,
+      "loss": 0.1989,
+      "step": 8980
+    },
+    {
+      "epoch": 6.760255927738051,
+      "grad_norm": 0.15648576617240906,
+      "learning_rate": 6.265435132231721e-06,
+      "loss": 0.1423,
+      "step": 8981
+    },
+    {
+      "epoch": 6.761008656379375,
+      "grad_norm": 0.17540343105793,
+      "learning_rate": 6.262976160199251e-06,
+      "loss": 0.2271,
+      "step": 8982
+    },
+    {
+      "epoch": 6.7617613850207,
+      "grad_norm": 0.1608407348394394,
+      "learning_rate": 6.260519510615781e-06,
+      "loss": 0.1637,
+      "step": 8983
+    },
+    {
+      "epoch": 6.762514113662025,
+      "grad_norm": 0.16721844673156738,
+      "learning_rate": 6.258065183749961e-06,
+      "loss": 0.1686,
+      "step": 8984
+    },
+    {
+      "epoch": 6.7632668423033495,
+      "grad_norm": 0.16745583713054657,
+      "learning_rate": 6.255613179870198e-06,
+      "loss": 0.1505,
+      "step": 8985
+    },
+    {
+      "epoch": 6.764019570944674,
+      "grad_norm": 0.15655948221683502,
+      "learning_rate": 6.253163499244634e-06,
+      "loss": 0.173,
+      "step": 8986
+    },
+    {
+      "epoch": 6.764772299585999,
+      "grad_norm": 0.18594081699848175,
+      "learning_rate": 6.250716142141172e-06,
+      "loss": 0.1447,
+      "step": 8987
+    },
+    {
+      "epoch": 6.765525028227324,
+      "grad_norm": 0.17839574813842773,
+      "learning_rate": 6.248271108827438e-06,
+      "loss": 0.1818,
+      "step": 8988
+    },
+    {
+      "epoch": 6.766277756868649,
+      "grad_norm": 0.1645810306072235,
+      "learning_rate": 6.245828399570831e-06,
+      "loss": 0.1447,
+      "step": 8989
+    },
+    {
+      "epoch": 6.767030485509974,
+      "grad_norm": 0.2065805345773697,
+      "learning_rate": 6.24338801463847e-06,
+      "loss": 0.1921,
+      "step": 8990
+    },
+    {
+      "epoch": 6.767783214151298,
+      "grad_norm": 0.18150584399700165,
+      "learning_rate": 6.240949954297238e-06,
+      "loss": 0.2041,
+      "step": 8991
+    },
+    {
+      "epoch": 6.7685359427926235,
+      "grad_norm": 0.17029710114002228,
+      "learning_rate": 6.238514218813755e-06,
+      "loss": 0.1298,
+      "step": 8992
+    },
+    {
+      "epoch": 6.769288671433948,
+      "grad_norm": 0.17308031022548676,
+      "learning_rate": 6.236080808454395e-06,
+      "loss": 0.1555,
+      "step": 8993
+    },
+    {
+      "epoch": 6.770041400075273,
+      "grad_norm": 0.1695803999900818,
+      "learning_rate": 6.2336497234852635e-06,
+      "loss": 0.1646,
+      "step": 8994
+    },
+    {
+      "epoch": 6.770794128716598,
+      "grad_norm": 0.18917560577392578,
+      "learning_rate": 6.231220964172226e-06,
+      "loss": 0.1811,
+      "step": 8995
+    },
+    {
+      "epoch": 6.771546857357922,
+      "grad_norm": 0.17721015214920044,
+      "learning_rate": 6.22879453078089e-06,
+      "loss": 0.1775,
+      "step": 8996
+    },
+    {
+      "epoch": 6.772299585999248,
+      "grad_norm": 0.1550753116607666,
+      "learning_rate": 6.226370423576598e-06,
+      "loss": 0.0867,
+      "step": 8997
+    },
+    {
+      "epoch": 6.773052314640572,
+      "grad_norm": 0.14397944509983063,
+      "learning_rate": 6.2239486428244535e-06,
+      "loss": 0.1094,
+      "step": 8998
+    },
+    {
+      "epoch": 6.773805043281897,
+      "grad_norm": 0.13827694952487946,
+      "learning_rate": 6.2215291887892976e-06,
+      "loss": 0.0989,
+      "step": 8999
+    },
+    {
+      "epoch": 6.774557771923222,
+      "grad_norm": 0.17225022614002228,
+      "learning_rate": 6.219112061735721e-06,
+      "loss": 0.1663,
+      "step": 9000
+    },
+    {
+      "epoch": 6.774557771923222,
+      "eval_loss": 0.20296046137809753,
+      "eval_runtime": 453.0206,
+      "eval_samples_per_second": 21.251,
+      "eval_steps_per_second": 0.664,
+      "step": 9000
+    },
+    {
+      "epoch": 6.775310500564546,
+      "grad_norm": 0.17923004925251007,
+      "learning_rate": 6.216697261928051e-06,
+      "loss": 0.1883,
+      "step": 9001
+    },
+    {
+      "epoch": 6.776063229205871,
+      "grad_norm": 0.18431873619556427,
+      "learning_rate": 6.214284789630373e-06,
+      "loss": 0.1772,
+      "step": 9002
+    },
+    {
+      "epoch": 6.776815957847196,
+      "grad_norm": 0.16281960904598236,
+      "learning_rate": 6.211874645106503e-06,
+      "loss": 0.1412,
+      "step": 9003
+    },
+    {
+      "epoch": 6.777568686488521,
+      "grad_norm": 0.17614658176898956,
+      "learning_rate": 6.209466828620018e-06,
+      "loss": 0.0762,
+      "step": 9004
+    },
+    {
+      "epoch": 6.778321415129845,
+      "grad_norm": 0.19096463918685913,
+      "learning_rate": 6.207061340434229e-06,
+      "loss": 0.252,
+      "step": 9005
+    },
+    {
+      "epoch": 6.779074143771171,
+      "grad_norm": 0.16211050748825073,
+      "learning_rate": 6.204658180812202e-06,
+      "loss": 0.1468,
+      "step": 9006
+    },
+    {
+      "epoch": 6.779826872412495,
+      "grad_norm": 0.17037135362625122,
+      "learning_rate": 6.202257350016736e-06,
+      "loss": 0.2328,
+      "step": 9007
+    },
+    {
+      "epoch": 6.78057960105382,
+      "grad_norm": 0.178330659866333,
+      "learning_rate": 6.19985884831039e-06,
+      "loss": 0.1643,
+      "step": 9008
+    },
+    {
+      "epoch": 6.781332329695145,
+      "grad_norm": 0.16516248881816864,
+      "learning_rate": 6.197462675955452e-06,
+      "loss": 0.1356,
+      "step": 9009
+    },
+    {
+      "epoch": 6.782085058336469,
+      "grad_norm": 0.17355838418006897,
+      "learning_rate": 6.1950688332139666e-06,
+      "loss": 0.1549,
+      "step": 9010
+    },
+    {
+      "epoch": 6.782837786977795,
+      "grad_norm": 0.14487875998020172,
+      "learning_rate": 6.192677320347721e-06,
+      "loss": 0.1036,
+      "step": 9011
+    },
+    {
+      "epoch": 6.783590515619119,
+      "grad_norm": 0.18972843885421753,
+      "learning_rate": 6.190288137618256e-06,
+      "loss": 0.2031,
+      "step": 9012
+    },
+    {
+      "epoch": 6.7843432442604445,
+      "grad_norm": 0.1692441701889038,
+      "learning_rate": 6.187901285286836e-06,
+      "loss": 0.1917,
+      "step": 9013
+    },
+    {
+      "epoch": 6.785095972901769,
+      "grad_norm": 0.18434494733810425,
+      "learning_rate": 6.185516763614492e-06,
+      "loss": 0.1707,
+      "step": 9014
+    },
+    {
+      "epoch": 6.7858487015430935,
+      "grad_norm": 0.18103021383285522,
+      "learning_rate": 6.183134572861985e-06,
+      "loss": 0.0865,
+      "step": 9015
+    },
+    {
+      "epoch": 6.786601430184419,
+      "grad_norm": 0.18425416946411133,
+      "learning_rate": 6.180754713289835e-06,
+      "loss": 0.2043,
+      "step": 9016
+    },
+    {
+      "epoch": 6.787354158825743,
+      "grad_norm": 0.1902814656496048,
+      "learning_rate": 6.178377185158293e-06,
+      "loss": 0.2034,
+      "step": 9017
+    },
+    {
+      "epoch": 6.788106887467068,
+      "grad_norm": 0.18200334906578064,
+      "learning_rate": 6.176001988727372e-06,
+      "loss": 0.2088,
+      "step": 9018
+    },
+    {
+      "epoch": 6.788859616108393,
+      "grad_norm": 0.15432654321193695,
+      "learning_rate": 6.173629124256809e-06,
+      "loss": 0.1204,
+      "step": 9019
+    },
+    {
+      "epoch": 6.789612344749718,
+      "grad_norm": 0.1695530116558075,
+      "learning_rate": 6.171258592006099e-06,
+      "loss": 0.1584,
+      "step": 9020
+    },
+    {
+      "epoch": 6.790365073391042,
+      "grad_norm": 0.17367853224277496,
+      "learning_rate": 6.168890392234489e-06,
+      "loss": 0.1684,
+      "step": 9021
+    },
+    {
+      "epoch": 6.7911178020323675,
+      "grad_norm": 0.1636940836906433,
+      "learning_rate": 6.1665245252009485e-06,
+      "loss": 0.1481,
+      "step": 9022
+    },
+    {
+      "epoch": 6.791870530673692,
+      "grad_norm": 0.1792275756597519,
+      "learning_rate": 6.1641609911642116e-06,
+      "loss": 0.1375,
+      "step": 9023
+    },
+    {
+      "epoch": 6.792623259315017,
+      "grad_norm": 0.1610170602798462,
+      "learning_rate": 6.161799790382752e-06,
+      "loss": 0.1775,
+      "step": 9024
+    },
+    {
+      "epoch": 6.793375987956342,
+      "grad_norm": 0.1757025271654129,
+      "learning_rate": 6.159440923114789e-06,
+      "loss": 0.1847,
+      "step": 9025
+    },
+    {
+      "epoch": 6.794128716597666,
+      "grad_norm": 0.1533709019422531,
+      "learning_rate": 6.157084389618277e-06,
+      "loss": 0.0865,
+      "step": 9026
+    },
+    {
+      "epoch": 6.794881445238992,
+      "grad_norm": 0.16795764863491058,
+      "learning_rate": 6.154730190150933e-06,
+      "loss": 0.0876,
+      "step": 9027
+    },
+    {
+      "epoch": 6.795634173880316,
+      "grad_norm": 0.14703668653964996,
+      "learning_rate": 6.152378324970198e-06,
+      "loss": 0.1185,
+      "step": 9028
+    },
+    {
+      "epoch": 6.796386902521641,
+      "grad_norm": 0.19223375618457794,
+      "learning_rate": 6.150028794333276e-06,
+      "loss": 0.1714,
+      "step": 9029
+    },
+    {
+      "epoch": 6.797139631162966,
+      "grad_norm": 0.18021082878112793,
+      "learning_rate": 6.147681598497104e-06,
+      "loss": 0.2418,
+      "step": 9030
+    },
+    {
+      "epoch": 6.79789235980429,
+      "grad_norm": 0.1874581277370453,
+      "learning_rate": 6.145336737718375e-06,
+      "loss": 0.1637,
+      "step": 9031
+    },
+    {
+      "epoch": 6.798645088445616,
+      "grad_norm": 0.14787736535072327,
+      "learning_rate": 6.142994212253508e-06,
+      "loss": 0.1106,
+      "step": 9032
+    },
+    {
+      "epoch": 6.79939781708694,
+      "grad_norm": 0.1690092831850052,
+      "learning_rate": 6.140654022358688e-06,
+      "loss": 0.1912,
+      "step": 9033
+    },
+    {
+      "epoch": 6.800150545728265,
+      "grad_norm": 0.18666937947273254,
+      "learning_rate": 6.138316168289828e-06,
+      "loss": 0.21,
+      "step": 9034
+    },
+    {
+      "epoch": 6.80090327436959,
+      "grad_norm": 0.14129406213760376,
+      "learning_rate": 6.135980650302599e-06,
+      "loss": 0.1069,
+      "step": 9035
+    },
+    {
+      "epoch": 6.8016560030109146,
+      "grad_norm": 0.2029493898153305,
+      "learning_rate": 6.1336474686524e-06,
+      "loss": 0.2444,
+      "step": 9036
+    },
+    {
+      "epoch": 6.802408731652239,
+      "grad_norm": 0.18821404874324799,
+      "learning_rate": 6.131316623594393e-06,
+      "loss": 0.1717,
+      "step": 9037
+    },
+    {
+      "epoch": 6.803161460293564,
+      "grad_norm": 0.19430384039878845,
+      "learning_rate": 6.128988115383472e-06,
+      "loss": 0.1817,
+      "step": 9038
+    },
+    {
+      "epoch": 6.803914188934889,
+      "grad_norm": 0.20247262716293335,
+      "learning_rate": 6.126661944274285e-06,
+      "loss": 0.1873,
+      "step": 9039
+    },
+    {
+      "epoch": 6.804666917576213,
+      "grad_norm": 0.210099995136261,
+      "learning_rate": 6.124338110521206e-06,
+      "loss": 0.1149,
+      "step": 9040
+    },
+    {
+      "epoch": 6.805419646217539,
+      "grad_norm": 0.15032555162906647,
+      "learning_rate": 6.122016614378378e-06,
+      "loss": 0.1462,
+      "step": 9041
+    },
+    {
+      "epoch": 6.806172374858863,
+      "grad_norm": 0.18035855889320374,
+      "learning_rate": 6.119697456099666e-06,
+      "loss": 0.1661,
+      "step": 9042
+    },
+    {
+      "epoch": 6.8069251035001885,
+      "grad_norm": 0.17341667413711548,
+      "learning_rate": 6.117380635938702e-06,
+      "loss": 0.2222,
+      "step": 9043
+    },
+    {
+      "epoch": 6.807677832141513,
+      "grad_norm": 0.16925868391990662,
+      "learning_rate": 6.115066154148842e-06,
+      "loss": 0.1509,
+      "step": 9044
+    },
+    {
+      "epoch": 6.8084305607828375,
+      "grad_norm": 0.18473222851753235,
+      "learning_rate": 6.112754010983192e-06,
+      "loss": 0.1989,
+      "step": 9045
+    },
+    {
+      "epoch": 6.809183289424163,
+      "grad_norm": 0.17586006224155426,
+      "learning_rate": 6.110444206694613e-06,
+      "loss": 0.1324,
+      "step": 9046
+    },
+    {
+      "epoch": 6.809936018065487,
+      "grad_norm": 0.16870157420635223,
+      "learning_rate": 6.108136741535694e-06,
+      "loss": 0.1547,
+      "step": 9047
+    },
+    {
+      "epoch": 6.810688746706813,
+      "grad_norm": 0.1771600842475891,
+      "learning_rate": 6.1058316157587815e-06,
+      "loss": 0.1139,
+      "step": 9048
+    },
+    {
+      "epoch": 6.811441475348137,
+      "grad_norm": 0.19170044362545013,
+      "learning_rate": 6.1035288296159526e-06,
+      "loss": 0.1868,
+      "step": 9049
+    },
+    {
+      "epoch": 6.812194203989462,
+      "grad_norm": 0.14203530550003052,
+      "learning_rate": 6.1012283833590465e-06,
+      "loss": 0.1236,
+      "step": 9050
+    },
+    {
+      "epoch": 6.812946932630787,
+      "grad_norm": 0.19455240666866302,
+      "learning_rate": 6.098930277239629e-06,
+      "loss": 0.1662,
+      "step": 9051
+    },
+    {
+      "epoch": 6.8136996612721115,
+      "grad_norm": 0.15385456383228302,
+      "learning_rate": 6.096634511509023e-06,
+      "loss": 0.2293,
+      "step": 9052
+    },
+    {
+      "epoch": 6.814452389913436,
+      "grad_norm": 0.19526728987693787,
+      "learning_rate": 6.094341086418286e-06,
+      "loss": 0.2342,
+      "step": 9053
+    },
+    {
+      "epoch": 6.815205118554761,
+      "grad_norm": 0.1651216745376587,
+      "learning_rate": 6.092050002218226e-06,
+      "loss": 0.1105,
+      "step": 9054
+    },
+    {
+      "epoch": 6.815957847196086,
+      "grad_norm": 0.16209134459495544,
+      "learning_rate": 6.089761259159385e-06,
+      "loss": 0.0631,
+      "step": 9055
+    },
+    {
+      "epoch": 6.81671057583741,
+      "grad_norm": 0.16872255504131317,
+      "learning_rate": 6.0874748574920706e-06,
+      "loss": 0.1386,
+      "step": 9056
+    },
+    {
+      "epoch": 6.817463304478736,
+      "grad_norm": 0.16399109363555908,
+      "learning_rate": 6.085190797466308e-06,
+      "loss": 0.125,
+      "step": 9057
+    },
+    {
+      "epoch": 6.81821603312006,
+      "grad_norm": 0.1943063586950302,
+      "learning_rate": 6.08290907933189e-06,
+      "loss": 0.2159,
+      "step": 9058
+    },
+    {
+      "epoch": 6.818968761761385,
+      "grad_norm": 0.15169170498847961,
+      "learning_rate": 6.080629703338327e-06,
+      "loss": 0.0958,
+      "step": 9059
+    },
+    {
+      "epoch": 6.81972149040271,
+      "grad_norm": 0.1672370582818985,
+      "learning_rate": 6.078352669734901e-06,
+      "loss": 0.1552,
+      "step": 9060
+    },
+    {
+      "epoch": 6.820474219044034,
+      "grad_norm": 0.1537681519985199,
+      "learning_rate": 6.076077978770613e-06,
+      "loss": 0.0943,
+      "step": 9061
+    },
+    {
+      "epoch": 6.82122694768536,
+      "grad_norm": 0.17460502684116364,
+      "learning_rate": 6.073805630694237e-06,
+      "loss": 0.1462,
+      "step": 9062
+    },
+    {
+      "epoch": 6.821979676326684,
+      "grad_norm": 0.1813414841890335,
+      "learning_rate": 6.071535625754255e-06,
+      "loss": 0.1382,
+      "step": 9063
+    },
+    {
+      "epoch": 6.822732404968009,
+      "grad_norm": 0.16990172863006592,
+      "learning_rate": 6.069267964198927e-06,
+      "loss": 0.1552,
+      "step": 9064
+    },
+    {
+      "epoch": 6.823485133609334,
+      "grad_norm": 0.17758101224899292,
+      "learning_rate": 6.0670026462762295e-06,
+      "loss": 0.2076,
+      "step": 9065
+    },
+    {
+      "epoch": 6.8242378622506585,
+      "grad_norm": 0.1634216457605362,
+      "learning_rate": 6.064739672233897e-06,
+      "loss": 0.1471,
+      "step": 9066
+    },
+    {
+      "epoch": 6.824990590891984,
+      "grad_norm": 0.16655908524990082,
+      "learning_rate": 6.06247904231941e-06,
+      "loss": 0.1916,
+      "step": 9067
+    },
+    {
+      "epoch": 6.825743319533308,
+      "grad_norm": 0.18505340814590454,
+      "learning_rate": 6.060220756779981e-06,
+      "loss": 0.2581,
+      "step": 9068
+    },
+    {
+      "epoch": 6.826496048174633,
+      "grad_norm": 0.21439802646636963,
+      "learning_rate": 6.057964815862579e-06,
+      "loss": 0.2105,
+      "step": 9069
+    },
+    {
+      "epoch": 6.827248776815958,
+      "grad_norm": 0.18505296111106873,
+      "learning_rate": 6.0557112198139026e-06,
+      "loss": 0.1692,
+      "step": 9070
+    },
+    {
+      "epoch": 6.828001505457283,
+      "grad_norm": 0.19905975461006165,
+      "learning_rate": 6.0534599688804105e-06,
+      "loss": 0.1668,
+      "step": 9071
+    },
+    {
+      "epoch": 6.828754234098607,
+      "grad_norm": 0.19586703181266785,
+      "learning_rate": 6.051211063308289e-06,
+      "loss": 0.1942,
+      "step": 9072
+    },
+    {
+      "epoch": 6.8295069627399325,
+      "grad_norm": 0.16823816299438477,
+      "learning_rate": 6.048964503343484e-06,
+      "loss": 0.1741,
+      "step": 9073
+    },
+    {
+      "epoch": 6.830259691381257,
+      "grad_norm": 0.16672897338867188,
+      "learning_rate": 6.04672028923166e-06,
+      "loss": 0.1126,
+      "step": 9074
+    },
+    {
+      "epoch": 6.8310124200225815,
+      "grad_norm": 0.17705610394477844,
+      "learning_rate": 6.044478421218257e-06,
+      "loss": 0.1463,
+      "step": 9075
+    },
+    {
+      "epoch": 6.831765148663907,
+      "grad_norm": 0.20422276854515076,
+      "learning_rate": 6.042238899548434e-06,
+      "loss": 0.1677,
+      "step": 9076
+    },
+    {
+      "epoch": 6.832517877305231,
+      "grad_norm": 0.18119825422763824,
+      "learning_rate": 6.040001724467109e-06,
+      "loss": 0.1847,
+      "step": 9077
+    },
+    {
+      "epoch": 6.833270605946556,
+      "grad_norm": 0.15586361289024353,
+      "learning_rate": 6.0377668962189235e-06,
+      "loss": 0.1476,
+      "step": 9078
+    },
+    {
+      "epoch": 6.834023334587881,
+      "grad_norm": 0.15529999136924744,
+      "learning_rate": 6.035534415048286e-06,
+      "loss": 0.092,
+      "step": 9079
+    },
+    {
+      "epoch": 6.834776063229206,
+      "grad_norm": 0.17334872484207153,
+      "learning_rate": 6.033304281199331e-06,
+      "loss": 0.2409,
+      "step": 9080
+    },
+    {
+      "epoch": 6.835528791870531,
+      "grad_norm": 0.18114148080348969,
+      "learning_rate": 6.031076494915946e-06,
+      "loss": 0.1647,
+      "step": 9081
+    },
+    {
+      "epoch": 6.8362815205118554,
+      "grad_norm": 0.1708449423313141,
+      "learning_rate": 6.028851056441758e-06,
+      "loss": 0.152,
+      "step": 9082
+    },
+    {
+      "epoch": 6.83703424915318,
+      "grad_norm": 0.17473837733268738,
+      "learning_rate": 6.026627966020139e-06,
+      "loss": 0.0953,
+      "step": 9083
+    },
+    {
+      "epoch": 6.837786977794505,
+      "grad_norm": 0.17218314111232758,
+      "learning_rate": 6.024407223894197e-06,
+      "loss": 0.0963,
+      "step": 9084
+    },
+    {
+      "epoch": 6.83853970643583,
+      "grad_norm": 0.14286787807941437,
+      "learning_rate": 6.022188830306797e-06,
+      "loss": 0.0973,
+      "step": 9085
+    },
+    {
+      "epoch": 6.839292435077155,
+      "grad_norm": 0.174488365650177,
+      "learning_rate": 6.0199727855005345e-06,
+      "loss": 0.1811,
+      "step": 9086
+    },
+    {
+      "epoch": 6.84004516371848,
+      "grad_norm": 0.15982212126255035,
+      "learning_rate": 6.017759089717749e-06,
+      "loss": 0.2153,
+      "step": 9087
+    },
+    {
+      "epoch": 6.840797892359804,
+      "grad_norm": 0.1669660210609436,
+      "learning_rate": 6.0155477432005345e-06,
+      "loss": 0.1651,
+      "step": 9088
+    },
+    {
+      "epoch": 6.841550621001129,
+      "grad_norm": 0.17226020991802216,
+      "learning_rate": 6.013338746190718e-06,
+      "loss": 0.2283,
+      "step": 9089
+    },
+    {
+      "epoch": 6.842303349642454,
+      "grad_norm": 0.16546054184436798,
+      "learning_rate": 6.011132098929871e-06,
+      "loss": 0.1944,
+      "step": 9090
+    },
+    {
+      "epoch": 6.843056078283778,
+      "grad_norm": 0.17987163364887238,
+      "learning_rate": 6.008927801659307e-06,
+      "loss": 0.2095,
+      "step": 9091
+    },
+    {
+      "epoch": 6.843808806925104,
+      "grad_norm": 0.1629629135131836,
+      "learning_rate": 6.006725854620093e-06,
+      "loss": 0.2546,
+      "step": 9092
+    },
+    {
+      "epoch": 6.844561535566428,
+      "grad_norm": 0.19408802688121796,
+      "learning_rate": 6.004526258053018e-06,
+      "loss": 0.1033,
+      "step": 9093
+    },
+    {
+      "epoch": 6.845314264207753,
+      "grad_norm": 0.17280739545822144,
+      "learning_rate": 6.0023290121986384e-06,
+      "loss": 0.1145,
+      "step": 9094
+    },
+    {
+      "epoch": 6.846066992849078,
+      "grad_norm": 0.19072900712490082,
+      "learning_rate": 6.000134117297234e-06,
+      "loss": 0.1922,
+      "step": 9095
+    },
+    {
+      "epoch": 6.8468197214904025,
+      "grad_norm": 0.2261296957731247,
+      "learning_rate": 5.997941573588843e-06,
+      "loss": 0.2231,
+      "step": 9096
+    },
+    {
+      "epoch": 6.847572450131728,
+      "grad_norm": 0.17243972420692444,
+      "learning_rate": 5.995751381313228e-06,
+      "loss": 0.171,
+      "step": 9097
+    },
+    {
+      "epoch": 6.848325178773052,
+      "grad_norm": 0.1819406896829605,
+      "learning_rate": 5.993563540709916e-06,
+      "loss": 0.1806,
+      "step": 9098
+    },
+    {
+      "epoch": 6.849077907414377,
+      "grad_norm": 0.19137223064899445,
+      "learning_rate": 5.991378052018154e-06,
+      "loss": 0.1553,
+      "step": 9099
+    },
+    {
+      "epoch": 6.849830636055702,
+      "grad_norm": 0.17236994206905365,
+      "learning_rate": 5.989194915476954e-06,
+      "loss": 0.1766,
+      "step": 9100
+    },
+    {
+      "epoch": 6.850583364697027,
+      "grad_norm": 0.16802355647087097,
+      "learning_rate": 5.987014131325057e-06,
+      "loss": 0.138,
+      "step": 9101
+    },
+    {
+      "epoch": 6.851336093338351,
+      "grad_norm": 0.18285518884658813,
+      "learning_rate": 5.98483569980095e-06,
+      "loss": 0.1961,
+      "step": 9102
+    },
+    {
+      "epoch": 6.8520888219796765,
+      "grad_norm": 0.14868606626987457,
+      "learning_rate": 5.982659621142863e-06,
+      "loss": 0.1286,
+      "step": 9103
+    },
+    {
+      "epoch": 6.852841550621001,
+      "grad_norm": 0.24087001383304596,
+      "learning_rate": 5.9804858955887725e-06,
+      "loss": 0.1544,
+      "step": 9104
+    },
+    {
+      "epoch": 6.853594279262326,
+      "grad_norm": 0.18202632665634155,
+      "learning_rate": 5.978314523376386e-06,
+      "loss": 0.0946,
+      "step": 9105
+    },
+    {
+      "epoch": 6.854347007903651,
+      "grad_norm": 0.1918644905090332,
+      "learning_rate": 5.9761455047431665e-06,
+      "loss": 0.343,
+      "step": 9106
+    },
+    {
+      "epoch": 6.855099736544975,
+      "grad_norm": 0.20113520324230194,
+      "learning_rate": 5.9739788399263134e-06,
+      "loss": 0.1639,
+      "step": 9107
+    },
+    {
+      "epoch": 6.855852465186301,
+      "grad_norm": 0.17498397827148438,
+      "learning_rate": 5.971814529162775e-06,
+      "loss": 0.195,
+      "step": 9108
+    },
+    {
+      "epoch": 6.856605193827625,
+      "grad_norm": 0.17755509912967682,
+      "learning_rate": 5.9696525726892274e-06,
+      "loss": 0.2267,
+      "step": 9109
+    },
+    {
+      "epoch": 6.85735792246895,
+      "grad_norm": 0.16548463702201843,
+      "learning_rate": 5.967492970742108e-06,
+      "loss": 0.1095,
+      "step": 9110
+    },
+    {
+      "epoch": 6.858110651110275,
+      "grad_norm": 0.17868390679359436,
+      "learning_rate": 5.965335723557582e-06,
+      "loss": 0.1817,
+      "step": 9111
+    },
+    {
+      "epoch": 6.858863379751599,
+      "grad_norm": 0.17369379103183746,
+      "learning_rate": 5.963180831371563e-06,
+      "loss": 0.2015,
+      "step": 9112
+    },
+    {
+      "epoch": 6.859616108392924,
+      "grad_norm": 0.2023226022720337,
+      "learning_rate": 5.961028294419709e-06,
+      "loss": 0.2094,
+      "step": 9113
+    },
+    {
+      "epoch": 6.860368837034249,
+      "grad_norm": 0.17929013073444366,
+      "learning_rate": 5.958878112937416e-06,
+      "loss": 0.1883,
+      "step": 9114
+    },
+    {
+      "epoch": 6.861121565675574,
+      "grad_norm": 0.19664649665355682,
+      "learning_rate": 5.95673028715983e-06,
+      "loss": 0.2559,
+      "step": 9115
+    },
+    {
+      "epoch": 6.861874294316899,
+      "grad_norm": 0.19021382927894592,
+      "learning_rate": 5.954584817321828e-06,
+      "loss": 0.1033,
+      "step": 9116
+    },
+    {
+      "epoch": 6.862627022958224,
+      "grad_norm": 0.19198784232139587,
+      "learning_rate": 5.952441703658037e-06,
+      "loss": 0.1658,
+      "step": 9117
+    },
+    {
+      "epoch": 6.863379751599548,
+      "grad_norm": 0.18536818027496338,
+      "learning_rate": 5.950300946402825e-06,
+      "loss": 0.1461,
+      "step": 9118
+    },
+    {
+      "epoch": 6.864132480240873,
+      "grad_norm": 0.17602196335792542,
+      "learning_rate": 5.948162545790301e-06,
+      "loss": 0.1345,
+      "step": 9119
+    },
+    {
+      "epoch": 6.864885208882198,
+      "grad_norm": 0.18133299052715302,
+      "learning_rate": 5.946026502054317e-06,
+      "loss": 0.1655,
+      "step": 9120
+    },
+    {
+      "epoch": 6.865637937523523,
+      "grad_norm": 0.1897277683019638,
+      "learning_rate": 5.943892815428474e-06,
+      "loss": 0.1978,
+      "step": 9121
+    },
+    {
+      "epoch": 6.866390666164848,
+      "grad_norm": 0.17590303719043732,
+      "learning_rate": 5.9417614861460995e-06,
+      "loss": 0.1998,
+      "step": 9122
+    },
+    {
+      "epoch": 6.867143394806172,
+      "grad_norm": 0.16365347802639008,
+      "learning_rate": 5.939632514440278e-06,
+      "loss": 0.1799,
+      "step": 9123
+    },
+    {
+      "epoch": 6.867896123447498,
+      "grad_norm": 0.15664416551589966,
+      "learning_rate": 5.937505900543828e-06,
+      "loss": 0.1615,
+      "step": 9124
+    },
+    {
+      "epoch": 6.868648852088822,
+      "grad_norm": 0.17882834374904633,
+      "learning_rate": 5.9353816446893144e-06,
+      "loss": 0.1726,
+      "step": 9125
+    },
+    {
+      "epoch": 6.8694015807301465,
+      "grad_norm": 0.15997828543186188,
+      "learning_rate": 5.933259747109042e-06,
+      "loss": 0.1313,
+      "step": 9126
+    },
+    {
+      "epoch": 6.870154309371472,
+      "grad_norm": 0.17075949907302856,
+      "learning_rate": 5.931140208035062e-06,
+      "loss": 0.1647,
+      "step": 9127
+    },
+    {
+      "epoch": 6.870907038012796,
+      "grad_norm": 0.1806129813194275,
+      "learning_rate": 5.929023027699156e-06,
+      "loss": 0.232,
+      "step": 9128
+    },
+    {
+      "epoch": 6.871659766654121,
+      "grad_norm": 0.13889382779598236,
+      "learning_rate": 5.926908206332864e-06,
+      "loss": 0.133,
+      "step": 9129
+    },
+    {
+      "epoch": 6.872412495295446,
+      "grad_norm": 0.17679063975811005,
+      "learning_rate": 5.924795744167453e-06,
+      "loss": 0.1628,
+      "step": 9130
+    },
+    {
+      "epoch": 6.873165223936771,
+      "grad_norm": 0.21719300746917725,
+      "learning_rate": 5.922685641433942e-06,
+      "loss": 0.1649,
+      "step": 9131
+    },
+    {
+      "epoch": 6.873917952578095,
+      "grad_norm": 0.17525354027748108,
+      "learning_rate": 5.920577898363088e-06,
+      "loss": 0.2053,
+      "step": 9132
+    },
+    {
+      "epoch": 6.8746706812194205,
+      "grad_norm": 0.15865714848041534,
+      "learning_rate": 5.918472515185393e-06,
+      "loss": 0.1869,
+      "step": 9133
+    },
+    {
+      "epoch": 6.875423409860745,
+      "grad_norm": 0.17898888885974884,
+      "learning_rate": 5.916369492131095e-06,
+      "loss": 0.1666,
+      "step": 9134
+    },
+    {
+      "epoch": 6.87617613850207,
+      "grad_norm": 0.9048988819122314,
+      "learning_rate": 5.914268829430177e-06,
+      "loss": 0.166,
+      "step": 9135
+    },
+    {
+      "epoch": 6.876928867143395,
+      "grad_norm": 0.17743374407291412,
+      "learning_rate": 5.912170527312371e-06,
+      "loss": 0.2588,
+      "step": 9136
+    },
+    {
+      "epoch": 6.877681595784719,
+      "grad_norm": 0.19911053776741028,
+      "learning_rate": 5.910074586007135e-06,
+      "loss": 0.1718,
+      "step": 9137
+    },
+    {
+      "epoch": 6.878434324426045,
+      "grad_norm": 0.17957210540771484,
+      "learning_rate": 5.907981005743681e-06,
+      "loss": 0.1407,
+      "step": 9138
+    },
+    {
+      "epoch": 6.879187053067369,
+      "grad_norm": 0.1615339070558548,
+      "learning_rate": 5.905889786750964e-06,
+      "loss": 0.1314,
+      "step": 9139
+    },
+    {
+      "epoch": 6.8799397817086945,
+      "grad_norm": 0.18534255027770996,
+      "learning_rate": 5.903800929257675e-06,
+      "loss": 0.2383,
+      "step": 9140
+    },
+    {
+      "epoch": 6.880692510350019,
+      "grad_norm": 0.17687661945819855,
+      "learning_rate": 5.901714433492244e-06,
+      "loss": 0.1515,
+      "step": 9141
+    },
+    {
+      "epoch": 6.881445238991343,
+      "grad_norm": 0.17865508794784546,
+      "learning_rate": 5.899630299682854e-06,
+      "loss": 0.1821,
+      "step": 9142
+    },
+    {
+      "epoch": 6.882197967632669,
+      "grad_norm": 0.17505647242069244,
+      "learning_rate": 5.897548528057415e-06,
+      "loss": 0.1928,
+      "step": 9143
+    },
+    {
+      "epoch": 6.882950696273993,
+      "grad_norm": 0.1526002138853073,
+      "learning_rate": 5.8954691188435895e-06,
+      "loss": 0.0966,
+      "step": 9144
+    },
+    {
+      "epoch": 6.883703424915318,
+      "grad_norm": 0.17759129405021667,
+      "learning_rate": 5.8933920722687816e-06,
+      "loss": 0.1706,
+      "step": 9145
+    },
+    {
+      "epoch": 6.884456153556643,
+      "grad_norm": 0.18283486366271973,
+      "learning_rate": 5.891317388560131e-06,
+      "loss": 0.2001,
+      "step": 9146
+    },
+    {
+      "epoch": 6.885208882197968,
+      "grad_norm": 0.1634177714586258,
+      "learning_rate": 5.889245067944522e-06,
+      "loss": 0.1454,
+      "step": 9147
+    },
+    {
+      "epoch": 6.885961610839292,
+      "grad_norm": 0.18654844164848328,
+      "learning_rate": 5.887175110648585e-06,
+      "loss": 0.2313,
+      "step": 9148
+    },
+    {
+      "epoch": 6.886714339480617,
+      "grad_norm": 0.20342320203781128,
+      "learning_rate": 5.885107516898678e-06,
+      "loss": 0.2158,
+      "step": 9149
+    },
+    {
+      "epoch": 6.887467068121942,
+      "grad_norm": 0.14795415103435516,
+      "learning_rate": 5.883042286920918e-06,
+      "loss": 0.1896,
+      "step": 9150
+    },
+    {
+      "epoch": 6.888219796763266,
+      "grad_norm": 0.17844241857528687,
+      "learning_rate": 5.88097942094115e-06,
+      "loss": 0.1326,
+      "step": 9151
+    },
+    {
+      "epoch": 6.888972525404592,
+      "grad_norm": 0.18412765860557556,
+      "learning_rate": 5.878918919184974e-06,
+      "loss": 0.1047,
+      "step": 9152
+    },
+    {
+      "epoch": 6.889725254045916,
+      "grad_norm": 0.18018552660942078,
+      "learning_rate": 5.876860781877715e-06,
+      "loss": 0.1591,
+      "step": 9153
+    },
+    {
+      "epoch": 6.8904779826872415,
+      "grad_norm": 0.16454093158245087,
+      "learning_rate": 5.874805009244453e-06,
+      "loss": 0.1135,
+      "step": 9154
+    },
+    {
+      "epoch": 6.891230711328566,
+      "grad_norm": 0.19804099202156067,
+      "learning_rate": 5.87275160151e-06,
+      "loss": 0.1372,
+      "step": 9155
+    },
+    {
+      "epoch": 6.8919834399698905,
+      "grad_norm": 0.17198461294174194,
+      "learning_rate": 5.870700558898916e-06,
+      "loss": 0.217,
+      "step": 9156
+    },
+    {
+      "epoch": 6.892736168611216,
+      "grad_norm": 0.17346464097499847,
+      "learning_rate": 5.868651881635498e-06,
+      "loss": 0.2072,
+      "step": 9157
+    },
+    {
+      "epoch": 6.89348889725254,
+      "grad_norm": 0.16746307909488678,
+      "learning_rate": 5.866605569943794e-06,
+      "loss": 0.2346,
+      "step": 9158
+    },
+    {
+      "epoch": 6.894241625893866,
+      "grad_norm": 0.18652644753456116,
+      "learning_rate": 5.864561624047574e-06,
+      "loss": 0.1561,
+      "step": 9159
+    },
+    {
+      "epoch": 6.89499435453519,
+      "grad_norm": 0.18331921100616455,
+      "learning_rate": 5.862520044170367e-06,
+      "loss": 0.1763,
+      "step": 9160
+    },
+    {
+      "epoch": 6.895747083176515,
+      "grad_norm": 0.1801430732011795,
+      "learning_rate": 5.860480830535439e-06,
+      "loss": 0.1667,
+      "step": 9161
+    },
+    {
+      "epoch": 6.89649981181784,
+      "grad_norm": 0.15705016255378723,
+      "learning_rate": 5.858443983365791e-06,
+      "loss": 0.1297,
+      "step": 9162
+    },
+    {
+      "epoch": 6.8972525404591645,
+      "grad_norm": 0.15624496340751648,
+      "learning_rate": 5.85640950288417e-06,
+      "loss": 0.2064,
+      "step": 9163
+    },
+    {
+      "epoch": 6.898005269100489,
+      "grad_norm": 0.16970805823802948,
+      "learning_rate": 5.854377389313067e-06,
+      "loss": 0.2378,
+      "step": 9164
+    },
+    {
+      "epoch": 6.898757997741814,
+      "grad_norm": 0.18475955724716187,
+      "learning_rate": 5.852347642874711e-06,
+      "loss": 0.2622,
+      "step": 9165
+    },
+    {
+      "epoch": 6.899510726383139,
+      "grad_norm": 0.18405461311340332,
+      "learning_rate": 5.850320263791065e-06,
+      "loss": 0.1916,
+      "step": 9166
+    },
+    {
+      "epoch": 6.900263455024463,
+      "grad_norm": 0.187552809715271,
+      "learning_rate": 5.848295252283852e-06,
+      "loss": 0.2575,
+      "step": 9167
+    },
+    {
+      "epoch": 6.901016183665789,
+      "grad_norm": 0.18161293864250183,
+      "learning_rate": 5.846272608574511e-06,
+      "loss": 0.2518,
+      "step": 9168
+    },
+    {
+      "epoch": 6.901768912307113,
+      "grad_norm": 0.19358642399311066,
+      "learning_rate": 5.844252332884247e-06,
+      "loss": 0.2126,
+      "step": 9169
+    },
+    {
+      "epoch": 6.9025216409484385,
+      "grad_norm": 0.16061607003211975,
+      "learning_rate": 5.842234425433983e-06,
+      "loss": 0.1359,
+      "step": 9170
+    },
+    {
+      "epoch": 6.903274369589763,
+      "grad_norm": 0.16984976828098297,
+      "learning_rate": 5.840218886444406e-06,
+      "loss": 0.2086,
+      "step": 9171
+    },
+    {
+      "epoch": 6.904027098231087,
+      "grad_norm": 0.16013413667678833,
+      "learning_rate": 5.838205716135921e-06,
+      "loss": 0.2129,
+      "step": 9172
+    },
+    {
+      "epoch": 6.904779826872413,
+      "grad_norm": 0.1796564757823944,
+      "learning_rate": 5.836194914728697e-06,
+      "loss": 0.1698,
+      "step": 9173
+    },
+    {
+      "epoch": 6.905532555513737,
+      "grad_norm": 0.17107151448726654,
+      "learning_rate": 5.834186482442622e-06,
+      "loss": 0.1489,
+      "step": 9174
+    },
+    {
+      "epoch": 6.906285284155063,
+      "grad_norm": 0.17684851586818695,
+      "learning_rate": 5.83218041949734e-06,
+      "loss": 0.1441,
+      "step": 9175
+    },
+    {
+      "epoch": 6.907038012796387,
+      "grad_norm": 0.16075344383716583,
+      "learning_rate": 5.830176726112227e-06,
+      "loss": 0.1802,
+      "step": 9176
+    },
+    {
+      "epoch": 6.9077907414377115,
+      "grad_norm": 0.15691833198070526,
+      "learning_rate": 5.828175402506414e-06,
+      "loss": 0.1199,
+      "step": 9177
+    },
+    {
+      "epoch": 6.908543470079037,
+      "grad_norm": 0.1673429161310196,
+      "learning_rate": 5.826176448898751e-06,
+      "loss": 0.1124,
+      "step": 9178
+    },
+    {
+      "epoch": 6.909296198720361,
+      "grad_norm": 0.16875813901424408,
+      "learning_rate": 5.824179865507848e-06,
+      "loss": 0.2695,
+      "step": 9179
+    },
+    {
+      "epoch": 6.910048927361686,
+      "grad_norm": 0.17945249378681183,
+      "learning_rate": 5.8221856525520435e-06,
+      "loss": 0.1107,
+      "step": 9180
+    },
+    {
+      "epoch": 6.910801656003011,
+      "grad_norm": 0.1684263050556183,
+      "learning_rate": 5.820193810249423e-06,
+      "loss": 0.18,
+      "step": 9181
+    },
+    {
+      "epoch": 6.911554384644336,
+      "grad_norm": 0.15749810636043549,
+      "learning_rate": 5.8182043388178144e-06,
+      "loss": 0.0896,
+      "step": 9182
+    },
+    {
+      "epoch": 6.91230711328566,
+      "grad_norm": 0.16866309940814972,
+      "learning_rate": 5.816217238474782e-06,
+      "loss": 0.1628,
+      "step": 9183
+    },
+    {
+      "epoch": 6.9130598419269855,
+      "grad_norm": 0.16996467113494873,
+      "learning_rate": 5.814232509437631e-06,
+      "loss": 0.1371,
+      "step": 9184
+    },
+    {
+      "epoch": 6.91381257056831,
+      "grad_norm": 0.19131989777088165,
+      "learning_rate": 5.812250151923408e-06,
+      "loss": 0.2334,
+      "step": 9185
+    },
+    {
+      "epoch": 6.9145652992096345,
+      "grad_norm": 0.17434820532798767,
+      "learning_rate": 5.810270166148901e-06,
+      "loss": 0.2287,
+      "step": 9186
+    },
+    {
+      "epoch": 6.91531802785096,
+      "grad_norm": 0.16808070242404938,
+      "learning_rate": 5.8082925523306385e-06,
+      "loss": 0.1957,
+      "step": 9187
+    },
+    {
+      "epoch": 6.916070756492284,
+      "grad_norm": 0.18075785040855408,
+      "learning_rate": 5.806317310684892e-06,
+      "loss": 0.169,
+      "step": 9188
+    },
+    {
+      "epoch": 6.91682348513361,
+      "grad_norm": 0.16849488019943237,
+      "learning_rate": 5.804344441427665e-06,
+      "loss": 0.1637,
+      "step": 9189
+    },
+    {
+      "epoch": 6.917576213774934,
+      "grad_norm": 0.1901453286409378,
+      "learning_rate": 5.802373944774714e-06,
+      "loss": 0.1704,
+      "step": 9190
+    },
+    {
+      "epoch": 6.918328942416259,
+      "grad_norm": 0.20044827461242676,
+      "learning_rate": 5.800405820941522e-06,
+      "loss": 0.1658,
+      "step": 9191
+    },
+    {
+      "epoch": 6.919081671057584,
+      "grad_norm": 0.1631947159767151,
+      "learning_rate": 5.798440070143331e-06,
+      "loss": 0.1585,
+      "step": 9192
+    },
+    {
+      "epoch": 6.9198343996989085,
+      "grad_norm": 0.16352154314517975,
+      "learning_rate": 5.796476692595102e-06,
+      "loss": 0.1631,
+      "step": 9193
+    },
+    {
+      "epoch": 6.920587128340234,
+      "grad_norm": 0.17681702971458435,
+      "learning_rate": 5.794515688511552e-06,
+      "loss": 0.2031,
+      "step": 9194
+    },
+    {
+      "epoch": 6.921339856981558,
+      "grad_norm": 0.16840511560440063,
+      "learning_rate": 5.792557058107131e-06,
+      "loss": 0.1617,
+      "step": 9195
+    },
+    {
+      "epoch": 6.922092585622883,
+      "grad_norm": 0.1847878247499466,
+      "learning_rate": 5.790600801596037e-06,
+      "loss": 0.1183,
+      "step": 9196
+    },
+    {
+      "epoch": 6.922845314264208,
+      "grad_norm": 0.1784508228302002,
+      "learning_rate": 5.788646919192195e-06,
+      "loss": 0.1345,
+      "step": 9197
+    },
+    {
+      "epoch": 6.923598042905533,
+      "grad_norm": 0.19986358284950256,
+      "learning_rate": 5.786695411109289e-06,
+      "loss": 0.1237,
+      "step": 9198
+    },
+    {
+      "epoch": 6.924350771546857,
+      "grad_norm": 0.15233035385608673,
+      "learning_rate": 5.784746277560721e-06,
+      "loss": 0.0956,
+      "step": 9199
+    },
+    {
+      "epoch": 6.925103500188182,
+      "grad_norm": 0.18262962996959686,
+      "learning_rate": 5.782799518759658e-06,
+      "loss": 0.1638,
+      "step": 9200
+    },
+    {
+      "epoch": 6.925103500188182,
+      "eval_loss": 0.2020898312330246,
+      "eval_runtime": 453.0023,
+      "eval_samples_per_second": 21.252,
+      "eval_steps_per_second": 0.664,
+      "step": 9200
+    },
+    {
+      "epoch": 6.925856228829507,
+      "grad_norm": 0.13648352026939392,
+      "learning_rate": 5.7808551349189826e-06,
+      "loss": 0.1061,
+      "step": 9201
+    },
+    {
+      "epoch": 6.926608957470831,
+      "grad_norm": 0.1881500631570816,
+      "learning_rate": 5.778913126251339e-06,
+      "loss": 0.1933,
+      "step": 9202
+    },
+    {
+      "epoch": 6.927361686112157,
+      "grad_norm": 0.16280503571033478,
+      "learning_rate": 5.776973492969097e-06,
+      "loss": 0.1688,
+      "step": 9203
+    },
+    {
+      "epoch": 6.928114414753481,
+      "grad_norm": 0.18366216123104095,
+      "learning_rate": 5.775036235284377e-06,
+      "loss": 0.1759,
+      "step": 9204
+    },
+    {
+      "epoch": 6.928867143394806,
+      "grad_norm": 0.16828042268753052,
+      "learning_rate": 5.773101353409026e-06,
+      "loss": 0.1579,
+      "step": 9205
+    },
+    {
+      "epoch": 6.929619872036131,
+      "grad_norm": 0.16235437989234924,
+      "learning_rate": 5.771168847554648e-06,
+      "loss": 0.1404,
+      "step": 9206
+    },
+    {
+      "epoch": 6.9303726006774555,
+      "grad_norm": 0.17618657648563385,
+      "learning_rate": 5.769238717932579e-06,
+      "loss": 0.1437,
+      "step": 9207
+    },
+    {
+      "epoch": 6.931125329318781,
+      "grad_norm": 0.15962640941143036,
+      "learning_rate": 5.767310964753888e-06,
+      "loss": 0.0724,
+      "step": 9208
+    },
+    {
+      "epoch": 6.931878057960105,
+      "grad_norm": 0.16006723046302795,
+      "learning_rate": 5.765385588229395e-06,
+      "loss": 0.1845,
+      "step": 9209
+    },
+    {
+      "epoch": 6.93263078660143,
+      "grad_norm": 0.17647448182106018,
+      "learning_rate": 5.763462588569658e-06,
+      "loss": 0.1718,
+      "step": 9210
+    },
+    {
+      "epoch": 6.933383515242755,
+      "grad_norm": 0.1782313883304596,
+      "learning_rate": 5.7615419659849726e-06,
+      "loss": 0.157,
+      "step": 9211
+    },
+    {
+      "epoch": 6.93413624388408,
+      "grad_norm": 0.1872044801712036,
+      "learning_rate": 5.759623720685374e-06,
+      "loss": 0.1565,
+      "step": 9212
+    },
+    {
+      "epoch": 6.934888972525405,
+      "grad_norm": 0.16115252673625946,
+      "learning_rate": 5.757707852880639e-06,
+      "loss": 0.1955,
+      "step": 9213
+    },
+    {
+      "epoch": 6.9356417011667295,
+      "grad_norm": 0.1743275374174118,
+      "learning_rate": 5.755794362780283e-06,
+      "loss": 0.1888,
+      "step": 9214
+    },
+    {
+      "epoch": 6.936394429808054,
+      "grad_norm": 0.2027321308851242,
+      "learning_rate": 5.753883250593564e-06,
+      "loss": 0.1942,
+      "step": 9215
+    },
+    {
+      "epoch": 6.937147158449379,
+      "grad_norm": 0.16405415534973145,
+      "learning_rate": 5.751974516529477e-06,
+      "loss": 0.1846,
+      "step": 9216
+    },
+    {
+      "epoch": 6.937899887090704,
+      "grad_norm": 0.16241998970508575,
+      "learning_rate": 5.750068160796765e-06,
+      "loss": 0.1332,
+      "step": 9217
+    },
+    {
+      "epoch": 6.938652615732028,
+      "grad_norm": 0.1787491887807846,
+      "learning_rate": 5.748164183603892e-06,
+      "loss": 0.1264,
+      "step": 9218
+    },
+    {
+      "epoch": 6.939405344373354,
+      "grad_norm": 0.19587033987045288,
+      "learning_rate": 5.746262585159086e-06,
+      "loss": 0.1299,
+      "step": 9219
+    },
+    {
+      "epoch": 6.940158073014678,
+      "grad_norm": 0.15615347027778625,
+      "learning_rate": 5.7443633656702915e-06,
+      "loss": 0.1503,
+      "step": 9220
+    },
+    {
+      "epoch": 6.940910801656003,
+      "grad_norm": 0.17196132242679596,
+      "learning_rate": 5.742466525345213e-06,
+      "loss": 0.1544,
+      "step": 9221
+    },
+    {
+      "epoch": 6.941663530297328,
+      "grad_norm": 0.17525126039981842,
+      "learning_rate": 5.740572064391282e-06,
+      "loss": 0.1475,
+      "step": 9222
+    },
+    {
+      "epoch": 6.942416258938652,
+      "grad_norm": 0.16196881234645844,
+      "learning_rate": 5.738679983015679e-06,
+      "loss": 0.1302,
+      "step": 9223
+    },
+    {
+      "epoch": 6.943168987579978,
+      "grad_norm": 0.18296115100383759,
+      "learning_rate": 5.736790281425311e-06,
+      "loss": 0.1268,
+      "step": 9224
+    },
+    {
+      "epoch": 6.943921716221302,
+      "grad_norm": 0.19891373813152313,
+      "learning_rate": 5.73490295982684e-06,
+      "loss": 0.1557,
+      "step": 9225
+    },
+    {
+      "epoch": 6.944674444862627,
+      "grad_norm": 0.1656951755285263,
+      "learning_rate": 5.733018018426657e-06,
+      "loss": 0.1583,
+      "step": 9226
+    },
+    {
+      "epoch": 6.945427173503952,
+      "grad_norm": 0.14729690551757812,
+      "learning_rate": 5.731135457430895e-06,
+      "loss": 0.1173,
+      "step": 9227
+    },
+    {
+      "epoch": 6.946179902145277,
+      "grad_norm": 0.17173771560192108,
+      "learning_rate": 5.729255277045432e-06,
+      "loss": 0.1225,
+      "step": 9228
+    },
+    {
+      "epoch": 6.946932630786601,
+      "grad_norm": 0.17815105617046356,
+      "learning_rate": 5.727377477475884e-06,
+      "loss": 0.2064,
+      "step": 9229
+    },
+    {
+      "epoch": 6.947685359427926,
+      "grad_norm": 0.16346144676208496,
+      "learning_rate": 5.725502058927595e-06,
+      "loss": 0.0751,
+      "step": 9230
+    },
+    {
+      "epoch": 6.948438088069251,
+      "grad_norm": 0.16651810705661774,
+      "learning_rate": 5.723629021605666e-06,
+      "loss": 0.1231,
+      "step": 9231
+    },
+    {
+      "epoch": 6.949190816710576,
+      "grad_norm": 0.16479770839214325,
+      "learning_rate": 5.7217583657149285e-06,
+      "loss": 0.1133,
+      "step": 9232
+    },
+    {
+      "epoch": 6.949943545351901,
+      "grad_norm": 0.1664014309644699,
+      "learning_rate": 5.71989009145995e-06,
+      "loss": 0.2667,
+      "step": 9233
+    },
+    {
+      "epoch": 6.950696273993225,
+      "grad_norm": 0.1760845184326172,
+      "learning_rate": 5.718024199045046e-06,
+      "loss": 0.1476,
+      "step": 9234
+    },
+    {
+      "epoch": 6.951449002634551,
+      "grad_norm": 0.1764766424894333,
+      "learning_rate": 5.716160688674265e-06,
+      "loss": 0.1208,
+      "step": 9235
+    },
+    {
+      "epoch": 6.952201731275875,
+      "grad_norm": 0.19393615424633026,
+      "learning_rate": 5.714299560551404e-06,
+      "loss": 0.1792,
+      "step": 9236
+    },
+    {
+      "epoch": 6.9529544599171995,
+      "grad_norm": 0.17603832483291626,
+      "learning_rate": 5.712440814879985e-06,
+      "loss": 0.1855,
+      "step": 9237
+    },
+    {
+      "epoch": 6.953707188558525,
+      "grad_norm": 0.17262300848960876,
+      "learning_rate": 5.710584451863284e-06,
+      "loss": 0.1692,
+      "step": 9238
+    },
+    {
+      "epoch": 6.954459917199849,
+      "grad_norm": 0.1750895380973816,
+      "learning_rate": 5.708730471704307e-06,
+      "loss": 0.1353,
+      "step": 9239
+    },
+    {
+      "epoch": 6.955212645841174,
+      "grad_norm": 0.18765677511692047,
+      "learning_rate": 5.706878874605801e-06,
+      "loss": 0.1792,
+      "step": 9240
+    },
+    {
+      "epoch": 6.955965374482499,
+      "grad_norm": 0.17620089650154114,
+      "learning_rate": 5.705029660770255e-06,
+      "loss": 0.1626,
+      "step": 9241
+    },
+    {
+      "epoch": 6.956718103123824,
+      "grad_norm": 0.18388959765434265,
+      "learning_rate": 5.703182830399902e-06,
+      "loss": 0.1912,
+      "step": 9242
+    },
+    {
+      "epoch": 6.957470831765149,
+      "grad_norm": 0.15224073827266693,
+      "learning_rate": 5.7013383836967e-06,
+      "loss": 0.1249,
+      "step": 9243
+    },
+    {
+      "epoch": 6.9582235604064735,
+      "grad_norm": 0.1907045543193817,
+      "learning_rate": 5.69949632086236e-06,
+      "loss": 0.1798,
+      "step": 9244
+    },
+    {
+      "epoch": 6.958976289047798,
+      "grad_norm": 0.18667931854724884,
+      "learning_rate": 5.697656642098323e-06,
+      "loss": 0.1637,
+      "step": 9245
+    },
+    {
+      "epoch": 6.959729017689123,
+      "grad_norm": 0.15744712948799133,
+      "learning_rate": 5.69581934760578e-06,
+      "loss": 0.1085,
+      "step": 9246
+    },
+    {
+      "epoch": 6.960481746330448,
+      "grad_norm": 0.19210852682590485,
+      "learning_rate": 5.693984437585648e-06,
+      "loss": 0.1675,
+      "step": 9247
+    },
+    {
+      "epoch": 6.961234474971773,
+      "grad_norm": 0.15454590320587158,
+      "learning_rate": 5.692151912238596e-06,
+      "loss": 0.1282,
+      "step": 9248
+    },
+    {
+      "epoch": 6.961987203613098,
+      "grad_norm": 0.1653377264738083,
+      "learning_rate": 5.69032177176502e-06,
+      "loss": 0.1498,
+      "step": 9249
+    },
+    {
+      "epoch": 6.962739932254422,
+      "grad_norm": 0.1896493285894394,
+      "learning_rate": 5.688494016365067e-06,
+      "loss": 0.1783,
+      "step": 9250
+    },
+    {
+      "epoch": 6.9634926608957475,
+      "grad_norm": 0.17318212985992432,
+      "learning_rate": 5.686668646238616e-06,
+      "loss": 0.2221,
+      "step": 9251
+    },
+    {
+      "epoch": 6.964245389537072,
+      "grad_norm": 0.1716877967119217,
+      "learning_rate": 5.684845661585283e-06,
+      "loss": 0.1701,
+      "step": 9252
+    },
+    {
+      "epoch": 6.964998118178396,
+      "grad_norm": 0.17897970974445343,
+      "learning_rate": 5.68302506260443e-06,
+      "loss": 0.2106,
+      "step": 9253
+    },
+    {
+      "epoch": 6.965750846819722,
+      "grad_norm": 0.16309291124343872,
+      "learning_rate": 5.681206849495157e-06,
+      "loss": 0.178,
+      "step": 9254
+    },
+    {
+      "epoch": 6.966503575461046,
+      "grad_norm": 0.1524188220500946,
+      "learning_rate": 5.679391022456299e-06,
+      "loss": 0.1476,
+      "step": 9255
+    },
+    {
+      "epoch": 6.967256304102371,
+      "grad_norm": 0.18793624639511108,
+      "learning_rate": 5.67757758168643e-06,
+      "loss": 0.1942,
+      "step": 9256
+    },
+    {
+      "epoch": 6.968009032743696,
+      "grad_norm": 0.18678423762321472,
+      "learning_rate": 5.67576652738387e-06,
+      "loss": 0.1822,
+      "step": 9257
+    },
+    {
+      "epoch": 6.968761761385021,
+      "grad_norm": 0.1656944453716278,
+      "learning_rate": 5.673957859746667e-06,
+      "loss": 0.1418,
+      "step": 9258
+    },
+    {
+      "epoch": 6.969514490026345,
+      "grad_norm": 0.18654775619506836,
+      "learning_rate": 5.672151578972621e-06,
+      "loss": 0.1934,
+      "step": 9259
+    },
+    {
+      "epoch": 6.97026721866767,
+      "grad_norm": 0.1765357255935669,
+      "learning_rate": 5.670347685259256e-06,
+      "loss": 0.1129,
+      "step": 9260
+    },
+    {
+      "epoch": 6.971019947308995,
+      "grad_norm": 0.14810730516910553,
+      "learning_rate": 5.668546178803852e-06,
+      "loss": 0.1329,
+      "step": 9261
+    },
+    {
+      "epoch": 6.97177267595032,
+      "grad_norm": 0.17247073352336884,
+      "learning_rate": 5.666747059803413e-06,
+      "loss": 0.1833,
+      "step": 9262
+    },
+    {
+      "epoch": 6.972525404591645,
+      "grad_norm": 0.16399025917053223,
+      "learning_rate": 5.664950328454692e-06,
+      "loss": 0.1391,
+      "step": 9263
+    },
+    {
+      "epoch": 6.973278133232969,
+      "grad_norm": 0.16239909827709198,
+      "learning_rate": 5.663155984954171e-06,
+      "loss": 0.1272,
+      "step": 9264
+    },
+    {
+      "epoch": 6.9740308618742946,
+      "grad_norm": 0.188712015748024,
+      "learning_rate": 5.66136402949808e-06,
+      "loss": 0.1468,
+      "step": 9265
+    },
+    {
+      "epoch": 6.974783590515619,
+      "grad_norm": 0.1939750611782074,
+      "learning_rate": 5.659574462282387e-06,
+      "loss": 0.1695,
+      "step": 9266
+    },
+    {
+      "epoch": 6.975536319156944,
+      "grad_norm": 0.1532604992389679,
+      "learning_rate": 5.657787283502795e-06,
+      "loss": 0.111,
+      "step": 9267
+    },
+    {
+      "epoch": 6.976289047798269,
+      "grad_norm": 0.17853257060050964,
+      "learning_rate": 5.656002493354744e-06,
+      "loss": 0.1575,
+      "step": 9268
+    },
+    {
+      "epoch": 6.977041776439593,
+      "grad_norm": 0.18428535759449005,
+      "learning_rate": 5.654220092033422e-06,
+      "loss": 0.2164,
+      "step": 9269
+    },
+    {
+      "epoch": 6.977794505080919,
+      "grad_norm": 0.18450269103050232,
+      "learning_rate": 5.652440079733742e-06,
+      "loss": 0.1906,
+      "step": 9270
+    },
+    {
+      "epoch": 6.978547233722243,
+      "grad_norm": 0.17932608723640442,
+      "learning_rate": 5.65066245665037e-06,
+      "loss": 0.1883,
+      "step": 9271
+    },
+    {
+      "epoch": 6.979299962363568,
+      "grad_norm": 0.18452100455760956,
+      "learning_rate": 5.6488872229777e-06,
+      "loss": 0.2075,
+      "step": 9272
+    },
+    {
+      "epoch": 6.980052691004893,
+      "grad_norm": 0.16156788170337677,
+      "learning_rate": 5.6471143789098734e-06,
+      "loss": 0.1899,
+      "step": 9273
+    },
+    {
+      "epoch": 6.9808054196462175,
+      "grad_norm": 0.2013237625360489,
+      "learning_rate": 5.6453439246407614e-06,
+      "loss": 0.2129,
+      "step": 9274
+    },
+    {
+      "epoch": 6.981558148287542,
+      "grad_norm": 0.17515034973621368,
+      "learning_rate": 5.64357586036398e-06,
+      "loss": 0.1568,
+      "step": 9275
+    },
+    {
+      "epoch": 6.982310876928867,
+      "grad_norm": 0.17702360451221466,
+      "learning_rate": 5.641810186272881e-06,
+      "loss": 0.1716,
+      "step": 9276
+    },
+    {
+      "epoch": 6.983063605570192,
+      "grad_norm": 0.1638128161430359,
+      "learning_rate": 5.640046902560557e-06,
+      "loss": 0.1325,
+      "step": 9277
+    },
+    {
+      "epoch": 6.983816334211516,
+      "grad_norm": 0.16660210490226746,
+      "learning_rate": 5.638286009419836e-06,
+      "loss": 0.1608,
+      "step": 9278
+    },
+    {
+      "epoch": 6.984569062852842,
+      "grad_norm": 0.1820458471775055,
+      "learning_rate": 5.636527507043292e-06,
+      "loss": 0.2046,
+      "step": 9279
+    },
+    {
+      "epoch": 6.985321791494166,
+      "grad_norm": 0.16904503107070923,
+      "learning_rate": 5.634771395623228e-06,
+      "loss": 0.1688,
+      "step": 9280
+    },
+    {
+      "epoch": 6.9860745201354915,
+      "grad_norm": 0.17705582082271576,
+      "learning_rate": 5.633017675351688e-06,
+      "loss": 0.1646,
+      "step": 9281
+    },
+    {
+      "epoch": 6.986827248776816,
+      "grad_norm": 0.16776296496391296,
+      "learning_rate": 5.6312663464204615e-06,
+      "loss": 0.1255,
+      "step": 9282
+    },
+    {
+      "epoch": 6.98757997741814,
+      "grad_norm": 0.17528966069221497,
+      "learning_rate": 5.629517409021065e-06,
+      "loss": 0.1192,
+      "step": 9283
+    },
+    {
+      "epoch": 6.988332706059466,
+      "grad_norm": 0.15749572217464447,
+      "learning_rate": 5.627770863344763e-06,
+      "loss": 0.1651,
+      "step": 9284
+    },
+    {
+      "epoch": 6.98908543470079,
+      "grad_norm": 0.17295561730861664,
+      "learning_rate": 5.626026709582556e-06,
+      "loss": 0.1557,
+      "step": 9285
+    },
+    {
+      "epoch": 6.989838163342116,
+      "grad_norm": 0.16981877386569977,
+      "learning_rate": 5.624284947925185e-06,
+      "loss": 0.1008,
+      "step": 9286
+    },
+    {
+      "epoch": 6.99059089198344,
+      "grad_norm": 0.17399053275585175,
+      "learning_rate": 5.622545578563115e-06,
+      "loss": 0.1996,
+      "step": 9287
+    },
+    {
+      "epoch": 6.9913436206247646,
+      "grad_norm": 0.17937052249908447,
+      "learning_rate": 5.620808601686577e-06,
+      "loss": 0.1574,
+      "step": 9288
+    },
+    {
+      "epoch": 6.99209634926609,
+      "grad_norm": 0.165865957736969,
+      "learning_rate": 5.619074017485508e-06,
+      "loss": 0.1997,
+      "step": 9289
+    },
+    {
+      "epoch": 6.992849077907414,
+      "grad_norm": 0.1675519049167633,
+      "learning_rate": 5.617341826149612e-06,
+      "loss": 0.1952,
+      "step": 9290
+    },
+    {
+      "epoch": 6.993601806548739,
+      "grad_norm": 0.16915032267570496,
+      "learning_rate": 5.6156120278683085e-06,
+      "loss": 0.1988,
+      "step": 9291
+    },
+    {
+      "epoch": 6.994354535190064,
+      "grad_norm": 0.1617073118686676,
+      "learning_rate": 5.613884622830775e-06,
+      "loss": 0.1321,
+      "step": 9292
+    },
+    {
+      "epoch": 6.995107263831389,
+      "grad_norm": 0.16865842044353485,
+      "learning_rate": 5.612159611225912e-06,
+      "loss": 0.1239,
+      "step": 9293
+    },
+    {
+      "epoch": 6.995859992472713,
+      "grad_norm": 0.17428216338157654,
+      "learning_rate": 5.610436993242367e-06,
+      "loss": 0.1764,
+      "step": 9294
+    },
+    {
+      "epoch": 6.9966127211140385,
+      "grad_norm": 0.14767979085445404,
+      "learning_rate": 5.608716769068523e-06,
+      "loss": 0.1478,
+      "step": 9295
+    },
+    {
+      "epoch": 6.997365449755363,
+      "grad_norm": 0.18555033206939697,
+      "learning_rate": 5.606998938892498e-06,
+      "loss": 0.177,
+      "step": 9296
+    },
+    {
+      "epoch": 6.998118178396688,
+      "grad_norm": 0.19722111523151398,
+      "learning_rate": 5.605283502902151e-06,
+      "loss": 0.1413,
+      "step": 9297
+    },
+    {
+      "epoch": 6.998870907038013,
+      "grad_norm": 0.1775757223367691,
+      "learning_rate": 5.603570461285089e-06,
+      "loss": 0.1649,
+      "step": 9298
+    },
+    {
+      "epoch": 6.999623635679337,
+      "grad_norm": 0.1544857919216156,
+      "learning_rate": 5.601859814228634e-06,
+      "loss": 0.1851,
+      "step": 9299
+    },
+    {
+      "epoch": 7.000376364320663,
+      "grad_norm": 0.17472431063652039,
+      "learning_rate": 5.600151561919871e-06,
+      "loss": 0.141,
+      "step": 9300
+    },
+    {
+      "epoch": 7.001129092961987,
+      "grad_norm": 0.17086473107337952,
+      "learning_rate": 5.598445704545605e-06,
+      "loss": 0.1145,
+      "step": 9301
+    },
+    {
+      "epoch": 7.001881821603312,
+      "grad_norm": 0.15730687975883484,
+      "learning_rate": 5.596742242292387e-06,
+      "loss": 0.1566,
+      "step": 9302
+    },
+    {
+      "epoch": 7.002634550244637,
+      "grad_norm": 0.18831485509872437,
+      "learning_rate": 5.5950411753465065e-06,
+      "loss": 0.1498,
+      "step": 9303
+    },
+    {
+      "epoch": 7.0033872788859615,
+      "grad_norm": 0.17723087966442108,
+      "learning_rate": 5.5933425038939925e-06,
+      "loss": 0.1553,
+      "step": 9304
+    },
+    {
+      "epoch": 7.004140007527287,
+      "grad_norm": 0.1723754107952118,
+      "learning_rate": 5.5916462281206035e-06,
+      "loss": 0.1598,
+      "step": 9305
+    },
+    {
+      "epoch": 7.004892736168611,
+      "grad_norm": 0.16898663341999054,
+      "learning_rate": 5.589952348211846e-06,
+      "loss": 0.0986,
+      "step": 9306
+    },
+    {
+      "epoch": 7.005645464809936,
+      "grad_norm": 0.19343961775302887,
+      "learning_rate": 5.588260864352958e-06,
+      "loss": 0.1471,
+      "step": 9307
+    },
+    {
+      "epoch": 7.006398193451261,
+      "grad_norm": 0.18113486468791962,
+      "learning_rate": 5.586571776728916e-06,
+      "loss": 0.1768,
+      "step": 9308
+    },
+    {
+      "epoch": 7.007150922092586,
+      "grad_norm": 0.17991642653942108,
+      "learning_rate": 5.5848850855244435e-06,
+      "loss": 0.2002,
+      "step": 9309
+    },
+    {
+      "epoch": 7.00790365073391,
+      "grad_norm": 0.15144942700862885,
+      "learning_rate": 5.583200790923981e-06,
+      "loss": 0.0902,
+      "step": 9310
+    },
+    {
+      "epoch": 7.0086563793752354,
+      "grad_norm": 0.16912738978862762,
+      "learning_rate": 5.581518893111735e-06,
+      "loss": 0.1644,
+      "step": 9311
+    },
+    {
+      "epoch": 7.00940910801656,
+      "grad_norm": 0.17693501710891724,
+      "learning_rate": 5.579839392271627e-06,
+      "loss": 0.2059,
+      "step": 9312
+    },
+    {
+      "epoch": 7.010161836657884,
+      "grad_norm": 0.15419958531856537,
+      "learning_rate": 5.5781622885873265e-06,
+      "loss": 0.1854,
+      "step": 9313
+    },
+    {
+      "epoch": 7.01091456529921,
+      "grad_norm": 0.16303493082523346,
+      "learning_rate": 5.576487582242238e-06,
+      "loss": 0.1074,
+      "step": 9314
+    },
+    {
+      "epoch": 7.011667293940534,
+      "grad_norm": 0.18634925782680511,
+      "learning_rate": 5.57481527341951e-06,
+      "loss": 0.1924,
+      "step": 9315
+    },
+    {
+      "epoch": 7.01242002258186,
+      "grad_norm": 0.18575862050056458,
+      "learning_rate": 5.573145362302012e-06,
+      "loss": 0.1509,
+      "step": 9316
+    },
+    {
+      "epoch": 7.013172751223184,
+      "grad_norm": 0.18476814031600952,
+      "learning_rate": 5.571477849072378e-06,
+      "loss": 0.1643,
+      "step": 9317
+    },
+    {
+      "epoch": 7.0139254798645085,
+      "grad_norm": 0.15633293986320496,
+      "learning_rate": 5.569812733912956e-06,
+      "loss": 0.1691,
+      "step": 9318
+    },
+    {
+      "epoch": 7.014678208505834,
+      "grad_norm": 0.1887543797492981,
+      "learning_rate": 5.568150017005843e-06,
+      "loss": 0.1429,
+      "step": 9319
+    },
+    {
+      "epoch": 7.015430937147158,
+      "grad_norm": 0.20987819135189056,
+      "learning_rate": 5.5664896985328656e-06,
+      "loss": 0.2201,
+      "step": 9320
+    },
+    {
+      "epoch": 7.016183665788483,
+      "grad_norm": 0.16370825469493866,
+      "learning_rate": 5.564831778675602e-06,
+      "loss": 0.162,
+      "step": 9321
+    },
+    {
+      "epoch": 7.016936394429808,
+      "grad_norm": 0.17419888079166412,
+      "learning_rate": 5.563176257615353e-06,
+      "loss": 0.1241,
+      "step": 9322
+    },
+    {
+      "epoch": 7.017689123071133,
+      "grad_norm": 0.1658007800579071,
+      "learning_rate": 5.561523135533168e-06,
+      "loss": 0.1729,
+      "step": 9323
+    },
+    {
+      "epoch": 7.018441851712458,
+      "grad_norm": 0.1622164100408554,
+      "learning_rate": 5.559872412609829e-06,
+      "loss": 0.1041,
+      "step": 9324
+    },
+    {
+      "epoch": 7.0191945803537825,
+      "grad_norm": 0.20278915762901306,
+      "learning_rate": 5.5582240890258566e-06,
+      "loss": 0.2277,
+      "step": 9325
+    },
+    {
+      "epoch": 7.019947308995107,
+      "grad_norm": 0.1833452433347702,
+      "learning_rate": 5.556578164961511e-06,
+      "loss": 0.1314,
+      "step": 9326
+    },
+    {
+      "epoch": 7.020700037636432,
+      "grad_norm": 0.18166440725326538,
+      "learning_rate": 5.554934640596782e-06,
+      "loss": 0.176,
+      "step": 9327
+    },
+    {
+      "epoch": 7.021452766277757,
+      "grad_norm": 0.18227902054786682,
+      "learning_rate": 5.55329351611141e-06,
+      "loss": 0.1204,
+      "step": 9328
+    },
+    {
+      "epoch": 7.022205494919081,
+      "grad_norm": 0.14926744997501373,
+      "learning_rate": 5.5516547916848614e-06,
+      "loss": 0.1449,
+      "step": 9329
+    },
+    {
+      "epoch": 7.022958223560407,
+      "grad_norm": 0.18078403174877167,
+      "learning_rate": 5.550018467496344e-06,
+      "loss": 0.1075,
+      "step": 9330
+    },
+    {
+      "epoch": 7.023710952201731,
+      "grad_norm": 0.20064757764339447,
+      "learning_rate": 5.5483845437248066e-06,
+      "loss": 0.2155,
+      "step": 9331
+    },
+    {
+      "epoch": 7.0244636808430565,
+      "grad_norm": 0.1807621717453003,
+      "learning_rate": 5.5467530205489335e-06,
+      "loss": 0.2253,
+      "step": 9332
+    },
+    {
+      "epoch": 7.025216409484381,
+      "grad_norm": 0.15824201703071594,
+      "learning_rate": 5.545123898147144e-06,
+      "loss": 0.0952,
+      "step": 9333
+    },
+    {
+      "epoch": 7.0259691381257054,
+      "grad_norm": 0.1521102339029312,
+      "learning_rate": 5.543497176697597e-06,
+      "loss": 0.1166,
+      "step": 9334
+    },
+    {
+      "epoch": 7.026721866767031,
+      "grad_norm": 0.17620034515857697,
+      "learning_rate": 5.541872856378183e-06,
+      "loss": 0.2824,
+      "step": 9335
+    },
+    {
+      "epoch": 7.027474595408355,
+      "grad_norm": 0.1674203723669052,
+      "learning_rate": 5.5402509373665445e-06,
+      "loss": 0.2017,
+      "step": 9336
+    },
+    {
+      "epoch": 7.02822732404968,
+      "grad_norm": 0.17830674350261688,
+      "learning_rate": 5.5386314198400444e-06,
+      "loss": 0.1842,
+      "step": 9337
+    },
+    {
+      "epoch": 7.028980052691005,
+      "grad_norm": 0.1481676548719406,
+      "learning_rate": 5.537014303975798e-06,
+      "loss": 0.1074,
+      "step": 9338
+    },
+    {
+      "epoch": 7.02973278133233,
+      "grad_norm": 0.17086362838745117,
+      "learning_rate": 5.535399589950642e-06,
+      "loss": 0.1681,
+      "step": 9339
+    },
+    {
+      "epoch": 7.030485509973654,
+      "grad_norm": 0.18874800205230713,
+      "learning_rate": 5.533787277941167e-06,
+      "loss": 0.1918,
+      "step": 9340
+    },
+    {
+      "epoch": 7.031238238614979,
+      "grad_norm": 0.16590076684951782,
+      "learning_rate": 5.532177368123685e-06,
+      "loss": 0.1771,
+      "step": 9341
+    },
+    {
+      "epoch": 7.031990967256304,
+      "grad_norm": 0.17269964516162872,
+      "learning_rate": 5.5305698606742615e-06,
+      "loss": 0.1029,
+      "step": 9342
+    },
+    {
+      "epoch": 7.032743695897629,
+      "grad_norm": 0.17551176249980927,
+      "learning_rate": 5.528964755768688e-06,
+      "loss": 0.1567,
+      "step": 9343
+    },
+    {
+      "epoch": 7.033496424538954,
+      "grad_norm": 0.1892547607421875,
+      "learning_rate": 5.527362053582495e-06,
+      "loss": 0.1739,
+      "step": 9344
+    },
+    {
+      "epoch": 7.034249153180278,
+      "grad_norm": 0.18011914193630219,
+      "learning_rate": 5.52576175429095e-06,
+      "loss": 0.2201,
+      "step": 9345
+    },
+    {
+      "epoch": 7.035001881821604,
+      "grad_norm": 0.18024246394634247,
+      "learning_rate": 5.524163858069067e-06,
+      "loss": 0.2325,
+      "step": 9346
+    },
+    {
+      "epoch": 7.035754610462928,
+      "grad_norm": 0.16431590914726257,
+      "learning_rate": 5.522568365091582e-06,
+      "loss": 0.145,
+      "step": 9347
+    },
+    {
+      "epoch": 7.0365073391042525,
+      "grad_norm": 0.18962688744068146,
+      "learning_rate": 5.5209752755329755e-06,
+      "loss": 0.1796,
+      "step": 9348
+    },
+    {
+      "epoch": 7.037260067745578,
+      "grad_norm": 0.1919524222612381,
+      "learning_rate": 5.519384589567468e-06,
+      "loss": 0.2803,
+      "step": 9349
+    },
+    {
+      "epoch": 7.038012796386902,
+      "grad_norm": 0.1601482331752777,
+      "learning_rate": 5.517796307369017e-06,
+      "loss": 0.1066,
+      "step": 9350
+    },
+    {
+      "epoch": 7.038765525028228,
+      "grad_norm": 0.15413956344127655,
+      "learning_rate": 5.516210429111312e-06,
+      "loss": 0.0787,
+      "step": 9351
+    },
+    {
+      "epoch": 7.039518253669552,
+      "grad_norm": 0.18582424521446228,
+      "learning_rate": 5.51462695496778e-06,
+      "loss": 0.1373,
+      "step": 9352
+    },
+    {
+      "epoch": 7.040270982310877,
+      "grad_norm": 0.18082261085510254,
+      "learning_rate": 5.513045885111591e-06,
+      "loss": 0.1993,
+      "step": 9353
+    },
+    {
+      "epoch": 7.041023710952202,
+      "grad_norm": 0.1613156944513321,
+      "learning_rate": 5.511467219715647e-06,
+      "loss": 0.1996,
+      "step": 9354
+    },
+    {
+      "epoch": 7.0417764395935265,
+      "grad_norm": 0.16548344492912292,
+      "learning_rate": 5.50989095895259e-06,
+      "loss": 0.1865,
+      "step": 9355
+    },
+    {
+      "epoch": 7.042529168234851,
+      "grad_norm": 0.1880880445241928,
+      "learning_rate": 5.5083171029947934e-06,
+      "loss": 0.1035,
+      "step": 9356
+    },
+    {
+      "epoch": 7.043281896876176,
+      "grad_norm": 0.16518759727478027,
+      "learning_rate": 5.506745652014379e-06,
+      "loss": 0.1422,
+      "step": 9357
+    },
+    {
+      "epoch": 7.044034625517501,
+      "grad_norm": 0.20452642440795898,
+      "learning_rate": 5.505176606183189e-06,
+      "loss": 0.2484,
+      "step": 9358
+    },
+    {
+      "epoch": 7.044787354158826,
+      "grad_norm": 0.18157097697257996,
+      "learning_rate": 5.503609965672819e-06,
+      "loss": 0.1736,
+      "step": 9359
+    },
+    {
+      "epoch": 7.045540082800151,
+      "grad_norm": 0.17769020795822144,
+      "learning_rate": 5.50204573065459e-06,
+      "loss": 0.189,
+      "step": 9360
+    },
+    {
+      "epoch": 7.046292811441475,
+      "grad_norm": 0.1697438806295395,
+      "learning_rate": 5.500483901299569e-06,
+      "loss": 0.1259,
+      "step": 9361
+    },
+    {
+      "epoch": 7.0470455400828005,
+      "grad_norm": 0.19293992221355438,
+      "learning_rate": 5.498924477778551e-06,
+      "loss": 0.1699,
+      "step": 9362
+    },
+    {
+      "epoch": 7.047798268724125,
+      "grad_norm": 0.16893503069877625,
+      "learning_rate": 5.497367460262075e-06,
+      "loss": 0.1252,
+      "step": 9363
+    },
+    {
+      "epoch": 7.048550997365449,
+      "grad_norm": 0.16943363845348358,
+      "learning_rate": 5.495812848920411e-06,
+      "loss": 0.1833,
+      "step": 9364
+    },
+    {
+      "epoch": 7.049303726006775,
+      "grad_norm": 0.1899724304676056,
+      "learning_rate": 5.494260643923576e-06,
+      "loss": 0.1134,
+      "step": 9365
+    },
+    {
+      "epoch": 7.050056454648099,
+      "grad_norm": 0.16093717515468597,
+      "learning_rate": 5.492710845441306e-06,
+      "loss": 0.155,
+      "step": 9366
+    },
+    {
+      "epoch": 7.050809183289424,
+      "grad_norm": 0.17229433357715607,
+      "learning_rate": 5.491163453643093e-06,
+      "loss": 0.1926,
+      "step": 9367
+    },
+    {
+      "epoch": 7.051561911930749,
+      "grad_norm": 0.1782997101545334,
+      "learning_rate": 5.489618468698154e-06,
+      "loss": 0.2212,
+      "step": 9368
+    },
+    {
+      "epoch": 7.052314640572074,
+      "grad_norm": 0.16203738749027252,
+      "learning_rate": 5.4880758907754515e-06,
+      "loss": 0.138,
+      "step": 9369
+    },
+    {
+      "epoch": 7.053067369213399,
+      "grad_norm": 0.18872782588005066,
+      "learning_rate": 5.486535720043671e-06,
+      "loss": 0.2337,
+      "step": 9370
+    },
+    {
+      "epoch": 7.053820097854723,
+      "grad_norm": 0.17964059114456177,
+      "learning_rate": 5.484997956671249e-06,
+      "loss": 0.2012,
+      "step": 9371
+    },
+    {
+      "epoch": 7.054572826496048,
+      "grad_norm": 0.18805034458637238,
+      "learning_rate": 5.48346260082635e-06,
+      "loss": 0.1805,
+      "step": 9372
+    },
+    {
+      "epoch": 7.055325555137373,
+      "grad_norm": 0.1607411950826645,
+      "learning_rate": 5.481929652676879e-06,
+      "loss": 0.1496,
+      "step": 9373
+    },
+    {
+      "epoch": 7.056078283778698,
+      "grad_norm": 0.1544840782880783,
+      "learning_rate": 5.4803991123904765e-06,
+      "loss": 0.0981,
+      "step": 9374
+    },
+    {
+      "epoch": 7.056831012420022,
+      "grad_norm": 0.1723361313343048,
+      "learning_rate": 5.4788709801345244e-06,
+      "loss": 0.1595,
+      "step": 9375
+    },
+    {
+      "epoch": 7.057583741061348,
+      "grad_norm": 0.17469173669815063,
+      "learning_rate": 5.477345256076135e-06,
+      "loss": 0.1565,
+      "step": 9376
+    },
+    {
+      "epoch": 7.058336469702672,
+      "grad_norm": 0.16086764633655548,
+      "learning_rate": 5.475821940382154e-06,
+      "loss": 0.1866,
+      "step": 9377
+    },
+    {
+      "epoch": 7.059089198343997,
+      "grad_norm": 0.17813463509082794,
+      "learning_rate": 5.474301033219176e-06,
+      "loss": 0.1428,
+      "step": 9378
+    },
+    {
+      "epoch": 7.059841926985322,
+      "grad_norm": 0.1688918173313141,
+      "learning_rate": 5.472782534753516e-06,
+      "loss": 0.208,
+      "step": 9379
+    },
+    {
+      "epoch": 7.060594655626646,
+      "grad_norm": 0.17677821218967438,
+      "learning_rate": 5.471266445151246e-06,
+      "loss": 0.1215,
+      "step": 9380
+    },
+    {
+      "epoch": 7.061347384267972,
+      "grad_norm": 0.17869319021701813,
+      "learning_rate": 5.469752764578154e-06,
+      "loss": 0.2057,
+      "step": 9381
+    },
+    {
+      "epoch": 7.062100112909296,
+      "grad_norm": 0.1812962293624878,
+      "learning_rate": 5.468241493199781e-06,
+      "loss": 0.1943,
+      "step": 9382
+    },
+    {
+      "epoch": 7.062852841550621,
+      "grad_norm": 0.17150941491127014,
+      "learning_rate": 5.466732631181395e-06,
+      "loss": 0.1358,
+      "step": 9383
+    },
+    {
+      "epoch": 7.063605570191946,
+      "grad_norm": 0.17403459548950195,
+      "learning_rate": 5.4652261786880005e-06,
+      "loss": 0.1356,
+      "step": 9384
+    },
+    {
+      "epoch": 7.0643582988332705,
+      "grad_norm": 0.17356498539447784,
+      "learning_rate": 5.463722135884342e-06,
+      "loss": 0.1291,
+      "step": 9385
+    },
+    {
+      "epoch": 7.065111027474595,
+      "grad_norm": 0.18072207272052765,
+      "learning_rate": 5.462220502934898e-06,
+      "loss": 0.191,
+      "step": 9386
+    },
+    {
+      "epoch": 7.06586375611592,
+      "grad_norm": 0.17007389664649963,
+      "learning_rate": 5.460721280003887e-06,
+      "loss": 0.1719,
+      "step": 9387
+    },
+    {
+      "epoch": 7.066616484757245,
+      "grad_norm": 0.15931229293346405,
+      "learning_rate": 5.459224467255264e-06,
+      "loss": 0.1393,
+      "step": 9388
+    },
+    {
+      "epoch": 7.06736921339857,
+      "grad_norm": 0.1810770481824875,
+      "learning_rate": 5.457730064852714e-06,
+      "loss": 0.2315,
+      "step": 9389
+    },
+    {
+      "epoch": 7.068121942039895,
+      "grad_norm": 0.15328064560890198,
+      "learning_rate": 5.456238072959664e-06,
+      "loss": 0.1,
+      "step": 9390
+    },
+    {
+      "epoch": 7.068874670681219,
+      "grad_norm": 0.18390515446662903,
+      "learning_rate": 5.454748491739275e-06,
+      "loss": 0.2345,
+      "step": 9391
+    },
+    {
+      "epoch": 7.0696273993225445,
+      "grad_norm": 0.1624535173177719,
+      "learning_rate": 5.453261321354447e-06,
+      "loss": 0.1122,
+      "step": 9392
+    },
+    {
+      "epoch": 7.070380127963869,
+      "grad_norm": 0.17523403465747833,
+      "learning_rate": 5.451776561967812e-06,
+      "loss": 0.2036,
+      "step": 9393
+    },
+    {
+      "epoch": 7.071132856605193,
+      "grad_norm": 0.16982907056808472,
+      "learning_rate": 5.450294213741748e-06,
+      "loss": 0.1682,
+      "step": 9394
+    },
+    {
+      "epoch": 7.071885585246519,
+      "grad_norm": 0.16768120229244232,
+      "learning_rate": 5.448814276838354e-06,
+      "loss": 0.1814,
+      "step": 9395
+    },
+    {
+      "epoch": 7.072638313887843,
+      "grad_norm": 0.19140028953552246,
+      "learning_rate": 5.447336751419478e-06,
+      "loss": 0.2308,
+      "step": 9396
+    },
+    {
+      "epoch": 7.073391042529169,
+      "grad_norm": 0.15683558583259583,
+      "learning_rate": 5.4458616376467e-06,
+      "loss": 0.1264,
+      "step": 9397
+    },
+    {
+      "epoch": 7.074143771170493,
+      "grad_norm": 0.1950828731060028,
+      "learning_rate": 5.444388935681336e-06,
+      "loss": 0.1626,
+      "step": 9398
+    },
+    {
+      "epoch": 7.074896499811818,
+      "grad_norm": 0.17747320234775543,
+      "learning_rate": 5.442918645684434e-06,
+      "loss": 0.1555,
+      "step": 9399
+    },
+    {
+      "epoch": 7.075649228453143,
+      "grad_norm": 0.15364092588424683,
+      "learning_rate": 5.44145076781679e-06,
+      "loss": 0.1081,
+      "step": 9400
+    },
+    {
+      "epoch": 7.075649228453143,
+      "eval_loss": 0.20089496672153473,
+      "eval_runtime": 452.8685,
+      "eval_samples_per_second": 21.258,
+      "eval_steps_per_second": 0.665,
+      "step": 9400
+    },
+    {
+      "epoch": 7.076401957094467,
+      "grad_norm": 0.1607252061367035,
+      "learning_rate": 5.439985302238924e-06,
+      "loss": 0.1004,
+      "step": 9401
+    },
+    {
+      "epoch": 7.077154685735792,
+      "grad_norm": 0.1780771017074585,
+      "learning_rate": 5.438522249111099e-06,
+      "loss": 0.1724,
+      "step": 9402
+    },
+    {
+      "epoch": 7.077907414377117,
+      "grad_norm": 0.18484535813331604,
+      "learning_rate": 5.437061608593315e-06,
+      "loss": 0.1966,
+      "step": 9403
+    },
+    {
+      "epoch": 7.078660143018442,
+      "grad_norm": 0.19618922472000122,
+      "learning_rate": 5.435603380845298e-06,
+      "loss": 0.2079,
+      "step": 9404
+    },
+    {
+      "epoch": 7.079412871659767,
+      "grad_norm": 0.15975233912467957,
+      "learning_rate": 5.434147566026524e-06,
+      "loss": 0.2004,
+      "step": 9405
+    },
+    {
+      "epoch": 7.0801656003010915,
+      "grad_norm": 0.16689433157444,
+      "learning_rate": 5.432694164296195e-06,
+      "loss": 0.1823,
+      "step": 9406
+    },
+    {
+      "epoch": 7.080918328942416,
+      "grad_norm": 0.17345786094665527,
+      "learning_rate": 5.431243175813258e-06,
+      "loss": 0.1422,
+      "step": 9407
+    },
+    {
+      "epoch": 7.081671057583741,
+      "grad_norm": 0.16389131546020508,
+      "learning_rate": 5.429794600736382e-06,
+      "loss": 0.2032,
+      "step": 9408
+    },
+    {
+      "epoch": 7.082423786225066,
+      "grad_norm": 0.18728072941303253,
+      "learning_rate": 5.428348439223993e-06,
+      "loss": 0.1104,
+      "step": 9409
+    },
+    {
+      "epoch": 7.08317651486639,
+      "grad_norm": 0.17989712953567505,
+      "learning_rate": 5.4269046914342304e-06,
+      "loss": 0.191,
+      "step": 9410
+    },
+    {
+      "epoch": 7.083929243507716,
+      "grad_norm": 0.1714271903038025,
+      "learning_rate": 5.425463357524986e-06,
+      "loss": 0.1916,
+      "step": 9411
+    },
+    {
+      "epoch": 7.08468197214904,
+      "grad_norm": 0.1867133229970932,
+      "learning_rate": 5.424024437653876e-06,
+      "loss": 0.2578,
+      "step": 9412
+    },
+    {
+      "epoch": 7.085434700790365,
+      "grad_norm": 0.15716256201267242,
+      "learning_rate": 5.422587931978266e-06,
+      "loss": 0.1797,
+      "step": 9413
+    },
+    {
+      "epoch": 7.08618742943169,
+      "grad_norm": 0.16613735258579254,
+      "learning_rate": 5.421153840655248e-06,
+      "loss": 0.1926,
+      "step": 9414
+    },
+    {
+      "epoch": 7.0869401580730145,
+      "grad_norm": 0.17237091064453125,
+      "learning_rate": 5.419722163841652e-06,
+      "loss": 0.1651,
+      "step": 9415
+    },
+    {
+      "epoch": 7.08769288671434,
+      "grad_norm": 0.17974120378494263,
+      "learning_rate": 5.4182929016940375e-06,
+      "loss": 0.2417,
+      "step": 9416
+    },
+    {
+      "epoch": 7.088445615355664,
+      "grad_norm": 0.16703961789608002,
+      "learning_rate": 5.416866054368718e-06,
+      "loss": 0.1186,
+      "step": 9417
+    },
+    {
+      "epoch": 7.089198343996989,
+      "grad_norm": 0.18551835417747498,
+      "learning_rate": 5.415441622021718e-06,
+      "loss": 0.1338,
+      "step": 9418
+    },
+    {
+      "epoch": 7.089951072638314,
+      "grad_norm": 0.1626303642988205,
+      "learning_rate": 5.414019604808826e-06,
+      "loss": 0.0918,
+      "step": 9419
+    },
+    {
+      "epoch": 7.090703801279639,
+      "grad_norm": 0.16569682955741882,
+      "learning_rate": 5.41260000288554e-06,
+      "loss": 0.1462,
+      "step": 9420
+    },
+    {
+      "epoch": 7.091456529920963,
+      "grad_norm": 0.15986298024654388,
+      "learning_rate": 5.411182816407113e-06,
+      "loss": 0.0848,
+      "step": 9421
+    },
+    {
+      "epoch": 7.0922092585622885,
+      "grad_norm": 0.15576709806919098,
+      "learning_rate": 5.409768045528521e-06,
+      "loss": 0.1528,
+      "step": 9422
+    },
+    {
+      "epoch": 7.092961987203613,
+      "grad_norm": 0.1761845350265503,
+      "learning_rate": 5.408355690404483e-06,
+      "loss": 0.1608,
+      "step": 9423
+    },
+    {
+      "epoch": 7.093714715844938,
+      "grad_norm": 0.17797081172466278,
+      "learning_rate": 5.40694575118945e-06,
+      "loss": 0.2171,
+      "step": 9424
+    },
+    {
+      "epoch": 7.094467444486263,
+      "grad_norm": 0.17349646985530853,
+      "learning_rate": 5.40553822803762e-06,
+      "loss": 0.1506,
+      "step": 9425
+    },
+    {
+      "epoch": 7.095220173127587,
+      "grad_norm": 0.1818084567785263,
+      "learning_rate": 5.404133121102906e-06,
+      "loss": 0.1473,
+      "step": 9426
+    },
+    {
+      "epoch": 7.095972901768913,
+      "grad_norm": 0.18147067725658417,
+      "learning_rate": 5.402730430538972e-06,
+      "loss": 0.1419,
+      "step": 9427
+    },
+    {
+      "epoch": 7.096725630410237,
+      "grad_norm": 0.1472322791814804,
+      "learning_rate": 5.401330156499219e-06,
+      "loss": 0.1356,
+      "step": 9428
+    },
+    {
+      "epoch": 7.0974783590515615,
+      "grad_norm": 0.18753895163536072,
+      "learning_rate": 5.399932299136775e-06,
+      "loss": 0.2007,
+      "step": 9429
+    },
+    {
+      "epoch": 7.098231087692887,
+      "grad_norm": 0.16332972049713135,
+      "learning_rate": 5.398536858604507e-06,
+      "loss": 0.109,
+      "step": 9430
+    },
+    {
+      "epoch": 7.098983816334211,
+      "grad_norm": 0.17688313126564026,
+      "learning_rate": 5.397143835055015e-06,
+      "loss": 0.166,
+      "step": 9431
+    },
+    {
+      "epoch": 7.099736544975537,
+      "grad_norm": 0.18036213517189026,
+      "learning_rate": 5.395753228640648e-06,
+      "loss": 0.1511,
+      "step": 9432
+    },
+    {
+      "epoch": 7.100489273616861,
+      "grad_norm": 0.21243293583393097,
+      "learning_rate": 5.3943650395134725e-06,
+      "loss": 0.2216,
+      "step": 9433
+    },
+    {
+      "epoch": 7.101242002258186,
+      "grad_norm": 0.17012172937393188,
+      "learning_rate": 5.392979267825303e-06,
+      "loss": 0.1553,
+      "step": 9434
+    },
+    {
+      "epoch": 7.101994730899511,
+      "grad_norm": 0.1817018687725067,
+      "learning_rate": 5.39159591372768e-06,
+      "loss": 0.2245,
+      "step": 9435
+    },
+    {
+      "epoch": 7.1027474595408355,
+      "grad_norm": 0.18702681362628937,
+      "learning_rate": 5.390214977371889e-06,
+      "loss": 0.2031,
+      "step": 9436
+    },
+    {
+      "epoch": 7.10350018818216,
+      "grad_norm": 0.15625205636024475,
+      "learning_rate": 5.388836458908947e-06,
+      "loss": 0.1245,
+      "step": 9437
+    },
+    {
+      "epoch": 7.104252916823485,
+      "grad_norm": 0.15788534283638,
+      "learning_rate": 5.387460358489607e-06,
+      "loss": 0.123,
+      "step": 9438
+    },
+    {
+      "epoch": 7.10500564546481,
+      "grad_norm": 0.17690280079841614,
+      "learning_rate": 5.386086676264355e-06,
+      "loss": 0.1829,
+      "step": 9439
+    },
+    {
+      "epoch": 7.105758374106134,
+      "grad_norm": 0.18451355397701263,
+      "learning_rate": 5.384715412383418e-06,
+      "loss": 0.1577,
+      "step": 9440
+    },
+    {
+      "epoch": 7.10651110274746,
+      "grad_norm": 0.19901305437088013,
+      "learning_rate": 5.383346566996752e-06,
+      "loss": 0.1911,
+      "step": 9441
+    },
+    {
+      "epoch": 7.107263831388784,
+      "grad_norm": 0.17947033047676086,
+      "learning_rate": 5.381980140254057e-06,
+      "loss": 0.19,
+      "step": 9442
+    },
+    {
+      "epoch": 7.1080165600301095,
+      "grad_norm": 0.1597910076379776,
+      "learning_rate": 5.3806161323047526e-06,
+      "loss": 0.1532,
+      "step": 9443
+    },
+    {
+      "epoch": 7.108769288671434,
+      "grad_norm": 0.17439018189907074,
+      "learning_rate": 5.37925454329802e-06,
+      "loss": 0.1055,
+      "step": 9444
+    },
+    {
+      "epoch": 7.1095220173127585,
+      "grad_norm": 0.20591242611408234,
+      "learning_rate": 5.3778953733827495e-06,
+      "loss": 0.2081,
+      "step": 9445
+    },
+    {
+      "epoch": 7.110274745954084,
+      "grad_norm": 0.1862751990556717,
+      "learning_rate": 5.376538622707581e-06,
+      "loss": 0.1696,
+      "step": 9446
+    },
+    {
+      "epoch": 7.111027474595408,
+      "grad_norm": 0.19853584468364716,
+      "learning_rate": 5.375184291420886e-06,
+      "loss": 0.2742,
+      "step": 9447
+    },
+    {
+      "epoch": 7.111780203236733,
+      "grad_norm": 0.1724155694246292,
+      "learning_rate": 5.37383237967077e-06,
+      "loss": 0.156,
+      "step": 9448
+    },
+    {
+      "epoch": 7.112532931878058,
+      "grad_norm": 0.17883935570716858,
+      "learning_rate": 5.372482887605084e-06,
+      "loss": 0.2254,
+      "step": 9449
+    },
+    {
+      "epoch": 7.113285660519383,
+      "grad_norm": 0.16639521718025208,
+      "learning_rate": 5.371135815371398e-06,
+      "loss": 0.1246,
+      "step": 9450
+    },
+    {
+      "epoch": 7.114038389160708,
+      "grad_norm": 0.15507738292217255,
+      "learning_rate": 5.36979116311703e-06,
+      "loss": 0.1171,
+      "step": 9451
+    },
+    {
+      "epoch": 7.114791117802032,
+      "grad_norm": 0.18975435197353363,
+      "learning_rate": 5.3684489309890265e-06,
+      "loss": 0.1668,
+      "step": 9452
+    },
+    {
+      "epoch": 7.115543846443357,
+      "grad_norm": 0.17154189944267273,
+      "learning_rate": 5.367109119134177e-06,
+      "loss": 0.096,
+      "step": 9453
+    },
+    {
+      "epoch": 7.116296575084682,
+      "grad_norm": 0.20695990324020386,
+      "learning_rate": 5.365771727698996e-06,
+      "loss": 0.2238,
+      "step": 9454
+    },
+    {
+      "epoch": 7.117049303726007,
+      "grad_norm": 0.157502219080925,
+      "learning_rate": 5.36443675682974e-06,
+      "loss": 0.1429,
+      "step": 9455
+    },
+    {
+      "epoch": 7.117802032367331,
+      "grad_norm": 0.1613231897354126,
+      "learning_rate": 5.363104206672399e-06,
+      "loss": 0.1526,
+      "step": 9456
+    },
+    {
+      "epoch": 7.118554761008657,
+      "grad_norm": 0.17134562134742737,
+      "learning_rate": 5.361774077372703e-06,
+      "loss": 0.0883,
+      "step": 9457
+    },
+    {
+      "epoch": 7.119307489649981,
+      "grad_norm": 0.1746046543121338,
+      "learning_rate": 5.360446369076107e-06,
+      "loss": 0.1348,
+      "step": 9458
+    },
+    {
+      "epoch": 7.120060218291306,
+      "grad_norm": 0.17342586815357208,
+      "learning_rate": 5.359121081927811e-06,
+      "loss": 0.1915,
+      "step": 9459
+    },
+    {
+      "epoch": 7.120812946932631,
+      "grad_norm": 0.17482800781726837,
+      "learning_rate": 5.357798216072743e-06,
+      "loss": 0.1541,
+      "step": 9460
+    },
+    {
+      "epoch": 7.121565675573955,
+      "grad_norm": 0.16541661322116852,
+      "learning_rate": 5.356477771655576e-06,
+      "loss": 0.123,
+      "step": 9461
+    },
+    {
+      "epoch": 7.122318404215281,
+      "grad_norm": 0.17010317742824554,
+      "learning_rate": 5.355159748820702e-06,
+      "loss": 0.1233,
+      "step": 9462
+    },
+    {
+      "epoch": 7.123071132856605,
+      "grad_norm": 0.1921461969614029,
+      "learning_rate": 5.353844147712268e-06,
+      "loss": 0.1861,
+      "step": 9463
+    },
+    {
+      "epoch": 7.12382386149793,
+      "grad_norm": 0.1657104194164276,
+      "learning_rate": 5.352530968474139e-06,
+      "loss": 0.2046,
+      "step": 9464
+    },
+    {
+      "epoch": 7.124576590139255,
+      "grad_norm": 0.14743275940418243,
+      "learning_rate": 5.3512202112499275e-06,
+      "loss": 0.1024,
+      "step": 9465
+    },
+    {
+      "epoch": 7.1253293187805795,
+      "grad_norm": 0.17397698760032654,
+      "learning_rate": 5.3499118761829715e-06,
+      "loss": 0.16,
+      "step": 9466
+    },
+    {
+      "epoch": 7.126082047421904,
+      "grad_norm": 0.16333118081092834,
+      "learning_rate": 5.3486059634163544e-06,
+      "loss": 0.1126,
+      "step": 9467
+    },
+    {
+      "epoch": 7.126834776063229,
+      "grad_norm": 0.1871376782655716,
+      "learning_rate": 5.347302473092881e-06,
+      "loss": 0.2319,
+      "step": 9468
+    },
+    {
+      "epoch": 7.127587504704554,
+      "grad_norm": 0.15200571715831757,
+      "learning_rate": 5.346001405355104e-06,
+      "loss": 0.1543,
+      "step": 9469
+    },
+    {
+      "epoch": 7.128340233345879,
+      "grad_norm": 0.1607501357793808,
+      "learning_rate": 5.344702760345303e-06,
+      "loss": 0.1921,
+      "step": 9470
+    },
+    {
+      "epoch": 7.129092961987204,
+      "grad_norm": 0.17596352100372314,
+      "learning_rate": 5.343406538205502e-06,
+      "loss": 0.1378,
+      "step": 9471
+    },
+    {
+      "epoch": 7.129845690628528,
+      "grad_norm": 0.16849951446056366,
+      "learning_rate": 5.342112739077449e-06,
+      "loss": 0.15,
+      "step": 9472
+    },
+    {
+      "epoch": 7.1305984192698535,
+      "grad_norm": 0.16680613160133362,
+      "learning_rate": 5.340821363102632e-06,
+      "loss": 0.1357,
+      "step": 9473
+    },
+    {
+      "epoch": 7.131351147911178,
+      "grad_norm": 0.1678377389907837,
+      "learning_rate": 5.339532410422277e-06,
+      "loss": 0.1445,
+      "step": 9474
+    },
+    {
+      "epoch": 7.132103876552502,
+      "grad_norm": 0.17459125816822052,
+      "learning_rate": 5.338245881177338e-06,
+      "loss": 0.2147,
+      "step": 9475
+    },
+    {
+      "epoch": 7.132856605193828,
+      "grad_norm": 0.1488466113805771,
+      "learning_rate": 5.3369617755085084e-06,
+      "loss": 0.1633,
+      "step": 9476
+    },
+    {
+      "epoch": 7.133609333835152,
+      "grad_norm": 0.16518634557724,
+      "learning_rate": 5.335680093556218e-06,
+      "loss": 0.1689,
+      "step": 9477
+    },
+    {
+      "epoch": 7.134362062476478,
+      "grad_norm": 0.1843031942844391,
+      "learning_rate": 5.334400835460631e-06,
+      "loss": 0.1622,
+      "step": 9478
+    },
+    {
+      "epoch": 7.135114791117802,
+      "grad_norm": 0.17122811079025269,
+      "learning_rate": 5.333124001361642e-06,
+      "loss": 0.1276,
+      "step": 9479
+    },
+    {
+      "epoch": 7.135867519759127,
+      "grad_norm": 0.14966925978660583,
+      "learning_rate": 5.331849591398886e-06,
+      "loss": 0.1146,
+      "step": 9480
+    },
+    {
+      "epoch": 7.136620248400452,
+      "grad_norm": 0.1894652396440506,
+      "learning_rate": 5.330577605711729e-06,
+      "loss": 0.1495,
+      "step": 9481
+    },
+    {
+      "epoch": 7.137372977041776,
+      "grad_norm": 0.16068127751350403,
+      "learning_rate": 5.329308044439272e-06,
+      "loss": 0.1429,
+      "step": 9482
+    },
+    {
+      "epoch": 7.138125705683101,
+      "grad_norm": 0.17114008963108063,
+      "learning_rate": 5.328040907720355e-06,
+      "loss": 0.153,
+      "step": 9483
+    },
+    {
+      "epoch": 7.138878434324426,
+      "grad_norm": 0.16028067469596863,
+      "learning_rate": 5.3267761956935494e-06,
+      "loss": 0.1633,
+      "step": 9484
+    },
+    {
+      "epoch": 7.139631162965751,
+      "grad_norm": 0.19149629771709442,
+      "learning_rate": 5.325513908497161e-06,
+      "loss": 0.1831,
+      "step": 9485
+    },
+    {
+      "epoch": 7.140383891607076,
+      "grad_norm": 0.19049705564975739,
+      "learning_rate": 5.3242540462692366e-06,
+      "loss": 0.1802,
+      "step": 9486
+    },
+    {
+      "epoch": 7.141136620248401,
+      "grad_norm": 0.16764770448207855,
+      "learning_rate": 5.322996609147544e-06,
+      "loss": 0.1213,
+      "step": 9487
+    },
+    {
+      "epoch": 7.141889348889725,
+      "grad_norm": 0.17411957681179047,
+      "learning_rate": 5.321741597269603e-06,
+      "loss": 0.1486,
+      "step": 9488
+    },
+    {
+      "epoch": 7.14264207753105,
+      "grad_norm": 0.17023666203022003,
+      "learning_rate": 5.320489010772654e-06,
+      "loss": 0.146,
+      "step": 9489
+    },
+    {
+      "epoch": 7.143394806172375,
+      "grad_norm": 0.16776102781295776,
+      "learning_rate": 5.319238849793683e-06,
+      "loss": 0.1604,
+      "step": 9490
+    },
+    {
+      "epoch": 7.144147534813699,
+      "grad_norm": 0.1911868453025818,
+      "learning_rate": 5.317991114469402e-06,
+      "loss": 0.2454,
+      "step": 9491
+    },
+    {
+      "epoch": 7.144900263455025,
+      "grad_norm": 0.16877219080924988,
+      "learning_rate": 5.316745804936264e-06,
+      "loss": 0.1842,
+      "step": 9492
+    },
+    {
+      "epoch": 7.145652992096349,
+      "grad_norm": 0.1780957281589508,
+      "learning_rate": 5.31550292133045e-06,
+      "loss": 0.2943,
+      "step": 9493
+    },
+    {
+      "epoch": 7.146405720737674,
+      "grad_norm": 0.14058493077754974,
+      "learning_rate": 5.314262463787883e-06,
+      "loss": 0.1189,
+      "step": 9494
+    },
+    {
+      "epoch": 7.147158449378999,
+      "grad_norm": 0.16874366998672485,
+      "learning_rate": 5.3130244324442174e-06,
+      "loss": 0.1726,
+      "step": 9495
+    },
+    {
+      "epoch": 7.1479111780203235,
+      "grad_norm": 0.17766699194908142,
+      "learning_rate": 5.311788827434845e-06,
+      "loss": 0.1367,
+      "step": 9496
+    },
+    {
+      "epoch": 7.148663906661649,
+      "grad_norm": 0.15857580304145813,
+      "learning_rate": 5.310555648894883e-06,
+      "loss": 0.1694,
+      "step": 9497
+    },
+    {
+      "epoch": 7.149416635302973,
+      "grad_norm": 0.16917510330677032,
+      "learning_rate": 5.309324896959195e-06,
+      "loss": 0.1471,
+      "step": 9498
+    },
+    {
+      "epoch": 7.150169363944298,
+      "grad_norm": 0.1758975088596344,
+      "learning_rate": 5.308096571762374e-06,
+      "loss": 0.1655,
+      "step": 9499
+    },
+    {
+      "epoch": 7.150922092585623,
+      "grad_norm": 0.1953122913837433,
+      "learning_rate": 5.3068706734387484e-06,
+      "loss": 0.1884,
+      "step": 9500
+    },
+    {
+      "epoch": 7.151674821226948,
+      "grad_norm": 0.1798885017633438,
+      "learning_rate": 5.305647202122375e-06,
+      "loss": 0.1619,
+      "step": 9501
+    },
+    {
+      "epoch": 7.152427549868272,
+      "grad_norm": 0.1928894966840744,
+      "learning_rate": 5.304426157947058e-06,
+      "loss": 0.1604,
+      "step": 9502
+    },
+    {
+      "epoch": 7.1531802785095975,
+      "grad_norm": 0.17254605889320374,
+      "learning_rate": 5.303207541046326e-06,
+      "loss": 0.2613,
+      "step": 9503
+    },
+    {
+      "epoch": 7.153933007150922,
+      "grad_norm": 0.17996151745319366,
+      "learning_rate": 5.301991351553444e-06,
+      "loss": 0.1391,
+      "step": 9504
+    },
+    {
+      "epoch": 7.154685735792247,
+      "grad_norm": 0.17072172462940216,
+      "learning_rate": 5.300777589601414e-06,
+      "loss": 0.1497,
+      "step": 9505
+    },
+    {
+      "epoch": 7.155438464433572,
+      "grad_norm": 0.18690721690654755,
+      "learning_rate": 5.2995662553229696e-06,
+      "loss": 0.1648,
+      "step": 9506
+    },
+    {
+      "epoch": 7.156191193074896,
+      "grad_norm": 0.153951033949852,
+      "learning_rate": 5.298357348850582e-06,
+      "loss": 0.1314,
+      "step": 9507
+    },
+    {
+      "epoch": 7.156943921716222,
+      "grad_norm": 0.17120668292045593,
+      "learning_rate": 5.297150870316455e-06,
+      "loss": 0.1341,
+      "step": 9508
+    },
+    {
+      "epoch": 7.157696650357546,
+      "grad_norm": 0.1578524261713028,
+      "learning_rate": 5.295946819852528e-06,
+      "loss": 0.1659,
+      "step": 9509
+    },
+    {
+      "epoch": 7.158449378998871,
+      "grad_norm": 0.18280094861984253,
+      "learning_rate": 5.2947451975904724e-06,
+      "loss": 0.1562,
+      "step": 9510
+    },
+    {
+      "epoch": 7.159202107640196,
+      "grad_norm": 0.16615550220012665,
+      "learning_rate": 5.2935460036617e-06,
+      "loss": 0.1285,
+      "step": 9511
+    },
+    {
+      "epoch": 7.15995483628152,
+      "grad_norm": 0.17711415886878967,
+      "learning_rate": 5.292349238197346e-06,
+      "loss": 0.1165,
+      "step": 9512
+    },
+    {
+      "epoch": 7.160707564922845,
+      "grad_norm": 0.1750057339668274,
+      "learning_rate": 5.291154901328293e-06,
+      "loss": 0.1741,
+      "step": 9513
+    },
+    {
+      "epoch": 7.16146029356417,
+      "grad_norm": 0.18474334478378296,
+      "learning_rate": 5.289962993185147e-06,
+      "loss": 0.1422,
+      "step": 9514
+    },
+    {
+      "epoch": 7.162213022205495,
+      "grad_norm": 0.19896110892295837,
+      "learning_rate": 5.28877351389826e-06,
+      "loss": 0.1572,
+      "step": 9515
+    },
+    {
+      "epoch": 7.16296575084682,
+      "grad_norm": 0.1540241688489914,
+      "learning_rate": 5.287586463597702e-06,
+      "loss": 0.1383,
+      "step": 9516
+    },
+    {
+      "epoch": 7.1637184794881446,
+      "grad_norm": 0.19747938215732574,
+      "learning_rate": 5.286401842413298e-06,
+      "loss": 0.2171,
+      "step": 9517
+    },
+    {
+      "epoch": 7.164471208129469,
+      "grad_norm": 0.19082152843475342,
+      "learning_rate": 5.2852196504745855e-06,
+      "loss": 0.1896,
+      "step": 9518
+    },
+    {
+      "epoch": 7.165223936770794,
+      "grad_norm": 0.17522890865802765,
+      "learning_rate": 5.284039887910857e-06,
+      "loss": 0.0731,
+      "step": 9519
+    },
+    {
+      "epoch": 7.165976665412119,
+      "grad_norm": 0.1754325032234192,
+      "learning_rate": 5.282862554851122e-06,
+      "loss": 0.1697,
+      "step": 9520
+    },
+    {
+      "epoch": 7.166729394053443,
+      "grad_norm": 0.17594395577907562,
+      "learning_rate": 5.281687651424138e-06,
+      "loss": 0.1557,
+      "step": 9521
+    },
+    {
+      "epoch": 7.167482122694769,
+      "grad_norm": 0.16388164460659027,
+      "learning_rate": 5.280515177758385e-06,
+      "loss": 0.183,
+      "step": 9522
+    },
+    {
+      "epoch": 7.168234851336093,
+      "grad_norm": 0.1743902862071991,
+      "learning_rate": 5.279345133982086e-06,
+      "loss": 0.1938,
+      "step": 9523
+    },
+    {
+      "epoch": 7.1689875799774185,
+      "grad_norm": 0.18898531794548035,
+      "learning_rate": 5.278177520223197e-06,
+      "loss": 0.1553,
+      "step": 9524
+    },
+    {
+      "epoch": 7.169740308618743,
+      "grad_norm": 0.1683388501405716,
+      "learning_rate": 5.277012336609403e-06,
+      "loss": 0.199,
+      "step": 9525
+    },
+    {
+      "epoch": 7.1704930372600675,
+      "grad_norm": 0.17419223487377167,
+      "learning_rate": 5.275849583268128e-06,
+      "loss": 0.2144,
+      "step": 9526
+    },
+    {
+      "epoch": 7.171245765901393,
+      "grad_norm": 0.16675056517124176,
+      "learning_rate": 5.274689260326529e-06,
+      "loss": 0.1754,
+      "step": 9527
+    },
+    {
+      "epoch": 7.171998494542717,
+      "grad_norm": 0.1836860030889511,
+      "learning_rate": 5.273531367911497e-06,
+      "loss": 0.1409,
+      "step": 9528
+    },
+    {
+      "epoch": 7.172751223184042,
+      "grad_norm": 0.20078541338443756,
+      "learning_rate": 5.272375906149657e-06,
+      "loss": 0.1576,
+      "step": 9529
+    },
+    {
+      "epoch": 7.173503951825367,
+      "grad_norm": 0.1910555064678192,
+      "learning_rate": 5.271222875167371e-06,
+      "loss": 0.115,
+      "step": 9530
+    },
+    {
+      "epoch": 7.174256680466692,
+      "grad_norm": 0.15633803606033325,
+      "learning_rate": 5.27007227509073e-06,
+      "loss": 0.1218,
+      "step": 9531
+    },
+    {
+      "epoch": 7.175009409108017,
+      "grad_norm": 0.15108543634414673,
+      "learning_rate": 5.268924106045563e-06,
+      "loss": 0.1236,
+      "step": 9532
+    },
+    {
+      "epoch": 7.1757621377493415,
+      "grad_norm": 0.19474563002586365,
+      "learning_rate": 5.267778368157432e-06,
+      "loss": 0.1707,
+      "step": 9533
+    },
+    {
+      "epoch": 7.176514866390666,
+      "grad_norm": 0.18451333045959473,
+      "learning_rate": 5.2666350615516325e-06,
+      "loss": 0.12,
+      "step": 9534
+    },
+    {
+      "epoch": 7.177267595031991,
+      "grad_norm": 0.16751253604888916,
+      "learning_rate": 5.265494186353195e-06,
+      "loss": 0.1631,
+      "step": 9535
+    },
+    {
+      "epoch": 7.178020323673316,
+      "grad_norm": 0.14936354756355286,
+      "learning_rate": 5.264355742686885e-06,
+      "loss": 0.0785,
+      "step": 9536
+    },
+    {
+      "epoch": 7.17877305231464,
+      "grad_norm": 0.1798544079065323,
+      "learning_rate": 5.2632197306772e-06,
+      "loss": 0.2276,
+      "step": 9537
+    },
+    {
+      "epoch": 7.179525780955966,
+      "grad_norm": 0.19570425152778625,
+      "learning_rate": 5.2620861504483734e-06,
+      "loss": 0.2326,
+      "step": 9538
+    },
+    {
+      "epoch": 7.18027850959729,
+      "grad_norm": 0.171380415558815,
+      "learning_rate": 5.260955002124369e-06,
+      "loss": 0.103,
+      "step": 9539
+    },
+    {
+      "epoch": 7.1810312382386146,
+      "grad_norm": 0.1864515244960785,
+      "learning_rate": 5.2598262858288924e-06,
+      "loss": 0.2047,
+      "step": 9540
+    },
+    {
+      "epoch": 7.18178396687994,
+      "grad_norm": 0.18331468105316162,
+      "learning_rate": 5.258700001685374e-06,
+      "loss": 0.1297,
+      "step": 9541
+    },
+    {
+      "epoch": 7.182536695521264,
+      "grad_norm": 0.15780189633369446,
+      "learning_rate": 5.257576149816986e-06,
+      "loss": 0.1877,
+      "step": 9542
+    },
+    {
+      "epoch": 7.18328942416259,
+      "grad_norm": 0.1663518100976944,
+      "learning_rate": 5.2564547303466265e-06,
+      "loss": 0.1282,
+      "step": 9543
+    },
+    {
+      "epoch": 7.184042152803914,
+      "grad_norm": 0.1572708785533905,
+      "learning_rate": 5.255335743396937e-06,
+      "loss": 0.1272,
+      "step": 9544
+    },
+    {
+      "epoch": 7.184794881445239,
+      "grad_norm": 0.153344988822937,
+      "learning_rate": 5.254219189090284e-06,
+      "loss": 0.1335,
+      "step": 9545
+    },
+    {
+      "epoch": 7.185547610086564,
+      "grad_norm": 0.16096769273281097,
+      "learning_rate": 5.253105067548775e-06,
+      "loss": 0.1634,
+      "step": 9546
+    },
+    {
+      "epoch": 7.1863003387278885,
+      "grad_norm": 0.17247819900512695,
+      "learning_rate": 5.2519933788942484e-06,
+      "loss": 0.2207,
+      "step": 9547
+    },
+    {
+      "epoch": 7.187053067369213,
+      "grad_norm": 0.1846633404493332,
+      "learning_rate": 5.250884123248277e-06,
+      "loss": 0.214,
+      "step": 9548
+    },
+    {
+      "epoch": 7.187805796010538,
+      "grad_norm": 0.16404688358306885,
+      "learning_rate": 5.249777300732168e-06,
+      "loss": 0.0962,
+      "step": 9549
+    },
+    {
+      "epoch": 7.188558524651863,
+      "grad_norm": 0.19543053209781647,
+      "learning_rate": 5.248672911466959e-06,
+      "loss": 0.1249,
+      "step": 9550
+    },
+    {
+      "epoch": 7.189311253293188,
+      "grad_norm": 0.20585796236991882,
+      "learning_rate": 5.2475709555734304e-06,
+      "loss": 0.134,
+      "step": 9551
+    },
+    {
+      "epoch": 7.190063981934513,
+      "grad_norm": 0.15885896980762482,
+      "learning_rate": 5.246471433172079e-06,
+      "loss": 0.1959,
+      "step": 9552
+    },
+    {
+      "epoch": 7.190816710575837,
+      "grad_norm": 0.17267978191375732,
+      "learning_rate": 5.245374344383159e-06,
+      "loss": 0.1755,
+      "step": 9553
+    },
+    {
+      "epoch": 7.1915694392171625,
+      "grad_norm": 0.16274522244930267,
+      "learning_rate": 5.244279689326639e-06,
+      "loss": 0.1323,
+      "step": 9554
+    },
+    {
+      "epoch": 7.192322167858487,
+      "grad_norm": 0.17078734934329987,
+      "learning_rate": 5.243187468122232e-06,
+      "loss": 0.1516,
+      "step": 9555
+    },
+    {
+      "epoch": 7.1930748964998115,
+      "grad_norm": 0.15448325872421265,
+      "learning_rate": 5.24209768088938e-06,
+      "loss": 0.0872,
+      "step": 9556
+    },
+    {
+      "epoch": 7.193827625141137,
+      "grad_norm": 0.1732197403907776,
+      "learning_rate": 5.241010327747263e-06,
+      "loss": 0.1527,
+      "step": 9557
+    },
+    {
+      "epoch": 7.194580353782461,
+      "grad_norm": 0.15944743156433105,
+      "learning_rate": 5.239925408814788e-06,
+      "loss": 0.1384,
+      "step": 9558
+    },
+    {
+      "epoch": 7.195333082423787,
+      "grad_norm": 0.15052148699760437,
+      "learning_rate": 5.238842924210607e-06,
+      "loss": 0.1475,
+      "step": 9559
+    },
+    {
+      "epoch": 7.196085811065111,
+      "grad_norm": 0.17674288153648376,
+      "learning_rate": 5.23776287405309e-06,
+      "loss": 0.2072,
+      "step": 9560
+    },
+    {
+      "epoch": 7.196838539706436,
+      "grad_norm": 0.17039060592651367,
+      "learning_rate": 5.236685258460358e-06,
+      "loss": 0.15,
+      "step": 9561
+    },
+    {
+      "epoch": 7.197591268347761,
+      "grad_norm": 0.14818966388702393,
+      "learning_rate": 5.2356100775502524e-06,
+      "loss": 0.1025,
+      "step": 9562
+    },
+    {
+      "epoch": 7.1983439969890854,
+      "grad_norm": 0.17303569614887238,
+      "learning_rate": 5.234537331440355e-06,
+      "loss": 0.1737,
+      "step": 9563
+    },
+    {
+      "epoch": 7.19909672563041,
+      "grad_norm": 0.1901473104953766,
+      "learning_rate": 5.2334670202479775e-06,
+      "loss": 0.2226,
+      "step": 9564
+    },
+    {
+      "epoch": 7.199849454271735,
+      "grad_norm": 0.18445239961147308,
+      "learning_rate": 5.232399144090171e-06,
+      "loss": 0.198,
+      "step": 9565
+    },
+    {
+      "epoch": 7.20060218291306,
+      "grad_norm": 0.16957108676433563,
+      "learning_rate": 5.231333703083716e-06,
+      "loss": 0.1428,
+      "step": 9566
+    },
+    {
+      "epoch": 7.201354911554384,
+      "grad_norm": 0.17735648155212402,
+      "learning_rate": 5.230270697345126e-06,
+      "loss": 0.1534,
+      "step": 9567
+    },
+    {
+      "epoch": 7.20210764019571,
+      "grad_norm": 0.15678903460502625,
+      "learning_rate": 5.2292101269906494e-06,
+      "loss": 0.0823,
+      "step": 9568
+    },
+    {
+      "epoch": 7.202860368837034,
+      "grad_norm": 0.16675294935703278,
+      "learning_rate": 5.2281519921362695e-06,
+      "loss": 0.1193,
+      "step": 9569
+    },
+    {
+      "epoch": 7.203613097478359,
+      "grad_norm": 0.17713117599487305,
+      "learning_rate": 5.227096292897707e-06,
+      "loss": 0.1584,
+      "step": 9570
+    },
+    {
+      "epoch": 7.204365826119684,
+      "grad_norm": 0.18780572712421417,
+      "learning_rate": 5.226043029390403e-06,
+      "loss": 0.2221,
+      "step": 9571
+    },
+    {
+      "epoch": 7.205118554761008,
+      "grad_norm": 0.18530713021755219,
+      "learning_rate": 5.2249922017295435e-06,
+      "loss": 0.1558,
+      "step": 9572
+    },
+    {
+      "epoch": 7.205871283402334,
+      "grad_norm": 0.16881896555423737,
+      "learning_rate": 5.22394381003005e-06,
+      "loss": 0.1414,
+      "step": 9573
+    },
+    {
+      "epoch": 7.206624012043658,
+      "grad_norm": 0.17679844796657562,
+      "learning_rate": 5.2228978544065695e-06,
+      "loss": 0.1883,
+      "step": 9574
+    },
+    {
+      "epoch": 7.207376740684983,
+      "grad_norm": 0.1645129770040512,
+      "learning_rate": 5.221854334973485e-06,
+      "loss": 0.2787,
+      "step": 9575
+    },
+    {
+      "epoch": 7.208129469326308,
+      "grad_norm": 0.15561729669570923,
+      "learning_rate": 5.2208132518449195e-06,
+      "loss": 0.0632,
+      "step": 9576
+    },
+    {
+      "epoch": 7.2088821979676325,
+      "grad_norm": 0.16195526719093323,
+      "learning_rate": 5.219774605134716e-06,
+      "loss": 0.17,
+      "step": 9577
+    },
+    {
+      "epoch": 7.209634926608958,
+      "grad_norm": 0.196675106883049,
+      "learning_rate": 5.218738394956466e-06,
+      "loss": 0.1593,
+      "step": 9578
+    },
+    {
+      "epoch": 7.210387655250282,
+      "grad_norm": 0.19529181718826294,
+      "learning_rate": 5.217704621423485e-06,
+      "loss": 0.1079,
+      "step": 9579
+    },
+    {
+      "epoch": 7.211140383891607,
+      "grad_norm": 0.17100384831428528,
+      "learning_rate": 5.216673284648828e-06,
+      "loss": 0.1182,
+      "step": 9580
+    },
+    {
+      "epoch": 7.211893112532932,
+      "grad_norm": 0.1780667006969452,
+      "learning_rate": 5.2156443847452735e-06,
+      "loss": 0.1577,
+      "step": 9581
+    },
+    {
+      "epoch": 7.212645841174257,
+      "grad_norm": 0.16650354862213135,
+      "learning_rate": 5.214617921825349e-06,
+      "loss": 0.137,
+      "step": 9582
+    },
+    {
+      "epoch": 7.213398569815581,
+      "grad_norm": 0.16963180899620056,
+      "learning_rate": 5.2135938960013e-06,
+      "loss": 0.1109,
+      "step": 9583
+    },
+    {
+      "epoch": 7.2141512984569065,
+      "grad_norm": 0.1947159469127655,
+      "learning_rate": 5.212572307385118e-06,
+      "loss": 0.1876,
+      "step": 9584
+    },
+    {
+      "epoch": 7.214904027098231,
+      "grad_norm": 0.16912515461444855,
+      "learning_rate": 5.2115531560885184e-06,
+      "loss": 0.1289,
+      "step": 9585
+    },
+    {
+      "epoch": 7.2156567557395555,
+      "grad_norm": 0.1664021909236908,
+      "learning_rate": 5.210536442222957e-06,
+      "loss": 0.151,
+      "step": 9586
+    },
+    {
+      "epoch": 7.216409484380881,
+      "grad_norm": 0.15198037028312683,
+      "learning_rate": 5.209522165899617e-06,
+      "loss": 0.115,
+      "step": 9587
+    },
+    {
+      "epoch": 7.217162213022205,
+      "grad_norm": 0.15778395533561707,
+      "learning_rate": 5.2085103272294215e-06,
+      "loss": 0.1571,
+      "step": 9588
+    },
+    {
+      "epoch": 7.217914941663531,
+      "grad_norm": 0.17603380978107452,
+      "learning_rate": 5.207500926323019e-06,
+      "loss": 0.2397,
+      "step": 9589
+    },
+    {
+      "epoch": 7.218667670304855,
+      "grad_norm": 0.18663322925567627,
+      "learning_rate": 5.206493963290802e-06,
+      "loss": 0.2552,
+      "step": 9590
+    },
+    {
+      "epoch": 7.21942039894618,
+      "grad_norm": 0.15369410812854767,
+      "learning_rate": 5.205489438242884e-06,
+      "loss": 0.1135,
+      "step": 9591
+    },
+    {
+      "epoch": 7.220173127587505,
+      "grad_norm": 0.174699068069458,
+      "learning_rate": 5.204487351289125e-06,
+      "loss": 0.2267,
+      "step": 9592
+    },
+    {
+      "epoch": 7.220925856228829,
+      "grad_norm": 0.16740386188030243,
+      "learning_rate": 5.2034877025391076e-06,
+      "loss": 0.2235,
+      "step": 9593
+    },
+    {
+      "epoch": 7.221678584870154,
+      "grad_norm": 0.16752369701862335,
+      "learning_rate": 5.202490492102152e-06,
+      "loss": 0.1247,
+      "step": 9594
+    },
+    {
+      "epoch": 7.222431313511479,
+      "grad_norm": 0.15724553167819977,
+      "learning_rate": 5.201495720087313e-06,
+      "loss": 0.1835,
+      "step": 9595
+    },
+    {
+      "epoch": 7.223184042152804,
+      "grad_norm": 0.16751445829868317,
+      "learning_rate": 5.200503386603375e-06,
+      "loss": 0.1576,
+      "step": 9596
+    },
+    {
+      "epoch": 7.223936770794129,
+      "grad_norm": 0.17532697319984436,
+      "learning_rate": 5.199513491758861e-06,
+      "loss": 0.2149,
+      "step": 9597
+    },
+    {
+      "epoch": 7.224689499435454,
+      "grad_norm": 0.17880091071128845,
+      "learning_rate": 5.198526035662023e-06,
+      "loss": 0.2471,
+      "step": 9598
+    },
+    {
+      "epoch": 7.225442228076778,
+      "grad_norm": 0.18646639585494995,
+      "learning_rate": 5.19754101842085e-06,
+      "loss": 0.1804,
+      "step": 9599
+    },
+    {
+      "epoch": 7.226194956718103,
+      "grad_norm": 0.1815239042043686,
+      "learning_rate": 5.196558440143059e-06,
+      "loss": 0.1295,
+      "step": 9600
+    },
+    {
+      "epoch": 7.226194956718103,
+      "eval_loss": 0.2012045681476593,
+      "eval_runtime": 452.7985,
+      "eval_samples_per_second": 21.261,
+      "eval_steps_per_second": 0.665,
+      "step": 9600
+    },
+    {
+      "epoch": 7.226947685359428,
+      "grad_norm": 0.18046434223651886,
+      "learning_rate": 5.195578300936105e-06,
+      "loss": 0.2613,
+      "step": 9601
+    },
+    {
+      "epoch": 7.227700414000752,
+      "grad_norm": 0.17731885612010956,
+      "learning_rate": 5.1946006009071704e-06,
+      "loss": 0.152,
+      "step": 9602
+    },
+    {
+      "epoch": 7.228453142642078,
+      "grad_norm": 0.17450009286403656,
+      "learning_rate": 5.19362534016318e-06,
+      "loss": 0.1993,
+      "step": 9603
+    },
+    {
+      "epoch": 7.229205871283402,
+      "grad_norm": 0.16135737299919128,
+      "learning_rate": 5.192652518810786e-06,
+      "loss": 0.1487,
+      "step": 9604
+    },
+    {
+      "epoch": 7.229958599924728,
+      "grad_norm": 0.17209556698799133,
+      "learning_rate": 5.191682136956376e-06,
+      "loss": 0.1767,
+      "step": 9605
+    },
+    {
+      "epoch": 7.230711328566052,
+      "grad_norm": 0.16481085121631622,
+      "learning_rate": 5.1907141947060635e-06,
+      "loss": 0.147,
+      "step": 9606
+    },
+    {
+      "epoch": 7.2314640572073765,
+      "grad_norm": 0.16940534114837646,
+      "learning_rate": 5.189748692165709e-06,
+      "loss": 0.1638,
+      "step": 9607
+    },
+    {
+      "epoch": 7.232216785848702,
+      "grad_norm": 0.17044676840305328,
+      "learning_rate": 5.1887856294408904e-06,
+      "loss": 0.1423,
+      "step": 9608
+    },
+    {
+      "epoch": 7.232969514490026,
+      "grad_norm": 0.17587774991989136,
+      "learning_rate": 5.187825006636933e-06,
+      "loss": 0.1303,
+      "step": 9609
+    },
+    {
+      "epoch": 7.233722243131351,
+      "grad_norm": 0.1805824190378189,
+      "learning_rate": 5.186866823858886e-06,
+      "loss": 0.1531,
+      "step": 9610
+    },
+    {
+      "epoch": 7.234474971772676,
+      "grad_norm": 0.1601659059524536,
+      "learning_rate": 5.185911081211538e-06,
+      "loss": 0.2044,
+      "step": 9611
+    },
+    {
+      "epoch": 7.235227700414001,
+      "grad_norm": 0.16830968856811523,
+      "learning_rate": 5.184957778799403e-06,
+      "loss": 0.1627,
+      "step": 9612
+    },
+    {
+      "epoch": 7.235980429055326,
+      "grad_norm": 0.1574537754058838,
+      "learning_rate": 5.184006916726739e-06,
+      "loss": 0.1983,
+      "step": 9613
+    },
+    {
+      "epoch": 7.2367331576966505,
+      "grad_norm": 0.1680692732334137,
+      "learning_rate": 5.1830584950975234e-06,
+      "loss": 0.1573,
+      "step": 9614
+    },
+    {
+      "epoch": 7.237485886337975,
+      "grad_norm": 0.16044075787067413,
+      "learning_rate": 5.182112514015478e-06,
+      "loss": 0.2191,
+      "step": 9615
+    },
+    {
+      "epoch": 7.2382386149793,
+      "grad_norm": 0.17655861377716064,
+      "learning_rate": 5.181168973584054e-06,
+      "loss": 0.1364,
+      "step": 9616
+    },
+    {
+      "epoch": 7.238991343620625,
+      "grad_norm": 0.17649157345294952,
+      "learning_rate": 5.180227873906438e-06,
+      "loss": 0.1849,
+      "step": 9617
+    },
+    {
+      "epoch": 7.239744072261949,
+      "grad_norm": 0.20300404727458954,
+      "learning_rate": 5.179289215085539e-06,
+      "loss": 0.1654,
+      "step": 9618
+    },
+    {
+      "epoch": 7.240496800903275,
+      "grad_norm": 0.18087223172187805,
+      "learning_rate": 5.178352997224016e-06,
+      "loss": 0.14,
+      "step": 9619
+    },
+    {
+      "epoch": 7.241249529544599,
+      "grad_norm": 0.18133968114852905,
+      "learning_rate": 5.177419220424251e-06,
+      "loss": 0.1299,
+      "step": 9620
+    },
+    {
+      "epoch": 7.242002258185924,
+      "grad_norm": 0.16634416580200195,
+      "learning_rate": 5.1764878847883564e-06,
+      "loss": 0.1996,
+      "step": 9621
+    },
+    {
+      "epoch": 7.242754986827249,
+      "grad_norm": 0.16974878311157227,
+      "learning_rate": 5.175558990418182e-06,
+      "loss": 0.2194,
+      "step": 9622
+    },
+    {
+      "epoch": 7.243507715468573,
+      "grad_norm": 0.15365953743457794,
+      "learning_rate": 5.174632537415315e-06,
+      "loss": 0.0644,
+      "step": 9623
+    },
+    {
+      "epoch": 7.244260444109899,
+      "grad_norm": 0.15909388661384583,
+      "learning_rate": 5.1737085258810655e-06,
+      "loss": 0.1959,
+      "step": 9624
+    },
+    {
+      "epoch": 7.245013172751223,
+      "grad_norm": 0.1891498863697052,
+      "learning_rate": 5.172786955916488e-06,
+      "loss": 0.1495,
+      "step": 9625
+    },
+    {
+      "epoch": 7.245765901392548,
+      "grad_norm": 0.17020896077156067,
+      "learning_rate": 5.171867827622358e-06,
+      "loss": 0.1304,
+      "step": 9626
+    },
+    {
+      "epoch": 7.246518630033873,
+      "grad_norm": 0.17524418234825134,
+      "learning_rate": 5.170951141099192e-06,
+      "loss": 0.1898,
+      "step": 9627
+    },
+    {
+      "epoch": 7.247271358675198,
+      "grad_norm": 0.1652647852897644,
+      "learning_rate": 5.170036896447241e-06,
+      "loss": 0.2064,
+      "step": 9628
+    },
+    {
+      "epoch": 7.248024087316522,
+      "grad_norm": 0.18469250202178955,
+      "learning_rate": 5.169125093766479e-06,
+      "loss": 0.2093,
+      "step": 9629
+    },
+    {
+      "epoch": 7.248776815957847,
+      "grad_norm": 0.1715327501296997,
+      "learning_rate": 5.168215733156623e-06,
+      "loss": 0.1161,
+      "step": 9630
+    },
+    {
+      "epoch": 7.249529544599172,
+      "grad_norm": 0.1738138645887375,
+      "learning_rate": 5.1673088147171214e-06,
+      "loss": 0.169,
+      "step": 9631
+    },
+    {
+      "epoch": 7.250282273240497,
+      "grad_norm": 0.17101417481899261,
+      "learning_rate": 5.166404338547151e-06,
+      "loss": 0.1357,
+      "step": 9632
+    },
+    {
+      "epoch": 7.251035001881822,
+      "grad_norm": 0.18316714465618134,
+      "learning_rate": 5.165502304745623e-06,
+      "loss": 0.2209,
+      "step": 9633
+    },
+    {
+      "epoch": 7.251787730523146,
+      "grad_norm": 0.19148802757263184,
+      "learning_rate": 5.164602713411183e-06,
+      "loss": 0.2166,
+      "step": 9634
+    },
+    {
+      "epoch": 7.2525404591644715,
+      "grad_norm": 0.17324025928974152,
+      "learning_rate": 5.163705564642209e-06,
+      "loss": 0.1183,
+      "step": 9635
+    },
+    {
+      "epoch": 7.253293187805796,
+      "grad_norm": 0.19636303186416626,
+      "learning_rate": 5.162810858536815e-06,
+      "loss": 0.218,
+      "step": 9636
+    },
+    {
+      "epoch": 7.2540459164471205,
+      "grad_norm": 0.1749131977558136,
+      "learning_rate": 5.16191859519284e-06,
+      "loss": 0.1783,
+      "step": 9637
+    },
+    {
+      "epoch": 7.254798645088446,
+      "grad_norm": 0.15334342420101166,
+      "learning_rate": 5.161028774707865e-06,
+      "loss": 0.0957,
+      "step": 9638
+    },
+    {
+      "epoch": 7.25555137372977,
+      "grad_norm": 0.19247205555438995,
+      "learning_rate": 5.160141397179195e-06,
+      "loss": 0.1306,
+      "step": 9639
+    },
+    {
+      "epoch": 7.256304102371095,
+      "grad_norm": 0.19373901188373566,
+      "learning_rate": 5.159256462703871e-06,
+      "loss": 0.1928,
+      "step": 9640
+    },
+    {
+      "epoch": 7.25705683101242,
+      "grad_norm": 0.1836824268102646,
+      "learning_rate": 5.1583739713786756e-06,
+      "loss": 0.2294,
+      "step": 9641
+    },
+    {
+      "epoch": 7.257809559653745,
+      "grad_norm": 0.20581217110157013,
+      "learning_rate": 5.1574939233001125e-06,
+      "loss": 0.2496,
+      "step": 9642
+    },
+    {
+      "epoch": 7.25856228829507,
+      "grad_norm": 0.18953441083431244,
+      "learning_rate": 5.156616318564422e-06,
+      "loss": 0.1181,
+      "step": 9643
+    },
+    {
+      "epoch": 7.2593150169363945,
+      "grad_norm": 0.2143981158733368,
+      "learning_rate": 5.155741157267578e-06,
+      "loss": 0.2459,
+      "step": 9644
+    },
+    {
+      "epoch": 7.260067745577719,
+      "grad_norm": 0.1819150596857071,
+      "learning_rate": 5.154868439505289e-06,
+      "loss": 0.1364,
+      "step": 9645
+    },
+    {
+      "epoch": 7.260820474219044,
+      "grad_norm": 0.1669853776693344,
+      "learning_rate": 5.153998165372988e-06,
+      "loss": 0.13,
+      "step": 9646
+    },
+    {
+      "epoch": 7.261573202860369,
+      "grad_norm": 0.174150750041008,
+      "learning_rate": 5.153130334965853e-06,
+      "loss": 0.2034,
+      "step": 9647
+    },
+    {
+      "epoch": 7.262325931501693,
+      "grad_norm": 0.17472559213638306,
+      "learning_rate": 5.1522649483787865e-06,
+      "loss": 0.1486,
+      "step": 9648
+    },
+    {
+      "epoch": 7.263078660143019,
+      "grad_norm": 0.17480334639549255,
+      "learning_rate": 5.151402005706426e-06,
+      "loss": 0.1768,
+      "step": 9649
+    },
+    {
+      "epoch": 7.263831388784343,
+      "grad_norm": 0.15730643272399902,
+      "learning_rate": 5.150541507043143e-06,
+      "loss": 0.1151,
+      "step": 9650
+    },
+    {
+      "epoch": 7.2645841174256685,
+      "grad_norm": 0.1577325463294983,
+      "learning_rate": 5.149683452483038e-06,
+      "loss": 0.0901,
+      "step": 9651
+    },
+    {
+      "epoch": 7.265336846066993,
+      "grad_norm": 0.16705699265003204,
+      "learning_rate": 5.1488278421199455e-06,
+      "loss": 0.1571,
+      "step": 9652
+    },
+    {
+      "epoch": 7.266089574708317,
+      "grad_norm": 0.14310643076896667,
+      "learning_rate": 5.147974676047436e-06,
+      "loss": 0.1055,
+      "step": 9653
+    },
+    {
+      "epoch": 7.266842303349643,
+      "grad_norm": 0.1696098893880844,
+      "learning_rate": 5.147123954358811e-06,
+      "loss": 0.1586,
+      "step": 9654
+    },
+    {
+      "epoch": 7.267595031990967,
+      "grad_norm": 0.17257997393608093,
+      "learning_rate": 5.1462756771471024e-06,
+      "loss": 0.2098,
+      "step": 9655
+    },
+    {
+      "epoch": 7.268347760632292,
+      "grad_norm": 0.18418802320957184,
+      "learning_rate": 5.145429844505078e-06,
+      "loss": 0.1538,
+      "step": 9656
+    },
+    {
+      "epoch": 7.269100489273617,
+      "grad_norm": 0.16968032717704773,
+      "learning_rate": 5.14458645652524e-06,
+      "loss": 0.133,
+      "step": 9657
+    },
+    {
+      "epoch": 7.2698532179149415,
+      "grad_norm": 0.18164776265621185,
+      "learning_rate": 5.143745513299811e-06,
+      "loss": 0.138,
+      "step": 9658
+    },
+    {
+      "epoch": 7.270605946556266,
+      "grad_norm": 0.16523751616477966,
+      "learning_rate": 5.142907014920764e-06,
+      "loss": 0.1409,
+      "step": 9659
+    },
+    {
+      "epoch": 7.271358675197591,
+      "grad_norm": 0.16007889807224274,
+      "learning_rate": 5.142070961479789e-06,
+      "loss": 0.1318,
+      "step": 9660
+    },
+    {
+      "epoch": 7.272111403838916,
+      "grad_norm": 0.17440903186798096,
+      "learning_rate": 5.141237353068322e-06,
+      "loss": 0.1454,
+      "step": 9661
+    },
+    {
+      "epoch": 7.272864132480241,
+      "grad_norm": 0.16515131294727325,
+      "learning_rate": 5.1404061897775215e-06,
+      "loss": 0.0817,
+      "step": 9662
+    },
+    {
+      "epoch": 7.273616861121566,
+      "grad_norm": 0.17611882090568542,
+      "learning_rate": 5.139577471698281e-06,
+      "loss": 0.1654,
+      "step": 9663
+    },
+    {
+      "epoch": 7.27436958976289,
+      "grad_norm": 0.2097768485546112,
+      "learning_rate": 5.138751198921232e-06,
+      "loss": 0.1591,
+      "step": 9664
+    },
+    {
+      "epoch": 7.2751223184042155,
+      "grad_norm": 0.17666275799274445,
+      "learning_rate": 5.137927371536732e-06,
+      "loss": 0.2409,
+      "step": 9665
+    },
+    {
+      "epoch": 7.27587504704554,
+      "grad_norm": 0.14883045852184296,
+      "learning_rate": 5.137105989634872e-06,
+      "loss": 0.1397,
+      "step": 9666
+    },
+    {
+      "epoch": 7.2766277756868645,
+      "grad_norm": 0.173545241355896,
+      "learning_rate": 5.136287053305483e-06,
+      "loss": 0.1451,
+      "step": 9667
+    },
+    {
+      "epoch": 7.27738050432819,
+      "grad_norm": 0.19130313396453857,
+      "learning_rate": 5.135470562638113e-06,
+      "loss": 0.1729,
+      "step": 9668
+    },
+    {
+      "epoch": 7.278133232969514,
+      "grad_norm": 0.19222310185432434,
+      "learning_rate": 5.134656517722061e-06,
+      "loss": 0.1975,
+      "step": 9669
+    },
+    {
+      "epoch": 7.27888596161084,
+      "grad_norm": 0.153669536113739,
+      "learning_rate": 5.133844918646348e-06,
+      "loss": 0.1009,
+      "step": 9670
+    },
+    {
+      "epoch": 7.279638690252164,
+      "grad_norm": 0.17142049968242645,
+      "learning_rate": 5.133035765499725e-06,
+      "loss": 0.1449,
+      "step": 9671
+    },
+    {
+      "epoch": 7.280391418893489,
+      "grad_norm": 0.17374397814273834,
+      "learning_rate": 5.132229058370682e-06,
+      "loss": 0.187,
+      "step": 9672
+    },
+    {
+      "epoch": 7.281144147534814,
+      "grad_norm": 0.1720137745141983,
+      "learning_rate": 5.13142479734744e-06,
+      "loss": 0.2039,
+      "step": 9673
+    },
+    {
+      "epoch": 7.2818968761761385,
+      "grad_norm": 0.1551455855369568,
+      "learning_rate": 5.130622982517953e-06,
+      "loss": 0.1316,
+      "step": 9674
+    },
+    {
+      "epoch": 7.282649604817463,
+      "grad_norm": 0.19383400678634644,
+      "learning_rate": 5.129823613969904e-06,
+      "loss": 0.1624,
+      "step": 9675
+    },
+    {
+      "epoch": 7.283402333458788,
+      "grad_norm": 0.17984820902347565,
+      "learning_rate": 5.129026691790712e-06,
+      "loss": 0.1835,
+      "step": 9676
+    },
+    {
+      "epoch": 7.284155062100113,
+      "grad_norm": 0.176526740193367,
+      "learning_rate": 5.128232216067524e-06,
+      "loss": 0.1009,
+      "step": 9677
+    },
+    {
+      "epoch": 7.284907790741438,
+      "grad_norm": 0.15534695982933044,
+      "learning_rate": 5.1274401868872315e-06,
+      "loss": 0.1405,
+      "step": 9678
+    },
+    {
+      "epoch": 7.285660519382763,
+      "grad_norm": 0.1539594829082489,
+      "learning_rate": 5.126650604336437e-06,
+      "loss": 0.1848,
+      "step": 9679
+    },
+    {
+      "epoch": 7.286413248024087,
+      "grad_norm": 0.17623071372509003,
+      "learning_rate": 5.125863468501497e-06,
+      "loss": 0.1162,
+      "step": 9680
+    },
+    {
+      "epoch": 7.287165976665412,
+      "grad_norm": 0.17455565929412842,
+      "learning_rate": 5.1250787794684885e-06,
+      "loss": 0.1892,
+      "step": 9681
+    },
+    {
+      "epoch": 7.287918705306737,
+      "grad_norm": 0.18523722887039185,
+      "learning_rate": 5.124296537323228e-06,
+      "loss": 0.1975,
+      "step": 9682
+    },
+    {
+      "epoch": 7.288671433948061,
+      "grad_norm": 0.16469398140907288,
+      "learning_rate": 5.123516742151253e-06,
+      "loss": 0.1802,
+      "step": 9683
+    },
+    {
+      "epoch": 7.289424162589387,
+      "grad_norm": 0.1606116145849228,
+      "learning_rate": 5.122739394037847e-06,
+      "loss": 0.1386,
+      "step": 9684
+    },
+    {
+      "epoch": 7.290176891230711,
+      "grad_norm": 0.1558232605457306,
+      "learning_rate": 5.1219644930680135e-06,
+      "loss": 0.1022,
+      "step": 9685
+    },
+    {
+      "epoch": 7.290929619872037,
+      "grad_norm": 0.19172774255275726,
+      "learning_rate": 5.121192039326502e-06,
+      "loss": 0.1722,
+      "step": 9686
+    },
+    {
+      "epoch": 7.291682348513361,
+      "grad_norm": 0.17343848943710327,
+      "learning_rate": 5.120422032897782e-06,
+      "loss": 0.1484,
+      "step": 9687
+    },
+    {
+      "epoch": 7.2924350771546855,
+      "grad_norm": 0.19179384410381317,
+      "learning_rate": 5.1196544738660605e-06,
+      "loss": 0.127,
+      "step": 9688
+    },
+    {
+      "epoch": 7.293187805796011,
+      "grad_norm": 0.16638930141925812,
+      "learning_rate": 5.118889362315277e-06,
+      "loss": 0.1091,
+      "step": 9689
+    },
+    {
+      "epoch": 7.293940534437335,
+      "grad_norm": 0.1666993945837021,
+      "learning_rate": 5.118126698329105e-06,
+      "loss": 0.1626,
+      "step": 9690
+    },
+    {
+      "epoch": 7.29469326307866,
+      "grad_norm": 0.20673085749149323,
+      "learning_rate": 5.117366481990944e-06,
+      "loss": 0.2354,
+      "step": 9691
+    },
+    {
+      "epoch": 7.295445991719985,
+      "grad_norm": 0.1743408590555191,
+      "learning_rate": 5.116608713383935e-06,
+      "loss": 0.1673,
+      "step": 9692
+    },
+    {
+      "epoch": 7.29619872036131,
+      "grad_norm": 0.18816038966178894,
+      "learning_rate": 5.1158533925909435e-06,
+      "loss": 0.1166,
+      "step": 9693
+    },
+    {
+      "epoch": 7.296951449002634,
+      "grad_norm": 0.165283665060997,
+      "learning_rate": 5.11510051969457e-06,
+      "loss": 0.203,
+      "step": 9694
+    },
+    {
+      "epoch": 7.2977041776439595,
+      "grad_norm": 0.15886792540550232,
+      "learning_rate": 5.114350094777151e-06,
+      "loss": 0.2002,
+      "step": 9695
+    },
+    {
+      "epoch": 7.298456906285284,
+      "grad_norm": 0.15822751820087433,
+      "learning_rate": 5.113602117920746e-06,
+      "loss": 0.1218,
+      "step": 9696
+    },
+    {
+      "epoch": 7.299209634926609,
+      "grad_norm": 0.1559489220380783,
+      "learning_rate": 5.11285658920716e-06,
+      "loss": 0.1267,
+      "step": 9697
+    },
+    {
+      "epoch": 7.299962363567934,
+      "grad_norm": 0.1642986685037613,
+      "learning_rate": 5.112113508717915e-06,
+      "loss": 0.1603,
+      "step": 9698
+    },
+    {
+      "epoch": 7.300715092209258,
+      "grad_norm": 0.20282308757305145,
+      "learning_rate": 5.111372876534281e-06,
+      "loss": 0.1972,
+      "step": 9699
+    },
+    {
+      "epoch": 7.301467820850584,
+      "grad_norm": 0.18710726499557495,
+      "learning_rate": 5.110634692737244e-06,
+      "loss": 0.1303,
+      "step": 9700
+    },
+    {
+      "epoch": 7.302220549491908,
+      "grad_norm": 0.17408272624015808,
+      "learning_rate": 5.109898957407537e-06,
+      "loss": 0.1858,
+      "step": 9701
+    },
+    {
+      "epoch": 7.302973278133233,
+      "grad_norm": 0.17210040986537933,
+      "learning_rate": 5.109165670625618e-06,
+      "loss": 0.1631,
+      "step": 9702
+    },
+    {
+      "epoch": 7.303726006774558,
+      "grad_norm": 0.1673329621553421,
+      "learning_rate": 5.108434832471676e-06,
+      "loss": 0.1666,
+      "step": 9703
+    },
+    {
+      "epoch": 7.304478735415882,
+      "grad_norm": 0.16123028099536896,
+      "learning_rate": 5.107706443025635e-06,
+      "loss": 0.1689,
+      "step": 9704
+    },
+    {
+      "epoch": 7.305231464057208,
+      "grad_norm": 0.1881340891122818,
+      "learning_rate": 5.106980502367152e-06,
+      "loss": 0.1863,
+      "step": 9705
+    },
+    {
+      "epoch": 7.305984192698532,
+      "grad_norm": 0.16289329528808594,
+      "learning_rate": 5.1062570105756125e-06,
+      "loss": 0.1654,
+      "step": 9706
+    },
+    {
+      "epoch": 7.306736921339857,
+      "grad_norm": 0.18195848166942596,
+      "learning_rate": 5.105535967730142e-06,
+      "loss": 0.2046,
+      "step": 9707
+    },
+    {
+      "epoch": 7.307489649981182,
+      "grad_norm": 0.16319413483142853,
+      "learning_rate": 5.104817373909584e-06,
+      "loss": 0.1611,
+      "step": 9708
+    },
+    {
+      "epoch": 7.308242378622507,
+      "grad_norm": 0.1900513470172882,
+      "learning_rate": 5.104101229192528e-06,
+      "loss": 0.229,
+      "step": 9709
+    },
+    {
+      "epoch": 7.308995107263831,
+      "grad_norm": 0.17418894171714783,
+      "learning_rate": 5.1033875336572905e-06,
+      "loss": 0.1882,
+      "step": 9710
+    },
+    {
+      "epoch": 7.309747835905156,
+      "grad_norm": 0.17933915555477142,
+      "learning_rate": 5.10267628738192e-06,
+      "loss": 0.1902,
+      "step": 9711
+    },
+    {
+      "epoch": 7.310500564546481,
+      "grad_norm": 0.18145540356636047,
+      "learning_rate": 5.101967490444197e-06,
+      "loss": 0.1543,
+      "step": 9712
+    },
+    {
+      "epoch": 7.311253293187805,
+      "grad_norm": 0.16726788878440857,
+      "learning_rate": 5.1012611429216325e-06,
+      "loss": 0.1717,
+      "step": 9713
+    },
+    {
+      "epoch": 7.312006021829131,
+      "grad_norm": 0.17700543999671936,
+      "learning_rate": 5.100557244891477e-06,
+      "loss": 0.1211,
+      "step": 9714
+    },
+    {
+      "epoch": 7.312758750470455,
+      "grad_norm": 0.171833336353302,
+      "learning_rate": 5.0998557964307e-06,
+      "loss": 0.1713,
+      "step": 9715
+    },
+    {
+      "epoch": 7.313511479111781,
+      "grad_norm": 0.1694435179233551,
+      "learning_rate": 5.099156797616018e-06,
+      "loss": 0.1387,
+      "step": 9716
+    },
+    {
+      "epoch": 7.314264207753105,
+      "grad_norm": 0.1775207221508026,
+      "learning_rate": 5.098460248523868e-06,
+      "loss": 0.2425,
+      "step": 9717
+    },
+    {
+      "epoch": 7.3150169363944295,
+      "grad_norm": 0.16320672631263733,
+      "learning_rate": 5.097766149230424e-06,
+      "loss": 0.1474,
+      "step": 9718
+    },
+    {
+      "epoch": 7.315769665035755,
+      "grad_norm": 0.18156519532203674,
+      "learning_rate": 5.097074499811594e-06,
+      "loss": 0.1584,
+      "step": 9719
+    },
+    {
+      "epoch": 7.316522393677079,
+      "grad_norm": 0.1632760614156723,
+      "learning_rate": 5.096385300343016e-06,
+      "loss": 0.1067,
+      "step": 9720
+    },
+    {
+      "epoch": 7.317275122318404,
+      "grad_norm": 0.176652729511261,
+      "learning_rate": 5.095698550900056e-06,
+      "loss": 0.1248,
+      "step": 9721
+    },
+    {
+      "epoch": 7.318027850959729,
+      "grad_norm": 0.17518632113933563,
+      "learning_rate": 5.09501425155782e-06,
+      "loss": 0.1559,
+      "step": 9722
+    },
+    {
+      "epoch": 7.318780579601054,
+      "grad_norm": 0.16845932602882385,
+      "learning_rate": 5.094332402391137e-06,
+      "loss": 0.1734,
+      "step": 9723
+    },
+    {
+      "epoch": 7.319533308242379,
+      "grad_norm": 0.17695966362953186,
+      "learning_rate": 5.093653003474579e-06,
+      "loss": 0.0917,
+      "step": 9724
+    },
+    {
+      "epoch": 7.3202860368837035,
+      "grad_norm": 0.17665793001651764,
+      "learning_rate": 5.092976054882439e-06,
+      "loss": 0.1818,
+      "step": 9725
+    },
+    {
+      "epoch": 7.321038765525028,
+      "grad_norm": 0.1742457151412964,
+      "learning_rate": 5.092301556688752e-06,
+      "loss": 0.1565,
+      "step": 9726
+    },
+    {
+      "epoch": 7.321791494166353,
+      "grad_norm": 0.1677614450454712,
+      "learning_rate": 5.091629508967273e-06,
+      "loss": 0.1974,
+      "step": 9727
+    },
+    {
+      "epoch": 7.322544222807678,
+      "grad_norm": 0.14712142944335938,
+      "learning_rate": 5.090959911791504e-06,
+      "loss": 0.1263,
+      "step": 9728
+    },
+    {
+      "epoch": 7.323296951449002,
+      "grad_norm": 0.16906876862049103,
+      "learning_rate": 5.0902927652346655e-06,
+      "loss": 0.2173,
+      "step": 9729
+    },
+    {
+      "epoch": 7.324049680090328,
+      "grad_norm": 0.17322172224521637,
+      "learning_rate": 5.0896280693697185e-06,
+      "loss": 0.2355,
+      "step": 9730
+    },
+    {
+      "epoch": 7.324802408731652,
+      "grad_norm": 0.15958693623542786,
+      "learning_rate": 5.088965824269351e-06,
+      "loss": 0.1159,
+      "step": 9731
+    },
+    {
+      "epoch": 7.325555137372977,
+      "grad_norm": 0.17918546497821808,
+      "learning_rate": 5.088306030005989e-06,
+      "loss": 0.1719,
+      "step": 9732
+    },
+    {
+      "epoch": 7.326307866014302,
+      "grad_norm": 0.18280020356178284,
+      "learning_rate": 5.0876486866517825e-06,
+      "loss": 0.2621,
+      "step": 9733
+    },
+    {
+      "epoch": 7.327060594655626,
+      "grad_norm": 0.15906819701194763,
+      "learning_rate": 5.086993794278619e-06,
+      "loss": 0.1079,
+      "step": 9734
+    },
+    {
+      "epoch": 7.327813323296952,
+      "grad_norm": 0.16760969161987305,
+      "learning_rate": 5.086341352958119e-06,
+      "loss": 0.1762,
+      "step": 9735
+    },
+    {
+      "epoch": 7.328566051938276,
+      "grad_norm": 0.16383200883865356,
+      "learning_rate": 5.0856913627616275e-06,
+      "loss": 0.1113,
+      "step": 9736
+    },
+    {
+      "epoch": 7.329318780579601,
+      "grad_norm": 0.20381316542625427,
+      "learning_rate": 5.085043823760231e-06,
+      "loss": 0.1825,
+      "step": 9737
+    },
+    {
+      "epoch": 7.330071509220926,
+      "grad_norm": 0.15747974812984467,
+      "learning_rate": 5.08439873602474e-06,
+      "loss": 0.1153,
+      "step": 9738
+    },
+    {
+      "epoch": 7.330824237862251,
+      "grad_norm": 0.2037978619337082,
+      "learning_rate": 5.0837560996257045e-06,
+      "loss": 0.1992,
+      "step": 9739
+    },
+    {
+      "epoch": 7.331576966503576,
+      "grad_norm": 0.19686850905418396,
+      "learning_rate": 5.083115914633396e-06,
+      "loss": 0.206,
+      "step": 9740
+    },
+    {
+      "epoch": 7.3323296951449,
+      "grad_norm": 0.16168154776096344,
+      "learning_rate": 5.08247818111783e-06,
+      "loss": 0.1804,
+      "step": 9741
+    },
+    {
+      "epoch": 7.333082423786225,
+      "grad_norm": 0.16215650737285614,
+      "learning_rate": 5.081842899148745e-06,
+      "loss": 0.1652,
+      "step": 9742
+    },
+    {
+      "epoch": 7.33383515242755,
+      "grad_norm": 0.16933904588222504,
+      "learning_rate": 5.081210068795615e-06,
+      "loss": 0.1732,
+      "step": 9743
+    },
+    {
+      "epoch": 7.334587881068875,
+      "grad_norm": 0.16179290413856506,
+      "learning_rate": 5.080579690127647e-06,
+      "loss": 0.1234,
+      "step": 9744
+    },
+    {
+      "epoch": 7.335340609710199,
+      "grad_norm": 0.16188035905361176,
+      "learning_rate": 5.079951763213777e-06,
+      "loss": 0.099,
+      "step": 9745
+    },
+    {
+      "epoch": 7.3360933383515246,
+      "grad_norm": 0.1701260358095169,
+      "learning_rate": 5.0793262881226735e-06,
+      "loss": 0.218,
+      "step": 9746
+    },
+    {
+      "epoch": 7.336846066992849,
+      "grad_norm": 0.1724553257226944,
+      "learning_rate": 5.078703264922738e-06,
+      "loss": 0.1877,
+      "step": 9747
+    },
+    {
+      "epoch": 7.3375987956341735,
+      "grad_norm": 0.1945074051618576,
+      "learning_rate": 5.078082693682104e-06,
+      "loss": 0.1064,
+      "step": 9748
+    },
+    {
+      "epoch": 7.338351524275499,
+      "grad_norm": 0.152532696723938,
+      "learning_rate": 5.077464574468635e-06,
+      "loss": 0.1112,
+      "step": 9749
+    },
+    {
+      "epoch": 7.339104252916823,
+      "grad_norm": 0.19756178557872772,
+      "learning_rate": 5.0768489073499295e-06,
+      "loss": 0.1755,
+      "step": 9750
+    },
+    {
+      "epoch": 7.339856981558149,
+      "grad_norm": 0.17954282462596893,
+      "learning_rate": 5.076235692393314e-06,
+      "loss": 0.1905,
+      "step": 9751
+    },
+    {
+      "epoch": 7.340609710199473,
+      "grad_norm": 0.16830159723758698,
+      "learning_rate": 5.07562492966585e-06,
+      "loss": 0.1072,
+      "step": 9752
+    },
+    {
+      "epoch": 7.341362438840798,
+      "grad_norm": 0.16587796807289124,
+      "learning_rate": 5.075016619234328e-06,
+      "loss": 0.1352,
+      "step": 9753
+    },
+    {
+      "epoch": 7.342115167482123,
+      "grad_norm": 0.17262671887874603,
+      "learning_rate": 5.074410761165275e-06,
+      "loss": 0.1617,
+      "step": 9754
+    },
+    {
+      "epoch": 7.3428678961234475,
+      "grad_norm": 0.19598762691020966,
+      "learning_rate": 5.073807355524944e-06,
+      "loss": 0.2035,
+      "step": 9755
+    },
+    {
+      "epoch": 7.343620624764772,
+      "grad_norm": 0.17963829636573792,
+      "learning_rate": 5.073206402379322e-06,
+      "loss": 0.1367,
+      "step": 9756
+    },
+    {
+      "epoch": 7.344373353406097,
+      "grad_norm": 0.15799349546432495,
+      "learning_rate": 5.072607901794131e-06,
+      "loss": 0.1636,
+      "step": 9757
+    },
+    {
+      "epoch": 7.345126082047422,
+      "grad_norm": 0.1709592342376709,
+      "learning_rate": 5.072011853834821e-06,
+      "loss": 0.0915,
+      "step": 9758
+    },
+    {
+      "epoch": 7.345878810688747,
+      "grad_norm": 0.18763123452663422,
+      "learning_rate": 5.071418258566572e-06,
+      "loss": 0.2115,
+      "step": 9759
+    },
+    {
+      "epoch": 7.346631539330072,
+      "grad_norm": 0.162696972489357,
+      "learning_rate": 5.070827116054304e-06,
+      "loss": 0.1419,
+      "step": 9760
+    },
+    {
+      "epoch": 7.347384267971396,
+      "grad_norm": 0.17529311776161194,
+      "learning_rate": 5.07023842636266e-06,
+      "loss": 0.1582,
+      "step": 9761
+    },
+    {
+      "epoch": 7.3481369966127215,
+      "grad_norm": 0.17838211357593536,
+      "learning_rate": 5.069652189556021e-06,
+      "loss": 0.1324,
+      "step": 9762
+    },
+    {
+      "epoch": 7.348889725254046,
+      "grad_norm": 0.16549493372440338,
+      "learning_rate": 5.069068405698493e-06,
+      "loss": 0.1561,
+      "step": 9763
+    },
+    {
+      "epoch": 7.34964245389537,
+      "grad_norm": 0.17745758593082428,
+      "learning_rate": 5.068487074853921e-06,
+      "loss": 0.1385,
+      "step": 9764
+    },
+    {
+      "epoch": 7.350395182536696,
+      "grad_norm": 0.17260216176509857,
+      "learning_rate": 5.067908197085876e-06,
+      "loss": 0.1681,
+      "step": 9765
+    },
+    {
+      "epoch": 7.35114791117802,
+      "grad_norm": 0.17730489373207092,
+      "learning_rate": 5.067331772457664e-06,
+      "loss": 0.1589,
+      "step": 9766
+    },
+    {
+      "epoch": 7.351900639819345,
+      "grad_norm": 0.18561460077762604,
+      "learning_rate": 5.066757801032324e-06,
+      "loss": 0.1726,
+      "step": 9767
+    },
+    {
+      "epoch": 7.35265336846067,
+      "grad_norm": 0.1531103402376175,
+      "learning_rate": 5.06618628287262e-06,
+      "loss": 0.0867,
+      "step": 9768
+    },
+    {
+      "epoch": 7.3534060971019946,
+      "grad_norm": 0.15109340846538544,
+      "learning_rate": 5.065617218041058e-06,
+      "loss": 0.1832,
+      "step": 9769
+    },
+    {
+      "epoch": 7.35415882574332,
+      "grad_norm": 0.16783109307289124,
+      "learning_rate": 5.065050606599867e-06,
+      "loss": 0.1559,
+      "step": 9770
+    },
+    {
+      "epoch": 7.354911554384644,
+      "grad_norm": 0.17415280640125275,
+      "learning_rate": 5.064486448611013e-06,
+      "loss": 0.1417,
+      "step": 9771
+    },
+    {
+      "epoch": 7.355664283025969,
+      "grad_norm": 0.15091277658939362,
+      "learning_rate": 5.0639247441361875e-06,
+      "loss": 0.1507,
+      "step": 9772
+    },
+    {
+      "epoch": 7.356417011667294,
+      "grad_norm": 0.18661390244960785,
+      "learning_rate": 5.063365493236819e-06,
+      "loss": 0.2434,
+      "step": 9773
+    },
+    {
+      "epoch": 7.357169740308619,
+      "grad_norm": 0.16362755000591278,
+      "learning_rate": 5.062808695974068e-06,
+      "loss": 0.2521,
+      "step": 9774
+    },
+    {
+      "epoch": 7.357922468949943,
+      "grad_norm": 0.1708521544933319,
+      "learning_rate": 5.062254352408825e-06,
+      "loss": 0.1666,
+      "step": 9775
+    },
+    {
+      "epoch": 7.3586751975912685,
+      "grad_norm": 0.16093102097511292,
+      "learning_rate": 5.061702462601713e-06,
+      "loss": 0.1808,
+      "step": 9776
+    },
+    {
+      "epoch": 7.359427926232593,
+      "grad_norm": 0.18480870127677917,
+      "learning_rate": 5.061153026613082e-06,
+      "loss": 0.2195,
+      "step": 9777
+    },
+    {
+      "epoch": 7.360180654873918,
+      "grad_norm": 0.17309071123600006,
+      "learning_rate": 5.060606044503018e-06,
+      "loss": 0.1028,
+      "step": 9778
+    },
+    {
+      "epoch": 7.360933383515243,
+      "grad_norm": 0.1574009507894516,
+      "learning_rate": 5.060061516331343e-06,
+      "loss": 0.1427,
+      "step": 9779
+    },
+    {
+      "epoch": 7.361686112156567,
+      "grad_norm": 0.18434906005859375,
+      "learning_rate": 5.059519442157602e-06,
+      "loss": 0.1608,
+      "step": 9780
+    },
+    {
+      "epoch": 7.362438840797893,
+      "grad_norm": 0.18128813803195953,
+      "learning_rate": 5.0589798220410735e-06,
+      "loss": 0.236,
+      "step": 9781
+    },
+    {
+      "epoch": 7.363191569439217,
+      "grad_norm": 0.15055593848228455,
+      "learning_rate": 5.058442656040776e-06,
+      "loss": 0.1822,
+      "step": 9782
+    },
+    {
+      "epoch": 7.363944298080542,
+      "grad_norm": 0.17016054689884186,
+      "learning_rate": 5.057907944215446e-06,
+      "loss": 0.1472,
+      "step": 9783
+    },
+    {
+      "epoch": 7.364697026721867,
+      "grad_norm": 0.1731010526418686,
+      "learning_rate": 5.057375686623562e-06,
+      "loss": 0.1229,
+      "step": 9784
+    },
+    {
+      "epoch": 7.3654497553631915,
+      "grad_norm": 0.1658627986907959,
+      "learning_rate": 5.056845883323334e-06,
+      "loss": 0.1309,
+      "step": 9785
+    },
+    {
+      "epoch": 7.366202484004516,
+      "grad_norm": 0.1848636120557785,
+      "learning_rate": 5.056318534372696e-06,
+      "loss": 0.179,
+      "step": 9786
+    },
+    {
+      "epoch": 7.366955212645841,
+      "grad_norm": 0.15824256837368011,
+      "learning_rate": 5.055793639829319e-06,
+      "loss": 0.2044,
+      "step": 9787
+    },
+    {
+      "epoch": 7.367707941287166,
+      "grad_norm": 0.17398875951766968,
+      "learning_rate": 5.055271199750605e-06,
+      "loss": 0.1593,
+      "step": 9788
+    },
+    {
+      "epoch": 7.368460669928491,
+      "grad_norm": 0.16598115861415863,
+      "learning_rate": 5.054751214193687e-06,
+      "loss": 0.124,
+      "step": 9789
+    },
+    {
+      "epoch": 7.369213398569816,
+      "grad_norm": 0.1559397131204605,
+      "learning_rate": 5.05423368321543e-06,
+      "loss": 0.0944,
+      "step": 9790
+    },
+    {
+      "epoch": 7.36996612721114,
+      "grad_norm": 0.1789567470550537,
+      "learning_rate": 5.053718606872433e-06,
+      "loss": 0.1636,
+      "step": 9791
+    },
+    {
+      "epoch": 7.3707188558524654,
+      "grad_norm": 0.1777769774198532,
+      "learning_rate": 5.053205985221021e-06,
+      "loss": 0.2289,
+      "step": 9792
+    },
+    {
+      "epoch": 7.37147158449379,
+      "grad_norm": 0.17517900466918945,
+      "learning_rate": 5.0526958183172534e-06,
+      "loss": 0.1857,
+      "step": 9793
+    },
+    {
+      "epoch": 7.372224313135114,
+      "grad_norm": 0.16545803844928741,
+      "learning_rate": 5.052188106216922e-06,
+      "loss": 0.1351,
+      "step": 9794
+    },
+    {
+      "epoch": 7.37297704177644,
+      "grad_norm": 0.17899641394615173,
+      "learning_rate": 5.0516828489755514e-06,
+      "loss": 0.1051,
+      "step": 9795
+    },
+    {
+      "epoch": 7.373729770417764,
+      "grad_norm": 0.19776751101016998,
+      "learning_rate": 5.051180046648391e-06,
+      "loss": 0.2423,
+      "step": 9796
+    },
+    {
+      "epoch": 7.37448249905909,
+      "grad_norm": 0.17004403471946716,
+      "learning_rate": 5.050679699290432e-06,
+      "loss": 0.1768,
+      "step": 9797
+    },
+    {
+      "epoch": 7.375235227700414,
+      "grad_norm": 0.15088164806365967,
+      "learning_rate": 5.050181806956388e-06,
+      "loss": 0.1269,
+      "step": 9798
+    },
+    {
+      "epoch": 7.3759879563417385,
+      "grad_norm": 0.16764067113399506,
+      "learning_rate": 5.049686369700707e-06,
+      "loss": 0.1055,
+      "step": 9799
+    },
+    {
+      "epoch": 7.376740684983064,
+      "grad_norm": 0.177651509642601,
+      "learning_rate": 5.049193387577574e-06,
+      "loss": 0.1016,
+      "step": 9800
+    },
+    {
+      "epoch": 7.376740684983064,
+      "eval_loss": 0.19921083748340607,
+      "eval_runtime": 452.743,
+      "eval_samples_per_second": 21.264,
+      "eval_steps_per_second": 0.665,
+      "step": 9800
+    },
+    {
+      "epoch": 7.377493413624388,
+      "grad_norm": 0.16572149097919464,
+      "learning_rate": 5.048702860640897e-06,
+      "loss": 0.1694,
+      "step": 9801
+    },
+    {
+      "epoch": 7.378246142265713,
+      "grad_norm": 0.16389374434947968,
+      "learning_rate": 5.048214788944318e-06,
+      "loss": 0.1284,
+      "step": 9802
+    },
+    {
+      "epoch": 7.378998870907038,
+      "grad_norm": 0.1878087818622589,
+      "learning_rate": 5.0477291725412154e-06,
+      "loss": 0.1599,
+      "step": 9803
+    },
+    {
+      "epoch": 7.379751599548363,
+      "grad_norm": 0.17075082659721375,
+      "learning_rate": 5.047246011484695e-06,
+      "loss": 0.1014,
+      "step": 9804
+    },
+    {
+      "epoch": 7.380504328189687,
+      "grad_norm": 0.17821316421031952,
+      "learning_rate": 5.046765305827592e-06,
+      "loss": 0.1638,
+      "step": 9805
+    },
+    {
+      "epoch": 7.3812570568310125,
+      "grad_norm": 0.19534040987491608,
+      "learning_rate": 5.046287055622477e-06,
+      "loss": 0.199,
+      "step": 9806
+    },
+    {
+      "epoch": 7.382009785472337,
+      "grad_norm": 0.15699464082717896,
+      "learning_rate": 5.045811260921653e-06,
+      "loss": 0.1894,
+      "step": 9807
+    },
+    {
+      "epoch": 7.382762514113662,
+      "grad_norm": 0.19357648491859436,
+      "learning_rate": 5.045337921777148e-06,
+      "loss": 0.1913,
+      "step": 9808
+    },
+    {
+      "epoch": 7.383515242754987,
+      "grad_norm": 0.19716566801071167,
+      "learning_rate": 5.04486703824073e-06,
+      "loss": 0.2025,
+      "step": 9809
+    },
+    {
+      "epoch": 7.384267971396311,
+      "grad_norm": 0.17190399765968323,
+      "learning_rate": 5.04439861036389e-06,
+      "loss": 0.1667,
+      "step": 9810
+    },
+    {
+      "epoch": 7.385020700037637,
+      "grad_norm": 0.1578807532787323,
+      "learning_rate": 5.043932638197858e-06,
+      "loss": 0.1648,
+      "step": 9811
+    },
+    {
+      "epoch": 7.385773428678961,
+      "grad_norm": 0.1732514202594757,
+      "learning_rate": 5.043469121793589e-06,
+      "loss": 0.203,
+      "step": 9812
+    },
+    {
+      "epoch": 7.3865261573202865,
+      "grad_norm": 0.16237963736057281,
+      "learning_rate": 5.043008061201774e-06,
+      "loss": 0.0955,
+      "step": 9813
+    },
+    {
+      "epoch": 7.387278885961611,
+      "grad_norm": 0.2059394270181656,
+      "learning_rate": 5.042549456472837e-06,
+      "loss": 0.141,
+      "step": 9814
+    },
+    {
+      "epoch": 7.3880316146029354,
+      "grad_norm": 0.15897466242313385,
+      "learning_rate": 5.042093307656922e-06,
+      "loss": 0.1401,
+      "step": 9815
+    },
+    {
+      "epoch": 7.388784343244261,
+      "grad_norm": 0.1868777722120285,
+      "learning_rate": 5.041639614803923e-06,
+      "loss": 0.1423,
+      "step": 9816
+    },
+    {
+      "epoch": 7.389537071885585,
+      "grad_norm": 0.17519281804561615,
+      "learning_rate": 5.041188377963447e-06,
+      "loss": 0.1218,
+      "step": 9817
+    },
+    {
+      "epoch": 7.39028980052691,
+      "grad_norm": 0.19020310044288635,
+      "learning_rate": 5.040739597184843e-06,
+      "loss": 0.1534,
+      "step": 9818
+    },
+    {
+      "epoch": 7.391042529168235,
+      "grad_norm": 0.19243499636650085,
+      "learning_rate": 5.0402932725171915e-06,
+      "loss": 0.1784,
+      "step": 9819
+    },
+    {
+      "epoch": 7.39179525780956,
+      "grad_norm": 0.16426114737987518,
+      "learning_rate": 5.0398494040093e-06,
+      "loss": 0.0724,
+      "step": 9820
+    },
+    {
+      "epoch": 7.392547986450884,
+      "grad_norm": 0.17870084941387177,
+      "learning_rate": 5.039407991709707e-06,
+      "loss": 0.1361,
+      "step": 9821
+    },
+    {
+      "epoch": 7.393300715092209,
+      "grad_norm": 0.20436504483222961,
+      "learning_rate": 5.038969035666691e-06,
+      "loss": 0.1257,
+      "step": 9822
+    },
+    {
+      "epoch": 7.394053443733534,
+      "grad_norm": 0.1575627624988556,
+      "learning_rate": 5.03853253592825e-06,
+      "loss": 0.0987,
+      "step": 9823
+    },
+    {
+      "epoch": 7.394806172374859,
+      "grad_norm": 0.17085757851600647,
+      "learning_rate": 5.038098492542119e-06,
+      "loss": 0.1564,
+      "step": 9824
+    },
+    {
+      "epoch": 7.395558901016184,
+      "grad_norm": 0.1830836534500122,
+      "learning_rate": 5.037666905555768e-06,
+      "loss": 0.1484,
+      "step": 9825
+    },
+    {
+      "epoch": 7.396311629657508,
+      "grad_norm": 0.16762550175189972,
+      "learning_rate": 5.037237775016389e-06,
+      "loss": 0.161,
+      "step": 9826
+    },
+    {
+      "epoch": 7.397064358298834,
+      "grad_norm": 0.17258968949317932,
+      "learning_rate": 5.03681110097092e-06,
+      "loss": 0.1423,
+      "step": 9827
+    },
+    {
+      "epoch": 7.397817086940158,
+      "grad_norm": 0.17142492532730103,
+      "learning_rate": 5.036386883466013e-06,
+      "loss": 0.1224,
+      "step": 9828
+    },
+    {
+      "epoch": 7.3985698155814825,
+      "grad_norm": 0.1914673000574112,
+      "learning_rate": 5.0359651225480626e-06,
+      "loss": 0.1478,
+      "step": 9829
+    },
+    {
+      "epoch": 7.399322544222808,
+      "grad_norm": 0.15470001101493835,
+      "learning_rate": 5.035545818263192e-06,
+      "loss": 0.0854,
+      "step": 9830
+    },
+    {
+      "epoch": 7.400075272864132,
+      "grad_norm": 0.18175916373729706,
+      "learning_rate": 5.035128970657256e-06,
+      "loss": 0.1706,
+      "step": 9831
+    },
+    {
+      "epoch": 7.400828001505458,
+      "grad_norm": 0.1948997676372528,
+      "learning_rate": 5.034714579775842e-06,
+      "loss": 0.1807,
+      "step": 9832
+    },
+    {
+      "epoch": 7.401580730146782,
+      "grad_norm": 0.16325083374977112,
+      "learning_rate": 5.034302645664266e-06,
+      "loss": 0.225,
+      "step": 9833
+    },
+    {
+      "epoch": 7.402333458788107,
+      "grad_norm": 0.1623910814523697,
+      "learning_rate": 5.033893168367573e-06,
+      "loss": 0.0867,
+      "step": 9834
+    },
+    {
+      "epoch": 7.403086187429432,
+      "grad_norm": 0.18087072670459747,
+      "learning_rate": 5.033486147930549e-06,
+      "loss": 0.121,
+      "step": 9835
+    },
+    {
+      "epoch": 7.4038389160707565,
+      "grad_norm": 0.1492723971605301,
+      "learning_rate": 5.033081584397699e-06,
+      "loss": 0.0731,
+      "step": 9836
+    },
+    {
+      "epoch": 7.404591644712081,
+      "grad_norm": 0.17601977288722992,
+      "learning_rate": 5.03267947781327e-06,
+      "loss": 0.2117,
+      "step": 9837
+    },
+    {
+      "epoch": 7.405344373353406,
+      "grad_norm": 0.16530084609985352,
+      "learning_rate": 5.0322798282212325e-06,
+      "loss": 0.1649,
+      "step": 9838
+    },
+    {
+      "epoch": 7.406097101994731,
+      "grad_norm": 0.18091288208961487,
+      "learning_rate": 5.031882635665297e-06,
+      "loss": 0.1783,
+      "step": 9839
+    },
+    {
+      "epoch": 7.406849830636055,
+      "grad_norm": 0.1792866289615631,
+      "learning_rate": 5.0314879001888915e-06,
+      "loss": 0.1376,
+      "step": 9840
+    },
+    {
+      "epoch": 7.407602559277381,
+      "grad_norm": 0.15493075549602509,
+      "learning_rate": 5.0310956218351915e-06,
+      "loss": 0.1526,
+      "step": 9841
+    },
+    {
+      "epoch": 7.408355287918705,
+      "grad_norm": 0.17027203738689423,
+      "learning_rate": 5.0307058006470915e-06,
+      "loss": 0.1378,
+      "step": 9842
+    },
+    {
+      "epoch": 7.4091080165600305,
+      "grad_norm": 0.18706917762756348,
+      "learning_rate": 5.030318436667224e-06,
+      "loss": 0.1851,
+      "step": 9843
+    },
+    {
+      "epoch": 7.409860745201355,
+      "grad_norm": 0.16297775506973267,
+      "learning_rate": 5.02993352993795e-06,
+      "loss": 0.1851,
+      "step": 9844
+    },
+    {
+      "epoch": 7.410613473842679,
+      "grad_norm": 0.15418070554733276,
+      "learning_rate": 5.02955108050136e-06,
+      "loss": 0.1133,
+      "step": 9845
+    },
+    {
+      "epoch": 7.411366202484005,
+      "grad_norm": 0.17261594533920288,
+      "learning_rate": 5.0291710883992805e-06,
+      "loss": 0.166,
+      "step": 9846
+    },
+    {
+      "epoch": 7.412118931125329,
+      "grad_norm": 0.18105727434158325,
+      "learning_rate": 5.028793553673269e-06,
+      "loss": 0.1623,
+      "step": 9847
+    },
+    {
+      "epoch": 7.412871659766654,
+      "grad_norm": 0.16574756801128387,
+      "learning_rate": 5.0284184763646076e-06,
+      "loss": 0.1438,
+      "step": 9848
+    },
+    {
+      "epoch": 7.413624388407979,
+      "grad_norm": 0.1597142368555069,
+      "learning_rate": 5.028045856514315e-06,
+      "loss": 0.165,
+      "step": 9849
+    },
+    {
+      "epoch": 7.414377117049304,
+      "grad_norm": 0.17988185584545135,
+      "learning_rate": 5.027675694163144e-06,
+      "loss": 0.1787,
+      "step": 9850
+    },
+    {
+      "epoch": 7.415129845690629,
+      "grad_norm": 0.1656467616558075,
+      "learning_rate": 5.02730798935157e-06,
+      "loss": 0.1058,
+      "step": 9851
+    },
+    {
+      "epoch": 7.415882574331953,
+      "grad_norm": 0.18206903338432312,
+      "learning_rate": 5.026942742119808e-06,
+      "loss": 0.1476,
+      "step": 9852
+    },
+    {
+      "epoch": 7.416635302973278,
+      "grad_norm": 0.17815183103084564,
+      "learning_rate": 5.026579952507801e-06,
+      "loss": 0.1508,
+      "step": 9853
+    },
+    {
+      "epoch": 7.417388031614603,
+      "grad_norm": 0.19604098796844482,
+      "learning_rate": 5.026219620555221e-06,
+      "loss": 0.2428,
+      "step": 9854
+    },
+    {
+      "epoch": 7.418140760255928,
+      "grad_norm": 0.1520511358976364,
+      "learning_rate": 5.025861746301474e-06,
+      "loss": 0.178,
+      "step": 9855
+    },
+    {
+      "epoch": 7.418893488897252,
+      "grad_norm": 0.17609646916389465,
+      "learning_rate": 5.025506329785698e-06,
+      "loss": 0.08,
+      "step": 9856
+    },
+    {
+      "epoch": 7.419646217538578,
+      "grad_norm": 0.17449119687080383,
+      "learning_rate": 5.02515337104676e-06,
+      "loss": 0.22,
+      "step": 9857
+    },
+    {
+      "epoch": 7.420398946179902,
+      "grad_norm": 0.16989514231681824,
+      "learning_rate": 5.024802870123257e-06,
+      "loss": 0.2161,
+      "step": 9858
+    },
+    {
+      "epoch": 7.4211516748212265,
+      "grad_norm": 0.1756111979484558,
+      "learning_rate": 5.024454827053523e-06,
+      "loss": 0.1498,
+      "step": 9859
+    },
+    {
+      "epoch": 7.421904403462552,
+      "grad_norm": 0.1825180947780609,
+      "learning_rate": 5.024109241875616e-06,
+      "loss": 0.1688,
+      "step": 9860
+    },
+    {
+      "epoch": 7.422657132103876,
+      "grad_norm": 0.19750112295150757,
+      "learning_rate": 5.023766114627331e-06,
+      "loss": 0.2511,
+      "step": 9861
+    },
+    {
+      "epoch": 7.423409860745202,
+      "grad_norm": 0.16141065955162048,
+      "learning_rate": 5.023425445346194e-06,
+      "loss": 0.1084,
+      "step": 9862
+    },
+    {
+      "epoch": 7.424162589386526,
+      "grad_norm": 0.1645076721906662,
+      "learning_rate": 5.023087234069454e-06,
+      "loss": 0.1855,
+      "step": 9863
+    },
+    {
+      "epoch": 7.424915318027851,
+      "grad_norm": 0.14803555607795715,
+      "learning_rate": 5.022751480834102e-06,
+      "loss": 0.1749,
+      "step": 9864
+    },
+    {
+      "epoch": 7.425668046669176,
+      "grad_norm": 0.16652458906173706,
+      "learning_rate": 5.022418185676852e-06,
+      "loss": 0.1229,
+      "step": 9865
+    },
+    {
+      "epoch": 7.4264207753105005,
+      "grad_norm": 0.1577867865562439,
+      "learning_rate": 5.022087348634158e-06,
+      "loss": 0.1995,
+      "step": 9866
+    },
+    {
+      "epoch": 7.427173503951825,
+      "grad_norm": 0.17088346183300018,
+      "learning_rate": 5.021758969742195e-06,
+      "loss": 0.1002,
+      "step": 9867
+    },
+    {
+      "epoch": 7.42792623259315,
+      "grad_norm": 0.1710389256477356,
+      "learning_rate": 5.021433049036875e-06,
+      "loss": 0.1541,
+      "step": 9868
+    },
+    {
+      "epoch": 7.428678961234475,
+      "grad_norm": 0.15967147052288055,
+      "learning_rate": 5.02110958655384e-06,
+      "loss": 0.1152,
+      "step": 9869
+    },
+    {
+      "epoch": 7.4294316898758,
+      "grad_norm": 0.16468025743961334,
+      "learning_rate": 5.0207885823284656e-06,
+      "loss": 0.2375,
+      "step": 9870
+    },
+    {
+      "epoch": 7.430184418517125,
+      "grad_norm": 0.17886069416999817,
+      "learning_rate": 5.020470036395855e-06,
+      "loss": 0.1665,
+      "step": 9871
+    },
+    {
+      "epoch": 7.430937147158449,
+      "grad_norm": 0.18458415567874908,
+      "learning_rate": 5.020153948790842e-06,
+      "loss": 0.1801,
+      "step": 9872
+    },
+    {
+      "epoch": 7.4316898757997745,
+      "grad_norm": 0.17392830550670624,
+      "learning_rate": 5.019840319547996e-06,
+      "loss": 0.143,
+      "step": 9873
+    },
+    {
+      "epoch": 7.432442604441099,
+      "grad_norm": 0.18404322862625122,
+      "learning_rate": 5.0195291487016154e-06,
+      "loss": 0.2193,
+      "step": 9874
+    },
+    {
+      "epoch": 7.433195333082423,
+      "grad_norm": 0.18353281915187836,
+      "learning_rate": 5.019220436285728e-06,
+      "loss": 0.1895,
+      "step": 9875
+    },
+    {
+      "epoch": 7.433948061723749,
+      "grad_norm": 0.18055714666843414,
+      "learning_rate": 5.018914182334094e-06,
+      "loss": 0.2396,
+      "step": 9876
+    },
+    {
+      "epoch": 7.434700790365073,
+      "grad_norm": 0.19978861510753632,
+      "learning_rate": 5.0186103868802035e-06,
+      "loss": 0.0981,
+      "step": 9877
+    },
+    {
+      "epoch": 7.435453519006398,
+      "grad_norm": 0.16415461897850037,
+      "learning_rate": 5.0183090499572844e-06,
+      "loss": 0.1249,
+      "step": 9878
+    },
+    {
+      "epoch": 7.436206247647723,
+      "grad_norm": 0.1704832762479782,
+      "learning_rate": 5.018010171598282e-06,
+      "loss": 0.1493,
+      "step": 9879
+    },
+    {
+      "epoch": 7.436958976289048,
+      "grad_norm": 0.17545543611049652,
+      "learning_rate": 5.017713751835889e-06,
+      "loss": 0.0993,
+      "step": 9880
+    },
+    {
+      "epoch": 7.437711704930373,
+      "grad_norm": 0.163346529006958,
+      "learning_rate": 5.017419790702519e-06,
+      "loss": 0.1751,
+      "step": 9881
+    },
+    {
+      "epoch": 7.438464433571697,
+      "grad_norm": 0.16161894798278809,
+      "learning_rate": 5.017128288230318e-06,
+      "loss": 0.1212,
+      "step": 9882
+    },
+    {
+      "epoch": 7.439217162213022,
+      "grad_norm": 0.16408632695674896,
+      "learning_rate": 5.016839244451164e-06,
+      "loss": 0.1243,
+      "step": 9883
+    },
+    {
+      "epoch": 7.439969890854347,
+      "grad_norm": 0.17883709073066711,
+      "learning_rate": 5.01655265939667e-06,
+      "loss": 0.2321,
+      "step": 9884
+    },
+    {
+      "epoch": 7.440722619495672,
+      "grad_norm": 0.17038625478744507,
+      "learning_rate": 5.016268533098171e-06,
+      "loss": 0.1753,
+      "step": 9885
+    },
+    {
+      "epoch": 7.441475348136997,
+      "grad_norm": 0.1613932102918625,
+      "learning_rate": 5.015986865586743e-06,
+      "loss": 0.1225,
+      "step": 9886
+    },
+    {
+      "epoch": 7.4422280767783215,
+      "grad_norm": 0.1901073008775711,
+      "learning_rate": 5.015707656893186e-06,
+      "loss": 0.1941,
+      "step": 9887
+    },
+    {
+      "epoch": 7.442980805419646,
+      "grad_norm": 0.17055179178714752,
+      "learning_rate": 5.015430907048035e-06,
+      "loss": 0.1989,
+      "step": 9888
+    },
+    {
+      "epoch": 7.443733534060971,
+      "grad_norm": 0.15681254863739014,
+      "learning_rate": 5.015156616081555e-06,
+      "loss": 0.1307,
+      "step": 9889
+    },
+    {
+      "epoch": 7.444486262702296,
+      "grad_norm": 0.15084208548069,
+      "learning_rate": 5.014884784023744e-06,
+      "loss": 0.1424,
+      "step": 9890
+    },
+    {
+      "epoch": 7.44523899134362,
+      "grad_norm": 0.18793655931949615,
+      "learning_rate": 5.014615410904326e-06,
+      "loss": 0.1556,
+      "step": 9891
+    },
+    {
+      "epoch": 7.445991719984946,
+      "grad_norm": 0.1453855037689209,
+      "learning_rate": 5.01434849675276e-06,
+      "loss": 0.1054,
+      "step": 9892
+    },
+    {
+      "epoch": 7.44674444862627,
+      "grad_norm": 0.1713564246892929,
+      "learning_rate": 5.014084041598236e-06,
+      "loss": 0.2131,
+      "step": 9893
+    },
+    {
+      "epoch": 7.447497177267595,
+      "grad_norm": 0.18135051429271698,
+      "learning_rate": 5.013822045469673e-06,
+      "loss": 0.2028,
+      "step": 9894
+    },
+    {
+      "epoch": 7.44824990590892,
+      "grad_norm": 0.18971742689609528,
+      "learning_rate": 5.013562508395723e-06,
+      "loss": 0.1709,
+      "step": 9895
+    },
+    {
+      "epoch": 7.4490026345502445,
+      "grad_norm": 0.16657386720180511,
+      "learning_rate": 5.013305430404771e-06,
+      "loss": 0.1802,
+      "step": 9896
+    },
+    {
+      "epoch": 7.44975536319157,
+      "grad_norm": 0.18605156242847443,
+      "learning_rate": 5.013050811524931e-06,
+      "loss": 0.2075,
+      "step": 9897
+    },
+    {
+      "epoch": 7.450508091832894,
+      "grad_norm": 0.15059299767017365,
+      "learning_rate": 5.012798651784043e-06,
+      "loss": 0.1163,
+      "step": 9898
+    },
+    {
+      "epoch": 7.451260820474219,
+      "grad_norm": 0.16712887585163116,
+      "learning_rate": 5.012548951209686e-06,
+      "loss": 0.1692,
+      "step": 9899
+    },
+    {
+      "epoch": 7.452013549115544,
+      "grad_norm": 0.18686091899871826,
+      "learning_rate": 5.012301709829164e-06,
+      "loss": 0.1208,
+      "step": 9900
+    },
+    {
+      "epoch": 7.452766277756869,
+      "grad_norm": 0.18380074203014374,
+      "learning_rate": 5.01205692766952e-06,
+      "loss": 0.2131,
+      "step": 9901
+    },
+    {
+      "epoch": 7.453519006398193,
+      "grad_norm": 0.1679258942604065,
+      "learning_rate": 5.011814604757521e-06,
+      "loss": 0.1761,
+      "step": 9902
+    },
+    {
+      "epoch": 7.4542717350395185,
+      "grad_norm": 0.18939349055290222,
+      "learning_rate": 5.011574741119665e-06,
+      "loss": 0.1785,
+      "step": 9903
+    },
+    {
+      "epoch": 7.455024463680843,
+      "grad_norm": 0.1904030293226242,
+      "learning_rate": 5.011337336782183e-06,
+      "loss": 0.1805,
+      "step": 9904
+    },
+    {
+      "epoch": 7.455777192322168,
+      "grad_norm": 0.1884366273880005,
+      "learning_rate": 5.01110239177104e-06,
+      "loss": 0.1218,
+      "step": 9905
+    },
+    {
+      "epoch": 7.456529920963493,
+      "grad_norm": 0.17975223064422607,
+      "learning_rate": 5.010869906111927e-06,
+      "loss": 0.1901,
+      "step": 9906
+    },
+    {
+      "epoch": 7.457282649604817,
+      "grad_norm": 0.15619507431983948,
+      "learning_rate": 5.0106398798302704e-06,
+      "loss": 0.1755,
+      "step": 9907
+    },
+    {
+      "epoch": 7.458035378246143,
+      "grad_norm": 0.1725010722875595,
+      "learning_rate": 5.010412312951225e-06,
+      "loss": 0.1747,
+      "step": 9908
+    },
+    {
+      "epoch": 7.458788106887467,
+      "grad_norm": 0.17267660796642303,
+      "learning_rate": 5.010187205499676e-06,
+      "loss": 0.1072,
+      "step": 9909
+    },
+    {
+      "epoch": 7.4595408355287915,
+      "grad_norm": 0.17998217046260834,
+      "learning_rate": 5.009964557500241e-06,
+      "loss": 0.1804,
+      "step": 9910
+    },
+    {
+      "epoch": 7.460293564170117,
+      "grad_norm": 0.1505386084318161,
+      "learning_rate": 5.009744368977269e-06,
+      "loss": 0.1413,
+      "step": 9911
+    },
+    {
+      "epoch": 7.461046292811441,
+      "grad_norm": 0.18383261561393738,
+      "learning_rate": 5.009526639954837e-06,
+      "loss": 0.1697,
+      "step": 9912
+    },
+    {
+      "epoch": 7.461799021452766,
+      "grad_norm": 0.16838715970516205,
+      "learning_rate": 5.00931137045676e-06,
+      "loss": 0.1371,
+      "step": 9913
+    },
+    {
+      "epoch": 7.462551750094091,
+      "grad_norm": 0.17266912758350372,
+      "learning_rate": 5.0090985605065756e-06,
+      "loss": 0.1343,
+      "step": 9914
+    },
+    {
+      "epoch": 7.463304478735416,
+      "grad_norm": 0.18254931271076202,
+      "learning_rate": 5.0088882101275604e-06,
+      "loss": 0.1714,
+      "step": 9915
+    },
+    {
+      "epoch": 7.464057207376741,
+      "grad_norm": 0.173904687166214,
+      "learning_rate": 5.008680319342714e-06,
+      "loss": 0.153,
+      "step": 9916
+    },
+    {
+      "epoch": 7.4648099360180655,
+      "grad_norm": 0.17871741950511932,
+      "learning_rate": 5.008474888174772e-06,
+      "loss": 0.1658,
+      "step": 9917
+    },
+    {
+      "epoch": 7.46556266465939,
+      "grad_norm": 0.17572855949401855,
+      "learning_rate": 5.0082719166462024e-06,
+      "loss": 0.1769,
+      "step": 9918
+    },
+    {
+      "epoch": 7.466315393300715,
+      "grad_norm": 0.16883620619773865,
+      "learning_rate": 5.0080714047792e-06,
+      "loss": 0.2053,
+      "step": 9919
+    },
+    {
+      "epoch": 7.46706812194204,
+      "grad_norm": 0.1781918704509735,
+      "learning_rate": 5.007873352595692e-06,
+      "loss": 0.1718,
+      "step": 9920
+    },
+    {
+      "epoch": 7.467820850583364,
+      "grad_norm": 0.15462233126163483,
+      "learning_rate": 5.007677760117338e-06,
+      "loss": 0.1527,
+      "step": 9921
+    },
+    {
+      "epoch": 7.46857357922469,
+      "grad_norm": 0.16755259037017822,
+      "learning_rate": 5.007484627365527e-06,
+      "loss": 0.1261,
+      "step": 9922
+    },
+    {
+      "epoch": 7.469326307866014,
+      "grad_norm": 0.18726646900177002,
+      "learning_rate": 5.007293954361381e-06,
+      "loss": 0.2454,
+      "step": 9923
+    },
+    {
+      "epoch": 7.4700790365073395,
+      "grad_norm": 0.16527734696865082,
+      "learning_rate": 5.007105741125752e-06,
+      "loss": 0.1213,
+      "step": 9924
+    },
+    {
+      "epoch": 7.470831765148664,
+      "grad_norm": 0.168714702129364,
+      "learning_rate": 5.00691998767922e-06,
+      "loss": 0.1104,
+      "step": 9925
+    },
+    {
+      "epoch": 7.4715844937899885,
+      "grad_norm": 0.1640613079071045,
+      "learning_rate": 5.006736694042099e-06,
+      "loss": 0.1086,
+      "step": 9926
+    },
+    {
+      "epoch": 7.472337222431314,
+      "grad_norm": 0.16224412620067596,
+      "learning_rate": 5.006555860234436e-06,
+      "loss": 0.0988,
+      "step": 9927
+    },
+    {
+      "epoch": 7.473089951072638,
+      "grad_norm": 0.15428531169891357,
+      "learning_rate": 5.006377486276007e-06,
+      "loss": 0.1706,
+      "step": 9928
+    },
+    {
+      "epoch": 7.473842679713963,
+      "grad_norm": 0.18764428794384003,
+      "learning_rate": 5.006201572186317e-06,
+      "loss": 0.185,
+      "step": 9929
+    },
+    {
+      "epoch": 7.474595408355288,
+      "grad_norm": 0.18537500500679016,
+      "learning_rate": 5.006028117984602e-06,
+      "loss": 0.19,
+      "step": 9930
+    },
+    {
+      "epoch": 7.475348136996613,
+      "grad_norm": 0.16120761632919312,
+      "learning_rate": 5.0058571236898366e-06,
+      "loss": 0.1225,
+      "step": 9931
+    },
+    {
+      "epoch": 7.476100865637937,
+      "grad_norm": 0.1707906275987625,
+      "learning_rate": 5.005688589320714e-06,
+      "loss": 0.1588,
+      "step": 9932
+    },
+    {
+      "epoch": 7.476853594279262,
+      "grad_norm": 0.16920123994350433,
+      "learning_rate": 5.0055225148956695e-06,
+      "loss": 0.1658,
+      "step": 9933
+    },
+    {
+      "epoch": 7.477606322920587,
+      "grad_norm": 0.16792544722557068,
+      "learning_rate": 5.005358900432859e-06,
+      "loss": 0.1423,
+      "step": 9934
+    },
+    {
+      "epoch": 7.478359051561912,
+      "grad_norm": 0.17540551722049713,
+      "learning_rate": 5.005197745950183e-06,
+      "loss": 0.277,
+      "step": 9935
+    },
+    {
+      "epoch": 7.479111780203237,
+      "grad_norm": 0.18997682631015778,
+      "learning_rate": 5.00503905146526e-06,
+      "loss": 0.1656,
+      "step": 9936
+    },
+    {
+      "epoch": 7.479864508844561,
+      "grad_norm": 0.17891763150691986,
+      "learning_rate": 5.004882816995443e-06,
+      "loss": 0.1235,
+      "step": 9937
+    },
+    {
+      "epoch": 7.480617237485887,
+      "grad_norm": 0.18761782348155975,
+      "learning_rate": 5.004729042557825e-06,
+      "loss": 0.1712,
+      "step": 9938
+    },
+    {
+      "epoch": 7.481369966127211,
+      "grad_norm": 0.19660481810569763,
+      "learning_rate": 5.0045777281692135e-06,
+      "loss": 0.2285,
+      "step": 9939
+    },
+    {
+      "epoch": 7.4821226947685355,
+      "grad_norm": 0.1804434210062027,
+      "learning_rate": 5.004428873846163e-06,
+      "loss": 0.2243,
+      "step": 9940
+    },
+    {
+      "epoch": 7.482875423409861,
+      "grad_norm": 0.17224887013435364,
+      "learning_rate": 5.004282479604949e-06,
+      "loss": 0.1342,
+      "step": 9941
+    },
+    {
+      "epoch": 7.483628152051185,
+      "grad_norm": 0.16320247948169708,
+      "learning_rate": 5.00413854546158e-06,
+      "loss": 0.1127,
+      "step": 9942
+    },
+    {
+      "epoch": 7.484380880692511,
+      "grad_norm": 0.1805265098810196,
+      "learning_rate": 5.003997071431799e-06,
+      "loss": 0.1649,
+      "step": 9943
+    },
+    {
+      "epoch": 7.485133609333835,
+      "grad_norm": 0.1902584433555603,
+      "learning_rate": 5.003858057531073e-06,
+      "loss": 0.2217,
+      "step": 9944
+    },
+    {
+      "epoch": 7.48588633797516,
+      "grad_norm": 0.17519675195217133,
+      "learning_rate": 5.0037215037746114e-06,
+      "loss": 0.1761,
+      "step": 9945
+    },
+    {
+      "epoch": 7.486639066616485,
+      "grad_norm": 0.16712571680545807,
+      "learning_rate": 5.00358741017734e-06,
+      "loss": 0.1551,
+      "step": 9946
+    },
+    {
+      "epoch": 7.4873917952578095,
+      "grad_norm": 0.18451783061027527,
+      "learning_rate": 5.003455776753929e-06,
+      "loss": 0.1986,
+      "step": 9947
+    },
+    {
+      "epoch": 7.488144523899134,
+      "grad_norm": 0.16739137470722198,
+      "learning_rate": 5.0033266035187705e-06,
+      "loss": 0.0983,
+      "step": 9948
+    },
+    {
+      "epoch": 7.488897252540459,
+      "grad_norm": 0.14789532124996185,
+      "learning_rate": 5.003199890485991e-06,
+      "loss": 0.2077,
+      "step": 9949
+    },
+    {
+      "epoch": 7.489649981181784,
+      "grad_norm": 0.19321487843990326,
+      "learning_rate": 5.003075637669448e-06,
+      "loss": 0.1605,
+      "step": 9950
+    },
+    {
+      "epoch": 7.490402709823109,
+      "grad_norm": 0.1644616574048996,
+      "learning_rate": 5.002953845082732e-06,
+      "loss": 0.1949,
+      "step": 9951
+    },
+    {
+      "epoch": 7.491155438464434,
+      "grad_norm": 0.17519335448741913,
+      "learning_rate": 5.002834512739158e-06,
+      "loss": 0.2447,
+      "step": 9952
+    },
+    {
+      "epoch": 7.491908167105758,
+      "grad_norm": 0.17243580520153046,
+      "learning_rate": 5.002717640651775e-06,
+      "loss": 0.1738,
+      "step": 9953
+    },
+    {
+      "epoch": 7.4926608957470835,
+      "grad_norm": 0.15753324329853058,
+      "learning_rate": 5.002603228833371e-06,
+      "loss": 0.1208,
+      "step": 9954
+    },
+    {
+      "epoch": 7.493413624388408,
+      "grad_norm": 0.17834587395191193,
+      "learning_rate": 5.00249127729645e-06,
+      "loss": 0.1521,
+      "step": 9955
+    },
+    {
+      "epoch": 7.494166353029732,
+      "grad_norm": 0.17092958092689514,
+      "learning_rate": 5.00238178605326e-06,
+      "loss": 0.0849,
+      "step": 9956
+    },
+    {
+      "epoch": 7.494919081671058,
+      "grad_norm": 0.17333608865737915,
+      "learning_rate": 5.002274755115775e-06,
+      "loss": 0.1178,
+      "step": 9957
+    },
+    {
+      "epoch": 7.495671810312382,
+      "grad_norm": 0.18905170261859894,
+      "learning_rate": 5.002170184495694e-06,
+      "loss": 0.1802,
+      "step": 9958
+    },
+    {
+      "epoch": 7.496424538953708,
+      "grad_norm": 0.16138394176959991,
+      "learning_rate": 5.002068074204459e-06,
+      "loss": 0.1923,
+      "step": 9959
+    },
+    {
+      "epoch": 7.497177267595032,
+      "grad_norm": 0.20963206887245178,
+      "learning_rate": 5.001968424253237e-06,
+      "loss": 0.3089,
+      "step": 9960
+    },
+    {
+      "epoch": 7.497929996236357,
+      "grad_norm": 0.1471710205078125,
+      "learning_rate": 5.001871234652919e-06,
+      "loss": 0.1566,
+      "step": 9961
+    },
+    {
+      "epoch": 7.498682724877682,
+      "grad_norm": 0.18470756709575653,
+      "learning_rate": 5.001776505414138e-06,
+      "loss": 0.1912,
+      "step": 9962
+    },
+    {
+      "epoch": 7.499435453519006,
+      "grad_norm": 0.17113684117794037,
+      "learning_rate": 5.0016842365472535e-06,
+      "loss": 0.1465,
+      "step": 9963
+    },
+    {
+      "epoch": 7.500188182160331,
+      "grad_norm": 0.174101322889328,
+      "learning_rate": 5.0015944280623546e-06,
+      "loss": 0.1952,
+      "step": 9964
+    },
+    {
+      "epoch": 7.500940910801656,
+      "grad_norm": 0.1573587954044342,
+      "learning_rate": 5.0015070799692635e-06,
+      "loss": 0.1104,
+      "step": 9965
+    },
+    {
+      "epoch": 7.501693639442981,
+      "grad_norm": 0.1777660995721817,
+      "learning_rate": 5.001422192277533e-06,
+      "loss": 0.1991,
+      "step": 9966
+    },
+    {
+      "epoch": 7.502446368084305,
+      "grad_norm": 0.18216216564178467,
+      "learning_rate": 5.001339764996442e-06,
+      "loss": 0.1875,
+      "step": 9967
+    },
+    {
+      "epoch": 7.503199096725631,
+      "grad_norm": 0.18464192748069763,
+      "learning_rate": 5.001259798135012e-06,
+      "loss": 0.1422,
+      "step": 9968
+    },
+    {
+      "epoch": 7.503951825366955,
+      "grad_norm": 0.1592063158750534,
+      "learning_rate": 5.001182291701983e-06,
+      "loss": 0.1501,
+      "step": 9969
+    },
+    {
+      "epoch": 7.5047045540082795,
+      "grad_norm": 0.17425912618637085,
+      "learning_rate": 5.001107245705831e-06,
+      "loss": 0.157,
+      "step": 9970
+    },
+    {
+      "epoch": 7.505457282649605,
+      "grad_norm": 0.16778632998466492,
+      "learning_rate": 5.001034660154766e-06,
+      "loss": 0.109,
+      "step": 9971
+    },
+    {
+      "epoch": 7.506210011290929,
+      "grad_norm": 0.18372581899166107,
+      "learning_rate": 5.000964535056722e-06,
+      "loss": 0.2092,
+      "step": 9972
+    },
+    {
+      "epoch": 7.506962739932255,
+      "grad_norm": 0.16823649406433105,
+      "learning_rate": 5.000896870419369e-06,
+      "loss": 0.1319,
+      "step": 9973
+    },
+    {
+      "epoch": 7.507715468573579,
+      "grad_norm": 0.19092750549316406,
+      "learning_rate": 5.00083166625011e-06,
+      "loss": 0.1644,
+      "step": 9974
+    },
+    {
+      "epoch": 7.508468197214904,
+      "grad_norm": 0.18487070500850677,
+      "learning_rate": 5.000768922556072e-06,
+      "loss": 0.1403,
+      "step": 9975
+    },
+    {
+      "epoch": 7.509220925856229,
+      "grad_norm": 0.17287452518939972,
+      "learning_rate": 5.000708639344118e-06,
+      "loss": 0.1559,
+      "step": 9976
+    },
+    {
+      "epoch": 7.5099736544975535,
+      "grad_norm": 0.16271868348121643,
+      "learning_rate": 5.0006508166208384e-06,
+      "loss": 0.1623,
+      "step": 9977
+    },
+    {
+      "epoch": 7.510726383138879,
+      "grad_norm": 0.17279645800590515,
+      "learning_rate": 5.000595454392559e-06,
+      "loss": 0.107,
+      "step": 9978
+    },
+    {
+      "epoch": 7.511479111780203,
+      "grad_norm": 0.17229324579238892,
+      "learning_rate": 5.0005425526653325e-06,
+      "loss": 0.0932,
+      "step": 9979
+    },
+    {
+      "epoch": 7.512231840421528,
+      "grad_norm": 0.18202410638332367,
+      "learning_rate": 5.000492111444948e-06,
+      "loss": 0.1639,
+      "step": 9980
+    },
+    {
+      "epoch": 7.512984569062853,
+      "grad_norm": 0.1862594336271286,
+      "learning_rate": 5.0004441307369155e-06,
+      "loss": 0.2421,
+      "step": 9981
+    },
+    {
+      "epoch": 7.513737297704178,
+      "grad_norm": 0.2099241018295288,
+      "learning_rate": 5.0003986105464864e-06,
+      "loss": 0.181,
+      "step": 9982
+    },
+    {
+      "epoch": 7.514490026345502,
+      "grad_norm": 0.1736823171377182,
+      "learning_rate": 5.000355550878637e-06,
+      "loss": 0.1816,
+      "step": 9983
+    },
+    {
+      "epoch": 7.5152427549868275,
+      "grad_norm": 0.18118782341480255,
+      "learning_rate": 5.000314951738077e-06,
+      "loss": 0.1517,
+      "step": 9984
+    },
+    {
+      "epoch": 7.515995483628152,
+      "grad_norm": 0.16741518676280975,
+      "learning_rate": 5.000276813129247e-06,
+      "loss": 0.1504,
+      "step": 9985
+    },
+    {
+      "epoch": 7.516748212269476,
+      "grad_norm": 0.18985189497470856,
+      "learning_rate": 5.000241135056316e-06,
+      "loss": 0.1504,
+      "step": 9986
+    },
+    {
+      "epoch": 7.517500940910802,
+      "grad_norm": 0.16710835695266724,
+      "learning_rate": 5.000207917523189e-06,
+      "loss": 0.15,
+      "step": 9987
+    },
+    {
+      "epoch": 7.518253669552126,
+      "grad_norm": 0.17216521501541138,
+      "learning_rate": 5.000177160533494e-06,
+      "loss": 0.1392,
+      "step": 9988
+    },
+    {
+      "epoch": 7.519006398193452,
+      "grad_norm": 0.1765749454498291,
+      "learning_rate": 5.000148864090595e-06,
+      "loss": 0.1631,
+      "step": 9989
+    },
+    {
+      "epoch": 7.519759126834776,
+      "grad_norm": 0.1582346111536026,
+      "learning_rate": 5.00012302819759e-06,
+      "loss": 0.2207,
+      "step": 9990
+    },
+    {
+      "epoch": 7.520511855476101,
+      "grad_norm": 0.14430560171604156,
+      "learning_rate": 5.000099652857304e-06,
+      "loss": 0.1063,
+      "step": 9991
+    },
+    {
+      "epoch": 7.521264584117426,
+      "grad_norm": 0.18231065571308136,
+      "learning_rate": 5.00007873807229e-06,
+      "loss": 0.1964,
+      "step": 9992
+    },
+    {
+      "epoch": 7.52201731275875,
+      "grad_norm": 0.1622893363237381,
+      "learning_rate": 5.000060283844839e-06,
+      "loss": 0.1064,
+      "step": 9993
+    },
+    {
+      "epoch": 7.522770041400076,
+      "grad_norm": 0.1678018718957901,
+      "learning_rate": 5.000044290176964e-06,
+      "loss": 0.2108,
+      "step": 9994
+    },
+    {
+      "epoch": 7.5235227700414,
+      "grad_norm": 0.15341037511825562,
+      "learning_rate": 5.000030757070421e-06,
+      "loss": 0.1657,
+      "step": 9995
+    },
+    {
+      "epoch": 7.524275498682725,
+      "grad_norm": 0.19295717775821686,
+      "learning_rate": 5.0000196845266845e-06,
+      "loss": 0.1967,
+      "step": 9996
+    },
+    {
+      "epoch": 7.52502822732405,
+      "grad_norm": 0.16695430874824524,
+      "learning_rate": 5.000011072546967e-06,
+      "loss": 0.1423,
+      "step": 9997
+    },
+    {
+      "epoch": 7.5257809559653746,
+      "grad_norm": 0.1833392232656479,
+      "learning_rate": 5.00000492113221e-06,
+      "loss": 0.2427,
+      "step": 9998
+    },
+    {
+      "epoch": 7.526533684606699,
+      "grad_norm": 0.15590594708919525,
+      "learning_rate": 5.000001230283086e-06,
+      "loss": 0.119,
+      "step": 9999
+    },
+    {
+      "epoch": 7.527286413248024,
+      "grad_norm": 0.17052628099918365,
+      "learning_rate": 5e-06,
+      "loss": 0.1621,
+      "step": 10000
+    },
+    {
+      "epoch": 7.527286413248024,
+      "eval_loss": 0.1988229751586914,
+      "eval_runtime": 452.7736,
+      "eval_samples_per_second": 21.262,
+      "eval_steps_per_second": 0.665,
+      "step": 10000
+    },
+    {
+      "epoch": 7.528039141889349,
+      "grad_norm": 0.16024969518184662,
+      "learning_rate": 5e-06,
+      "loss": 0.1718,
+      "step": 10001
+    },
+    {
+      "epoch": 7.528791870530673,
+      "grad_norm": 0.17039638757705688,
+      "learning_rate": 5e-06,
+      "loss": 0.24,
+      "step": 10002
+    },
+    {
+      "epoch": 7.529544599171999,
+      "grad_norm": 0.16877350211143494,
+      "learning_rate": 5e-06,
+      "loss": 0.1869,
+      "step": 10003
+    },
+    {
+      "epoch": 7.530297327813323,
+      "grad_norm": 0.16441795229911804,
+      "learning_rate": 5e-06,
+      "loss": 0.0772,
+      "step": 10004
+    },
+    {
+      "epoch": 7.531050056454648,
+      "grad_norm": 0.19938969612121582,
+      "learning_rate": 5e-06,
+      "loss": 0.1977,
+      "step": 10005
+    },
+    {
+      "epoch": 7.531802785095973,
+      "grad_norm": 0.1801726520061493,
+      "learning_rate": 5e-06,
+      "loss": 0.1265,
+      "step": 10006
+    },
+    {
+      "epoch": 7.5325555137372975,
+      "grad_norm": 0.17723821103572845,
+      "learning_rate": 5e-06,
+      "loss": 0.1378,
+      "step": 10007
+    },
+    {
+      "epoch": 7.533308242378623,
+      "grad_norm": 0.1768878996372223,
+      "learning_rate": 5e-06,
+      "loss": 0.1606,
+      "step": 10008
+    },
+    {
+      "epoch": 7.534060971019947,
+      "grad_norm": 0.19865447282791138,
+      "learning_rate": 5e-06,
+      "loss": 0.1995,
+      "step": 10009
+    },
+    {
+      "epoch": 7.534813699661272,
+      "grad_norm": 0.21433025598526,
+      "learning_rate": 5e-06,
+      "loss": 0.2278,
+      "step": 10010
+    },
+    {
+      "epoch": 7.535566428302597,
+      "grad_norm": 0.17212240397930145,
+      "learning_rate": 5e-06,
+      "loss": 0.1673,
+      "step": 10011
+    },
+    {
+      "epoch": 7.536319156943922,
+      "grad_norm": 0.1974761188030243,
+      "learning_rate": 5e-06,
+      "loss": 0.2384,
+      "step": 10012
+    },
+    {
+      "epoch": 7.537071885585247,
+      "grad_norm": 0.17007741332054138,
+      "learning_rate": 5e-06,
+      "loss": 0.1443,
+      "step": 10013
+    },
+    {
+      "epoch": 7.5378246142265715,
+      "grad_norm": 0.16719643771648407,
+      "learning_rate": 5e-06,
+      "loss": 0.1366,
+      "step": 10014
+    },
+    {
+      "epoch": 7.538577342867896,
+      "grad_norm": 0.17680832743644714,
+      "learning_rate": 5e-06,
+      "loss": 0.2001,
+      "step": 10015
+    },
+    {
+      "epoch": 7.539330071509221,
+      "grad_norm": 0.18232429027557373,
+      "learning_rate": 5e-06,
+      "loss": 0.2074,
+      "step": 10016
+    },
+    {
+      "epoch": 7.540082800150546,
+      "grad_norm": 0.13250592350959778,
+      "learning_rate": 5e-06,
+      "loss": 0.1558,
+      "step": 10017
+    },
+    {
+      "epoch": 7.54083552879187,
+      "grad_norm": 0.18771328032016754,
+      "learning_rate": 5e-06,
+      "loss": 0.1813,
+      "step": 10018
+    },
+    {
+      "epoch": 7.541588257433196,
+      "grad_norm": 0.1777888685464859,
+      "learning_rate": 5e-06,
+      "loss": 0.1532,
+      "step": 10019
+    },
+    {
+      "epoch": 7.54234098607452,
+      "grad_norm": 0.16563457250595093,
+      "learning_rate": 5e-06,
+      "loss": 0.1872,
+      "step": 10020
+    },
+    {
+      "epoch": 7.5430937147158446,
+      "grad_norm": 0.16147390007972717,
+      "learning_rate": 5e-06,
+      "loss": 0.145,
+      "step": 10021
+    },
+    {
+      "epoch": 7.54384644335717,
+      "grad_norm": 0.16373394429683685,
+      "learning_rate": 5e-06,
+      "loss": 0.2179,
+      "step": 10022
+    },
+    {
+      "epoch": 7.544599171998494,
+      "grad_norm": 0.1867450624704361,
+      "learning_rate": 5e-06,
+      "loss": 0.1474,
+      "step": 10023
+    },
+    {
+      "epoch": 7.545351900639819,
+      "grad_norm": 0.16407430171966553,
+      "learning_rate": 5e-06,
+      "loss": 0.1135,
+      "step": 10024
+    },
+    {
+      "epoch": 7.546104629281144,
+      "grad_norm": 0.16905532777309418,
+      "learning_rate": 5e-06,
+      "loss": 0.1322,
+      "step": 10025
+    },
+    {
+      "epoch": 7.546857357922469,
+      "grad_norm": 0.17296622693538666,
+      "learning_rate": 5e-06,
+      "loss": 0.1728,
+      "step": 10026
+    },
+    {
+      "epoch": 7.547610086563794,
+      "grad_norm": 0.1524672657251358,
+      "learning_rate": 5e-06,
+      "loss": 0.1178,
+      "step": 10027
+    },
+    {
+      "epoch": 7.5483628152051185,
+      "grad_norm": 0.1500646322965622,
+      "learning_rate": 5e-06,
+      "loss": 0.1614,
+      "step": 10028
+    },
+    {
+      "epoch": 7.549115543846443,
+      "grad_norm": 0.15586933493614197,
+      "learning_rate": 5e-06,
+      "loss": 0.106,
+      "step": 10029
+    },
+    {
+      "epoch": 7.549868272487768,
+      "grad_norm": 0.17268146574497223,
+      "learning_rate": 5e-06,
+      "loss": 0.1445,
+      "step": 10030
+    },
+    {
+      "epoch": 7.550621001129093,
+      "grad_norm": 0.17307054996490479,
+      "learning_rate": 5e-06,
+      "loss": 0.1381,
+      "step": 10031
+    },
+    {
+      "epoch": 7.551373729770418,
+      "grad_norm": 0.1604866236448288,
+      "learning_rate": 5e-06,
+      "loss": 0.1219,
+      "step": 10032
+    },
+    {
+      "epoch": 7.552126458411743,
+      "grad_norm": 0.16631588339805603,
+      "learning_rate": 5e-06,
+      "loss": 0.1823,
+      "step": 10033
+    },
+    {
+      "epoch": 7.552879187053067,
+      "grad_norm": 0.16278037428855896,
+      "learning_rate": 5e-06,
+      "loss": 0.2302,
+      "step": 10034
+    },
+    {
+      "epoch": 7.5536319156943925,
+      "grad_norm": 0.18914462625980377,
+      "learning_rate": 5e-06,
+      "loss": 0.1725,
+      "step": 10035
+    },
+    {
+      "epoch": 7.554384644335717,
+      "grad_norm": 0.21550263464450836,
+      "learning_rate": 5e-06,
+      "loss": 0.1892,
+      "step": 10036
+    },
+    {
+      "epoch": 7.5551373729770415,
+      "grad_norm": 0.16075345873832703,
+      "learning_rate": 5e-06,
+      "loss": 0.1038,
+      "step": 10037
+    },
+    {
+      "epoch": 7.555890101618367,
+      "grad_norm": 0.1765889823436737,
+      "learning_rate": 5e-06,
+      "loss": 0.1821,
+      "step": 10038
+    },
+    {
+      "epoch": 7.556642830259691,
+      "grad_norm": 0.17599597573280334,
+      "learning_rate": 5e-06,
+      "loss": 0.1727,
+      "step": 10039
+    },
+    {
+      "epoch": 7.557395558901016,
+      "grad_norm": 0.16151274740695953,
+      "learning_rate": 5e-06,
+      "loss": 0.1018,
+      "step": 10040
+    },
+    {
+      "epoch": 7.558148287542341,
+      "grad_norm": 0.17284367978572845,
+      "learning_rate": 5e-06,
+      "loss": 0.1664,
+      "step": 10041
+    },
+    {
+      "epoch": 7.558901016183666,
+      "grad_norm": 0.1703868955373764,
+      "learning_rate": 5e-06,
+      "loss": 0.1361,
+      "step": 10042
+    },
+    {
+      "epoch": 7.559653744824991,
+      "grad_norm": 0.1639912724494934,
+      "learning_rate": 5e-06,
+      "loss": 0.1474,
+      "step": 10043
+    },
+    {
+      "epoch": 7.5604064734663154,
+      "grad_norm": 0.1683533638715744,
+      "learning_rate": 5e-06,
+      "loss": 0.1528,
+      "step": 10044
+    },
+    {
+      "epoch": 7.56115920210764,
+      "grad_norm": 0.1682954728603363,
+      "learning_rate": 5e-06,
+      "loss": 0.1157,
+      "step": 10045
+    },
+    {
+      "epoch": 7.561911930748965,
+      "grad_norm": 0.1755424588918686,
+      "learning_rate": 5e-06,
+      "loss": 0.1236,
+      "step": 10046
+    },
+    {
+      "epoch": 7.56266465939029,
+      "grad_norm": 0.1690262109041214,
+      "learning_rate": 5e-06,
+      "loss": 0.0636,
+      "step": 10047
+    },
+    {
+      "epoch": 7.563417388031614,
+      "grad_norm": 0.18089617788791656,
+      "learning_rate": 5e-06,
+      "loss": 0.2583,
+      "step": 10048
+    },
+    {
+      "epoch": 7.56417011667294,
+      "grad_norm": 0.18302488327026367,
+      "learning_rate": 5e-06,
+      "loss": 0.1585,
+      "step": 10049
+    },
+    {
+      "epoch": 7.564922845314264,
+      "grad_norm": 0.1947031319141388,
+      "learning_rate": 5e-06,
+      "loss": 0.1483,
+      "step": 10050
+    },
+    {
+      "epoch": 7.565675573955589,
+      "grad_norm": 0.16583481431007385,
+      "learning_rate": 5e-06,
+      "loss": 0.1443,
+      "step": 10051
+    },
+    {
+      "epoch": 7.566428302596914,
+      "grad_norm": 0.15567465126514435,
+      "learning_rate": 5e-06,
+      "loss": 0.1173,
+      "step": 10052
+    },
+    {
+      "epoch": 7.567181031238238,
+      "grad_norm": 0.17674362659454346,
+      "learning_rate": 5e-06,
+      "loss": 0.171,
+      "step": 10053
+    },
+    {
+      "epoch": 7.567933759879564,
+      "grad_norm": 0.162286639213562,
+      "learning_rate": 5e-06,
+      "loss": 0.1362,
+      "step": 10054
+    },
+    {
+      "epoch": 7.568686488520888,
+      "grad_norm": 0.16481666266918182,
+      "learning_rate": 5e-06,
+      "loss": 0.1106,
+      "step": 10055
+    },
+    {
+      "epoch": 7.569439217162213,
+      "grad_norm": 0.18541088700294495,
+      "learning_rate": 5e-06,
+      "loss": 0.2316,
+      "step": 10056
+    },
+    {
+      "epoch": 7.570191945803538,
+      "grad_norm": 0.15407097339630127,
+      "learning_rate": 5e-06,
+      "loss": 0.0622,
+      "step": 10057
+    },
+    {
+      "epoch": 7.5709446744448625,
+      "grad_norm": 0.1627190262079239,
+      "learning_rate": 5e-06,
+      "loss": 0.137,
+      "step": 10058
+    },
+    {
+      "epoch": 7.571697403086187,
+      "grad_norm": 0.2000432014465332,
+      "learning_rate": 5e-06,
+      "loss": 0.1773,
+      "step": 10059
+    },
+    {
+      "epoch": 7.572450131727512,
+      "grad_norm": 0.17585667967796326,
+      "learning_rate": 5e-06,
+      "loss": 0.1503,
+      "step": 10060
+    },
+    {
+      "epoch": 7.573202860368837,
+      "grad_norm": 0.1917986422777176,
+      "learning_rate": 5e-06,
+      "loss": 0.2139,
+      "step": 10061
+    },
+    {
+      "epoch": 7.573955589010162,
+      "grad_norm": 0.1710434854030609,
+      "learning_rate": 5e-06,
+      "loss": 0.1454,
+      "step": 10062
+    },
+    {
+      "epoch": 7.574708317651487,
+      "grad_norm": 0.16758829355239868,
+      "learning_rate": 5e-06,
+      "loss": 0.1184,
+      "step": 10063
+    },
+    {
+      "epoch": 7.575461046292811,
+      "grad_norm": 0.16134800016880035,
+      "learning_rate": 5e-06,
+      "loss": 0.14,
+      "step": 10064
+    },
+    {
+      "epoch": 7.5762137749341365,
+      "grad_norm": 0.1823464334011078,
+      "learning_rate": 5e-06,
+      "loss": 0.0838,
+      "step": 10065
+    },
+    {
+      "epoch": 7.576966503575461,
+      "grad_norm": 0.14474011957645416,
+      "learning_rate": 5e-06,
+      "loss": 0.0774,
+      "step": 10066
+    },
+    {
+      "epoch": 7.577719232216786,
+      "grad_norm": 0.17133736610412598,
+      "learning_rate": 5e-06,
+      "loss": 0.1243,
+      "step": 10067
+    },
+    {
+      "epoch": 7.578471960858111,
+      "grad_norm": 0.1633196473121643,
+      "learning_rate": 5e-06,
+      "loss": 0.1459,
+      "step": 10068
+    },
+    {
+      "epoch": 7.579224689499435,
+      "grad_norm": 0.1703857183456421,
+      "learning_rate": 5e-06,
+      "loss": 0.1681,
+      "step": 10069
+    },
+    {
+      "epoch": 7.579977418140761,
+      "grad_norm": 0.1811559945344925,
+      "learning_rate": 5e-06,
+      "loss": 0.1491,
+      "step": 10070
+    },
+    {
+      "epoch": 7.580730146782085,
+      "grad_norm": 0.1737089306116104,
+      "learning_rate": 5e-06,
+      "loss": 0.1554,
+      "step": 10071
+    },
+    {
+      "epoch": 7.58148287542341,
+      "grad_norm": 0.14763757586479187,
+      "learning_rate": 5e-06,
+      "loss": 0.1108,
+      "step": 10072
+    },
+    {
+      "epoch": 7.582235604064735,
+      "grad_norm": 0.17833201587200165,
+      "learning_rate": 5e-06,
+      "loss": 0.1503,
+      "step": 10073
+    },
+    {
+      "epoch": 7.582988332706059,
+      "grad_norm": 0.1701297163963318,
+      "learning_rate": 5e-06,
+      "loss": 0.1051,
+      "step": 10074
+    },
+    {
+      "epoch": 7.583741061347384,
+      "grad_norm": 0.17935429513454437,
+      "learning_rate": 5e-06,
+      "loss": 0.133,
+      "step": 10075
+    },
+    {
+      "epoch": 7.584493789988709,
+      "grad_norm": 0.190237894654274,
+      "learning_rate": 5e-06,
+      "loss": 0.1754,
+      "step": 10076
+    },
+    {
+      "epoch": 7.585246518630034,
+      "grad_norm": 0.16840867698192596,
+      "learning_rate": 5e-06,
+      "loss": 0.1795,
+      "step": 10077
+    },
+    {
+      "epoch": 7.585999247271358,
+      "grad_norm": 0.15180407464504242,
+      "learning_rate": 5e-06,
+      "loss": 0.1548,
+      "step": 10078
+    },
+    {
+      "epoch": 7.586751975912684,
+      "grad_norm": 0.18263943493366241,
+      "learning_rate": 5e-06,
+      "loss": 0.1769,
+      "step": 10079
+    },
+    {
+      "epoch": 7.587504704554008,
+      "grad_norm": 0.19772593677043915,
+      "learning_rate": 5e-06,
+      "loss": 0.1413,
+      "step": 10080
+    },
+    {
+      "epoch": 7.588257433195333,
+      "grad_norm": 0.17342665791511536,
+      "learning_rate": 5e-06,
+      "loss": 0.1284,
+      "step": 10081
+    },
+    {
+      "epoch": 7.589010161836658,
+      "grad_norm": 0.18364092707633972,
+      "learning_rate": 5e-06,
+      "loss": 0.1172,
+      "step": 10082
+    },
+    {
+      "epoch": 7.589762890477982,
+      "grad_norm": 0.16361868381500244,
+      "learning_rate": 5e-06,
+      "loss": 0.2103,
+      "step": 10083
+    },
+    {
+      "epoch": 7.590515619119308,
+      "grad_norm": 0.18309196829795837,
+      "learning_rate": 5e-06,
+      "loss": 0.1789,
+      "step": 10084
+    },
+    {
+      "epoch": 7.591268347760632,
+      "grad_norm": 0.17171745002269745,
+      "learning_rate": 5e-06,
+      "loss": 0.1646,
+      "step": 10085
+    },
+    {
+      "epoch": 7.592021076401958,
+      "grad_norm": 0.18470463156700134,
+      "learning_rate": 5e-06,
+      "loss": 0.1109,
+      "step": 10086
+    },
+    {
+      "epoch": 7.592773805043282,
+      "grad_norm": 0.16599254310131073,
+      "learning_rate": 5e-06,
+      "loss": 0.166,
+      "step": 10087
+    },
+    {
+      "epoch": 7.5935265336846065,
+      "grad_norm": 0.18659912049770355,
+      "learning_rate": 5e-06,
+      "loss": 0.197,
+      "step": 10088
+    },
+    {
+      "epoch": 7.594279262325932,
+      "grad_norm": 0.18166089057922363,
+      "learning_rate": 5e-06,
+      "loss": 0.1585,
+      "step": 10089
+    },
+    {
+      "epoch": 7.595031990967256,
+      "grad_norm": 0.17359480261802673,
+      "learning_rate": 5e-06,
+      "loss": 0.2945,
+      "step": 10090
+    },
+    {
+      "epoch": 7.595784719608581,
+      "grad_norm": 0.1932349056005478,
+      "learning_rate": 5e-06,
+      "loss": 0.1993,
+      "step": 10091
+    },
+    {
+      "epoch": 7.596537448249906,
+      "grad_norm": 0.17216959595680237,
+      "learning_rate": 5e-06,
+      "loss": 0.086,
+      "step": 10092
+    },
+    {
+      "epoch": 7.597290176891231,
+      "grad_norm": 0.1791159063577652,
+      "learning_rate": 5e-06,
+      "loss": 0.2306,
+      "step": 10093
+    },
+    {
+      "epoch": 7.598042905532555,
+      "grad_norm": 0.170023575425148,
+      "learning_rate": 5e-06,
+      "loss": 0.2076,
+      "step": 10094
+    },
+    {
+      "epoch": 7.5987956341738805,
+      "grad_norm": 0.17678019404411316,
+      "learning_rate": 5e-06,
+      "loss": 0.1529,
+      "step": 10095
+    },
+    {
+      "epoch": 7.599548362815205,
+      "grad_norm": 0.16233713924884796,
+      "learning_rate": 5e-06,
+      "loss": 0.161,
+      "step": 10096
+    },
+    {
+      "epoch": 7.600301091456529,
+      "grad_norm": 0.15497015416622162,
+      "learning_rate": 5e-06,
+      "loss": 0.1947,
+      "step": 10097
+    },
+    {
+      "epoch": 7.601053820097855,
+      "grad_norm": 0.16411812603473663,
+      "learning_rate": 5e-06,
+      "loss": 0.2022,
+      "step": 10098
+    },
+    {
+      "epoch": 7.601806548739179,
+      "grad_norm": 0.166364386677742,
+      "learning_rate": 5e-06,
+      "loss": 0.1626,
+      "step": 10099
+    },
+    {
+      "epoch": 7.602559277380505,
+      "grad_norm": 0.17460018396377563,
+      "learning_rate": 5e-06,
+      "loss": 0.1549,
+      "step": 10100
+    },
+    {
+      "epoch": 7.603312006021829,
+      "grad_norm": 0.1734592318534851,
+      "learning_rate": 5e-06,
+      "loss": 0.1854,
+      "step": 10101
+    },
+    {
+      "epoch": 7.604064734663154,
+      "grad_norm": 0.17373596131801605,
+      "learning_rate": 5e-06,
+      "loss": 0.1615,
+      "step": 10102
+    },
+    {
+      "epoch": 7.604817463304479,
+      "grad_norm": 0.16388700902462006,
+      "learning_rate": 5e-06,
+      "loss": 0.101,
+      "step": 10103
+    },
+    {
+      "epoch": 7.605570191945803,
+      "grad_norm": 0.17638930678367615,
+      "learning_rate": 5e-06,
+      "loss": 0.1311,
+      "step": 10104
+    },
+    {
+      "epoch": 7.606322920587129,
+      "grad_norm": 0.16639016568660736,
+      "learning_rate": 5e-06,
+      "loss": 0.189,
+      "step": 10105
+    },
+    {
+      "epoch": 7.607075649228453,
+      "grad_norm": 0.1598782241344452,
+      "learning_rate": 5e-06,
+      "loss": 0.1233,
+      "step": 10106
+    },
+    {
+      "epoch": 7.607828377869778,
+      "grad_norm": 0.18665041029453278,
+      "learning_rate": 5e-06,
+      "loss": 0.1707,
+      "step": 10107
+    },
+    {
+      "epoch": 7.608581106511103,
+      "grad_norm": 0.16197679936885834,
+      "learning_rate": 5e-06,
+      "loss": 0.1617,
+      "step": 10108
+    },
+    {
+      "epoch": 7.609333835152428,
+      "grad_norm": 0.17873843014240265,
+      "learning_rate": 5e-06,
+      "loss": 0.1674,
+      "step": 10109
+    },
+    {
+      "epoch": 7.610086563793752,
+      "grad_norm": 0.16525158286094666,
+      "learning_rate": 5e-06,
+      "loss": 0.1832,
+      "step": 10110
+    },
+    {
+      "epoch": 7.610839292435077,
+      "grad_norm": 0.18297290802001953,
+      "learning_rate": 5e-06,
+      "loss": 0.1806,
+      "step": 10111
+    },
+    {
+      "epoch": 7.611592021076402,
+      "grad_norm": 0.16088932752609253,
+      "learning_rate": 5e-06,
+      "loss": 0.1681,
+      "step": 10112
+    },
+    {
+      "epoch": 7.612344749717726,
+      "grad_norm": 0.16538047790527344,
+      "learning_rate": 5e-06,
+      "loss": 0.1337,
+      "step": 10113
+    },
+    {
+      "epoch": 7.613097478359052,
+      "grad_norm": 0.17482824623584747,
+      "learning_rate": 5e-06,
+      "loss": 0.1475,
+      "step": 10114
+    },
+    {
+      "epoch": 7.613850207000376,
+      "grad_norm": 0.16961850225925446,
+      "learning_rate": 5e-06,
+      "loss": 0.1645,
+      "step": 10115
+    },
+    {
+      "epoch": 7.6146029356417015,
+      "grad_norm": 0.1669793426990509,
+      "learning_rate": 5e-06,
+      "loss": 0.1666,
+      "step": 10116
+    },
+    {
+      "epoch": 7.615355664283026,
+      "grad_norm": 0.18019744753837585,
+      "learning_rate": 5e-06,
+      "loss": 0.1676,
+      "step": 10117
+    },
+    {
+      "epoch": 7.6161083929243505,
+      "grad_norm": 0.16901430487632751,
+      "learning_rate": 5e-06,
+      "loss": 0.207,
+      "step": 10118
+    },
+    {
+      "epoch": 7.616861121565676,
+      "grad_norm": 0.18023233115673065,
+      "learning_rate": 5e-06,
+      "loss": 0.2106,
+      "step": 10119
+    },
+    {
+      "epoch": 7.617613850207,
+      "grad_norm": 0.1932206153869629,
+      "learning_rate": 5e-06,
+      "loss": 0.2126,
+      "step": 10120
+    },
+    {
+      "epoch": 7.618366578848325,
+      "grad_norm": 0.1858149766921997,
+      "learning_rate": 5e-06,
+      "loss": 0.1844,
+      "step": 10121
+    },
+    {
+      "epoch": 7.61911930748965,
+      "grad_norm": 0.15842871367931366,
+      "learning_rate": 5e-06,
+      "loss": 0.1549,
+      "step": 10122
+    },
+    {
+      "epoch": 7.619872036130975,
+      "grad_norm": 0.17585396766662598,
+      "learning_rate": 5e-06,
+      "loss": 0.1134,
+      "step": 10123
+    },
+    {
+      "epoch": 7.6206247647723,
+      "grad_norm": 0.16027332842350006,
+      "learning_rate": 5e-06,
+      "loss": 0.1076,
+      "step": 10124
+    },
+    {
+      "epoch": 7.6213774934136245,
+      "grad_norm": 0.18879950046539307,
+      "learning_rate": 5e-06,
+      "loss": 0.1981,
+      "step": 10125
+    },
+    {
+      "epoch": 7.622130222054949,
+      "grad_norm": 0.15220041573047638,
+      "learning_rate": 5e-06,
+      "loss": 0.1377,
+      "step": 10126
+    },
+    {
+      "epoch": 7.622882950696274,
+      "grad_norm": 0.17716729640960693,
+      "learning_rate": 5e-06,
+      "loss": 0.2175,
+      "step": 10127
+    },
+    {
+      "epoch": 7.623635679337599,
+      "grad_norm": 0.1777564287185669,
+      "learning_rate": 5e-06,
+      "loss": 0.1256,
+      "step": 10128
+    },
+    {
+      "epoch": 7.624388407978923,
+      "grad_norm": 0.16093100607395172,
+      "learning_rate": 5e-06,
+      "loss": 0.1174,
+      "step": 10129
+    },
+    {
+      "epoch": 7.625141136620249,
+      "grad_norm": 0.1876409351825714,
+      "learning_rate": 5e-06,
+      "loss": 0.1396,
+      "step": 10130
+    },
+    {
+      "epoch": 7.625893865261573,
+      "grad_norm": 0.16062882542610168,
+      "learning_rate": 5e-06,
+      "loss": 0.1303,
+      "step": 10131
+    },
+    {
+      "epoch": 7.626646593902898,
+      "grad_norm": 0.16671651601791382,
+      "learning_rate": 5e-06,
+      "loss": 0.1,
+      "step": 10132
+    },
+    {
+      "epoch": 7.627399322544223,
+      "grad_norm": 0.17106091976165771,
+      "learning_rate": 5e-06,
+      "loss": 0.1704,
+      "step": 10133
+    },
+    {
+      "epoch": 7.628152051185547,
+      "grad_norm": 0.16296666860580444,
+      "learning_rate": 5e-06,
+      "loss": 0.122,
+      "step": 10134
+    },
+    {
+      "epoch": 7.628904779826873,
+      "grad_norm": 0.1670026183128357,
+      "learning_rate": 5e-06,
+      "loss": 0.1711,
+      "step": 10135
+    },
+    {
+      "epoch": 7.629657508468197,
+      "grad_norm": 0.16115419566631317,
+      "learning_rate": 5e-06,
+      "loss": 0.1461,
+      "step": 10136
+    },
+    {
+      "epoch": 7.630410237109522,
+      "grad_norm": 0.18349644541740417,
+      "learning_rate": 5e-06,
+      "loss": 0.1568,
+      "step": 10137
+    },
+    {
+      "epoch": 7.631162965750847,
+      "grad_norm": 0.17542198300361633,
+      "learning_rate": 5e-06,
+      "loss": 0.1947,
+      "step": 10138
+    },
+    {
+      "epoch": 7.6319156943921715,
+      "grad_norm": 0.17402659356594086,
+      "learning_rate": 5e-06,
+      "loss": 0.1741,
+      "step": 10139
+    },
+    {
+      "epoch": 7.632668423033497,
+      "grad_norm": 0.15355907380580902,
+      "learning_rate": 5e-06,
+      "loss": 0.113,
+      "step": 10140
+    },
+    {
+      "epoch": 7.633421151674821,
+      "grad_norm": 0.1830025315284729,
+      "learning_rate": 5e-06,
+      "loss": 0.1468,
+      "step": 10141
+    },
+    {
+      "epoch": 7.634173880316146,
+      "grad_norm": 0.16280651092529297,
+      "learning_rate": 5e-06,
+      "loss": 0.1376,
+      "step": 10142
+    },
+    {
+      "epoch": 7.634926608957471,
+      "grad_norm": 0.16642452776432037,
+      "learning_rate": 5e-06,
+      "loss": 0.1698,
+      "step": 10143
+    },
+    {
+      "epoch": 7.635679337598796,
+      "grad_norm": 0.15370798110961914,
+      "learning_rate": 5e-06,
+      "loss": 0.1169,
+      "step": 10144
+    },
+    {
+      "epoch": 7.63643206624012,
+      "grad_norm": 0.15707337856292725,
+      "learning_rate": 5e-06,
+      "loss": 0.2051,
+      "step": 10145
+    },
+    {
+      "epoch": 7.6371847948814455,
+      "grad_norm": 0.15669076144695282,
+      "learning_rate": 5e-06,
+      "loss": 0.1399,
+      "step": 10146
+    },
+    {
+      "epoch": 7.63793752352277,
+      "grad_norm": 0.14862458407878876,
+      "learning_rate": 5e-06,
+      "loss": 0.1437,
+      "step": 10147
+    },
+    {
+      "epoch": 7.6386902521640945,
+      "grad_norm": 0.18630380928516388,
+      "learning_rate": 5e-06,
+      "loss": 0.1436,
+      "step": 10148
+    },
+    {
+      "epoch": 7.63944298080542,
+      "grad_norm": 0.1640799641609192,
+      "learning_rate": 5e-06,
+      "loss": 0.1113,
+      "step": 10149
+    },
+    {
+      "epoch": 7.640195709446744,
+      "grad_norm": 0.19054904580116272,
+      "learning_rate": 5e-06,
+      "loss": 0.1551,
+      "step": 10150
+    },
+    {
+      "epoch": 7.640948438088069,
+      "grad_norm": 0.17289690673351288,
+      "learning_rate": 5e-06,
+      "loss": 0.171,
+      "step": 10151
+    },
+    {
+      "epoch": 7.641701166729394,
+      "grad_norm": 0.17465031147003174,
+      "learning_rate": 5e-06,
+      "loss": 0.1548,
+      "step": 10152
+    },
+    {
+      "epoch": 7.642453895370719,
+      "grad_norm": 0.18534915149211884,
+      "learning_rate": 5e-06,
+      "loss": 0.2286,
+      "step": 10153
+    },
+    {
+      "epoch": 7.643206624012044,
+      "grad_norm": 0.1821305751800537,
+      "learning_rate": 5e-06,
+      "loss": 0.2095,
+      "step": 10154
+    },
+    {
+      "epoch": 7.6439593526533685,
+      "grad_norm": 0.16856952011585236,
+      "learning_rate": 5e-06,
+      "loss": 0.1535,
+      "step": 10155
+    },
+    {
+      "epoch": 7.644712081294693,
+      "grad_norm": 0.18349088728427887,
+      "learning_rate": 5e-06,
+      "loss": 0.1447,
+      "step": 10156
+    },
+    {
+      "epoch": 7.645464809936018,
+      "grad_norm": 0.17174296081066132,
+      "learning_rate": 5e-06,
+      "loss": 0.1493,
+      "step": 10157
+    },
+    {
+      "epoch": 7.646217538577343,
+      "grad_norm": 0.1743174046278,
+      "learning_rate": 5e-06,
+      "loss": 0.1546,
+      "step": 10158
+    },
+    {
+      "epoch": 7.646970267218668,
+      "grad_norm": 0.15557415783405304,
+      "learning_rate": 5e-06,
+      "loss": 0.2244,
+      "step": 10159
+    },
+    {
+      "epoch": 7.647722995859993,
+      "grad_norm": 0.15729065239429474,
+      "learning_rate": 5e-06,
+      "loss": 0.2047,
+      "step": 10160
+    },
+    {
+      "epoch": 7.648475724501317,
+      "grad_norm": 0.1687522530555725,
+      "learning_rate": 5e-06,
+      "loss": 0.14,
+      "step": 10161
+    },
+    {
+      "epoch": 7.649228453142642,
+      "grad_norm": 0.17867667973041534,
+      "learning_rate": 5e-06,
+      "loss": 0.1697,
+      "step": 10162
+    },
+    {
+      "epoch": 7.649981181783967,
+      "grad_norm": 0.16044336557388306,
+      "learning_rate": 5e-06,
+      "loss": 0.0897,
+      "step": 10163
+    },
+    {
+      "epoch": 7.650733910425291,
+      "grad_norm": 0.18732306361198425,
+      "learning_rate": 5e-06,
+      "loss": 0.1251,
+      "step": 10164
+    },
+    {
+      "epoch": 7.651486639066617,
+      "grad_norm": 0.16144660115242004,
+      "learning_rate": 5e-06,
+      "loss": 0.1467,
+      "step": 10165
+    },
+    {
+      "epoch": 7.652239367707941,
+      "grad_norm": 0.17832209169864655,
+      "learning_rate": 5e-06,
+      "loss": 0.2011,
+      "step": 10166
+    },
+    {
+      "epoch": 7.652992096349266,
+      "grad_norm": 0.16890659928321838,
+      "learning_rate": 5e-06,
+      "loss": 0.1638,
+      "step": 10167
+    },
+    {
+      "epoch": 7.653744824990591,
+      "grad_norm": 0.17658677697181702,
+      "learning_rate": 5e-06,
+      "loss": 0.1782,
+      "step": 10168
+    },
+    {
+      "epoch": 7.6544975536319155,
+      "grad_norm": 0.1625896692276001,
+      "learning_rate": 5e-06,
+      "loss": 0.1214,
+      "step": 10169
+    },
+    {
+      "epoch": 7.65525028227324,
+      "grad_norm": 0.18655715882778168,
+      "learning_rate": 5e-06,
+      "loss": 0.1764,
+      "step": 10170
+    },
+    {
+      "epoch": 7.656003010914565,
+      "grad_norm": 0.16755631566047668,
+      "learning_rate": 5e-06,
+      "loss": 0.1707,
+      "step": 10171
+    },
+    {
+      "epoch": 7.65675573955589,
+      "grad_norm": 0.1585579663515091,
+      "learning_rate": 5e-06,
+      "loss": 0.0896,
+      "step": 10172
+    },
+    {
+      "epoch": 7.657508468197215,
+      "grad_norm": 0.1703519970178604,
+      "learning_rate": 5e-06,
+      "loss": 0.1617,
+      "step": 10173
+    },
+    {
+      "epoch": 7.65826119683854,
+      "grad_norm": 0.1415853202342987,
+      "learning_rate": 5e-06,
+      "loss": 0.0542,
+      "step": 10174
+    },
+    {
+      "epoch": 7.659013925479864,
+      "grad_norm": 0.17952612042427063,
+      "learning_rate": 5e-06,
+      "loss": 0.1727,
+      "step": 10175
+    },
+    {
+      "epoch": 7.6597666541211895,
+      "grad_norm": 0.1806783527135849,
+      "learning_rate": 5e-06,
+      "loss": 0.1427,
+      "step": 10176
+    },
+    {
+      "epoch": 7.660519382762514,
+      "grad_norm": 0.17361170053482056,
+      "learning_rate": 5e-06,
+      "loss": 0.1611,
+      "step": 10177
+    },
+    {
+      "epoch": 7.661272111403839,
+      "grad_norm": 0.19658350944519043,
+      "learning_rate": 5e-06,
+      "loss": 0.2246,
+      "step": 10178
+    },
+    {
+      "epoch": 7.662024840045164,
+      "grad_norm": 0.1753070205450058,
+      "learning_rate": 5e-06,
+      "loss": 0.102,
+      "step": 10179
+    },
+    {
+      "epoch": 7.662777568686488,
+      "grad_norm": 0.19013842940330505,
+      "learning_rate": 5e-06,
+      "loss": 0.173,
+      "step": 10180
+    },
+    {
+      "epoch": 7.663530297327814,
+      "grad_norm": 0.21315689384937286,
+      "learning_rate": 5e-06,
+      "loss": 0.1258,
+      "step": 10181
+    },
+    {
+      "epoch": 7.664283025969138,
+      "grad_norm": 0.19774894416332245,
+      "learning_rate": 5e-06,
+      "loss": 0.1603,
+      "step": 10182
+    },
+    {
+      "epoch": 7.665035754610463,
+      "grad_norm": 0.19022132456302643,
+      "learning_rate": 5e-06,
+      "loss": 0.1505,
+      "step": 10183
+    },
+    {
+      "epoch": 7.665788483251788,
+      "grad_norm": 0.16330872476100922,
+      "learning_rate": 5e-06,
+      "loss": 0.1605,
+      "step": 10184
+    },
+    {
+      "epoch": 7.666541211893112,
+      "grad_norm": 0.18975785374641418,
+      "learning_rate": 5e-06,
+      "loss": 0.1762,
+      "step": 10185
+    },
+    {
+      "epoch": 7.667293940534437,
+      "grad_norm": 0.2076784074306488,
+      "learning_rate": 5e-06,
+      "loss": 0.1913,
+      "step": 10186
+    },
+    {
+      "epoch": 7.668046669175762,
+      "grad_norm": 0.15522760152816772,
+      "learning_rate": 5e-06,
+      "loss": 0.1588,
+      "step": 10187
+    },
+    {
+      "epoch": 7.668799397817087,
+      "grad_norm": 0.15816499292850494,
+      "learning_rate": 5e-06,
+      "loss": 0.1922,
+      "step": 10188
+    },
+    {
+      "epoch": 7.669552126458412,
+      "grad_norm": 0.17470891773700714,
+      "learning_rate": 5e-06,
+      "loss": 0.2591,
+      "step": 10189
+    },
+    {
+      "epoch": 7.670304855099737,
+      "grad_norm": 0.18902310729026794,
+      "learning_rate": 5e-06,
+      "loss": 0.0871,
+      "step": 10190
+    },
+    {
+      "epoch": 7.671057583741061,
+      "grad_norm": 0.1784159243106842,
+      "learning_rate": 5e-06,
+      "loss": 0.1527,
+      "step": 10191
+    },
+    {
+      "epoch": 7.671810312382386,
+      "grad_norm": 0.1858360320329666,
+      "learning_rate": 5e-06,
+      "loss": 0.1678,
+      "step": 10192
+    },
+    {
+      "epoch": 7.672563041023711,
+      "grad_norm": 0.18418079614639282,
+      "learning_rate": 5e-06,
+      "loss": 0.1879,
+      "step": 10193
+    },
+    {
+      "epoch": 7.673315769665035,
+      "grad_norm": 0.18741141259670258,
+      "learning_rate": 5e-06,
+      "loss": 0.2002,
+      "step": 10194
+    },
+    {
+      "epoch": 7.674068498306361,
+      "grad_norm": 0.20179195702075958,
+      "learning_rate": 5e-06,
+      "loss": 0.2496,
+      "step": 10195
+    },
+    {
+      "epoch": 7.674821226947685,
+      "grad_norm": 0.15322118997573853,
+      "learning_rate": 5e-06,
+      "loss": 0.1275,
+      "step": 10196
+    },
+    {
+      "epoch": 7.675573955589011,
+      "grad_norm": 0.1845404952764511,
+      "learning_rate": 5e-06,
+      "loss": 0.1226,
+      "step": 10197
+    },
+    {
+      "epoch": 7.676326684230335,
+      "grad_norm": 0.21822261810302734,
+      "learning_rate": 5e-06,
+      "loss": 0.2117,
+      "step": 10198
+    },
+    {
+      "epoch": 7.6770794128716595,
+      "grad_norm": 0.18466490507125854,
+      "learning_rate": 5e-06,
+      "loss": 0.1797,
+      "step": 10199
+    },
+    {
+      "epoch": 7.677832141512985,
+      "grad_norm": 0.18418292701244354,
+      "learning_rate": 5e-06,
+      "loss": 0.18,
+      "step": 10200
+    },
+    {
+      "epoch": 7.677832141512985,
+      "eval_loss": 0.19837842881679535,
+      "eval_runtime": 452.835,
+      "eval_samples_per_second": 21.259,
+      "eval_steps_per_second": 0.665,
+      "step": 10200
+    },
+    {
+      "epoch": 7.678584870154309,
+      "grad_norm": 0.14645805954933167,
+      "learning_rate": 5e-06,
+      "loss": 0.0977,
+      "step": 10201
+    },
+    {
+      "epoch": 7.679337598795634,
+      "grad_norm": 0.16831925511360168,
+      "learning_rate": 5e-06,
+      "loss": 0.0903,
+      "step": 10202
+    },
+    {
+      "epoch": 7.680090327436959,
+      "grad_norm": 0.16514895856380463,
+      "learning_rate": 5e-06,
+      "loss": 0.1304,
+      "step": 10203
+    },
+    {
+      "epoch": 7.680843056078284,
+      "grad_norm": 0.20418758690357208,
+      "learning_rate": 5e-06,
+      "loss": 0.1907,
+      "step": 10204
+    },
+    {
+      "epoch": 7.681595784719608,
+      "grad_norm": 0.18832139670848846,
+      "learning_rate": 5e-06,
+      "loss": 0.1526,
+      "step": 10205
+    },
+    {
+      "epoch": 7.6823485133609335,
+      "grad_norm": 0.18588022887706757,
+      "learning_rate": 5e-06,
+      "loss": 0.1263,
+      "step": 10206
+    },
+    {
+      "epoch": 7.683101242002258,
+      "grad_norm": 0.17003194987773895,
+      "learning_rate": 5e-06,
+      "loss": 0.199,
+      "step": 10207
+    },
+    {
+      "epoch": 7.683853970643583,
+      "grad_norm": 0.16737908124923706,
+      "learning_rate": 5e-06,
+      "loss": 0.173,
+      "step": 10208
+    },
+    {
+      "epoch": 7.684606699284908,
+      "grad_norm": 0.1661984920501709,
+      "learning_rate": 5e-06,
+      "loss": 0.1014,
+      "step": 10209
+    },
+    {
+      "epoch": 7.685359427926232,
+      "grad_norm": 0.18010088801383972,
+      "learning_rate": 5e-06,
+      "loss": 0.1095,
+      "step": 10210
+    },
+    {
+      "epoch": 7.686112156567558,
+      "grad_norm": 0.17269495129585266,
+      "learning_rate": 5e-06,
+      "loss": 0.141,
+      "step": 10211
+    },
+    {
+      "epoch": 7.686864885208882,
+      "grad_norm": 0.17514966428279877,
+      "learning_rate": 5e-06,
+      "loss": 0.2038,
+      "step": 10212
+    },
+    {
+      "epoch": 7.6876176138502075,
+      "grad_norm": 0.1670428067445755,
+      "learning_rate": 5e-06,
+      "loss": 0.084,
+      "step": 10213
+    },
+    {
+      "epoch": 7.688370342491532,
+      "grad_norm": 0.17655318975448608,
+      "learning_rate": 5e-06,
+      "loss": 0.1317,
+      "step": 10214
+    },
+    {
+      "epoch": 7.689123071132856,
+      "grad_norm": 0.17500950396060944,
+      "learning_rate": 5e-06,
+      "loss": 0.2293,
+      "step": 10215
+    },
+    {
+      "epoch": 7.689875799774182,
+      "grad_norm": 0.1754717081785202,
+      "learning_rate": 5e-06,
+      "loss": 0.1202,
+      "step": 10216
+    },
+    {
+      "epoch": 7.690628528415506,
+      "grad_norm": 0.14711077511310577,
+      "learning_rate": 5e-06,
+      "loss": 0.1171,
+      "step": 10217
+    },
+    {
+      "epoch": 7.691381257056831,
+      "grad_norm": 0.159993976354599,
+      "learning_rate": 5e-06,
+      "loss": 0.1043,
+      "step": 10218
+    },
+    {
+      "epoch": 7.692133985698156,
+      "grad_norm": 0.18543241918087006,
+      "learning_rate": 5e-06,
+      "loss": 0.119,
+      "step": 10219
+    },
+    {
+      "epoch": 7.692886714339481,
+      "grad_norm": 0.18161047995090485,
+      "learning_rate": 5e-06,
+      "loss": 0.1702,
+      "step": 10220
+    },
+    {
+      "epoch": 7.693639442980805,
+      "grad_norm": 0.1608254611492157,
+      "learning_rate": 5e-06,
+      "loss": 0.2563,
+      "step": 10221
+    },
+    {
+      "epoch": 7.69439217162213,
+      "grad_norm": 0.15180781483650208,
+      "learning_rate": 5e-06,
+      "loss": 0.1191,
+      "step": 10222
+    },
+    {
+      "epoch": 7.695144900263455,
+      "grad_norm": 0.16920755803585052,
+      "learning_rate": 5e-06,
+      "loss": 0.11,
+      "step": 10223
+    },
+    {
+      "epoch": 7.695897628904779,
+      "grad_norm": 0.1656620353460312,
+      "learning_rate": 5e-06,
+      "loss": 0.2058,
+      "step": 10224
+    },
+    {
+      "epoch": 7.696650357546105,
+      "grad_norm": 0.15082162618637085,
+      "learning_rate": 5e-06,
+      "loss": 0.1695,
+      "step": 10225
+    },
+    {
+      "epoch": 7.697403086187429,
+      "grad_norm": 0.15827736258506775,
+      "learning_rate": 5e-06,
+      "loss": 0.1484,
+      "step": 10226
+    },
+    {
+      "epoch": 7.6981558148287546,
+      "grad_norm": 0.15714897215366364,
+      "learning_rate": 5e-06,
+      "loss": 0.1238,
+      "step": 10227
+    },
+    {
+      "epoch": 7.698908543470079,
+      "grad_norm": 0.17947465181350708,
+      "learning_rate": 5e-06,
+      "loss": 0.151,
+      "step": 10228
+    },
+    {
+      "epoch": 7.6996612721114035,
+      "grad_norm": 0.1524287313222885,
+      "learning_rate": 5e-06,
+      "loss": 0.1415,
+      "step": 10229
+    },
+    {
+      "epoch": 7.700414000752729,
+      "grad_norm": 0.18127386271953583,
+      "learning_rate": 5e-06,
+      "loss": 0.1751,
+      "step": 10230
+    },
+    {
+      "epoch": 7.701166729394053,
+      "grad_norm": 0.17653772234916687,
+      "learning_rate": 5e-06,
+      "loss": 0.1479,
+      "step": 10231
+    },
+    {
+      "epoch": 7.701919458035379,
+      "grad_norm": 0.18813283741474152,
+      "learning_rate": 5e-06,
+      "loss": 0.1291,
+      "step": 10232
+    },
+    {
+      "epoch": 7.702672186676703,
+      "grad_norm": 0.16585977375507355,
+      "learning_rate": 5e-06,
+      "loss": 0.1367,
+      "step": 10233
+    },
+    {
+      "epoch": 7.703424915318028,
+      "grad_norm": 0.18522359430789948,
+      "learning_rate": 5e-06,
+      "loss": 0.1603,
+      "step": 10234
+    },
+    {
+      "epoch": 7.704177643959353,
+      "grad_norm": 0.17551812529563904,
+      "learning_rate": 5e-06,
+      "loss": 0.2122,
+      "step": 10235
+    },
+    {
+      "epoch": 7.7049303726006775,
+      "grad_norm": 0.17284555733203888,
+      "learning_rate": 5e-06,
+      "loss": 0.2002,
+      "step": 10236
+    },
+    {
+      "epoch": 7.705683101242002,
+      "grad_norm": 0.18537695705890656,
+      "learning_rate": 5e-06,
+      "loss": 0.1986,
+      "step": 10237
+    },
+    {
+      "epoch": 7.706435829883327,
+      "grad_norm": 0.1609324961900711,
+      "learning_rate": 5e-06,
+      "loss": 0.1504,
+      "step": 10238
+    },
+    {
+      "epoch": 7.707188558524652,
+      "grad_norm": 0.17668621242046356,
+      "learning_rate": 5e-06,
+      "loss": 0.1515,
+      "step": 10239
+    },
+    {
+      "epoch": 7.707941287165976,
+      "grad_norm": 0.16594433784484863,
+      "learning_rate": 5e-06,
+      "loss": 0.1576,
+      "step": 10240
+    },
+    {
+      "epoch": 7.708694015807302,
+      "grad_norm": 0.18601712584495544,
+      "learning_rate": 5e-06,
+      "loss": 0.191,
+      "step": 10241
+    },
+    {
+      "epoch": 7.709446744448626,
+      "grad_norm": 0.16759604215621948,
+      "learning_rate": 5e-06,
+      "loss": 0.1194,
+      "step": 10242
+    },
+    {
+      "epoch": 7.7101994730899515,
+      "grad_norm": 0.18750518560409546,
+      "learning_rate": 5e-06,
+      "loss": 0.1766,
+      "step": 10243
+    },
+    {
+      "epoch": 7.710952201731276,
+      "grad_norm": 0.17961017787456512,
+      "learning_rate": 5e-06,
+      "loss": 0.1995,
+      "step": 10244
+    },
+    {
+      "epoch": 7.7117049303726,
+      "grad_norm": 0.14580363035202026,
+      "learning_rate": 5e-06,
+      "loss": 0.1189,
+      "step": 10245
+    },
+    {
+      "epoch": 7.712457659013926,
+      "grad_norm": 0.1767502874135971,
+      "learning_rate": 5e-06,
+      "loss": 0.1464,
+      "step": 10246
+    },
+    {
+      "epoch": 7.71321038765525,
+      "grad_norm": 0.16863499581813812,
+      "learning_rate": 5e-06,
+      "loss": 0.1573,
+      "step": 10247
+    },
+    {
+      "epoch": 7.713963116296575,
+      "grad_norm": 0.17143991589546204,
+      "learning_rate": 5e-06,
+      "loss": 0.1879,
+      "step": 10248
+    },
+    {
+      "epoch": 7.7147158449379,
+      "grad_norm": 0.172263503074646,
+      "learning_rate": 5e-06,
+      "loss": 0.1798,
+      "step": 10249
+    },
+    {
+      "epoch": 7.7154685735792246,
+      "grad_norm": 0.17227372527122498,
+      "learning_rate": 5e-06,
+      "loss": 0.2019,
+      "step": 10250
+    },
+    {
+      "epoch": 7.71622130222055,
+      "grad_norm": 0.1632464975118637,
+      "learning_rate": 5e-06,
+      "loss": 0.1728,
+      "step": 10251
+    },
+    {
+      "epoch": 7.716974030861874,
+      "grad_norm": 0.17440201342105865,
+      "learning_rate": 5e-06,
+      "loss": 0.2388,
+      "step": 10252
+    },
+    {
+      "epoch": 7.717726759503199,
+      "grad_norm": 0.17707385122776031,
+      "learning_rate": 5e-06,
+      "loss": 0.1765,
+      "step": 10253
+    },
+    {
+      "epoch": 7.718479488144524,
+      "grad_norm": 0.14085997641086578,
+      "learning_rate": 5e-06,
+      "loss": 0.1712,
+      "step": 10254
+    },
+    {
+      "epoch": 7.719232216785849,
+      "grad_norm": 0.16969536244869232,
+      "learning_rate": 5e-06,
+      "loss": 0.1507,
+      "step": 10255
+    },
+    {
+      "epoch": 7.719984945427173,
+      "grad_norm": 0.18230673670768738,
+      "learning_rate": 5e-06,
+      "loss": 0.1868,
+      "step": 10256
+    },
+    {
+      "epoch": 7.7207376740684985,
+      "grad_norm": 0.1610090285539627,
+      "learning_rate": 5e-06,
+      "loss": 0.1036,
+      "step": 10257
+    },
+    {
+      "epoch": 7.721490402709823,
+      "grad_norm": 0.17480257153511047,
+      "learning_rate": 5e-06,
+      "loss": 0.1571,
+      "step": 10258
+    },
+    {
+      "epoch": 7.7222431313511475,
+      "grad_norm": 0.16723628342151642,
+      "learning_rate": 5e-06,
+      "loss": 0.1579,
+      "step": 10259
+    },
+    {
+      "epoch": 7.722995859992473,
+      "grad_norm": 0.15936505794525146,
+      "learning_rate": 5e-06,
+      "loss": 0.1547,
+      "step": 10260
+    },
+    {
+      "epoch": 7.723748588633797,
+      "grad_norm": 0.1634572297334671,
+      "learning_rate": 5e-06,
+      "loss": 0.1773,
+      "step": 10261
+    },
+    {
+      "epoch": 7.724501317275123,
+      "grad_norm": 0.19120050966739655,
+      "learning_rate": 5e-06,
+      "loss": 0.1942,
+      "step": 10262
+    },
+    {
+      "epoch": 7.725254045916447,
+      "grad_norm": 0.16615411639213562,
+      "learning_rate": 5e-06,
+      "loss": 0.1274,
+      "step": 10263
+    },
+    {
+      "epoch": 7.726006774557772,
+      "grad_norm": 0.16659227013587952,
+      "learning_rate": 5e-06,
+      "loss": 0.1033,
+      "step": 10264
+    },
+    {
+      "epoch": 7.726759503199097,
+      "grad_norm": 0.15850894153118134,
+      "learning_rate": 5e-06,
+      "loss": 0.127,
+      "step": 10265
+    },
+    {
+      "epoch": 7.7275122318404215,
+      "grad_norm": 0.17687661945819855,
+      "learning_rate": 5e-06,
+      "loss": 0.1728,
+      "step": 10266
+    },
+    {
+      "epoch": 7.728264960481747,
+      "grad_norm": 0.1645738035440445,
+      "learning_rate": 5e-06,
+      "loss": 0.0999,
+      "step": 10267
+    },
+    {
+      "epoch": 7.729017689123071,
+      "grad_norm": 0.19899751245975494,
+      "learning_rate": 5e-06,
+      "loss": 0.2737,
+      "step": 10268
+    },
+    {
+      "epoch": 7.729770417764396,
+      "grad_norm": 0.17848017811775208,
+      "learning_rate": 5e-06,
+      "loss": 0.1694,
+      "step": 10269
+    },
+    {
+      "epoch": 7.730523146405721,
+      "grad_norm": 0.2043374478816986,
+      "learning_rate": 5e-06,
+      "loss": 0.2027,
+      "step": 10270
+    },
+    {
+      "epoch": 7.731275875047046,
+      "grad_norm": 0.20339149236679077,
+      "learning_rate": 5e-06,
+      "loss": 0.1598,
+      "step": 10271
+    },
+    {
+      "epoch": 7.73202860368837,
+      "grad_norm": 0.15343308448791504,
+      "learning_rate": 5e-06,
+      "loss": 0.1471,
+      "step": 10272
+    },
+    {
+      "epoch": 7.7327813323296954,
+      "grad_norm": 0.18557438254356384,
+      "learning_rate": 5e-06,
+      "loss": 0.1925,
+      "step": 10273
+    },
+    {
+      "epoch": 7.73353406097102,
+      "grad_norm": 0.17043089866638184,
+      "learning_rate": 5e-06,
+      "loss": 0.1677,
+      "step": 10274
+    },
+    {
+      "epoch": 7.734286789612344,
+      "grad_norm": 0.1873033046722412,
+      "learning_rate": 5e-06,
+      "loss": 0.1632,
+      "step": 10275
+    },
+    {
+      "epoch": 7.73503951825367,
+      "grad_norm": 0.16172587871551514,
+      "learning_rate": 5e-06,
+      "loss": 0.1687,
+      "step": 10276
+    },
+    {
+      "epoch": 7.735792246894994,
+      "grad_norm": 0.16095639765262604,
+      "learning_rate": 5e-06,
+      "loss": 0.1693,
+      "step": 10277
+    },
+    {
+      "epoch": 7.736544975536319,
+      "grad_norm": 0.18036577105522156,
+      "learning_rate": 5e-06,
+      "loss": 0.125,
+      "step": 10278
+    },
+    {
+      "epoch": 7.737297704177644,
+      "grad_norm": 0.17019149661064148,
+      "learning_rate": 5e-06,
+      "loss": 0.1216,
+      "step": 10279
+    },
+    {
+      "epoch": 7.7380504328189685,
+      "grad_norm": 0.17849524319171906,
+      "learning_rate": 5e-06,
+      "loss": 0.1419,
+      "step": 10280
+    },
+    {
+      "epoch": 7.738803161460294,
+      "grad_norm": 0.1815095841884613,
+      "learning_rate": 5e-06,
+      "loss": 0.1769,
+      "step": 10281
+    },
+    {
+      "epoch": 7.739555890101618,
+      "grad_norm": 0.16946256160736084,
+      "learning_rate": 5e-06,
+      "loss": 0.2107,
+      "step": 10282
+    },
+    {
+      "epoch": 7.740308618742943,
+      "grad_norm": 0.1686309576034546,
+      "learning_rate": 5e-06,
+      "loss": 0.1331,
+      "step": 10283
+    },
+    {
+      "epoch": 7.741061347384268,
+      "grad_norm": 0.20104096829891205,
+      "learning_rate": 5e-06,
+      "loss": 0.1934,
+      "step": 10284
+    },
+    {
+      "epoch": 7.741814076025593,
+      "grad_norm": 0.1820465326309204,
+      "learning_rate": 5e-06,
+      "loss": 0.1414,
+      "step": 10285
+    },
+    {
+      "epoch": 7.742566804666918,
+      "grad_norm": 0.1819499135017395,
+      "learning_rate": 5e-06,
+      "loss": 0.1712,
+      "step": 10286
+    },
+    {
+      "epoch": 7.7433195333082425,
+      "grad_norm": 0.17196711897850037,
+      "learning_rate": 5e-06,
+      "loss": 0.2591,
+      "step": 10287
+    },
+    {
+      "epoch": 7.744072261949567,
+      "grad_norm": 0.17126864194869995,
+      "learning_rate": 5e-06,
+      "loss": 0.1537,
+      "step": 10288
+    },
+    {
+      "epoch": 7.744824990590892,
+      "grad_norm": 0.1403571218252182,
+      "learning_rate": 5e-06,
+      "loss": 0.0884,
+      "step": 10289
+    },
+    {
+      "epoch": 7.745577719232217,
+      "grad_norm": 0.17338363826274872,
+      "learning_rate": 5e-06,
+      "loss": 0.1058,
+      "step": 10290
+    },
+    {
+      "epoch": 7.746330447873541,
+      "grad_norm": 0.18447190523147583,
+      "learning_rate": 5e-06,
+      "loss": 0.191,
+      "step": 10291
+    },
+    {
+      "epoch": 7.747083176514867,
+      "grad_norm": 0.17602355778217316,
+      "learning_rate": 5e-06,
+      "loss": 0.153,
+      "step": 10292
+    },
+    {
+      "epoch": 7.747835905156191,
+      "grad_norm": 0.1724979132413864,
+      "learning_rate": 5e-06,
+      "loss": 0.1684,
+      "step": 10293
+    },
+    {
+      "epoch": 7.748588633797516,
+      "grad_norm": 0.18783001601696014,
+      "learning_rate": 5e-06,
+      "loss": 0.1925,
+      "step": 10294
+    },
+    {
+      "epoch": 7.749341362438841,
+      "grad_norm": 0.1795843541622162,
+      "learning_rate": 5e-06,
+      "loss": 0.1745,
+      "step": 10295
+    },
+    {
+      "epoch": 7.7500940910801654,
+      "grad_norm": 0.15006543695926666,
+      "learning_rate": 5e-06,
+      "loss": 0.1454,
+      "step": 10296
+    },
+    {
+      "epoch": 7.75084681972149,
+      "grad_norm": 0.16154341399669647,
+      "learning_rate": 5e-06,
+      "loss": 0.1609,
+      "step": 10297
+    },
+    {
+      "epoch": 7.751599548362815,
+      "grad_norm": 0.17190568149089813,
+      "learning_rate": 5e-06,
+      "loss": 0.1602,
+      "step": 10298
+    },
+    {
+      "epoch": 7.75235227700414,
+      "grad_norm": 0.17064689099788666,
+      "learning_rate": 5e-06,
+      "loss": 0.1494,
+      "step": 10299
+    },
+    {
+      "epoch": 7.753105005645465,
+      "grad_norm": 0.18974347412586212,
+      "learning_rate": 5e-06,
+      "loss": 0.2194,
+      "step": 10300
+    },
+    {
+      "epoch": 7.75385773428679,
+      "grad_norm": 0.16442415118217468,
+      "learning_rate": 5e-06,
+      "loss": 0.2244,
+      "step": 10301
+    },
+    {
+      "epoch": 7.754610462928114,
+      "grad_norm": 0.19078846275806427,
+      "learning_rate": 5e-06,
+      "loss": 0.1565,
+      "step": 10302
+    },
+    {
+      "epoch": 7.755363191569439,
+      "grad_norm": 0.16911016404628754,
+      "learning_rate": 5e-06,
+      "loss": 0.1422,
+      "step": 10303
+    },
+    {
+      "epoch": 7.756115920210764,
+      "grad_norm": 0.1715373545885086,
+      "learning_rate": 5e-06,
+      "loss": 0.116,
+      "step": 10304
+    },
+    {
+      "epoch": 7.756868648852089,
+      "grad_norm": 0.1992502063512802,
+      "learning_rate": 5e-06,
+      "loss": 0.2203,
+      "step": 10305
+    },
+    {
+      "epoch": 7.757621377493414,
+      "grad_norm": 0.15100455284118652,
+      "learning_rate": 5e-06,
+      "loss": 0.1703,
+      "step": 10306
+    },
+    {
+      "epoch": 7.758374106134738,
+      "grad_norm": 0.16790249943733215,
+      "learning_rate": 5e-06,
+      "loss": 0.1747,
+      "step": 10307
+    },
+    {
+      "epoch": 7.759126834776064,
+      "grad_norm": 0.1746111661195755,
+      "learning_rate": 5e-06,
+      "loss": 0.1613,
+      "step": 10308
+    },
+    {
+      "epoch": 7.759879563417388,
+      "grad_norm": 0.17482581734657288,
+      "learning_rate": 5e-06,
+      "loss": 0.1572,
+      "step": 10309
+    },
+    {
+      "epoch": 7.7606322920587125,
+      "grad_norm": 0.20785056054592133,
+      "learning_rate": 5e-06,
+      "loss": 0.151,
+      "step": 10310
+    },
+    {
+      "epoch": 7.761385020700038,
+      "grad_norm": 0.16471293568611145,
+      "learning_rate": 5e-06,
+      "loss": 0.2272,
+      "step": 10311
+    },
+    {
+      "epoch": 7.762137749341362,
+      "grad_norm": 0.18622630834579468,
+      "learning_rate": 5e-06,
+      "loss": 0.1289,
+      "step": 10312
+    },
+    {
+      "epoch": 7.762890477982687,
+      "grad_norm": 0.14304865896701813,
+      "learning_rate": 5e-06,
+      "loss": 0.175,
+      "step": 10313
+    },
+    {
+      "epoch": 7.763643206624012,
+      "grad_norm": 0.1656215786933899,
+      "learning_rate": 5e-06,
+      "loss": 0.1263,
+      "step": 10314
+    },
+    {
+      "epoch": 7.764395935265337,
+      "grad_norm": 0.15901640057563782,
+      "learning_rate": 5e-06,
+      "loss": 0.1099,
+      "step": 10315
+    },
+    {
+      "epoch": 7.765148663906662,
+      "grad_norm": 0.17110630869865417,
+      "learning_rate": 5e-06,
+      "loss": 0.1138,
+      "step": 10316
+    },
+    {
+      "epoch": 7.7659013925479865,
+      "grad_norm": 0.16580626368522644,
+      "learning_rate": 5e-06,
+      "loss": 0.2227,
+      "step": 10317
+    },
+    {
+      "epoch": 7.766654121189311,
+      "grad_norm": 0.17129012942314148,
+      "learning_rate": 5e-06,
+      "loss": 0.2092,
+      "step": 10318
+    },
+    {
+      "epoch": 7.767406849830636,
+      "grad_norm": 0.17254896461963654,
+      "learning_rate": 5e-06,
+      "loss": 0.1935,
+      "step": 10319
+    },
+    {
+      "epoch": 7.768159578471961,
+      "grad_norm": 0.18132539093494415,
+      "learning_rate": 5e-06,
+      "loss": 0.1585,
+      "step": 10320
+    },
+    {
+      "epoch": 7.768912307113285,
+      "grad_norm": 0.181572824716568,
+      "learning_rate": 5e-06,
+      "loss": 0.1761,
+      "step": 10321
+    },
+    {
+      "epoch": 7.769665035754611,
+      "grad_norm": 0.15714643895626068,
+      "learning_rate": 5e-06,
+      "loss": 0.1737,
+      "step": 10322
+    },
+    {
+      "epoch": 7.770417764395935,
+      "grad_norm": 0.17446421086788177,
+      "learning_rate": 5e-06,
+      "loss": 0.1808,
+      "step": 10323
+    },
+    {
+      "epoch": 7.7711704930372605,
+      "grad_norm": 0.16685134172439575,
+      "learning_rate": 5e-06,
+      "loss": 0.1876,
+      "step": 10324
+    },
+    {
+      "epoch": 7.771923221678585,
+      "grad_norm": 0.18994049727916718,
+      "learning_rate": 5e-06,
+      "loss": 0.1532,
+      "step": 10325
+    },
+    {
+      "epoch": 7.772675950319909,
+      "grad_norm": 0.1746072769165039,
+      "learning_rate": 5e-06,
+      "loss": 0.1357,
+      "step": 10326
+    },
+    {
+      "epoch": 7.773428678961235,
+      "grad_norm": 0.17079931497573853,
+      "learning_rate": 5e-06,
+      "loss": 0.192,
+      "step": 10327
+    },
+    {
+      "epoch": 7.774181407602559,
+      "grad_norm": 0.1643812507390976,
+      "learning_rate": 5e-06,
+      "loss": 0.1709,
+      "step": 10328
+    },
+    {
+      "epoch": 7.774934136243884,
+      "grad_norm": 0.16573260724544525,
+      "learning_rate": 5e-06,
+      "loss": 0.1078,
+      "step": 10329
+    },
+    {
+      "epoch": 7.775686864885209,
+      "grad_norm": 0.17891660332679749,
+      "learning_rate": 5e-06,
+      "loss": 0.1783,
+      "step": 10330
+    },
+    {
+      "epoch": 7.776439593526534,
+      "grad_norm": 0.19811101257801056,
+      "learning_rate": 5e-06,
+      "loss": 0.1711,
+      "step": 10331
+    },
+    {
+      "epoch": 7.777192322167858,
+      "grad_norm": 0.15231749415397644,
+      "learning_rate": 5e-06,
+      "loss": 0.1447,
+      "step": 10332
+    },
+    {
+      "epoch": 7.777945050809183,
+      "grad_norm": 0.14011304080486298,
+      "learning_rate": 5e-06,
+      "loss": 0.1367,
+      "step": 10333
+    },
+    {
+      "epoch": 7.778697779450508,
+      "grad_norm": 0.16580460965633392,
+      "learning_rate": 5e-06,
+      "loss": 0.2053,
+      "step": 10334
+    },
+    {
+      "epoch": 7.779450508091833,
+      "grad_norm": 0.17972160875797272,
+      "learning_rate": 5e-06,
+      "loss": 0.2049,
+      "step": 10335
+    },
+    {
+      "epoch": 7.780203236733158,
+      "grad_norm": 0.17674484848976135,
+      "learning_rate": 5e-06,
+      "loss": 0.1748,
+      "step": 10336
+    },
+    {
+      "epoch": 7.780955965374482,
+      "grad_norm": 0.1871851235628128,
+      "learning_rate": 5e-06,
+      "loss": 0.1573,
+      "step": 10337
+    },
+    {
+      "epoch": 7.781708694015808,
+      "grad_norm": 0.1712276190519333,
+      "learning_rate": 5e-06,
+      "loss": 0.2184,
+      "step": 10338
+    },
+    {
+      "epoch": 7.782461422657132,
+      "grad_norm": 0.15674850344657898,
+      "learning_rate": 5e-06,
+      "loss": 0.1461,
+      "step": 10339
+    },
+    {
+      "epoch": 7.783214151298457,
+      "grad_norm": 0.16217398643493652,
+      "learning_rate": 5e-06,
+      "loss": 0.1651,
+      "step": 10340
+    },
+    {
+      "epoch": 7.783966879939782,
+      "grad_norm": 0.14669685065746307,
+      "learning_rate": 5e-06,
+      "loss": 0.1043,
+      "step": 10341
+    },
+    {
+      "epoch": 7.784719608581106,
+      "grad_norm": 0.1548566371202469,
+      "learning_rate": 5e-06,
+      "loss": 0.1407,
+      "step": 10342
+    },
+    {
+      "epoch": 7.785472337222432,
+      "grad_norm": 0.16309213638305664,
+      "learning_rate": 5e-06,
+      "loss": 0.119,
+      "step": 10343
+    },
+    {
+      "epoch": 7.786225065863756,
+      "grad_norm": 0.20370380580425262,
+      "learning_rate": 5e-06,
+      "loss": 0.1093,
+      "step": 10344
+    },
+    {
+      "epoch": 7.786977794505081,
+      "grad_norm": 0.1760915070772171,
+      "learning_rate": 5e-06,
+      "loss": 0.1214,
+      "step": 10345
+    },
+    {
+      "epoch": 7.787730523146406,
+      "grad_norm": 0.20487353205680847,
+      "learning_rate": 5e-06,
+      "loss": 0.1594,
+      "step": 10346
+    },
+    {
+      "epoch": 7.7884832517877305,
+      "grad_norm": 0.1681225597858429,
+      "learning_rate": 5e-06,
+      "loss": 0.1976,
+      "step": 10347
+    },
+    {
+      "epoch": 7.789235980429055,
+      "grad_norm": 0.18367356061935425,
+      "learning_rate": 5e-06,
+      "loss": 0.0919,
+      "step": 10348
+    },
+    {
+      "epoch": 7.78998870907038,
+      "grad_norm": 0.16083267331123352,
+      "learning_rate": 5e-06,
+      "loss": 0.1634,
+      "step": 10349
+    },
+    {
+      "epoch": 7.790741437711705,
+      "grad_norm": 0.16733670234680176,
+      "learning_rate": 5e-06,
+      "loss": 0.0691,
+      "step": 10350
+    },
+    {
+      "epoch": 7.791494166353029,
+      "grad_norm": 0.18109707534313202,
+      "learning_rate": 5e-06,
+      "loss": 0.1654,
+      "step": 10351
+    },
+    {
+      "epoch": 7.792246894994355,
+      "grad_norm": 0.16392925381660461,
+      "learning_rate": 5e-06,
+      "loss": 0.0976,
+      "step": 10352
+    },
+    {
+      "epoch": 7.792999623635679,
+      "grad_norm": 0.1983296424150467,
+      "learning_rate": 5e-06,
+      "loss": 0.1326,
+      "step": 10353
+    },
+    {
+      "epoch": 7.7937523522770045,
+      "grad_norm": 0.1879493147134781,
+      "learning_rate": 5e-06,
+      "loss": 0.2091,
+      "step": 10354
+    },
+    {
+      "epoch": 7.794505080918329,
+      "grad_norm": 0.17135076224803925,
+      "learning_rate": 5e-06,
+      "loss": 0.1262,
+      "step": 10355
+    },
+    {
+      "epoch": 7.795257809559653,
+      "grad_norm": 0.15505287051200867,
+      "learning_rate": 5e-06,
+      "loss": 0.1179,
+      "step": 10356
+    },
+    {
+      "epoch": 7.796010538200979,
+      "grad_norm": 0.1678730994462967,
+      "learning_rate": 5e-06,
+      "loss": 0.0913,
+      "step": 10357
+    },
+    {
+      "epoch": 7.796763266842303,
+      "grad_norm": 0.18194587528705597,
+      "learning_rate": 5e-06,
+      "loss": 0.2458,
+      "step": 10358
+    },
+    {
+      "epoch": 7.797515995483629,
+      "grad_norm": 0.1741657555103302,
+      "learning_rate": 5e-06,
+      "loss": 0.2033,
+      "step": 10359
+    },
+    {
+      "epoch": 7.798268724124953,
+      "grad_norm": 0.1460656225681305,
+      "learning_rate": 5e-06,
+      "loss": 0.0696,
+      "step": 10360
+    },
+    {
+      "epoch": 7.799021452766278,
+      "grad_norm": 0.20530493557453156,
+      "learning_rate": 5e-06,
+      "loss": 0.1658,
+      "step": 10361
+    },
+    {
+      "epoch": 7.799774181407603,
+      "grad_norm": 0.16781647503376007,
+      "learning_rate": 5e-06,
+      "loss": 0.1043,
+      "step": 10362
+    },
+    {
+      "epoch": 7.800526910048927,
+      "grad_norm": 0.1709718257188797,
+      "learning_rate": 5e-06,
+      "loss": 0.2386,
+      "step": 10363
+    },
+    {
+      "epoch": 7.801279638690252,
+      "grad_norm": 0.18925030529499054,
+      "learning_rate": 5e-06,
+      "loss": 0.1992,
+      "step": 10364
+    },
+    {
+      "epoch": 7.802032367331577,
+      "grad_norm": 0.16756018996238708,
+      "learning_rate": 5e-06,
+      "loss": 0.1513,
+      "step": 10365
+    },
+    {
+      "epoch": 7.802785095972902,
+      "grad_norm": 0.20215168595314026,
+      "learning_rate": 5e-06,
+      "loss": 0.2283,
+      "step": 10366
+    },
+    {
+      "epoch": 7.803537824614226,
+      "grad_norm": 0.19303341209888458,
+      "learning_rate": 5e-06,
+      "loss": 0.1968,
+      "step": 10367
+    },
+    {
+      "epoch": 7.8042905532555515,
+      "grad_norm": 0.16820453107357025,
+      "learning_rate": 5e-06,
+      "loss": 0.1529,
+      "step": 10368
+    },
+    {
+      "epoch": 7.805043281896876,
+      "grad_norm": 0.16673751175403595,
+      "learning_rate": 5e-06,
+      "loss": 0.0842,
+      "step": 10369
+    },
+    {
+      "epoch": 7.8057960105382005,
+      "grad_norm": 0.17101863026618958,
+      "learning_rate": 5e-06,
+      "loss": 0.1306,
+      "step": 10370
+    },
+    {
+      "epoch": 7.806548739179526,
+      "grad_norm": 0.18310880661010742,
+      "learning_rate": 5e-06,
+      "loss": 0.1854,
+      "step": 10371
+    },
+    {
+      "epoch": 7.80730146782085,
+      "grad_norm": 0.18397407233715057,
+      "learning_rate": 5e-06,
+      "loss": 0.2165,
+      "step": 10372
+    },
+    {
+      "epoch": 7.808054196462176,
+      "grad_norm": 0.17076116800308228,
+      "learning_rate": 5e-06,
+      "loss": 0.1875,
+      "step": 10373
+    },
+    {
+      "epoch": 7.8088069251035,
+      "grad_norm": 0.15115900337696075,
+      "learning_rate": 5e-06,
+      "loss": 0.0872,
+      "step": 10374
+    },
+    {
+      "epoch": 7.809559653744825,
+      "grad_norm": 0.18752150237560272,
+      "learning_rate": 5e-06,
+      "loss": 0.1788,
+      "step": 10375
+    },
+    {
+      "epoch": 7.81031238238615,
+      "grad_norm": 0.17193639278411865,
+      "learning_rate": 5e-06,
+      "loss": 0.1702,
+      "step": 10376
+    },
+    {
+      "epoch": 7.8110651110274745,
+      "grad_norm": 0.17135556042194366,
+      "learning_rate": 5e-06,
+      "loss": 0.2184,
+      "step": 10377
+    },
+    {
+      "epoch": 7.8118178396688,
+      "grad_norm": 0.18226070702075958,
+      "learning_rate": 5e-06,
+      "loss": 0.1635,
+      "step": 10378
+    },
+    {
+      "epoch": 7.812570568310124,
+      "grad_norm": 0.17731226980686188,
+      "learning_rate": 5e-06,
+      "loss": 0.1601,
+      "step": 10379
+    },
+    {
+      "epoch": 7.813323296951449,
+      "grad_norm": 0.16258448362350464,
+      "learning_rate": 5e-06,
+      "loss": 0.1674,
+      "step": 10380
+    },
+    {
+      "epoch": 7.814076025592774,
+      "grad_norm": 0.171820268034935,
+      "learning_rate": 5e-06,
+      "loss": 0.154,
+      "step": 10381
+    },
+    {
+      "epoch": 7.814828754234099,
+      "grad_norm": 0.17103268206119537,
+      "learning_rate": 5e-06,
+      "loss": 0.1781,
+      "step": 10382
+    },
+    {
+      "epoch": 7.815581482875423,
+      "grad_norm": 0.1976029872894287,
+      "learning_rate": 5e-06,
+      "loss": 0.1651,
+      "step": 10383
+    },
+    {
+      "epoch": 7.8163342115167485,
+      "grad_norm": 0.1459185779094696,
+      "learning_rate": 5e-06,
+      "loss": 0.0969,
+      "step": 10384
+    },
+    {
+      "epoch": 7.817086940158073,
+      "grad_norm": 0.15525341033935547,
+      "learning_rate": 5e-06,
+      "loss": 0.1145,
+      "step": 10385
+    },
+    {
+      "epoch": 7.817839668799397,
+      "grad_norm": 0.14352558553218842,
+      "learning_rate": 5e-06,
+      "loss": 0.1253,
+      "step": 10386
+    },
+    {
+      "epoch": 7.818592397440723,
+      "grad_norm": 0.16209010779857635,
+      "learning_rate": 5e-06,
+      "loss": 0.1252,
+      "step": 10387
+    },
+    {
+      "epoch": 7.819345126082047,
+      "grad_norm": 0.1773180216550827,
+      "learning_rate": 5e-06,
+      "loss": 0.1632,
+      "step": 10388
+    },
+    {
+      "epoch": 7.820097854723373,
+      "grad_norm": 0.1780671775341034,
+      "learning_rate": 5e-06,
+      "loss": 0.2336,
+      "step": 10389
+    },
+    {
+      "epoch": 7.820850583364697,
+      "grad_norm": 0.18120189011096954,
+      "learning_rate": 5e-06,
+      "loss": 0.2018,
+      "step": 10390
+    },
+    {
+      "epoch": 7.8216033120060215,
+      "grad_norm": 0.16347140073776245,
+      "learning_rate": 5e-06,
+      "loss": 0.1405,
+      "step": 10391
+    },
+    {
+      "epoch": 7.822356040647347,
+      "grad_norm": 0.17415007948875427,
+      "learning_rate": 5e-06,
+      "loss": 0.1212,
+      "step": 10392
+    },
+    {
+      "epoch": 7.823108769288671,
+      "grad_norm": 0.1848256140947342,
+      "learning_rate": 5e-06,
+      "loss": 0.2547,
+      "step": 10393
+    },
+    {
+      "epoch": 7.823861497929996,
+      "grad_norm": 0.17619389295578003,
+      "learning_rate": 5e-06,
+      "loss": 0.1529,
+      "step": 10394
+    },
+    {
+      "epoch": 7.824614226571321,
+      "grad_norm": 0.17182141542434692,
+      "learning_rate": 5e-06,
+      "loss": 0.1653,
+      "step": 10395
+    },
+    {
+      "epoch": 7.825366955212646,
+      "grad_norm": 0.16498325765132904,
+      "learning_rate": 5e-06,
+      "loss": 0.1627,
+      "step": 10396
+    },
+    {
+      "epoch": 7.826119683853971,
+      "grad_norm": 0.18097122013568878,
+      "learning_rate": 5e-06,
+      "loss": 0.1445,
+      "step": 10397
+    },
+    {
+      "epoch": 7.8268724124952955,
+      "grad_norm": 0.16501609981060028,
+      "learning_rate": 5e-06,
+      "loss": 0.1224,
+      "step": 10398
+    },
+    {
+      "epoch": 7.82762514113662,
+      "grad_norm": 0.17505331337451935,
+      "learning_rate": 5e-06,
+      "loss": 0.2026,
+      "step": 10399
+    },
+    {
+      "epoch": 7.828377869777945,
+      "grad_norm": 0.19068992137908936,
+      "learning_rate": 5e-06,
+      "loss": 0.1931,
+      "step": 10400
+    },
+    {
+      "epoch": 7.828377869777945,
+      "eval_loss": 0.19818571209907532,
+      "eval_runtime": 452.6633,
+      "eval_samples_per_second": 21.267,
+      "eval_steps_per_second": 0.665,
+      "step": 10400
+    },
+    {
+      "epoch": 7.82913059841927,
+      "grad_norm": 0.1794004887342453,
+      "learning_rate": 5e-06,
+      "loss": 0.1392,
+      "step": 10401
+    },
+    {
+      "epoch": 7.829883327060594,
+      "grad_norm": 0.1763933002948761,
+      "learning_rate": 5e-06,
+      "loss": 0.153,
+      "step": 10402
+    },
+    {
+      "epoch": 7.83063605570192,
+      "grad_norm": 0.15707562863826752,
+      "learning_rate": 5e-06,
+      "loss": 0.1143,
+      "step": 10403
+    },
+    {
+      "epoch": 7.831388784343244,
+      "grad_norm": 0.1744367927312851,
+      "learning_rate": 5e-06,
+      "loss": 0.1443,
+      "step": 10404
+    },
+    {
+      "epoch": 7.832141512984569,
+      "grad_norm": 0.17122827470302582,
+      "learning_rate": 5e-06,
+      "loss": 0.1341,
+      "step": 10405
+    },
+    {
+      "epoch": 7.832894241625894,
+      "grad_norm": 0.18515749275684357,
+      "learning_rate": 5e-06,
+      "loss": 0.2047,
+      "step": 10406
+    },
+    {
+      "epoch": 7.8336469702672185,
+      "grad_norm": 0.1854933649301529,
+      "learning_rate": 5e-06,
+      "loss": 0.1839,
+      "step": 10407
+    },
+    {
+      "epoch": 7.834399698908544,
+      "grad_norm": 0.186584010720253,
+      "learning_rate": 5e-06,
+      "loss": 0.1399,
+      "step": 10408
+    },
+    {
+      "epoch": 7.835152427549868,
+      "grad_norm": 0.1616971492767334,
+      "learning_rate": 5e-06,
+      "loss": 0.1488,
+      "step": 10409
+    },
+    {
+      "epoch": 7.835905156191193,
+      "grad_norm": 0.160986989736557,
+      "learning_rate": 5e-06,
+      "loss": 0.1194,
+      "step": 10410
+    },
+    {
+      "epoch": 7.836657884832518,
+      "grad_norm": 0.19459660351276398,
+      "learning_rate": 5e-06,
+      "loss": 0.1347,
+      "step": 10411
+    },
+    {
+      "epoch": 7.837410613473843,
+      "grad_norm": 0.15528905391693115,
+      "learning_rate": 5e-06,
+      "loss": 0.1239,
+      "step": 10412
+    },
+    {
+      "epoch": 7.838163342115168,
+      "grad_norm": 0.19032682478427887,
+      "learning_rate": 5e-06,
+      "loss": 0.1684,
+      "step": 10413
+    },
+    {
+      "epoch": 7.838916070756492,
+      "grad_norm": 0.1790112406015396,
+      "learning_rate": 5e-06,
+      "loss": 0.1904,
+      "step": 10414
+    },
+    {
+      "epoch": 7.839668799397817,
+      "grad_norm": 0.19741500914096832,
+      "learning_rate": 5e-06,
+      "loss": 0.1683,
+      "step": 10415
+    },
+    {
+      "epoch": 7.840421528039142,
+      "grad_norm": 0.1775272935628891,
+      "learning_rate": 5e-06,
+      "loss": 0.1228,
+      "step": 10416
+    },
+    {
+      "epoch": 7.841174256680467,
+      "grad_norm": 0.1805524080991745,
+      "learning_rate": 5e-06,
+      "loss": 0.2001,
+      "step": 10417
+    },
+    {
+      "epoch": 7.841926985321791,
+      "grad_norm": 0.15846870839595795,
+      "learning_rate": 5e-06,
+      "loss": 0.1319,
+      "step": 10418
+    },
+    {
+      "epoch": 7.842679713963117,
+      "grad_norm": 0.18190769851207733,
+      "learning_rate": 5e-06,
+      "loss": 0.2197,
+      "step": 10419
+    },
+    {
+      "epoch": 7.843432442604441,
+      "grad_norm": 0.14791758358478546,
+      "learning_rate": 5e-06,
+      "loss": 0.1029,
+      "step": 10420
+    },
+    {
+      "epoch": 7.8441851712457655,
+      "grad_norm": 0.15850667655467987,
+      "learning_rate": 5e-06,
+      "loss": 0.1409,
+      "step": 10421
+    },
+    {
+      "epoch": 7.844937899887091,
+      "grad_norm": 0.17383241653442383,
+      "learning_rate": 5e-06,
+      "loss": 0.1122,
+      "step": 10422
+    },
+    {
+      "epoch": 7.845690628528415,
+      "grad_norm": 0.18016475439071655,
+      "learning_rate": 5e-06,
+      "loss": 0.1315,
+      "step": 10423
+    },
+    {
+      "epoch": 7.84644335716974,
+      "grad_norm": 0.1857885718345642,
+      "learning_rate": 5e-06,
+      "loss": 0.1979,
+      "step": 10424
+    },
+    {
+      "epoch": 7.847196085811065,
+      "grad_norm": 0.21279148757457733,
+      "learning_rate": 5e-06,
+      "loss": 0.1575,
+      "step": 10425
+    },
+    {
+      "epoch": 7.84794881445239,
+      "grad_norm": 0.15992777049541473,
+      "learning_rate": 5e-06,
+      "loss": 0.1784,
+      "step": 10426
+    },
+    {
+      "epoch": 7.848701543093715,
+      "grad_norm": 0.1506795883178711,
+      "learning_rate": 5e-06,
+      "loss": 0.1469,
+      "step": 10427
+    },
+    {
+      "epoch": 7.8494542717350395,
+      "grad_norm": 0.17269858717918396,
+      "learning_rate": 5e-06,
+      "loss": 0.1054,
+      "step": 10428
+    },
+    {
+      "epoch": 7.850207000376364,
+      "grad_norm": 0.1593117117881775,
+      "learning_rate": 5e-06,
+      "loss": 0.2329,
+      "step": 10429
+    },
+    {
+      "epoch": 7.850959729017689,
+      "grad_norm": 0.16969916224479675,
+      "learning_rate": 5e-06,
+      "loss": 0.1438,
+      "step": 10430
+    },
+    {
+      "epoch": 7.851712457659014,
+      "grad_norm": 0.1795244812965393,
+      "learning_rate": 5e-06,
+      "loss": 0.1989,
+      "step": 10431
+    },
+    {
+      "epoch": 7.852465186300339,
+      "grad_norm": 0.15926940739154816,
+      "learning_rate": 5e-06,
+      "loss": 0.1287,
+      "step": 10432
+    },
+    {
+      "epoch": 7.853217914941664,
+      "grad_norm": 0.17952078580856323,
+      "learning_rate": 5e-06,
+      "loss": 0.1445,
+      "step": 10433
+    },
+    {
+      "epoch": 7.853970643582988,
+      "grad_norm": 0.1785321831703186,
+      "learning_rate": 5e-06,
+      "loss": 0.0931,
+      "step": 10434
+    },
+    {
+      "epoch": 7.8547233722243135,
+      "grad_norm": 0.16301198303699493,
+      "learning_rate": 5e-06,
+      "loss": 0.1308,
+      "step": 10435
+    },
+    {
+      "epoch": 7.855476100865638,
+      "grad_norm": 0.1719595491886139,
+      "learning_rate": 5e-06,
+      "loss": 0.148,
+      "step": 10436
+    },
+    {
+      "epoch": 7.856228829506962,
+      "grad_norm": 0.16155248880386353,
+      "learning_rate": 5e-06,
+      "loss": 0.1307,
+      "step": 10437
+    },
+    {
+      "epoch": 7.856981558148288,
+      "grad_norm": 0.17858895659446716,
+      "learning_rate": 5e-06,
+      "loss": 0.2295,
+      "step": 10438
+    },
+    {
+      "epoch": 7.857734286789612,
+      "grad_norm": 0.16424810886383057,
+      "learning_rate": 5e-06,
+      "loss": 0.1386,
+      "step": 10439
+    },
+    {
+      "epoch": 7.858487015430937,
+      "grad_norm": 0.19934043288230896,
+      "learning_rate": 5e-06,
+      "loss": 0.197,
+      "step": 10440
+    },
+    {
+      "epoch": 7.859239744072262,
+      "grad_norm": 0.2046423852443695,
+      "learning_rate": 5e-06,
+      "loss": 0.1775,
+      "step": 10441
+    },
+    {
+      "epoch": 7.859992472713587,
+      "grad_norm": 0.17153207957744598,
+      "learning_rate": 5e-06,
+      "loss": 0.1862,
+      "step": 10442
+    },
+    {
+      "epoch": 7.860745201354911,
+      "grad_norm": 0.15543897449970245,
+      "learning_rate": 5e-06,
+      "loss": 0.153,
+      "step": 10443
+    },
+    {
+      "epoch": 7.861497929996236,
+      "grad_norm": 0.1680585891008377,
+      "learning_rate": 5e-06,
+      "loss": 0.1893,
+      "step": 10444
+    },
+    {
+      "epoch": 7.862250658637561,
+      "grad_norm": 0.16364221274852753,
+      "learning_rate": 5e-06,
+      "loss": 0.0724,
+      "step": 10445
+    },
+    {
+      "epoch": 7.863003387278886,
+      "grad_norm": 0.15542180836200714,
+      "learning_rate": 5e-06,
+      "loss": 0.0681,
+      "step": 10446
+    },
+    {
+      "epoch": 7.863756115920211,
+      "grad_norm": 0.1923418939113617,
+      "learning_rate": 5e-06,
+      "loss": 0.1315,
+      "step": 10447
+    },
+    {
+      "epoch": 7.864508844561535,
+      "grad_norm": 0.15882927179336548,
+      "learning_rate": 5e-06,
+      "loss": 0.1447,
+      "step": 10448
+    },
+    {
+      "epoch": 7.865261573202861,
+      "grad_norm": 0.16925033926963806,
+      "learning_rate": 5e-06,
+      "loss": 0.1418,
+      "step": 10449
+    },
+    {
+      "epoch": 7.866014301844185,
+      "grad_norm": 0.15866747498512268,
+      "learning_rate": 5e-06,
+      "loss": 0.1335,
+      "step": 10450
+    },
+    {
+      "epoch": 7.86676703048551,
+      "grad_norm": 0.17920218408107758,
+      "learning_rate": 5e-06,
+      "loss": 0.2068,
+      "step": 10451
+    },
+    {
+      "epoch": 7.867519759126835,
+      "grad_norm": 0.17149214446544647,
+      "learning_rate": 5e-06,
+      "loss": 0.1306,
+      "step": 10452
+    },
+    {
+      "epoch": 7.868272487768159,
+      "grad_norm": 0.1933361142873764,
+      "learning_rate": 5e-06,
+      "loss": 0.18,
+      "step": 10453
+    },
+    {
+      "epoch": 7.869025216409485,
+      "grad_norm": 0.1735745370388031,
+      "learning_rate": 5e-06,
+      "loss": 0.0942,
+      "step": 10454
+    },
+    {
+      "epoch": 7.869777945050809,
+      "grad_norm": 0.17176131904125214,
+      "learning_rate": 5e-06,
+      "loss": 0.114,
+      "step": 10455
+    },
+    {
+      "epoch": 7.870530673692134,
+      "grad_norm": 0.17469358444213867,
+      "learning_rate": 5e-06,
+      "loss": 0.168,
+      "step": 10456
+    },
+    {
+      "epoch": 7.871283402333459,
+      "grad_norm": 0.17488901317119598,
+      "learning_rate": 5e-06,
+      "loss": 0.1081,
+      "step": 10457
+    },
+    {
+      "epoch": 7.8720361309747835,
+      "grad_norm": 0.1927560269832611,
+      "learning_rate": 5e-06,
+      "loss": 0.1646,
+      "step": 10458
+    },
+    {
+      "epoch": 7.872788859616108,
+      "grad_norm": 0.1609954833984375,
+      "learning_rate": 5e-06,
+      "loss": 0.2303,
+      "step": 10459
+    },
+    {
+      "epoch": 7.873541588257433,
+      "grad_norm": 0.1409468948841095,
+      "learning_rate": 5e-06,
+      "loss": 0.0554,
+      "step": 10460
+    },
+    {
+      "epoch": 7.874294316898758,
+      "grad_norm": 0.1843308061361313,
+      "learning_rate": 5e-06,
+      "loss": 0.1669,
+      "step": 10461
+    },
+    {
+      "epoch": 7.875047045540083,
+      "grad_norm": 0.1815812736749649,
+      "learning_rate": 5e-06,
+      "loss": 0.1867,
+      "step": 10462
+    },
+    {
+      "epoch": 7.875799774181408,
+      "grad_norm": 0.17086006700992584,
+      "learning_rate": 5e-06,
+      "loss": 0.1188,
+      "step": 10463
+    },
+    {
+      "epoch": 7.876552502822732,
+      "grad_norm": 0.1839863657951355,
+      "learning_rate": 5e-06,
+      "loss": 0.1998,
+      "step": 10464
+    },
+    {
+      "epoch": 7.8773052314640575,
+      "grad_norm": 0.175687775015831,
+      "learning_rate": 5e-06,
+      "loss": 0.1014,
+      "step": 10465
+    },
+    {
+      "epoch": 7.878057960105382,
+      "grad_norm": 0.21282726526260376,
+      "learning_rate": 5e-06,
+      "loss": 0.1665,
+      "step": 10466
+    },
+    {
+      "epoch": 7.878810688746706,
+      "grad_norm": 0.17153622210025787,
+      "learning_rate": 5e-06,
+      "loss": 0.1068,
+      "step": 10467
+    },
+    {
+      "epoch": 7.879563417388032,
+      "grad_norm": 0.17556731402873993,
+      "learning_rate": 5e-06,
+      "loss": 0.1625,
+      "step": 10468
+    },
+    {
+      "epoch": 7.880316146029356,
+      "grad_norm": 0.15831153094768524,
+      "learning_rate": 5e-06,
+      "loss": 0.1336,
+      "step": 10469
+    },
+    {
+      "epoch": 7.881068874670682,
+      "grad_norm": 0.1563747525215149,
+      "learning_rate": 5e-06,
+      "loss": 0.188,
+      "step": 10470
+    },
+    {
+      "epoch": 7.881821603312006,
+      "grad_norm": 0.184896782040596,
+      "learning_rate": 5e-06,
+      "loss": 0.1109,
+      "step": 10471
+    },
+    {
+      "epoch": 7.882574331953331,
+      "grad_norm": 0.17667023837566376,
+      "learning_rate": 5e-06,
+      "loss": 0.1774,
+      "step": 10472
+    },
+    {
+      "epoch": 7.883327060594656,
+      "grad_norm": 0.16022704541683197,
+      "learning_rate": 5e-06,
+      "loss": 0.1531,
+      "step": 10473
+    },
+    {
+      "epoch": 7.88407978923598,
+      "grad_norm": 0.1909102350473404,
+      "learning_rate": 5e-06,
+      "loss": 0.1976,
+      "step": 10474
+    },
+    {
+      "epoch": 7.884832517877305,
+      "grad_norm": 0.16358067095279694,
+      "learning_rate": 5e-06,
+      "loss": 0.1571,
+      "step": 10475
+    },
+    {
+      "epoch": 7.88558524651863,
+      "grad_norm": 0.16922885179519653,
+      "learning_rate": 5e-06,
+      "loss": 0.1035,
+      "step": 10476
+    },
+    {
+      "epoch": 7.886337975159955,
+      "grad_norm": 0.18916679918766022,
+      "learning_rate": 5e-06,
+      "loss": 0.1347,
+      "step": 10477
+    },
+    {
+      "epoch": 7.887090703801279,
+      "grad_norm": 0.16750644147396088,
+      "learning_rate": 5e-06,
+      "loss": 0.1579,
+      "step": 10478
+    },
+    {
+      "epoch": 7.8878434324426046,
+      "grad_norm": 0.1799430102109909,
+      "learning_rate": 5e-06,
+      "loss": 0.1716,
+      "step": 10479
+    },
+    {
+      "epoch": 7.888596161083929,
+      "grad_norm": 0.16659456491470337,
+      "learning_rate": 5e-06,
+      "loss": 0.1453,
+      "step": 10480
+    },
+    {
+      "epoch": 7.889348889725254,
+      "grad_norm": 0.1856493502855301,
+      "learning_rate": 5e-06,
+      "loss": 0.1696,
+      "step": 10481
+    },
+    {
+      "epoch": 7.890101618366579,
+      "grad_norm": 0.185565784573555,
+      "learning_rate": 5e-06,
+      "loss": 0.1271,
+      "step": 10482
+    },
+    {
+      "epoch": 7.890854347007903,
+      "grad_norm": 0.1839095503091812,
+      "learning_rate": 5e-06,
+      "loss": 0.1146,
+      "step": 10483
+    },
+    {
+      "epoch": 7.891607075649229,
+      "grad_norm": 0.17781230807304382,
+      "learning_rate": 5e-06,
+      "loss": 0.2387,
+      "step": 10484
+    },
+    {
+      "epoch": 7.892359804290553,
+      "grad_norm": 0.1826009899377823,
+      "learning_rate": 5e-06,
+      "loss": 0.1028,
+      "step": 10485
+    },
+    {
+      "epoch": 7.8931125329318785,
+      "grad_norm": 0.17720139026641846,
+      "learning_rate": 5e-06,
+      "loss": 0.2117,
+      "step": 10486
+    },
+    {
+      "epoch": 7.893865261573203,
+      "grad_norm": 0.18146395683288574,
+      "learning_rate": 5e-06,
+      "loss": 0.2196,
+      "step": 10487
+    },
+    {
+      "epoch": 7.8946179902145275,
+      "grad_norm": 0.1726980358362198,
+      "learning_rate": 5e-06,
+      "loss": 0.1941,
+      "step": 10488
+    },
+    {
+      "epoch": 7.895370718855853,
+      "grad_norm": 0.161845400929451,
+      "learning_rate": 5e-06,
+      "loss": 0.1483,
+      "step": 10489
+    },
+    {
+      "epoch": 7.896123447497177,
+      "grad_norm": 0.170225128531456,
+      "learning_rate": 5e-06,
+      "loss": 0.1594,
+      "step": 10490
+    },
+    {
+      "epoch": 7.896876176138502,
+      "grad_norm": 0.16558890044689178,
+      "learning_rate": 5e-06,
+      "loss": 0.1293,
+      "step": 10491
+    },
+    {
+      "epoch": 7.897628904779827,
+      "grad_norm": 0.18863724172115326,
+      "learning_rate": 5e-06,
+      "loss": 0.1324,
+      "step": 10492
+    },
+    {
+      "epoch": 7.898381633421152,
+      "grad_norm": 0.18392132222652435,
+      "learning_rate": 5e-06,
+      "loss": 0.1535,
+      "step": 10493
+    },
+    {
+      "epoch": 7.899134362062476,
+      "grad_norm": 0.19429652392864227,
+      "learning_rate": 5e-06,
+      "loss": 0.1326,
+      "step": 10494
+    },
+    {
+      "epoch": 7.8998870907038015,
+      "grad_norm": 0.1570025384426117,
+      "learning_rate": 5e-06,
+      "loss": 0.1397,
+      "step": 10495
+    },
+    {
+      "epoch": 7.900639819345126,
+      "grad_norm": 0.18436144292354584,
+      "learning_rate": 5e-06,
+      "loss": 0.1793,
+      "step": 10496
+    },
+    {
+      "epoch": 7.90139254798645,
+      "grad_norm": 0.18264584243297577,
+      "learning_rate": 5e-06,
+      "loss": 0.1285,
+      "step": 10497
+    },
+    {
+      "epoch": 7.902145276627776,
+      "grad_norm": 0.17773638665676117,
+      "learning_rate": 5e-06,
+      "loss": 0.0997,
+      "step": 10498
+    },
+    {
+      "epoch": 7.9028980052691,
+      "grad_norm": 0.15895211696624756,
+      "learning_rate": 5e-06,
+      "loss": 0.1576,
+      "step": 10499
+    },
+    {
+      "epoch": 7.903650733910426,
+      "grad_norm": 0.16698002815246582,
+      "learning_rate": 5e-06,
+      "loss": 0.2495,
+      "step": 10500
+    },
+    {
+      "epoch": 7.90440346255175,
+      "grad_norm": 0.1806250810623169,
+      "learning_rate": 5e-06,
+      "loss": 0.0968,
+      "step": 10501
+    },
+    {
+      "epoch": 7.9051561911930746,
+      "grad_norm": 0.15625673532485962,
+      "learning_rate": 5e-06,
+      "loss": 0.0966,
+      "step": 10502
+    },
+    {
+      "epoch": 7.9059089198344,
+      "grad_norm": 0.17032566666603088,
+      "learning_rate": 5e-06,
+      "loss": 0.09,
+      "step": 10503
+    },
+    {
+      "epoch": 7.906661648475724,
+      "grad_norm": 0.2014005184173584,
+      "learning_rate": 5e-06,
+      "loss": 0.177,
+      "step": 10504
+    },
+    {
+      "epoch": 7.90741437711705,
+      "grad_norm": 0.15814617276191711,
+      "learning_rate": 5e-06,
+      "loss": 0.1406,
+      "step": 10505
+    },
+    {
+      "epoch": 7.908167105758374,
+      "grad_norm": 0.15488660335540771,
+      "learning_rate": 5e-06,
+      "loss": 0.1272,
+      "step": 10506
+    },
+    {
+      "epoch": 7.908919834399699,
+      "grad_norm": 0.1889989823102951,
+      "learning_rate": 5e-06,
+      "loss": 0.1456,
+      "step": 10507
+    },
+    {
+      "epoch": 7.909672563041024,
+      "grad_norm": 0.18405841290950775,
+      "learning_rate": 5e-06,
+      "loss": 0.1577,
+      "step": 10508
+    },
+    {
+      "epoch": 7.9104252916823485,
+      "grad_norm": 0.19077877700328827,
+      "learning_rate": 5e-06,
+      "loss": 0.2314,
+      "step": 10509
+    },
+    {
+      "epoch": 7.911178020323673,
+      "grad_norm": 0.17922422289848328,
+      "learning_rate": 5e-06,
+      "loss": 0.1475,
+      "step": 10510
+    },
+    {
+      "epoch": 7.911930748964998,
+      "grad_norm": 0.16838297247886658,
+      "learning_rate": 5e-06,
+      "loss": 0.1469,
+      "step": 10511
+    },
+    {
+      "epoch": 7.912683477606323,
+      "grad_norm": 0.1814161241054535,
+      "learning_rate": 5e-06,
+      "loss": 0.1736,
+      "step": 10512
+    },
+    {
+      "epoch": 7.913436206247647,
+      "grad_norm": 0.18554408848285675,
+      "learning_rate": 5e-06,
+      "loss": 0.1682,
+      "step": 10513
+    },
+    {
+      "epoch": 7.914188934888973,
+      "grad_norm": 0.1563807725906372,
+      "learning_rate": 5e-06,
+      "loss": 0.1775,
+      "step": 10514
+    },
+    {
+      "epoch": 7.914941663530297,
+      "grad_norm": 0.1748068779706955,
+      "learning_rate": 5e-06,
+      "loss": 0.21,
+      "step": 10515
+    },
+    {
+      "epoch": 7.9156943921716225,
+      "grad_norm": 0.17410998046398163,
+      "learning_rate": 5e-06,
+      "loss": 0.0757,
+      "step": 10516
+    },
+    {
+      "epoch": 7.916447120812947,
+      "grad_norm": 0.18172843754291534,
+      "learning_rate": 5e-06,
+      "loss": 0.1352,
+      "step": 10517
+    },
+    {
+      "epoch": 7.9171998494542715,
+      "grad_norm": 0.1822301149368286,
+      "learning_rate": 5e-06,
+      "loss": 0.1119,
+      "step": 10518
+    },
+    {
+      "epoch": 7.917952578095597,
+      "grad_norm": 0.21517564356327057,
+      "learning_rate": 5e-06,
+      "loss": 0.2436,
+      "step": 10519
+    },
+    {
+      "epoch": 7.918705306736921,
+      "grad_norm": 0.20552216470241547,
+      "learning_rate": 5e-06,
+      "loss": 0.1941,
+      "step": 10520
+    },
+    {
+      "epoch": 7.919458035378246,
+      "grad_norm": 0.16693642735481262,
+      "learning_rate": 5e-06,
+      "loss": 0.0999,
+      "step": 10521
+    },
+    {
+      "epoch": 7.920210764019571,
+      "grad_norm": 0.1856323778629303,
+      "learning_rate": 5e-06,
+      "loss": 0.1803,
+      "step": 10522
+    },
+    {
+      "epoch": 7.920963492660896,
+      "grad_norm": 0.17901113629341125,
+      "learning_rate": 5e-06,
+      "loss": 0.1382,
+      "step": 10523
+    },
+    {
+      "epoch": 7.921716221302221,
+      "grad_norm": 0.16558252274990082,
+      "learning_rate": 5e-06,
+      "loss": 0.1071,
+      "step": 10524
+    },
+    {
+      "epoch": 7.9224689499435454,
+      "grad_norm": 0.18385158479213715,
+      "learning_rate": 5e-06,
+      "loss": 0.1076,
+      "step": 10525
+    },
+    {
+      "epoch": 7.92322167858487,
+      "grad_norm": 0.2025388479232788,
+      "learning_rate": 5e-06,
+      "loss": 0.2267,
+      "step": 10526
+    },
+    {
+      "epoch": 7.923974407226195,
+      "grad_norm": 0.16210821270942688,
+      "learning_rate": 5e-06,
+      "loss": 0.1284,
+      "step": 10527
+    },
+    {
+      "epoch": 7.92472713586752,
+      "grad_norm": 0.16442638635635376,
+      "learning_rate": 5e-06,
+      "loss": 0.1335,
+      "step": 10528
+    },
+    {
+      "epoch": 7.925479864508844,
+      "grad_norm": 0.1446906477212906,
+      "learning_rate": 5e-06,
+      "loss": 0.1214,
+      "step": 10529
+    },
+    {
+      "epoch": 7.92623259315017,
+      "grad_norm": 0.16478650271892548,
+      "learning_rate": 5e-06,
+      "loss": 0.141,
+      "step": 10530
+    },
+    {
+      "epoch": 7.926985321791494,
+      "grad_norm": 0.18487758934497833,
+      "learning_rate": 5e-06,
+      "loss": 0.1497,
+      "step": 10531
+    },
+    {
+      "epoch": 7.9277380504328185,
+      "grad_norm": 0.23268680274486542,
+      "learning_rate": 5e-06,
+      "loss": 0.1981,
+      "step": 10532
+    },
+    {
+      "epoch": 7.928490779074144,
+      "grad_norm": 0.16414576768875122,
+      "learning_rate": 5e-06,
+      "loss": 0.1126,
+      "step": 10533
+    },
+    {
+      "epoch": 7.929243507715468,
+      "grad_norm": 0.16639381647109985,
+      "learning_rate": 5e-06,
+      "loss": 0.1378,
+      "step": 10534
+    },
+    {
+      "epoch": 7.929996236356794,
+      "grad_norm": 0.157688707113266,
+      "learning_rate": 5e-06,
+      "loss": 0.1179,
+      "step": 10535
+    },
+    {
+      "epoch": 7.930748964998118,
+      "grad_norm": 0.1511930227279663,
+      "learning_rate": 5e-06,
+      "loss": 0.1349,
+      "step": 10536
+    },
+    {
+      "epoch": 7.931501693639443,
+      "grad_norm": 0.17040586471557617,
+      "learning_rate": 5e-06,
+      "loss": 0.1381,
+      "step": 10537
+    },
+    {
+      "epoch": 7.932254422280768,
+      "grad_norm": 0.18015047907829285,
+      "learning_rate": 5e-06,
+      "loss": 0.1921,
+      "step": 10538
+    },
+    {
+      "epoch": 7.9330071509220925,
+      "grad_norm": 0.2003893405199051,
+      "learning_rate": 5e-06,
+      "loss": 0.1458,
+      "step": 10539
+    },
+    {
+      "epoch": 7.933759879563418,
+      "grad_norm": 0.15853774547576904,
+      "learning_rate": 5e-06,
+      "loss": 0.1116,
+      "step": 10540
+    },
+    {
+      "epoch": 7.934512608204742,
+      "grad_norm": 0.1946670562028885,
+      "learning_rate": 5e-06,
+      "loss": 0.1031,
+      "step": 10541
+    },
+    {
+      "epoch": 7.935265336846067,
+      "grad_norm": 0.16106371581554413,
+      "learning_rate": 5e-06,
+      "loss": 0.125,
+      "step": 10542
+    },
+    {
+      "epoch": 7.936018065487392,
+      "grad_norm": 0.13691405951976776,
+      "learning_rate": 5e-06,
+      "loss": 0.1319,
+      "step": 10543
+    },
+    {
+      "epoch": 7.936770794128717,
+      "grad_norm": 0.1929137259721756,
+      "learning_rate": 5e-06,
+      "loss": 0.2549,
+      "step": 10544
+    },
+    {
+      "epoch": 7.937523522770041,
+      "grad_norm": 0.15326285362243652,
+      "learning_rate": 5e-06,
+      "loss": 0.1014,
+      "step": 10545
+    },
+    {
+      "epoch": 7.9382762514113665,
+      "grad_norm": 0.1716601550579071,
+      "learning_rate": 5e-06,
+      "loss": 0.1833,
+      "step": 10546
+    },
+    {
+      "epoch": 7.939028980052691,
+      "grad_norm": 0.16845718026161194,
+      "learning_rate": 5e-06,
+      "loss": 0.1728,
+      "step": 10547
+    },
+    {
+      "epoch": 7.9397817086940154,
+      "grad_norm": 0.15758025646209717,
+      "learning_rate": 5e-06,
+      "loss": 0.1378,
+      "step": 10548
+    },
+    {
+      "epoch": 7.940534437335341,
+      "grad_norm": 0.17937950789928436,
+      "learning_rate": 5e-06,
+      "loss": 0.1387,
+      "step": 10549
+    },
+    {
+      "epoch": 7.941287165976665,
+      "grad_norm": 0.18237029016017914,
+      "learning_rate": 5e-06,
+      "loss": 0.2198,
+      "step": 10550
+    },
+    {
+      "epoch": 7.94203989461799,
+      "grad_norm": 0.18162941932678223,
+      "learning_rate": 5e-06,
+      "loss": 0.1758,
+      "step": 10551
+    },
+    {
+      "epoch": 7.942792623259315,
+      "grad_norm": 0.1751306802034378,
+      "learning_rate": 5e-06,
+      "loss": 0.1744,
+      "step": 10552
+    },
+    {
+      "epoch": 7.94354535190064,
+      "grad_norm": 0.14871300756931305,
+      "learning_rate": 5e-06,
+      "loss": 0.1289,
+      "step": 10553
+    },
+    {
+      "epoch": 7.944298080541965,
+      "grad_norm": 0.16699321568012238,
+      "learning_rate": 5e-06,
+      "loss": 0.1835,
+      "step": 10554
+    },
+    {
+      "epoch": 7.945050809183289,
+      "grad_norm": 0.1653067171573639,
+      "learning_rate": 5e-06,
+      "loss": 0.1218,
+      "step": 10555
+    },
+    {
+      "epoch": 7.945803537824614,
+      "grad_norm": 0.16796046495437622,
+      "learning_rate": 5e-06,
+      "loss": 0.185,
+      "step": 10556
+    },
+    {
+      "epoch": 7.946556266465939,
+      "grad_norm": 0.2014586478471756,
+      "learning_rate": 5e-06,
+      "loss": 0.2364,
+      "step": 10557
+    },
+    {
+      "epoch": 7.947308995107264,
+      "grad_norm": 0.15926901996135712,
+      "learning_rate": 5e-06,
+      "loss": 0.1503,
+      "step": 10558
+    },
+    {
+      "epoch": 7.948061723748589,
+      "grad_norm": 0.18757709860801697,
+      "learning_rate": 5e-06,
+      "loss": 0.1545,
+      "step": 10559
+    },
+    {
+      "epoch": 7.948814452389914,
+      "grad_norm": 0.17633816599845886,
+      "learning_rate": 5e-06,
+      "loss": 0.1388,
+      "step": 10560
+    },
+    {
+      "epoch": 7.949567181031238,
+      "grad_norm": 0.17023156583309174,
+      "learning_rate": 5e-06,
+      "loss": 0.1336,
+      "step": 10561
+    },
+    {
+      "epoch": 7.950319909672563,
+      "grad_norm": 0.18111641705036163,
+      "learning_rate": 5e-06,
+      "loss": 0.1355,
+      "step": 10562
+    },
+    {
+      "epoch": 7.951072638313888,
+      "grad_norm": 0.15987521409988403,
+      "learning_rate": 5e-06,
+      "loss": 0.1149,
+      "step": 10563
+    },
+    {
+      "epoch": 7.951825366955212,
+      "grad_norm": 0.15509887039661407,
+      "learning_rate": 5e-06,
+      "loss": 0.1523,
+      "step": 10564
+    },
+    {
+      "epoch": 7.952578095596538,
+      "grad_norm": 0.1717749387025833,
+      "learning_rate": 5e-06,
+      "loss": 0.1212,
+      "step": 10565
+    },
+    {
+      "epoch": 7.953330824237862,
+      "grad_norm": 0.1804409623146057,
+      "learning_rate": 5e-06,
+      "loss": 0.1187,
+      "step": 10566
+    },
+    {
+      "epoch": 7.954083552879187,
+      "grad_norm": 0.1932348608970642,
+      "learning_rate": 5e-06,
+      "loss": 0.1114,
+      "step": 10567
+    },
+    {
+      "epoch": 7.954836281520512,
+      "grad_norm": 0.16805307567119598,
+      "learning_rate": 5e-06,
+      "loss": 0.1243,
+      "step": 10568
+    },
+    {
+      "epoch": 7.9555890101618365,
+      "grad_norm": 0.21111133694648743,
+      "learning_rate": 5e-06,
+      "loss": 0.0787,
+      "step": 10569
+    },
+    {
+      "epoch": 7.956341738803161,
+      "grad_norm": 0.1738795042037964,
+      "learning_rate": 5e-06,
+      "loss": 0.1861,
+      "step": 10570
+    },
+    {
+      "epoch": 7.957094467444486,
+      "grad_norm": 0.1832365244626999,
+      "learning_rate": 5e-06,
+      "loss": 0.1499,
+      "step": 10571
+    },
+    {
+      "epoch": 7.957847196085811,
+      "grad_norm": 0.16532491147518158,
+      "learning_rate": 5e-06,
+      "loss": 0.1208,
+      "step": 10572
+    },
+    {
+      "epoch": 7.958599924727136,
+      "grad_norm": 0.1635986715555191,
+      "learning_rate": 5e-06,
+      "loss": 0.1879,
+      "step": 10573
+    },
+    {
+      "epoch": 7.959352653368461,
+      "grad_norm": 0.1712152063846588,
+      "learning_rate": 5e-06,
+      "loss": 0.1169,
+      "step": 10574
+    },
+    {
+      "epoch": 7.960105382009785,
+      "grad_norm": 0.17568036913871765,
+      "learning_rate": 5e-06,
+      "loss": 0.1565,
+      "step": 10575
+    },
+    {
+      "epoch": 7.9608581106511105,
+      "grad_norm": 0.19220015406608582,
+      "learning_rate": 5e-06,
+      "loss": 0.1198,
+      "step": 10576
+    },
+    {
+      "epoch": 7.961610839292435,
+      "grad_norm": 0.20094189047813416,
+      "learning_rate": 5e-06,
+      "loss": 0.2523,
+      "step": 10577
+    },
+    {
+      "epoch": 7.96236356793376,
+      "grad_norm": 0.16207584738731384,
+      "learning_rate": 5e-06,
+      "loss": 0.0804,
+      "step": 10578
+    },
+    {
+      "epoch": 7.963116296575085,
+      "grad_norm": 0.18669313192367554,
+      "learning_rate": 5e-06,
+      "loss": 0.1813,
+      "step": 10579
+    },
+    {
+      "epoch": 7.963869025216409,
+      "grad_norm": 0.15465959906578064,
+      "learning_rate": 5e-06,
+      "loss": 0.1328,
+      "step": 10580
+    },
+    {
+      "epoch": 7.964621753857735,
+      "grad_norm": 0.171953946352005,
+      "learning_rate": 5e-06,
+      "loss": 0.0524,
+      "step": 10581
+    },
+    {
+      "epoch": 7.965374482499059,
+      "grad_norm": 0.14263807237148285,
+      "learning_rate": 5e-06,
+      "loss": 0.084,
+      "step": 10582
+    },
+    {
+      "epoch": 7.966127211140384,
+      "grad_norm": 0.16820864379405975,
+      "learning_rate": 5e-06,
+      "loss": 0.2264,
+      "step": 10583
+    },
+    {
+      "epoch": 7.966879939781709,
+      "grad_norm": 0.17590157687664032,
+      "learning_rate": 5e-06,
+      "loss": 0.1779,
+      "step": 10584
+    },
+    {
+      "epoch": 7.967632668423033,
+      "grad_norm": 0.17069044709205627,
+      "learning_rate": 5e-06,
+      "loss": 0.1449,
+      "step": 10585
+    },
+    {
+      "epoch": 7.968385397064358,
+      "grad_norm": 0.17646531760692596,
+      "learning_rate": 5e-06,
+      "loss": 0.0899,
+      "step": 10586
+    },
+    {
+      "epoch": 7.969138125705683,
+      "grad_norm": 0.16775958240032196,
+      "learning_rate": 5e-06,
+      "loss": 0.1162,
+      "step": 10587
+    },
+    {
+      "epoch": 7.969890854347008,
+      "grad_norm": 0.19422702491283417,
+      "learning_rate": 5e-06,
+      "loss": 0.1162,
+      "step": 10588
+    },
+    {
+      "epoch": 7.970643582988333,
+      "grad_norm": 0.20378142595291138,
+      "learning_rate": 5e-06,
+      "loss": 0.1932,
+      "step": 10589
+    },
+    {
+      "epoch": 7.971396311629658,
+      "grad_norm": 0.18244822323322296,
+      "learning_rate": 5e-06,
+      "loss": 0.229,
+      "step": 10590
+    },
+    {
+      "epoch": 7.972149040270982,
+      "grad_norm": 0.16533920168876648,
+      "learning_rate": 5e-06,
+      "loss": 0.1809,
+      "step": 10591
+    },
+    {
+      "epoch": 7.972901768912307,
+      "grad_norm": 0.1713644415140152,
+      "learning_rate": 5e-06,
+      "loss": 0.0987,
+      "step": 10592
+    },
+    {
+      "epoch": 7.973654497553632,
+      "grad_norm": 0.18008393049240112,
+      "learning_rate": 5e-06,
+      "loss": 0.1369,
+      "step": 10593
+    },
+    {
+      "epoch": 7.974407226194956,
+      "grad_norm": 0.16723570227622986,
+      "learning_rate": 5e-06,
+      "loss": 0.1061,
+      "step": 10594
+    },
+    {
+      "epoch": 7.975159954836282,
+      "grad_norm": 0.1908866912126541,
+      "learning_rate": 5e-06,
+      "loss": 0.1807,
+      "step": 10595
+    },
+    {
+      "epoch": 7.975912683477606,
+      "grad_norm": 0.16977928578853607,
+      "learning_rate": 5e-06,
+      "loss": 0.1695,
+      "step": 10596
+    },
+    {
+      "epoch": 7.9766654121189315,
+      "grad_norm": 0.1724618673324585,
+      "learning_rate": 5e-06,
+      "loss": 0.1353,
+      "step": 10597
+    },
+    {
+      "epoch": 7.977418140760256,
+      "grad_norm": 0.17652849853038788,
+      "learning_rate": 5e-06,
+      "loss": 0.1889,
+      "step": 10598
+    },
+    {
+      "epoch": 7.9781708694015805,
+      "grad_norm": 0.16223205626010895,
+      "learning_rate": 5e-06,
+      "loss": 0.1732,
+      "step": 10599
+    },
+    {
+      "epoch": 7.978923598042906,
+      "grad_norm": 0.1871464103460312,
+      "learning_rate": 5e-06,
+      "loss": 0.1954,
+      "step": 10600
+    },
+    {
+      "epoch": 7.978923598042906,
+      "eval_loss": 0.19740720093250275,
+      "eval_runtime": 452.6865,
+      "eval_samples_per_second": 21.266,
+      "eval_steps_per_second": 0.665,
+      "step": 10600
+    },
+    {
+      "epoch": 7.97967632668423,
+      "grad_norm": 0.1866469383239746,
+      "learning_rate": 5e-06,
+      "loss": 0.174,
+      "step": 10601
+    },
+    {
+      "epoch": 7.980429055325555,
+      "grad_norm": 0.17138336598873138,
+      "learning_rate": 5e-06,
+      "loss": 0.156,
+      "step": 10602
+    },
+    {
+      "epoch": 7.98118178396688,
+      "grad_norm": 0.16959266364574432,
+      "learning_rate": 5e-06,
+      "loss": 0.1419,
+      "step": 10603
+    },
+    {
+      "epoch": 7.981934512608205,
+      "grad_norm": 0.16892686486244202,
+      "learning_rate": 5e-06,
+      "loss": 0.1754,
+      "step": 10604
+    },
+    {
+      "epoch": 7.982687241249529,
+      "grad_norm": 0.18671923875808716,
+      "learning_rate": 5e-06,
+      "loss": 0.1413,
+      "step": 10605
+    },
+    {
+      "epoch": 7.9834399698908545,
+      "grad_norm": 0.15117092430591583,
+      "learning_rate": 5e-06,
+      "loss": 0.1532,
+      "step": 10606
+    },
+    {
+      "epoch": 7.984192698532179,
+      "grad_norm": 0.18648268282413483,
+      "learning_rate": 5e-06,
+      "loss": 0.1567,
+      "step": 10607
+    },
+    {
+      "epoch": 7.984945427173504,
+      "grad_norm": 0.17605070769786835,
+      "learning_rate": 5e-06,
+      "loss": 0.2271,
+      "step": 10608
+    },
+    {
+      "epoch": 7.985698155814829,
+      "grad_norm": 0.18296979367733002,
+      "learning_rate": 5e-06,
+      "loss": 0.1409,
+      "step": 10609
+    },
+    {
+      "epoch": 7.986450884456153,
+      "grad_norm": 0.15655933320522308,
+      "learning_rate": 5e-06,
+      "loss": 0.0988,
+      "step": 10610
+    },
+    {
+      "epoch": 7.987203613097479,
+      "grad_norm": 0.1737181395292282,
+      "learning_rate": 5e-06,
+      "loss": 0.2087,
+      "step": 10611
+    },
+    {
+      "epoch": 7.987956341738803,
+      "grad_norm": 0.16622112691402435,
+      "learning_rate": 5e-06,
+      "loss": 0.1625,
+      "step": 10612
+    },
+    {
+      "epoch": 7.9887090703801285,
+      "grad_norm": 0.19708025455474854,
+      "learning_rate": 5e-06,
+      "loss": 0.1631,
+      "step": 10613
+    },
+    {
+      "epoch": 7.989461799021453,
+      "grad_norm": 0.17799140512943268,
+      "learning_rate": 5e-06,
+      "loss": 0.2451,
+      "step": 10614
+    },
+    {
+      "epoch": 7.990214527662777,
+      "grad_norm": 0.18216712772846222,
+      "learning_rate": 5e-06,
+      "loss": 0.2623,
+      "step": 10615
+    },
+    {
+      "epoch": 7.990967256304103,
+      "grad_norm": 0.15291103720664978,
+      "learning_rate": 5e-06,
+      "loss": 0.1457,
+      "step": 10616
+    },
+    {
+      "epoch": 7.991719984945427,
+      "grad_norm": 0.14999130368232727,
+      "learning_rate": 5e-06,
+      "loss": 0.1337,
+      "step": 10617
+    },
+    {
+      "epoch": 7.992472713586752,
+      "grad_norm": 0.1689845621585846,
+      "learning_rate": 5e-06,
+      "loss": 0.0949,
+      "step": 10618
+    },
+    {
+      "epoch": 7.993225442228077,
+      "grad_norm": 0.1772058606147766,
+      "learning_rate": 5e-06,
+      "loss": 0.1744,
+      "step": 10619
+    },
+    {
+      "epoch": 7.9939781708694015,
+      "grad_norm": 0.17137643694877625,
+      "learning_rate": 5e-06,
+      "loss": 0.141,
+      "step": 10620
+    },
+    {
+      "epoch": 7.994730899510726,
+      "grad_norm": 0.16360481083393097,
+      "learning_rate": 5e-06,
+      "loss": 0.1594,
+      "step": 10621
+    },
+    {
+      "epoch": 7.995483628152051,
+      "grad_norm": 0.16909976303577423,
+      "learning_rate": 5e-06,
+      "loss": 0.1085,
+      "step": 10622
+    },
+    {
+      "epoch": 7.996236356793376,
+      "grad_norm": 0.17772410809993744,
+      "learning_rate": 5e-06,
+      "loss": 0.0997,
+      "step": 10623
+    },
+    {
+      "epoch": 7.9969890854347,
+      "grad_norm": 0.19466081261634827,
+      "learning_rate": 5e-06,
+      "loss": 0.1458,
+      "step": 10624
+    },
+    {
+      "epoch": 7.997741814076026,
+      "grad_norm": 0.16064828634262085,
+      "learning_rate": 5e-06,
+      "loss": 0.1241,
+      "step": 10625
+    },
+    {
+      "epoch": 7.99849454271735,
+      "grad_norm": 0.1656101942062378,
+      "learning_rate": 5e-06,
+      "loss": 0.1478,
+      "step": 10626
+    },
+    {
+      "epoch": 7.9992472713586755,
+      "grad_norm": 0.1678565889596939,
+      "learning_rate": 5e-06,
+      "loss": 0.1415,
+      "step": 10627
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 0.18174228072166443,
+      "learning_rate": 5e-06,
+      "loss": 0.1606,
+      "step": 10628
+    },
+    {
+      "epoch": 8.000752728641325,
+      "grad_norm": 0.1844407021999359,
+      "learning_rate": 5e-06,
+      "loss": 0.242,
+      "step": 10629
+    },
+    {
+      "epoch": 8.001505457282649,
+      "grad_norm": 0.15887047350406647,
+      "learning_rate": 5e-06,
+      "loss": 0.1408,
+      "step": 10630
+    },
+    {
+      "epoch": 8.002258185923974,
+      "grad_norm": 0.21817681193351746,
+      "learning_rate": 5e-06,
+      "loss": 0.1291,
+      "step": 10631
+    },
+    {
+      "epoch": 8.0030109145653,
+      "grad_norm": 0.19823695719242096,
+      "learning_rate": 5e-06,
+      "loss": 0.1532,
+      "step": 10632
+    },
+    {
+      "epoch": 8.003763643206623,
+      "grad_norm": 0.1538822203874588,
+      "learning_rate": 5e-06,
+      "loss": 0.0516,
+      "step": 10633
+    },
+    {
+      "epoch": 8.004516371847949,
+      "grad_norm": 0.18516603112220764,
+      "learning_rate": 5e-06,
+      "loss": 0.1577,
+      "step": 10634
+    },
+    {
+      "epoch": 8.005269100489274,
+      "grad_norm": 0.16762830317020416,
+      "learning_rate": 5e-06,
+      "loss": 0.106,
+      "step": 10635
+    },
+    {
+      "epoch": 8.006021829130598,
+      "grad_norm": 0.18713317811489105,
+      "learning_rate": 5e-06,
+      "loss": 0.143,
+      "step": 10636
+    },
+    {
+      "epoch": 8.006774557771923,
+      "grad_norm": 0.18204712867736816,
+      "learning_rate": 5e-06,
+      "loss": 0.1571,
+      "step": 10637
+    },
+    {
+      "epoch": 8.007527286413248,
+      "grad_norm": 0.16423673927783966,
+      "learning_rate": 5e-06,
+      "loss": 0.1094,
+      "step": 10638
+    },
+    {
+      "epoch": 8.008280015054574,
+      "grad_norm": 0.1943371593952179,
+      "learning_rate": 5e-06,
+      "loss": 0.224,
+      "step": 10639
+    },
+    {
+      "epoch": 8.009032743695897,
+      "grad_norm": 0.16207876801490784,
+      "learning_rate": 5e-06,
+      "loss": 0.0688,
+      "step": 10640
+    },
+    {
+      "epoch": 8.009785472337223,
+      "grad_norm": 0.16787156462669373,
+      "learning_rate": 5e-06,
+      "loss": 0.1641,
+      "step": 10641
+    },
+    {
+      "epoch": 8.010538200978548,
+      "grad_norm": 0.1860729455947876,
+      "learning_rate": 5e-06,
+      "loss": 0.203,
+      "step": 10642
+    },
+    {
+      "epoch": 8.011290929619872,
+      "grad_norm": 0.20577116310596466,
+      "learning_rate": 5e-06,
+      "loss": 0.2234,
+      "step": 10643
+    },
+    {
+      "epoch": 8.012043658261197,
+      "grad_norm": 0.15283389389514923,
+      "learning_rate": 5e-06,
+      "loss": 0.1493,
+      "step": 10644
+    },
+    {
+      "epoch": 8.012796386902522,
+      "grad_norm": 0.1890982687473297,
+      "learning_rate": 5e-06,
+      "loss": 0.169,
+      "step": 10645
+    },
+    {
+      "epoch": 8.013549115543846,
+      "grad_norm": 0.193354070186615,
+      "learning_rate": 5e-06,
+      "loss": 0.1459,
+      "step": 10646
+    },
+    {
+      "epoch": 8.014301844185171,
+      "grad_norm": 0.1982319951057434,
+      "learning_rate": 5e-06,
+      "loss": 0.1138,
+      "step": 10647
+    },
+    {
+      "epoch": 8.015054572826497,
+      "grad_norm": 0.1419096738100052,
+      "learning_rate": 5e-06,
+      "loss": 0.1156,
+      "step": 10648
+    },
+    {
+      "epoch": 8.01580730146782,
+      "grad_norm": 0.1925135999917984,
+      "learning_rate": 5e-06,
+      "loss": 0.1846,
+      "step": 10649
+    },
+    {
+      "epoch": 8.016560030109146,
+      "grad_norm": 0.17908842861652374,
+      "learning_rate": 5e-06,
+      "loss": 0.1552,
+      "step": 10650
+    },
+    {
+      "epoch": 8.017312758750471,
+      "grad_norm": 0.20299533009529114,
+      "learning_rate": 5e-06,
+      "loss": 0.2341,
+      "step": 10651
+    },
+    {
+      "epoch": 8.018065487391794,
+      "grad_norm": 0.18140071630477905,
+      "learning_rate": 5e-06,
+      "loss": 0.1289,
+      "step": 10652
+    },
+    {
+      "epoch": 8.01881821603312,
+      "grad_norm": 0.19177433848381042,
+      "learning_rate": 5e-06,
+      "loss": 0.1654,
+      "step": 10653
+    },
+    {
+      "epoch": 8.019570944674445,
+      "grad_norm": 0.19179493188858032,
+      "learning_rate": 5e-06,
+      "loss": 0.2797,
+      "step": 10654
+    },
+    {
+      "epoch": 8.020323673315769,
+      "grad_norm": 0.16953332722187042,
+      "learning_rate": 5e-06,
+      "loss": 0.1536,
+      "step": 10655
+    },
+    {
+      "epoch": 8.021076401957094,
+      "grad_norm": 0.18465059995651245,
+      "learning_rate": 5e-06,
+      "loss": 0.1264,
+      "step": 10656
+    },
+    {
+      "epoch": 8.02182913059842,
+      "grad_norm": 0.14861585199832916,
+      "learning_rate": 5e-06,
+      "loss": 0.139,
+      "step": 10657
+    },
+    {
+      "epoch": 8.022581859239745,
+      "grad_norm": 0.18550367653369904,
+      "learning_rate": 5e-06,
+      "loss": 0.1168,
+      "step": 10658
+    },
+    {
+      "epoch": 8.023334587881068,
+      "grad_norm": 0.17864544689655304,
+      "learning_rate": 5e-06,
+      "loss": 0.1997,
+      "step": 10659
+    },
+    {
+      "epoch": 8.024087316522394,
+      "grad_norm": 0.15698134899139404,
+      "learning_rate": 5e-06,
+      "loss": 0.1696,
+      "step": 10660
+    },
+    {
+      "epoch": 8.02484004516372,
+      "grad_norm": 0.18193718791007996,
+      "learning_rate": 5e-06,
+      "loss": 0.117,
+      "step": 10661
+    },
+    {
+      "epoch": 8.025592773805043,
+      "grad_norm": 0.18827304244041443,
+      "learning_rate": 5e-06,
+      "loss": 0.2202,
+      "step": 10662
+    },
+    {
+      "epoch": 8.026345502446368,
+      "grad_norm": 0.18730641901493073,
+      "learning_rate": 5e-06,
+      "loss": 0.1608,
+      "step": 10663
+    },
+    {
+      "epoch": 8.027098231087693,
+      "grad_norm": 0.18850228190422058,
+      "learning_rate": 5e-06,
+      "loss": 0.1603,
+      "step": 10664
+    },
+    {
+      "epoch": 8.027850959729017,
+      "grad_norm": 0.15155206620693207,
+      "learning_rate": 5e-06,
+      "loss": 0.0905,
+      "step": 10665
+    },
+    {
+      "epoch": 8.028603688370342,
+      "grad_norm": 0.1664169728755951,
+      "learning_rate": 5e-06,
+      "loss": 0.1761,
+      "step": 10666
+    },
+    {
+      "epoch": 8.029356417011668,
+      "grad_norm": 0.18711531162261963,
+      "learning_rate": 5e-06,
+      "loss": 0.1638,
+      "step": 10667
+    },
+    {
+      "epoch": 8.030109145652991,
+      "grad_norm": 0.1682155728340149,
+      "learning_rate": 5e-06,
+      "loss": 0.1409,
+      "step": 10668
+    },
+    {
+      "epoch": 8.030861874294317,
+      "grad_norm": 0.17405566573143005,
+      "learning_rate": 5e-06,
+      "loss": 0.1231,
+      "step": 10669
+    },
+    {
+      "epoch": 8.031614602935642,
+      "grad_norm": 0.1669795662164688,
+      "learning_rate": 5e-06,
+      "loss": 0.1505,
+      "step": 10670
+    },
+    {
+      "epoch": 8.032367331576966,
+      "grad_norm": 0.17117708921432495,
+      "learning_rate": 5e-06,
+      "loss": 0.1529,
+      "step": 10671
+    },
+    {
+      "epoch": 8.033120060218291,
+      "grad_norm": 0.18405644595623016,
+      "learning_rate": 5e-06,
+      "loss": 0.184,
+      "step": 10672
+    },
+    {
+      "epoch": 8.033872788859616,
+      "grad_norm": 0.18388938903808594,
+      "learning_rate": 5e-06,
+      "loss": 0.2188,
+      "step": 10673
+    },
+    {
+      "epoch": 8.034625517500942,
+      "grad_norm": 0.15948596596717834,
+      "learning_rate": 5e-06,
+      "loss": 0.1587,
+      "step": 10674
+    },
+    {
+      "epoch": 8.035378246142265,
+      "grad_norm": 0.22103038430213928,
+      "learning_rate": 5e-06,
+      "loss": 0.1923,
+      "step": 10675
+    },
+    {
+      "epoch": 8.03613097478359,
+      "grad_norm": 0.15976698696613312,
+      "learning_rate": 5e-06,
+      "loss": 0.1823,
+      "step": 10676
+    },
+    {
+      "epoch": 8.036883703424916,
+      "grad_norm": 0.1774054765701294,
+      "learning_rate": 5e-06,
+      "loss": 0.2372,
+      "step": 10677
+    },
+    {
+      "epoch": 8.03763643206624,
+      "grad_norm": 0.1654359996318817,
+      "learning_rate": 5e-06,
+      "loss": 0.1903,
+      "step": 10678
+    },
+    {
+      "epoch": 8.038389160707565,
+      "grad_norm": 0.1709703803062439,
+      "learning_rate": 5e-06,
+      "loss": 0.1972,
+      "step": 10679
+    },
+    {
+      "epoch": 8.03914188934889,
+      "grad_norm": 0.149202361702919,
+      "learning_rate": 5e-06,
+      "loss": 0.0665,
+      "step": 10680
+    },
+    {
+      "epoch": 8.039894617990214,
+      "grad_norm": 0.18032830953598022,
+      "learning_rate": 5e-06,
+      "loss": 0.1579,
+      "step": 10681
+    },
+    {
+      "epoch": 8.04064734663154,
+      "grad_norm": 0.1681692898273468,
+      "learning_rate": 5e-06,
+      "loss": 0.1524,
+      "step": 10682
+    },
+    {
+      "epoch": 8.041400075272865,
+      "grad_norm": 0.1712620109319687,
+      "learning_rate": 5e-06,
+      "loss": 0.2011,
+      "step": 10683
+    },
+    {
+      "epoch": 8.042152803914188,
+      "grad_norm": 0.17087076604366302,
+      "learning_rate": 5e-06,
+      "loss": 0.1684,
+      "step": 10684
+    },
+    {
+      "epoch": 8.042905532555514,
+      "grad_norm": 0.18466046452522278,
+      "learning_rate": 5e-06,
+      "loss": 0.1427,
+      "step": 10685
+    },
+    {
+      "epoch": 8.043658261196839,
+      "grad_norm": 0.17822198569774628,
+      "learning_rate": 5e-06,
+      "loss": 0.1645,
+      "step": 10686
+    },
+    {
+      "epoch": 8.044410989838163,
+      "grad_norm": 0.18281406164169312,
+      "learning_rate": 5e-06,
+      "loss": 0.1383,
+      "step": 10687
+    },
+    {
+      "epoch": 8.045163718479488,
+      "grad_norm": 0.1915549486875534,
+      "learning_rate": 5e-06,
+      "loss": 0.1164,
+      "step": 10688
+    },
+    {
+      "epoch": 8.045916447120813,
+      "grad_norm": 0.17325779795646667,
+      "learning_rate": 5e-06,
+      "loss": 0.1618,
+      "step": 10689
+    },
+    {
+      "epoch": 8.046669175762137,
+      "grad_norm": 0.17526401579380035,
+      "learning_rate": 5e-06,
+      "loss": 0.1087,
+      "step": 10690
+    },
+    {
+      "epoch": 8.047421904403462,
+      "grad_norm": 0.1986013501882553,
+      "learning_rate": 5e-06,
+      "loss": 0.1459,
+      "step": 10691
+    },
+    {
+      "epoch": 8.048174633044788,
+      "grad_norm": 0.18960563838481903,
+      "learning_rate": 5e-06,
+      "loss": 0.2284,
+      "step": 10692
+    },
+    {
+      "epoch": 8.048927361686113,
+      "grad_norm": 0.1699245572090149,
+      "learning_rate": 5e-06,
+      "loss": 0.1613,
+      "step": 10693
+    },
+    {
+      "epoch": 8.049680090327437,
+      "grad_norm": 0.16296923160552979,
+      "learning_rate": 5e-06,
+      "loss": 0.16,
+      "step": 10694
+    },
+    {
+      "epoch": 8.050432818968762,
+      "grad_norm": 0.16481544077396393,
+      "learning_rate": 5e-06,
+      "loss": 0.1654,
+      "step": 10695
+    },
+    {
+      "epoch": 8.051185547610087,
+      "grad_norm": 0.21641308069229126,
+      "learning_rate": 5e-06,
+      "loss": 0.1634,
+      "step": 10696
+    },
+    {
+      "epoch": 8.051938276251411,
+      "grad_norm": 0.15854217112064362,
+      "learning_rate": 5e-06,
+      "loss": 0.1244,
+      "step": 10697
+    },
+    {
+      "epoch": 8.052691004892736,
+      "grad_norm": 0.17587560415267944,
+      "learning_rate": 5e-06,
+      "loss": 0.1071,
+      "step": 10698
+    },
+    {
+      "epoch": 8.053443733534062,
+      "grad_norm": 0.17923671007156372,
+      "learning_rate": 5e-06,
+      "loss": 0.1611,
+      "step": 10699
+    },
+    {
+      "epoch": 8.054196462175385,
+      "grad_norm": 0.17735326290130615,
+      "learning_rate": 5e-06,
+      "loss": 0.2493,
+      "step": 10700
+    },
+    {
+      "epoch": 8.05494919081671,
+      "grad_norm": 0.17546100914478302,
+      "learning_rate": 5e-06,
+      "loss": 0.1598,
+      "step": 10701
+    },
+    {
+      "epoch": 8.055701919458036,
+      "grad_norm": 0.1545703113079071,
+      "learning_rate": 5e-06,
+      "loss": 0.1479,
+      "step": 10702
+    },
+    {
+      "epoch": 8.05645464809936,
+      "grad_norm": 0.1609795093536377,
+      "learning_rate": 5e-06,
+      "loss": 0.1357,
+      "step": 10703
+    },
+    {
+      "epoch": 8.057207376740685,
+      "grad_norm": 0.16388747096061707,
+      "learning_rate": 5e-06,
+      "loss": 0.1536,
+      "step": 10704
+    },
+    {
+      "epoch": 8.05796010538201,
+      "grad_norm": 0.16111882030963898,
+      "learning_rate": 5e-06,
+      "loss": 0.1288,
+      "step": 10705
+    },
+    {
+      "epoch": 8.058712834023334,
+      "grad_norm": 0.17281420528888702,
+      "learning_rate": 5e-06,
+      "loss": 0.1676,
+      "step": 10706
+    },
+    {
+      "epoch": 8.05946556266466,
+      "grad_norm": 0.19774341583251953,
+      "learning_rate": 5e-06,
+      "loss": 0.1343,
+      "step": 10707
+    },
+    {
+      "epoch": 8.060218291305985,
+      "grad_norm": 0.16577792167663574,
+      "learning_rate": 5e-06,
+      "loss": 0.1878,
+      "step": 10708
+    },
+    {
+      "epoch": 8.060971019947308,
+      "grad_norm": 0.1640148162841797,
+      "learning_rate": 5e-06,
+      "loss": 0.1061,
+      "step": 10709
+    },
+    {
+      "epoch": 8.061723748588633,
+      "grad_norm": 0.1640256941318512,
+      "learning_rate": 5e-06,
+      "loss": 0.1576,
+      "step": 10710
+    },
+    {
+      "epoch": 8.062476477229959,
+      "grad_norm": 0.18218058347702026,
+      "learning_rate": 5e-06,
+      "loss": 0.1165,
+      "step": 10711
+    },
+    {
+      "epoch": 8.063229205871284,
+      "grad_norm": 0.18125255405902863,
+      "learning_rate": 5e-06,
+      "loss": 0.1788,
+      "step": 10712
+    },
+    {
+      "epoch": 8.063981934512608,
+      "grad_norm": 0.19308142364025116,
+      "learning_rate": 5e-06,
+      "loss": 0.1727,
+      "step": 10713
+    },
+    {
+      "epoch": 8.064734663153933,
+      "grad_norm": 0.1763375997543335,
+      "learning_rate": 5e-06,
+      "loss": 0.1211,
+      "step": 10714
+    },
+    {
+      "epoch": 8.065487391795259,
+      "grad_norm": 0.19433194398880005,
+      "learning_rate": 5e-06,
+      "loss": 0.1845,
+      "step": 10715
+    },
+    {
+      "epoch": 8.066240120436582,
+      "grad_norm": 0.18100354075431824,
+      "learning_rate": 5e-06,
+      "loss": 0.1331,
+      "step": 10716
+    },
+    {
+      "epoch": 8.066992849077907,
+      "grad_norm": 0.1634449064731598,
+      "learning_rate": 5e-06,
+      "loss": 0.1974,
+      "step": 10717
+    },
+    {
+      "epoch": 8.067745577719233,
+      "grad_norm": 0.1707880049943924,
+      "learning_rate": 5e-06,
+      "loss": 0.0896,
+      "step": 10718
+    },
+    {
+      "epoch": 8.068498306360556,
+      "grad_norm": 0.18306659162044525,
+      "learning_rate": 5e-06,
+      "loss": 0.1226,
+      "step": 10719
+    },
+    {
+      "epoch": 8.069251035001882,
+      "grad_norm": 0.17817042768001556,
+      "learning_rate": 5e-06,
+      "loss": 0.1447,
+      "step": 10720
+    },
+    {
+      "epoch": 8.070003763643207,
+      "grad_norm": 0.18481534719467163,
+      "learning_rate": 5e-06,
+      "loss": 0.1605,
+      "step": 10721
+    },
+    {
+      "epoch": 8.07075649228453,
+      "grad_norm": 0.18828417360782623,
+      "learning_rate": 5e-06,
+      "loss": 0.1833,
+      "step": 10722
+    },
+    {
+      "epoch": 8.071509220925856,
+      "grad_norm": 0.1719844788312912,
+      "learning_rate": 5e-06,
+      "loss": 0.1093,
+      "step": 10723
+    },
+    {
+      "epoch": 8.072261949567181,
+      "grad_norm": 0.18762044608592987,
+      "learning_rate": 5e-06,
+      "loss": 0.1458,
+      "step": 10724
+    },
+    {
+      "epoch": 8.073014678208505,
+      "grad_norm": 0.1716400682926178,
+      "learning_rate": 5e-06,
+      "loss": 0.2393,
+      "step": 10725
+    },
+    {
+      "epoch": 8.07376740684983,
+      "grad_norm": 0.1606491208076477,
+      "learning_rate": 5e-06,
+      "loss": 0.1153,
+      "step": 10726
+    },
+    {
+      "epoch": 8.074520135491156,
+      "grad_norm": 0.19190862774848938,
+      "learning_rate": 5e-06,
+      "loss": 0.2282,
+      "step": 10727
+    },
+    {
+      "epoch": 8.07527286413248,
+      "grad_norm": 0.17450731992721558,
+      "learning_rate": 5e-06,
+      "loss": 0.1209,
+      "step": 10728
+    },
+    {
+      "epoch": 8.076025592773805,
+      "grad_norm": 0.15961089730262756,
+      "learning_rate": 5e-06,
+      "loss": 0.1196,
+      "step": 10729
+    },
+    {
+      "epoch": 8.07677832141513,
+      "grad_norm": 0.1682020127773285,
+      "learning_rate": 5e-06,
+      "loss": 0.1961,
+      "step": 10730
+    },
+    {
+      "epoch": 8.077531050056455,
+      "grad_norm": 0.17712602019309998,
+      "learning_rate": 5e-06,
+      "loss": 0.1486,
+      "step": 10731
+    },
+    {
+      "epoch": 8.078283778697779,
+      "grad_norm": 0.1537756323814392,
+      "learning_rate": 5e-06,
+      "loss": 0.0609,
+      "step": 10732
+    },
+    {
+      "epoch": 8.079036507339104,
+      "grad_norm": 0.17843692004680634,
+      "learning_rate": 5e-06,
+      "loss": 0.1168,
+      "step": 10733
+    },
+    {
+      "epoch": 8.07978923598043,
+      "grad_norm": 0.1581287980079651,
+      "learning_rate": 5e-06,
+      "loss": 0.0729,
+      "step": 10734
+    },
+    {
+      "epoch": 8.080541964621753,
+      "grad_norm": 0.19811628758907318,
+      "learning_rate": 5e-06,
+      "loss": 0.1815,
+      "step": 10735
+    },
+    {
+      "epoch": 8.081294693263079,
+      "grad_norm": 0.16969351470470428,
+      "learning_rate": 5e-06,
+      "loss": 0.1155,
+      "step": 10736
+    },
+    {
+      "epoch": 8.082047421904404,
+      "grad_norm": 0.16764800250530243,
+      "learning_rate": 5e-06,
+      "loss": 0.1326,
+      "step": 10737
+    },
+    {
+      "epoch": 8.082800150545728,
+      "grad_norm": 0.17908219993114471,
+      "learning_rate": 5e-06,
+      "loss": 0.2155,
+      "step": 10738
+    },
+    {
+      "epoch": 8.083552879187053,
+      "grad_norm": 0.18137872219085693,
+      "learning_rate": 5e-06,
+      "loss": 0.2113,
+      "step": 10739
+    },
+    {
+      "epoch": 8.084305607828378,
+      "grad_norm": 0.1344897747039795,
+      "learning_rate": 5e-06,
+      "loss": 0.055,
+      "step": 10740
+    },
+    {
+      "epoch": 8.085058336469702,
+      "grad_norm": 0.16347843408584595,
+      "learning_rate": 5e-06,
+      "loss": 0.287,
+      "step": 10741
+    },
+    {
+      "epoch": 8.085811065111027,
+      "grad_norm": 0.17334505915641785,
+      "learning_rate": 5e-06,
+      "loss": 0.1445,
+      "step": 10742
+    },
+    {
+      "epoch": 8.086563793752353,
+      "grad_norm": 0.14747369289398193,
+      "learning_rate": 5e-06,
+      "loss": 0.162,
+      "step": 10743
+    },
+    {
+      "epoch": 8.087316522393676,
+      "grad_norm": 0.189052015542984,
+      "learning_rate": 5e-06,
+      "loss": 0.1664,
+      "step": 10744
+    },
+    {
+      "epoch": 8.088069251035002,
+      "grad_norm": 0.1883181929588318,
+      "learning_rate": 5e-06,
+      "loss": 0.1789,
+      "step": 10745
+    },
+    {
+      "epoch": 8.088821979676327,
+      "grad_norm": 0.16795703768730164,
+      "learning_rate": 5e-06,
+      "loss": 0.1271,
+      "step": 10746
+    },
+    {
+      "epoch": 8.089574708317652,
+      "grad_norm": 0.19345951080322266,
+      "learning_rate": 5e-06,
+      "loss": 0.2435,
+      "step": 10747
+    },
+    {
+      "epoch": 8.090327436958976,
+      "grad_norm": 0.17437128722667694,
+      "learning_rate": 5e-06,
+      "loss": 0.1362,
+      "step": 10748
+    },
+    {
+      "epoch": 8.091080165600301,
+      "grad_norm": 0.169234499335289,
+      "learning_rate": 5e-06,
+      "loss": 0.094,
+      "step": 10749
+    },
+    {
+      "epoch": 8.091832894241627,
+      "grad_norm": 0.19944234192371368,
+      "learning_rate": 5e-06,
+      "loss": 0.2167,
+      "step": 10750
+    },
+    {
+      "epoch": 8.09258562288295,
+      "grad_norm": 0.2026865929365158,
+      "learning_rate": 5e-06,
+      "loss": 0.2078,
+      "step": 10751
+    },
+    {
+      "epoch": 8.093338351524276,
+      "grad_norm": 0.1965821534395218,
+      "learning_rate": 5e-06,
+      "loss": 0.2944,
+      "step": 10752
+    },
+    {
+      "epoch": 8.094091080165601,
+      "grad_norm": 0.17559179663658142,
+      "learning_rate": 5e-06,
+      "loss": 0.1687,
+      "step": 10753
+    },
+    {
+      "epoch": 8.094843808806925,
+      "grad_norm": 0.19114050269126892,
+      "learning_rate": 5e-06,
+      "loss": 0.2081,
+      "step": 10754
+    },
+    {
+      "epoch": 8.09559653744825,
+      "grad_norm": 0.17163698375225067,
+      "learning_rate": 5e-06,
+      "loss": 0.1681,
+      "step": 10755
+    },
+    {
+      "epoch": 8.096349266089575,
+      "grad_norm": 0.18734736740589142,
+      "learning_rate": 5e-06,
+      "loss": 0.183,
+      "step": 10756
+    },
+    {
+      "epoch": 8.097101994730899,
+      "grad_norm": 0.17331726849079132,
+      "learning_rate": 5e-06,
+      "loss": 0.2045,
+      "step": 10757
+    },
+    {
+      "epoch": 8.097854723372224,
+      "grad_norm": 0.195328488945961,
+      "learning_rate": 5e-06,
+      "loss": 0.2165,
+      "step": 10758
+    },
+    {
+      "epoch": 8.09860745201355,
+      "grad_norm": 0.21146562695503235,
+      "learning_rate": 5e-06,
+      "loss": 0.214,
+      "step": 10759
+    },
+    {
+      "epoch": 8.099360180654873,
+      "grad_norm": 0.17612530291080475,
+      "learning_rate": 5e-06,
+      "loss": 0.1371,
+      "step": 10760
+    },
+    {
+      "epoch": 8.100112909296199,
+      "grad_norm": 0.18318124115467072,
+      "learning_rate": 5e-06,
+      "loss": 0.2229,
+      "step": 10761
+    },
+    {
+      "epoch": 8.100865637937524,
+      "grad_norm": 0.17089836299419403,
+      "learning_rate": 5e-06,
+      "loss": 0.1143,
+      "step": 10762
+    },
+    {
+      "epoch": 8.101618366578847,
+      "grad_norm": 0.18303056061267853,
+      "learning_rate": 5e-06,
+      "loss": 0.1286,
+      "step": 10763
+    },
+    {
+      "epoch": 8.102371095220173,
+      "grad_norm": 0.17785151302814484,
+      "learning_rate": 5e-06,
+      "loss": 0.0996,
+      "step": 10764
+    },
+    {
+      "epoch": 8.103123823861498,
+      "grad_norm": 0.1483810544013977,
+      "learning_rate": 5e-06,
+      "loss": 0.083,
+      "step": 10765
+    },
+    {
+      "epoch": 8.103876552502824,
+      "grad_norm": 0.16529348492622375,
+      "learning_rate": 5e-06,
+      "loss": 0.1163,
+      "step": 10766
+    },
+    {
+      "epoch": 8.104629281144147,
+      "grad_norm": 0.16549812257289886,
+      "learning_rate": 5e-06,
+      "loss": 0.1311,
+      "step": 10767
+    },
+    {
+      "epoch": 8.105382009785473,
+      "grad_norm": 0.1952773928642273,
+      "learning_rate": 5e-06,
+      "loss": 0.1691,
+      "step": 10768
+    },
+    {
+      "epoch": 8.106134738426798,
+      "grad_norm": 0.18981978297233582,
+      "learning_rate": 5e-06,
+      "loss": 0.1384,
+      "step": 10769
+    },
+    {
+      "epoch": 8.106887467068121,
+      "grad_norm": 0.16729788482189178,
+      "learning_rate": 5e-06,
+      "loss": 0.1349,
+      "step": 10770
+    },
+    {
+      "epoch": 8.107640195709447,
+      "grad_norm": 0.17599791288375854,
+      "learning_rate": 5e-06,
+      "loss": 0.1024,
+      "step": 10771
+    },
+    {
+      "epoch": 8.108392924350772,
+      "grad_norm": 0.17743521928787231,
+      "learning_rate": 5e-06,
+      "loss": 0.116,
+      "step": 10772
+    },
+    {
+      "epoch": 8.109145652992096,
+      "grad_norm": 0.14970463514328003,
+      "learning_rate": 5e-06,
+      "loss": 0.1024,
+      "step": 10773
+    },
+    {
+      "epoch": 8.109898381633421,
+      "grad_norm": 0.1884056031703949,
+      "learning_rate": 5e-06,
+      "loss": 0.1426,
+      "step": 10774
+    },
+    {
+      "epoch": 8.110651110274747,
+      "grad_norm": 0.1625315397977829,
+      "learning_rate": 5e-06,
+      "loss": 0.1709,
+      "step": 10775
+    },
+    {
+      "epoch": 8.11140383891607,
+      "grad_norm": 0.1951439380645752,
+      "learning_rate": 5e-06,
+      "loss": 0.2328,
+      "step": 10776
+    },
+    {
+      "epoch": 8.112156567557395,
+      "grad_norm": 0.17812028527259827,
+      "learning_rate": 5e-06,
+      "loss": 0.1463,
+      "step": 10777
+    },
+    {
+      "epoch": 8.11290929619872,
+      "grad_norm": 0.18914024531841278,
+      "learning_rate": 5e-06,
+      "loss": 0.3004,
+      "step": 10778
+    },
+    {
+      "epoch": 8.113662024840044,
+      "grad_norm": 0.1666121631860733,
+      "learning_rate": 5e-06,
+      "loss": 0.1407,
+      "step": 10779
+    },
+    {
+      "epoch": 8.11441475348137,
+      "grad_norm": 0.16638872027397156,
+      "learning_rate": 5e-06,
+      "loss": 0.1419,
+      "step": 10780
+    },
+    {
+      "epoch": 8.115167482122695,
+      "grad_norm": 0.1907833367586136,
+      "learning_rate": 5e-06,
+      "loss": 0.1565,
+      "step": 10781
+    },
+    {
+      "epoch": 8.115920210764019,
+      "grad_norm": 0.18581214547157288,
+      "learning_rate": 5e-06,
+      "loss": 0.1282,
+      "step": 10782
+    },
+    {
+      "epoch": 8.116672939405344,
+      "grad_norm": 0.16492260992527008,
+      "learning_rate": 5e-06,
+      "loss": 0.2021,
+      "step": 10783
+    },
+    {
+      "epoch": 8.11742566804667,
+      "grad_norm": 0.1878097504377365,
+      "learning_rate": 5e-06,
+      "loss": 0.1995,
+      "step": 10784
+    },
+    {
+      "epoch": 8.118178396687995,
+      "grad_norm": 0.1787305772304535,
+      "learning_rate": 5e-06,
+      "loss": 0.1682,
+      "step": 10785
+    },
+    {
+      "epoch": 8.118931125329318,
+      "grad_norm": 0.1793331503868103,
+      "learning_rate": 5e-06,
+      "loss": 0.2236,
+      "step": 10786
+    },
+    {
+      "epoch": 8.119683853970644,
+      "grad_norm": 0.1731233447790146,
+      "learning_rate": 5e-06,
+      "loss": 0.1791,
+      "step": 10787
+    },
+    {
+      "epoch": 8.120436582611969,
+      "grad_norm": 0.17654608190059662,
+      "learning_rate": 5e-06,
+      "loss": 0.1693,
+      "step": 10788
+    },
+    {
+      "epoch": 8.121189311253293,
+      "grad_norm": 0.16943086683750153,
+      "learning_rate": 5e-06,
+      "loss": 0.1595,
+      "step": 10789
+    },
+    {
+      "epoch": 8.121942039894618,
+      "grad_norm": 0.1883063018321991,
+      "learning_rate": 5e-06,
+      "loss": 0.1677,
+      "step": 10790
+    },
+    {
+      "epoch": 8.122694768535943,
+      "grad_norm": 0.15545739233493805,
+      "learning_rate": 5e-06,
+      "loss": 0.1408,
+      "step": 10791
+    },
+    {
+      "epoch": 8.123447497177267,
+      "grad_norm": 0.16020528972148895,
+      "learning_rate": 5e-06,
+      "loss": 0.111,
+      "step": 10792
+    },
+    {
+      "epoch": 8.124200225818592,
+      "grad_norm": 0.14879874885082245,
+      "learning_rate": 5e-06,
+      "loss": 0.1496,
+      "step": 10793
+    },
+    {
+      "epoch": 8.124952954459918,
+      "grad_norm": 0.1610066145658493,
+      "learning_rate": 5e-06,
+      "loss": 0.0992,
+      "step": 10794
+    },
+    {
+      "epoch": 8.125705683101241,
+      "grad_norm": 0.1720704734325409,
+      "learning_rate": 5e-06,
+      "loss": 0.1153,
+      "step": 10795
+    },
+    {
+      "epoch": 8.126458411742567,
+      "grad_norm": 0.19761766493320465,
+      "learning_rate": 5e-06,
+      "loss": 0.1219,
+      "step": 10796
+    },
+    {
+      "epoch": 8.127211140383892,
+      "grad_norm": 0.1575680524110794,
+      "learning_rate": 5e-06,
+      "loss": 0.1147,
+      "step": 10797
+    },
+    {
+      "epoch": 8.127963869025216,
+      "grad_norm": 0.19569289684295654,
+      "learning_rate": 5e-06,
+      "loss": 0.2191,
+      "step": 10798
+    },
+    {
+      "epoch": 8.128716597666541,
+      "grad_norm": 0.16167078912258148,
+      "learning_rate": 5e-06,
+      "loss": 0.0994,
+      "step": 10799
+    },
+    {
+      "epoch": 8.129469326307866,
+      "grad_norm": 0.18629218637943268,
+      "learning_rate": 5e-06,
+      "loss": 0.1466,
+      "step": 10800
+    },
+    {
+      "epoch": 8.129469326307866,
+      "eval_loss": 0.196060448884964,
+      "eval_runtime": 452.5932,
+      "eval_samples_per_second": 21.271,
+      "eval_steps_per_second": 0.665,
+      "step": 10800
+    },
+    {
+      "epoch": 8.13022205494919,
+      "grad_norm": 0.18631887435913086,
+      "learning_rate": 5e-06,
+      "loss": 0.2238,
+      "step": 10801
+    },
+    {
+      "epoch": 8.130974783590515,
+      "grad_norm": 0.17437227070331573,
+      "learning_rate": 5e-06,
+      "loss": 0.1421,
+      "step": 10802
+    },
+    {
+      "epoch": 8.13172751223184,
+      "grad_norm": 0.14336897432804108,
+      "learning_rate": 5e-06,
+      "loss": 0.133,
+      "step": 10803
+    },
+    {
+      "epoch": 8.132480240873166,
+      "grad_norm": 0.15870413184165955,
+      "learning_rate": 5e-06,
+      "loss": 0.1193,
+      "step": 10804
+    },
+    {
+      "epoch": 8.13323296951449,
+      "grad_norm": 0.17215222120285034,
+      "learning_rate": 5e-06,
+      "loss": 0.1195,
+      "step": 10805
+    },
+    {
+      "epoch": 8.133985698155815,
+      "grad_norm": 0.2129744291305542,
+      "learning_rate": 5e-06,
+      "loss": 0.147,
+      "step": 10806
+    },
+    {
+      "epoch": 8.13473842679714,
+      "grad_norm": 0.1763717532157898,
+      "learning_rate": 5e-06,
+      "loss": 0.149,
+      "step": 10807
+    },
+    {
+      "epoch": 8.135491155438464,
+      "grad_norm": 0.16883905231952667,
+      "learning_rate": 5e-06,
+      "loss": 0.1684,
+      "step": 10808
+    },
+    {
+      "epoch": 8.13624388407979,
+      "grad_norm": 0.1961241513490677,
+      "learning_rate": 5e-06,
+      "loss": 0.1877,
+      "step": 10809
+    },
+    {
+      "epoch": 8.136996612721115,
+      "grad_norm": 0.18482686579227448,
+      "learning_rate": 5e-06,
+      "loss": 0.1836,
+      "step": 10810
+    },
+    {
+      "epoch": 8.137749341362438,
+      "grad_norm": 0.15650351345539093,
+      "learning_rate": 5e-06,
+      "loss": 0.1142,
+      "step": 10811
+    },
+    {
+      "epoch": 8.138502070003764,
+      "grad_norm": 0.17809109389781952,
+      "learning_rate": 5e-06,
+      "loss": 0.125,
+      "step": 10812
+    },
+    {
+      "epoch": 8.139254798645089,
+      "grad_norm": 0.16597093641757965,
+      "learning_rate": 5e-06,
+      "loss": 0.0926,
+      "step": 10813
+    },
+    {
+      "epoch": 8.140007527286413,
+      "grad_norm": 0.1846981942653656,
+      "learning_rate": 5e-06,
+      "loss": 0.1957,
+      "step": 10814
+    },
+    {
+      "epoch": 8.140760255927738,
+      "grad_norm": 0.17407666146755219,
+      "learning_rate": 5e-06,
+      "loss": 0.1608,
+      "step": 10815
+    },
+    {
+      "epoch": 8.141512984569063,
+      "grad_norm": 0.17276622354984283,
+      "learning_rate": 5e-06,
+      "loss": 0.1831,
+      "step": 10816
+    },
+    {
+      "epoch": 8.142265713210387,
+      "grad_norm": 0.1669469028711319,
+      "learning_rate": 5e-06,
+      "loss": 0.1336,
+      "step": 10817
+    },
+    {
+      "epoch": 8.143018441851712,
+      "grad_norm": 0.19390259683132172,
+      "learning_rate": 5e-06,
+      "loss": 0.1357,
+      "step": 10818
+    },
+    {
+      "epoch": 8.143771170493038,
+      "grad_norm": 0.1624075323343277,
+      "learning_rate": 5e-06,
+      "loss": 0.1683,
+      "step": 10819
+    },
+    {
+      "epoch": 8.144523899134363,
+      "grad_norm": 0.19062034785747528,
+      "learning_rate": 5e-06,
+      "loss": 0.2031,
+      "step": 10820
+    },
+    {
+      "epoch": 8.145276627775687,
+      "grad_norm": 0.16741545498371124,
+      "learning_rate": 5e-06,
+      "loss": 0.1202,
+      "step": 10821
+    },
+    {
+      "epoch": 8.146029356417012,
+      "grad_norm": 0.14864568412303925,
+      "learning_rate": 5e-06,
+      "loss": 0.1937,
+      "step": 10822
+    },
+    {
+      "epoch": 8.146782085058337,
+      "grad_norm": 0.17304150760173798,
+      "learning_rate": 5e-06,
+      "loss": 0.1276,
+      "step": 10823
+    },
+    {
+      "epoch": 8.14753481369966,
+      "grad_norm": 0.1566981077194214,
+      "learning_rate": 5e-06,
+      "loss": 0.1205,
+      "step": 10824
+    },
+    {
+      "epoch": 8.148287542340986,
+      "grad_norm": 0.1867920160293579,
+      "learning_rate": 5e-06,
+      "loss": 0.1474,
+      "step": 10825
+    },
+    {
+      "epoch": 8.149040270982312,
+      "grad_norm": 0.1694052666425705,
+      "learning_rate": 5e-06,
+      "loss": 0.0861,
+      "step": 10826
+    },
+    {
+      "epoch": 8.149792999623635,
+      "grad_norm": 0.16226324439048767,
+      "learning_rate": 5e-06,
+      "loss": 0.1075,
+      "step": 10827
+    },
+    {
+      "epoch": 8.15054572826496,
+      "grad_norm": 0.161004438996315,
+      "learning_rate": 5e-06,
+      "loss": 0.1518,
+      "step": 10828
+    },
+    {
+      "epoch": 8.151298456906286,
+      "grad_norm": 0.18044564127922058,
+      "learning_rate": 5e-06,
+      "loss": 0.1944,
+      "step": 10829
+    },
+    {
+      "epoch": 8.15205118554761,
+      "grad_norm": 0.2000689059495926,
+      "learning_rate": 5e-06,
+      "loss": 0.189,
+      "step": 10830
+    },
+    {
+      "epoch": 8.152803914188935,
+      "grad_norm": 0.16476072371006012,
+      "learning_rate": 5e-06,
+      "loss": 0.1173,
+      "step": 10831
+    },
+    {
+      "epoch": 8.15355664283026,
+      "grad_norm": 0.18060484528541565,
+      "learning_rate": 5e-06,
+      "loss": 0.137,
+      "step": 10832
+    },
+    {
+      "epoch": 8.154309371471584,
+      "grad_norm": 0.18942871689796448,
+      "learning_rate": 5e-06,
+      "loss": 0.2022,
+      "step": 10833
+    },
+    {
+      "epoch": 8.155062100112909,
+      "grad_norm": 0.16734462976455688,
+      "learning_rate": 5e-06,
+      "loss": 0.1512,
+      "step": 10834
+    },
+    {
+      "epoch": 8.155814828754234,
+      "grad_norm": 0.17190393805503845,
+      "learning_rate": 5e-06,
+      "loss": 0.1393,
+      "step": 10835
+    },
+    {
+      "epoch": 8.156567557395558,
+      "grad_norm": 0.18805740773677826,
+      "learning_rate": 5e-06,
+      "loss": 0.1501,
+      "step": 10836
+    },
+    {
+      "epoch": 8.157320286036883,
+      "grad_norm": 0.17832568287849426,
+      "learning_rate": 5e-06,
+      "loss": 0.1687,
+      "step": 10837
+    },
+    {
+      "epoch": 8.158073014678209,
+      "grad_norm": 0.1723102629184723,
+      "learning_rate": 5e-06,
+      "loss": 0.1674,
+      "step": 10838
+    },
+    {
+      "epoch": 8.158825743319534,
+      "grad_norm": 0.1463172733783722,
+      "learning_rate": 5e-06,
+      "loss": 0.1003,
+      "step": 10839
+    },
+    {
+      "epoch": 8.159578471960858,
+      "grad_norm": 0.1719013750553131,
+      "learning_rate": 5e-06,
+      "loss": 0.0838,
+      "step": 10840
+    },
+    {
+      "epoch": 8.160331200602183,
+      "grad_norm": 0.1541757583618164,
+      "learning_rate": 5e-06,
+      "loss": 0.1534,
+      "step": 10841
+    },
+    {
+      "epoch": 8.161083929243508,
+      "grad_norm": 0.17780712246894836,
+      "learning_rate": 5e-06,
+      "loss": 0.1381,
+      "step": 10842
+    },
+    {
+      "epoch": 8.161836657884832,
+      "grad_norm": 0.17498205602169037,
+      "learning_rate": 5e-06,
+      "loss": 0.1778,
+      "step": 10843
+    },
+    {
+      "epoch": 8.162589386526157,
+      "grad_norm": 0.18235526978969574,
+      "learning_rate": 5e-06,
+      "loss": 0.1838,
+      "step": 10844
+    },
+    {
+      "epoch": 8.163342115167483,
+      "grad_norm": 0.18820124864578247,
+      "learning_rate": 5e-06,
+      "loss": 0.1799,
+      "step": 10845
+    },
+    {
+      "epoch": 8.164094843808806,
+      "grad_norm": 0.38464921712875366,
+      "learning_rate": 5e-06,
+      "loss": 0.1975,
+      "step": 10846
+    },
+    {
+      "epoch": 8.164847572450132,
+      "grad_norm": 0.18513062596321106,
+      "learning_rate": 5e-06,
+      "loss": 0.284,
+      "step": 10847
+    },
+    {
+      "epoch": 8.165600301091457,
+      "grad_norm": 0.180914968252182,
+      "learning_rate": 5e-06,
+      "loss": 0.1743,
+      "step": 10848
+    },
+    {
+      "epoch": 8.16635302973278,
+      "grad_norm": 0.1792019158601761,
+      "learning_rate": 5e-06,
+      "loss": 0.2342,
+      "step": 10849
+    },
+    {
+      "epoch": 8.167105758374106,
+      "grad_norm": 0.18300962448120117,
+      "learning_rate": 5e-06,
+      "loss": 0.1216,
+      "step": 10850
+    },
+    {
+      "epoch": 8.167858487015431,
+      "grad_norm": 0.1705649048089981,
+      "learning_rate": 5e-06,
+      "loss": 0.1649,
+      "step": 10851
+    },
+    {
+      "epoch": 8.168611215656755,
+      "grad_norm": 0.18503698706626892,
+      "learning_rate": 5e-06,
+      "loss": 0.2234,
+      "step": 10852
+    },
+    {
+      "epoch": 8.16936394429808,
+      "grad_norm": 0.17623291909694672,
+      "learning_rate": 5e-06,
+      "loss": 0.0968,
+      "step": 10853
+    },
+    {
+      "epoch": 8.170116672939406,
+      "grad_norm": 0.16241592168807983,
+      "learning_rate": 5e-06,
+      "loss": 0.1557,
+      "step": 10854
+    },
+    {
+      "epoch": 8.17086940158073,
+      "grad_norm": 0.15707921981811523,
+      "learning_rate": 5e-06,
+      "loss": 0.1007,
+      "step": 10855
+    },
+    {
+      "epoch": 8.171622130222055,
+      "grad_norm": 0.17524369060993195,
+      "learning_rate": 5e-06,
+      "loss": 0.1916,
+      "step": 10856
+    },
+    {
+      "epoch": 8.17237485886338,
+      "grad_norm": 0.1992543339729309,
+      "learning_rate": 5e-06,
+      "loss": 0.218,
+      "step": 10857
+    },
+    {
+      "epoch": 8.173127587504705,
+      "grad_norm": 0.1705479919910431,
+      "learning_rate": 5e-06,
+      "loss": 0.1511,
+      "step": 10858
+    },
+    {
+      "epoch": 8.173880316146029,
+      "grad_norm": 0.1619301289319992,
+      "learning_rate": 5e-06,
+      "loss": 0.1072,
+      "step": 10859
+    },
+    {
+      "epoch": 8.174633044787354,
+      "grad_norm": 0.17832180857658386,
+      "learning_rate": 5e-06,
+      "loss": 0.1005,
+      "step": 10860
+    },
+    {
+      "epoch": 8.17538577342868,
+      "grad_norm": 0.16439136862754822,
+      "learning_rate": 5e-06,
+      "loss": 0.1215,
+      "step": 10861
+    },
+    {
+      "epoch": 8.176138502070003,
+      "grad_norm": 0.17468136548995972,
+      "learning_rate": 5e-06,
+      "loss": 0.1186,
+      "step": 10862
+    },
+    {
+      "epoch": 8.176891230711329,
+      "grad_norm": 0.1962159425020218,
+      "learning_rate": 5e-06,
+      "loss": 0.1714,
+      "step": 10863
+    },
+    {
+      "epoch": 8.177643959352654,
+      "grad_norm": 0.14799098670482635,
+      "learning_rate": 5e-06,
+      "loss": 0.0663,
+      "step": 10864
+    },
+    {
+      "epoch": 8.178396687993978,
+      "grad_norm": 0.15644697844982147,
+      "learning_rate": 5e-06,
+      "loss": 0.1771,
+      "step": 10865
+    },
+    {
+      "epoch": 8.179149416635303,
+      "grad_norm": 0.1915024369955063,
+      "learning_rate": 5e-06,
+      "loss": 0.1701,
+      "step": 10866
+    },
+    {
+      "epoch": 8.179902145276628,
+      "grad_norm": 0.1742694228887558,
+      "learning_rate": 5e-06,
+      "loss": 0.1385,
+      "step": 10867
+    },
+    {
+      "epoch": 8.180654873917952,
+      "grad_norm": 0.1782456636428833,
+      "learning_rate": 5e-06,
+      "loss": 0.1087,
+      "step": 10868
+    },
+    {
+      "epoch": 8.181407602559277,
+      "grad_norm": 0.18547964096069336,
+      "learning_rate": 5e-06,
+      "loss": 0.1748,
+      "step": 10869
+    },
+    {
+      "epoch": 8.182160331200603,
+      "grad_norm": 0.17726488411426544,
+      "learning_rate": 5e-06,
+      "loss": 0.2047,
+      "step": 10870
+    },
+    {
+      "epoch": 8.182913059841926,
+      "grad_norm": 0.19256168603897095,
+      "learning_rate": 5e-06,
+      "loss": 0.192,
+      "step": 10871
+    },
+    {
+      "epoch": 8.183665788483252,
+      "grad_norm": 0.20020323991775513,
+      "learning_rate": 5e-06,
+      "loss": 0.1707,
+      "step": 10872
+    },
+    {
+      "epoch": 8.184418517124577,
+      "grad_norm": 0.16664308309555054,
+      "learning_rate": 5e-06,
+      "loss": 0.1892,
+      "step": 10873
+    },
+    {
+      "epoch": 8.1851712457659,
+      "grad_norm": 0.18491105735301971,
+      "learning_rate": 5e-06,
+      "loss": 0.1671,
+      "step": 10874
+    },
+    {
+      "epoch": 8.185923974407226,
+      "grad_norm": 0.15934066474437714,
+      "learning_rate": 5e-06,
+      "loss": 0.1187,
+      "step": 10875
+    },
+    {
+      "epoch": 8.186676703048551,
+      "grad_norm": 0.1674882024526596,
+      "learning_rate": 5e-06,
+      "loss": 0.154,
+      "step": 10876
+    },
+    {
+      "epoch": 8.187429431689877,
+      "grad_norm": 0.18928243219852448,
+      "learning_rate": 5e-06,
+      "loss": 0.108,
+      "step": 10877
+    },
+    {
+      "epoch": 8.1881821603312,
+      "grad_norm": 0.17549735307693481,
+      "learning_rate": 5e-06,
+      "loss": 0.1218,
+      "step": 10878
+    },
+    {
+      "epoch": 8.188934888972526,
+      "grad_norm": 0.16252374649047852,
+      "learning_rate": 5e-06,
+      "loss": 0.0858,
+      "step": 10879
+    },
+    {
+      "epoch": 8.189687617613851,
+      "grad_norm": 0.197111576795578,
+      "learning_rate": 5e-06,
+      "loss": 0.1841,
+      "step": 10880
+    },
+    {
+      "epoch": 8.190440346255174,
+      "grad_norm": 0.1651911586523056,
+      "learning_rate": 5e-06,
+      "loss": 0.1653,
+      "step": 10881
+    },
+    {
+      "epoch": 8.1911930748965,
+      "grad_norm": 0.17212766408920288,
+      "learning_rate": 5e-06,
+      "loss": 0.1588,
+      "step": 10882
+    },
+    {
+      "epoch": 8.191945803537825,
+      "grad_norm": 0.1623649150133133,
+      "learning_rate": 5e-06,
+      "loss": 0.1712,
+      "step": 10883
+    },
+    {
+      "epoch": 8.192698532179149,
+      "grad_norm": 0.16035649180412292,
+      "learning_rate": 5e-06,
+      "loss": 0.1405,
+      "step": 10884
+    },
+    {
+      "epoch": 8.193451260820474,
+      "grad_norm": 0.180393248796463,
+      "learning_rate": 5e-06,
+      "loss": 0.1579,
+      "step": 10885
+    },
+    {
+      "epoch": 8.1942039894618,
+      "grad_norm": 0.17022670805454254,
+      "learning_rate": 5e-06,
+      "loss": 0.0991,
+      "step": 10886
+    },
+    {
+      "epoch": 8.194956718103123,
+      "grad_norm": 0.17003504931926727,
+      "learning_rate": 5e-06,
+      "loss": 0.1945,
+      "step": 10887
+    },
+    {
+      "epoch": 8.195709446744448,
+      "grad_norm": 0.17917050421237946,
+      "learning_rate": 5e-06,
+      "loss": 0.1979,
+      "step": 10888
+    },
+    {
+      "epoch": 8.196462175385774,
+      "grad_norm": 0.1656339019536972,
+      "learning_rate": 5e-06,
+      "loss": 0.123,
+      "step": 10889
+    },
+    {
+      "epoch": 8.197214904027097,
+      "grad_norm": 0.20602059364318848,
+      "learning_rate": 5e-06,
+      "loss": 0.1349,
+      "step": 10890
+    },
+    {
+      "epoch": 8.197967632668423,
+      "grad_norm": 0.1751329004764557,
+      "learning_rate": 5e-06,
+      "loss": 0.1715,
+      "step": 10891
+    },
+    {
+      "epoch": 8.198720361309748,
+      "grad_norm": 0.1820261925458908,
+      "learning_rate": 5e-06,
+      "loss": 0.172,
+      "step": 10892
+    },
+    {
+      "epoch": 8.199473089951073,
+      "grad_norm": 0.17554132640361786,
+      "learning_rate": 5e-06,
+      "loss": 0.1599,
+      "step": 10893
+    },
+    {
+      "epoch": 8.200225818592397,
+      "grad_norm": 0.16628657281398773,
+      "learning_rate": 5e-06,
+      "loss": 0.1601,
+      "step": 10894
+    },
+    {
+      "epoch": 8.200978547233722,
+      "grad_norm": 0.19929681718349457,
+      "learning_rate": 5e-06,
+      "loss": 0.233,
+      "step": 10895
+    },
+    {
+      "epoch": 8.201731275875048,
+      "grad_norm": 0.16984374821186066,
+      "learning_rate": 5e-06,
+      "loss": 0.1325,
+      "step": 10896
+    },
+    {
+      "epoch": 8.202484004516371,
+      "grad_norm": 0.1862315982580185,
+      "learning_rate": 5e-06,
+      "loss": 0.1433,
+      "step": 10897
+    },
+    {
+      "epoch": 8.203236733157697,
+      "grad_norm": 0.16899442672729492,
+      "learning_rate": 5e-06,
+      "loss": 0.111,
+      "step": 10898
+    },
+    {
+      "epoch": 8.203989461799022,
+      "grad_norm": 0.1592492014169693,
+      "learning_rate": 5e-06,
+      "loss": 0.16,
+      "step": 10899
+    },
+    {
+      "epoch": 8.204742190440346,
+      "grad_norm": 0.1587124913930893,
+      "learning_rate": 5e-06,
+      "loss": 0.1716,
+      "step": 10900
+    },
+    {
+      "epoch": 8.205494919081671,
+      "grad_norm": 0.16246333718299866,
+      "learning_rate": 5e-06,
+      "loss": 0.1085,
+      "step": 10901
+    },
+    {
+      "epoch": 8.206247647722996,
+      "grad_norm": 0.15986712276935577,
+      "learning_rate": 5e-06,
+      "loss": 0.0922,
+      "step": 10902
+    },
+    {
+      "epoch": 8.20700037636432,
+      "grad_norm": 0.161284938454628,
+      "learning_rate": 5e-06,
+      "loss": 0.0677,
+      "step": 10903
+    },
+    {
+      "epoch": 8.207753105005645,
+      "grad_norm": 0.1963070034980774,
+      "learning_rate": 5e-06,
+      "loss": 0.1123,
+      "step": 10904
+    },
+    {
+      "epoch": 8.20850583364697,
+      "grad_norm": 0.18046998977661133,
+      "learning_rate": 5e-06,
+      "loss": 0.1277,
+      "step": 10905
+    },
+    {
+      "epoch": 8.209258562288294,
+      "grad_norm": 0.17406493425369263,
+      "learning_rate": 5e-06,
+      "loss": 0.1665,
+      "step": 10906
+    },
+    {
+      "epoch": 8.21001129092962,
+      "grad_norm": 0.17353889346122742,
+      "learning_rate": 5e-06,
+      "loss": 0.2438,
+      "step": 10907
+    },
+    {
+      "epoch": 8.210764019570945,
+      "grad_norm": 0.19513355195522308,
+      "learning_rate": 5e-06,
+      "loss": 0.1876,
+      "step": 10908
+    },
+    {
+      "epoch": 8.211516748212269,
+      "grad_norm": 0.17707398533821106,
+      "learning_rate": 5e-06,
+      "loss": 0.1329,
+      "step": 10909
+    },
+    {
+      "epoch": 8.212269476853594,
+      "grad_norm": 0.18105575442314148,
+      "learning_rate": 5e-06,
+      "loss": 0.1417,
+      "step": 10910
+    },
+    {
+      "epoch": 8.21302220549492,
+      "grad_norm": 0.20567820966243744,
+      "learning_rate": 5e-06,
+      "loss": 0.1751,
+      "step": 10911
+    },
+    {
+      "epoch": 8.213774934136245,
+      "grad_norm": 0.16047224402427673,
+      "learning_rate": 5e-06,
+      "loss": 0.0929,
+      "step": 10912
+    },
+    {
+      "epoch": 8.214527662777568,
+      "grad_norm": 0.1790238618850708,
+      "learning_rate": 5e-06,
+      "loss": 0.1812,
+      "step": 10913
+    },
+    {
+      "epoch": 8.215280391418894,
+      "grad_norm": 0.16848281025886536,
+      "learning_rate": 5e-06,
+      "loss": 0.1783,
+      "step": 10914
+    },
+    {
+      "epoch": 8.216033120060219,
+      "grad_norm": 0.15376292169094086,
+      "learning_rate": 5e-06,
+      "loss": 0.1797,
+      "step": 10915
+    },
+    {
+      "epoch": 8.216785848701543,
+      "grad_norm": 0.17515933513641357,
+      "learning_rate": 5e-06,
+      "loss": 0.2145,
+      "step": 10916
+    },
+    {
+      "epoch": 8.217538577342868,
+      "grad_norm": 0.1803903877735138,
+      "learning_rate": 5e-06,
+      "loss": 0.1585,
+      "step": 10917
+    },
+    {
+      "epoch": 8.218291305984193,
+      "grad_norm": 0.17155683040618896,
+      "learning_rate": 5e-06,
+      "loss": 0.1553,
+      "step": 10918
+    },
+    {
+      "epoch": 8.219044034625517,
+      "grad_norm": 0.13111692667007446,
+      "learning_rate": 5e-06,
+      "loss": 0.1042,
+      "step": 10919
+    },
+    {
+      "epoch": 8.219796763266842,
+      "grad_norm": 0.19313782453536987,
+      "learning_rate": 5e-06,
+      "loss": 0.1571,
+      "step": 10920
+    },
+    {
+      "epoch": 8.220549491908168,
+      "grad_norm": 0.17609134316444397,
+      "learning_rate": 5e-06,
+      "loss": 0.0855,
+      "step": 10921
+    },
+    {
+      "epoch": 8.221302220549491,
+      "grad_norm": 0.18568545579910278,
+      "learning_rate": 5e-06,
+      "loss": 0.15,
+      "step": 10922
+    },
+    {
+      "epoch": 8.222054949190817,
+      "grad_norm": 0.19606609642505646,
+      "learning_rate": 5e-06,
+      "loss": 0.2174,
+      "step": 10923
+    },
+    {
+      "epoch": 8.222807677832142,
+      "grad_norm": 0.18922735750675201,
+      "learning_rate": 5e-06,
+      "loss": 0.2078,
+      "step": 10924
+    },
+    {
+      "epoch": 8.223560406473466,
+      "grad_norm": 0.18644821643829346,
+      "learning_rate": 5e-06,
+      "loss": 0.098,
+      "step": 10925
+    },
+    {
+      "epoch": 8.224313135114791,
+      "grad_norm": 0.16484367847442627,
+      "learning_rate": 5e-06,
+      "loss": 0.1166,
+      "step": 10926
+    },
+    {
+      "epoch": 8.225065863756116,
+      "grad_norm": 0.1661226898431778,
+      "learning_rate": 5e-06,
+      "loss": 0.1105,
+      "step": 10927
+    },
+    {
+      "epoch": 8.225818592397442,
+      "grad_norm": 0.16170276701450348,
+      "learning_rate": 5e-06,
+      "loss": 0.1698,
+      "step": 10928
+    },
+    {
+      "epoch": 8.226571321038765,
+      "grad_norm": 0.18491925299167633,
+      "learning_rate": 5e-06,
+      "loss": 0.1713,
+      "step": 10929
+    },
+    {
+      "epoch": 8.22732404968009,
+      "grad_norm": 0.21090978384017944,
+      "learning_rate": 5e-06,
+      "loss": 0.1845,
+      "step": 10930
+    },
+    {
+      "epoch": 8.228076778321416,
+      "grad_norm": 0.1558266282081604,
+      "learning_rate": 5e-06,
+      "loss": 0.1021,
+      "step": 10931
+    },
+    {
+      "epoch": 8.22882950696274,
+      "grad_norm": 0.19794422388076782,
+      "learning_rate": 5e-06,
+      "loss": 0.2557,
+      "step": 10932
+    },
+    {
+      "epoch": 8.229582235604065,
+      "grad_norm": 0.17799097299575806,
+      "learning_rate": 5e-06,
+      "loss": 0.2597,
+      "step": 10933
+    },
+    {
+      "epoch": 8.23033496424539,
+      "grad_norm": 0.14939013123512268,
+      "learning_rate": 5e-06,
+      "loss": 0.1116,
+      "step": 10934
+    },
+    {
+      "epoch": 8.231087692886714,
+      "grad_norm": 0.1628924310207367,
+      "learning_rate": 5e-06,
+      "loss": 0.2227,
+      "step": 10935
+    },
+    {
+      "epoch": 8.23184042152804,
+      "grad_norm": 0.1822328418493271,
+      "learning_rate": 5e-06,
+      "loss": 0.103,
+      "step": 10936
+    },
+    {
+      "epoch": 8.232593150169365,
+      "grad_norm": 0.1597937047481537,
+      "learning_rate": 5e-06,
+      "loss": 0.0763,
+      "step": 10937
+    },
+    {
+      "epoch": 8.233345878810688,
+      "grad_norm": 0.19745661318302155,
+      "learning_rate": 5e-06,
+      "loss": 0.2046,
+      "step": 10938
+    },
+    {
+      "epoch": 8.234098607452013,
+      "grad_norm": 0.1853310614824295,
+      "learning_rate": 5e-06,
+      "loss": 0.1915,
+      "step": 10939
+    },
+    {
+      "epoch": 8.234851336093339,
+      "grad_norm": 0.1923995465040207,
+      "learning_rate": 5e-06,
+      "loss": 0.1393,
+      "step": 10940
+    },
+    {
+      "epoch": 8.235604064734662,
+      "grad_norm": 0.17474669218063354,
+      "learning_rate": 5e-06,
+      "loss": 0.2175,
+      "step": 10941
+    },
+    {
+      "epoch": 8.236356793375988,
+      "grad_norm": 0.1774951070547104,
+      "learning_rate": 5e-06,
+      "loss": 0.12,
+      "step": 10942
+    },
+    {
+      "epoch": 8.237109522017313,
+      "grad_norm": 0.14927475154399872,
+      "learning_rate": 5e-06,
+      "loss": 0.116,
+      "step": 10943
+    },
+    {
+      "epoch": 8.237862250658637,
+      "grad_norm": 0.19204656779766083,
+      "learning_rate": 5e-06,
+      "loss": 0.1864,
+      "step": 10944
+    },
+    {
+      "epoch": 8.238614979299962,
+      "grad_norm": 0.19612698256969452,
+      "learning_rate": 5e-06,
+      "loss": 0.1615,
+      "step": 10945
+    },
+    {
+      "epoch": 8.239367707941287,
+      "grad_norm": 0.17342349886894226,
+      "learning_rate": 5e-06,
+      "loss": 0.1218,
+      "step": 10946
+    },
+    {
+      "epoch": 8.240120436582613,
+      "grad_norm": 0.1817307472229004,
+      "learning_rate": 5e-06,
+      "loss": 0.186,
+      "step": 10947
+    },
+    {
+      "epoch": 8.240873165223936,
+      "grad_norm": 0.1828354299068451,
+      "learning_rate": 5e-06,
+      "loss": 0.1614,
+      "step": 10948
+    },
+    {
+      "epoch": 8.241625893865262,
+      "grad_norm": 0.17086198925971985,
+      "learning_rate": 5e-06,
+      "loss": 0.1302,
+      "step": 10949
+    },
+    {
+      "epoch": 8.242378622506587,
+      "grad_norm": 0.1553080826997757,
+      "learning_rate": 5e-06,
+      "loss": 0.1303,
+      "step": 10950
+    },
+    {
+      "epoch": 8.24313135114791,
+      "grad_norm": 0.18501976132392883,
+      "learning_rate": 5e-06,
+      "loss": 0.1459,
+      "step": 10951
+    },
+    {
+      "epoch": 8.243884079789236,
+      "grad_norm": 0.17576372623443604,
+      "learning_rate": 5e-06,
+      "loss": 0.1821,
+      "step": 10952
+    },
+    {
+      "epoch": 8.244636808430561,
+      "grad_norm": 0.1645433008670807,
+      "learning_rate": 5e-06,
+      "loss": 0.1154,
+      "step": 10953
+    },
+    {
+      "epoch": 8.245389537071885,
+      "grad_norm": 0.18414008617401123,
+      "learning_rate": 5e-06,
+      "loss": 0.2117,
+      "step": 10954
+    },
+    {
+      "epoch": 8.24614226571321,
+      "grad_norm": 0.16355347633361816,
+      "learning_rate": 5e-06,
+      "loss": 0.1855,
+      "step": 10955
+    },
+    {
+      "epoch": 8.246894994354536,
+      "grad_norm": 0.16933304071426392,
+      "learning_rate": 5e-06,
+      "loss": 0.1273,
+      "step": 10956
+    },
+    {
+      "epoch": 8.24764772299586,
+      "grad_norm": 0.1899043470621109,
+      "learning_rate": 5e-06,
+      "loss": 0.2334,
+      "step": 10957
+    },
+    {
+      "epoch": 8.248400451637185,
+      "grad_norm": 0.16845405101776123,
+      "learning_rate": 5e-06,
+      "loss": 0.2571,
+      "step": 10958
+    },
+    {
+      "epoch": 8.24915318027851,
+      "grad_norm": 0.19368374347686768,
+      "learning_rate": 5e-06,
+      "loss": 0.2109,
+      "step": 10959
+    },
+    {
+      "epoch": 8.249905908919834,
+      "grad_norm": 0.19910354912281036,
+      "learning_rate": 5e-06,
+      "loss": 0.1642,
+      "step": 10960
+    },
+    {
+      "epoch": 8.250658637561159,
+      "grad_norm": 0.16622884571552277,
+      "learning_rate": 5e-06,
+      "loss": 0.1337,
+      "step": 10961
+    },
+    {
+      "epoch": 8.251411366202484,
+      "grad_norm": 0.16200710833072662,
+      "learning_rate": 5e-06,
+      "loss": 0.0969,
+      "step": 10962
+    },
+    {
+      "epoch": 8.252164094843808,
+      "grad_norm": 0.15621739625930786,
+      "learning_rate": 5e-06,
+      "loss": 0.0855,
+      "step": 10963
+    },
+    {
+      "epoch": 8.252916823485133,
+      "grad_norm": 0.1815962940454483,
+      "learning_rate": 5e-06,
+      "loss": 0.1391,
+      "step": 10964
+    },
+    {
+      "epoch": 8.253669552126459,
+      "grad_norm": 0.17860320210456848,
+      "learning_rate": 5e-06,
+      "loss": 0.1835,
+      "step": 10965
+    },
+    {
+      "epoch": 8.254422280767784,
+      "grad_norm": 0.17425961792469025,
+      "learning_rate": 5e-06,
+      "loss": 0.089,
+      "step": 10966
+    },
+    {
+      "epoch": 8.255175009409108,
+      "grad_norm": 0.15591152012348175,
+      "learning_rate": 5e-06,
+      "loss": 0.0936,
+      "step": 10967
+    },
+    {
+      "epoch": 8.255927738050433,
+      "grad_norm": 0.15591144561767578,
+      "learning_rate": 5e-06,
+      "loss": 0.1069,
+      "step": 10968
+    },
+    {
+      "epoch": 8.256680466691758,
+      "grad_norm": 0.16513299942016602,
+      "learning_rate": 5e-06,
+      "loss": 0.1554,
+      "step": 10969
+    },
+    {
+      "epoch": 8.257433195333082,
+      "grad_norm": 0.15755723416805267,
+      "learning_rate": 5e-06,
+      "loss": 0.1491,
+      "step": 10970
+    },
+    {
+      "epoch": 8.258185923974407,
+      "grad_norm": 0.1647329330444336,
+      "learning_rate": 5e-06,
+      "loss": 0.1659,
+      "step": 10971
+    },
+    {
+      "epoch": 8.258938652615733,
+      "grad_norm": 0.19134005904197693,
+      "learning_rate": 5e-06,
+      "loss": 0.1498,
+      "step": 10972
+    },
+    {
+      "epoch": 8.259691381257056,
+      "grad_norm": 0.15781567990779877,
+      "learning_rate": 5e-06,
+      "loss": 0.0688,
+      "step": 10973
+    },
+    {
+      "epoch": 8.260444109898382,
+      "grad_norm": 0.18034695088863373,
+      "learning_rate": 5e-06,
+      "loss": 0.1785,
+      "step": 10974
+    },
+    {
+      "epoch": 8.261196838539707,
+      "grad_norm": 0.1789560616016388,
+      "learning_rate": 5e-06,
+      "loss": 0.1536,
+      "step": 10975
+    },
+    {
+      "epoch": 8.26194956718103,
+      "grad_norm": 0.15441444516181946,
+      "learning_rate": 5e-06,
+      "loss": 0.1211,
+      "step": 10976
+    },
+    {
+      "epoch": 8.262702295822356,
+      "grad_norm": 0.1509503275156021,
+      "learning_rate": 5e-06,
+      "loss": 0.0503,
+      "step": 10977
+    },
+    {
+      "epoch": 8.263455024463681,
+      "grad_norm": 0.17037037014961243,
+      "learning_rate": 5e-06,
+      "loss": 0.1315,
+      "step": 10978
+    },
+    {
+      "epoch": 8.264207753105005,
+      "grad_norm": 0.15150867402553558,
+      "learning_rate": 5e-06,
+      "loss": 0.1568,
+      "step": 10979
+    },
+    {
+      "epoch": 8.26496048174633,
+      "grad_norm": 0.1710224449634552,
+      "learning_rate": 5e-06,
+      "loss": 0.1209,
+      "step": 10980
+    },
+    {
+      "epoch": 8.265713210387656,
+      "grad_norm": 0.18621079623699188,
+      "learning_rate": 5e-06,
+      "loss": 0.2027,
+      "step": 10981
+    },
+    {
+      "epoch": 8.26646593902898,
+      "grad_norm": 0.17228387296199799,
+      "learning_rate": 5e-06,
+      "loss": 0.2238,
+      "step": 10982
+    },
+    {
+      "epoch": 8.267218667670305,
+      "grad_norm": 0.15854886174201965,
+      "learning_rate": 5e-06,
+      "loss": 0.0934,
+      "step": 10983
+    },
+    {
+      "epoch": 8.26797139631163,
+      "grad_norm": 0.17735835909843445,
+      "learning_rate": 5e-06,
+      "loss": 0.1293,
+      "step": 10984
+    },
+    {
+      "epoch": 8.268724124952955,
+      "grad_norm": 0.15839701890945435,
+      "learning_rate": 5e-06,
+      "loss": 0.164,
+      "step": 10985
+    },
+    {
+      "epoch": 8.269476853594279,
+      "grad_norm": 0.16102933883666992,
+      "learning_rate": 5e-06,
+      "loss": 0.1939,
+      "step": 10986
+    },
+    {
+      "epoch": 8.270229582235604,
+      "grad_norm": 0.1507130265235901,
+      "learning_rate": 5e-06,
+      "loss": 0.1253,
+      "step": 10987
+    },
+    {
+      "epoch": 8.27098231087693,
+      "grad_norm": 0.17814628779888153,
+      "learning_rate": 5e-06,
+      "loss": 0.142,
+      "step": 10988
+    },
+    {
+      "epoch": 8.271735039518253,
+      "grad_norm": 0.16417206823825836,
+      "learning_rate": 5e-06,
+      "loss": 0.0941,
+      "step": 10989
+    },
+    {
+      "epoch": 8.272487768159579,
+      "grad_norm": 0.15023556351661682,
+      "learning_rate": 5e-06,
+      "loss": 0.1366,
+      "step": 10990
+    },
+    {
+      "epoch": 8.273240496800904,
+      "grad_norm": 0.18644344806671143,
+      "learning_rate": 5e-06,
+      "loss": 0.2258,
+      "step": 10991
+    },
+    {
+      "epoch": 8.273993225442227,
+      "grad_norm": 0.1668095886707306,
+      "learning_rate": 5e-06,
+      "loss": 0.1311,
+      "step": 10992
+    },
+    {
+      "epoch": 8.274745954083553,
+      "grad_norm": 0.21029342710971832,
+      "learning_rate": 5e-06,
+      "loss": 0.2347,
+      "step": 10993
+    },
+    {
+      "epoch": 8.275498682724878,
+      "grad_norm": 0.16934935748577118,
+      "learning_rate": 5e-06,
+      "loss": 0.1025,
+      "step": 10994
+    },
+    {
+      "epoch": 8.276251411366202,
+      "grad_norm": 0.16781793534755707,
+      "learning_rate": 5e-06,
+      "loss": 0.1588,
+      "step": 10995
+    },
+    {
+      "epoch": 8.277004140007527,
+      "grad_norm": 0.18266648054122925,
+      "learning_rate": 5e-06,
+      "loss": 0.0964,
+      "step": 10996
+    },
+    {
+      "epoch": 8.277756868648853,
+      "grad_norm": 0.16234950721263885,
+      "learning_rate": 5e-06,
+      "loss": 0.1495,
+      "step": 10997
+    },
+    {
+      "epoch": 8.278509597290176,
+      "grad_norm": 0.2050333023071289,
+      "learning_rate": 5e-06,
+      "loss": 0.2216,
+      "step": 10998
+    },
+    {
+      "epoch": 8.279262325931501,
+      "grad_norm": 0.17918281257152557,
+      "learning_rate": 5e-06,
+      "loss": 0.1608,
+      "step": 10999
+    },
+    {
+      "epoch": 8.280015054572827,
+      "grad_norm": 0.16942082345485687,
+      "learning_rate": 5e-06,
+      "loss": 0.1167,
+      "step": 11000
+    },
+    {
+      "epoch": 8.280015054572827,
+      "eval_loss": 0.1953597515821457,
+      "eval_runtime": 452.888,
+      "eval_samples_per_second": 21.257,
+      "eval_steps_per_second": 0.665,
+      "step": 11000
+    },
+    {
+      "epoch": 8.280767783214152,
+      "grad_norm": 0.1660325676202774,
+      "learning_rate": 5e-06,
+      "loss": 0.1015,
+      "step": 11001
+    },
+    {
+      "epoch": 8.281520511855476,
+      "grad_norm": 0.18826332688331604,
+      "learning_rate": 5e-06,
+      "loss": 0.1673,
+      "step": 11002
+    },
+    {
+      "epoch": 8.282273240496801,
+      "grad_norm": 0.1719111055135727,
+      "learning_rate": 5e-06,
+      "loss": 0.1373,
+      "step": 11003
+    },
+    {
+      "epoch": 8.283025969138126,
+      "grad_norm": 0.18992769718170166,
+      "learning_rate": 5e-06,
+      "loss": 0.1618,
+      "step": 11004
+    },
+    {
+      "epoch": 8.28377869777945,
+      "grad_norm": 0.20485910773277283,
+      "learning_rate": 5e-06,
+      "loss": 0.1642,
+      "step": 11005
+    },
+    {
+      "epoch": 8.284531426420775,
+      "grad_norm": 0.1711505502462387,
+      "learning_rate": 5e-06,
+      "loss": 0.1045,
+      "step": 11006
+    },
+    {
+      "epoch": 8.2852841550621,
+      "grad_norm": 0.1939704716205597,
+      "learning_rate": 5e-06,
+      "loss": 0.16,
+      "step": 11007
+    },
+    {
+      "epoch": 8.286036883703424,
+      "grad_norm": 0.17641495168209076,
+      "learning_rate": 5e-06,
+      "loss": 0.1542,
+      "step": 11008
+    },
+    {
+      "epoch": 8.28678961234475,
+      "grad_norm": 0.15367750823497772,
+      "learning_rate": 5e-06,
+      "loss": 0.1796,
+      "step": 11009
+    },
+    {
+      "epoch": 8.287542340986075,
+      "grad_norm": 0.14976699650287628,
+      "learning_rate": 5e-06,
+      "loss": 0.1222,
+      "step": 11010
+    },
+    {
+      "epoch": 8.288295069627399,
+      "grad_norm": 0.21439386904239655,
+      "learning_rate": 5e-06,
+      "loss": 0.1813,
+      "step": 11011
+    },
+    {
+      "epoch": 8.289047798268724,
+      "grad_norm": 0.19663594663143158,
+      "learning_rate": 5e-06,
+      "loss": 0.1036,
+      "step": 11012
+    },
+    {
+      "epoch": 8.28980052691005,
+      "grad_norm": 0.14143528044223785,
+      "learning_rate": 5e-06,
+      "loss": 0.1544,
+      "step": 11013
+    },
+    {
+      "epoch": 8.290553255551373,
+      "grad_norm": 0.19017012417316437,
+      "learning_rate": 5e-06,
+      "loss": 0.1331,
+      "step": 11014
+    },
+    {
+      "epoch": 8.291305984192698,
+      "grad_norm": 0.1896316409111023,
+      "learning_rate": 5e-06,
+      "loss": 0.1786,
+      "step": 11015
+    },
+    {
+      "epoch": 8.292058712834024,
+      "grad_norm": 0.18924808502197266,
+      "learning_rate": 5e-06,
+      "loss": 0.1816,
+      "step": 11016
+    },
+    {
+      "epoch": 8.292811441475347,
+      "grad_norm": 0.15612755715847015,
+      "learning_rate": 5e-06,
+      "loss": 0.1211,
+      "step": 11017
+    },
+    {
+      "epoch": 8.293564170116673,
+      "grad_norm": 0.1963435858488083,
+      "learning_rate": 5e-06,
+      "loss": 0.2668,
+      "step": 11018
+    },
+    {
+      "epoch": 8.294316898757998,
+      "grad_norm": 0.16544198989868164,
+      "learning_rate": 5e-06,
+      "loss": 0.1459,
+      "step": 11019
+    },
+    {
+      "epoch": 8.295069627399322,
+      "grad_norm": 0.182954341173172,
+      "learning_rate": 5e-06,
+      "loss": 0.1946,
+      "step": 11020
+    },
+    {
+      "epoch": 8.295822356040647,
+      "grad_norm": 0.18318608403205872,
+      "learning_rate": 5e-06,
+      "loss": 0.1566,
+      "step": 11021
+    },
+    {
+      "epoch": 8.296575084681972,
+      "grad_norm": 0.17218956351280212,
+      "learning_rate": 5e-06,
+      "loss": 0.1136,
+      "step": 11022
+    },
+    {
+      "epoch": 8.297327813323298,
+      "grad_norm": 0.17204846441745758,
+      "learning_rate": 5e-06,
+      "loss": 0.1218,
+      "step": 11023
+    },
+    {
+      "epoch": 8.298080541964621,
+      "grad_norm": 0.17869402468204498,
+      "learning_rate": 5e-06,
+      "loss": 0.1809,
+      "step": 11024
+    },
+    {
+      "epoch": 8.298833270605947,
+      "grad_norm": 0.16719095408916473,
+      "learning_rate": 5e-06,
+      "loss": 0.1855,
+      "step": 11025
+    },
+    {
+      "epoch": 8.299585999247272,
+      "grad_norm": 0.1604928970336914,
+      "learning_rate": 5e-06,
+      "loss": 0.1637,
+      "step": 11026
+    },
+    {
+      "epoch": 8.300338727888596,
+      "grad_norm": 0.1837502121925354,
+      "learning_rate": 5e-06,
+      "loss": 0.1472,
+      "step": 11027
+    },
+    {
+      "epoch": 8.301091456529921,
+      "grad_norm": 0.16713504493236542,
+      "learning_rate": 5e-06,
+      "loss": 0.079,
+      "step": 11028
+    },
+    {
+      "epoch": 8.301844185171246,
+      "grad_norm": 0.18427948653697968,
+      "learning_rate": 5e-06,
+      "loss": 0.1905,
+      "step": 11029
+    },
+    {
+      "epoch": 8.30259691381257,
+      "grad_norm": 0.1853123903274536,
+      "learning_rate": 5e-06,
+      "loss": 0.1898,
+      "step": 11030
+    },
+    {
+      "epoch": 8.303349642453895,
+      "grad_norm": 0.1772039532661438,
+      "learning_rate": 5e-06,
+      "loss": 0.2093,
+      "step": 11031
+    },
+    {
+      "epoch": 8.30410237109522,
+      "grad_norm": 0.19239722192287445,
+      "learning_rate": 5e-06,
+      "loss": 0.1535,
+      "step": 11032
+    },
+    {
+      "epoch": 8.304855099736544,
+      "grad_norm": 0.17521290481090546,
+      "learning_rate": 5e-06,
+      "loss": 0.2372,
+      "step": 11033
+    },
+    {
+      "epoch": 8.30560782837787,
+      "grad_norm": 0.17544904351234436,
+      "learning_rate": 5e-06,
+      "loss": 0.1659,
+      "step": 11034
+    },
+    {
+      "epoch": 8.306360557019195,
+      "grad_norm": 0.17918133735656738,
+      "learning_rate": 5e-06,
+      "loss": 0.1141,
+      "step": 11035
+    },
+    {
+      "epoch": 8.307113285660519,
+      "grad_norm": 0.18101105093955994,
+      "learning_rate": 5e-06,
+      "loss": 0.1498,
+      "step": 11036
+    },
+    {
+      "epoch": 8.307866014301844,
+      "grad_norm": 0.22221817076206207,
+      "learning_rate": 5e-06,
+      "loss": 0.1907,
+      "step": 11037
+    },
+    {
+      "epoch": 8.30861874294317,
+      "grad_norm": 0.17911024391651154,
+      "learning_rate": 5e-06,
+      "loss": 0.1357,
+      "step": 11038
+    },
+    {
+      "epoch": 8.309371471584495,
+      "grad_norm": 0.1622738391160965,
+      "learning_rate": 5e-06,
+      "loss": 0.1098,
+      "step": 11039
+    },
+    {
+      "epoch": 8.310124200225818,
+      "grad_norm": 0.19653767347335815,
+      "learning_rate": 5e-06,
+      "loss": 0.2015,
+      "step": 11040
+    },
+    {
+      "epoch": 8.310876928867144,
+      "grad_norm": 0.2102191299200058,
+      "learning_rate": 5e-06,
+      "loss": 0.1242,
+      "step": 11041
+    },
+    {
+      "epoch": 8.311629657508469,
+      "grad_norm": 0.15700532495975494,
+      "learning_rate": 5e-06,
+      "loss": 0.2332,
+      "step": 11042
+    },
+    {
+      "epoch": 8.312382386149793,
+      "grad_norm": 0.1774825155735016,
+      "learning_rate": 5e-06,
+      "loss": 0.1894,
+      "step": 11043
+    },
+    {
+      "epoch": 8.313135114791118,
+      "grad_norm": 0.1855405867099762,
+      "learning_rate": 5e-06,
+      "loss": 0.2274,
+      "step": 11044
+    },
+    {
+      "epoch": 8.313887843432443,
+      "grad_norm": 0.17532843351364136,
+      "learning_rate": 5e-06,
+      "loss": 0.1574,
+      "step": 11045
+    },
+    {
+      "epoch": 8.314640572073767,
+      "grad_norm": 0.16603243350982666,
+      "learning_rate": 5e-06,
+      "loss": 0.1599,
+      "step": 11046
+    },
+    {
+      "epoch": 8.315393300715092,
+      "grad_norm": 0.176155686378479,
+      "learning_rate": 5e-06,
+      "loss": 0.1134,
+      "step": 11047
+    },
+    {
+      "epoch": 8.316146029356418,
+      "grad_norm": 0.2062234729528427,
+      "learning_rate": 5e-06,
+      "loss": 0.1552,
+      "step": 11048
+    },
+    {
+      "epoch": 8.316898757997741,
+      "grad_norm": 0.17361503839492798,
+      "learning_rate": 5e-06,
+      "loss": 0.1583,
+      "step": 11049
+    },
+    {
+      "epoch": 8.317651486639067,
+      "grad_norm": 0.1468573808670044,
+      "learning_rate": 5e-06,
+      "loss": 0.0967,
+      "step": 11050
+    },
+    {
+      "epoch": 8.318404215280392,
+      "grad_norm": 0.15953968465328217,
+      "learning_rate": 5e-06,
+      "loss": 0.1212,
+      "step": 11051
+    },
+    {
+      "epoch": 8.319156943921715,
+      "grad_norm": 0.17000249028205872,
+      "learning_rate": 5e-06,
+      "loss": 0.1016,
+      "step": 11052
+    },
+    {
+      "epoch": 8.31990967256304,
+      "grad_norm": 0.15021781623363495,
+      "learning_rate": 5e-06,
+      "loss": 0.097,
+      "step": 11053
+    },
+    {
+      "epoch": 8.320662401204366,
+      "grad_norm": 0.18808038532733917,
+      "learning_rate": 5e-06,
+      "loss": 0.1863,
+      "step": 11054
+    },
+    {
+      "epoch": 8.32141512984569,
+      "grad_norm": 0.18276125192642212,
+      "learning_rate": 5e-06,
+      "loss": 0.2008,
+      "step": 11055
+    },
+    {
+      "epoch": 8.322167858487015,
+      "grad_norm": 0.15892204642295837,
+      "learning_rate": 5e-06,
+      "loss": 0.0614,
+      "step": 11056
+    },
+    {
+      "epoch": 8.32292058712834,
+      "grad_norm": 0.14978578686714172,
+      "learning_rate": 5e-06,
+      "loss": 0.1189,
+      "step": 11057
+    },
+    {
+      "epoch": 8.323673315769666,
+      "grad_norm": 0.1763635128736496,
+      "learning_rate": 5e-06,
+      "loss": 0.1362,
+      "step": 11058
+    },
+    {
+      "epoch": 8.32442604441099,
+      "grad_norm": 0.16082248091697693,
+      "learning_rate": 5e-06,
+      "loss": 0.1088,
+      "step": 11059
+    },
+    {
+      "epoch": 8.325178773052315,
+      "grad_norm": 0.2053227424621582,
+      "learning_rate": 5e-06,
+      "loss": 0.1734,
+      "step": 11060
+    },
+    {
+      "epoch": 8.32593150169364,
+      "grad_norm": 0.15393947064876556,
+      "learning_rate": 5e-06,
+      "loss": 0.123,
+      "step": 11061
+    },
+    {
+      "epoch": 8.326684230334964,
+      "grad_norm": 0.17694091796875,
+      "learning_rate": 5e-06,
+      "loss": 0.1411,
+      "step": 11062
+    },
+    {
+      "epoch": 8.327436958976289,
+      "grad_norm": 0.16282865405082703,
+      "learning_rate": 5e-06,
+      "loss": 0.1619,
+      "step": 11063
+    },
+    {
+      "epoch": 8.328189687617614,
+      "grad_norm": 0.18792004883289337,
+      "learning_rate": 5e-06,
+      "loss": 0.137,
+      "step": 11064
+    },
+    {
+      "epoch": 8.328942416258938,
+      "grad_norm": 0.14409877359867096,
+      "learning_rate": 5e-06,
+      "loss": 0.1402,
+      "step": 11065
+    },
+    {
+      "epoch": 8.329695144900263,
+      "grad_norm": 0.17222870886325836,
+      "learning_rate": 5e-06,
+      "loss": 0.1796,
+      "step": 11066
+    },
+    {
+      "epoch": 8.330447873541589,
+      "grad_norm": 0.17420059442520142,
+      "learning_rate": 5e-06,
+      "loss": 0.0945,
+      "step": 11067
+    },
+    {
+      "epoch": 8.331200602182912,
+      "grad_norm": 0.163681760430336,
+      "learning_rate": 5e-06,
+      "loss": 0.0946,
+      "step": 11068
+    },
+    {
+      "epoch": 8.331953330824238,
+      "grad_norm": 0.1926737278699875,
+      "learning_rate": 5e-06,
+      "loss": 0.2281,
+      "step": 11069
+    },
+    {
+      "epoch": 8.332706059465563,
+      "grad_norm": 0.16940459609031677,
+      "learning_rate": 5e-06,
+      "loss": 0.1068,
+      "step": 11070
+    },
+    {
+      "epoch": 8.333458788106887,
+      "grad_norm": 0.18529342114925385,
+      "learning_rate": 5e-06,
+      "loss": 0.1243,
+      "step": 11071
+    },
+    {
+      "epoch": 8.334211516748212,
+      "grad_norm": 0.1494583636522293,
+      "learning_rate": 5e-06,
+      "loss": 0.1397,
+      "step": 11072
+    },
+    {
+      "epoch": 8.334964245389537,
+      "grad_norm": 0.1636829525232315,
+      "learning_rate": 5e-06,
+      "loss": 0.1218,
+      "step": 11073
+    },
+    {
+      "epoch": 8.335716974030863,
+      "grad_norm": 0.16170799732208252,
+      "learning_rate": 5e-06,
+      "loss": 0.1157,
+      "step": 11074
+    },
+    {
+      "epoch": 8.336469702672186,
+      "grad_norm": 0.17650072276592255,
+      "learning_rate": 5e-06,
+      "loss": 0.2394,
+      "step": 11075
+    },
+    {
+      "epoch": 8.337222431313512,
+      "grad_norm": 0.18219806253910065,
+      "learning_rate": 5e-06,
+      "loss": 0.1269,
+      "step": 11076
+    },
+    {
+      "epoch": 8.337975159954837,
+      "grad_norm": 0.20506206154823303,
+      "learning_rate": 5e-06,
+      "loss": 0.198,
+      "step": 11077
+    },
+    {
+      "epoch": 8.33872788859616,
+      "grad_norm": 0.16829092800617218,
+      "learning_rate": 5e-06,
+      "loss": 0.1107,
+      "step": 11078
+    },
+    {
+      "epoch": 8.339480617237486,
+      "grad_norm": 0.19007010757923126,
+      "learning_rate": 5e-06,
+      "loss": 0.2383,
+      "step": 11079
+    },
+    {
+      "epoch": 8.340233345878811,
+      "grad_norm": 0.16067758202552795,
+      "learning_rate": 5e-06,
+      "loss": 0.093,
+      "step": 11080
+    },
+    {
+      "epoch": 8.340986074520135,
+      "grad_norm": 0.1714971363544464,
+      "learning_rate": 5e-06,
+      "loss": 0.0977,
+      "step": 11081
+    },
+    {
+      "epoch": 8.34173880316146,
+      "grad_norm": 0.1685871183872223,
+      "learning_rate": 5e-06,
+      "loss": 0.157,
+      "step": 11082
+    },
+    {
+      "epoch": 8.342491531802786,
+      "grad_norm": 0.1716592162847519,
+      "learning_rate": 5e-06,
+      "loss": 0.1403,
+      "step": 11083
+    },
+    {
+      "epoch": 8.34324426044411,
+      "grad_norm": 0.19482578337192535,
+      "learning_rate": 5e-06,
+      "loss": 0.189,
+      "step": 11084
+    },
+    {
+      "epoch": 8.343996989085435,
+      "grad_norm": 0.16923439502716064,
+      "learning_rate": 5e-06,
+      "loss": 0.1883,
+      "step": 11085
+    },
+    {
+      "epoch": 8.34474971772676,
+      "grad_norm": 0.16578848659992218,
+      "learning_rate": 5e-06,
+      "loss": 0.125,
+      "step": 11086
+    },
+    {
+      "epoch": 8.345502446368084,
+      "grad_norm": 0.17798274755477905,
+      "learning_rate": 5e-06,
+      "loss": 0.1477,
+      "step": 11087
+    },
+    {
+      "epoch": 8.346255175009409,
+      "grad_norm": 0.18254312872886658,
+      "learning_rate": 5e-06,
+      "loss": 0.1566,
+      "step": 11088
+    },
+    {
+      "epoch": 8.347007903650734,
+      "grad_norm": 0.17902685701847076,
+      "learning_rate": 5e-06,
+      "loss": 0.1422,
+      "step": 11089
+    },
+    {
+      "epoch": 8.347760632292058,
+      "grad_norm": 0.1792583465576172,
+      "learning_rate": 5e-06,
+      "loss": 0.1372,
+      "step": 11090
+    },
+    {
+      "epoch": 8.348513360933383,
+      "grad_norm": 0.17781338095664978,
+      "learning_rate": 5e-06,
+      "loss": 0.2078,
+      "step": 11091
+    },
+    {
+      "epoch": 8.349266089574709,
+      "grad_norm": 0.2111758440732956,
+      "learning_rate": 5e-06,
+      "loss": 0.235,
+      "step": 11092
+    },
+    {
+      "epoch": 8.350018818216034,
+      "grad_norm": 0.16331560909748077,
+      "learning_rate": 5e-06,
+      "loss": 0.1262,
+      "step": 11093
+    },
+    {
+      "epoch": 8.350771546857358,
+      "grad_norm": 0.17370444536209106,
+      "learning_rate": 5e-06,
+      "loss": 0.1736,
+      "step": 11094
+    },
+    {
+      "epoch": 8.351524275498683,
+      "grad_norm": 0.1889808177947998,
+      "learning_rate": 5e-06,
+      "loss": 0.2507,
+      "step": 11095
+    },
+    {
+      "epoch": 8.352277004140008,
+      "grad_norm": 0.20081447064876556,
+      "learning_rate": 5e-06,
+      "loss": 0.2427,
+      "step": 11096
+    },
+    {
+      "epoch": 8.353029732781332,
+      "grad_norm": 0.1774754822254181,
+      "learning_rate": 5e-06,
+      "loss": 0.1934,
+      "step": 11097
+    },
+    {
+      "epoch": 8.353782461422657,
+      "grad_norm": 0.16311922669410706,
+      "learning_rate": 5e-06,
+      "loss": 0.1899,
+      "step": 11098
+    },
+    {
+      "epoch": 8.354535190063983,
+      "grad_norm": 0.1568531095981598,
+      "learning_rate": 5e-06,
+      "loss": 0.1361,
+      "step": 11099
+    },
+    {
+      "epoch": 8.355287918705306,
+      "grad_norm": 0.1640501171350479,
+      "learning_rate": 5e-06,
+      "loss": 0.1086,
+      "step": 11100
+    },
+    {
+      "epoch": 8.356040647346632,
+      "grad_norm": 0.1656324565410614,
+      "learning_rate": 5e-06,
+      "loss": 0.0975,
+      "step": 11101
+    },
+    {
+      "epoch": 8.356793375987957,
+      "grad_norm": 0.16784673929214478,
+      "learning_rate": 5e-06,
+      "loss": 0.201,
+      "step": 11102
+    },
+    {
+      "epoch": 8.35754610462928,
+      "grad_norm": 0.16762857139110565,
+      "learning_rate": 5e-06,
+      "loss": 0.1821,
+      "step": 11103
+    },
+    {
+      "epoch": 8.358298833270606,
+      "grad_norm": 0.1898474395275116,
+      "learning_rate": 5e-06,
+      "loss": 0.1187,
+      "step": 11104
+    },
+    {
+      "epoch": 8.359051561911931,
+      "grad_norm": 0.20514728128910065,
+      "learning_rate": 5e-06,
+      "loss": 0.1979,
+      "step": 11105
+    },
+    {
+      "epoch": 8.359804290553255,
+      "grad_norm": 0.1586589366197586,
+      "learning_rate": 5e-06,
+      "loss": 0.0972,
+      "step": 11106
+    },
+    {
+      "epoch": 8.36055701919458,
+      "grad_norm": 0.1662803292274475,
+      "learning_rate": 5e-06,
+      "loss": 0.1451,
+      "step": 11107
+    },
+    {
+      "epoch": 8.361309747835906,
+      "grad_norm": 0.17054776847362518,
+      "learning_rate": 5e-06,
+      "loss": 0.1439,
+      "step": 11108
+    },
+    {
+      "epoch": 8.362062476477229,
+      "grad_norm": 0.18851162493228912,
+      "learning_rate": 5e-06,
+      "loss": 0.2073,
+      "step": 11109
+    },
+    {
+      "epoch": 8.362815205118554,
+      "grad_norm": 0.17925497889518738,
+      "learning_rate": 5e-06,
+      "loss": 0.2249,
+      "step": 11110
+    },
+    {
+      "epoch": 8.36356793375988,
+      "grad_norm": 0.16740640997886658,
+      "learning_rate": 5e-06,
+      "loss": 0.1303,
+      "step": 11111
+    },
+    {
+      "epoch": 8.364320662401205,
+      "grad_norm": 0.1662808209657669,
+      "learning_rate": 5e-06,
+      "loss": 0.1969,
+      "step": 11112
+    },
+    {
+      "epoch": 8.365073391042529,
+      "grad_norm": 0.15173839032649994,
+      "learning_rate": 5e-06,
+      "loss": 0.0788,
+      "step": 11113
+    },
+    {
+      "epoch": 8.365826119683854,
+      "grad_norm": 0.1821129471063614,
+      "learning_rate": 5e-06,
+      "loss": 0.0988,
+      "step": 11114
+    },
+    {
+      "epoch": 8.36657884832518,
+      "grad_norm": 0.17887891829013824,
+      "learning_rate": 5e-06,
+      "loss": 0.2091,
+      "step": 11115
+    },
+    {
+      "epoch": 8.367331576966503,
+      "grad_norm": 0.20403096079826355,
+      "learning_rate": 5e-06,
+      "loss": 0.154,
+      "step": 11116
+    },
+    {
+      "epoch": 8.368084305607828,
+      "grad_norm": 0.17366720736026764,
+      "learning_rate": 5e-06,
+      "loss": 0.1576,
+      "step": 11117
+    },
+    {
+      "epoch": 8.368837034249154,
+      "grad_norm": 0.18191829323768616,
+      "learning_rate": 5e-06,
+      "loss": 0.1914,
+      "step": 11118
+    },
+    {
+      "epoch": 8.369589762890477,
+      "grad_norm": 0.17939655482769012,
+      "learning_rate": 5e-06,
+      "loss": 0.2032,
+      "step": 11119
+    },
+    {
+      "epoch": 8.370342491531803,
+      "grad_norm": 0.17404380440711975,
+      "learning_rate": 5e-06,
+      "loss": 0.1493,
+      "step": 11120
+    },
+    {
+      "epoch": 8.371095220173128,
+      "grad_norm": 0.14199712872505188,
+      "learning_rate": 5e-06,
+      "loss": 0.109,
+      "step": 11121
+    },
+    {
+      "epoch": 8.371847948814452,
+      "grad_norm": 0.17375826835632324,
+      "learning_rate": 5e-06,
+      "loss": 0.1688,
+      "step": 11122
+    },
+    {
+      "epoch": 8.372600677455777,
+      "grad_norm": 0.17276796698570251,
+      "learning_rate": 5e-06,
+      "loss": 0.1213,
+      "step": 11123
+    },
+    {
+      "epoch": 8.373353406097102,
+      "grad_norm": 0.17240475118160248,
+      "learning_rate": 5e-06,
+      "loss": 0.1298,
+      "step": 11124
+    },
+    {
+      "epoch": 8.374106134738426,
+      "grad_norm": 0.1841440051794052,
+      "learning_rate": 5e-06,
+      "loss": 0.2315,
+      "step": 11125
+    },
+    {
+      "epoch": 8.374858863379751,
+      "grad_norm": 0.1602412760257721,
+      "learning_rate": 5e-06,
+      "loss": 0.1349,
+      "step": 11126
+    },
+    {
+      "epoch": 8.375611592021077,
+      "grad_norm": 0.17091158032417297,
+      "learning_rate": 5e-06,
+      "loss": 0.1576,
+      "step": 11127
+    },
+    {
+      "epoch": 8.3763643206624,
+      "grad_norm": 0.18945929408073425,
+      "learning_rate": 5e-06,
+      "loss": 0.1427,
+      "step": 11128
+    },
+    {
+      "epoch": 8.377117049303726,
+      "grad_norm": 0.1569545567035675,
+      "learning_rate": 5e-06,
+      "loss": 0.1232,
+      "step": 11129
+    },
+    {
+      "epoch": 8.377869777945051,
+      "grad_norm": 0.20928041636943817,
+      "learning_rate": 5e-06,
+      "loss": 0.1617,
+      "step": 11130
+    },
+    {
+      "epoch": 8.378622506586376,
+      "grad_norm": 0.18268349766731262,
+      "learning_rate": 5e-06,
+      "loss": 0.1755,
+      "step": 11131
+    },
+    {
+      "epoch": 8.3793752352277,
+      "grad_norm": 0.15208205580711365,
+      "learning_rate": 5e-06,
+      "loss": 0.1018,
+      "step": 11132
+    },
+    {
+      "epoch": 8.380127963869025,
+      "grad_norm": 0.1758023202419281,
+      "learning_rate": 5e-06,
+      "loss": 0.1063,
+      "step": 11133
+    },
+    {
+      "epoch": 8.38088069251035,
+      "grad_norm": 0.17405903339385986,
+      "learning_rate": 5e-06,
+      "loss": 0.1723,
+      "step": 11134
+    },
+    {
+      "epoch": 8.381633421151674,
+      "grad_norm": 0.14736349880695343,
+      "learning_rate": 5e-06,
+      "loss": 0.1441,
+      "step": 11135
+    },
+    {
+      "epoch": 8.382386149793,
+      "grad_norm": 0.1818760335445404,
+      "learning_rate": 5e-06,
+      "loss": 0.2196,
+      "step": 11136
+    },
+    {
+      "epoch": 8.383138878434325,
+      "grad_norm": 0.16185910999774933,
+      "learning_rate": 5e-06,
+      "loss": 0.1674,
+      "step": 11137
+    },
+    {
+      "epoch": 8.383891607075649,
+      "grad_norm": 0.1584905982017517,
+      "learning_rate": 5e-06,
+      "loss": 0.1791,
+      "step": 11138
+    },
+    {
+      "epoch": 8.384644335716974,
+      "grad_norm": 0.14299485087394714,
+      "learning_rate": 5e-06,
+      "loss": 0.1176,
+      "step": 11139
+    },
+    {
+      "epoch": 8.3853970643583,
+      "grad_norm": 0.1888038069009781,
+      "learning_rate": 5e-06,
+      "loss": 0.1695,
+      "step": 11140
+    },
+    {
+      "epoch": 8.386149792999623,
+      "grad_norm": 0.18491166830062866,
+      "learning_rate": 5e-06,
+      "loss": 0.1846,
+      "step": 11141
+    },
+    {
+      "epoch": 8.386902521640948,
+      "grad_norm": 0.17367327213287354,
+      "learning_rate": 5e-06,
+      "loss": 0.1127,
+      "step": 11142
+    },
+    {
+      "epoch": 8.387655250282274,
+      "grad_norm": 0.1652689427137375,
+      "learning_rate": 5e-06,
+      "loss": 0.1339,
+      "step": 11143
+    },
+    {
+      "epoch": 8.388407978923597,
+      "grad_norm": 0.17843475937843323,
+      "learning_rate": 5e-06,
+      "loss": 0.0986,
+      "step": 11144
+    },
+    {
+      "epoch": 8.389160707564923,
+      "grad_norm": 0.17072570323944092,
+      "learning_rate": 5e-06,
+      "loss": 0.1083,
+      "step": 11145
+    },
+    {
+      "epoch": 8.389913436206248,
+      "grad_norm": 0.16896885633468628,
+      "learning_rate": 5e-06,
+      "loss": 0.1888,
+      "step": 11146
+    },
+    {
+      "epoch": 8.390666164847573,
+      "grad_norm": 0.17673015594482422,
+      "learning_rate": 5e-06,
+      "loss": 0.1748,
+      "step": 11147
+    },
+    {
+      "epoch": 8.391418893488897,
+      "grad_norm": 0.16553696990013123,
+      "learning_rate": 5e-06,
+      "loss": 0.1501,
+      "step": 11148
+    },
+    {
+      "epoch": 8.392171622130222,
+      "grad_norm": 0.17955462634563446,
+      "learning_rate": 5e-06,
+      "loss": 0.1631,
+      "step": 11149
+    },
+    {
+      "epoch": 8.392924350771548,
+      "grad_norm": 0.19030405580997467,
+      "learning_rate": 5e-06,
+      "loss": 0.1596,
+      "step": 11150
+    },
+    {
+      "epoch": 8.393677079412871,
+      "grad_norm": 0.1542009562253952,
+      "learning_rate": 5e-06,
+      "loss": 0.1688,
+      "step": 11151
+    },
+    {
+      "epoch": 8.394429808054197,
+      "grad_norm": 0.17817698419094086,
+      "learning_rate": 5e-06,
+      "loss": 0.1468,
+      "step": 11152
+    },
+    {
+      "epoch": 8.395182536695522,
+      "grad_norm": 0.19852964580059052,
+      "learning_rate": 5e-06,
+      "loss": 0.2044,
+      "step": 11153
+    },
+    {
+      "epoch": 8.395935265336846,
+      "grad_norm": 0.16506610810756683,
+      "learning_rate": 5e-06,
+      "loss": 0.1954,
+      "step": 11154
+    },
+    {
+      "epoch": 8.396687993978171,
+      "grad_norm": 0.16025662422180176,
+      "learning_rate": 5e-06,
+      "loss": 0.1159,
+      "step": 11155
+    },
+    {
+      "epoch": 8.397440722619496,
+      "grad_norm": 0.16123132407665253,
+      "learning_rate": 5e-06,
+      "loss": 0.1383,
+      "step": 11156
+    },
+    {
+      "epoch": 8.39819345126082,
+      "grad_norm": 0.20959557592868805,
+      "learning_rate": 5e-06,
+      "loss": 0.2502,
+      "step": 11157
+    },
+    {
+      "epoch": 8.398946179902145,
+      "grad_norm": 0.18460339307785034,
+      "learning_rate": 5e-06,
+      "loss": 0.2658,
+      "step": 11158
+    },
+    {
+      "epoch": 8.39969890854347,
+      "grad_norm": 0.15705427527427673,
+      "learning_rate": 5e-06,
+      "loss": 0.0854,
+      "step": 11159
+    },
+    {
+      "epoch": 8.400451637184794,
+      "grad_norm": 0.19334763288497925,
+      "learning_rate": 5e-06,
+      "loss": 0.2029,
+      "step": 11160
+    },
+    {
+      "epoch": 8.40120436582612,
+      "grad_norm": 0.18397556245326996,
+      "learning_rate": 5e-06,
+      "loss": 0.1722,
+      "step": 11161
+    },
+    {
+      "epoch": 8.401957094467445,
+      "grad_norm": 0.14603297412395477,
+      "learning_rate": 5e-06,
+      "loss": 0.0917,
+      "step": 11162
+    },
+    {
+      "epoch": 8.402709823108768,
+      "grad_norm": 0.19652481377124786,
+      "learning_rate": 5e-06,
+      "loss": 0.1815,
+      "step": 11163
+    },
+    {
+      "epoch": 8.403462551750094,
+      "grad_norm": 0.18062511086463928,
+      "learning_rate": 5e-06,
+      "loss": 0.1636,
+      "step": 11164
+    },
+    {
+      "epoch": 8.40421528039142,
+      "grad_norm": 0.16811951994895935,
+      "learning_rate": 5e-06,
+      "loss": 0.1493,
+      "step": 11165
+    },
+    {
+      "epoch": 8.404968009032745,
+      "grad_norm": 0.18884176015853882,
+      "learning_rate": 5e-06,
+      "loss": 0.1569,
+      "step": 11166
+    },
+    {
+      "epoch": 8.405720737674068,
+      "grad_norm": 0.18175804615020752,
+      "learning_rate": 5e-06,
+      "loss": 0.1845,
+      "step": 11167
+    },
+    {
+      "epoch": 8.406473466315393,
+      "grad_norm": 0.20202110707759857,
+      "learning_rate": 5e-06,
+      "loss": 0.1528,
+      "step": 11168
+    },
+    {
+      "epoch": 8.407226194956719,
+      "grad_norm": 0.1643306463956833,
+      "learning_rate": 5e-06,
+      "loss": 0.1666,
+      "step": 11169
+    },
+    {
+      "epoch": 8.407978923598042,
+      "grad_norm": 0.1947314590215683,
+      "learning_rate": 5e-06,
+      "loss": 0.1657,
+      "step": 11170
+    },
+    {
+      "epoch": 8.408731652239368,
+      "grad_norm": 0.15470540523529053,
+      "learning_rate": 5e-06,
+      "loss": 0.097,
+      "step": 11171
+    },
+    {
+      "epoch": 8.409484380880693,
+      "grad_norm": 0.1761370748281479,
+      "learning_rate": 5e-06,
+      "loss": 0.1458,
+      "step": 11172
+    },
+    {
+      "epoch": 8.410237109522017,
+      "grad_norm": 0.16433586180210114,
+      "learning_rate": 5e-06,
+      "loss": 0.1262,
+      "step": 11173
+    },
+    {
+      "epoch": 8.410989838163342,
+      "grad_norm": 0.15021730959415436,
+      "learning_rate": 5e-06,
+      "loss": 0.0901,
+      "step": 11174
+    },
+    {
+      "epoch": 8.411742566804667,
+      "grad_norm": 0.17215389013290405,
+      "learning_rate": 5e-06,
+      "loss": 0.1207,
+      "step": 11175
+    },
+    {
+      "epoch": 8.412495295445991,
+      "grad_norm": 0.1933157742023468,
+      "learning_rate": 5e-06,
+      "loss": 0.2023,
+      "step": 11176
+    },
+    {
+      "epoch": 8.413248024087316,
+      "grad_norm": 0.16012001037597656,
+      "learning_rate": 5e-06,
+      "loss": 0.1242,
+      "step": 11177
+    },
+    {
+      "epoch": 8.414000752728642,
+      "grad_norm": 0.17401725053787231,
+      "learning_rate": 5e-06,
+      "loss": 0.1659,
+      "step": 11178
+    },
+    {
+      "epoch": 8.414753481369965,
+      "grad_norm": 0.15385101735591888,
+      "learning_rate": 5e-06,
+      "loss": 0.1094,
+      "step": 11179
+    },
+    {
+      "epoch": 8.41550621001129,
+      "grad_norm": 0.16413326561450958,
+      "learning_rate": 5e-06,
+      "loss": 0.1383,
+      "step": 11180
+    },
+    {
+      "epoch": 8.416258938652616,
+      "grad_norm": 0.166517972946167,
+      "learning_rate": 5e-06,
+      "loss": 0.1277,
+      "step": 11181
+    },
+    {
+      "epoch": 8.41701166729394,
+      "grad_norm": 0.17401081323623657,
+      "learning_rate": 5e-06,
+      "loss": 0.1677,
+      "step": 11182
+    },
+    {
+      "epoch": 8.417764395935265,
+      "grad_norm": 0.18627037107944489,
+      "learning_rate": 5e-06,
+      "loss": 0.1623,
+      "step": 11183
+    },
+    {
+      "epoch": 8.41851712457659,
+      "grad_norm": 0.2137485295534134,
+      "learning_rate": 5e-06,
+      "loss": 0.1799,
+      "step": 11184
+    },
+    {
+      "epoch": 8.419269853217916,
+      "grad_norm": 0.21773777902126312,
+      "learning_rate": 5e-06,
+      "loss": 0.1377,
+      "step": 11185
+    },
+    {
+      "epoch": 8.42002258185924,
+      "grad_norm": 0.13875602185726166,
+      "learning_rate": 5e-06,
+      "loss": 0.0592,
+      "step": 11186
+    },
+    {
+      "epoch": 8.420775310500565,
+      "grad_norm": 0.19865421950817108,
+      "learning_rate": 5e-06,
+      "loss": 0.1802,
+      "step": 11187
+    },
+    {
+      "epoch": 8.42152803914189,
+      "grad_norm": 0.1878143697977066,
+      "learning_rate": 5e-06,
+      "loss": 0.1251,
+      "step": 11188
+    },
+    {
+      "epoch": 8.422280767783214,
+      "grad_norm": 0.1717860996723175,
+      "learning_rate": 5e-06,
+      "loss": 0.1436,
+      "step": 11189
+    },
+    {
+      "epoch": 8.423033496424539,
+      "grad_norm": 0.19132544100284576,
+      "learning_rate": 5e-06,
+      "loss": 0.2113,
+      "step": 11190
+    },
+    {
+      "epoch": 8.423786225065864,
+      "grad_norm": 0.18597713112831116,
+      "learning_rate": 5e-06,
+      "loss": 0.1619,
+      "step": 11191
+    },
+    {
+      "epoch": 8.424538953707188,
+      "grad_norm": 0.15287895500659943,
+      "learning_rate": 5e-06,
+      "loss": 0.0843,
+      "step": 11192
+    },
+    {
+      "epoch": 8.425291682348513,
+      "grad_norm": 0.1923738718032837,
+      "learning_rate": 5e-06,
+      "loss": 0.1472,
+      "step": 11193
+    },
+    {
+      "epoch": 8.426044410989839,
+      "grad_norm": 0.17207881808280945,
+      "learning_rate": 5e-06,
+      "loss": 0.1946,
+      "step": 11194
+    },
+    {
+      "epoch": 8.426797139631162,
+      "grad_norm": 0.19896718859672546,
+      "learning_rate": 5e-06,
+      "loss": 0.1635,
+      "step": 11195
+    },
+    {
+      "epoch": 8.427549868272488,
+      "grad_norm": 0.20108747482299805,
+      "learning_rate": 5e-06,
+      "loss": 0.1626,
+      "step": 11196
+    },
+    {
+      "epoch": 8.428302596913813,
+      "grad_norm": 0.17759418487548828,
+      "learning_rate": 5e-06,
+      "loss": 0.2113,
+      "step": 11197
+    },
+    {
+      "epoch": 8.429055325555137,
+      "grad_norm": 0.15633781254291534,
+      "learning_rate": 5e-06,
+      "loss": 0.1104,
+      "step": 11198
+    },
+    {
+      "epoch": 8.429808054196462,
+      "grad_norm": 0.16796551644802094,
+      "learning_rate": 5e-06,
+      "loss": 0.0745,
+      "step": 11199
+    },
+    {
+      "epoch": 8.430560782837787,
+      "grad_norm": 0.1663864701986313,
+      "learning_rate": 5e-06,
+      "loss": 0.2135,
+      "step": 11200
+    },
+    {
+      "epoch": 8.430560782837787,
+      "eval_loss": 0.1949397772550583,
+      "eval_runtime": 452.8909,
+      "eval_samples_per_second": 21.257,
+      "eval_steps_per_second": 0.665,
+      "step": 11200
+    },
+    {
+      "epoch": 8.431313511479111,
+      "grad_norm": 0.1772366315126419,
+      "learning_rate": 5e-06,
+      "loss": 0.185,
+      "step": 11201
+    },
+    {
+      "epoch": 8.432066240120436,
+      "grad_norm": 0.18573597073554993,
+      "learning_rate": 5e-06,
+      "loss": 0.2397,
+      "step": 11202
+    },
+    {
+      "epoch": 8.432818968761762,
+      "grad_norm": 0.18835359811782837,
+      "learning_rate": 5e-06,
+      "loss": 0.1442,
+      "step": 11203
+    },
+    {
+      "epoch": 8.433571697403087,
+      "grad_norm": 0.18673522770404816,
+      "learning_rate": 5e-06,
+      "loss": 0.1401,
+      "step": 11204
+    },
+    {
+      "epoch": 8.43432442604441,
+      "grad_norm": 0.19533681869506836,
+      "learning_rate": 5e-06,
+      "loss": 0.1853,
+      "step": 11205
+    },
+    {
+      "epoch": 8.435077154685736,
+      "grad_norm": 0.17201679944992065,
+      "learning_rate": 5e-06,
+      "loss": 0.1245,
+      "step": 11206
+    },
+    {
+      "epoch": 8.435829883327061,
+      "grad_norm": 0.17692914605140686,
+      "learning_rate": 5e-06,
+      "loss": 0.25,
+      "step": 11207
+    },
+    {
+      "epoch": 8.436582611968385,
+      "grad_norm": 0.177909255027771,
+      "learning_rate": 5e-06,
+      "loss": 0.1691,
+      "step": 11208
+    },
+    {
+      "epoch": 8.43733534060971,
+      "grad_norm": 0.16745661199092865,
+      "learning_rate": 5e-06,
+      "loss": 0.1067,
+      "step": 11209
+    },
+    {
+      "epoch": 8.438088069251036,
+      "grad_norm": 0.20880572497844696,
+      "learning_rate": 5e-06,
+      "loss": 0.1125,
+      "step": 11210
+    },
+    {
+      "epoch": 8.43884079789236,
+      "grad_norm": 0.1683230698108673,
+      "learning_rate": 5e-06,
+      "loss": 0.1176,
+      "step": 11211
+    },
+    {
+      "epoch": 8.439593526533685,
+      "grad_norm": 0.17667382955551147,
+      "learning_rate": 5e-06,
+      "loss": 0.1526,
+      "step": 11212
+    },
+    {
+      "epoch": 8.44034625517501,
+      "grad_norm": 0.17150604724884033,
+      "learning_rate": 5e-06,
+      "loss": 0.1671,
+      "step": 11213
+    },
+    {
+      "epoch": 8.441098983816334,
+      "grad_norm": 0.20692507922649384,
+      "learning_rate": 5e-06,
+      "loss": 0.1855,
+      "step": 11214
+    },
+    {
+      "epoch": 8.441851712457659,
+      "grad_norm": 0.1665118932723999,
+      "learning_rate": 5e-06,
+      "loss": 0.0931,
+      "step": 11215
+    },
+    {
+      "epoch": 8.442604441098984,
+      "grad_norm": 0.17363613843917847,
+      "learning_rate": 5e-06,
+      "loss": 0.118,
+      "step": 11216
+    },
+    {
+      "epoch": 8.443357169740308,
+      "grad_norm": 0.21057696640491486,
+      "learning_rate": 5e-06,
+      "loss": 0.1432,
+      "step": 11217
+    },
+    {
+      "epoch": 8.444109898381633,
+      "grad_norm": 0.19384048879146576,
+      "learning_rate": 5e-06,
+      "loss": 0.1598,
+      "step": 11218
+    },
+    {
+      "epoch": 8.444862627022959,
+      "grad_norm": 0.1655665636062622,
+      "learning_rate": 5e-06,
+      "loss": 0.1786,
+      "step": 11219
+    },
+    {
+      "epoch": 8.445615355664284,
+      "grad_norm": 0.18828095495700836,
+      "learning_rate": 5e-06,
+      "loss": 0.1779,
+      "step": 11220
+    },
+    {
+      "epoch": 8.446368084305607,
+      "grad_norm": 0.16776598989963531,
+      "learning_rate": 5e-06,
+      "loss": 0.1369,
+      "step": 11221
+    },
+    {
+      "epoch": 8.447120812946933,
+      "grad_norm": 0.17282681167125702,
+      "learning_rate": 5e-06,
+      "loss": 0.163,
+      "step": 11222
+    },
+    {
+      "epoch": 8.447873541588258,
+      "grad_norm": 0.19399863481521606,
+      "learning_rate": 5e-06,
+      "loss": 0.2033,
+      "step": 11223
+    },
+    {
+      "epoch": 8.448626270229582,
+      "grad_norm": 0.18377827107906342,
+      "learning_rate": 5e-06,
+      "loss": 0.1448,
+      "step": 11224
+    },
+    {
+      "epoch": 8.449378998870907,
+      "grad_norm": 0.17250105738639832,
+      "learning_rate": 5e-06,
+      "loss": 0.1121,
+      "step": 11225
+    },
+    {
+      "epoch": 8.450131727512233,
+      "grad_norm": 0.18329060077667236,
+      "learning_rate": 5e-06,
+      "loss": 0.1333,
+      "step": 11226
+    },
+    {
+      "epoch": 8.450884456153556,
+      "grad_norm": 0.1597832888364792,
+      "learning_rate": 5e-06,
+      "loss": 0.1161,
+      "step": 11227
+    },
+    {
+      "epoch": 8.451637184794881,
+      "grad_norm": 0.19052523374557495,
+      "learning_rate": 5e-06,
+      "loss": 0.1614,
+      "step": 11228
+    },
+    {
+      "epoch": 8.452389913436207,
+      "grad_norm": 0.17982028424739838,
+      "learning_rate": 5e-06,
+      "loss": 0.1257,
+      "step": 11229
+    },
+    {
+      "epoch": 8.45314264207753,
+      "grad_norm": 0.17906947433948517,
+      "learning_rate": 5e-06,
+      "loss": 0.1025,
+      "step": 11230
+    },
+    {
+      "epoch": 8.453895370718856,
+      "grad_norm": 0.17206379771232605,
+      "learning_rate": 5e-06,
+      "loss": 0.0977,
+      "step": 11231
+    },
+    {
+      "epoch": 8.454648099360181,
+      "grad_norm": 0.17871931195259094,
+      "learning_rate": 5e-06,
+      "loss": 0.0965,
+      "step": 11232
+    },
+    {
+      "epoch": 8.455400828001505,
+      "grad_norm": 0.17160283029079437,
+      "learning_rate": 5e-06,
+      "loss": 0.1785,
+      "step": 11233
+    },
+    {
+      "epoch": 8.45615355664283,
+      "grad_norm": 0.16544251143932343,
+      "learning_rate": 5e-06,
+      "loss": 0.0955,
+      "step": 11234
+    },
+    {
+      "epoch": 8.456906285284155,
+      "grad_norm": 0.1968957632780075,
+      "learning_rate": 5e-06,
+      "loss": 0.1888,
+      "step": 11235
+    },
+    {
+      "epoch": 8.457659013925479,
+      "grad_norm": 0.16234752535820007,
+      "learning_rate": 5e-06,
+      "loss": 0.1351,
+      "step": 11236
+    },
+    {
+      "epoch": 8.458411742566804,
+      "grad_norm": 0.15348941087722778,
+      "learning_rate": 5e-06,
+      "loss": 0.1276,
+      "step": 11237
+    },
+    {
+      "epoch": 8.45916447120813,
+      "grad_norm": 0.16275297105312347,
+      "learning_rate": 5e-06,
+      "loss": 0.1962,
+      "step": 11238
+    },
+    {
+      "epoch": 8.459917199849455,
+      "grad_norm": 0.1761731058359146,
+      "learning_rate": 5e-06,
+      "loss": 0.1258,
+      "step": 11239
+    },
+    {
+      "epoch": 8.460669928490779,
+      "grad_norm": 0.179912731051445,
+      "learning_rate": 5e-06,
+      "loss": 0.1557,
+      "step": 11240
+    },
+    {
+      "epoch": 8.461422657132104,
+      "grad_norm": 0.15325690805912018,
+      "learning_rate": 5e-06,
+      "loss": 0.146,
+      "step": 11241
+    },
+    {
+      "epoch": 8.46217538577343,
+      "grad_norm": 0.16031669080257416,
+      "learning_rate": 5e-06,
+      "loss": 0.068,
+      "step": 11242
+    },
+    {
+      "epoch": 8.462928114414753,
+      "grad_norm": 0.19811101257801056,
+      "learning_rate": 5e-06,
+      "loss": 0.1936,
+      "step": 11243
+    },
+    {
+      "epoch": 8.463680843056078,
+      "grad_norm": 0.16754043102264404,
+      "learning_rate": 5e-06,
+      "loss": 0.1343,
+      "step": 11244
+    },
+    {
+      "epoch": 8.464433571697404,
+      "grad_norm": 0.16842305660247803,
+      "learning_rate": 5e-06,
+      "loss": 0.1154,
+      "step": 11245
+    },
+    {
+      "epoch": 8.465186300338727,
+      "grad_norm": 0.18578293919563293,
+      "learning_rate": 5e-06,
+      "loss": 0.1435,
+      "step": 11246
+    },
+    {
+      "epoch": 8.465939028980053,
+      "grad_norm": 0.17664505541324615,
+      "learning_rate": 5e-06,
+      "loss": 0.0894,
+      "step": 11247
+    },
+    {
+      "epoch": 8.466691757621378,
+      "grad_norm": 0.18836906552314758,
+      "learning_rate": 5e-06,
+      "loss": 0.1232,
+      "step": 11248
+    },
+    {
+      "epoch": 8.467444486262702,
+      "grad_norm": 0.15609723329544067,
+      "learning_rate": 5e-06,
+      "loss": 0.1081,
+      "step": 11249
+    },
+    {
+      "epoch": 8.468197214904027,
+      "grad_norm": 0.1903337687253952,
+      "learning_rate": 5e-06,
+      "loss": 0.1727,
+      "step": 11250
+    },
+    {
+      "epoch": 8.468949943545352,
+      "grad_norm": 0.16118167340755463,
+      "learning_rate": 5e-06,
+      "loss": 0.1759,
+      "step": 11251
+    },
+    {
+      "epoch": 8.469702672186676,
+      "grad_norm": 0.20199370384216309,
+      "learning_rate": 5e-06,
+      "loss": 0.2226,
+      "step": 11252
+    },
+    {
+      "epoch": 8.470455400828001,
+      "grad_norm": 0.16716256737709045,
+      "learning_rate": 5e-06,
+      "loss": 0.1372,
+      "step": 11253
+    },
+    {
+      "epoch": 8.471208129469327,
+      "grad_norm": 0.1675744652748108,
+      "learning_rate": 5e-06,
+      "loss": 0.1049,
+      "step": 11254
+    },
+    {
+      "epoch": 8.471960858110652,
+      "grad_norm": 0.16285213828086853,
+      "learning_rate": 5e-06,
+      "loss": 0.1528,
+      "step": 11255
+    },
+    {
+      "epoch": 8.472713586751976,
+      "grad_norm": 0.1973881721496582,
+      "learning_rate": 5e-06,
+      "loss": 0.1582,
+      "step": 11256
+    },
+    {
+      "epoch": 8.473466315393301,
+      "grad_norm": 0.15338030457496643,
+      "learning_rate": 5e-06,
+      "loss": 0.1021,
+      "step": 11257
+    },
+    {
+      "epoch": 8.474219044034626,
+      "grad_norm": 0.17477814853191376,
+      "learning_rate": 5e-06,
+      "loss": 0.1579,
+      "step": 11258
+    },
+    {
+      "epoch": 8.47497177267595,
+      "grad_norm": 0.1546858549118042,
+      "learning_rate": 5e-06,
+      "loss": 0.064,
+      "step": 11259
+    },
+    {
+      "epoch": 8.475724501317275,
+      "grad_norm": 0.17499621212482452,
+      "learning_rate": 5e-06,
+      "loss": 0.2108,
+      "step": 11260
+    },
+    {
+      "epoch": 8.4764772299586,
+      "grad_norm": 0.17075885832309723,
+      "learning_rate": 5e-06,
+      "loss": 0.1786,
+      "step": 11261
+    },
+    {
+      "epoch": 8.477229958599924,
+      "grad_norm": 0.18749092519283295,
+      "learning_rate": 5e-06,
+      "loss": 0.1586,
+      "step": 11262
+    },
+    {
+      "epoch": 8.47798268724125,
+      "grad_norm": 0.18322212994098663,
+      "learning_rate": 5e-06,
+      "loss": 0.1722,
+      "step": 11263
+    },
+    {
+      "epoch": 8.478735415882575,
+      "grad_norm": 0.17481370270252228,
+      "learning_rate": 5e-06,
+      "loss": 0.1207,
+      "step": 11264
+    },
+    {
+      "epoch": 8.479488144523899,
+      "grad_norm": 0.16892766952514648,
+      "learning_rate": 5e-06,
+      "loss": 0.1635,
+      "step": 11265
+    },
+    {
+      "epoch": 8.480240873165224,
+      "grad_norm": 0.1910858154296875,
+      "learning_rate": 5e-06,
+      "loss": 0.156,
+      "step": 11266
+    },
+    {
+      "epoch": 8.48099360180655,
+      "grad_norm": 0.1789727658033371,
+      "learning_rate": 5e-06,
+      "loss": 0.2783,
+      "step": 11267
+    },
+    {
+      "epoch": 8.481746330447873,
+      "grad_norm": 0.16258060932159424,
+      "learning_rate": 5e-06,
+      "loss": 0.1465,
+      "step": 11268
+    },
+    {
+      "epoch": 8.482499059089198,
+      "grad_norm": 0.19626271724700928,
+      "learning_rate": 5e-06,
+      "loss": 0.1749,
+      "step": 11269
+    },
+    {
+      "epoch": 8.483251787730524,
+      "grad_norm": 0.18268834054470062,
+      "learning_rate": 5e-06,
+      "loss": 0.1355,
+      "step": 11270
+    },
+    {
+      "epoch": 8.484004516371847,
+      "grad_norm": 0.16548137366771698,
+      "learning_rate": 5e-06,
+      "loss": 0.1116,
+      "step": 11271
+    },
+    {
+      "epoch": 8.484757245013173,
+      "grad_norm": 0.17995671927928925,
+      "learning_rate": 5e-06,
+      "loss": 0.1301,
+      "step": 11272
+    },
+    {
+      "epoch": 8.485509973654498,
+      "grad_norm": 0.14712229371070862,
+      "learning_rate": 5e-06,
+      "loss": 0.1302,
+      "step": 11273
+    },
+    {
+      "epoch": 8.486262702295821,
+      "grad_norm": 0.17674894630908966,
+      "learning_rate": 5e-06,
+      "loss": 0.1153,
+      "step": 11274
+    },
+    {
+      "epoch": 8.487015430937147,
+      "grad_norm": 0.19749018549919128,
+      "learning_rate": 5e-06,
+      "loss": 0.1488,
+      "step": 11275
+    },
+    {
+      "epoch": 8.487768159578472,
+      "grad_norm": 0.17472365498542786,
+      "learning_rate": 5e-06,
+      "loss": 0.1069,
+      "step": 11276
+    },
+    {
+      "epoch": 8.488520888219798,
+      "grad_norm": 0.19628962874412537,
+      "learning_rate": 5e-06,
+      "loss": 0.2093,
+      "step": 11277
+    },
+    {
+      "epoch": 8.489273616861121,
+      "grad_norm": 0.19072028994560242,
+      "learning_rate": 5e-06,
+      "loss": 0.2391,
+      "step": 11278
+    },
+    {
+      "epoch": 8.490026345502447,
+      "grad_norm": 0.17401820421218872,
+      "learning_rate": 5e-06,
+      "loss": 0.131,
+      "step": 11279
+    },
+    {
+      "epoch": 8.490779074143772,
+      "grad_norm": 0.15460304915905,
+      "learning_rate": 5e-06,
+      "loss": 0.0944,
+      "step": 11280
+    },
+    {
+      "epoch": 8.491531802785095,
+      "grad_norm": 0.16564026474952698,
+      "learning_rate": 5e-06,
+      "loss": 0.2008,
+      "step": 11281
+    },
+    {
+      "epoch": 8.49228453142642,
+      "grad_norm": 0.1666959971189499,
+      "learning_rate": 5e-06,
+      "loss": 0.123,
+      "step": 11282
+    },
+    {
+      "epoch": 8.493037260067746,
+      "grad_norm": 0.17606599628925323,
+      "learning_rate": 5e-06,
+      "loss": 0.1638,
+      "step": 11283
+    },
+    {
+      "epoch": 8.49378998870907,
+      "grad_norm": 0.17597255110740662,
+      "learning_rate": 5e-06,
+      "loss": 0.1336,
+      "step": 11284
+    },
+    {
+      "epoch": 8.494542717350395,
+      "grad_norm": 0.1715712994337082,
+      "learning_rate": 5e-06,
+      "loss": 0.2075,
+      "step": 11285
+    },
+    {
+      "epoch": 8.49529544599172,
+      "grad_norm": 0.2012930065393448,
+      "learning_rate": 5e-06,
+      "loss": 0.1696,
+      "step": 11286
+    },
+    {
+      "epoch": 8.496048174633044,
+      "grad_norm": 0.19203422963619232,
+      "learning_rate": 5e-06,
+      "loss": 0.1462,
+      "step": 11287
+    },
+    {
+      "epoch": 8.49680090327437,
+      "grad_norm": 0.17072857916355133,
+      "learning_rate": 5e-06,
+      "loss": 0.122,
+      "step": 11288
+    },
+    {
+      "epoch": 8.497553631915695,
+      "grad_norm": 0.17201568186283112,
+      "learning_rate": 5e-06,
+      "loss": 0.1385,
+      "step": 11289
+    },
+    {
+      "epoch": 8.498306360557018,
+      "grad_norm": 0.17942199110984802,
+      "learning_rate": 5e-06,
+      "loss": 0.1486,
+      "step": 11290
+    },
+    {
+      "epoch": 8.499059089198344,
+      "grad_norm": 0.18472112715244293,
+      "learning_rate": 5e-06,
+      "loss": 0.2156,
+      "step": 11291
+    },
+    {
+      "epoch": 8.499811817839669,
+      "grad_norm": 0.16491149365901947,
+      "learning_rate": 5e-06,
+      "loss": 0.2238,
+      "step": 11292
+    },
+    {
+      "epoch": 8.500564546480994,
+      "grad_norm": 0.1648290604352951,
+      "learning_rate": 5e-06,
+      "loss": 0.0981,
+      "step": 11293
+    },
+    {
+      "epoch": 8.501317275122318,
+      "grad_norm": 0.17943470180034637,
+      "learning_rate": 5e-06,
+      "loss": 0.2307,
+      "step": 11294
+    },
+    {
+      "epoch": 8.502070003763643,
+      "grad_norm": 0.19792670011520386,
+      "learning_rate": 5e-06,
+      "loss": 0.1935,
+      "step": 11295
+    },
+    {
+      "epoch": 8.502822732404969,
+      "grad_norm": 0.1777549535036087,
+      "learning_rate": 5e-06,
+      "loss": 0.1645,
+      "step": 11296
+    },
+    {
+      "epoch": 8.503575461046292,
+      "grad_norm": 0.1971159726381302,
+      "learning_rate": 5e-06,
+      "loss": 0.1454,
+      "step": 11297
+    },
+    {
+      "epoch": 8.504328189687618,
+      "grad_norm": 0.19153417646884918,
+      "learning_rate": 5e-06,
+      "loss": 0.1759,
+      "step": 11298
+    },
+    {
+      "epoch": 8.505080918328943,
+      "grad_norm": 0.19767266511917114,
+      "learning_rate": 5e-06,
+      "loss": 0.1643,
+      "step": 11299
+    },
+    {
+      "epoch": 8.505833646970267,
+      "grad_norm": 0.14840786159038544,
+      "learning_rate": 5e-06,
+      "loss": 0.1293,
+      "step": 11300
+    },
+    {
+      "epoch": 8.506586375611592,
+      "grad_norm": 0.1850418597459793,
+      "learning_rate": 5e-06,
+      "loss": 0.2154,
+      "step": 11301
+    },
+    {
+      "epoch": 8.507339104252917,
+      "grad_norm": 0.1997561901807785,
+      "learning_rate": 5e-06,
+      "loss": 0.1132,
+      "step": 11302
+    },
+    {
+      "epoch": 8.508091832894241,
+      "grad_norm": 0.17268769443035126,
+      "learning_rate": 5e-06,
+      "loss": 0.1889,
+      "step": 11303
+    },
+    {
+      "epoch": 8.508844561535566,
+      "grad_norm": 0.17487579584121704,
+      "learning_rate": 5e-06,
+      "loss": 0.191,
+      "step": 11304
+    },
+    {
+      "epoch": 8.509597290176892,
+      "grad_norm": 0.1814659833908081,
+      "learning_rate": 5e-06,
+      "loss": 0.1898,
+      "step": 11305
+    },
+    {
+      "epoch": 8.510350018818215,
+      "grad_norm": 0.1862671822309494,
+      "learning_rate": 5e-06,
+      "loss": 0.152,
+      "step": 11306
+    },
+    {
+      "epoch": 8.51110274745954,
+      "grad_norm": 0.14932675659656525,
+      "learning_rate": 5e-06,
+      "loss": 0.0836,
+      "step": 11307
+    },
+    {
+      "epoch": 8.511855476100866,
+      "grad_norm": 0.16370300948619843,
+      "learning_rate": 5e-06,
+      "loss": 0.1754,
+      "step": 11308
+    },
+    {
+      "epoch": 8.51260820474219,
+      "grad_norm": 0.1798919290304184,
+      "learning_rate": 5e-06,
+      "loss": 0.1541,
+      "step": 11309
+    },
+    {
+      "epoch": 8.513360933383515,
+      "grad_norm": 0.1771814376115799,
+      "learning_rate": 5e-06,
+      "loss": 0.1363,
+      "step": 11310
+    },
+    {
+      "epoch": 8.51411366202484,
+      "grad_norm": 0.17049679160118103,
+      "learning_rate": 5e-06,
+      "loss": 0.128,
+      "step": 11311
+    },
+    {
+      "epoch": 8.514866390666164,
+      "grad_norm": 0.1785392463207245,
+      "learning_rate": 5e-06,
+      "loss": 0.1954,
+      "step": 11312
+    },
+    {
+      "epoch": 8.51561911930749,
+      "grad_norm": 0.17913012206554413,
+      "learning_rate": 5e-06,
+      "loss": 0.2308,
+      "step": 11313
+    },
+    {
+      "epoch": 8.516371847948815,
+      "grad_norm": 0.16916127502918243,
+      "learning_rate": 5e-06,
+      "loss": 0.1544,
+      "step": 11314
+    },
+    {
+      "epoch": 8.51712457659014,
+      "grad_norm": 0.17164994776248932,
+      "learning_rate": 5e-06,
+      "loss": 0.2095,
+      "step": 11315
+    },
+    {
+      "epoch": 8.517877305231464,
+      "grad_norm": 0.1787753850221634,
+      "learning_rate": 5e-06,
+      "loss": 0.213,
+      "step": 11316
+    },
+    {
+      "epoch": 8.518630033872789,
+      "grad_norm": 0.15688453614711761,
+      "learning_rate": 5e-06,
+      "loss": 0.1195,
+      "step": 11317
+    },
+    {
+      "epoch": 8.519382762514114,
+      "grad_norm": 0.1766117364168167,
+      "learning_rate": 5e-06,
+      "loss": 0.1255,
+      "step": 11318
+    },
+    {
+      "epoch": 8.520135491155438,
+      "grad_norm": 0.17056958377361298,
+      "learning_rate": 5e-06,
+      "loss": 0.1255,
+      "step": 11319
+    },
+    {
+      "epoch": 8.520888219796763,
+      "grad_norm": 0.17584750056266785,
+      "learning_rate": 5e-06,
+      "loss": 0.1497,
+      "step": 11320
+    },
+    {
+      "epoch": 8.521640948438089,
+      "grad_norm": 0.18644702434539795,
+      "learning_rate": 5e-06,
+      "loss": 0.1694,
+      "step": 11321
+    },
+    {
+      "epoch": 8.522393677079412,
+      "grad_norm": 0.17894309759140015,
+      "learning_rate": 5e-06,
+      "loss": 0.1068,
+      "step": 11322
+    },
+    {
+      "epoch": 8.523146405720738,
+      "grad_norm": 0.1597880870103836,
+      "learning_rate": 5e-06,
+      "loss": 0.1447,
+      "step": 11323
+    },
+    {
+      "epoch": 8.523899134362063,
+      "grad_norm": 0.15783676505088806,
+      "learning_rate": 5e-06,
+      "loss": 0.2156,
+      "step": 11324
+    },
+    {
+      "epoch": 8.524651863003387,
+      "grad_norm": 0.17795048654079437,
+      "learning_rate": 5e-06,
+      "loss": 0.2235,
+      "step": 11325
+    },
+    {
+      "epoch": 8.525404591644712,
+      "grad_norm": 0.14452147483825684,
+      "learning_rate": 5e-06,
+      "loss": 0.1015,
+      "step": 11326
+    },
+    {
+      "epoch": 8.526157320286037,
+      "grad_norm": 0.19353477656841278,
+      "learning_rate": 5e-06,
+      "loss": 0.195,
+      "step": 11327
+    },
+    {
+      "epoch": 8.526910048927363,
+      "grad_norm": 0.19640696048736572,
+      "learning_rate": 5e-06,
+      "loss": 0.1429,
+      "step": 11328
+    },
+    {
+      "epoch": 8.527662777568686,
+      "grad_norm": 0.1729712188243866,
+      "learning_rate": 5e-06,
+      "loss": 0.1367,
+      "step": 11329
+    },
+    {
+      "epoch": 8.528415506210012,
+      "grad_norm": 0.1551416963338852,
+      "learning_rate": 5e-06,
+      "loss": 0.092,
+      "step": 11330
+    },
+    {
+      "epoch": 8.529168234851337,
+      "grad_norm": 0.15890157222747803,
+      "learning_rate": 5e-06,
+      "loss": 0.12,
+      "step": 11331
+    },
+    {
+      "epoch": 8.52992096349266,
+      "grad_norm": 0.15337473154067993,
+      "learning_rate": 5e-06,
+      "loss": 0.1643,
+      "step": 11332
+    },
+    {
+      "epoch": 8.530673692133986,
+      "grad_norm": 0.167642280459404,
+      "learning_rate": 5e-06,
+      "loss": 0.1706,
+      "step": 11333
+    },
+    {
+      "epoch": 8.531426420775311,
+      "grad_norm": 0.19506096839904785,
+      "learning_rate": 5e-06,
+      "loss": 0.2332,
+      "step": 11334
+    },
+    {
+      "epoch": 8.532179149416635,
+      "grad_norm": 0.17994891107082367,
+      "learning_rate": 5e-06,
+      "loss": 0.1763,
+      "step": 11335
+    },
+    {
+      "epoch": 8.53293187805796,
+      "grad_norm": 0.18111246824264526,
+      "learning_rate": 5e-06,
+      "loss": 0.1361,
+      "step": 11336
+    },
+    {
+      "epoch": 8.533684606699286,
+      "grad_norm": 0.16485299170017242,
+      "learning_rate": 5e-06,
+      "loss": 0.1174,
+      "step": 11337
+    },
+    {
+      "epoch": 8.534437335340609,
+      "grad_norm": 0.16698047518730164,
+      "learning_rate": 5e-06,
+      "loss": 0.1005,
+      "step": 11338
+    },
+    {
+      "epoch": 8.535190063981934,
+      "grad_norm": 0.17713086307048798,
+      "learning_rate": 5e-06,
+      "loss": 0.1763,
+      "step": 11339
+    },
+    {
+      "epoch": 8.53594279262326,
+      "grad_norm": 0.19514504075050354,
+      "learning_rate": 5e-06,
+      "loss": 0.1978,
+      "step": 11340
+    },
+    {
+      "epoch": 8.536695521264583,
+      "grad_norm": 0.1678309589624405,
+      "learning_rate": 5e-06,
+      "loss": 0.1773,
+      "step": 11341
+    },
+    {
+      "epoch": 8.537448249905909,
+      "grad_norm": 0.16383737325668335,
+      "learning_rate": 5e-06,
+      "loss": 0.1279,
+      "step": 11342
+    },
+    {
+      "epoch": 8.538200978547234,
+      "grad_norm": 0.22369453310966492,
+      "learning_rate": 5e-06,
+      "loss": 0.2263,
+      "step": 11343
+    },
+    {
+      "epoch": 8.538953707188558,
+      "grad_norm": 0.20071621239185333,
+      "learning_rate": 5e-06,
+      "loss": 0.1443,
+      "step": 11344
+    },
+    {
+      "epoch": 8.539706435829883,
+      "grad_norm": 0.18960809707641602,
+      "learning_rate": 5e-06,
+      "loss": 0.1476,
+      "step": 11345
+    },
+    {
+      "epoch": 8.540459164471208,
+      "grad_norm": 0.17722751200199127,
+      "learning_rate": 5e-06,
+      "loss": 0.1235,
+      "step": 11346
+    },
+    {
+      "epoch": 8.541211893112532,
+      "grad_norm": 0.1789807677268982,
+      "learning_rate": 5e-06,
+      "loss": 0.2156,
+      "step": 11347
+    },
+    {
+      "epoch": 8.541964621753857,
+      "grad_norm": 0.17920222878456116,
+      "learning_rate": 5e-06,
+      "loss": 0.1134,
+      "step": 11348
+    },
+    {
+      "epoch": 8.542717350395183,
+      "grad_norm": 0.19392059743404388,
+      "learning_rate": 5e-06,
+      "loss": 0.1855,
+      "step": 11349
+    },
+    {
+      "epoch": 8.543470079036508,
+      "grad_norm": 0.1641637533903122,
+      "learning_rate": 5e-06,
+      "loss": 0.1249,
+      "step": 11350
+    },
+    {
+      "epoch": 8.544222807677832,
+      "grad_norm": 0.1673748642206192,
+      "learning_rate": 5e-06,
+      "loss": 0.1141,
+      "step": 11351
+    },
+    {
+      "epoch": 8.544975536319157,
+      "grad_norm": 0.19397848844528198,
+      "learning_rate": 5e-06,
+      "loss": 0.2005,
+      "step": 11352
+    },
+    {
+      "epoch": 8.545728264960482,
+      "grad_norm": 0.17721404135227203,
+      "learning_rate": 5e-06,
+      "loss": 0.1563,
+      "step": 11353
+    },
+    {
+      "epoch": 8.546480993601806,
+      "grad_norm": 0.2026022970676422,
+      "learning_rate": 5e-06,
+      "loss": 0.1595,
+      "step": 11354
+    },
+    {
+      "epoch": 8.547233722243131,
+      "grad_norm": 0.17273041605949402,
+      "learning_rate": 5e-06,
+      "loss": 0.1785,
+      "step": 11355
+    },
+    {
+      "epoch": 8.547986450884457,
+      "grad_norm": 0.16318009793758392,
+      "learning_rate": 5e-06,
+      "loss": 0.1186,
+      "step": 11356
+    },
+    {
+      "epoch": 8.54873917952578,
+      "grad_norm": 0.16952666640281677,
+      "learning_rate": 5e-06,
+      "loss": 0.2117,
+      "step": 11357
+    },
+    {
+      "epoch": 8.549491908167106,
+      "grad_norm": 0.1696176826953888,
+      "learning_rate": 5e-06,
+      "loss": 0.2222,
+      "step": 11358
+    },
+    {
+      "epoch": 8.550244636808431,
+      "grad_norm": 0.19320346415042877,
+      "learning_rate": 5e-06,
+      "loss": 0.1052,
+      "step": 11359
+    },
+    {
+      "epoch": 8.550997365449755,
+      "grad_norm": 0.1874830573797226,
+      "learning_rate": 5e-06,
+      "loss": 0.2122,
+      "step": 11360
+    },
+    {
+      "epoch": 8.55175009409108,
+      "grad_norm": 0.17460043728351593,
+      "learning_rate": 5e-06,
+      "loss": 0.1751,
+      "step": 11361
+    },
+    {
+      "epoch": 8.552502822732405,
+      "grad_norm": 0.1880190074443817,
+      "learning_rate": 5e-06,
+      "loss": 0.2273,
+      "step": 11362
+    },
+    {
+      "epoch": 8.553255551373729,
+      "grad_norm": 0.1571013331413269,
+      "learning_rate": 5e-06,
+      "loss": 0.1205,
+      "step": 11363
+    },
+    {
+      "epoch": 8.554008280015054,
+      "grad_norm": 0.17279693484306335,
+      "learning_rate": 5e-06,
+      "loss": 0.1807,
+      "step": 11364
+    },
+    {
+      "epoch": 8.55476100865638,
+      "grad_norm": 0.19891144335269928,
+      "learning_rate": 5e-06,
+      "loss": 0.1249,
+      "step": 11365
+    },
+    {
+      "epoch": 8.555513737297705,
+      "grad_norm": 0.19624356925487518,
+      "learning_rate": 5e-06,
+      "loss": 0.1468,
+      "step": 11366
+    },
+    {
+      "epoch": 8.556266465939029,
+      "grad_norm": 0.17273813486099243,
+      "learning_rate": 5e-06,
+      "loss": 0.1297,
+      "step": 11367
+    },
+    {
+      "epoch": 8.557019194580354,
+      "grad_norm": 0.1849258542060852,
+      "learning_rate": 5e-06,
+      "loss": 0.1704,
+      "step": 11368
+    },
+    {
+      "epoch": 8.55777192322168,
+      "grad_norm": 0.18634352087974548,
+      "learning_rate": 5e-06,
+      "loss": 0.1086,
+      "step": 11369
+    },
+    {
+      "epoch": 8.558524651863003,
+      "grad_norm": 0.18552805483341217,
+      "learning_rate": 5e-06,
+      "loss": 0.2451,
+      "step": 11370
+    },
+    {
+      "epoch": 8.559277380504328,
+      "grad_norm": 0.16005554795265198,
+      "learning_rate": 5e-06,
+      "loss": 0.1164,
+      "step": 11371
+    },
+    {
+      "epoch": 8.560030109145654,
+      "grad_norm": 0.17147855460643768,
+      "learning_rate": 5e-06,
+      "loss": 0.0848,
+      "step": 11372
+    },
+    {
+      "epoch": 8.560782837786977,
+      "grad_norm": 0.17575836181640625,
+      "learning_rate": 5e-06,
+      "loss": 0.1744,
+      "step": 11373
+    },
+    {
+      "epoch": 8.561535566428303,
+      "grad_norm": 0.19313274323940277,
+      "learning_rate": 5e-06,
+      "loss": 0.1583,
+      "step": 11374
+    },
+    {
+      "epoch": 8.562288295069628,
+      "grad_norm": 0.18360058963298798,
+      "learning_rate": 5e-06,
+      "loss": 0.1657,
+      "step": 11375
+    },
+    {
+      "epoch": 8.563041023710952,
+      "grad_norm": 0.18885838985443115,
+      "learning_rate": 5e-06,
+      "loss": 0.1159,
+      "step": 11376
+    },
+    {
+      "epoch": 8.563793752352277,
+      "grad_norm": 0.20081937313079834,
+      "learning_rate": 5e-06,
+      "loss": 0.1572,
+      "step": 11377
+    },
+    {
+      "epoch": 8.564546480993602,
+      "grad_norm": 0.1504679173231125,
+      "learning_rate": 5e-06,
+      "loss": 0.1778,
+      "step": 11378
+    },
+    {
+      "epoch": 8.565299209634926,
+      "grad_norm": 0.20227310061454773,
+      "learning_rate": 5e-06,
+      "loss": 0.1873,
+      "step": 11379
+    },
+    {
+      "epoch": 8.566051938276251,
+      "grad_norm": 0.17509128153324127,
+      "learning_rate": 5e-06,
+      "loss": 0.1412,
+      "step": 11380
+    },
+    {
+      "epoch": 8.566804666917577,
+      "grad_norm": 0.18683962523937225,
+      "learning_rate": 5e-06,
+      "loss": 0.0967,
+      "step": 11381
+    },
+    {
+      "epoch": 8.5675573955589,
+      "grad_norm": 0.1784738302230835,
+      "learning_rate": 5e-06,
+      "loss": 0.193,
+      "step": 11382
+    },
+    {
+      "epoch": 8.568310124200226,
+      "grad_norm": 0.1630382537841797,
+      "learning_rate": 5e-06,
+      "loss": 0.0705,
+      "step": 11383
+    },
+    {
+      "epoch": 8.569062852841551,
+      "grad_norm": 0.19708123803138733,
+      "learning_rate": 5e-06,
+      "loss": 0.1726,
+      "step": 11384
+    },
+    {
+      "epoch": 8.569815581482876,
+      "grad_norm": 0.1740908920764923,
+      "learning_rate": 5e-06,
+      "loss": 0.208,
+      "step": 11385
+    },
+    {
+      "epoch": 8.5705683101242,
+      "grad_norm": 0.1791400909423828,
+      "learning_rate": 5e-06,
+      "loss": 0.2177,
+      "step": 11386
+    },
+    {
+      "epoch": 8.571321038765525,
+      "grad_norm": 0.165521040558815,
+      "learning_rate": 5e-06,
+      "loss": 0.1432,
+      "step": 11387
+    },
+    {
+      "epoch": 8.57207376740685,
+      "grad_norm": 0.1732538938522339,
+      "learning_rate": 5e-06,
+      "loss": 0.1578,
+      "step": 11388
+    },
+    {
+      "epoch": 8.572826496048174,
+      "grad_norm": 0.17469194531440735,
+      "learning_rate": 5e-06,
+      "loss": 0.1207,
+      "step": 11389
+    },
+    {
+      "epoch": 8.5735792246895,
+      "grad_norm": 0.15996406972408295,
+      "learning_rate": 5e-06,
+      "loss": 0.1128,
+      "step": 11390
+    },
+    {
+      "epoch": 8.574331953330825,
+      "grad_norm": 0.18981397151947021,
+      "learning_rate": 5e-06,
+      "loss": 0.153,
+      "step": 11391
+    },
+    {
+      "epoch": 8.575084681972148,
+      "grad_norm": 0.17507779598236084,
+      "learning_rate": 5e-06,
+      "loss": 0.1328,
+      "step": 11392
+    },
+    {
+      "epoch": 8.575837410613474,
+      "grad_norm": 0.18244794011116028,
+      "learning_rate": 5e-06,
+      "loss": 0.2014,
+      "step": 11393
+    },
+    {
+      "epoch": 8.5765901392548,
+      "grad_norm": 0.18919619917869568,
+      "learning_rate": 5e-06,
+      "loss": 0.1573,
+      "step": 11394
+    },
+    {
+      "epoch": 8.577342867896123,
+      "grad_norm": 0.18050818145275116,
+      "learning_rate": 5e-06,
+      "loss": 0.1572,
+      "step": 11395
+    },
+    {
+      "epoch": 8.578095596537448,
+      "grad_norm": 0.17140372097492218,
+      "learning_rate": 5e-06,
+      "loss": 0.1336,
+      "step": 11396
+    },
+    {
+      "epoch": 8.578848325178773,
+      "grad_norm": 0.17198918759822845,
+      "learning_rate": 5e-06,
+      "loss": 0.2018,
+      "step": 11397
+    },
+    {
+      "epoch": 8.579601053820097,
+      "grad_norm": 0.18288381397724152,
+      "learning_rate": 5e-06,
+      "loss": 0.1041,
+      "step": 11398
+    },
+    {
+      "epoch": 8.580353782461422,
+      "grad_norm": 0.18016619980335236,
+      "learning_rate": 5e-06,
+      "loss": 0.1259,
+      "step": 11399
+    },
+    {
+      "epoch": 8.581106511102748,
+      "grad_norm": 0.183480903506279,
+      "learning_rate": 5e-06,
+      "loss": 0.1914,
+      "step": 11400
+    },
+    {
+      "epoch": 8.581106511102748,
+      "eval_loss": 0.19513721764087677,
+      "eval_runtime": 452.83,
+      "eval_samples_per_second": 21.26,
+      "eval_steps_per_second": 0.665,
+      "step": 11400
+    },
+    {
+      "epoch": 8.581859239744073,
+      "grad_norm": 0.17412447929382324,
+      "learning_rate": 5e-06,
+      "loss": 0.1723,
+      "step": 11401
+    },
+    {
+      "epoch": 8.582611968385397,
+      "grad_norm": 0.18699198961257935,
+      "learning_rate": 5e-06,
+      "loss": 0.2023,
+      "step": 11402
+    },
+    {
+      "epoch": 8.583364697026722,
+      "grad_norm": 0.1625889539718628,
+      "learning_rate": 5e-06,
+      "loss": 0.1548,
+      "step": 11403
+    },
+    {
+      "epoch": 8.584117425668047,
+      "grad_norm": 0.17415238916873932,
+      "learning_rate": 5e-06,
+      "loss": 0.1547,
+      "step": 11404
+    },
+    {
+      "epoch": 8.584870154309371,
+      "grad_norm": 0.16461002826690674,
+      "learning_rate": 5e-06,
+      "loss": 0.1672,
+      "step": 11405
+    },
+    {
+      "epoch": 8.585622882950696,
+      "grad_norm": 0.1502087265253067,
+      "learning_rate": 5e-06,
+      "loss": 0.1115,
+      "step": 11406
+    },
+    {
+      "epoch": 8.586375611592022,
+      "grad_norm": 0.168661966919899,
+      "learning_rate": 5e-06,
+      "loss": 0.1199,
+      "step": 11407
+    },
+    {
+      "epoch": 8.587128340233345,
+      "grad_norm": 0.18326051533222198,
+      "learning_rate": 5e-06,
+      "loss": 0.132,
+      "step": 11408
+    },
+    {
+      "epoch": 8.58788106887467,
+      "grad_norm": 0.18357670307159424,
+      "learning_rate": 5e-06,
+      "loss": 0.206,
+      "step": 11409
+    },
+    {
+      "epoch": 8.588633797515996,
+      "grad_norm": 0.19694103300571442,
+      "learning_rate": 5e-06,
+      "loss": 0.1511,
+      "step": 11410
+    },
+    {
+      "epoch": 8.58938652615732,
+      "grad_norm": 0.1490551233291626,
+      "learning_rate": 5e-06,
+      "loss": 0.1378,
+      "step": 11411
+    },
+    {
+      "epoch": 8.590139254798645,
+      "grad_norm": 0.19134385883808136,
+      "learning_rate": 5e-06,
+      "loss": 0.2132,
+      "step": 11412
+    },
+    {
+      "epoch": 8.59089198343997,
+      "grad_norm": 0.18635550141334534,
+      "learning_rate": 5e-06,
+      "loss": 0.119,
+      "step": 11413
+    },
+    {
+      "epoch": 8.591644712081294,
+      "grad_norm": 0.19997571408748627,
+      "learning_rate": 5e-06,
+      "loss": 0.1534,
+      "step": 11414
+    },
+    {
+      "epoch": 8.59239744072262,
+      "grad_norm": 0.19177241623401642,
+      "learning_rate": 5e-06,
+      "loss": 0.1498,
+      "step": 11415
+    },
+    {
+      "epoch": 8.593150169363945,
+      "grad_norm": 0.17763429880142212,
+      "learning_rate": 5e-06,
+      "loss": 0.1597,
+      "step": 11416
+    },
+    {
+      "epoch": 8.593902898005268,
+      "grad_norm": 0.17597289383411407,
+      "learning_rate": 5e-06,
+      "loss": 0.0935,
+      "step": 11417
+    },
+    {
+      "epoch": 8.594655626646594,
+      "grad_norm": 0.19342632591724396,
+      "learning_rate": 5e-06,
+      "loss": 0.2547,
+      "step": 11418
+    },
+    {
+      "epoch": 8.595408355287919,
+      "grad_norm": 0.1617271900177002,
+      "learning_rate": 5e-06,
+      "loss": 0.1372,
+      "step": 11419
+    },
+    {
+      "epoch": 8.596161083929243,
+      "grad_norm": 0.17886410653591156,
+      "learning_rate": 5e-06,
+      "loss": 0.1676,
+      "step": 11420
+    },
+    {
+      "epoch": 8.596913812570568,
+      "grad_norm": 0.15053698420524597,
+      "learning_rate": 5e-06,
+      "loss": 0.1156,
+      "step": 11421
+    },
+    {
+      "epoch": 8.597666541211893,
+      "grad_norm": 0.17399388551712036,
+      "learning_rate": 5e-06,
+      "loss": 0.1375,
+      "step": 11422
+    },
+    {
+      "epoch": 8.598419269853219,
+      "grad_norm": 0.15048864483833313,
+      "learning_rate": 5e-06,
+      "loss": 0.1662,
+      "step": 11423
+    },
+    {
+      "epoch": 8.599171998494542,
+      "grad_norm": 0.17066171765327454,
+      "learning_rate": 5e-06,
+      "loss": 0.1504,
+      "step": 11424
+    },
+    {
+      "epoch": 8.599924727135868,
+      "grad_norm": 0.20302227139472961,
+      "learning_rate": 5e-06,
+      "loss": 0.1392,
+      "step": 11425
+    },
+    {
+      "epoch": 8.600677455777193,
+      "grad_norm": 0.18629521131515503,
+      "learning_rate": 5e-06,
+      "loss": 0.1252,
+      "step": 11426
+    },
+    {
+      "epoch": 8.601430184418517,
+      "grad_norm": 0.19982405006885529,
+      "learning_rate": 5e-06,
+      "loss": 0.1375,
+      "step": 11427
+    },
+    {
+      "epoch": 8.602182913059842,
+      "grad_norm": 0.16598166525363922,
+      "learning_rate": 5e-06,
+      "loss": 0.113,
+      "step": 11428
+    },
+    {
+      "epoch": 8.602935641701167,
+      "grad_norm": 0.15098942816257477,
+      "learning_rate": 5e-06,
+      "loss": 0.1105,
+      "step": 11429
+    },
+    {
+      "epoch": 8.603688370342491,
+      "grad_norm": 0.1857132613658905,
+      "learning_rate": 5e-06,
+      "loss": 0.1843,
+      "step": 11430
+    },
+    {
+      "epoch": 8.604441098983816,
+      "grad_norm": 0.16452182829380035,
+      "learning_rate": 5e-06,
+      "loss": 0.1331,
+      "step": 11431
+    },
+    {
+      "epoch": 8.605193827625142,
+      "grad_norm": 0.1940489560365677,
+      "learning_rate": 5e-06,
+      "loss": 0.1881,
+      "step": 11432
+    },
+    {
+      "epoch": 8.605946556266465,
+      "grad_norm": 0.1672956943511963,
+      "learning_rate": 5e-06,
+      "loss": 0.1232,
+      "step": 11433
+    },
+    {
+      "epoch": 8.60669928490779,
+      "grad_norm": 0.18360552191734314,
+      "learning_rate": 5e-06,
+      "loss": 0.157,
+      "step": 11434
+    },
+    {
+      "epoch": 8.607452013549116,
+      "grad_norm": 0.19415254890918732,
+      "learning_rate": 5e-06,
+      "loss": 0.1393,
+      "step": 11435
+    },
+    {
+      "epoch": 8.60820474219044,
+      "grad_norm": 0.1761964112520218,
+      "learning_rate": 5e-06,
+      "loss": 0.1248,
+      "step": 11436
+    },
+    {
+      "epoch": 8.608957470831765,
+      "grad_norm": 0.17533673346042633,
+      "learning_rate": 5e-06,
+      "loss": 0.133,
+      "step": 11437
+    },
+    {
+      "epoch": 8.60971019947309,
+      "grad_norm": 0.17509356141090393,
+      "learning_rate": 5e-06,
+      "loss": 0.1286,
+      "step": 11438
+    },
+    {
+      "epoch": 8.610462928114416,
+      "grad_norm": 0.1640540361404419,
+      "learning_rate": 5e-06,
+      "loss": 0.1034,
+      "step": 11439
+    },
+    {
+      "epoch": 8.61121565675574,
+      "grad_norm": 0.16864259541034698,
+      "learning_rate": 5e-06,
+      "loss": 0.1275,
+      "step": 11440
+    },
+    {
+      "epoch": 8.611968385397065,
+      "grad_norm": 0.1741812527179718,
+      "learning_rate": 5e-06,
+      "loss": 0.1672,
+      "step": 11441
+    },
+    {
+      "epoch": 8.61272111403839,
+      "grad_norm": 0.18165290355682373,
+      "learning_rate": 5e-06,
+      "loss": 0.2073,
+      "step": 11442
+    },
+    {
+      "epoch": 8.613473842679713,
+      "grad_norm": 0.17803309857845306,
+      "learning_rate": 5e-06,
+      "loss": 0.1577,
+      "step": 11443
+    },
+    {
+      "epoch": 8.614226571321039,
+      "grad_norm": 0.1534479558467865,
+      "learning_rate": 5e-06,
+      "loss": 0.1469,
+      "step": 11444
+    },
+    {
+      "epoch": 8.614979299962364,
+      "grad_norm": 0.1806640475988388,
+      "learning_rate": 5e-06,
+      "loss": 0.2285,
+      "step": 11445
+    },
+    {
+      "epoch": 8.615732028603688,
+      "grad_norm": 0.16727712750434875,
+      "learning_rate": 5e-06,
+      "loss": 0.1289,
+      "step": 11446
+    },
+    {
+      "epoch": 8.616484757245013,
+      "grad_norm": 0.1679805964231491,
+      "learning_rate": 5e-06,
+      "loss": 0.2045,
+      "step": 11447
+    },
+    {
+      "epoch": 8.617237485886339,
+      "grad_norm": 0.19871145486831665,
+      "learning_rate": 5e-06,
+      "loss": 0.1664,
+      "step": 11448
+    },
+    {
+      "epoch": 8.617990214527662,
+      "grad_norm": 0.16007335484027863,
+      "learning_rate": 5e-06,
+      "loss": 0.1693,
+      "step": 11449
+    },
+    {
+      "epoch": 8.618742943168987,
+      "grad_norm": 0.1516086906194687,
+      "learning_rate": 5e-06,
+      "loss": 0.1601,
+      "step": 11450
+    },
+    {
+      "epoch": 8.619495671810313,
+      "grad_norm": 0.16230417788028717,
+      "learning_rate": 5e-06,
+      "loss": 0.1163,
+      "step": 11451
+    },
+    {
+      "epoch": 8.620248400451636,
+      "grad_norm": 0.19119225442409515,
+      "learning_rate": 5e-06,
+      "loss": 0.1611,
+      "step": 11452
+    },
+    {
+      "epoch": 8.621001129092962,
+      "grad_norm": 0.16915260255336761,
+      "learning_rate": 5e-06,
+      "loss": 0.1452,
+      "step": 11453
+    },
+    {
+      "epoch": 8.621753857734287,
+      "grad_norm": 0.18227896094322205,
+      "learning_rate": 5e-06,
+      "loss": 0.1801,
+      "step": 11454
+    },
+    {
+      "epoch": 8.62250658637561,
+      "grad_norm": 0.1766354739665985,
+      "learning_rate": 5e-06,
+      "loss": 0.1822,
+      "step": 11455
+    },
+    {
+      "epoch": 8.623259315016936,
+      "grad_norm": 0.17780008912086487,
+      "learning_rate": 5e-06,
+      "loss": 0.1434,
+      "step": 11456
+    },
+    {
+      "epoch": 8.624012043658261,
+      "grad_norm": 0.1567412167787552,
+      "learning_rate": 5e-06,
+      "loss": 0.1116,
+      "step": 11457
+    },
+    {
+      "epoch": 8.624764772299587,
+      "grad_norm": 0.15317733585834503,
+      "learning_rate": 5e-06,
+      "loss": 0.0897,
+      "step": 11458
+    },
+    {
+      "epoch": 8.62551750094091,
+      "grad_norm": 0.18116892874240875,
+      "learning_rate": 5e-06,
+      "loss": 0.1999,
+      "step": 11459
+    },
+    {
+      "epoch": 8.626270229582236,
+      "grad_norm": 0.1611555963754654,
+      "learning_rate": 5e-06,
+      "loss": 0.0739,
+      "step": 11460
+    },
+    {
+      "epoch": 8.627022958223561,
+      "grad_norm": 0.1751524657011032,
+      "learning_rate": 5e-06,
+      "loss": 0.1744,
+      "step": 11461
+    },
+    {
+      "epoch": 8.627775686864885,
+      "grad_norm": 0.2142149657011032,
+      "learning_rate": 5e-06,
+      "loss": 0.188,
+      "step": 11462
+    },
+    {
+      "epoch": 8.62852841550621,
+      "grad_norm": 0.1760539412498474,
+      "learning_rate": 5e-06,
+      "loss": 0.1313,
+      "step": 11463
+    },
+    {
+      "epoch": 8.629281144147535,
+      "grad_norm": 0.19913864135742188,
+      "learning_rate": 5e-06,
+      "loss": 0.127,
+      "step": 11464
+    },
+    {
+      "epoch": 8.630033872788859,
+      "grad_norm": 0.1742306351661682,
+      "learning_rate": 5e-06,
+      "loss": 0.173,
+      "step": 11465
+    },
+    {
+      "epoch": 8.630786601430184,
+      "grad_norm": 0.15365144610404968,
+      "learning_rate": 5e-06,
+      "loss": 0.1382,
+      "step": 11466
+    },
+    {
+      "epoch": 8.63153933007151,
+      "grad_norm": 0.18592211604118347,
+      "learning_rate": 5e-06,
+      "loss": 0.1515,
+      "step": 11467
+    },
+    {
+      "epoch": 8.632292058712833,
+      "grad_norm": 0.18099895119667053,
+      "learning_rate": 5e-06,
+      "loss": 0.1078,
+      "step": 11468
+    },
+    {
+      "epoch": 8.633044787354159,
+      "grad_norm": 0.16946259140968323,
+      "learning_rate": 5e-06,
+      "loss": 0.1373,
+      "step": 11469
+    },
+    {
+      "epoch": 8.633797515995484,
+      "grad_norm": 0.18694928288459778,
+      "learning_rate": 5e-06,
+      "loss": 0.2392,
+      "step": 11470
+    },
+    {
+      "epoch": 8.634550244636808,
+      "grad_norm": 0.17207539081573486,
+      "learning_rate": 5e-06,
+      "loss": 0.2817,
+      "step": 11471
+    },
+    {
+      "epoch": 8.635302973278133,
+      "grad_norm": 0.1802719384431839,
+      "learning_rate": 5e-06,
+      "loss": 0.1639,
+      "step": 11472
+    },
+    {
+      "epoch": 8.636055701919458,
+      "grad_norm": 0.18269065022468567,
+      "learning_rate": 5e-06,
+      "loss": 0.1736,
+      "step": 11473
+    },
+    {
+      "epoch": 8.636808430560784,
+      "grad_norm": 0.1662711203098297,
+      "learning_rate": 5e-06,
+      "loss": 0.1091,
+      "step": 11474
+    },
+    {
+      "epoch": 8.637561159202107,
+      "grad_norm": 0.17383907735347748,
+      "learning_rate": 5e-06,
+      "loss": 0.1476,
+      "step": 11475
+    },
+    {
+      "epoch": 8.638313887843433,
+      "grad_norm": 0.1790684461593628,
+      "learning_rate": 5e-06,
+      "loss": 0.1365,
+      "step": 11476
+    },
+    {
+      "epoch": 8.639066616484758,
+      "grad_norm": 0.1701682209968567,
+      "learning_rate": 5e-06,
+      "loss": 0.1426,
+      "step": 11477
+    },
+    {
+      "epoch": 8.639819345126082,
+      "grad_norm": 0.16385656595230103,
+      "learning_rate": 5e-06,
+      "loss": 0.0842,
+      "step": 11478
+    },
+    {
+      "epoch": 8.640572073767407,
+      "grad_norm": 0.17595744132995605,
+      "learning_rate": 5e-06,
+      "loss": 0.1708,
+      "step": 11479
+    },
+    {
+      "epoch": 8.641324802408732,
+      "grad_norm": 0.16111032664775848,
+      "learning_rate": 5e-06,
+      "loss": 0.1729,
+      "step": 11480
+    },
+    {
+      "epoch": 8.642077531050056,
+      "grad_norm": 0.19082608819007874,
+      "learning_rate": 5e-06,
+      "loss": 0.1507,
+      "step": 11481
+    },
+    {
+      "epoch": 8.642830259691381,
+      "grad_norm": 0.1677676886320114,
+      "learning_rate": 5e-06,
+      "loss": 0.1583,
+      "step": 11482
+    },
+    {
+      "epoch": 8.643582988332707,
+      "grad_norm": 0.19175618886947632,
+      "learning_rate": 5e-06,
+      "loss": 0.1867,
+      "step": 11483
+    },
+    {
+      "epoch": 8.64433571697403,
+      "grad_norm": 0.20233233273029327,
+      "learning_rate": 5e-06,
+      "loss": 0.189,
+      "step": 11484
+    },
+    {
+      "epoch": 8.645088445615356,
+      "grad_norm": 0.16466301679611206,
+      "learning_rate": 5e-06,
+      "loss": 0.1256,
+      "step": 11485
+    },
+    {
+      "epoch": 8.645841174256681,
+      "grad_norm": 0.15737171471118927,
+      "learning_rate": 5e-06,
+      "loss": 0.1524,
+      "step": 11486
+    },
+    {
+      "epoch": 8.646593902898005,
+      "grad_norm": 0.1643225997686386,
+      "learning_rate": 5e-06,
+      "loss": 0.1186,
+      "step": 11487
+    },
+    {
+      "epoch": 8.64734663153933,
+      "grad_norm": 0.1897178292274475,
+      "learning_rate": 5e-06,
+      "loss": 0.1806,
+      "step": 11488
+    },
+    {
+      "epoch": 8.648099360180655,
+      "grad_norm": 0.16329315304756165,
+      "learning_rate": 5e-06,
+      "loss": 0.1059,
+      "step": 11489
+    },
+    {
+      "epoch": 8.648852088821979,
+      "grad_norm": 0.15344883501529694,
+      "learning_rate": 5e-06,
+      "loss": 0.1267,
+      "step": 11490
+    },
+    {
+      "epoch": 8.649604817463304,
+      "grad_norm": 0.19413438439369202,
+      "learning_rate": 5e-06,
+      "loss": 0.2258,
+      "step": 11491
+    },
+    {
+      "epoch": 8.65035754610463,
+      "grad_norm": 0.17104081809520721,
+      "learning_rate": 5e-06,
+      "loss": 0.1418,
+      "step": 11492
+    },
+    {
+      "epoch": 8.651110274745953,
+      "grad_norm": 0.1803814023733139,
+      "learning_rate": 5e-06,
+      "loss": 0.1457,
+      "step": 11493
+    },
+    {
+      "epoch": 8.651863003387279,
+      "grad_norm": 0.17292703688144684,
+      "learning_rate": 5e-06,
+      "loss": 0.1165,
+      "step": 11494
+    },
+    {
+      "epoch": 8.652615732028604,
+      "grad_norm": 0.1650870442390442,
+      "learning_rate": 5e-06,
+      "loss": 0.103,
+      "step": 11495
+    },
+    {
+      "epoch": 8.65336846066993,
+      "grad_norm": 0.185072124004364,
+      "learning_rate": 5e-06,
+      "loss": 0.2695,
+      "step": 11496
+    },
+    {
+      "epoch": 8.654121189311253,
+      "grad_norm": 0.20105595886707306,
+      "learning_rate": 5e-06,
+      "loss": 0.1929,
+      "step": 11497
+    },
+    {
+      "epoch": 8.654873917952578,
+      "grad_norm": 0.16316749155521393,
+      "learning_rate": 5e-06,
+      "loss": 0.1579,
+      "step": 11498
+    },
+    {
+      "epoch": 8.655626646593904,
+      "grad_norm": 0.20423679053783417,
+      "learning_rate": 5e-06,
+      "loss": 0.1595,
+      "step": 11499
+    },
+    {
+      "epoch": 8.656379375235227,
+      "grad_norm": 0.16487275063991547,
+      "learning_rate": 5e-06,
+      "loss": 0.1356,
+      "step": 11500
+    },
+    {
+      "epoch": 8.657132103876553,
+      "grad_norm": 0.18261252343654633,
+      "learning_rate": 5e-06,
+      "loss": 0.1031,
+      "step": 11501
+    },
+    {
+      "epoch": 8.657884832517878,
+      "grad_norm": 0.18189609050750732,
+      "learning_rate": 5e-06,
+      "loss": 0.1783,
+      "step": 11502
+    },
+    {
+      "epoch": 8.658637561159201,
+      "grad_norm": 0.17144127190113068,
+      "learning_rate": 5e-06,
+      "loss": 0.1192,
+      "step": 11503
+    },
+    {
+      "epoch": 8.659390289800527,
+      "grad_norm": 0.20256419479846954,
+      "learning_rate": 5e-06,
+      "loss": 0.1651,
+      "step": 11504
+    },
+    {
+      "epoch": 8.660143018441852,
+      "grad_norm": 0.15673960745334625,
+      "learning_rate": 5e-06,
+      "loss": 0.1221,
+      "step": 11505
+    },
+    {
+      "epoch": 8.660895747083176,
+      "grad_norm": 0.17831334471702576,
+      "learning_rate": 5e-06,
+      "loss": 0.1283,
+      "step": 11506
+    },
+    {
+      "epoch": 8.661648475724501,
+      "grad_norm": 0.18922026455402374,
+      "learning_rate": 5e-06,
+      "loss": 0.2126,
+      "step": 11507
+    },
+    {
+      "epoch": 8.662401204365827,
+      "grad_norm": 0.1776096522808075,
+      "learning_rate": 5e-06,
+      "loss": 0.1398,
+      "step": 11508
+    },
+    {
+      "epoch": 8.663153933007152,
+      "grad_norm": 0.15730491280555725,
+      "learning_rate": 5e-06,
+      "loss": 0.172,
+      "step": 11509
+    },
+    {
+      "epoch": 8.663906661648475,
+      "grad_norm": 0.17607882618904114,
+      "learning_rate": 5e-06,
+      "loss": 0.0909,
+      "step": 11510
+    },
+    {
+      "epoch": 8.6646593902898,
+      "grad_norm": 0.16638946533203125,
+      "learning_rate": 5e-06,
+      "loss": 0.2015,
+      "step": 11511
+    },
+    {
+      "epoch": 8.665412118931126,
+      "grad_norm": 0.17191894352436066,
+      "learning_rate": 5e-06,
+      "loss": 0.1768,
+      "step": 11512
+    },
+    {
+      "epoch": 8.66616484757245,
+      "grad_norm": 0.14547140896320343,
+      "learning_rate": 5e-06,
+      "loss": 0.099,
+      "step": 11513
+    },
+    {
+      "epoch": 8.666917576213775,
+      "grad_norm": 0.16987855732440948,
+      "learning_rate": 5e-06,
+      "loss": 0.1077,
+      "step": 11514
+    },
+    {
+      "epoch": 8.6676703048551,
+      "grad_norm": 0.17288650572299957,
+      "learning_rate": 5e-06,
+      "loss": 0.1615,
+      "step": 11515
+    },
+    {
+      "epoch": 8.668423033496424,
+      "grad_norm": 0.20750176906585693,
+      "learning_rate": 5e-06,
+      "loss": 0.155,
+      "step": 11516
+    },
+    {
+      "epoch": 8.66917576213775,
+      "grad_norm": 0.17126862704753876,
+      "learning_rate": 5e-06,
+      "loss": 0.1841,
+      "step": 11517
+    },
+    {
+      "epoch": 8.669928490779075,
+      "grad_norm": 0.1766023486852646,
+      "learning_rate": 5e-06,
+      "loss": 0.1351,
+      "step": 11518
+    },
+    {
+      "epoch": 8.670681219420398,
+      "grad_norm": 0.19740121066570282,
+      "learning_rate": 5e-06,
+      "loss": 0.2025,
+      "step": 11519
+    },
+    {
+      "epoch": 8.671433948061724,
+      "grad_norm": 0.1690657138824463,
+      "learning_rate": 5e-06,
+      "loss": 0.0766,
+      "step": 11520
+    },
+    {
+      "epoch": 8.672186676703049,
+      "grad_norm": 0.1639159619808197,
+      "learning_rate": 5e-06,
+      "loss": 0.0752,
+      "step": 11521
+    },
+    {
+      "epoch": 8.672939405344373,
+      "grad_norm": 0.16176150739192963,
+      "learning_rate": 5e-06,
+      "loss": 0.1284,
+      "step": 11522
+    },
+    {
+      "epoch": 8.673692133985698,
+      "grad_norm": 0.2112165242433548,
+      "learning_rate": 5e-06,
+      "loss": 0.2091,
+      "step": 11523
+    },
+    {
+      "epoch": 8.674444862627023,
+      "grad_norm": 0.172660693526268,
+      "learning_rate": 5e-06,
+      "loss": 0.175,
+      "step": 11524
+    },
+    {
+      "epoch": 8.675197591268347,
+      "grad_norm": 0.17162655293941498,
+      "learning_rate": 5e-06,
+      "loss": 0.0987,
+      "step": 11525
+    },
+    {
+      "epoch": 8.675950319909672,
+      "grad_norm": 0.17457370460033417,
+      "learning_rate": 5e-06,
+      "loss": 0.1546,
+      "step": 11526
+    },
+    {
+      "epoch": 8.676703048550998,
+      "grad_norm": 0.1804584264755249,
+      "learning_rate": 5e-06,
+      "loss": 0.1629,
+      "step": 11527
+    },
+    {
+      "epoch": 8.677455777192321,
+      "grad_norm": 0.16726204752922058,
+      "learning_rate": 5e-06,
+      "loss": 0.1221,
+      "step": 11528
+    },
+    {
+      "epoch": 8.678208505833647,
+      "grad_norm": 0.16219601035118103,
+      "learning_rate": 5e-06,
+      "loss": 0.0895,
+      "step": 11529
+    },
+    {
+      "epoch": 8.678961234474972,
+      "grad_norm": 0.19064722955226898,
+      "learning_rate": 5e-06,
+      "loss": 0.1169,
+      "step": 11530
+    },
+    {
+      "epoch": 8.679713963116297,
+      "grad_norm": 0.17507235705852509,
+      "learning_rate": 5e-06,
+      "loss": 0.1944,
+      "step": 11531
+    },
+    {
+      "epoch": 8.680466691757621,
+      "grad_norm": 0.1909978836774826,
+      "learning_rate": 5e-06,
+      "loss": 0.1961,
+      "step": 11532
+    },
+    {
+      "epoch": 8.681219420398946,
+      "grad_norm": 0.16220979392528534,
+      "learning_rate": 5e-06,
+      "loss": 0.154,
+      "step": 11533
+    },
+    {
+      "epoch": 8.681972149040272,
+      "grad_norm": 0.1503327190876007,
+      "learning_rate": 5e-06,
+      "loss": 0.1257,
+      "step": 11534
+    },
+    {
+      "epoch": 8.682724877681595,
+      "grad_norm": 0.1951753944158554,
+      "learning_rate": 5e-06,
+      "loss": 0.1726,
+      "step": 11535
+    },
+    {
+      "epoch": 8.68347760632292,
+      "grad_norm": 0.1546325832605362,
+      "learning_rate": 5e-06,
+      "loss": 0.1286,
+      "step": 11536
+    },
+    {
+      "epoch": 8.684230334964246,
+      "grad_norm": 0.14618071913719177,
+      "learning_rate": 5e-06,
+      "loss": 0.0775,
+      "step": 11537
+    },
+    {
+      "epoch": 8.68498306360557,
+      "grad_norm": 0.17948102951049805,
+      "learning_rate": 5e-06,
+      "loss": 0.1522,
+      "step": 11538
+    },
+    {
+      "epoch": 8.685735792246895,
+      "grad_norm": 0.15199892222881317,
+      "learning_rate": 5e-06,
+      "loss": 0.1119,
+      "step": 11539
+    },
+    {
+      "epoch": 8.68648852088822,
+      "grad_norm": 0.17153166234493256,
+      "learning_rate": 5e-06,
+      "loss": 0.1144,
+      "step": 11540
+    },
+    {
+      "epoch": 8.687241249529544,
+      "grad_norm": 0.1735125631093979,
+      "learning_rate": 5e-06,
+      "loss": 0.1998,
+      "step": 11541
+    },
+    {
+      "epoch": 8.68799397817087,
+      "grad_norm": 0.1543084979057312,
+      "learning_rate": 5e-06,
+      "loss": 0.0831,
+      "step": 11542
+    },
+    {
+      "epoch": 8.688746706812195,
+      "grad_norm": 0.17269720137119293,
+      "learning_rate": 5e-06,
+      "loss": 0.1226,
+      "step": 11543
+    },
+    {
+      "epoch": 8.689499435453518,
+      "grad_norm": 0.21620385348796844,
+      "learning_rate": 5e-06,
+      "loss": 0.2789,
+      "step": 11544
+    },
+    {
+      "epoch": 8.690252164094844,
+      "grad_norm": 0.17970852553844452,
+      "learning_rate": 5e-06,
+      "loss": 0.153,
+      "step": 11545
+    },
+    {
+      "epoch": 8.691004892736169,
+      "grad_norm": 0.15820300579071045,
+      "learning_rate": 5e-06,
+      "loss": 0.1649,
+      "step": 11546
+    },
+    {
+      "epoch": 8.691757621377494,
+      "grad_norm": 0.19228938221931458,
+      "learning_rate": 5e-06,
+      "loss": 0.1741,
+      "step": 11547
+    },
+    {
+      "epoch": 8.692510350018818,
+      "grad_norm": 0.1688135266304016,
+      "learning_rate": 5e-06,
+      "loss": 0.1364,
+      "step": 11548
+    },
+    {
+      "epoch": 8.693263078660143,
+      "grad_norm": 0.1456788182258606,
+      "learning_rate": 5e-06,
+      "loss": 0.0789,
+      "step": 11549
+    },
+    {
+      "epoch": 8.694015807301469,
+      "grad_norm": 0.18288423120975494,
+      "learning_rate": 5e-06,
+      "loss": 0.1566,
+      "step": 11550
+    },
+    {
+      "epoch": 8.694768535942792,
+      "grad_norm": 0.16576485335826874,
+      "learning_rate": 5e-06,
+      "loss": 0.1184,
+      "step": 11551
+    },
+    {
+      "epoch": 8.695521264584118,
+      "grad_norm": 0.1842995136976242,
+      "learning_rate": 5e-06,
+      "loss": 0.1371,
+      "step": 11552
+    },
+    {
+      "epoch": 8.696273993225443,
+      "grad_norm": 0.16772200167179108,
+      "learning_rate": 5e-06,
+      "loss": 0.0891,
+      "step": 11553
+    },
+    {
+      "epoch": 8.697026721866767,
+      "grad_norm": 0.16334524750709534,
+      "learning_rate": 5e-06,
+      "loss": 0.0799,
+      "step": 11554
+    },
+    {
+      "epoch": 8.697779450508092,
+      "grad_norm": 0.18477056920528412,
+      "learning_rate": 5e-06,
+      "loss": 0.1571,
+      "step": 11555
+    },
+    {
+      "epoch": 8.698532179149417,
+      "grad_norm": 0.15539947152137756,
+      "learning_rate": 5e-06,
+      "loss": 0.1138,
+      "step": 11556
+    },
+    {
+      "epoch": 8.69928490779074,
+      "grad_norm": 0.15815219283103943,
+      "learning_rate": 5e-06,
+      "loss": 0.1149,
+      "step": 11557
+    },
+    {
+      "epoch": 8.700037636432066,
+      "grad_norm": 0.15403205156326294,
+      "learning_rate": 5e-06,
+      "loss": 0.1169,
+      "step": 11558
+    },
+    {
+      "epoch": 8.700790365073392,
+      "grad_norm": 0.18574979901313782,
+      "learning_rate": 5e-06,
+      "loss": 0.1743,
+      "step": 11559
+    },
+    {
+      "epoch": 8.701543093714715,
+      "grad_norm": 0.12924526631832123,
+      "learning_rate": 5e-06,
+      "loss": 0.042,
+      "step": 11560
+    },
+    {
+      "epoch": 8.70229582235604,
+      "grad_norm": 0.17704974114894867,
+      "learning_rate": 5e-06,
+      "loss": 0.1586,
+      "step": 11561
+    },
+    {
+      "epoch": 8.703048550997366,
+      "grad_norm": 0.18806658685207367,
+      "learning_rate": 5e-06,
+      "loss": 0.1335,
+      "step": 11562
+    },
+    {
+      "epoch": 8.70380127963869,
+      "grad_norm": 0.197265163064003,
+      "learning_rate": 5e-06,
+      "loss": 0.1338,
+      "step": 11563
+    },
+    {
+      "epoch": 8.704554008280015,
+      "grad_norm": 0.1859433650970459,
+      "learning_rate": 5e-06,
+      "loss": 0.1289,
+      "step": 11564
+    },
+    {
+      "epoch": 8.70530673692134,
+      "grad_norm": 0.18596608936786652,
+      "learning_rate": 5e-06,
+      "loss": 0.1403,
+      "step": 11565
+    },
+    {
+      "epoch": 8.706059465562664,
+      "grad_norm": 0.17328904569149017,
+      "learning_rate": 5e-06,
+      "loss": 0.1461,
+      "step": 11566
+    },
+    {
+      "epoch": 8.706812194203989,
+      "grad_norm": 0.1596156507730484,
+      "learning_rate": 5e-06,
+      "loss": 0.1097,
+      "step": 11567
+    },
+    {
+      "epoch": 8.707564922845314,
+      "grad_norm": 0.1394525021314621,
+      "learning_rate": 5e-06,
+      "loss": 0.0791,
+      "step": 11568
+    },
+    {
+      "epoch": 8.70831765148664,
+      "grad_norm": 0.15767619013786316,
+      "learning_rate": 5e-06,
+      "loss": 0.0694,
+      "step": 11569
+    },
+    {
+      "epoch": 8.709070380127963,
+      "grad_norm": 0.19156001508235931,
+      "learning_rate": 5e-06,
+      "loss": 0.2108,
+      "step": 11570
+    },
+    {
+      "epoch": 8.709823108769289,
+      "grad_norm": 0.17392012476921082,
+      "learning_rate": 5e-06,
+      "loss": 0.1376,
+      "step": 11571
+    },
+    {
+      "epoch": 8.710575837410614,
+      "grad_norm": 0.17081691324710846,
+      "learning_rate": 5e-06,
+      "loss": 0.1879,
+      "step": 11572
+    },
+    {
+      "epoch": 8.711328566051938,
+      "grad_norm": 0.17662520706653595,
+      "learning_rate": 5e-06,
+      "loss": 0.1425,
+      "step": 11573
+    },
+    {
+      "epoch": 8.712081294693263,
+      "grad_norm": 0.18858225643634796,
+      "learning_rate": 5e-06,
+      "loss": 0.1773,
+      "step": 11574
+    },
+    {
+      "epoch": 8.712834023334588,
+      "grad_norm": 0.1818973273038864,
+      "learning_rate": 5e-06,
+      "loss": 0.1638,
+      "step": 11575
+    },
+    {
+      "epoch": 8.713586751975912,
+      "grad_norm": 0.17694467306137085,
+      "learning_rate": 5e-06,
+      "loss": 0.1724,
+      "step": 11576
+    },
+    {
+      "epoch": 8.714339480617237,
+      "grad_norm": 0.189173623919487,
+      "learning_rate": 5e-06,
+      "loss": 0.1789,
+      "step": 11577
+    },
+    {
+      "epoch": 8.715092209258563,
+      "grad_norm": 0.15780600905418396,
+      "learning_rate": 5e-06,
+      "loss": 0.1562,
+      "step": 11578
+    },
+    {
+      "epoch": 8.715844937899886,
+      "grad_norm": 0.14270776510238647,
+      "learning_rate": 5e-06,
+      "loss": 0.0566,
+      "step": 11579
+    },
+    {
+      "epoch": 8.716597666541212,
+      "grad_norm": 0.18730922043323517,
+      "learning_rate": 5e-06,
+      "loss": 0.1434,
+      "step": 11580
+    },
+    {
+      "epoch": 8.717350395182537,
+      "grad_norm": 0.1655704379081726,
+      "learning_rate": 5e-06,
+      "loss": 0.1415,
+      "step": 11581
+    },
+    {
+      "epoch": 8.718103123823862,
+      "grad_norm": 0.1710081845521927,
+      "learning_rate": 5e-06,
+      "loss": 0.1824,
+      "step": 11582
+    },
+    {
+      "epoch": 8.718855852465186,
+      "grad_norm": 0.17957602441310883,
+      "learning_rate": 5e-06,
+      "loss": 0.2056,
+      "step": 11583
+    },
+    {
+      "epoch": 8.719608581106511,
+      "grad_norm": 0.17531469464302063,
+      "learning_rate": 5e-06,
+      "loss": 0.137,
+      "step": 11584
+    },
+    {
+      "epoch": 8.720361309747837,
+      "grad_norm": 0.18287605047225952,
+      "learning_rate": 5e-06,
+      "loss": 0.1473,
+      "step": 11585
+    },
+    {
+      "epoch": 8.72111403838916,
+      "grad_norm": 0.16177624464035034,
+      "learning_rate": 5e-06,
+      "loss": 0.1104,
+      "step": 11586
+    },
+    {
+      "epoch": 8.721866767030486,
+      "grad_norm": 0.1985306590795517,
+      "learning_rate": 5e-06,
+      "loss": 0.2133,
+      "step": 11587
+    },
+    {
+      "epoch": 8.722619495671811,
+      "grad_norm": 0.18272294104099274,
+      "learning_rate": 5e-06,
+      "loss": 0.1049,
+      "step": 11588
+    },
+    {
+      "epoch": 8.723372224313135,
+      "grad_norm": 0.15185199677944183,
+      "learning_rate": 5e-06,
+      "loss": 0.0894,
+      "step": 11589
+    },
+    {
+      "epoch": 8.72412495295446,
+      "grad_norm": 0.17312023043632507,
+      "learning_rate": 5e-06,
+      "loss": 0.1829,
+      "step": 11590
+    },
+    {
+      "epoch": 8.724877681595785,
+      "grad_norm": 0.2082335501909256,
+      "learning_rate": 5e-06,
+      "loss": 0.166,
+      "step": 11591
+    },
+    {
+      "epoch": 8.725630410237109,
+      "grad_norm": 0.1896756887435913,
+      "learning_rate": 5e-06,
+      "loss": 0.0933,
+      "step": 11592
+    },
+    {
+      "epoch": 8.726383138878434,
+      "grad_norm": 0.17242474853992462,
+      "learning_rate": 5e-06,
+      "loss": 0.1157,
+      "step": 11593
+    },
+    {
+      "epoch": 8.72713586751976,
+      "grad_norm": 0.18695366382598877,
+      "learning_rate": 5e-06,
+      "loss": 0.1161,
+      "step": 11594
+    },
+    {
+      "epoch": 8.727888596161083,
+      "grad_norm": 0.18048220872879028,
+      "learning_rate": 5e-06,
+      "loss": 0.18,
+      "step": 11595
+    },
+    {
+      "epoch": 8.728641324802409,
+      "grad_norm": 0.20634709298610687,
+      "learning_rate": 5e-06,
+      "loss": 0.1558,
+      "step": 11596
+    },
+    {
+      "epoch": 8.729394053443734,
+      "grad_norm": 0.19754402339458466,
+      "learning_rate": 5e-06,
+      "loss": 0.1442,
+      "step": 11597
+    },
+    {
+      "epoch": 8.730146782085058,
+      "grad_norm": 0.1496960073709488,
+      "learning_rate": 5e-06,
+      "loss": 0.0825,
+      "step": 11598
+    },
+    {
+      "epoch": 8.730899510726383,
+      "grad_norm": 0.18135398626327515,
+      "learning_rate": 5e-06,
+      "loss": 0.087,
+      "step": 11599
+    },
+    {
+      "epoch": 8.731652239367708,
+      "grad_norm": 0.1653597354888916,
+      "learning_rate": 5e-06,
+      "loss": 0.1731,
+      "step": 11600
+    },
+    {
+      "epoch": 8.731652239367708,
+      "eval_loss": 0.19419583678245544,
+      "eval_runtime": 454.3388,
+      "eval_samples_per_second": 21.189,
+      "eval_steps_per_second": 0.663,
+      "step": 11600
+    },
+    {
+      "epoch": 8.732404968009032,
+      "grad_norm": 0.15393000841140747,
+      "learning_rate": 5e-06,
+      "loss": 0.1613,
+      "step": 11601
+    },
+    {
+      "epoch": 8.733157696650357,
+      "grad_norm": 0.16956664621829987,
+      "learning_rate": 5e-06,
+      "loss": 0.1721,
+      "step": 11602
+    },
+    {
+      "epoch": 8.733910425291683,
+      "grad_norm": 0.17878805100917816,
+      "learning_rate": 5e-06,
+      "loss": 0.1189,
+      "step": 11603
+    },
+    {
+      "epoch": 8.734663153933008,
+      "grad_norm": 0.1666754186153412,
+      "learning_rate": 5e-06,
+      "loss": 0.1143,
+      "step": 11604
+    },
+    {
+      "epoch": 8.735415882574332,
+      "grad_norm": 0.18971672654151917,
+      "learning_rate": 5e-06,
+      "loss": 0.2095,
+      "step": 11605
+    },
+    {
+      "epoch": 8.736168611215657,
+      "grad_norm": 0.16517223417758942,
+      "learning_rate": 5e-06,
+      "loss": 0.1709,
+      "step": 11606
+    },
+    {
+      "epoch": 8.736921339856982,
+      "grad_norm": 0.15930336713790894,
+      "learning_rate": 5e-06,
+      "loss": 0.1935,
+      "step": 11607
+    },
+    {
+      "epoch": 8.737674068498306,
+      "grad_norm": 0.15533670783042908,
+      "learning_rate": 5e-06,
+      "loss": 0.1059,
+      "step": 11608
+    },
+    {
+      "epoch": 8.738426797139631,
+      "grad_norm": 0.1885155290365219,
+      "learning_rate": 5e-06,
+      "loss": 0.1846,
+      "step": 11609
+    },
+    {
+      "epoch": 8.739179525780957,
+      "grad_norm": 0.15098822116851807,
+      "learning_rate": 5e-06,
+      "loss": 0.2014,
+      "step": 11610
+    },
+    {
+      "epoch": 8.73993225442228,
+      "grad_norm": 0.16912713646888733,
+      "learning_rate": 5e-06,
+      "loss": 0.2745,
+      "step": 11611
+    },
+    {
+      "epoch": 8.740684983063606,
+      "grad_norm": 0.17330411076545715,
+      "learning_rate": 5e-06,
+      "loss": 0.0932,
+      "step": 11612
+    },
+    {
+      "epoch": 8.741437711704931,
+      "grad_norm": 0.15642060339450836,
+      "learning_rate": 5e-06,
+      "loss": 0.2122,
+      "step": 11613
+    },
+    {
+      "epoch": 8.742190440346254,
+      "grad_norm": 0.18003900349140167,
+      "learning_rate": 5e-06,
+      "loss": 0.166,
+      "step": 11614
+    },
+    {
+      "epoch": 8.74294316898758,
+      "grad_norm": 0.1675892174243927,
+      "learning_rate": 5e-06,
+      "loss": 0.1624,
+      "step": 11615
+    },
+    {
+      "epoch": 8.743695897628905,
+      "grad_norm": 0.2081669569015503,
+      "learning_rate": 5e-06,
+      "loss": 0.1694,
+      "step": 11616
+    },
+    {
+      "epoch": 8.744448626270229,
+      "grad_norm": 0.15357185900211334,
+      "learning_rate": 5e-06,
+      "loss": 0.0726,
+      "step": 11617
+    },
+    {
+      "epoch": 8.745201354911554,
+      "grad_norm": 0.18141578137874603,
+      "learning_rate": 5e-06,
+      "loss": 0.1604,
+      "step": 11618
+    },
+    {
+      "epoch": 8.74595408355288,
+      "grad_norm": 0.19866059720516205,
+      "learning_rate": 5e-06,
+      "loss": 0.1429,
+      "step": 11619
+    },
+    {
+      "epoch": 8.746706812194205,
+      "grad_norm": 0.17338448762893677,
+      "learning_rate": 5e-06,
+      "loss": 0.1251,
+      "step": 11620
+    },
+    {
+      "epoch": 8.747459540835528,
+      "grad_norm": 0.15079787373542786,
+      "learning_rate": 5e-06,
+      "loss": 0.1774,
+      "step": 11621
+    },
+    {
+      "epoch": 8.748212269476854,
+      "grad_norm": 0.18239563703536987,
+      "learning_rate": 5e-06,
+      "loss": 0.1096,
+      "step": 11622
+    },
+    {
+      "epoch": 8.74896499811818,
+      "grad_norm": 0.18291619420051575,
+      "learning_rate": 5e-06,
+      "loss": 0.1806,
+      "step": 11623
+    },
+    {
+      "epoch": 8.749717726759503,
+      "grad_norm": 0.17350444197654724,
+      "learning_rate": 5e-06,
+      "loss": 0.141,
+      "step": 11624
+    },
+    {
+      "epoch": 8.750470455400828,
+      "grad_norm": 0.1798190176486969,
+      "learning_rate": 5e-06,
+      "loss": 0.1946,
+      "step": 11625
+    },
+    {
+      "epoch": 8.751223184042153,
+      "grad_norm": 0.17472705245018005,
+      "learning_rate": 5e-06,
+      "loss": 0.1256,
+      "step": 11626
+    },
+    {
+      "epoch": 8.751975912683477,
+      "grad_norm": 0.1700902134180069,
+      "learning_rate": 5e-06,
+      "loss": 0.0948,
+      "step": 11627
+    },
+    {
+      "epoch": 8.752728641324802,
+      "grad_norm": 0.19266298413276672,
+      "learning_rate": 5e-06,
+      "loss": 0.1155,
+      "step": 11628
+    },
+    {
+      "epoch": 8.753481369966128,
+      "grad_norm": 0.18155914545059204,
+      "learning_rate": 5e-06,
+      "loss": 0.1468,
+      "step": 11629
+    },
+    {
+      "epoch": 8.754234098607451,
+      "grad_norm": 0.1854536086320877,
+      "learning_rate": 5e-06,
+      "loss": 0.1127,
+      "step": 11630
+    },
+    {
+      "epoch": 8.754986827248777,
+      "grad_norm": 0.19632095098495483,
+      "learning_rate": 5e-06,
+      "loss": 0.183,
+      "step": 11631
+    },
+    {
+      "epoch": 8.755739555890102,
+      "grad_norm": 0.17589707672595978,
+      "learning_rate": 5e-06,
+      "loss": 0.1672,
+      "step": 11632
+    },
+    {
+      "epoch": 8.756492284531426,
+      "grad_norm": 0.18389002978801727,
+      "learning_rate": 5e-06,
+      "loss": 0.1107,
+      "step": 11633
+    },
+    {
+      "epoch": 8.757245013172751,
+      "grad_norm": 0.16406840085983276,
+      "learning_rate": 5e-06,
+      "loss": 0.1426,
+      "step": 11634
+    },
+    {
+      "epoch": 8.757997741814076,
+      "grad_norm": 0.20047275722026825,
+      "learning_rate": 5e-06,
+      "loss": 0.1919,
+      "step": 11635
+    },
+    {
+      "epoch": 8.7587504704554,
+      "grad_norm": 0.15766297280788422,
+      "learning_rate": 5e-06,
+      "loss": 0.1582,
+      "step": 11636
+    },
+    {
+      "epoch": 8.759503199096725,
+      "grad_norm": 0.17845821380615234,
+      "learning_rate": 5e-06,
+      "loss": 0.2207,
+      "step": 11637
+    },
+    {
+      "epoch": 8.76025592773805,
+      "grad_norm": 0.17664466798305511,
+      "learning_rate": 5e-06,
+      "loss": 0.1595,
+      "step": 11638
+    },
+    {
+      "epoch": 8.761008656379374,
+      "grad_norm": 0.1867150366306305,
+      "learning_rate": 5e-06,
+      "loss": 0.139,
+      "step": 11639
+    },
+    {
+      "epoch": 8.7617613850207,
+      "grad_norm": 0.18164502084255219,
+      "learning_rate": 5e-06,
+      "loss": 0.1459,
+      "step": 11640
+    },
+    {
+      "epoch": 8.762514113662025,
+      "grad_norm": 0.1745232194662094,
+      "learning_rate": 5e-06,
+      "loss": 0.0952,
+      "step": 11641
+    },
+    {
+      "epoch": 8.76326684230335,
+      "grad_norm": 0.17645031213760376,
+      "learning_rate": 5e-06,
+      "loss": 0.1854,
+      "step": 11642
+    },
+    {
+      "epoch": 8.764019570944674,
+      "grad_norm": 0.17036162316799164,
+      "learning_rate": 5e-06,
+      "loss": 0.155,
+      "step": 11643
+    },
+    {
+      "epoch": 8.764772299586,
+      "grad_norm": 0.17214803397655487,
+      "learning_rate": 5e-06,
+      "loss": 0.1523,
+      "step": 11644
+    },
+    {
+      "epoch": 8.765525028227325,
+      "grad_norm": 0.17745302617549896,
+      "learning_rate": 5e-06,
+      "loss": 0.1367,
+      "step": 11645
+    },
+    {
+      "epoch": 8.766277756868648,
+      "grad_norm": 0.1900198608636856,
+      "learning_rate": 5e-06,
+      "loss": 0.2242,
+      "step": 11646
+    },
+    {
+      "epoch": 8.767030485509974,
+      "grad_norm": 0.1570509970188141,
+      "learning_rate": 5e-06,
+      "loss": 0.1497,
+      "step": 11647
+    },
+    {
+      "epoch": 8.767783214151299,
+      "grad_norm": 0.16097445785999298,
+      "learning_rate": 5e-06,
+      "loss": 0.1041,
+      "step": 11648
+    },
+    {
+      "epoch": 8.768535942792623,
+      "grad_norm": 0.16799132525920868,
+      "learning_rate": 5e-06,
+      "loss": 0.2487,
+      "step": 11649
+    },
+    {
+      "epoch": 8.769288671433948,
+      "grad_norm": 0.16023597121238708,
+      "learning_rate": 5e-06,
+      "loss": 0.1077,
+      "step": 11650
+    },
+    {
+      "epoch": 8.770041400075273,
+      "grad_norm": 0.18137367069721222,
+      "learning_rate": 5e-06,
+      "loss": 0.1948,
+      "step": 11651
+    },
+    {
+      "epoch": 8.770794128716597,
+      "grad_norm": 0.17695070803165436,
+      "learning_rate": 5e-06,
+      "loss": 0.1473,
+      "step": 11652
+    },
+    {
+      "epoch": 8.771546857357922,
+      "grad_norm": 0.18603317439556122,
+      "learning_rate": 5e-06,
+      "loss": 0.1372,
+      "step": 11653
+    },
+    {
+      "epoch": 8.772299585999248,
+      "grad_norm": 0.20703646540641785,
+      "learning_rate": 5e-06,
+      "loss": 0.1973,
+      "step": 11654
+    },
+    {
+      "epoch": 8.773052314640573,
+      "grad_norm": 0.16157718002796173,
+      "learning_rate": 5e-06,
+      "loss": 0.1569,
+      "step": 11655
+    },
+    {
+      "epoch": 8.773805043281897,
+      "grad_norm": 0.18260557949543,
+      "learning_rate": 5e-06,
+      "loss": 0.1196,
+      "step": 11656
+    },
+    {
+      "epoch": 8.774557771923222,
+      "grad_norm": 0.16140805184841156,
+      "learning_rate": 5e-06,
+      "loss": 0.173,
+      "step": 11657
+    },
+    {
+      "epoch": 8.775310500564547,
+      "grad_norm": 0.1879730522632599,
+      "learning_rate": 5e-06,
+      "loss": 0.1414,
+      "step": 11658
+    },
+    {
+      "epoch": 8.776063229205871,
+      "grad_norm": 0.17017552256584167,
+      "learning_rate": 5e-06,
+      "loss": 0.1383,
+      "step": 11659
+    },
+    {
+      "epoch": 8.776815957847196,
+      "grad_norm": 0.1841881424188614,
+      "learning_rate": 5e-06,
+      "loss": 0.1806,
+      "step": 11660
+    },
+    {
+      "epoch": 8.777568686488522,
+      "grad_norm": 0.16339220106601715,
+      "learning_rate": 5e-06,
+      "loss": 0.1818,
+      "step": 11661
+    },
+    {
+      "epoch": 8.778321415129845,
+      "grad_norm": 0.16489194333553314,
+      "learning_rate": 5e-06,
+      "loss": 0.1112,
+      "step": 11662
+    },
+    {
+      "epoch": 8.77907414377117,
+      "grad_norm": 0.16481399536132812,
+      "learning_rate": 5e-06,
+      "loss": 0.1502,
+      "step": 11663
+    },
+    {
+      "epoch": 8.779826872412496,
+      "grad_norm": 0.22317542135715485,
+      "learning_rate": 5e-06,
+      "loss": 0.1518,
+      "step": 11664
+    },
+    {
+      "epoch": 8.78057960105382,
+      "grad_norm": 0.16699910163879395,
+      "learning_rate": 5e-06,
+      "loss": 0.0942,
+      "step": 11665
+    },
+    {
+      "epoch": 8.781332329695145,
+      "grad_norm": 0.17036215960979462,
+      "learning_rate": 5e-06,
+      "loss": 0.1814,
+      "step": 11666
+    },
+    {
+      "epoch": 8.78208505833647,
+      "grad_norm": 0.17571017146110535,
+      "learning_rate": 5e-06,
+      "loss": 0.0908,
+      "step": 11667
+    },
+    {
+      "epoch": 8.782837786977794,
+      "grad_norm": 0.2048608660697937,
+      "learning_rate": 5e-06,
+      "loss": 0.1839,
+      "step": 11668
+    },
+    {
+      "epoch": 8.78359051561912,
+      "grad_norm": 0.151102215051651,
+      "learning_rate": 5e-06,
+      "loss": 0.0993,
+      "step": 11669
+    },
+    {
+      "epoch": 8.784343244260445,
+      "grad_norm": 0.17819207906723022,
+      "learning_rate": 5e-06,
+      "loss": 0.1349,
+      "step": 11670
+    },
+    {
+      "epoch": 8.785095972901768,
+      "grad_norm": 0.21112219989299774,
+      "learning_rate": 5e-06,
+      "loss": 0.2039,
+      "step": 11671
+    },
+    {
+      "epoch": 8.785848701543093,
+      "grad_norm": 0.19261351227760315,
+      "learning_rate": 5e-06,
+      "loss": 0.1638,
+      "step": 11672
+    },
+    {
+      "epoch": 8.786601430184419,
+      "grad_norm": 0.18100933730602264,
+      "learning_rate": 5e-06,
+      "loss": 0.1186,
+      "step": 11673
+    },
+    {
+      "epoch": 8.787354158825742,
+      "grad_norm": 0.1921839565038681,
+      "learning_rate": 5e-06,
+      "loss": 0.1248,
+      "step": 11674
+    },
+    {
+      "epoch": 8.788106887467068,
+      "grad_norm": 0.1868467926979065,
+      "learning_rate": 5e-06,
+      "loss": 0.14,
+      "step": 11675
+    },
+    {
+      "epoch": 8.788859616108393,
+      "grad_norm": 0.18618936836719513,
+      "learning_rate": 5e-06,
+      "loss": 0.1727,
+      "step": 11676
+    },
+    {
+      "epoch": 8.789612344749719,
+      "grad_norm": 0.1894720047712326,
+      "learning_rate": 5e-06,
+      "loss": 0.2373,
+      "step": 11677
+    },
+    {
+      "epoch": 8.790365073391042,
+      "grad_norm": 0.16531729698181152,
+      "learning_rate": 5e-06,
+      "loss": 0.1373,
+      "step": 11678
+    },
+    {
+      "epoch": 8.791117802032367,
+      "grad_norm": 0.17315079271793365,
+      "learning_rate": 5e-06,
+      "loss": 0.1708,
+      "step": 11679
+    },
+    {
+      "epoch": 8.791870530673693,
+      "grad_norm": 0.16998104751110077,
+      "learning_rate": 5e-06,
+      "loss": 0.1333,
+      "step": 11680
+    },
+    {
+      "epoch": 8.792623259315016,
+      "grad_norm": 0.18419742584228516,
+      "learning_rate": 5e-06,
+      "loss": 0.1486,
+      "step": 11681
+    },
+    {
+      "epoch": 8.793375987956342,
+      "grad_norm": 0.17905883491039276,
+      "learning_rate": 5e-06,
+      "loss": 0.1577,
+      "step": 11682
+    },
+    {
+      "epoch": 8.794128716597667,
+      "grad_norm": 0.15159547328948975,
+      "learning_rate": 5e-06,
+      "loss": 0.1408,
+      "step": 11683
+    },
+    {
+      "epoch": 8.79488144523899,
+      "grad_norm": 0.19283230602741241,
+      "learning_rate": 5e-06,
+      "loss": 0.1846,
+      "step": 11684
+    },
+    {
+      "epoch": 8.795634173880316,
+      "grad_norm": 0.17410078644752502,
+      "learning_rate": 5e-06,
+      "loss": 0.1533,
+      "step": 11685
+    },
+    {
+      "epoch": 8.796386902521641,
+      "grad_norm": 0.1543518304824829,
+      "learning_rate": 5e-06,
+      "loss": 0.1356,
+      "step": 11686
+    },
+    {
+      "epoch": 8.797139631162965,
+      "grad_norm": 0.16924652457237244,
+      "learning_rate": 5e-06,
+      "loss": 0.1394,
+      "step": 11687
+    },
+    {
+      "epoch": 8.79789235980429,
+      "grad_norm": 0.17706367373466492,
+      "learning_rate": 5e-06,
+      "loss": 0.1753,
+      "step": 11688
+    },
+    {
+      "epoch": 8.798645088445616,
+      "grad_norm": 0.16994525492191315,
+      "learning_rate": 5e-06,
+      "loss": 0.1152,
+      "step": 11689
+    },
+    {
+      "epoch": 8.79939781708694,
+      "grad_norm": 0.17897270619869232,
+      "learning_rate": 5e-06,
+      "loss": 0.1461,
+      "step": 11690
+    },
+    {
+      "epoch": 8.800150545728265,
+      "grad_norm": 0.17803911864757538,
+      "learning_rate": 5e-06,
+      "loss": 0.1712,
+      "step": 11691
+    },
+    {
+      "epoch": 8.80090327436959,
+      "grad_norm": 0.21522057056427002,
+      "learning_rate": 5e-06,
+      "loss": 0.1914,
+      "step": 11692
+    },
+    {
+      "epoch": 8.801656003010915,
+      "grad_norm": 0.19013050198554993,
+      "learning_rate": 5e-06,
+      "loss": 0.1564,
+      "step": 11693
+    },
+    {
+      "epoch": 8.802408731652239,
+      "grad_norm": 0.1987210065126419,
+      "learning_rate": 5e-06,
+      "loss": 0.1724,
+      "step": 11694
+    },
+    {
+      "epoch": 8.803161460293564,
+      "grad_norm": 0.18683166801929474,
+      "learning_rate": 5e-06,
+      "loss": 0.1486,
+      "step": 11695
+    },
+    {
+      "epoch": 8.80391418893489,
+      "grad_norm": 0.19038240611553192,
+      "learning_rate": 5e-06,
+      "loss": 0.1845,
+      "step": 11696
+    },
+    {
+      "epoch": 8.804666917576213,
+      "grad_norm": 0.1690780073404312,
+      "learning_rate": 5e-06,
+      "loss": 0.1459,
+      "step": 11697
+    },
+    {
+      "epoch": 8.805419646217539,
+      "grad_norm": 0.2137841135263443,
+      "learning_rate": 5e-06,
+      "loss": 0.1835,
+      "step": 11698
+    },
+    {
+      "epoch": 8.806172374858864,
+      "grad_norm": 0.15853315591812134,
+      "learning_rate": 5e-06,
+      "loss": 0.1438,
+      "step": 11699
+    },
+    {
+      "epoch": 8.806925103500188,
+      "grad_norm": 0.17323370277881622,
+      "learning_rate": 5e-06,
+      "loss": 0.1027,
+      "step": 11700
+    },
+    {
+      "epoch": 8.807677832141513,
+      "grad_norm": 0.1900046318769455,
+      "learning_rate": 5e-06,
+      "loss": 0.168,
+      "step": 11701
+    },
+    {
+      "epoch": 8.808430560782838,
+      "grad_norm": 0.16749128699302673,
+      "learning_rate": 5e-06,
+      "loss": 0.2157,
+      "step": 11702
+    },
+    {
+      "epoch": 8.809183289424162,
+      "grad_norm": 0.17805172502994537,
+      "learning_rate": 5e-06,
+      "loss": 0.2243,
+      "step": 11703
+    },
+    {
+      "epoch": 8.809936018065487,
+      "grad_norm": 0.19864016771316528,
+      "learning_rate": 5e-06,
+      "loss": 0.1708,
+      "step": 11704
+    },
+    {
+      "epoch": 8.810688746706813,
+      "grad_norm": 0.1526234894990921,
+      "learning_rate": 5e-06,
+      "loss": 0.2186,
+      "step": 11705
+    },
+    {
+      "epoch": 8.811441475348136,
+      "grad_norm": 0.17951662838459015,
+      "learning_rate": 5e-06,
+      "loss": 0.1592,
+      "step": 11706
+    },
+    {
+      "epoch": 8.812194203989462,
+      "grad_norm": 0.16301481425762177,
+      "learning_rate": 5e-06,
+      "loss": 0.1886,
+      "step": 11707
+    },
+    {
+      "epoch": 8.812946932630787,
+      "grad_norm": 0.18087202310562134,
+      "learning_rate": 5e-06,
+      "loss": 0.1616,
+      "step": 11708
+    },
+    {
+      "epoch": 8.81369966127211,
+      "grad_norm": 0.1738184541463852,
+      "learning_rate": 5e-06,
+      "loss": 0.1263,
+      "step": 11709
+    },
+    {
+      "epoch": 8.814452389913436,
+      "grad_norm": 0.18246763944625854,
+      "learning_rate": 5e-06,
+      "loss": 0.1484,
+      "step": 11710
+    },
+    {
+      "epoch": 8.815205118554761,
+      "grad_norm": 0.20570321381092072,
+      "learning_rate": 5e-06,
+      "loss": 0.1364,
+      "step": 11711
+    },
+    {
+      "epoch": 8.815957847196085,
+      "grad_norm": 0.16805686056613922,
+      "learning_rate": 5e-06,
+      "loss": 0.1612,
+      "step": 11712
+    },
+    {
+      "epoch": 8.81671057583741,
+      "grad_norm": 0.1509520709514618,
+      "learning_rate": 5e-06,
+      "loss": 0.1314,
+      "step": 11713
+    },
+    {
+      "epoch": 8.817463304478736,
+      "grad_norm": 0.18332022428512573,
+      "learning_rate": 5e-06,
+      "loss": 0.1516,
+      "step": 11714
+    },
+    {
+      "epoch": 8.818216033120061,
+      "grad_norm": 0.19314078986644745,
+      "learning_rate": 5e-06,
+      "loss": 0.2151,
+      "step": 11715
+    },
+    {
+      "epoch": 8.818968761761385,
+      "grad_norm": 0.18336336314678192,
+      "learning_rate": 5e-06,
+      "loss": 0.1548,
+      "step": 11716
+    },
+    {
+      "epoch": 8.81972149040271,
+      "grad_norm": 0.18982209265232086,
+      "learning_rate": 5e-06,
+      "loss": 0.1421,
+      "step": 11717
+    },
+    {
+      "epoch": 8.820474219044035,
+      "grad_norm": 0.20082926750183105,
+      "learning_rate": 5e-06,
+      "loss": 0.1564,
+      "step": 11718
+    },
+    {
+      "epoch": 8.821226947685359,
+      "grad_norm": 0.2230597734451294,
+      "learning_rate": 5e-06,
+      "loss": 0.118,
+      "step": 11719
+    },
+    {
+      "epoch": 8.821979676326684,
+      "grad_norm": 0.18549251556396484,
+      "learning_rate": 5e-06,
+      "loss": 0.1801,
+      "step": 11720
+    },
+    {
+      "epoch": 8.82273240496801,
+      "grad_norm": 0.17268063127994537,
+      "learning_rate": 5e-06,
+      "loss": 0.1586,
+      "step": 11721
+    },
+    {
+      "epoch": 8.823485133609333,
+      "grad_norm": 0.19823817908763885,
+      "learning_rate": 5e-06,
+      "loss": 0.1158,
+      "step": 11722
+    },
+    {
+      "epoch": 8.824237862250659,
+      "grad_norm": 0.1883580982685089,
+      "learning_rate": 5e-06,
+      "loss": 0.1554,
+      "step": 11723
+    },
+    {
+      "epoch": 8.824990590891984,
+      "grad_norm": 0.20897682011127472,
+      "learning_rate": 5e-06,
+      "loss": 0.1801,
+      "step": 11724
+    },
+    {
+      "epoch": 8.825743319533307,
+      "grad_norm": 0.1544288992881775,
+      "learning_rate": 5e-06,
+      "loss": 0.0934,
+      "step": 11725
+    },
+    {
+      "epoch": 8.826496048174633,
+      "grad_norm": 0.14718890190124512,
+      "learning_rate": 5e-06,
+      "loss": 0.1137,
+      "step": 11726
+    },
+    {
+      "epoch": 8.827248776815958,
+      "grad_norm": 0.17500944435596466,
+      "learning_rate": 5e-06,
+      "loss": 0.1465,
+      "step": 11727
+    },
+    {
+      "epoch": 8.828001505457284,
+      "grad_norm": 0.22084517776966095,
+      "learning_rate": 5e-06,
+      "loss": 0.2651,
+      "step": 11728
+    },
+    {
+      "epoch": 8.828754234098607,
+      "grad_norm": 0.1936887949705124,
+      "learning_rate": 5e-06,
+      "loss": 0.1434,
+      "step": 11729
+    },
+    {
+      "epoch": 8.829506962739933,
+      "grad_norm": 0.16417622566223145,
+      "learning_rate": 5e-06,
+      "loss": 0.123,
+      "step": 11730
+    },
+    {
+      "epoch": 8.830259691381258,
+      "grad_norm": 0.17403170466423035,
+      "learning_rate": 5e-06,
+      "loss": 0.1348,
+      "step": 11731
+    },
+    {
+      "epoch": 8.831012420022581,
+      "grad_norm": 0.1818130612373352,
+      "learning_rate": 5e-06,
+      "loss": 0.1472,
+      "step": 11732
+    },
+    {
+      "epoch": 8.831765148663907,
+      "grad_norm": 0.14244620501995087,
+      "learning_rate": 5e-06,
+      "loss": 0.1029,
+      "step": 11733
+    },
+    {
+      "epoch": 8.832517877305232,
+      "grad_norm": 0.19228389859199524,
+      "learning_rate": 5e-06,
+      "loss": 0.1512,
+      "step": 11734
+    },
+    {
+      "epoch": 8.833270605946556,
+      "grad_norm": 0.1848372220993042,
+      "learning_rate": 5e-06,
+      "loss": 0.0821,
+      "step": 11735
+    },
+    {
+      "epoch": 8.834023334587881,
+      "grad_norm": 0.199353888630867,
+      "learning_rate": 5e-06,
+      "loss": 0.2546,
+      "step": 11736
+    },
+    {
+      "epoch": 8.834776063229206,
+      "grad_norm": 0.19025391340255737,
+      "learning_rate": 5e-06,
+      "loss": 0.1541,
+      "step": 11737
+    },
+    {
+      "epoch": 8.83552879187053,
+      "grad_norm": 0.18119597434997559,
+      "learning_rate": 5e-06,
+      "loss": 0.118,
+      "step": 11738
+    },
+    {
+      "epoch": 8.836281520511855,
+      "grad_norm": 0.19007112085819244,
+      "learning_rate": 5e-06,
+      "loss": 0.1407,
+      "step": 11739
+    },
+    {
+      "epoch": 8.83703424915318,
+      "grad_norm": 0.18464913964271545,
+      "learning_rate": 5e-06,
+      "loss": 0.1276,
+      "step": 11740
+    },
+    {
+      "epoch": 8.837786977794504,
+      "grad_norm": 0.1830991953611374,
+      "learning_rate": 5e-06,
+      "loss": 0.1347,
+      "step": 11741
+    },
+    {
+      "epoch": 8.83853970643583,
+      "grad_norm": 0.18338651955127716,
+      "learning_rate": 5e-06,
+      "loss": 0.1243,
+      "step": 11742
+    },
+    {
+      "epoch": 8.839292435077155,
+      "grad_norm": 0.16346631944179535,
+      "learning_rate": 5e-06,
+      "loss": 0.1781,
+      "step": 11743
+    },
+    {
+      "epoch": 8.840045163718479,
+      "grad_norm": 0.16194523870944977,
+      "learning_rate": 5e-06,
+      "loss": 0.152,
+      "step": 11744
+    },
+    {
+      "epoch": 8.840797892359804,
+      "grad_norm": 0.1930667757987976,
+      "learning_rate": 5e-06,
+      "loss": 0.1753,
+      "step": 11745
+    },
+    {
+      "epoch": 8.84155062100113,
+      "grad_norm": 0.15233972668647766,
+      "learning_rate": 5e-06,
+      "loss": 0.1485,
+      "step": 11746
+    },
+    {
+      "epoch": 8.842303349642453,
+      "grad_norm": 0.17848673462867737,
+      "learning_rate": 5e-06,
+      "loss": 0.2116,
+      "step": 11747
+    },
+    {
+      "epoch": 8.843056078283778,
+      "grad_norm": 0.16692833602428436,
+      "learning_rate": 5e-06,
+      "loss": 0.165,
+      "step": 11748
+    },
+    {
+      "epoch": 8.843808806925104,
+      "grad_norm": 0.16354449093341827,
+      "learning_rate": 5e-06,
+      "loss": 0.1082,
+      "step": 11749
+    },
+    {
+      "epoch": 8.844561535566429,
+      "grad_norm": 0.18093915283679962,
+      "learning_rate": 5e-06,
+      "loss": 0.164,
+      "step": 11750
+    },
+    {
+      "epoch": 8.845314264207753,
+      "grad_norm": 0.17687125504016876,
+      "learning_rate": 5e-06,
+      "loss": 0.1531,
+      "step": 11751
+    },
+    {
+      "epoch": 8.846066992849078,
+      "grad_norm": 0.19536571204662323,
+      "learning_rate": 5e-06,
+      "loss": 0.179,
+      "step": 11752
+    },
+    {
+      "epoch": 8.846819721490403,
+      "grad_norm": 0.17410778999328613,
+      "learning_rate": 5e-06,
+      "loss": 0.1332,
+      "step": 11753
+    },
+    {
+      "epoch": 8.847572450131727,
+      "grad_norm": 0.19186529517173767,
+      "learning_rate": 5e-06,
+      "loss": 0.1715,
+      "step": 11754
+    },
+    {
+      "epoch": 8.848325178773052,
+      "grad_norm": 0.18934406340122223,
+      "learning_rate": 5e-06,
+      "loss": 0.1733,
+      "step": 11755
+    },
+    {
+      "epoch": 8.849077907414378,
+      "grad_norm": 0.1787523478269577,
+      "learning_rate": 5e-06,
+      "loss": 0.1927,
+      "step": 11756
+    },
+    {
+      "epoch": 8.849830636055701,
+      "grad_norm": 0.18363237380981445,
+      "learning_rate": 5e-06,
+      "loss": 0.1936,
+      "step": 11757
+    },
+    {
+      "epoch": 8.850583364697027,
+      "grad_norm": 0.21074816584587097,
+      "learning_rate": 5e-06,
+      "loss": 0.2402,
+      "step": 11758
+    },
+    {
+      "epoch": 8.851336093338352,
+      "grad_norm": 0.17756055295467377,
+      "learning_rate": 5e-06,
+      "loss": 0.1882,
+      "step": 11759
+    },
+    {
+      "epoch": 8.852088821979676,
+      "grad_norm": 0.19215095043182373,
+      "learning_rate": 5e-06,
+      "loss": 0.1537,
+      "step": 11760
+    },
+    {
+      "epoch": 8.852841550621001,
+      "grad_norm": 0.1907804012298584,
+      "learning_rate": 5e-06,
+      "loss": 0.1581,
+      "step": 11761
+    },
+    {
+      "epoch": 8.853594279262326,
+      "grad_norm": 0.1742510050535202,
+      "learning_rate": 5e-06,
+      "loss": 0.2351,
+      "step": 11762
+    },
+    {
+      "epoch": 8.85434700790365,
+      "grad_norm": 0.16143493354320526,
+      "learning_rate": 5e-06,
+      "loss": 0.2049,
+      "step": 11763
+    },
+    {
+      "epoch": 8.855099736544975,
+      "grad_norm": 0.20004698634147644,
+      "learning_rate": 5e-06,
+      "loss": 0.2062,
+      "step": 11764
+    },
+    {
+      "epoch": 8.8558524651863,
+      "grad_norm": 0.17395012080669403,
+      "learning_rate": 5e-06,
+      "loss": 0.2148,
+      "step": 11765
+    },
+    {
+      "epoch": 8.856605193827626,
+      "grad_norm": 0.17450839281082153,
+      "learning_rate": 5e-06,
+      "loss": 0.1644,
+      "step": 11766
+    },
+    {
+      "epoch": 8.85735792246895,
+      "grad_norm": 0.16925694048404694,
+      "learning_rate": 5e-06,
+      "loss": 0.1348,
+      "step": 11767
+    },
+    {
+      "epoch": 8.858110651110275,
+      "grad_norm": 0.17106890678405762,
+      "learning_rate": 5e-06,
+      "loss": 0.1247,
+      "step": 11768
+    },
+    {
+      "epoch": 8.8588633797516,
+      "grad_norm": 0.1796817183494568,
+      "learning_rate": 5e-06,
+      "loss": 0.1466,
+      "step": 11769
+    },
+    {
+      "epoch": 8.859616108392924,
+      "grad_norm": 0.19518984854221344,
+      "learning_rate": 5e-06,
+      "loss": 0.1708,
+      "step": 11770
+    },
+    {
+      "epoch": 8.86036883703425,
+      "grad_norm": 0.1936909407377243,
+      "learning_rate": 5e-06,
+      "loss": 0.1736,
+      "step": 11771
+    },
+    {
+      "epoch": 8.861121565675575,
+      "grad_norm": 0.20622529089450836,
+      "learning_rate": 5e-06,
+      "loss": 0.1544,
+      "step": 11772
+    },
+    {
+      "epoch": 8.861874294316898,
+      "grad_norm": 0.15953341126441956,
+      "learning_rate": 5e-06,
+      "loss": 0.2011,
+      "step": 11773
+    },
+    {
+      "epoch": 8.862627022958224,
+      "grad_norm": 0.18223220109939575,
+      "learning_rate": 5e-06,
+      "loss": 0.1318,
+      "step": 11774
+    },
+    {
+      "epoch": 8.863379751599549,
+      "grad_norm": 0.19629202783107758,
+      "learning_rate": 5e-06,
+      "loss": 0.1845,
+      "step": 11775
+    },
+    {
+      "epoch": 8.864132480240873,
+      "grad_norm": 0.16880595684051514,
+      "learning_rate": 5e-06,
+      "loss": 0.1497,
+      "step": 11776
+    },
+    {
+      "epoch": 8.864885208882198,
+      "grad_norm": 0.15330307185649872,
+      "learning_rate": 5e-06,
+      "loss": 0.137,
+      "step": 11777
+    },
+    {
+      "epoch": 8.865637937523523,
+      "grad_norm": 0.18872800469398499,
+      "learning_rate": 5e-06,
+      "loss": 0.2098,
+      "step": 11778
+    },
+    {
+      "epoch": 8.866390666164847,
+      "grad_norm": 0.16175252199172974,
+      "learning_rate": 5e-06,
+      "loss": 0.1125,
+      "step": 11779
+    },
+    {
+      "epoch": 8.867143394806172,
+      "grad_norm": 0.19765952229499817,
+      "learning_rate": 5e-06,
+      "loss": 0.1896,
+      "step": 11780
+    },
+    {
+      "epoch": 8.867896123447498,
+      "grad_norm": 0.18436235189437866,
+      "learning_rate": 5e-06,
+      "loss": 0.1583,
+      "step": 11781
+    },
+    {
+      "epoch": 8.868648852088821,
+      "grad_norm": 0.1616498976945877,
+      "learning_rate": 5e-06,
+      "loss": 0.1632,
+      "step": 11782
+    },
+    {
+      "epoch": 8.869401580730147,
+      "grad_norm": 0.17287583649158478,
+      "learning_rate": 5e-06,
+      "loss": 0.1715,
+      "step": 11783
+    },
+    {
+      "epoch": 8.870154309371472,
+      "grad_norm": 0.15332061052322388,
+      "learning_rate": 5e-06,
+      "loss": 0.0806,
+      "step": 11784
+    },
+    {
+      "epoch": 8.870907038012795,
+      "grad_norm": 0.18374748528003693,
+      "learning_rate": 5e-06,
+      "loss": 0.1154,
+      "step": 11785
+    },
+    {
+      "epoch": 8.87165976665412,
+      "grad_norm": 0.16204285621643066,
+      "learning_rate": 5e-06,
+      "loss": 0.0821,
+      "step": 11786
+    },
+    {
+      "epoch": 8.872412495295446,
+      "grad_norm": 0.17513510584831238,
+      "learning_rate": 5e-06,
+      "loss": 0.1297,
+      "step": 11787
+    },
+    {
+      "epoch": 8.873165223936772,
+      "grad_norm": 0.16438938677310944,
+      "learning_rate": 5e-06,
+      "loss": 0.1251,
+      "step": 11788
+    },
+    {
+      "epoch": 8.873917952578095,
+      "grad_norm": 0.18415729701519012,
+      "learning_rate": 5e-06,
+      "loss": 0.1708,
+      "step": 11789
+    },
+    {
+      "epoch": 8.87467068121942,
+      "grad_norm": 0.15218622982501984,
+      "learning_rate": 5e-06,
+      "loss": 0.1678,
+      "step": 11790
+    },
+    {
+      "epoch": 8.875423409860746,
+      "grad_norm": 0.19419732689857483,
+      "learning_rate": 5e-06,
+      "loss": 0.2173,
+      "step": 11791
+    },
+    {
+      "epoch": 8.87617613850207,
+      "grad_norm": 0.19110365211963654,
+      "learning_rate": 5e-06,
+      "loss": 0.203,
+      "step": 11792
+    },
+    {
+      "epoch": 8.876928867143395,
+      "grad_norm": 0.17707471549510956,
+      "learning_rate": 5e-06,
+      "loss": 0.1617,
+      "step": 11793
+    },
+    {
+      "epoch": 8.87768159578472,
+      "grad_norm": 0.18072764575481415,
+      "learning_rate": 5e-06,
+      "loss": 0.2283,
+      "step": 11794
+    },
+    {
+      "epoch": 8.878434324426044,
+      "grad_norm": 0.16602873802185059,
+      "learning_rate": 5e-06,
+      "loss": 0.0984,
+      "step": 11795
+    },
+    {
+      "epoch": 8.879187053067369,
+      "grad_norm": 0.18353211879730225,
+      "learning_rate": 5e-06,
+      "loss": 0.1187,
+      "step": 11796
+    },
+    {
+      "epoch": 8.879939781708694,
+      "grad_norm": 0.205003559589386,
+      "learning_rate": 5e-06,
+      "loss": 0.2209,
+      "step": 11797
+    },
+    {
+      "epoch": 8.880692510350018,
+      "grad_norm": 0.17502473294734955,
+      "learning_rate": 5e-06,
+      "loss": 0.1261,
+      "step": 11798
+    },
+    {
+      "epoch": 8.881445238991343,
+      "grad_norm": 0.20646178722381592,
+      "learning_rate": 5e-06,
+      "loss": 0.1515,
+      "step": 11799
+    },
+    {
+      "epoch": 8.882197967632669,
+      "grad_norm": 0.1774919331073761,
+      "learning_rate": 5e-06,
+      "loss": 0.1943,
+      "step": 11800
+    },
+    {
+      "epoch": 8.882197967632669,
+      "eval_loss": 0.19354763627052307,
+      "eval_runtime": 452.7095,
+      "eval_samples_per_second": 21.265,
+      "eval_steps_per_second": 0.665,
+      "step": 11800
+    },
+    {
+      "epoch": 8.882950696273994,
+      "grad_norm": 0.173604354262352,
+      "learning_rate": 5e-06,
+      "loss": 0.1468,
+      "step": 11801
+    },
+    {
+      "epoch": 8.883703424915318,
+      "grad_norm": 0.1687273383140564,
+      "learning_rate": 5e-06,
+      "loss": 0.1664,
+      "step": 11802
+    },
+    {
+      "epoch": 8.884456153556643,
+      "grad_norm": 0.17445111274719238,
+      "learning_rate": 5e-06,
+      "loss": 0.1952,
+      "step": 11803
+    },
+    {
+      "epoch": 8.885208882197968,
+      "grad_norm": 0.16556765139102936,
+      "learning_rate": 5e-06,
+      "loss": 0.15,
+      "step": 11804
+    },
+    {
+      "epoch": 8.885961610839292,
+      "grad_norm": 0.17929010093212128,
+      "learning_rate": 5e-06,
+      "loss": 0.1253,
+      "step": 11805
+    },
+    {
+      "epoch": 8.886714339480617,
+      "grad_norm": 0.17018303275108337,
+      "learning_rate": 5e-06,
+      "loss": 0.1049,
+      "step": 11806
+    },
+    {
+      "epoch": 8.887467068121943,
+      "grad_norm": 0.1931876242160797,
+      "learning_rate": 5e-06,
+      "loss": 0.1622,
+      "step": 11807
+    },
+    {
+      "epoch": 8.888219796763266,
+      "grad_norm": 0.18168127536773682,
+      "learning_rate": 5e-06,
+      "loss": 0.1674,
+      "step": 11808
+    },
+    {
+      "epoch": 8.888972525404592,
+      "grad_norm": 0.18205946683883667,
+      "learning_rate": 5e-06,
+      "loss": 0.1746,
+      "step": 11809
+    },
+    {
+      "epoch": 8.889725254045917,
+      "grad_norm": 0.1811320185661316,
+      "learning_rate": 5e-06,
+      "loss": 0.1175,
+      "step": 11810
+    },
+    {
+      "epoch": 8.89047798268724,
+      "grad_norm": 0.16282421350479126,
+      "learning_rate": 5e-06,
+      "loss": 0.1658,
+      "step": 11811
+    },
+    {
+      "epoch": 8.891230711328566,
+      "grad_norm": 0.21012309193611145,
+      "learning_rate": 5e-06,
+      "loss": 0.1767,
+      "step": 11812
+    },
+    {
+      "epoch": 8.891983439969891,
+      "grad_norm": 0.164343923330307,
+      "learning_rate": 5e-06,
+      "loss": 0.2029,
+      "step": 11813
+    },
+    {
+      "epoch": 8.892736168611215,
+      "grad_norm": 0.17726941406726837,
+      "learning_rate": 5e-06,
+      "loss": 0.1747,
+      "step": 11814
+    },
+    {
+      "epoch": 8.89348889725254,
+      "grad_norm": 0.160500168800354,
+      "learning_rate": 5e-06,
+      "loss": 0.1262,
+      "step": 11815
+    },
+    {
+      "epoch": 8.894241625893866,
+      "grad_norm": 0.16256028413772583,
+      "learning_rate": 5e-06,
+      "loss": 0.1564,
+      "step": 11816
+    },
+    {
+      "epoch": 8.89499435453519,
+      "grad_norm": 0.1970326006412506,
+      "learning_rate": 5e-06,
+      "loss": 0.1775,
+      "step": 11817
+    },
+    {
+      "epoch": 8.895747083176515,
+      "grad_norm": 0.17011556029319763,
+      "learning_rate": 5e-06,
+      "loss": 0.1761,
+      "step": 11818
+    },
+    {
+      "epoch": 8.89649981181784,
+      "grad_norm": 0.19023136794567108,
+      "learning_rate": 5e-06,
+      "loss": 0.184,
+      "step": 11819
+    },
+    {
+      "epoch": 8.897252540459164,
+      "grad_norm": 0.19650644063949585,
+      "learning_rate": 5e-06,
+      "loss": 0.1387,
+      "step": 11820
+    },
+    {
+      "epoch": 8.898005269100489,
+      "grad_norm": 0.18941757082939148,
+      "learning_rate": 5e-06,
+      "loss": 0.1387,
+      "step": 11821
+    },
+    {
+      "epoch": 8.898757997741814,
+      "grad_norm": 0.1886523962020874,
+      "learning_rate": 5e-06,
+      "loss": 0.2194,
+      "step": 11822
+    },
+    {
+      "epoch": 8.89951072638314,
+      "grad_norm": 0.171735480427742,
+      "learning_rate": 5e-06,
+      "loss": 0.1156,
+      "step": 11823
+    },
+    {
+      "epoch": 8.900263455024463,
+      "grad_norm": 0.15688437223434448,
+      "learning_rate": 5e-06,
+      "loss": 0.0913,
+      "step": 11824
+    },
+    {
+      "epoch": 8.901016183665789,
+      "grad_norm": 0.16523554921150208,
+      "learning_rate": 5e-06,
+      "loss": 0.082,
+      "step": 11825
+    },
+    {
+      "epoch": 8.901768912307114,
+      "grad_norm": 0.22653061151504517,
+      "learning_rate": 5e-06,
+      "loss": 0.182,
+      "step": 11826
+    },
+    {
+      "epoch": 8.902521640948438,
+      "grad_norm": 0.17501500248908997,
+      "learning_rate": 5e-06,
+      "loss": 0.177,
+      "step": 11827
+    },
+    {
+      "epoch": 8.903274369589763,
+      "grad_norm": 0.15794745087623596,
+      "learning_rate": 5e-06,
+      "loss": 0.1483,
+      "step": 11828
+    },
+    {
+      "epoch": 8.904027098231088,
+      "grad_norm": 0.1651669591665268,
+      "learning_rate": 5e-06,
+      "loss": 0.092,
+      "step": 11829
+    },
+    {
+      "epoch": 8.904779826872412,
+      "grad_norm": 0.1993076503276825,
+      "learning_rate": 5e-06,
+      "loss": 0.1479,
+      "step": 11830
+    },
+    {
+      "epoch": 8.905532555513737,
+      "grad_norm": 0.18269336223602295,
+      "learning_rate": 5e-06,
+      "loss": 0.1315,
+      "step": 11831
+    },
+    {
+      "epoch": 8.906285284155063,
+      "grad_norm": 0.17535845935344696,
+      "learning_rate": 5e-06,
+      "loss": 0.1391,
+      "step": 11832
+    },
+    {
+      "epoch": 8.907038012796386,
+      "grad_norm": 0.19157323241233826,
+      "learning_rate": 5e-06,
+      "loss": 0.1742,
+      "step": 11833
+    },
+    {
+      "epoch": 8.907790741437712,
+      "grad_norm": 0.23719558119773865,
+      "learning_rate": 5e-06,
+      "loss": 0.2294,
+      "step": 11834
+    },
+    {
+      "epoch": 8.908543470079037,
+      "grad_norm": 0.16055700182914734,
+      "learning_rate": 5e-06,
+      "loss": 0.1446,
+      "step": 11835
+    },
+    {
+      "epoch": 8.90929619872036,
+      "grad_norm": 0.17068053781986237,
+      "learning_rate": 5e-06,
+      "loss": 0.1741,
+      "step": 11836
+    },
+    {
+      "epoch": 8.910048927361686,
+      "grad_norm": 0.17213885486125946,
+      "learning_rate": 5e-06,
+      "loss": 0.1276,
+      "step": 11837
+    },
+    {
+      "epoch": 8.910801656003011,
+      "grad_norm": 0.17770534753799438,
+      "learning_rate": 5e-06,
+      "loss": 0.1687,
+      "step": 11838
+    },
+    {
+      "epoch": 8.911554384644337,
+      "grad_norm": 0.16638332605361938,
+      "learning_rate": 5e-06,
+      "loss": 0.0942,
+      "step": 11839
+    },
+    {
+      "epoch": 8.91230711328566,
+      "grad_norm": 0.19274519383907318,
+      "learning_rate": 5e-06,
+      "loss": 0.15,
+      "step": 11840
+    },
+    {
+      "epoch": 8.913059841926986,
+      "grad_norm": 0.1963534951210022,
+      "learning_rate": 5e-06,
+      "loss": 0.1617,
+      "step": 11841
+    },
+    {
+      "epoch": 8.913812570568311,
+      "grad_norm": 0.17717215418815613,
+      "learning_rate": 5e-06,
+      "loss": 0.145,
+      "step": 11842
+    },
+    {
+      "epoch": 8.914565299209634,
+      "grad_norm": 0.16930082440376282,
+      "learning_rate": 5e-06,
+      "loss": 0.135,
+      "step": 11843
+    },
+    {
+      "epoch": 8.91531802785096,
+      "grad_norm": 0.1811845302581787,
+      "learning_rate": 5e-06,
+      "loss": 0.168,
+      "step": 11844
+    },
+    {
+      "epoch": 8.916070756492285,
+      "grad_norm": 0.17354819178581238,
+      "learning_rate": 5e-06,
+      "loss": 0.1084,
+      "step": 11845
+    },
+    {
+      "epoch": 8.916823485133609,
+      "grad_norm": 0.1834772676229477,
+      "learning_rate": 5e-06,
+      "loss": 0.1726,
+      "step": 11846
+    },
+    {
+      "epoch": 8.917576213774934,
+      "grad_norm": 0.18721860647201538,
+      "learning_rate": 5e-06,
+      "loss": 0.1624,
+      "step": 11847
+    },
+    {
+      "epoch": 8.91832894241626,
+      "grad_norm": 0.18124884366989136,
+      "learning_rate": 5e-06,
+      "loss": 0.1791,
+      "step": 11848
+    },
+    {
+      "epoch": 8.919081671057583,
+      "grad_norm": 0.18834614753723145,
+      "learning_rate": 5e-06,
+      "loss": 0.1558,
+      "step": 11849
+    },
+    {
+      "epoch": 8.919834399698908,
+      "grad_norm": 0.16549237072467804,
+      "learning_rate": 5e-06,
+      "loss": 0.2305,
+      "step": 11850
+    },
+    {
+      "epoch": 8.920587128340234,
+      "grad_norm": 0.18469175696372986,
+      "learning_rate": 5e-06,
+      "loss": 0.1309,
+      "step": 11851
+    },
+    {
+      "epoch": 8.921339856981557,
+      "grad_norm": 0.16259720921516418,
+      "learning_rate": 5e-06,
+      "loss": 0.2061,
+      "step": 11852
+    },
+    {
+      "epoch": 8.922092585622883,
+      "grad_norm": 0.16517768800258636,
+      "learning_rate": 5e-06,
+      "loss": 0.1236,
+      "step": 11853
+    },
+    {
+      "epoch": 8.922845314264208,
+      "grad_norm": 0.17075006663799286,
+      "learning_rate": 5e-06,
+      "loss": 0.1882,
+      "step": 11854
+    },
+    {
+      "epoch": 8.923598042905532,
+      "grad_norm": 0.15812206268310547,
+      "learning_rate": 5e-06,
+      "loss": 0.1686,
+      "step": 11855
+    },
+    {
+      "epoch": 8.924350771546857,
+      "grad_norm": 0.18838423490524292,
+      "learning_rate": 5e-06,
+      "loss": 0.2285,
+      "step": 11856
+    },
+    {
+      "epoch": 8.925103500188182,
+      "grad_norm": 0.17621509730815887,
+      "learning_rate": 5e-06,
+      "loss": 0.135,
+      "step": 11857
+    },
+    {
+      "epoch": 8.925856228829506,
+      "grad_norm": 0.1834542155265808,
+      "learning_rate": 5e-06,
+      "loss": 0.1308,
+      "step": 11858
+    },
+    {
+      "epoch": 8.926608957470831,
+      "grad_norm": 0.16746188700199127,
+      "learning_rate": 5e-06,
+      "loss": 0.1369,
+      "step": 11859
+    },
+    {
+      "epoch": 8.927361686112157,
+      "grad_norm": 0.15655604004859924,
+      "learning_rate": 5e-06,
+      "loss": 0.1295,
+      "step": 11860
+    },
+    {
+      "epoch": 8.928114414753482,
+      "grad_norm": 0.191427543759346,
+      "learning_rate": 5e-06,
+      "loss": 0.1764,
+      "step": 11861
+    },
+    {
+      "epoch": 8.928867143394806,
+      "grad_norm": 0.17524616420269012,
+      "learning_rate": 5e-06,
+      "loss": 0.102,
+      "step": 11862
+    },
+    {
+      "epoch": 8.929619872036131,
+      "grad_norm": 0.18278086185455322,
+      "learning_rate": 5e-06,
+      "loss": 0.1745,
+      "step": 11863
+    },
+    {
+      "epoch": 8.930372600677456,
+      "grad_norm": 0.16895221173763275,
+      "learning_rate": 5e-06,
+      "loss": 0.1579,
+      "step": 11864
+    },
+    {
+      "epoch": 8.93112532931878,
+      "grad_norm": 0.1637403964996338,
+      "learning_rate": 5e-06,
+      "loss": 0.1717,
+      "step": 11865
+    },
+    {
+      "epoch": 8.931878057960105,
+      "grad_norm": 0.18127627670764923,
+      "learning_rate": 5e-06,
+      "loss": 0.196,
+      "step": 11866
+    },
+    {
+      "epoch": 8.93263078660143,
+      "grad_norm": 0.1758916825056076,
+      "learning_rate": 5e-06,
+      "loss": 0.0979,
+      "step": 11867
+    },
+    {
+      "epoch": 8.933383515242754,
+      "grad_norm": 0.18173472583293915,
+      "learning_rate": 5e-06,
+      "loss": 0.1613,
+      "step": 11868
+    },
+    {
+      "epoch": 8.93413624388408,
+      "grad_norm": 0.16113556921482086,
+      "learning_rate": 5e-06,
+      "loss": 0.1384,
+      "step": 11869
+    },
+    {
+      "epoch": 8.934888972525405,
+      "grad_norm": 0.1876997947692871,
+      "learning_rate": 5e-06,
+      "loss": 0.1635,
+      "step": 11870
+    },
+    {
+      "epoch": 8.935641701166729,
+      "grad_norm": 0.1901046335697174,
+      "learning_rate": 5e-06,
+      "loss": 0.1203,
+      "step": 11871
+    },
+    {
+      "epoch": 8.936394429808054,
+      "grad_norm": 0.1459839940071106,
+      "learning_rate": 5e-06,
+      "loss": 0.1454,
+      "step": 11872
+    },
+    {
+      "epoch": 8.93714715844938,
+      "grad_norm": 0.18076254427433014,
+      "learning_rate": 5e-06,
+      "loss": 0.1798,
+      "step": 11873
+    },
+    {
+      "epoch": 8.937899887090705,
+      "grad_norm": 0.1746046245098114,
+      "learning_rate": 5e-06,
+      "loss": 0.101,
+      "step": 11874
+    },
+    {
+      "epoch": 8.938652615732028,
+      "grad_norm": 0.21190625429153442,
+      "learning_rate": 5e-06,
+      "loss": 0.1943,
+      "step": 11875
+    },
+    {
+      "epoch": 8.939405344373354,
+      "grad_norm": 0.1814216524362564,
+      "learning_rate": 5e-06,
+      "loss": 0.1876,
+      "step": 11876
+    },
+    {
+      "epoch": 8.940158073014679,
+      "grad_norm": 0.16758081316947937,
+      "learning_rate": 5e-06,
+      "loss": 0.1571,
+      "step": 11877
+    },
+    {
+      "epoch": 8.940910801656003,
+      "grad_norm": 0.1685776561498642,
+      "learning_rate": 5e-06,
+      "loss": 0.1296,
+      "step": 11878
+    },
+    {
+      "epoch": 8.941663530297328,
+      "grad_norm": 0.1602701097726822,
+      "learning_rate": 5e-06,
+      "loss": 0.1364,
+      "step": 11879
+    },
+    {
+      "epoch": 8.942416258938653,
+      "grad_norm": 0.20016252994537354,
+      "learning_rate": 5e-06,
+      "loss": 0.1871,
+      "step": 11880
+    },
+    {
+      "epoch": 8.943168987579977,
+      "grad_norm": 0.1928292214870453,
+      "learning_rate": 5e-06,
+      "loss": 0.2336,
+      "step": 11881
+    },
+    {
+      "epoch": 8.943921716221302,
+      "grad_norm": 0.17544451355934143,
+      "learning_rate": 5e-06,
+      "loss": 0.1759,
+      "step": 11882
+    },
+    {
+      "epoch": 8.944674444862628,
+      "grad_norm": 0.165851429104805,
+      "learning_rate": 5e-06,
+      "loss": 0.1389,
+      "step": 11883
+    },
+    {
+      "epoch": 8.945427173503951,
+      "grad_norm": 0.19666868448257446,
+      "learning_rate": 5e-06,
+      "loss": 0.1452,
+      "step": 11884
+    },
+    {
+      "epoch": 8.946179902145277,
+      "grad_norm": 0.16486501693725586,
+      "learning_rate": 5e-06,
+      "loss": 0.1058,
+      "step": 11885
+    },
+    {
+      "epoch": 8.946932630786602,
+      "grad_norm": 0.1694200187921524,
+      "learning_rate": 5e-06,
+      "loss": 0.1403,
+      "step": 11886
+    },
+    {
+      "epoch": 8.947685359427926,
+      "grad_norm": 0.17212772369384766,
+      "learning_rate": 5e-06,
+      "loss": 0.0675,
+      "step": 11887
+    },
+    {
+      "epoch": 8.948438088069251,
+      "grad_norm": 0.17751049995422363,
+      "learning_rate": 5e-06,
+      "loss": 0.129,
+      "step": 11888
+    },
+    {
+      "epoch": 8.949190816710576,
+      "grad_norm": 0.14849898219108582,
+      "learning_rate": 5e-06,
+      "loss": 0.1104,
+      "step": 11889
+    },
+    {
+      "epoch": 8.9499435453519,
+      "grad_norm": 0.16848379373550415,
+      "learning_rate": 5e-06,
+      "loss": 0.1394,
+      "step": 11890
+    },
+    {
+      "epoch": 8.950696273993225,
+      "grad_norm": 0.1694124937057495,
+      "learning_rate": 5e-06,
+      "loss": 0.1484,
+      "step": 11891
+    },
+    {
+      "epoch": 8.95144900263455,
+      "grad_norm": 0.19761042296886444,
+      "learning_rate": 5e-06,
+      "loss": 0.2157,
+      "step": 11892
+    },
+    {
+      "epoch": 8.952201731275874,
+      "grad_norm": 0.19928912818431854,
+      "learning_rate": 5e-06,
+      "loss": 0.1912,
+      "step": 11893
+    },
+    {
+      "epoch": 8.9529544599172,
+      "grad_norm": 0.1539842188358307,
+      "learning_rate": 5e-06,
+      "loss": 0.1486,
+      "step": 11894
+    },
+    {
+      "epoch": 8.953707188558525,
+      "grad_norm": 0.1819932609796524,
+      "learning_rate": 5e-06,
+      "loss": 0.1539,
+      "step": 11895
+    },
+    {
+      "epoch": 8.95445991719985,
+      "grad_norm": 0.1882627308368683,
+      "learning_rate": 5e-06,
+      "loss": 0.156,
+      "step": 11896
+    },
+    {
+      "epoch": 8.955212645841174,
+      "grad_norm": 0.17244301736354828,
+      "learning_rate": 5e-06,
+      "loss": 0.2613,
+      "step": 11897
+    },
+    {
+      "epoch": 8.9559653744825,
+      "grad_norm": 0.16105450689792633,
+      "learning_rate": 5e-06,
+      "loss": 0.2432,
+      "step": 11898
+    },
+    {
+      "epoch": 8.956718103123825,
+      "grad_norm": 0.16366705298423767,
+      "learning_rate": 5e-06,
+      "loss": 0.143,
+      "step": 11899
+    },
+    {
+      "epoch": 8.957470831765148,
+      "grad_norm": 0.1751248836517334,
+      "learning_rate": 5e-06,
+      "loss": 0.1297,
+      "step": 11900
+    },
+    {
+      "epoch": 8.958223560406473,
+      "grad_norm": 0.1880803406238556,
+      "learning_rate": 5e-06,
+      "loss": 0.193,
+      "step": 11901
+    },
+    {
+      "epoch": 8.958976289047799,
+      "grad_norm": 0.17100541293621063,
+      "learning_rate": 5e-06,
+      "loss": 0.216,
+      "step": 11902
+    },
+    {
+      "epoch": 8.959729017689122,
+      "grad_norm": 0.205710768699646,
+      "learning_rate": 5e-06,
+      "loss": 0.171,
+      "step": 11903
+    },
+    {
+      "epoch": 8.960481746330448,
+      "grad_norm": 0.16469530761241913,
+      "learning_rate": 5e-06,
+      "loss": 0.226,
+      "step": 11904
+    },
+    {
+      "epoch": 8.961234474971773,
+      "grad_norm": 0.16700541973114014,
+      "learning_rate": 5e-06,
+      "loss": 0.0811,
+      "step": 11905
+    },
+    {
+      "epoch": 8.961987203613097,
+      "grad_norm": 0.17498938739299774,
+      "learning_rate": 5e-06,
+      "loss": 0.133,
+      "step": 11906
+    },
+    {
+      "epoch": 8.962739932254422,
+      "grad_norm": 0.1732226461172104,
+      "learning_rate": 5e-06,
+      "loss": 0.1134,
+      "step": 11907
+    },
+    {
+      "epoch": 8.963492660895747,
+      "grad_norm": 0.17992785573005676,
+      "learning_rate": 5e-06,
+      "loss": 0.2753,
+      "step": 11908
+    },
+    {
+      "epoch": 8.964245389537071,
+      "grad_norm": 0.1713476926088333,
+      "learning_rate": 5e-06,
+      "loss": 0.14,
+      "step": 11909
+    },
+    {
+      "epoch": 8.964998118178396,
+      "grad_norm": 0.18164068460464478,
+      "learning_rate": 5e-06,
+      "loss": 0.1238,
+      "step": 11910
+    },
+    {
+      "epoch": 8.965750846819722,
+      "grad_norm": 0.17195403575897217,
+      "learning_rate": 5e-06,
+      "loss": 0.1196,
+      "step": 11911
+    },
+    {
+      "epoch": 8.966503575461047,
+      "grad_norm": 0.18191063404083252,
+      "learning_rate": 5e-06,
+      "loss": 0.1535,
+      "step": 11912
+    },
+    {
+      "epoch": 8.96725630410237,
+      "grad_norm": 0.14541253447532654,
+      "learning_rate": 5e-06,
+      "loss": 0.1157,
+      "step": 11913
+    },
+    {
+      "epoch": 8.968009032743696,
+      "grad_norm": 0.19183069467544556,
+      "learning_rate": 5e-06,
+      "loss": 0.3376,
+      "step": 11914
+    },
+    {
+      "epoch": 8.968761761385021,
+      "grad_norm": 0.16781258583068848,
+      "learning_rate": 5e-06,
+      "loss": 0.0826,
+      "step": 11915
+    },
+    {
+      "epoch": 8.969514490026345,
+      "grad_norm": 0.18170441687107086,
+      "learning_rate": 5e-06,
+      "loss": 0.1839,
+      "step": 11916
+    },
+    {
+      "epoch": 8.97026721866767,
+      "grad_norm": 0.1898699700832367,
+      "learning_rate": 5e-06,
+      "loss": 0.1974,
+      "step": 11917
+    },
+    {
+      "epoch": 8.971019947308996,
+      "grad_norm": 0.1729084849357605,
+      "learning_rate": 5e-06,
+      "loss": 0.2012,
+      "step": 11918
+    },
+    {
+      "epoch": 8.97177267595032,
+      "grad_norm": 0.15374675393104553,
+      "learning_rate": 5e-06,
+      "loss": 0.1357,
+      "step": 11919
+    },
+    {
+      "epoch": 8.972525404591645,
+      "grad_norm": 0.18405117094516754,
+      "learning_rate": 5e-06,
+      "loss": 0.1629,
+      "step": 11920
+    },
+    {
+      "epoch": 8.97327813323297,
+      "grad_norm": 0.17919611930847168,
+      "learning_rate": 5e-06,
+      "loss": 0.1431,
+      "step": 11921
+    },
+    {
+      "epoch": 8.974030861874294,
+      "grad_norm": 0.16845791041851044,
+      "learning_rate": 5e-06,
+      "loss": 0.1858,
+      "step": 11922
+    },
+    {
+      "epoch": 8.974783590515619,
+      "grad_norm": 0.1806272268295288,
+      "learning_rate": 5e-06,
+      "loss": 0.1316,
+      "step": 11923
+    },
+    {
+      "epoch": 8.975536319156944,
+      "grad_norm": 0.1789119988679886,
+      "learning_rate": 5e-06,
+      "loss": 0.1533,
+      "step": 11924
+    },
+    {
+      "epoch": 8.976289047798268,
+      "grad_norm": 0.17656229436397552,
+      "learning_rate": 5e-06,
+      "loss": 0.1729,
+      "step": 11925
+    },
+    {
+      "epoch": 8.977041776439593,
+      "grad_norm": 0.17932197451591492,
+      "learning_rate": 5e-06,
+      "loss": 0.1474,
+      "step": 11926
+    },
+    {
+      "epoch": 8.977794505080919,
+      "grad_norm": 0.16752265393733978,
+      "learning_rate": 5e-06,
+      "loss": 0.1152,
+      "step": 11927
+    },
+    {
+      "epoch": 8.978547233722242,
+      "grad_norm": 0.17056146264076233,
+      "learning_rate": 5e-06,
+      "loss": 0.0971,
+      "step": 11928
+    },
+    {
+      "epoch": 8.979299962363568,
+      "grad_norm": 0.19938895106315613,
+      "learning_rate": 5e-06,
+      "loss": 0.15,
+      "step": 11929
+    },
+    {
+      "epoch": 8.980052691004893,
+      "grad_norm": 0.17023980617523193,
+      "learning_rate": 5e-06,
+      "loss": 0.122,
+      "step": 11930
+    },
+    {
+      "epoch": 8.980805419646218,
+      "grad_norm": 0.23474697768688202,
+      "learning_rate": 5e-06,
+      "loss": 0.2171,
+      "step": 11931
+    },
+    {
+      "epoch": 8.981558148287542,
+      "grad_norm": 0.17006070911884308,
+      "learning_rate": 5e-06,
+      "loss": 0.183,
+      "step": 11932
+    },
+    {
+      "epoch": 8.982310876928867,
+      "grad_norm": 0.1608887016773224,
+      "learning_rate": 5e-06,
+      "loss": 0.0896,
+      "step": 11933
+    },
+    {
+      "epoch": 8.983063605570193,
+      "grad_norm": 0.19091439247131348,
+      "learning_rate": 5e-06,
+      "loss": 0.138,
+      "step": 11934
+    },
+    {
+      "epoch": 8.983816334211516,
+      "grad_norm": 0.20576831698417664,
+      "learning_rate": 5e-06,
+      "loss": 0.1805,
+      "step": 11935
+    },
+    {
+      "epoch": 8.984569062852842,
+      "grad_norm": 0.19022385776042938,
+      "learning_rate": 5e-06,
+      "loss": 0.2494,
+      "step": 11936
+    },
+    {
+      "epoch": 8.985321791494167,
+      "grad_norm": 0.18603339791297913,
+      "learning_rate": 5e-06,
+      "loss": 0.1841,
+      "step": 11937
+    },
+    {
+      "epoch": 8.98607452013549,
+      "grad_norm": 0.17628805339336395,
+      "learning_rate": 5e-06,
+      "loss": 0.1656,
+      "step": 11938
+    },
+    {
+      "epoch": 8.986827248776816,
+      "grad_norm": 0.17086617648601532,
+      "learning_rate": 5e-06,
+      "loss": 0.2073,
+      "step": 11939
+    },
+    {
+      "epoch": 8.987579977418141,
+      "grad_norm": 0.1938999742269516,
+      "learning_rate": 5e-06,
+      "loss": 0.142,
+      "step": 11940
+    },
+    {
+      "epoch": 8.988332706059465,
+      "grad_norm": 0.16878709197044373,
+      "learning_rate": 5e-06,
+      "loss": 0.1302,
+      "step": 11941
+    },
+    {
+      "epoch": 8.98908543470079,
+      "grad_norm": 0.18766766786575317,
+      "learning_rate": 5e-06,
+      "loss": 0.1596,
+      "step": 11942
+    },
+    {
+      "epoch": 8.989838163342116,
+      "grad_norm": 0.19408480823040009,
+      "learning_rate": 5e-06,
+      "loss": 0.1928,
+      "step": 11943
+    },
+    {
+      "epoch": 8.99059089198344,
+      "grad_norm": 0.1811264008283615,
+      "learning_rate": 5e-06,
+      "loss": 0.1682,
+      "step": 11944
+    },
+    {
+      "epoch": 8.991343620624765,
+      "grad_norm": 0.17553269863128662,
+      "learning_rate": 5e-06,
+      "loss": 0.1344,
+      "step": 11945
+    },
+    {
+      "epoch": 8.99209634926609,
+      "grad_norm": 0.18646866083145142,
+      "learning_rate": 5e-06,
+      "loss": 0.1238,
+      "step": 11946
+    },
+    {
+      "epoch": 8.992849077907415,
+      "grad_norm": 0.1862083226442337,
+      "learning_rate": 5e-06,
+      "loss": 0.2136,
+      "step": 11947
+    },
+    {
+      "epoch": 8.993601806548739,
+      "grad_norm": 0.1635519415140152,
+      "learning_rate": 5e-06,
+      "loss": 0.1649,
+      "step": 11948
+    },
+    {
+      "epoch": 8.994354535190064,
+      "grad_norm": 0.17185160517692566,
+      "learning_rate": 5e-06,
+      "loss": 0.1023,
+      "step": 11949
+    },
+    {
+      "epoch": 8.99510726383139,
+      "grad_norm": 0.16967397928237915,
+      "learning_rate": 5e-06,
+      "loss": 0.1298,
+      "step": 11950
+    },
+    {
+      "epoch": 8.995859992472713,
+      "grad_norm": 0.18602409958839417,
+      "learning_rate": 5e-06,
+      "loss": 0.0988,
+      "step": 11951
+    },
+    {
+      "epoch": 8.996612721114039,
+      "grad_norm": 0.1789250671863556,
+      "learning_rate": 5e-06,
+      "loss": 0.1646,
+      "step": 11952
+    },
+    {
+      "epoch": 8.997365449755364,
+      "grad_norm": 0.19670698046684265,
+      "learning_rate": 5e-06,
+      "loss": 0.1914,
+      "step": 11953
+    },
+    {
+      "epoch": 8.998118178396687,
+      "grad_norm": 0.18681655824184418,
+      "learning_rate": 5e-06,
+      "loss": 0.088,
+      "step": 11954
+    },
+    {
+      "epoch": 8.998870907038013,
+      "grad_norm": 0.1846577227115631,
+      "learning_rate": 5e-06,
+      "loss": 0.1386,
+      "step": 11955
+    },
+    {
+      "epoch": 8.999623635679338,
+      "grad_norm": 0.183318629860878,
+      "learning_rate": 5e-06,
+      "loss": 0.1249,
+      "step": 11956
+    },
+    {
+      "epoch": 9.000376364320662,
+      "grad_norm": 0.1582030951976776,
+      "learning_rate": 5e-06,
+      "loss": 0.163,
+      "step": 11957
+    },
+    {
+      "epoch": 9.001129092961987,
+      "grad_norm": 0.19558945298194885,
+      "learning_rate": 5e-06,
+      "loss": 0.1464,
+      "step": 11958
+    },
+    {
+      "epoch": 9.001881821603313,
+      "grad_norm": 0.17661327123641968,
+      "learning_rate": 5e-06,
+      "loss": 0.1485,
+      "step": 11959
+    },
+    {
+      "epoch": 9.002634550244636,
+      "grad_norm": 0.19474953413009644,
+      "learning_rate": 5e-06,
+      "loss": 0.1205,
+      "step": 11960
+    },
+    {
+      "epoch": 9.003387278885961,
+      "grad_norm": 0.1876097470521927,
+      "learning_rate": 5e-06,
+      "loss": 0.1626,
+      "step": 11961
+    },
+    {
+      "epoch": 9.004140007527287,
+      "grad_norm": 0.19243982434272766,
+      "learning_rate": 5e-06,
+      "loss": 0.1462,
+      "step": 11962
+    },
+    {
+      "epoch": 9.00489273616861,
+      "grad_norm": 0.16257978975772858,
+      "learning_rate": 5e-06,
+      "loss": 0.1361,
+      "step": 11963
+    },
+    {
+      "epoch": 9.005645464809936,
+      "grad_norm": 0.16735024750232697,
+      "learning_rate": 5e-06,
+      "loss": 0.1702,
+      "step": 11964
+    },
+    {
+      "epoch": 9.006398193451261,
+      "grad_norm": 0.18958741426467896,
+      "learning_rate": 5e-06,
+      "loss": 0.1621,
+      "step": 11965
+    },
+    {
+      "epoch": 9.007150922092586,
+      "grad_norm": 0.1505969613790512,
+      "learning_rate": 5e-06,
+      "loss": 0.0764,
+      "step": 11966
+    },
+    {
+      "epoch": 9.00790365073391,
+      "grad_norm": 0.1575021743774414,
+      "learning_rate": 5e-06,
+      "loss": 0.0897,
+      "step": 11967
+    },
+    {
+      "epoch": 9.008656379375235,
+      "grad_norm": 0.1902446150779724,
+      "learning_rate": 5e-06,
+      "loss": 0.1745,
+      "step": 11968
+    },
+    {
+      "epoch": 9.00940910801656,
+      "grad_norm": 0.18540407717227936,
+      "learning_rate": 5e-06,
+      "loss": 0.1973,
+      "step": 11969
+    },
+    {
+      "epoch": 9.010161836657884,
+      "grad_norm": 0.1851707398891449,
+      "learning_rate": 5e-06,
+      "loss": 0.1433,
+      "step": 11970
+    },
+    {
+      "epoch": 9.01091456529921,
+      "grad_norm": 0.1726054847240448,
+      "learning_rate": 5e-06,
+      "loss": 0.1398,
+      "step": 11971
+    },
+    {
+      "epoch": 9.011667293940535,
+      "grad_norm": 0.18599416315555573,
+      "learning_rate": 5e-06,
+      "loss": 0.1458,
+      "step": 11972
+    },
+    {
+      "epoch": 9.012420022581859,
+      "grad_norm": 0.16732558608055115,
+      "learning_rate": 5e-06,
+      "loss": 0.1589,
+      "step": 11973
+    },
+    {
+      "epoch": 9.013172751223184,
+      "grad_norm": 0.18023140728473663,
+      "learning_rate": 5e-06,
+      "loss": 0.1895,
+      "step": 11974
+    },
+    {
+      "epoch": 9.01392547986451,
+      "grad_norm": 0.17694298923015594,
+      "learning_rate": 5e-06,
+      "loss": 0.1905,
+      "step": 11975
+    },
+    {
+      "epoch": 9.014678208505833,
+      "grad_norm": 0.21718236804008484,
+      "learning_rate": 5e-06,
+      "loss": 0.1366,
+      "step": 11976
+    },
+    {
+      "epoch": 9.015430937147158,
+      "grad_norm": 0.2092822939157486,
+      "learning_rate": 5e-06,
+      "loss": 0.1559,
+      "step": 11977
+    },
+    {
+      "epoch": 9.016183665788484,
+      "grad_norm": 0.17409001290798187,
+      "learning_rate": 5e-06,
+      "loss": 0.1432,
+      "step": 11978
+    },
+    {
+      "epoch": 9.016936394429807,
+      "grad_norm": 0.17608729004859924,
+      "learning_rate": 5e-06,
+      "loss": 0.1565,
+      "step": 11979
+    },
+    {
+      "epoch": 9.017689123071133,
+      "grad_norm": 0.17913685739040375,
+      "learning_rate": 5e-06,
+      "loss": 0.2039,
+      "step": 11980
+    },
+    {
+      "epoch": 9.018441851712458,
+      "grad_norm": 0.18648011982440948,
+      "learning_rate": 5e-06,
+      "loss": 0.2418,
+      "step": 11981
+    },
+    {
+      "epoch": 9.019194580353782,
+      "grad_norm": 0.2064848691225052,
+      "learning_rate": 5e-06,
+      "loss": 0.1866,
+      "step": 11982
+    },
+    {
+      "epoch": 9.019947308995107,
+      "grad_norm": 0.18000181019306183,
+      "learning_rate": 5e-06,
+      "loss": 0.1447,
+      "step": 11983
+    },
+    {
+      "epoch": 9.020700037636432,
+      "grad_norm": 0.2501806914806366,
+      "learning_rate": 5e-06,
+      "loss": 0.234,
+      "step": 11984
+    },
+    {
+      "epoch": 9.021452766277758,
+      "grad_norm": 0.17210984230041504,
+      "learning_rate": 5e-06,
+      "loss": 0.1168,
+      "step": 11985
+    },
+    {
+      "epoch": 9.022205494919081,
+      "grad_norm": 0.2096197009086609,
+      "learning_rate": 5e-06,
+      "loss": 0.2189,
+      "step": 11986
+    },
+    {
+      "epoch": 9.022958223560407,
+      "grad_norm": 0.18832245469093323,
+      "learning_rate": 5e-06,
+      "loss": 0.1872,
+      "step": 11987
+    },
+    {
+      "epoch": 9.023710952201732,
+      "grad_norm": 0.16595140099525452,
+      "learning_rate": 5e-06,
+      "loss": 0.1164,
+      "step": 11988
+    },
+    {
+      "epoch": 9.024463680843056,
+      "grad_norm": 0.14988423883914948,
+      "learning_rate": 5e-06,
+      "loss": 0.0845,
+      "step": 11989
+    },
+    {
+      "epoch": 9.025216409484381,
+      "grad_norm": 0.17763236165046692,
+      "learning_rate": 5e-06,
+      "loss": 0.1537,
+      "step": 11990
+    },
+    {
+      "epoch": 9.025969138125706,
+      "grad_norm": 0.18617308139801025,
+      "learning_rate": 5e-06,
+      "loss": 0.1306,
+      "step": 11991
+    },
+    {
+      "epoch": 9.02672186676703,
+      "grad_norm": 0.18348513543605804,
+      "learning_rate": 5e-06,
+      "loss": 0.122,
+      "step": 11992
+    },
+    {
+      "epoch": 9.027474595408355,
+      "grad_norm": 0.1787821501493454,
+      "learning_rate": 5e-06,
+      "loss": 0.1629,
+      "step": 11993
+    },
+    {
+      "epoch": 9.02822732404968,
+      "grad_norm": 0.16575726866722107,
+      "learning_rate": 5e-06,
+      "loss": 0.2559,
+      "step": 11994
+    },
+    {
+      "epoch": 9.028980052691004,
+      "grad_norm": 0.16215752065181732,
+      "learning_rate": 5e-06,
+      "loss": 0.0981,
+      "step": 11995
+    },
+    {
+      "epoch": 9.02973278133233,
+      "grad_norm": 0.1783803403377533,
+      "learning_rate": 5e-06,
+      "loss": 0.1856,
+      "step": 11996
+    },
+    {
+      "epoch": 9.030485509973655,
+      "grad_norm": 0.16801045835018158,
+      "learning_rate": 5e-06,
+      "loss": 0.1718,
+      "step": 11997
+    },
+    {
+      "epoch": 9.031238238614979,
+      "grad_norm": 0.194411501288414,
+      "learning_rate": 5e-06,
+      "loss": 0.16,
+      "step": 11998
+    },
+    {
+      "epoch": 9.031990967256304,
+      "grad_norm": 0.16710485517978668,
+      "learning_rate": 5e-06,
+      "loss": 0.1346,
+      "step": 11999
+    },
+    {
+      "epoch": 9.03274369589763,
+      "grad_norm": 0.19147342443466187,
+      "learning_rate": 5e-06,
+      "loss": 0.1796,
+      "step": 12000
+    },
+    {
+      "epoch": 9.03274369589763,
+      "eval_loss": 0.19307510554790497,
+      "eval_runtime": 452.7895,
+      "eval_samples_per_second": 21.262,
+      "eval_steps_per_second": 0.665,
+      "step": 12000
+    },
+    {
+      "epoch": 9.033496424538953,
+      "grad_norm": 0.1988474428653717,
+      "learning_rate": 5e-06,
+      "loss": 0.1793,
+      "step": 12001
+    },
+    {
+      "epoch": 9.034249153180278,
+      "grad_norm": 0.22546648979187012,
+      "learning_rate": 5e-06,
+      "loss": 0.2018,
+      "step": 12002
+    },
+    {
+      "epoch": 9.035001881821604,
+      "grad_norm": 0.15144550800323486,
+      "learning_rate": 5e-06,
+      "loss": 0.0854,
+      "step": 12003
+    },
+    {
+      "epoch": 9.035754610462929,
+      "grad_norm": 0.17141476273536682,
+      "learning_rate": 5e-06,
+      "loss": 0.1742,
+      "step": 12004
+    },
+    {
+      "epoch": 9.036507339104253,
+      "grad_norm": 0.17484451830387115,
+      "learning_rate": 5e-06,
+      "loss": 0.1501,
+      "step": 12005
+    },
+    {
+      "epoch": 9.037260067745578,
+      "grad_norm": 0.18813495337963104,
+      "learning_rate": 5e-06,
+      "loss": 0.1655,
+      "step": 12006
+    },
+    {
+      "epoch": 9.038012796386903,
+      "grad_norm": 0.18255552649497986,
+      "learning_rate": 5e-06,
+      "loss": 0.1413,
+      "step": 12007
+    },
+    {
+      "epoch": 9.038765525028227,
+      "grad_norm": 0.20620527863502502,
+      "learning_rate": 5e-06,
+      "loss": 0.1773,
+      "step": 12008
+    },
+    {
+      "epoch": 9.039518253669552,
+      "grad_norm": 0.20360732078552246,
+      "learning_rate": 5e-06,
+      "loss": 0.2301,
+      "step": 12009
+    },
+    {
+      "epoch": 9.040270982310878,
+      "grad_norm": 0.16804543137550354,
+      "learning_rate": 5e-06,
+      "loss": 0.1448,
+      "step": 12010
+    },
+    {
+      "epoch": 9.041023710952201,
+      "grad_norm": 0.15182465314865112,
+      "learning_rate": 5e-06,
+      "loss": 0.1026,
+      "step": 12011
+    },
+    {
+      "epoch": 9.041776439593527,
+      "grad_norm": 0.17732980847358704,
+      "learning_rate": 5e-06,
+      "loss": 0.1227,
+      "step": 12012
+    },
+    {
+      "epoch": 9.042529168234852,
+      "grad_norm": 0.1618821620941162,
+      "learning_rate": 5e-06,
+      "loss": 0.1436,
+      "step": 12013
+    },
+    {
+      "epoch": 9.043281896876175,
+      "grad_norm": 0.13824725151062012,
+      "learning_rate": 5e-06,
+      "loss": 0.0739,
+      "step": 12014
+    },
+    {
+      "epoch": 9.0440346255175,
+      "grad_norm": 0.2058016061782837,
+      "learning_rate": 5e-06,
+      "loss": 0.1441,
+      "step": 12015
+    },
+    {
+      "epoch": 9.044787354158826,
+      "grad_norm": 0.16844139993190765,
+      "learning_rate": 5e-06,
+      "loss": 0.1501,
+      "step": 12016
+    },
+    {
+      "epoch": 9.04554008280015,
+      "grad_norm": 0.19450227916240692,
+      "learning_rate": 5e-06,
+      "loss": 0.0839,
+      "step": 12017
+    },
+    {
+      "epoch": 9.046292811441475,
+      "grad_norm": 0.1588466465473175,
+      "learning_rate": 5e-06,
+      "loss": 0.1239,
+      "step": 12018
+    },
+    {
+      "epoch": 9.0470455400828,
+      "grad_norm": 0.20148245990276337,
+      "learning_rate": 5e-06,
+      "loss": 0.218,
+      "step": 12019
+    },
+    {
+      "epoch": 9.047798268724124,
+      "grad_norm": 0.16490045189857483,
+      "learning_rate": 5e-06,
+      "loss": 0.1066,
+      "step": 12020
+    },
+    {
+      "epoch": 9.04855099736545,
+      "grad_norm": 0.18478776514530182,
+      "learning_rate": 5e-06,
+      "loss": 0.2192,
+      "step": 12021
+    },
+    {
+      "epoch": 9.049303726006775,
+      "grad_norm": 0.1753319650888443,
+      "learning_rate": 5e-06,
+      "loss": 0.1255,
+      "step": 12022
+    },
+    {
+      "epoch": 9.0500564546481,
+      "grad_norm": 0.17415745556354523,
+      "learning_rate": 5e-06,
+      "loss": 0.1396,
+      "step": 12023
+    },
+    {
+      "epoch": 9.050809183289424,
+      "grad_norm": 0.1922602504491806,
+      "learning_rate": 5e-06,
+      "loss": 0.1082,
+      "step": 12024
+    },
+    {
+      "epoch": 9.051561911930749,
+      "grad_norm": 0.17829406261444092,
+      "learning_rate": 5e-06,
+      "loss": 0.1783,
+      "step": 12025
+    },
+    {
+      "epoch": 9.052314640572074,
+      "grad_norm": 0.16621790826320648,
+      "learning_rate": 5e-06,
+      "loss": 0.1468,
+      "step": 12026
+    },
+    {
+      "epoch": 9.053067369213398,
+      "grad_norm": 0.1642482876777649,
+      "learning_rate": 5e-06,
+      "loss": 0.2053,
+      "step": 12027
+    },
+    {
+      "epoch": 9.053820097854723,
+      "grad_norm": 0.18733897805213928,
+      "learning_rate": 5e-06,
+      "loss": 0.1471,
+      "step": 12028
+    },
+    {
+      "epoch": 9.054572826496049,
+      "grad_norm": 0.17364995181560516,
+      "learning_rate": 5e-06,
+      "loss": 0.111,
+      "step": 12029
+    },
+    {
+      "epoch": 9.055325555137372,
+      "grad_norm": 0.18387560546398163,
+      "learning_rate": 5e-06,
+      "loss": 0.1755,
+      "step": 12030
+    },
+    {
+      "epoch": 9.056078283778698,
+      "grad_norm": 0.177209734916687,
+      "learning_rate": 5e-06,
+      "loss": 0.1393,
+      "step": 12031
+    },
+    {
+      "epoch": 9.056831012420023,
+      "grad_norm": 0.18568842113018036,
+      "learning_rate": 5e-06,
+      "loss": 0.1647,
+      "step": 12032
+    },
+    {
+      "epoch": 9.057583741061347,
+      "grad_norm": 0.19300997257232666,
+      "learning_rate": 5e-06,
+      "loss": 0.1602,
+      "step": 12033
+    },
+    {
+      "epoch": 9.058336469702672,
+      "grad_norm": 0.17698754370212555,
+      "learning_rate": 5e-06,
+      "loss": 0.1084,
+      "step": 12034
+    },
+    {
+      "epoch": 9.059089198343997,
+      "grad_norm": 0.17903625965118408,
+      "learning_rate": 5e-06,
+      "loss": 0.1279,
+      "step": 12035
+    },
+    {
+      "epoch": 9.059841926985321,
+      "grad_norm": 0.17870208621025085,
+      "learning_rate": 5e-06,
+      "loss": 0.2078,
+      "step": 12036
+    },
+    {
+      "epoch": 9.060594655626646,
+      "grad_norm": 0.1735912561416626,
+      "learning_rate": 5e-06,
+      "loss": 0.1434,
+      "step": 12037
+    },
+    {
+      "epoch": 9.061347384267972,
+      "grad_norm": 0.1827407032251358,
+      "learning_rate": 5e-06,
+      "loss": 0.176,
+      "step": 12038
+    },
+    {
+      "epoch": 9.062100112909297,
+      "grad_norm": 0.17381033301353455,
+      "learning_rate": 5e-06,
+      "loss": 0.1742,
+      "step": 12039
+    },
+    {
+      "epoch": 9.06285284155062,
+      "grad_norm": 0.19929072260856628,
+      "learning_rate": 5e-06,
+      "loss": 0.1707,
+      "step": 12040
+    },
+    {
+      "epoch": 9.063605570191946,
+      "grad_norm": 0.1731848120689392,
+      "learning_rate": 5e-06,
+      "loss": 0.1925,
+      "step": 12041
+    },
+    {
+      "epoch": 9.064358298833271,
+      "grad_norm": 0.15413598716259003,
+      "learning_rate": 5e-06,
+      "loss": 0.1023,
+      "step": 12042
+    },
+    {
+      "epoch": 9.065111027474595,
+      "grad_norm": 0.20478560030460358,
+      "learning_rate": 5e-06,
+      "loss": 0.1661,
+      "step": 12043
+    },
+    {
+      "epoch": 9.06586375611592,
+      "grad_norm": 0.18150170147418976,
+      "learning_rate": 5e-06,
+      "loss": 0.1606,
+      "step": 12044
+    },
+    {
+      "epoch": 9.066616484757246,
+      "grad_norm": 0.20821811258792877,
+      "learning_rate": 5e-06,
+      "loss": 0.1598,
+      "step": 12045
+    },
+    {
+      "epoch": 9.06736921339857,
+      "grad_norm": 0.15857036411762238,
+      "learning_rate": 5e-06,
+      "loss": 0.152,
+      "step": 12046
+    },
+    {
+      "epoch": 9.068121942039895,
+      "grad_norm": 0.18485942482948303,
+      "learning_rate": 5e-06,
+      "loss": 0.1574,
+      "step": 12047
+    },
+    {
+      "epoch": 9.06887467068122,
+      "grad_norm": 0.17328369617462158,
+      "learning_rate": 5e-06,
+      "loss": 0.1732,
+      "step": 12048
+    },
+    {
+      "epoch": 9.069627399322544,
+      "grad_norm": 0.18031658232212067,
+      "learning_rate": 5e-06,
+      "loss": 0.1515,
+      "step": 12049
+    },
+    {
+      "epoch": 9.070380127963869,
+      "grad_norm": 0.1707950234413147,
+      "learning_rate": 5e-06,
+      "loss": 0.1287,
+      "step": 12050
+    },
+    {
+      "epoch": 9.071132856605194,
+      "grad_norm": 0.1699761599302292,
+      "learning_rate": 5e-06,
+      "loss": 0.1275,
+      "step": 12051
+    },
+    {
+      "epoch": 9.071885585246518,
+      "grad_norm": 0.175688236951828,
+      "learning_rate": 5e-06,
+      "loss": 0.1779,
+      "step": 12052
+    },
+    {
+      "epoch": 9.072638313887843,
+      "grad_norm": 0.14653447270393372,
+      "learning_rate": 5e-06,
+      "loss": 0.1081,
+      "step": 12053
+    },
+    {
+      "epoch": 9.073391042529169,
+      "grad_norm": 0.17708353698253632,
+      "learning_rate": 5e-06,
+      "loss": 0.1746,
+      "step": 12054
+    },
+    {
+      "epoch": 9.074143771170492,
+      "grad_norm": 0.18872617185115814,
+      "learning_rate": 5e-06,
+      "loss": 0.1316,
+      "step": 12055
+    },
+    {
+      "epoch": 9.074896499811818,
+      "grad_norm": 0.20532667636871338,
+      "learning_rate": 5e-06,
+      "loss": 0.1659,
+      "step": 12056
+    },
+    {
+      "epoch": 9.075649228453143,
+      "grad_norm": 0.16380855441093445,
+      "learning_rate": 5e-06,
+      "loss": 0.1347,
+      "step": 12057
+    },
+    {
+      "epoch": 9.076401957094468,
+      "grad_norm": 0.1632387787103653,
+      "learning_rate": 5e-06,
+      "loss": 0.1814,
+      "step": 12058
+    },
+    {
+      "epoch": 9.077154685735792,
+      "grad_norm": 0.19038081169128418,
+      "learning_rate": 5e-06,
+      "loss": 0.1607,
+      "step": 12059
+    },
+    {
+      "epoch": 9.077907414377117,
+      "grad_norm": 0.19483956694602966,
+      "learning_rate": 5e-06,
+      "loss": 0.1763,
+      "step": 12060
+    },
+    {
+      "epoch": 9.078660143018443,
+      "grad_norm": 0.1668921560049057,
+      "learning_rate": 5e-06,
+      "loss": 0.1796,
+      "step": 12061
+    },
+    {
+      "epoch": 9.079412871659766,
+      "grad_norm": 0.17538844048976898,
+      "learning_rate": 5e-06,
+      "loss": 0.1491,
+      "step": 12062
+    },
+    {
+      "epoch": 9.080165600301092,
+      "grad_norm": 0.18697863817214966,
+      "learning_rate": 5e-06,
+      "loss": 0.2025,
+      "step": 12063
+    },
+    {
+      "epoch": 9.080918328942417,
+      "grad_norm": 0.17618438601493835,
+      "learning_rate": 5e-06,
+      "loss": 0.0883,
+      "step": 12064
+    },
+    {
+      "epoch": 9.08167105758374,
+      "grad_norm": 0.1673395186662674,
+      "learning_rate": 5e-06,
+      "loss": 0.1641,
+      "step": 12065
+    },
+    {
+      "epoch": 9.082423786225066,
+      "grad_norm": 0.17029914259910583,
+      "learning_rate": 5e-06,
+      "loss": 0.1602,
+      "step": 12066
+    },
+    {
+      "epoch": 9.083176514866391,
+      "grad_norm": 0.17818878591060638,
+      "learning_rate": 5e-06,
+      "loss": 0.114,
+      "step": 12067
+    },
+    {
+      "epoch": 9.083929243507715,
+      "grad_norm": 0.1916058510541916,
+      "learning_rate": 5e-06,
+      "loss": 0.1406,
+      "step": 12068
+    },
+    {
+      "epoch": 9.08468197214904,
+      "grad_norm": 0.17452698945999146,
+      "learning_rate": 5e-06,
+      "loss": 0.1792,
+      "step": 12069
+    },
+    {
+      "epoch": 9.085434700790366,
+      "grad_norm": 0.16613924503326416,
+      "learning_rate": 5e-06,
+      "loss": 0.1878,
+      "step": 12070
+    },
+    {
+      "epoch": 9.086187429431689,
+      "grad_norm": 0.17577297985553741,
+      "learning_rate": 5e-06,
+      "loss": 0.1856,
+      "step": 12071
+    },
+    {
+      "epoch": 9.086940158073014,
+      "grad_norm": 0.16173891723155975,
+      "learning_rate": 5e-06,
+      "loss": 0.0962,
+      "step": 12072
+    },
+    {
+      "epoch": 9.08769288671434,
+      "grad_norm": 0.15545636415481567,
+      "learning_rate": 5e-06,
+      "loss": 0.1944,
+      "step": 12073
+    },
+    {
+      "epoch": 9.088445615355663,
+      "grad_norm": 0.16321530938148499,
+      "learning_rate": 5e-06,
+      "loss": 0.153,
+      "step": 12074
+    },
+    {
+      "epoch": 9.089198343996989,
+      "grad_norm": 0.18522909283638,
+      "learning_rate": 5e-06,
+      "loss": 0.1888,
+      "step": 12075
+    },
+    {
+      "epoch": 9.089951072638314,
+      "grad_norm": 0.16677603125572205,
+      "learning_rate": 5e-06,
+      "loss": 0.1651,
+      "step": 12076
+    },
+    {
+      "epoch": 9.09070380127964,
+      "grad_norm": 0.17138929665088654,
+      "learning_rate": 5e-06,
+      "loss": 0.1285,
+      "step": 12077
+    },
+    {
+      "epoch": 9.091456529920963,
+      "grad_norm": 0.17052499949932098,
+      "learning_rate": 5e-06,
+      "loss": 0.1721,
+      "step": 12078
+    },
+    {
+      "epoch": 9.092209258562288,
+      "grad_norm": 0.16914887726306915,
+      "learning_rate": 5e-06,
+      "loss": 0.1718,
+      "step": 12079
+    },
+    {
+      "epoch": 9.092961987203614,
+      "grad_norm": 0.16532574594020844,
+      "learning_rate": 5e-06,
+      "loss": 0.0773,
+      "step": 12080
+    },
+    {
+      "epoch": 9.093714715844937,
+      "grad_norm": 0.18121908605098724,
+      "learning_rate": 5e-06,
+      "loss": 0.163,
+      "step": 12081
+    },
+    {
+      "epoch": 9.094467444486263,
+      "grad_norm": 0.17850525677204132,
+      "learning_rate": 5e-06,
+      "loss": 0.1377,
+      "step": 12082
+    },
+    {
+      "epoch": 9.095220173127588,
+      "grad_norm": 0.17374378442764282,
+      "learning_rate": 5e-06,
+      "loss": 0.122,
+      "step": 12083
+    },
+    {
+      "epoch": 9.095972901768912,
+      "grad_norm": 0.1950574666261673,
+      "learning_rate": 5e-06,
+      "loss": 0.1819,
+      "step": 12084
+    },
+    {
+      "epoch": 9.096725630410237,
+      "grad_norm": 0.1836615651845932,
+      "learning_rate": 5e-06,
+      "loss": 0.199,
+      "step": 12085
+    },
+    {
+      "epoch": 9.097478359051562,
+      "grad_norm": 0.17404590547084808,
+      "learning_rate": 5e-06,
+      "loss": 0.1365,
+      "step": 12086
+    },
+    {
+      "epoch": 9.098231087692886,
+      "grad_norm": 0.20815683901309967,
+      "learning_rate": 5e-06,
+      "loss": 0.1694,
+      "step": 12087
+    },
+    {
+      "epoch": 9.098983816334211,
+      "grad_norm": 0.1521454155445099,
+      "learning_rate": 5e-06,
+      "loss": 0.1521,
+      "step": 12088
+    },
+    {
+      "epoch": 9.099736544975537,
+      "grad_norm": 0.16391384601593018,
+      "learning_rate": 5e-06,
+      "loss": 0.0855,
+      "step": 12089
+    },
+    {
+      "epoch": 9.10048927361686,
+      "grad_norm": 0.20221920311450958,
+      "learning_rate": 5e-06,
+      "loss": 0.184,
+      "step": 12090
+    },
+    {
+      "epoch": 9.101242002258186,
+      "grad_norm": 0.19065435230731964,
+      "learning_rate": 5e-06,
+      "loss": 0.1624,
+      "step": 12091
+    },
+    {
+      "epoch": 9.101994730899511,
+      "grad_norm": 0.16278809309005737,
+      "learning_rate": 5e-06,
+      "loss": 0.1289,
+      "step": 12092
+    },
+    {
+      "epoch": 9.102747459540836,
+      "grad_norm": 0.1757497489452362,
+      "learning_rate": 5e-06,
+      "loss": 0.1959,
+      "step": 12093
+    },
+    {
+      "epoch": 9.10350018818216,
+      "grad_norm": 0.19447718560695648,
+      "learning_rate": 5e-06,
+      "loss": 0.1826,
+      "step": 12094
+    },
+    {
+      "epoch": 9.104252916823485,
+      "grad_norm": 0.1871030330657959,
+      "learning_rate": 5e-06,
+      "loss": 0.1464,
+      "step": 12095
+    },
+    {
+      "epoch": 9.10500564546481,
+      "grad_norm": 0.17390578985214233,
+      "learning_rate": 5e-06,
+      "loss": 0.1241,
+      "step": 12096
+    },
+    {
+      "epoch": 9.105758374106134,
+      "grad_norm": 0.1793605089187622,
+      "learning_rate": 5e-06,
+      "loss": 0.1436,
+      "step": 12097
+    },
+    {
+      "epoch": 9.10651110274746,
+      "grad_norm": 0.19909431040287018,
+      "learning_rate": 5e-06,
+      "loss": 0.1747,
+      "step": 12098
+    },
+    {
+      "epoch": 9.107263831388785,
+      "grad_norm": 0.17408527433872223,
+      "learning_rate": 5e-06,
+      "loss": 0.2236,
+      "step": 12099
+    },
+    {
+      "epoch": 9.108016560030109,
+      "grad_norm": 0.17412647604942322,
+      "learning_rate": 5e-06,
+      "loss": 0.1744,
+      "step": 12100
+    },
+    {
+      "epoch": 9.108769288671434,
+      "grad_norm": 0.16274356842041016,
+      "learning_rate": 5e-06,
+      "loss": 0.1739,
+      "step": 12101
+    },
+    {
+      "epoch": 9.10952201731276,
+      "grad_norm": 0.1984359323978424,
+      "learning_rate": 5e-06,
+      "loss": 0.1916,
+      "step": 12102
+    },
+    {
+      "epoch": 9.110274745954083,
+      "grad_norm": 0.1895211935043335,
+      "learning_rate": 5e-06,
+      "loss": 0.1484,
+      "step": 12103
+    },
+    {
+      "epoch": 9.111027474595408,
+      "grad_norm": 0.1778009682893753,
+      "learning_rate": 5e-06,
+      "loss": 0.1108,
+      "step": 12104
+    },
+    {
+      "epoch": 9.111780203236734,
+      "grad_norm": 0.19090090692043304,
+      "learning_rate": 5e-06,
+      "loss": 0.1563,
+      "step": 12105
+    },
+    {
+      "epoch": 9.112532931878057,
+      "grad_norm": 0.2210940271615982,
+      "learning_rate": 5e-06,
+      "loss": 0.1553,
+      "step": 12106
+    },
+    {
+      "epoch": 9.113285660519383,
+      "grad_norm": 0.18952931463718414,
+      "learning_rate": 5e-06,
+      "loss": 0.1366,
+      "step": 12107
+    },
+    {
+      "epoch": 9.114038389160708,
+      "grad_norm": 0.17843444645404816,
+      "learning_rate": 5e-06,
+      "loss": 0.1041,
+      "step": 12108
+    },
+    {
+      "epoch": 9.114791117802032,
+      "grad_norm": 0.137284517288208,
+      "learning_rate": 5e-06,
+      "loss": 0.1395,
+      "step": 12109
+    },
+    {
+      "epoch": 9.115543846443357,
+      "grad_norm": 0.18002720177173615,
+      "learning_rate": 5e-06,
+      "loss": 0.154,
+      "step": 12110
+    },
+    {
+      "epoch": 9.116296575084682,
+      "grad_norm": 0.19388499855995178,
+      "learning_rate": 5e-06,
+      "loss": 0.1374,
+      "step": 12111
+    },
+    {
+      "epoch": 9.117049303726008,
+      "grad_norm": 0.19794303178787231,
+      "learning_rate": 5e-06,
+      "loss": 0.1197,
+      "step": 12112
+    },
+    {
+      "epoch": 9.117802032367331,
+      "grad_norm": 0.18766602873802185,
+      "learning_rate": 5e-06,
+      "loss": 0.2018,
+      "step": 12113
+    },
+    {
+      "epoch": 9.118554761008657,
+      "grad_norm": 0.170462965965271,
+      "learning_rate": 5e-06,
+      "loss": 0.1006,
+      "step": 12114
+    },
+    {
+      "epoch": 9.119307489649982,
+      "grad_norm": 0.1688939481973648,
+      "learning_rate": 5e-06,
+      "loss": 0.1929,
+      "step": 12115
+    },
+    {
+      "epoch": 9.120060218291306,
+      "grad_norm": 0.16863776743412018,
+      "learning_rate": 5e-06,
+      "loss": 0.1367,
+      "step": 12116
+    },
+    {
+      "epoch": 9.120812946932631,
+      "grad_norm": 0.18111096322536469,
+      "learning_rate": 5e-06,
+      "loss": 0.0988,
+      "step": 12117
+    },
+    {
+      "epoch": 9.121565675573956,
+      "grad_norm": 0.18180328607559204,
+      "learning_rate": 5e-06,
+      "loss": 0.2428,
+      "step": 12118
+    },
+    {
+      "epoch": 9.12231840421528,
+      "grad_norm": 0.17224127054214478,
+      "learning_rate": 5e-06,
+      "loss": 0.1021,
+      "step": 12119
+    },
+    {
+      "epoch": 9.123071132856605,
+      "grad_norm": 0.18409858644008636,
+      "learning_rate": 5e-06,
+      "loss": 0.166,
+      "step": 12120
+    },
+    {
+      "epoch": 9.12382386149793,
+      "grad_norm": 0.1737634241580963,
+      "learning_rate": 5e-06,
+      "loss": 0.1359,
+      "step": 12121
+    },
+    {
+      "epoch": 9.124576590139254,
+      "grad_norm": 0.15860335528850555,
+      "learning_rate": 5e-06,
+      "loss": 0.151,
+      "step": 12122
+    },
+    {
+      "epoch": 9.12532931878058,
+      "grad_norm": 0.1735897958278656,
+      "learning_rate": 5e-06,
+      "loss": 0.1772,
+      "step": 12123
+    },
+    {
+      "epoch": 9.126082047421905,
+      "grad_norm": 0.19043254852294922,
+      "learning_rate": 5e-06,
+      "loss": 0.1991,
+      "step": 12124
+    },
+    {
+      "epoch": 9.126834776063228,
+      "grad_norm": 0.18686321377754211,
+      "learning_rate": 5e-06,
+      "loss": 0.0971,
+      "step": 12125
+    },
+    {
+      "epoch": 9.127587504704554,
+      "grad_norm": 0.1500331163406372,
+      "learning_rate": 5e-06,
+      "loss": 0.1515,
+      "step": 12126
+    },
+    {
+      "epoch": 9.12834023334588,
+      "grad_norm": 0.19108040630817413,
+      "learning_rate": 5e-06,
+      "loss": 0.2068,
+      "step": 12127
+    },
+    {
+      "epoch": 9.129092961987203,
+      "grad_norm": 0.1728270947933197,
+      "learning_rate": 5e-06,
+      "loss": 0.1759,
+      "step": 12128
+    },
+    {
+      "epoch": 9.129845690628528,
+      "grad_norm": 0.16385722160339355,
+      "learning_rate": 5e-06,
+      "loss": 0.1665,
+      "step": 12129
+    },
+    {
+      "epoch": 9.130598419269853,
+      "grad_norm": 0.183256596326828,
+      "learning_rate": 5e-06,
+      "loss": 0.2181,
+      "step": 12130
+    },
+    {
+      "epoch": 9.131351147911179,
+      "grad_norm": 0.16733382642269135,
+      "learning_rate": 5e-06,
+      "loss": 0.1328,
+      "step": 12131
+    },
+    {
+      "epoch": 9.132103876552502,
+      "grad_norm": 0.18604975938796997,
+      "learning_rate": 5e-06,
+      "loss": 0.1339,
+      "step": 12132
+    },
+    {
+      "epoch": 9.132856605193828,
+      "grad_norm": 0.17076362669467926,
+      "learning_rate": 5e-06,
+      "loss": 0.1621,
+      "step": 12133
+    },
+    {
+      "epoch": 9.133609333835153,
+      "grad_norm": 0.18888922035694122,
+      "learning_rate": 5e-06,
+      "loss": 0.1524,
+      "step": 12134
+    },
+    {
+      "epoch": 9.134362062476477,
+      "grad_norm": 0.19153538346290588,
+      "learning_rate": 5e-06,
+      "loss": 0.118,
+      "step": 12135
+    },
+    {
+      "epoch": 9.135114791117802,
+      "grad_norm": 0.15402813255786896,
+      "learning_rate": 5e-06,
+      "loss": 0.1658,
+      "step": 12136
+    },
+    {
+      "epoch": 9.135867519759127,
+      "grad_norm": 0.18001653254032135,
+      "learning_rate": 5e-06,
+      "loss": 0.1421,
+      "step": 12137
+    },
+    {
+      "epoch": 9.136620248400451,
+      "grad_norm": 0.18926165997982025,
+      "learning_rate": 5e-06,
+      "loss": 0.1946,
+      "step": 12138
+    },
+    {
+      "epoch": 9.137372977041776,
+      "grad_norm": 0.17520387470722198,
+      "learning_rate": 5e-06,
+      "loss": 0.1138,
+      "step": 12139
+    },
+    {
+      "epoch": 9.138125705683102,
+      "grad_norm": 0.16199107468128204,
+      "learning_rate": 5e-06,
+      "loss": 0.1126,
+      "step": 12140
+    },
+    {
+      "epoch": 9.138878434324425,
+      "grad_norm": 0.15779761970043182,
+      "learning_rate": 5e-06,
+      "loss": 0.0688,
+      "step": 12141
+    },
+    {
+      "epoch": 9.13963116296575,
+      "grad_norm": 0.16519878804683685,
+      "learning_rate": 5e-06,
+      "loss": 0.1223,
+      "step": 12142
+    },
+    {
+      "epoch": 9.140383891607076,
+      "grad_norm": 0.1853569895029068,
+      "learning_rate": 5e-06,
+      "loss": 0.146,
+      "step": 12143
+    },
+    {
+      "epoch": 9.1411366202484,
+      "grad_norm": 0.17410539090633392,
+      "learning_rate": 5e-06,
+      "loss": 0.1466,
+      "step": 12144
+    },
+    {
+      "epoch": 9.141889348889725,
+      "grad_norm": 0.16672082245349884,
+      "learning_rate": 5e-06,
+      "loss": 0.1371,
+      "step": 12145
+    },
+    {
+      "epoch": 9.14264207753105,
+      "grad_norm": 0.1848076730966568,
+      "learning_rate": 5e-06,
+      "loss": 0.2382,
+      "step": 12146
+    },
+    {
+      "epoch": 9.143394806172374,
+      "grad_norm": 0.17518550157546997,
+      "learning_rate": 5e-06,
+      "loss": 0.1587,
+      "step": 12147
+    },
+    {
+      "epoch": 9.1441475348137,
+      "grad_norm": 0.17237800359725952,
+      "learning_rate": 5e-06,
+      "loss": 0.2567,
+      "step": 12148
+    },
+    {
+      "epoch": 9.144900263455025,
+      "grad_norm": 0.16969022154808044,
+      "learning_rate": 5e-06,
+      "loss": 0.185,
+      "step": 12149
+    },
+    {
+      "epoch": 9.14565299209635,
+      "grad_norm": 0.19565916061401367,
+      "learning_rate": 5e-06,
+      "loss": 0.1404,
+      "step": 12150
+    },
+    {
+      "epoch": 9.146405720737674,
+      "grad_norm": 0.17231880128383636,
+      "learning_rate": 5e-06,
+      "loss": 0.0822,
+      "step": 12151
+    },
+    {
+      "epoch": 9.147158449378999,
+      "grad_norm": 0.19148479402065277,
+      "learning_rate": 5e-06,
+      "loss": 0.1498,
+      "step": 12152
+    },
+    {
+      "epoch": 9.147911178020324,
+      "grad_norm": 0.18761277198791504,
+      "learning_rate": 5e-06,
+      "loss": 0.1658,
+      "step": 12153
+    },
+    {
+      "epoch": 9.148663906661648,
+      "grad_norm": 0.18690793216228485,
+      "learning_rate": 5e-06,
+      "loss": 0.1141,
+      "step": 12154
+    },
+    {
+      "epoch": 9.149416635302973,
+      "grad_norm": 0.1719900220632553,
+      "learning_rate": 5e-06,
+      "loss": 0.1272,
+      "step": 12155
+    },
+    {
+      "epoch": 9.150169363944299,
+      "grad_norm": 0.1836354285478592,
+      "learning_rate": 5e-06,
+      "loss": 0.1598,
+      "step": 12156
+    },
+    {
+      "epoch": 9.150922092585622,
+      "grad_norm": 0.20235329866409302,
+      "learning_rate": 5e-06,
+      "loss": 0.1612,
+      "step": 12157
+    },
+    {
+      "epoch": 9.151674821226948,
+      "grad_norm": 0.16599774360656738,
+      "learning_rate": 5e-06,
+      "loss": 0.094,
+      "step": 12158
+    },
+    {
+      "epoch": 9.152427549868273,
+      "grad_norm": 0.19561392068862915,
+      "learning_rate": 5e-06,
+      "loss": 0.2011,
+      "step": 12159
+    },
+    {
+      "epoch": 9.153180278509597,
+      "grad_norm": 0.16940073668956757,
+      "learning_rate": 5e-06,
+      "loss": 0.1133,
+      "step": 12160
+    },
+    {
+      "epoch": 9.153933007150922,
+      "grad_norm": 0.20006638765335083,
+      "learning_rate": 5e-06,
+      "loss": 0.151,
+      "step": 12161
+    },
+    {
+      "epoch": 9.154685735792247,
+      "grad_norm": 0.18394073843955994,
+      "learning_rate": 5e-06,
+      "loss": 0.1193,
+      "step": 12162
+    },
+    {
+      "epoch": 9.155438464433571,
+      "grad_norm": 0.16561679542064667,
+      "learning_rate": 5e-06,
+      "loss": 0.1534,
+      "step": 12163
+    },
+    {
+      "epoch": 9.156191193074896,
+      "grad_norm": 0.17251929640769958,
+      "learning_rate": 5e-06,
+      "loss": 0.1266,
+      "step": 12164
+    },
+    {
+      "epoch": 9.156943921716222,
+      "grad_norm": 0.17270588874816895,
+      "learning_rate": 5e-06,
+      "loss": 0.1638,
+      "step": 12165
+    },
+    {
+      "epoch": 9.157696650357547,
+      "grad_norm": 0.18162013590335846,
+      "learning_rate": 5e-06,
+      "loss": 0.113,
+      "step": 12166
+    },
+    {
+      "epoch": 9.15844937899887,
+      "grad_norm": 0.1741899698972702,
+      "learning_rate": 5e-06,
+      "loss": 0.1765,
+      "step": 12167
+    },
+    {
+      "epoch": 9.159202107640196,
+      "grad_norm": 0.1583966314792633,
+      "learning_rate": 5e-06,
+      "loss": 0.1637,
+      "step": 12168
+    },
+    {
+      "epoch": 9.159954836281521,
+      "grad_norm": 0.1845666617155075,
+      "learning_rate": 5e-06,
+      "loss": 0.1183,
+      "step": 12169
+    },
+    {
+      "epoch": 9.160707564922845,
+      "grad_norm": 0.17952556908130646,
+      "learning_rate": 5e-06,
+      "loss": 0.1391,
+      "step": 12170
+    },
+    {
+      "epoch": 9.16146029356417,
+      "grad_norm": 0.19190183281898499,
+      "learning_rate": 5e-06,
+      "loss": 0.1576,
+      "step": 12171
+    },
+    {
+      "epoch": 9.162213022205496,
+      "grad_norm": 0.18243436515331268,
+      "learning_rate": 5e-06,
+      "loss": 0.2129,
+      "step": 12172
+    },
+    {
+      "epoch": 9.16296575084682,
+      "grad_norm": 0.17646770179271698,
+      "learning_rate": 5e-06,
+      "loss": 0.1426,
+      "step": 12173
+    },
+    {
+      "epoch": 9.163718479488145,
+      "grad_norm": 0.17348280549049377,
+      "learning_rate": 5e-06,
+      "loss": 0.1414,
+      "step": 12174
+    },
+    {
+      "epoch": 9.16447120812947,
+      "grad_norm": 0.18699373304843903,
+      "learning_rate": 5e-06,
+      "loss": 0.1523,
+      "step": 12175
+    },
+    {
+      "epoch": 9.165223936770794,
+      "grad_norm": 0.18530891835689545,
+      "learning_rate": 5e-06,
+      "loss": 0.231,
+      "step": 12176
+    },
+    {
+      "epoch": 9.165976665412119,
+      "grad_norm": 0.21589422225952148,
+      "learning_rate": 5e-06,
+      "loss": 0.2143,
+      "step": 12177
+    },
+    {
+      "epoch": 9.166729394053444,
+      "grad_norm": 0.16693034768104553,
+      "learning_rate": 5e-06,
+      "loss": 0.0644,
+      "step": 12178
+    },
+    {
+      "epoch": 9.167482122694768,
+      "grad_norm": 0.1960817128419876,
+      "learning_rate": 5e-06,
+      "loss": 0.1601,
+      "step": 12179
+    },
+    {
+      "epoch": 9.168234851336093,
+      "grad_norm": 0.1884528398513794,
+      "learning_rate": 5e-06,
+      "loss": 0.1421,
+      "step": 12180
+    },
+    {
+      "epoch": 9.168987579977419,
+      "grad_norm": 0.1836431920528412,
+      "learning_rate": 5e-06,
+      "loss": 0.1507,
+      "step": 12181
+    },
+    {
+      "epoch": 9.169740308618742,
+      "grad_norm": 0.17566810548305511,
+      "learning_rate": 5e-06,
+      "loss": 0.1474,
+      "step": 12182
+    },
+    {
+      "epoch": 9.170493037260067,
+      "grad_norm": 0.1899881213903427,
+      "learning_rate": 5e-06,
+      "loss": 0.1398,
+      "step": 12183
+    },
+    {
+      "epoch": 9.171245765901393,
+      "grad_norm": 0.20858432352542877,
+      "learning_rate": 5e-06,
+      "loss": 0.2598,
+      "step": 12184
+    },
+    {
+      "epoch": 9.171998494542718,
+      "grad_norm": 0.2005435675382614,
+      "learning_rate": 5e-06,
+      "loss": 0.171,
+      "step": 12185
+    },
+    {
+      "epoch": 9.172751223184042,
+      "grad_norm": 0.18290257453918457,
+      "learning_rate": 5e-06,
+      "loss": 0.1743,
+      "step": 12186
+    },
+    {
+      "epoch": 9.173503951825367,
+      "grad_norm": 0.17971278727054596,
+      "learning_rate": 5e-06,
+      "loss": 0.1443,
+      "step": 12187
+    },
+    {
+      "epoch": 9.174256680466693,
+      "grad_norm": 0.17876321077346802,
+      "learning_rate": 5e-06,
+      "loss": 0.1559,
+      "step": 12188
+    },
+    {
+      "epoch": 9.175009409108016,
+      "grad_norm": 0.17292268574237823,
+      "learning_rate": 5e-06,
+      "loss": 0.1197,
+      "step": 12189
+    },
+    {
+      "epoch": 9.175762137749341,
+      "grad_norm": 0.19772806763648987,
+      "learning_rate": 5e-06,
+      "loss": 0.1622,
+      "step": 12190
+    },
+    {
+      "epoch": 9.176514866390667,
+      "grad_norm": 0.2043067216873169,
+      "learning_rate": 5e-06,
+      "loss": 0.1947,
+      "step": 12191
+    },
+    {
+      "epoch": 9.17726759503199,
+      "grad_norm": 0.18532057106494904,
+      "learning_rate": 5e-06,
+      "loss": 0.1152,
+      "step": 12192
+    },
+    {
+      "epoch": 9.178020323673316,
+      "grad_norm": 0.17972327768802643,
+      "learning_rate": 5e-06,
+      "loss": 0.1838,
+      "step": 12193
+    },
+    {
+      "epoch": 9.178773052314641,
+      "grad_norm": 0.18920035660266876,
+      "learning_rate": 5e-06,
+      "loss": 0.1321,
+      "step": 12194
+    },
+    {
+      "epoch": 9.179525780955965,
+      "grad_norm": 0.17022979259490967,
+      "learning_rate": 5e-06,
+      "loss": 0.1395,
+      "step": 12195
+    },
+    {
+      "epoch": 9.18027850959729,
+      "grad_norm": 0.18125756084918976,
+      "learning_rate": 5e-06,
+      "loss": 0.1218,
+      "step": 12196
+    },
+    {
+      "epoch": 9.181031238238615,
+      "grad_norm": 0.17871402204036713,
+      "learning_rate": 5e-06,
+      "loss": 0.1808,
+      "step": 12197
+    },
+    {
+      "epoch": 9.181783966879939,
+      "grad_norm": 0.1832483410835266,
+      "learning_rate": 5e-06,
+      "loss": 0.2013,
+      "step": 12198
+    },
+    {
+      "epoch": 9.182536695521264,
+      "grad_norm": 0.16339628398418427,
+      "learning_rate": 5e-06,
+      "loss": 0.1123,
+      "step": 12199
+    },
+    {
+      "epoch": 9.18328942416259,
+      "grad_norm": 0.1609547734260559,
+      "learning_rate": 5e-06,
+      "loss": 0.0961,
+      "step": 12200
+    },
+    {
+      "epoch": 9.18328942416259,
+      "eval_loss": 0.19309543073177338,
+      "eval_runtime": 452.7551,
+      "eval_samples_per_second": 21.263,
+      "eval_steps_per_second": 0.665,
+      "step": 12200
+    },
+    {
+      "epoch": 9.184042152803913,
+      "grad_norm": 0.18373984098434448,
+      "learning_rate": 5e-06,
+      "loss": 0.1207,
+      "step": 12201
+    },
+    {
+      "epoch": 9.184794881445239,
+      "grad_norm": 0.1821109503507614,
+      "learning_rate": 5e-06,
+      "loss": 0.2675,
+      "step": 12202
+    },
+    {
+      "epoch": 9.185547610086564,
+      "grad_norm": 0.17688369750976562,
+      "learning_rate": 5e-06,
+      "loss": 0.1123,
+      "step": 12203
+    },
+    {
+      "epoch": 9.18630033872789,
+      "grad_norm": 0.17624647915363312,
+      "learning_rate": 5e-06,
+      "loss": 0.1522,
+      "step": 12204
+    },
+    {
+      "epoch": 9.187053067369213,
+      "grad_norm": 0.18776406347751617,
+      "learning_rate": 5e-06,
+      "loss": 0.1257,
+      "step": 12205
+    },
+    {
+      "epoch": 9.187805796010538,
+      "grad_norm": 0.17656566202640533,
+      "learning_rate": 5e-06,
+      "loss": 0.1766,
+      "step": 12206
+    },
+    {
+      "epoch": 9.188558524651864,
+      "grad_norm": 0.17458917200565338,
+      "learning_rate": 5e-06,
+      "loss": 0.1696,
+      "step": 12207
+    },
+    {
+      "epoch": 9.189311253293187,
+      "grad_norm": 0.1603238880634308,
+      "learning_rate": 5e-06,
+      "loss": 0.1195,
+      "step": 12208
+    },
+    {
+      "epoch": 9.190063981934513,
+      "grad_norm": 0.1862368881702423,
+      "learning_rate": 5e-06,
+      "loss": 0.1719,
+      "step": 12209
+    },
+    {
+      "epoch": 9.190816710575838,
+      "grad_norm": 0.15873181819915771,
+      "learning_rate": 5e-06,
+      "loss": 0.1101,
+      "step": 12210
+    },
+    {
+      "epoch": 9.191569439217162,
+      "grad_norm": 0.17826661467552185,
+      "learning_rate": 5e-06,
+      "loss": 0.138,
+      "step": 12211
+    },
+    {
+      "epoch": 9.192322167858487,
+      "grad_norm": 0.17449380457401276,
+      "learning_rate": 5e-06,
+      "loss": 0.124,
+      "step": 12212
+    },
+    {
+      "epoch": 9.193074896499812,
+      "grad_norm": 0.1489475816488266,
+      "learning_rate": 5e-06,
+      "loss": 0.1379,
+      "step": 12213
+    },
+    {
+      "epoch": 9.193827625141136,
+      "grad_norm": 0.18734894692897797,
+      "learning_rate": 5e-06,
+      "loss": 0.145,
+      "step": 12214
+    },
+    {
+      "epoch": 9.194580353782461,
+      "grad_norm": 0.1624721735715866,
+      "learning_rate": 5e-06,
+      "loss": 0.1279,
+      "step": 12215
+    },
+    {
+      "epoch": 9.195333082423787,
+      "grad_norm": 0.16508036851882935,
+      "learning_rate": 5e-06,
+      "loss": 0.1382,
+      "step": 12216
+    },
+    {
+      "epoch": 9.19608581106511,
+      "grad_norm": 0.15386727452278137,
+      "learning_rate": 5e-06,
+      "loss": 0.1096,
+      "step": 12217
+    },
+    {
+      "epoch": 9.196838539706436,
+      "grad_norm": 0.16598500311374664,
+      "learning_rate": 5e-06,
+      "loss": 0.1659,
+      "step": 12218
+    },
+    {
+      "epoch": 9.197591268347761,
+      "grad_norm": 0.15869170427322388,
+      "learning_rate": 5e-06,
+      "loss": 0.1221,
+      "step": 12219
+    },
+    {
+      "epoch": 9.198343996989085,
+      "grad_norm": 0.16631732881069183,
+      "learning_rate": 5e-06,
+      "loss": 0.105,
+      "step": 12220
+    },
+    {
+      "epoch": 9.19909672563041,
+      "grad_norm": 0.15560567378997803,
+      "learning_rate": 5e-06,
+      "loss": 0.0643,
+      "step": 12221
+    },
+    {
+      "epoch": 9.199849454271735,
+      "grad_norm": 0.16518917679786682,
+      "learning_rate": 5e-06,
+      "loss": 0.1074,
+      "step": 12222
+    },
+    {
+      "epoch": 9.20060218291306,
+      "grad_norm": 0.16875779628753662,
+      "learning_rate": 5e-06,
+      "loss": 0.1813,
+      "step": 12223
+    },
+    {
+      "epoch": 9.201354911554384,
+      "grad_norm": 0.15938973426818848,
+      "learning_rate": 5e-06,
+      "loss": 0.117,
+      "step": 12224
+    },
+    {
+      "epoch": 9.20210764019571,
+      "grad_norm": 0.21160349249839783,
+      "learning_rate": 5e-06,
+      "loss": 0.1809,
+      "step": 12225
+    },
+    {
+      "epoch": 9.202860368837035,
+      "grad_norm": 0.17452974617481232,
+      "learning_rate": 5e-06,
+      "loss": 0.1199,
+      "step": 12226
+    },
+    {
+      "epoch": 9.203613097478359,
+      "grad_norm": 0.1823432892560959,
+      "learning_rate": 5e-06,
+      "loss": 0.2059,
+      "step": 12227
+    },
+    {
+      "epoch": 9.204365826119684,
+      "grad_norm": 0.16334307193756104,
+      "learning_rate": 5e-06,
+      "loss": 0.0968,
+      "step": 12228
+    },
+    {
+      "epoch": 9.20511855476101,
+      "grad_norm": 0.1583770215511322,
+      "learning_rate": 5e-06,
+      "loss": 0.1207,
+      "step": 12229
+    },
+    {
+      "epoch": 9.205871283402333,
+      "grad_norm": 0.18351957201957703,
+      "learning_rate": 5e-06,
+      "loss": 0.1344,
+      "step": 12230
+    },
+    {
+      "epoch": 9.206624012043658,
+      "grad_norm": 0.17029443383216858,
+      "learning_rate": 5e-06,
+      "loss": 0.1752,
+      "step": 12231
+    },
+    {
+      "epoch": 9.207376740684984,
+      "grad_norm": 0.15474137663841248,
+      "learning_rate": 5e-06,
+      "loss": 0.1278,
+      "step": 12232
+    },
+    {
+      "epoch": 9.208129469326307,
+      "grad_norm": 0.16447603702545166,
+      "learning_rate": 5e-06,
+      "loss": 0.1967,
+      "step": 12233
+    },
+    {
+      "epoch": 9.208882197967633,
+      "grad_norm": 0.17110107839107513,
+      "learning_rate": 5e-06,
+      "loss": 0.1348,
+      "step": 12234
+    },
+    {
+      "epoch": 9.209634926608958,
+      "grad_norm": 0.1900234818458557,
+      "learning_rate": 5e-06,
+      "loss": 0.2672,
+      "step": 12235
+    },
+    {
+      "epoch": 9.210387655250281,
+      "grad_norm": 0.18940794467926025,
+      "learning_rate": 5e-06,
+      "loss": 0.1839,
+      "step": 12236
+    },
+    {
+      "epoch": 9.211140383891607,
+      "grad_norm": 0.21533313393592834,
+      "learning_rate": 5e-06,
+      "loss": 0.1818,
+      "step": 12237
+    },
+    {
+      "epoch": 9.211893112532932,
+      "grad_norm": 0.19424092769622803,
+      "learning_rate": 5e-06,
+      "loss": 0.2397,
+      "step": 12238
+    },
+    {
+      "epoch": 9.212645841174258,
+      "grad_norm": 0.16546447575092316,
+      "learning_rate": 5e-06,
+      "loss": 0.1298,
+      "step": 12239
+    },
+    {
+      "epoch": 9.213398569815581,
+      "grad_norm": 0.17105208337306976,
+      "learning_rate": 5e-06,
+      "loss": 0.0899,
+      "step": 12240
+    },
+    {
+      "epoch": 9.214151298456907,
+      "grad_norm": 0.2014627456665039,
+      "learning_rate": 5e-06,
+      "loss": 0.1865,
+      "step": 12241
+    },
+    {
+      "epoch": 9.214904027098232,
+      "grad_norm": 0.17722272872924805,
+      "learning_rate": 5e-06,
+      "loss": 0.1886,
+      "step": 12242
+    },
+    {
+      "epoch": 9.215656755739555,
+      "grad_norm": 0.1865025907754898,
+      "learning_rate": 5e-06,
+      "loss": 0.192,
+      "step": 12243
+    },
+    {
+      "epoch": 9.21640948438088,
+      "grad_norm": 0.16530701518058777,
+      "learning_rate": 5e-06,
+      "loss": 0.1658,
+      "step": 12244
+    },
+    {
+      "epoch": 9.217162213022206,
+      "grad_norm": 0.1934792399406433,
+      "learning_rate": 5e-06,
+      "loss": 0.1724,
+      "step": 12245
+    },
+    {
+      "epoch": 9.21791494166353,
+      "grad_norm": 0.1878943145275116,
+      "learning_rate": 5e-06,
+      "loss": 0.1996,
+      "step": 12246
+    },
+    {
+      "epoch": 9.218667670304855,
+      "grad_norm": 0.19426359236240387,
+      "learning_rate": 5e-06,
+      "loss": 0.3243,
+      "step": 12247
+    },
+    {
+      "epoch": 9.21942039894618,
+      "grad_norm": 0.18304991722106934,
+      "learning_rate": 5e-06,
+      "loss": 0.0915,
+      "step": 12248
+    },
+    {
+      "epoch": 9.220173127587504,
+      "grad_norm": 0.15666250884532928,
+      "learning_rate": 5e-06,
+      "loss": 0.1979,
+      "step": 12249
+    },
+    {
+      "epoch": 9.22092585622883,
+      "grad_norm": 0.1564083695411682,
+      "learning_rate": 5e-06,
+      "loss": 0.133,
+      "step": 12250
+    },
+    {
+      "epoch": 9.221678584870155,
+      "grad_norm": 0.16318075358867645,
+      "learning_rate": 5e-06,
+      "loss": 0.1662,
+      "step": 12251
+    },
+    {
+      "epoch": 9.222431313511478,
+      "grad_norm": 0.1750221997499466,
+      "learning_rate": 5e-06,
+      "loss": 0.1497,
+      "step": 12252
+    },
+    {
+      "epoch": 9.223184042152804,
+      "grad_norm": 0.15790952742099762,
+      "learning_rate": 5e-06,
+      "loss": 0.1283,
+      "step": 12253
+    },
+    {
+      "epoch": 9.223936770794129,
+      "grad_norm": 0.19544269144535065,
+      "learning_rate": 5e-06,
+      "loss": 0.216,
+      "step": 12254
+    },
+    {
+      "epoch": 9.224689499435453,
+      "grad_norm": 0.18065418303012848,
+      "learning_rate": 5e-06,
+      "loss": 0.0904,
+      "step": 12255
+    },
+    {
+      "epoch": 9.225442228076778,
+      "grad_norm": 0.17539039254188538,
+      "learning_rate": 5e-06,
+      "loss": 0.1492,
+      "step": 12256
+    },
+    {
+      "epoch": 9.226194956718103,
+      "grad_norm": 0.17329205572605133,
+      "learning_rate": 5e-06,
+      "loss": 0.1618,
+      "step": 12257
+    },
+    {
+      "epoch": 9.226947685359429,
+      "grad_norm": 0.18917538225650787,
+      "learning_rate": 5e-06,
+      "loss": 0.1265,
+      "step": 12258
+    },
+    {
+      "epoch": 9.227700414000752,
+      "grad_norm": 0.15498659014701843,
+      "learning_rate": 5e-06,
+      "loss": 0.103,
+      "step": 12259
+    },
+    {
+      "epoch": 9.228453142642078,
+      "grad_norm": 0.20118220150470734,
+      "learning_rate": 5e-06,
+      "loss": 0.2606,
+      "step": 12260
+    },
+    {
+      "epoch": 9.229205871283403,
+      "grad_norm": 0.1609196811914444,
+      "learning_rate": 5e-06,
+      "loss": 0.1199,
+      "step": 12261
+    },
+    {
+      "epoch": 9.229958599924727,
+      "grad_norm": 0.1727282553911209,
+      "learning_rate": 5e-06,
+      "loss": 0.1127,
+      "step": 12262
+    },
+    {
+      "epoch": 9.230711328566052,
+      "grad_norm": 0.19606392085552216,
+      "learning_rate": 5e-06,
+      "loss": 0.1507,
+      "step": 12263
+    },
+    {
+      "epoch": 9.231464057207377,
+      "grad_norm": 0.20071150362491608,
+      "learning_rate": 5e-06,
+      "loss": 0.1243,
+      "step": 12264
+    },
+    {
+      "epoch": 9.232216785848701,
+      "grad_norm": 0.194985032081604,
+      "learning_rate": 5e-06,
+      "loss": 0.1774,
+      "step": 12265
+    },
+    {
+      "epoch": 9.232969514490026,
+      "grad_norm": 0.16140896081924438,
+      "learning_rate": 5e-06,
+      "loss": 0.1791,
+      "step": 12266
+    },
+    {
+      "epoch": 9.233722243131352,
+      "grad_norm": 0.1979992836713791,
+      "learning_rate": 5e-06,
+      "loss": 0.21,
+      "step": 12267
+    },
+    {
+      "epoch": 9.234474971772675,
+      "grad_norm": 0.15845422446727753,
+      "learning_rate": 5e-06,
+      "loss": 0.1126,
+      "step": 12268
+    },
+    {
+      "epoch": 9.235227700414,
+      "grad_norm": 0.20766966044902802,
+      "learning_rate": 5e-06,
+      "loss": 0.1398,
+      "step": 12269
+    },
+    {
+      "epoch": 9.235980429055326,
+      "grad_norm": 0.15628869831562042,
+      "learning_rate": 5e-06,
+      "loss": 0.0774,
+      "step": 12270
+    },
+    {
+      "epoch": 9.23673315769665,
+      "grad_norm": 0.17685779929161072,
+      "learning_rate": 5e-06,
+      "loss": 0.1357,
+      "step": 12271
+    },
+    {
+      "epoch": 9.237485886337975,
+      "grad_norm": 0.1676325798034668,
+      "learning_rate": 5e-06,
+      "loss": 0.128,
+      "step": 12272
+    },
+    {
+      "epoch": 9.2382386149793,
+      "grad_norm": 0.18048600852489471,
+      "learning_rate": 5e-06,
+      "loss": 0.1557,
+      "step": 12273
+    },
+    {
+      "epoch": 9.238991343620624,
+      "grad_norm": 0.17989549040794373,
+      "learning_rate": 5e-06,
+      "loss": 0.1627,
+      "step": 12274
+    },
+    {
+      "epoch": 9.23974407226195,
+      "grad_norm": 0.16081131994724274,
+      "learning_rate": 5e-06,
+      "loss": 0.1439,
+      "step": 12275
+    },
+    {
+      "epoch": 9.240496800903275,
+      "grad_norm": 0.17601363360881805,
+      "learning_rate": 5e-06,
+      "loss": 0.1403,
+      "step": 12276
+    },
+    {
+      "epoch": 9.2412495295446,
+      "grad_norm": 0.19725826382637024,
+      "learning_rate": 5e-06,
+      "loss": 0.1351,
+      "step": 12277
+    },
+    {
+      "epoch": 9.242002258185924,
+      "grad_norm": 0.2168322503566742,
+      "learning_rate": 5e-06,
+      "loss": 0.1552,
+      "step": 12278
+    },
+    {
+      "epoch": 9.242754986827249,
+      "grad_norm": 0.15330666303634644,
+      "learning_rate": 5e-06,
+      "loss": 0.0976,
+      "step": 12279
+    },
+    {
+      "epoch": 9.243507715468574,
+      "grad_norm": 0.2372134029865265,
+      "learning_rate": 5e-06,
+      "loss": 0.2308,
+      "step": 12280
+    },
+    {
+      "epoch": 9.244260444109898,
+      "grad_norm": 0.2064095437526703,
+      "learning_rate": 5e-06,
+      "loss": 0.1936,
+      "step": 12281
+    },
+    {
+      "epoch": 9.245013172751223,
+      "grad_norm": 0.18899600207805634,
+      "learning_rate": 5e-06,
+      "loss": 0.1719,
+      "step": 12282
+    },
+    {
+      "epoch": 9.245765901392549,
+      "grad_norm": 0.1637261062860489,
+      "learning_rate": 5e-06,
+      "loss": 0.1022,
+      "step": 12283
+    },
+    {
+      "epoch": 9.246518630033872,
+      "grad_norm": 0.19466784596443176,
+      "learning_rate": 5e-06,
+      "loss": 0.2827,
+      "step": 12284
+    },
+    {
+      "epoch": 9.247271358675198,
+      "grad_norm": 0.13572128117084503,
+      "learning_rate": 5e-06,
+      "loss": 0.07,
+      "step": 12285
+    },
+    {
+      "epoch": 9.248024087316523,
+      "grad_norm": 0.17627239227294922,
+      "learning_rate": 5e-06,
+      "loss": 0.0964,
+      "step": 12286
+    },
+    {
+      "epoch": 9.248776815957847,
+      "grad_norm": 0.1612519770860672,
+      "learning_rate": 5e-06,
+      "loss": 0.1415,
+      "step": 12287
+    },
+    {
+      "epoch": 9.249529544599172,
+      "grad_norm": 0.2816060781478882,
+      "learning_rate": 5e-06,
+      "loss": 0.1656,
+      "step": 12288
+    },
+    {
+      "epoch": 9.250282273240497,
+      "grad_norm": 0.1772184520959854,
+      "learning_rate": 5e-06,
+      "loss": 0.1399,
+      "step": 12289
+    },
+    {
+      "epoch": 9.25103500188182,
+      "grad_norm": 0.16242025792598724,
+      "learning_rate": 5e-06,
+      "loss": 0.152,
+      "step": 12290
+    },
+    {
+      "epoch": 9.251787730523146,
+      "grad_norm": 0.19073472917079926,
+      "learning_rate": 5e-06,
+      "loss": 0.1393,
+      "step": 12291
+    },
+    {
+      "epoch": 9.252540459164472,
+      "grad_norm": 0.22973519563674927,
+      "learning_rate": 5e-06,
+      "loss": 0.1847,
+      "step": 12292
+    },
+    {
+      "epoch": 9.253293187805795,
+      "grad_norm": 0.17171047627925873,
+      "learning_rate": 5e-06,
+      "loss": 0.0873,
+      "step": 12293
+    },
+    {
+      "epoch": 9.25404591644712,
+      "grad_norm": 0.16917107999324799,
+      "learning_rate": 5e-06,
+      "loss": 0.1818,
+      "step": 12294
+    },
+    {
+      "epoch": 9.254798645088446,
+      "grad_norm": 0.19794589281082153,
+      "learning_rate": 5e-06,
+      "loss": 0.1776,
+      "step": 12295
+    },
+    {
+      "epoch": 9.255551373729771,
+      "grad_norm": 0.20039452612400055,
+      "learning_rate": 5e-06,
+      "loss": 0.1594,
+      "step": 12296
+    },
+    {
+      "epoch": 9.256304102371095,
+      "grad_norm": 0.17312850058078766,
+      "learning_rate": 5e-06,
+      "loss": 0.1711,
+      "step": 12297
+    },
+    {
+      "epoch": 9.25705683101242,
+      "grad_norm": 0.1865558922290802,
+      "learning_rate": 5e-06,
+      "loss": 0.1447,
+      "step": 12298
+    },
+    {
+      "epoch": 9.257809559653746,
+      "grad_norm": 0.16269046068191528,
+      "learning_rate": 5e-06,
+      "loss": 0.126,
+      "step": 12299
+    },
+    {
+      "epoch": 9.258562288295069,
+      "grad_norm": 0.19035567343235016,
+      "learning_rate": 5e-06,
+      "loss": 0.1447,
+      "step": 12300
+    },
+    {
+      "epoch": 9.259315016936394,
+      "grad_norm": 0.17222285270690918,
+      "learning_rate": 5e-06,
+      "loss": 0.1219,
+      "step": 12301
+    },
+    {
+      "epoch": 9.26006774557772,
+      "grad_norm": 0.17666374146938324,
+      "learning_rate": 5e-06,
+      "loss": 0.1483,
+      "step": 12302
+    },
+    {
+      "epoch": 9.260820474219043,
+      "grad_norm": 0.1933794617652893,
+      "learning_rate": 5e-06,
+      "loss": 0.1497,
+      "step": 12303
+    },
+    {
+      "epoch": 9.261573202860369,
+      "grad_norm": 0.17424657940864563,
+      "learning_rate": 5e-06,
+      "loss": 0.1209,
+      "step": 12304
+    },
+    {
+      "epoch": 9.262325931501694,
+      "grad_norm": 0.18126751482486725,
+      "learning_rate": 5e-06,
+      "loss": 0.1661,
+      "step": 12305
+    },
+    {
+      "epoch": 9.263078660143018,
+      "grad_norm": 0.20961707830429077,
+      "learning_rate": 5e-06,
+      "loss": 0.1595,
+      "step": 12306
+    },
+    {
+      "epoch": 9.263831388784343,
+      "grad_norm": 0.15868844091892242,
+      "learning_rate": 5e-06,
+      "loss": 0.1475,
+      "step": 12307
+    },
+    {
+      "epoch": 9.264584117425668,
+      "grad_norm": 0.16545122861862183,
+      "learning_rate": 5e-06,
+      "loss": 0.1051,
+      "step": 12308
+    },
+    {
+      "epoch": 9.265336846066992,
+      "grad_norm": 0.1793530285358429,
+      "learning_rate": 5e-06,
+      "loss": 0.1349,
+      "step": 12309
+    },
+    {
+      "epoch": 9.266089574708317,
+      "grad_norm": 0.18125271797180176,
+      "learning_rate": 5e-06,
+      "loss": 0.1556,
+      "step": 12310
+    },
+    {
+      "epoch": 9.266842303349643,
+      "grad_norm": 0.17835870385169983,
+      "learning_rate": 5e-06,
+      "loss": 0.1902,
+      "step": 12311
+    },
+    {
+      "epoch": 9.267595031990968,
+      "grad_norm": 0.16554927825927734,
+      "learning_rate": 5e-06,
+      "loss": 0.0845,
+      "step": 12312
+    },
+    {
+      "epoch": 9.268347760632292,
+      "grad_norm": 0.18556509912014008,
+      "learning_rate": 5e-06,
+      "loss": 0.2317,
+      "step": 12313
+    },
+    {
+      "epoch": 9.269100489273617,
+      "grad_norm": 0.2032296359539032,
+      "learning_rate": 5e-06,
+      "loss": 0.1385,
+      "step": 12314
+    },
+    {
+      "epoch": 9.269853217914942,
+      "grad_norm": 0.20240536332130432,
+      "learning_rate": 5e-06,
+      "loss": 0.1517,
+      "step": 12315
+    },
+    {
+      "epoch": 9.270605946556266,
+      "grad_norm": 0.1811200976371765,
+      "learning_rate": 5e-06,
+      "loss": 0.1816,
+      "step": 12316
+    },
+    {
+      "epoch": 9.271358675197591,
+      "grad_norm": 0.17529600858688354,
+      "learning_rate": 5e-06,
+      "loss": 0.1742,
+      "step": 12317
+    },
+    {
+      "epoch": 9.272111403838917,
+      "grad_norm": 0.1727580428123474,
+      "learning_rate": 5e-06,
+      "loss": 0.1566,
+      "step": 12318
+    },
+    {
+      "epoch": 9.27286413248024,
+      "grad_norm": 0.1744164526462555,
+      "learning_rate": 5e-06,
+      "loss": 0.1302,
+      "step": 12319
+    },
+    {
+      "epoch": 9.273616861121566,
+      "grad_norm": 0.14160189032554626,
+      "learning_rate": 5e-06,
+      "loss": 0.106,
+      "step": 12320
+    },
+    {
+      "epoch": 9.274369589762891,
+      "grad_norm": 0.18156449496746063,
+      "learning_rate": 5e-06,
+      "loss": 0.1246,
+      "step": 12321
+    },
+    {
+      "epoch": 9.275122318404215,
+      "grad_norm": 0.16957050561904907,
+      "learning_rate": 5e-06,
+      "loss": 0.0985,
+      "step": 12322
+    },
+    {
+      "epoch": 9.27587504704554,
+      "grad_norm": 0.20607160031795502,
+      "learning_rate": 5e-06,
+      "loss": 0.2968,
+      "step": 12323
+    },
+    {
+      "epoch": 9.276627775686865,
+      "grad_norm": 0.16204535961151123,
+      "learning_rate": 5e-06,
+      "loss": 0.1799,
+      "step": 12324
+    },
+    {
+      "epoch": 9.277380504328189,
+      "grad_norm": 0.17203371226787567,
+      "learning_rate": 5e-06,
+      "loss": 0.1652,
+      "step": 12325
+    },
+    {
+      "epoch": 9.278133232969514,
+      "grad_norm": 0.16442640125751495,
+      "learning_rate": 5e-06,
+      "loss": 0.1432,
+      "step": 12326
+    },
+    {
+      "epoch": 9.27888596161084,
+      "grad_norm": 0.18222428858280182,
+      "learning_rate": 5e-06,
+      "loss": 0.1829,
+      "step": 12327
+    },
+    {
+      "epoch": 9.279638690252163,
+      "grad_norm": 0.19879581034183502,
+      "learning_rate": 5e-06,
+      "loss": 0.1537,
+      "step": 12328
+    },
+    {
+      "epoch": 9.280391418893489,
+      "grad_norm": 0.1984388679265976,
+      "learning_rate": 5e-06,
+      "loss": 0.1283,
+      "step": 12329
+    },
+    {
+      "epoch": 9.281144147534814,
+      "grad_norm": 0.19671528041362762,
+      "learning_rate": 5e-06,
+      "loss": 0.1645,
+      "step": 12330
+    },
+    {
+      "epoch": 9.28189687617614,
+      "grad_norm": 0.1699669063091278,
+      "learning_rate": 5e-06,
+      "loss": 0.0861,
+      "step": 12331
+    },
+    {
+      "epoch": 9.282649604817463,
+      "grad_norm": 0.1747102588415146,
+      "learning_rate": 5e-06,
+      "loss": 0.1757,
+      "step": 12332
+    },
+    {
+      "epoch": 9.283402333458788,
+      "grad_norm": 0.17560963332653046,
+      "learning_rate": 5e-06,
+      "loss": 0.1851,
+      "step": 12333
+    },
+    {
+      "epoch": 9.284155062100114,
+      "grad_norm": 0.17312385141849518,
+      "learning_rate": 5e-06,
+      "loss": 0.0881,
+      "step": 12334
+    },
+    {
+      "epoch": 9.284907790741437,
+      "grad_norm": 0.17359857261180878,
+      "learning_rate": 5e-06,
+      "loss": 0.1531,
+      "step": 12335
+    },
+    {
+      "epoch": 9.285660519382763,
+      "grad_norm": 0.1628965586423874,
+      "learning_rate": 5e-06,
+      "loss": 0.0886,
+      "step": 12336
+    },
+    {
+      "epoch": 9.286413248024088,
+      "grad_norm": 0.1920742392539978,
+      "learning_rate": 5e-06,
+      "loss": 0.1924,
+      "step": 12337
+    },
+    {
+      "epoch": 9.287165976665412,
+      "grad_norm": 0.18349751830101013,
+      "learning_rate": 5e-06,
+      "loss": 0.1652,
+      "step": 12338
+    },
+    {
+      "epoch": 9.287918705306737,
+      "grad_norm": 0.2063753306865692,
+      "learning_rate": 5e-06,
+      "loss": 0.1366,
+      "step": 12339
+    },
+    {
+      "epoch": 9.288671433948062,
+      "grad_norm": 0.17109020054340363,
+      "learning_rate": 5e-06,
+      "loss": 0.1369,
+      "step": 12340
+    },
+    {
+      "epoch": 9.289424162589386,
+      "grad_norm": 0.19071412086486816,
+      "learning_rate": 5e-06,
+      "loss": 0.1671,
+      "step": 12341
+    },
+    {
+      "epoch": 9.290176891230711,
+      "grad_norm": 0.18061143159866333,
+      "learning_rate": 5e-06,
+      "loss": 0.1393,
+      "step": 12342
+    },
+    {
+      "epoch": 9.290929619872037,
+      "grad_norm": 0.18462273478507996,
+      "learning_rate": 5e-06,
+      "loss": 0.1036,
+      "step": 12343
+    },
+    {
+      "epoch": 9.29168234851336,
+      "grad_norm": 0.14508739113807678,
+      "learning_rate": 5e-06,
+      "loss": 0.1504,
+      "step": 12344
+    },
+    {
+      "epoch": 9.292435077154686,
+      "grad_norm": 0.1777936965227127,
+      "learning_rate": 5e-06,
+      "loss": 0.1186,
+      "step": 12345
+    },
+    {
+      "epoch": 9.293187805796011,
+      "grad_norm": 0.18063850700855255,
+      "learning_rate": 5e-06,
+      "loss": 0.1346,
+      "step": 12346
+    },
+    {
+      "epoch": 9.293940534437336,
+      "grad_norm": 0.20724476873874664,
+      "learning_rate": 5e-06,
+      "loss": 0.1535,
+      "step": 12347
+    },
+    {
+      "epoch": 9.29469326307866,
+      "grad_norm": 0.1773209273815155,
+      "learning_rate": 5e-06,
+      "loss": 0.1426,
+      "step": 12348
+    },
+    {
+      "epoch": 9.295445991719985,
+      "grad_norm": 0.17217890918254852,
+      "learning_rate": 5e-06,
+      "loss": 0.2149,
+      "step": 12349
+    },
+    {
+      "epoch": 9.29619872036131,
+      "grad_norm": 0.17098887264728546,
+      "learning_rate": 5e-06,
+      "loss": 0.1665,
+      "step": 12350
+    },
+    {
+      "epoch": 9.296951449002634,
+      "grad_norm": 0.19439935684204102,
+      "learning_rate": 5e-06,
+      "loss": 0.1862,
+      "step": 12351
+    },
+    {
+      "epoch": 9.29770417764396,
+      "grad_norm": 0.20866985619068146,
+      "learning_rate": 5e-06,
+      "loss": 0.1524,
+      "step": 12352
+    },
+    {
+      "epoch": 9.298456906285285,
+      "grad_norm": 0.19233675301074982,
+      "learning_rate": 5e-06,
+      "loss": 0.1813,
+      "step": 12353
+    },
+    {
+      "epoch": 9.299209634926608,
+      "grad_norm": 0.18526336550712585,
+      "learning_rate": 5e-06,
+      "loss": 0.1925,
+      "step": 12354
+    },
+    {
+      "epoch": 9.299962363567934,
+      "grad_norm": 0.16537921130657196,
+      "learning_rate": 5e-06,
+      "loss": 0.1147,
+      "step": 12355
+    },
+    {
+      "epoch": 9.30071509220926,
+      "grad_norm": 0.20120200514793396,
+      "learning_rate": 5e-06,
+      "loss": 0.1409,
+      "step": 12356
+    },
+    {
+      "epoch": 9.301467820850583,
+      "grad_norm": 0.16884943842887878,
+      "learning_rate": 5e-06,
+      "loss": 0.1896,
+      "step": 12357
+    },
+    {
+      "epoch": 9.302220549491908,
+      "grad_norm": 0.15924686193466187,
+      "learning_rate": 5e-06,
+      "loss": 0.0873,
+      "step": 12358
+    },
+    {
+      "epoch": 9.302973278133233,
+      "grad_norm": 0.16669225692749023,
+      "learning_rate": 5e-06,
+      "loss": 0.1807,
+      "step": 12359
+    },
+    {
+      "epoch": 9.303726006774557,
+      "grad_norm": 0.18712013959884644,
+      "learning_rate": 5e-06,
+      "loss": 0.115,
+      "step": 12360
+    },
+    {
+      "epoch": 9.304478735415882,
+      "grad_norm": 0.19456863403320312,
+      "learning_rate": 5e-06,
+      "loss": 0.1861,
+      "step": 12361
+    },
+    {
+      "epoch": 9.305231464057208,
+      "grad_norm": 0.16287466883659363,
+      "learning_rate": 5e-06,
+      "loss": 0.1634,
+      "step": 12362
+    },
+    {
+      "epoch": 9.305984192698531,
+      "grad_norm": 0.18350280821323395,
+      "learning_rate": 5e-06,
+      "loss": 0.2013,
+      "step": 12363
+    },
+    {
+      "epoch": 9.306736921339857,
+      "grad_norm": 0.18913929164409637,
+      "learning_rate": 5e-06,
+      "loss": 0.1131,
+      "step": 12364
+    },
+    {
+      "epoch": 9.307489649981182,
+      "grad_norm": 0.1838071048259735,
+      "learning_rate": 5e-06,
+      "loss": 0.1144,
+      "step": 12365
+    },
+    {
+      "epoch": 9.308242378622506,
+      "grad_norm": 0.1922542303800583,
+      "learning_rate": 5e-06,
+      "loss": 0.149,
+      "step": 12366
+    },
+    {
+      "epoch": 9.308995107263831,
+      "grad_norm": 0.18328674137592316,
+      "learning_rate": 5e-06,
+      "loss": 0.1566,
+      "step": 12367
+    },
+    {
+      "epoch": 9.309747835905156,
+      "grad_norm": 0.1994955837726593,
+      "learning_rate": 5e-06,
+      "loss": 0.2009,
+      "step": 12368
+    },
+    {
+      "epoch": 9.310500564546482,
+      "grad_norm": 0.15146517753601074,
+      "learning_rate": 5e-06,
+      "loss": 0.179,
+      "step": 12369
+    },
+    {
+      "epoch": 9.311253293187805,
+      "grad_norm": 0.1690344214439392,
+      "learning_rate": 5e-06,
+      "loss": 0.1817,
+      "step": 12370
+    },
+    {
+      "epoch": 9.31200602182913,
+      "grad_norm": 0.17183808982372284,
+      "learning_rate": 5e-06,
+      "loss": 0.152,
+      "step": 12371
+    },
+    {
+      "epoch": 9.312758750470456,
+      "grad_norm": 0.15340791642665863,
+      "learning_rate": 5e-06,
+      "loss": 0.0796,
+      "step": 12372
+    },
+    {
+      "epoch": 9.31351147911178,
+      "grad_norm": 0.1818641871213913,
+      "learning_rate": 5e-06,
+      "loss": 0.1635,
+      "step": 12373
+    },
+    {
+      "epoch": 9.314264207753105,
+      "grad_norm": 0.20287224650382996,
+      "learning_rate": 5e-06,
+      "loss": 0.1954,
+      "step": 12374
+    },
+    {
+      "epoch": 9.31501693639443,
+      "grad_norm": 0.1907271146774292,
+      "learning_rate": 5e-06,
+      "loss": 0.1051,
+      "step": 12375
+    },
+    {
+      "epoch": 9.315769665035754,
+      "grad_norm": 0.17315669357776642,
+      "learning_rate": 5e-06,
+      "loss": 0.1797,
+      "step": 12376
+    },
+    {
+      "epoch": 9.31652239367708,
+      "grad_norm": 0.16171294450759888,
+      "learning_rate": 5e-06,
+      "loss": 0.1641,
+      "step": 12377
+    },
+    {
+      "epoch": 9.317275122318405,
+      "grad_norm": 0.1783173382282257,
+      "learning_rate": 5e-06,
+      "loss": 0.1615,
+      "step": 12378
+    },
+    {
+      "epoch": 9.318027850959728,
+      "grad_norm": 0.18778525292873383,
+      "learning_rate": 5e-06,
+      "loss": 0.1673,
+      "step": 12379
+    },
+    {
+      "epoch": 9.318780579601054,
+      "grad_norm": 0.16983413696289062,
+      "learning_rate": 5e-06,
+      "loss": 0.1969,
+      "step": 12380
+    },
+    {
+      "epoch": 9.319533308242379,
+      "grad_norm": 0.205088809132576,
+      "learning_rate": 5e-06,
+      "loss": 0.2037,
+      "step": 12381
+    },
+    {
+      "epoch": 9.320286036883703,
+      "grad_norm": 0.17913778126239777,
+      "learning_rate": 5e-06,
+      "loss": 0.1431,
+      "step": 12382
+    },
+    {
+      "epoch": 9.321038765525028,
+      "grad_norm": 0.22450973093509674,
+      "learning_rate": 5e-06,
+      "loss": 0.2146,
+      "step": 12383
+    },
+    {
+      "epoch": 9.321791494166353,
+      "grad_norm": 0.1841527819633484,
+      "learning_rate": 5e-06,
+      "loss": 0.1184,
+      "step": 12384
+    },
+    {
+      "epoch": 9.322544222807679,
+      "grad_norm": 0.16446062922477722,
+      "learning_rate": 5e-06,
+      "loss": 0.1401,
+      "step": 12385
+    },
+    {
+      "epoch": 9.323296951449002,
+      "grad_norm": 0.19487178325653076,
+      "learning_rate": 5e-06,
+      "loss": 0.1306,
+      "step": 12386
+    },
+    {
+      "epoch": 9.324049680090328,
+      "grad_norm": 0.1974627524614334,
+      "learning_rate": 5e-06,
+      "loss": 0.1748,
+      "step": 12387
+    },
+    {
+      "epoch": 9.324802408731653,
+      "grad_norm": 0.1740226149559021,
+      "learning_rate": 5e-06,
+      "loss": 0.1366,
+      "step": 12388
+    },
+    {
+      "epoch": 9.325555137372977,
+      "grad_norm": 0.1841689497232437,
+      "learning_rate": 5e-06,
+      "loss": 0.1361,
+      "step": 12389
+    },
+    {
+      "epoch": 9.326307866014302,
+      "grad_norm": 0.1798611730337143,
+      "learning_rate": 5e-06,
+      "loss": 0.1063,
+      "step": 12390
+    },
+    {
+      "epoch": 9.327060594655627,
+      "grad_norm": 0.17424480617046356,
+      "learning_rate": 5e-06,
+      "loss": 0.1565,
+      "step": 12391
+    },
+    {
+      "epoch": 9.327813323296951,
+      "grad_norm": 0.16999340057373047,
+      "learning_rate": 5e-06,
+      "loss": 0.1538,
+      "step": 12392
+    },
+    {
+      "epoch": 9.328566051938276,
+      "grad_norm": 0.15942683815956116,
+      "learning_rate": 5e-06,
+      "loss": 0.1315,
+      "step": 12393
+    },
+    {
+      "epoch": 9.329318780579602,
+      "grad_norm": 0.1827930510044098,
+      "learning_rate": 5e-06,
+      "loss": 0.0908,
+      "step": 12394
+    },
+    {
+      "epoch": 9.330071509220925,
+      "grad_norm": 0.18602386116981506,
+      "learning_rate": 5e-06,
+      "loss": 0.1319,
+      "step": 12395
+    },
+    {
+      "epoch": 9.33082423786225,
+      "grad_norm": 0.16540715098381042,
+      "learning_rate": 5e-06,
+      "loss": 0.1138,
+      "step": 12396
+    },
+    {
+      "epoch": 9.331576966503576,
+      "grad_norm": 0.1853342205286026,
+      "learning_rate": 5e-06,
+      "loss": 0.1545,
+      "step": 12397
+    },
+    {
+      "epoch": 9.3323296951449,
+      "grad_norm": 0.1699950098991394,
+      "learning_rate": 5e-06,
+      "loss": 0.1569,
+      "step": 12398
+    },
+    {
+      "epoch": 9.333082423786225,
+      "grad_norm": 0.17882707715034485,
+      "learning_rate": 5e-06,
+      "loss": 0.1378,
+      "step": 12399
+    },
+    {
+      "epoch": 9.33383515242755,
+      "grad_norm": 0.18407827615737915,
+      "learning_rate": 5e-06,
+      "loss": 0.2467,
+      "step": 12400
+    },
+    {
+      "epoch": 9.33383515242755,
+      "eval_loss": 0.19272536039352417,
+      "eval_runtime": 453.0137,
+      "eval_samples_per_second": 21.251,
+      "eval_steps_per_second": 0.664,
+      "step": 12400
+    },
+    {
+      "epoch": 9.334587881068874,
+      "grad_norm": 0.17275668680667877,
+      "learning_rate": 5e-06,
+      "loss": 0.1608,
+      "step": 12401
+    },
+    {
+      "epoch": 9.3353406097102,
+      "grad_norm": 0.16451796889305115,
+      "learning_rate": 5e-06,
+      "loss": 0.1026,
+      "step": 12402
+    },
+    {
+      "epoch": 9.336093338351525,
+      "grad_norm": 0.13866785168647766,
+      "learning_rate": 5e-06,
+      "loss": 0.0862,
+      "step": 12403
+    },
+    {
+      "epoch": 9.33684606699285,
+      "grad_norm": 0.1720796674489975,
+      "learning_rate": 5e-06,
+      "loss": 0.1305,
+      "step": 12404
+    },
+    {
+      "epoch": 9.337598795634173,
+      "grad_norm": 0.1912587285041809,
+      "learning_rate": 5e-06,
+      "loss": 0.1663,
+      "step": 12405
+    },
+    {
+      "epoch": 9.338351524275499,
+      "grad_norm": 0.17615258693695068,
+      "learning_rate": 5e-06,
+      "loss": 0.0782,
+      "step": 12406
+    },
+    {
+      "epoch": 9.339104252916824,
+      "grad_norm": 0.1989985853433609,
+      "learning_rate": 5e-06,
+      "loss": 0.2059,
+      "step": 12407
+    },
+    {
+      "epoch": 9.339856981558148,
+      "grad_norm": 0.1751166433095932,
+      "learning_rate": 5e-06,
+      "loss": 0.1357,
+      "step": 12408
+    },
+    {
+      "epoch": 9.340609710199473,
+      "grad_norm": 0.17803756892681122,
+      "learning_rate": 5e-06,
+      "loss": 0.1118,
+      "step": 12409
+    },
+    {
+      "epoch": 9.341362438840799,
+      "grad_norm": 0.20125515758991241,
+      "learning_rate": 5e-06,
+      "loss": 0.1524,
+      "step": 12410
+    },
+    {
+      "epoch": 9.342115167482122,
+      "grad_norm": 0.17577631771564484,
+      "learning_rate": 5e-06,
+      "loss": 0.1526,
+      "step": 12411
+    },
+    {
+      "epoch": 9.342867896123447,
+      "grad_norm": 0.18308188021183014,
+      "learning_rate": 5e-06,
+      "loss": 0.1284,
+      "step": 12412
+    },
+    {
+      "epoch": 9.343620624764773,
+      "grad_norm": 0.16216348111629486,
+      "learning_rate": 5e-06,
+      "loss": 0.1361,
+      "step": 12413
+    },
+    {
+      "epoch": 9.344373353406096,
+      "grad_norm": 0.17648792266845703,
+      "learning_rate": 5e-06,
+      "loss": 0.1383,
+      "step": 12414
+    },
+    {
+      "epoch": 9.345126082047422,
+      "grad_norm": 0.1770739108324051,
+      "learning_rate": 5e-06,
+      "loss": 0.1304,
+      "step": 12415
+    },
+    {
+      "epoch": 9.345878810688747,
+      "grad_norm": 0.20246385037899017,
+      "learning_rate": 5e-06,
+      "loss": 0.1773,
+      "step": 12416
+    },
+    {
+      "epoch": 9.34663153933007,
+      "grad_norm": 0.1649981290102005,
+      "learning_rate": 5e-06,
+      "loss": 0.1595,
+      "step": 12417
+    },
+    {
+      "epoch": 9.347384267971396,
+      "grad_norm": 0.18727564811706543,
+      "learning_rate": 5e-06,
+      "loss": 0.1527,
+      "step": 12418
+    },
+    {
+      "epoch": 9.348136996612721,
+      "grad_norm": 0.16877035796642303,
+      "learning_rate": 5e-06,
+      "loss": 0.1238,
+      "step": 12419
+    },
+    {
+      "epoch": 9.348889725254047,
+      "grad_norm": 0.16647550463676453,
+      "learning_rate": 5e-06,
+      "loss": 0.148,
+      "step": 12420
+    },
+    {
+      "epoch": 9.34964245389537,
+      "grad_norm": 0.18942157924175262,
+      "learning_rate": 5e-06,
+      "loss": 0.1479,
+      "step": 12421
+    },
+    {
+      "epoch": 9.350395182536696,
+      "grad_norm": 0.16959527134895325,
+      "learning_rate": 5e-06,
+      "loss": 0.056,
+      "step": 12422
+    },
+    {
+      "epoch": 9.351147911178021,
+      "grad_norm": 0.19092075526714325,
+      "learning_rate": 5e-06,
+      "loss": 0.1187,
+      "step": 12423
+    },
+    {
+      "epoch": 9.351900639819345,
+      "grad_norm": 0.1660452038049698,
+      "learning_rate": 5e-06,
+      "loss": 0.1946,
+      "step": 12424
+    },
+    {
+      "epoch": 9.35265336846067,
+      "grad_norm": 0.20573556423187256,
+      "learning_rate": 5e-06,
+      "loss": 0.1456,
+      "step": 12425
+    },
+    {
+      "epoch": 9.353406097101995,
+      "grad_norm": 0.17798249423503876,
+      "learning_rate": 5e-06,
+      "loss": 0.0927,
+      "step": 12426
+    },
+    {
+      "epoch": 9.354158825743319,
+      "grad_norm": 0.16149353981018066,
+      "learning_rate": 5e-06,
+      "loss": 0.0949,
+      "step": 12427
+    },
+    {
+      "epoch": 9.354911554384644,
+      "grad_norm": 0.18357323110103607,
+      "learning_rate": 5e-06,
+      "loss": 0.1095,
+      "step": 12428
+    },
+    {
+      "epoch": 9.35566428302597,
+      "grad_norm": 0.1969509869813919,
+      "learning_rate": 5e-06,
+      "loss": 0.1725,
+      "step": 12429
+    },
+    {
+      "epoch": 9.356417011667293,
+      "grad_norm": 0.16283974051475525,
+      "learning_rate": 5e-06,
+      "loss": 0.1766,
+      "step": 12430
+    },
+    {
+      "epoch": 9.357169740308619,
+      "grad_norm": 0.18659637868404388,
+      "learning_rate": 5e-06,
+      "loss": 0.1475,
+      "step": 12431
+    },
+    {
+      "epoch": 9.357922468949944,
+      "grad_norm": 0.16655559837818146,
+      "learning_rate": 5e-06,
+      "loss": 0.0825,
+      "step": 12432
+    },
+    {
+      "epoch": 9.358675197591268,
+      "grad_norm": 0.1706191450357437,
+      "learning_rate": 5e-06,
+      "loss": 0.0849,
+      "step": 12433
+    },
+    {
+      "epoch": 9.359427926232593,
+      "grad_norm": 0.17576958239078522,
+      "learning_rate": 5e-06,
+      "loss": 0.1177,
+      "step": 12434
+    },
+    {
+      "epoch": 9.360180654873918,
+      "grad_norm": 0.19654624164104462,
+      "learning_rate": 5e-06,
+      "loss": 0.2173,
+      "step": 12435
+    },
+    {
+      "epoch": 9.360933383515242,
+      "grad_norm": 0.18346701562404633,
+      "learning_rate": 5e-06,
+      "loss": 0.2115,
+      "step": 12436
+    },
+    {
+      "epoch": 9.361686112156567,
+      "grad_norm": 0.1809374988079071,
+      "learning_rate": 5e-06,
+      "loss": 0.1681,
+      "step": 12437
+    },
+    {
+      "epoch": 9.362438840797893,
+      "grad_norm": 0.18268536031246185,
+      "learning_rate": 5e-06,
+      "loss": 0.155,
+      "step": 12438
+    },
+    {
+      "epoch": 9.363191569439216,
+      "grad_norm": 0.16440971195697784,
+      "learning_rate": 5e-06,
+      "loss": 0.1765,
+      "step": 12439
+    },
+    {
+      "epoch": 9.363944298080542,
+      "grad_norm": 0.19820912182331085,
+      "learning_rate": 5e-06,
+      "loss": 0.1961,
+      "step": 12440
+    },
+    {
+      "epoch": 9.364697026721867,
+      "grad_norm": 0.17176085710525513,
+      "learning_rate": 5e-06,
+      "loss": 0.2002,
+      "step": 12441
+    },
+    {
+      "epoch": 9.365449755363192,
+      "grad_norm": 0.18465957045555115,
+      "learning_rate": 5e-06,
+      "loss": 0.1913,
+      "step": 12442
+    },
+    {
+      "epoch": 9.366202484004516,
+      "grad_norm": 0.1767406314611435,
+      "learning_rate": 5e-06,
+      "loss": 0.1668,
+      "step": 12443
+    },
+    {
+      "epoch": 9.366955212645841,
+      "grad_norm": 0.16899657249450684,
+      "learning_rate": 5e-06,
+      "loss": 0.1276,
+      "step": 12444
+    },
+    {
+      "epoch": 9.367707941287167,
+      "grad_norm": 0.1793767213821411,
+      "learning_rate": 5e-06,
+      "loss": 0.1096,
+      "step": 12445
+    },
+    {
+      "epoch": 9.36846066992849,
+      "grad_norm": 0.1886938512325287,
+      "learning_rate": 5e-06,
+      "loss": 0.195,
+      "step": 12446
+    },
+    {
+      "epoch": 9.369213398569816,
+      "grad_norm": 0.18064574897289276,
+      "learning_rate": 5e-06,
+      "loss": 0.1584,
+      "step": 12447
+    },
+    {
+      "epoch": 9.369966127211141,
+      "grad_norm": 0.16064827144145966,
+      "learning_rate": 5e-06,
+      "loss": 0.1671,
+      "step": 12448
+    },
+    {
+      "epoch": 9.370718855852465,
+      "grad_norm": 0.1963978111743927,
+      "learning_rate": 5e-06,
+      "loss": 0.1395,
+      "step": 12449
+    },
+    {
+      "epoch": 9.37147158449379,
+      "grad_norm": 0.1847323477268219,
+      "learning_rate": 5e-06,
+      "loss": 0.1638,
+      "step": 12450
+    },
+    {
+      "epoch": 9.372224313135115,
+      "grad_norm": 0.16962143778800964,
+      "learning_rate": 5e-06,
+      "loss": 0.1657,
+      "step": 12451
+    },
+    {
+      "epoch": 9.372977041776439,
+      "grad_norm": 0.14879141747951508,
+      "learning_rate": 5e-06,
+      "loss": 0.0683,
+      "step": 12452
+    },
+    {
+      "epoch": 9.373729770417764,
+      "grad_norm": 0.1728792041540146,
+      "learning_rate": 5e-06,
+      "loss": 0.1178,
+      "step": 12453
+    },
+    {
+      "epoch": 9.37448249905909,
+      "grad_norm": 0.1584572046995163,
+      "learning_rate": 5e-06,
+      "loss": 0.0996,
+      "step": 12454
+    },
+    {
+      "epoch": 9.375235227700413,
+      "grad_norm": 0.17407071590423584,
+      "learning_rate": 5e-06,
+      "loss": 0.1352,
+      "step": 12455
+    },
+    {
+      "epoch": 9.375987956341739,
+      "grad_norm": 0.18558354675769806,
+      "learning_rate": 5e-06,
+      "loss": 0.1248,
+      "step": 12456
+    },
+    {
+      "epoch": 9.376740684983064,
+      "grad_norm": 0.17762690782546997,
+      "learning_rate": 5e-06,
+      "loss": 0.1644,
+      "step": 12457
+    },
+    {
+      "epoch": 9.37749341362439,
+      "grad_norm": 0.16983097791671753,
+      "learning_rate": 5e-06,
+      "loss": 0.1273,
+      "step": 12458
+    },
+    {
+      "epoch": 9.378246142265713,
+      "grad_norm": 0.18893837928771973,
+      "learning_rate": 5e-06,
+      "loss": 0.1676,
+      "step": 12459
+    },
+    {
+      "epoch": 9.378998870907038,
+      "grad_norm": 0.1761825978755951,
+      "learning_rate": 5e-06,
+      "loss": 0.1496,
+      "step": 12460
+    },
+    {
+      "epoch": 9.379751599548364,
+      "grad_norm": 0.19398389756679535,
+      "learning_rate": 5e-06,
+      "loss": 0.1776,
+      "step": 12461
+    },
+    {
+      "epoch": 9.380504328189687,
+      "grad_norm": 0.19648756086826324,
+      "learning_rate": 5e-06,
+      "loss": 0.1509,
+      "step": 12462
+    },
+    {
+      "epoch": 9.381257056831013,
+      "grad_norm": 0.16647085547447205,
+      "learning_rate": 5e-06,
+      "loss": 0.1769,
+      "step": 12463
+    },
+    {
+      "epoch": 9.382009785472338,
+      "grad_norm": 0.17056842148303986,
+      "learning_rate": 5e-06,
+      "loss": 0.1891,
+      "step": 12464
+    },
+    {
+      "epoch": 9.382762514113661,
+      "grad_norm": 0.16157355904579163,
+      "learning_rate": 5e-06,
+      "loss": 0.0763,
+      "step": 12465
+    },
+    {
+      "epoch": 9.383515242754987,
+      "grad_norm": 0.17868538200855255,
+      "learning_rate": 5e-06,
+      "loss": 0.1733,
+      "step": 12466
+    },
+    {
+      "epoch": 9.384267971396312,
+      "grad_norm": 0.15932448208332062,
+      "learning_rate": 5e-06,
+      "loss": 0.0975,
+      "step": 12467
+    },
+    {
+      "epoch": 9.385020700037636,
+      "grad_norm": 0.18492689728736877,
+      "learning_rate": 5e-06,
+      "loss": 0.1624,
+      "step": 12468
+    },
+    {
+      "epoch": 9.385773428678961,
+      "grad_norm": 0.2041931450366974,
+      "learning_rate": 5e-06,
+      "loss": 0.198,
+      "step": 12469
+    },
+    {
+      "epoch": 9.386526157320287,
+      "grad_norm": 0.18623551726341248,
+      "learning_rate": 5e-06,
+      "loss": 0.1662,
+      "step": 12470
+    },
+    {
+      "epoch": 9.38727888596161,
+      "grad_norm": 0.16694237291812897,
+      "learning_rate": 5e-06,
+      "loss": 0.0916,
+      "step": 12471
+    },
+    {
+      "epoch": 9.388031614602935,
+      "grad_norm": 0.1746043711900711,
+      "learning_rate": 5e-06,
+      "loss": 0.1076,
+      "step": 12472
+    },
+    {
+      "epoch": 9.38878434324426,
+      "grad_norm": 0.1801327019929886,
+      "learning_rate": 5e-06,
+      "loss": 0.1916,
+      "step": 12473
+    },
+    {
+      "epoch": 9.389537071885584,
+      "grad_norm": 0.1930829882621765,
+      "learning_rate": 5e-06,
+      "loss": 0.1425,
+      "step": 12474
+    },
+    {
+      "epoch": 9.39028980052691,
+      "grad_norm": 0.19482870399951935,
+      "learning_rate": 5e-06,
+      "loss": 0.2245,
+      "step": 12475
+    },
+    {
+      "epoch": 9.391042529168235,
+      "grad_norm": 0.17944084107875824,
+      "learning_rate": 5e-06,
+      "loss": 0.1833,
+      "step": 12476
+    },
+    {
+      "epoch": 9.39179525780956,
+      "grad_norm": 0.21474814414978027,
+      "learning_rate": 5e-06,
+      "loss": 0.1843,
+      "step": 12477
+    },
+    {
+      "epoch": 9.392547986450884,
+      "grad_norm": 0.19127865135669708,
+      "learning_rate": 5e-06,
+      "loss": 0.1944,
+      "step": 12478
+    },
+    {
+      "epoch": 9.39330071509221,
+      "grad_norm": 0.18602025508880615,
+      "learning_rate": 5e-06,
+      "loss": 0.175,
+      "step": 12479
+    },
+    {
+      "epoch": 9.394053443733535,
+      "grad_norm": 0.19035854935646057,
+      "learning_rate": 5e-06,
+      "loss": 0.1646,
+      "step": 12480
+    },
+    {
+      "epoch": 9.394806172374858,
+      "grad_norm": 0.16152794659137726,
+      "learning_rate": 5e-06,
+      "loss": 0.1329,
+      "step": 12481
+    },
+    {
+      "epoch": 9.395558901016184,
+      "grad_norm": 0.16721445322036743,
+      "learning_rate": 5e-06,
+      "loss": 0.1106,
+      "step": 12482
+    },
+    {
+      "epoch": 9.396311629657509,
+      "grad_norm": 0.18146821856498718,
+      "learning_rate": 5e-06,
+      "loss": 0.1024,
+      "step": 12483
+    },
+    {
+      "epoch": 9.397064358298833,
+      "grad_norm": 0.22571930289268494,
+      "learning_rate": 5e-06,
+      "loss": 0.1487,
+      "step": 12484
+    },
+    {
+      "epoch": 9.397817086940158,
+      "grad_norm": 0.23162078857421875,
+      "learning_rate": 5e-06,
+      "loss": 0.1815,
+      "step": 12485
+    },
+    {
+      "epoch": 9.398569815581483,
+      "grad_norm": 0.1720045953989029,
+      "learning_rate": 5e-06,
+      "loss": 0.1941,
+      "step": 12486
+    },
+    {
+      "epoch": 9.399322544222807,
+      "grad_norm": 0.18282577395439148,
+      "learning_rate": 5e-06,
+      "loss": 0.1526,
+      "step": 12487
+    },
+    {
+      "epoch": 9.400075272864132,
+      "grad_norm": 0.17296597361564636,
+      "learning_rate": 5e-06,
+      "loss": 0.1765,
+      "step": 12488
+    },
+    {
+      "epoch": 9.400828001505458,
+      "grad_norm": 0.2086920291185379,
+      "learning_rate": 5e-06,
+      "loss": 0.1627,
+      "step": 12489
+    },
+    {
+      "epoch": 9.401580730146781,
+      "grad_norm": 0.19452804327011108,
+      "learning_rate": 5e-06,
+      "loss": 0.1629,
+      "step": 12490
+    },
+    {
+      "epoch": 9.402333458788107,
+      "grad_norm": 0.16943788528442383,
+      "learning_rate": 5e-06,
+      "loss": 0.1864,
+      "step": 12491
+    },
+    {
+      "epoch": 9.403086187429432,
+      "grad_norm": 0.1946740746498108,
+      "learning_rate": 5e-06,
+      "loss": 0.1251,
+      "step": 12492
+    },
+    {
+      "epoch": 9.403838916070757,
+      "grad_norm": 0.1932232826948166,
+      "learning_rate": 5e-06,
+      "loss": 0.1184,
+      "step": 12493
+    },
+    {
+      "epoch": 9.404591644712081,
+      "grad_norm": 0.16324259340763092,
+      "learning_rate": 5e-06,
+      "loss": 0.1091,
+      "step": 12494
+    },
+    {
+      "epoch": 9.405344373353406,
+      "grad_norm": 0.20881254971027374,
+      "learning_rate": 5e-06,
+      "loss": 0.1893,
+      "step": 12495
+    },
+    {
+      "epoch": 9.406097101994732,
+      "grad_norm": 0.16017180681228638,
+      "learning_rate": 5e-06,
+      "loss": 0.0698,
+      "step": 12496
+    },
+    {
+      "epoch": 9.406849830636055,
+      "grad_norm": 0.17400237917900085,
+      "learning_rate": 5e-06,
+      "loss": 0.1202,
+      "step": 12497
+    },
+    {
+      "epoch": 9.40760255927738,
+      "grad_norm": 0.17291490733623505,
+      "learning_rate": 5e-06,
+      "loss": 0.169,
+      "step": 12498
+    },
+    {
+      "epoch": 9.408355287918706,
+      "grad_norm": 0.16540738940238953,
+      "learning_rate": 5e-06,
+      "loss": 0.1906,
+      "step": 12499
+    },
+    {
+      "epoch": 9.40910801656003,
+      "grad_norm": 0.18478401005268097,
+      "learning_rate": 5e-06,
+      "loss": 0.1786,
+      "step": 12500
+    },
+    {
+      "epoch": 9.409860745201355,
+      "grad_norm": 0.16684596240520477,
+      "learning_rate": 5e-06,
+      "loss": 0.1184,
+      "step": 12501
+    },
+    {
+      "epoch": 9.41061347384268,
+      "grad_norm": 0.1840030550956726,
+      "learning_rate": 5e-06,
+      "loss": 0.1827,
+      "step": 12502
+    },
+    {
+      "epoch": 9.411366202484004,
+      "grad_norm": 0.18321962654590607,
+      "learning_rate": 5e-06,
+      "loss": 0.1526,
+      "step": 12503
+    },
+    {
+      "epoch": 9.41211893112533,
+      "grad_norm": 0.1683277040719986,
+      "learning_rate": 5e-06,
+      "loss": 0.1291,
+      "step": 12504
+    },
+    {
+      "epoch": 9.412871659766655,
+      "grad_norm": 0.16169039905071259,
+      "learning_rate": 5e-06,
+      "loss": 0.1685,
+      "step": 12505
+    },
+    {
+      "epoch": 9.413624388407978,
+      "grad_norm": 0.16548600792884827,
+      "learning_rate": 5e-06,
+      "loss": 0.2185,
+      "step": 12506
+    },
+    {
+      "epoch": 9.414377117049304,
+      "grad_norm": 0.18172818422317505,
+      "learning_rate": 5e-06,
+      "loss": 0.1717,
+      "step": 12507
+    },
+    {
+      "epoch": 9.415129845690629,
+      "grad_norm": 0.16921134293079376,
+      "learning_rate": 5e-06,
+      "loss": 0.1482,
+      "step": 12508
+    },
+    {
+      "epoch": 9.415882574331953,
+      "grad_norm": 0.1908293068408966,
+      "learning_rate": 5e-06,
+      "loss": 0.1689,
+      "step": 12509
+    },
+    {
+      "epoch": 9.416635302973278,
+      "grad_norm": 0.16891705989837646,
+      "learning_rate": 5e-06,
+      "loss": 0.1074,
+      "step": 12510
+    },
+    {
+      "epoch": 9.417388031614603,
+      "grad_norm": 0.1626226156949997,
+      "learning_rate": 5e-06,
+      "loss": 0.0766,
+      "step": 12511
+    },
+    {
+      "epoch": 9.418140760255927,
+      "grad_norm": 0.16896575689315796,
+      "learning_rate": 5e-06,
+      "loss": 0.0856,
+      "step": 12512
+    },
+    {
+      "epoch": 9.418893488897252,
+      "grad_norm": 0.17056293785572052,
+      "learning_rate": 5e-06,
+      "loss": 0.1231,
+      "step": 12513
+    },
+    {
+      "epoch": 9.419646217538578,
+      "grad_norm": 0.21083857119083405,
+      "learning_rate": 5e-06,
+      "loss": 0.1629,
+      "step": 12514
+    },
+    {
+      "epoch": 9.420398946179903,
+      "grad_norm": 0.1661134660243988,
+      "learning_rate": 5e-06,
+      "loss": 0.107,
+      "step": 12515
+    },
+    {
+      "epoch": 9.421151674821227,
+      "grad_norm": 0.1972878873348236,
+      "learning_rate": 5e-06,
+      "loss": 0.1912,
+      "step": 12516
+    },
+    {
+      "epoch": 9.421904403462552,
+      "grad_norm": 0.17488929629325867,
+      "learning_rate": 5e-06,
+      "loss": 0.1546,
+      "step": 12517
+    },
+    {
+      "epoch": 9.422657132103877,
+      "grad_norm": 0.17049431800842285,
+      "learning_rate": 5e-06,
+      "loss": 0.1243,
+      "step": 12518
+    },
+    {
+      "epoch": 9.4234098607452,
+      "grad_norm": 0.17184659838676453,
+      "learning_rate": 5e-06,
+      "loss": 0.1362,
+      "step": 12519
+    },
+    {
+      "epoch": 9.424162589386526,
+      "grad_norm": 0.18684852123260498,
+      "learning_rate": 5e-06,
+      "loss": 0.1817,
+      "step": 12520
+    },
+    {
+      "epoch": 9.424915318027852,
+      "grad_norm": 0.18497949838638306,
+      "learning_rate": 5e-06,
+      "loss": 0.1346,
+      "step": 12521
+    },
+    {
+      "epoch": 9.425668046669175,
+      "grad_norm": 0.18118883669376373,
+      "learning_rate": 5e-06,
+      "loss": 0.1696,
+      "step": 12522
+    },
+    {
+      "epoch": 9.4264207753105,
+      "grad_norm": 0.17882047593593597,
+      "learning_rate": 5e-06,
+      "loss": 0.1335,
+      "step": 12523
+    },
+    {
+      "epoch": 9.427173503951826,
+      "grad_norm": 0.16197362542152405,
+      "learning_rate": 5e-06,
+      "loss": 0.1348,
+      "step": 12524
+    },
+    {
+      "epoch": 9.42792623259315,
+      "grad_norm": 0.16416355967521667,
+      "learning_rate": 5e-06,
+      "loss": 0.1106,
+      "step": 12525
+    },
+    {
+      "epoch": 9.428678961234475,
+      "grad_norm": 0.1942494809627533,
+      "learning_rate": 5e-06,
+      "loss": 0.1269,
+      "step": 12526
+    },
+    {
+      "epoch": 9.4294316898758,
+      "grad_norm": 0.1870129555463791,
+      "learning_rate": 5e-06,
+      "loss": 0.1605,
+      "step": 12527
+    },
+    {
+      "epoch": 9.430184418517124,
+      "grad_norm": 0.1633111983537674,
+      "learning_rate": 5e-06,
+      "loss": 0.1141,
+      "step": 12528
+    },
+    {
+      "epoch": 9.430937147158449,
+      "grad_norm": 0.16715744137763977,
+      "learning_rate": 5e-06,
+      "loss": 0.1506,
+      "step": 12529
+    },
+    {
+      "epoch": 9.431689875799774,
+      "grad_norm": 0.18693231046199799,
+      "learning_rate": 5e-06,
+      "loss": 0.175,
+      "step": 12530
+    },
+    {
+      "epoch": 9.4324426044411,
+      "grad_norm": 0.18449442088603973,
+      "learning_rate": 5e-06,
+      "loss": 0.1847,
+      "step": 12531
+    },
+    {
+      "epoch": 9.433195333082423,
+      "grad_norm": 0.184881791472435,
+      "learning_rate": 5e-06,
+      "loss": 0.1681,
+      "step": 12532
+    },
+    {
+      "epoch": 9.433948061723749,
+      "grad_norm": 0.16803991794586182,
+      "learning_rate": 5e-06,
+      "loss": 0.0903,
+      "step": 12533
+    },
+    {
+      "epoch": 9.434700790365074,
+      "grad_norm": 0.1671995371580124,
+      "learning_rate": 5e-06,
+      "loss": 0.1388,
+      "step": 12534
+    },
+    {
+      "epoch": 9.435453519006398,
+      "grad_norm": 0.20307239890098572,
+      "learning_rate": 5e-06,
+      "loss": 0.1667,
+      "step": 12535
+    },
+    {
+      "epoch": 9.436206247647723,
+      "grad_norm": 0.1566038727760315,
+      "learning_rate": 5e-06,
+      "loss": 0.0886,
+      "step": 12536
+    },
+    {
+      "epoch": 9.436958976289048,
+      "grad_norm": 0.17045238614082336,
+      "learning_rate": 5e-06,
+      "loss": 0.1049,
+      "step": 12537
+    },
+    {
+      "epoch": 9.437711704930372,
+      "grad_norm": 0.16857630014419556,
+      "learning_rate": 5e-06,
+      "loss": 0.1896,
+      "step": 12538
+    },
+    {
+      "epoch": 9.438464433571697,
+      "grad_norm": 0.17150327563285828,
+      "learning_rate": 5e-06,
+      "loss": 0.1312,
+      "step": 12539
+    },
+    {
+      "epoch": 9.439217162213023,
+      "grad_norm": 0.21722692251205444,
+      "learning_rate": 5e-06,
+      "loss": 0.222,
+      "step": 12540
+    },
+    {
+      "epoch": 9.439969890854346,
+      "grad_norm": 0.17659194767475128,
+      "learning_rate": 5e-06,
+      "loss": 0.1281,
+      "step": 12541
+    },
+    {
+      "epoch": 9.440722619495672,
+      "grad_norm": 0.18674342334270477,
+      "learning_rate": 5e-06,
+      "loss": 0.1724,
+      "step": 12542
+    },
+    {
+      "epoch": 9.441475348136997,
+      "grad_norm": 0.19115057587623596,
+      "learning_rate": 5e-06,
+      "loss": 0.167,
+      "step": 12543
+    },
+    {
+      "epoch": 9.44222807677832,
+      "grad_norm": 0.18175572156906128,
+      "learning_rate": 5e-06,
+      "loss": 0.2071,
+      "step": 12544
+    },
+    {
+      "epoch": 9.442980805419646,
+      "grad_norm": 0.1842178851366043,
+      "learning_rate": 5e-06,
+      "loss": 0.2185,
+      "step": 12545
+    },
+    {
+      "epoch": 9.443733534060971,
+      "grad_norm": 0.17799831926822662,
+      "learning_rate": 5e-06,
+      "loss": 0.0688,
+      "step": 12546
+    },
+    {
+      "epoch": 9.444486262702295,
+      "grad_norm": 0.19066062569618225,
+      "learning_rate": 5e-06,
+      "loss": 0.1173,
+      "step": 12547
+    },
+    {
+      "epoch": 9.44523899134362,
+      "grad_norm": 0.1562148481607437,
+      "learning_rate": 5e-06,
+      "loss": 0.1254,
+      "step": 12548
+    },
+    {
+      "epoch": 9.445991719984946,
+      "grad_norm": 0.16267499327659607,
+      "learning_rate": 5e-06,
+      "loss": 0.1215,
+      "step": 12549
+    },
+    {
+      "epoch": 9.446744448626271,
+      "grad_norm": 0.15512672066688538,
+      "learning_rate": 5e-06,
+      "loss": 0.128,
+      "step": 12550
+    },
+    {
+      "epoch": 9.447497177267595,
+      "grad_norm": 0.19150015711784363,
+      "learning_rate": 5e-06,
+      "loss": 0.151,
+      "step": 12551
+    },
+    {
+      "epoch": 9.44824990590892,
+      "grad_norm": 0.17326807975769043,
+      "learning_rate": 5e-06,
+      "loss": 0.1467,
+      "step": 12552
+    },
+    {
+      "epoch": 9.449002634550245,
+      "grad_norm": 0.17222610116004944,
+      "learning_rate": 5e-06,
+      "loss": 0.1384,
+      "step": 12553
+    },
+    {
+      "epoch": 9.449755363191569,
+      "grad_norm": 0.1812913566827774,
+      "learning_rate": 5e-06,
+      "loss": 0.1864,
+      "step": 12554
+    },
+    {
+      "epoch": 9.450508091832894,
+      "grad_norm": 0.1616690754890442,
+      "learning_rate": 5e-06,
+      "loss": 0.1895,
+      "step": 12555
+    },
+    {
+      "epoch": 9.45126082047422,
+      "grad_norm": 0.17902402579784393,
+      "learning_rate": 5e-06,
+      "loss": 0.169,
+      "step": 12556
+    },
+    {
+      "epoch": 9.452013549115543,
+      "grad_norm": 0.18908187747001648,
+      "learning_rate": 5e-06,
+      "loss": 0.1648,
+      "step": 12557
+    },
+    {
+      "epoch": 9.452766277756869,
+      "grad_norm": 0.18106858432292938,
+      "learning_rate": 5e-06,
+      "loss": 0.1864,
+      "step": 12558
+    },
+    {
+      "epoch": 9.453519006398194,
+      "grad_norm": 0.16842184960842133,
+      "learning_rate": 5e-06,
+      "loss": 0.1778,
+      "step": 12559
+    },
+    {
+      "epoch": 9.454271735039518,
+      "grad_norm": 0.18540748953819275,
+      "learning_rate": 5e-06,
+      "loss": 0.1209,
+      "step": 12560
+    },
+    {
+      "epoch": 9.455024463680843,
+      "grad_norm": 0.1718510538339615,
+      "learning_rate": 5e-06,
+      "loss": 0.1009,
+      "step": 12561
+    },
+    {
+      "epoch": 9.455777192322168,
+      "grad_norm": 0.1806427538394928,
+      "learning_rate": 5e-06,
+      "loss": 0.1776,
+      "step": 12562
+    },
+    {
+      "epoch": 9.456529920963492,
+      "grad_norm": 0.17154844105243683,
+      "learning_rate": 5e-06,
+      "loss": 0.1383,
+      "step": 12563
+    },
+    {
+      "epoch": 9.457282649604817,
+      "grad_norm": 0.17885948717594147,
+      "learning_rate": 5e-06,
+      "loss": 0.1512,
+      "step": 12564
+    },
+    {
+      "epoch": 9.458035378246143,
+      "grad_norm": 0.1540403962135315,
+      "learning_rate": 5e-06,
+      "loss": 0.166,
+      "step": 12565
+    },
+    {
+      "epoch": 9.458788106887468,
+      "grad_norm": 0.1842186599969864,
+      "learning_rate": 5e-06,
+      "loss": 0.2214,
+      "step": 12566
+    },
+    {
+      "epoch": 9.459540835528792,
+      "grad_norm": 0.1798463761806488,
+      "learning_rate": 5e-06,
+      "loss": 0.1995,
+      "step": 12567
+    },
+    {
+      "epoch": 9.460293564170117,
+      "grad_norm": 0.18761536478996277,
+      "learning_rate": 5e-06,
+      "loss": 0.1816,
+      "step": 12568
+    },
+    {
+      "epoch": 9.461046292811442,
+      "grad_norm": 0.17413495481014252,
+      "learning_rate": 5e-06,
+      "loss": 0.1879,
+      "step": 12569
+    },
+    {
+      "epoch": 9.461799021452766,
+      "grad_norm": 0.17660245299339294,
+      "learning_rate": 5e-06,
+      "loss": 0.1283,
+      "step": 12570
+    },
+    {
+      "epoch": 9.462551750094091,
+      "grad_norm": 0.20075149834156036,
+      "learning_rate": 5e-06,
+      "loss": 0.0999,
+      "step": 12571
+    },
+    {
+      "epoch": 9.463304478735417,
+      "grad_norm": 0.17064321041107178,
+      "learning_rate": 5e-06,
+      "loss": 0.1205,
+      "step": 12572
+    },
+    {
+      "epoch": 9.46405720737674,
+      "grad_norm": 0.2005404233932495,
+      "learning_rate": 5e-06,
+      "loss": 0.1431,
+      "step": 12573
+    },
+    {
+      "epoch": 9.464809936018066,
+      "grad_norm": 0.2064259797334671,
+      "learning_rate": 5e-06,
+      "loss": 0.1546,
+      "step": 12574
+    },
+    {
+      "epoch": 9.465562664659391,
+      "grad_norm": 0.17377828061580658,
+      "learning_rate": 5e-06,
+      "loss": 0.1569,
+      "step": 12575
+    },
+    {
+      "epoch": 9.466315393300714,
+      "grad_norm": 0.20576892793178558,
+      "learning_rate": 5e-06,
+      "loss": 0.1635,
+      "step": 12576
+    },
+    {
+      "epoch": 9.46706812194204,
+      "grad_norm": 0.15557193756103516,
+      "learning_rate": 5e-06,
+      "loss": 0.1319,
+      "step": 12577
+    },
+    {
+      "epoch": 9.467820850583365,
+      "grad_norm": 0.1619342565536499,
+      "learning_rate": 5e-06,
+      "loss": 0.1421,
+      "step": 12578
+    },
+    {
+      "epoch": 9.468573579224689,
+      "grad_norm": 0.1709316372871399,
+      "learning_rate": 5e-06,
+      "loss": 0.1303,
+      "step": 12579
+    },
+    {
+      "epoch": 9.469326307866014,
+      "grad_norm": 0.1685350090265274,
+      "learning_rate": 5e-06,
+      "loss": 0.0794,
+      "step": 12580
+    },
+    {
+      "epoch": 9.47007903650734,
+      "grad_norm": 0.14887475967407227,
+      "learning_rate": 5e-06,
+      "loss": 0.0927,
+      "step": 12581
+    },
+    {
+      "epoch": 9.470831765148663,
+      "grad_norm": 0.1740163415670395,
+      "learning_rate": 5e-06,
+      "loss": 0.1301,
+      "step": 12582
+    },
+    {
+      "epoch": 9.471584493789988,
+      "grad_norm": 0.17079435288906097,
+      "learning_rate": 5e-06,
+      "loss": 0.1436,
+      "step": 12583
+    },
+    {
+      "epoch": 9.472337222431314,
+      "grad_norm": 0.15865088999271393,
+      "learning_rate": 5e-06,
+      "loss": 0.1191,
+      "step": 12584
+    },
+    {
+      "epoch": 9.473089951072637,
+      "grad_norm": 0.20147015154361725,
+      "learning_rate": 5e-06,
+      "loss": 0.1609,
+      "step": 12585
+    },
+    {
+      "epoch": 9.473842679713963,
+      "grad_norm": 0.17530618607997894,
+      "learning_rate": 5e-06,
+      "loss": 0.1762,
+      "step": 12586
+    },
+    {
+      "epoch": 9.474595408355288,
+      "grad_norm": 0.1748683601617813,
+      "learning_rate": 5e-06,
+      "loss": 0.1341,
+      "step": 12587
+    },
+    {
+      "epoch": 9.475348136996613,
+      "grad_norm": 0.1516442447900772,
+      "learning_rate": 5e-06,
+      "loss": 0.0814,
+      "step": 12588
+    },
+    {
+      "epoch": 9.476100865637937,
+      "grad_norm": 0.17594434320926666,
+      "learning_rate": 5e-06,
+      "loss": 0.107,
+      "step": 12589
+    },
+    {
+      "epoch": 9.476853594279262,
+      "grad_norm": 0.18250566720962524,
+      "learning_rate": 5e-06,
+      "loss": 0.1707,
+      "step": 12590
+    },
+    {
+      "epoch": 9.477606322920588,
+      "grad_norm": 0.19906556606292725,
+      "learning_rate": 5e-06,
+      "loss": 0.1993,
+      "step": 12591
+    },
+    {
+      "epoch": 9.478359051561911,
+      "grad_norm": 0.16571053862571716,
+      "learning_rate": 5e-06,
+      "loss": 0.1497,
+      "step": 12592
+    },
+    {
+      "epoch": 9.479111780203237,
+      "grad_norm": 0.18261781334877014,
+      "learning_rate": 5e-06,
+      "loss": 0.1693,
+      "step": 12593
+    },
+    {
+      "epoch": 9.479864508844562,
+      "grad_norm": 0.2055646777153015,
+      "learning_rate": 5e-06,
+      "loss": 0.1093,
+      "step": 12594
+    },
+    {
+      "epoch": 9.480617237485886,
+      "grad_norm": 0.16207347810268402,
+      "learning_rate": 5e-06,
+      "loss": 0.1776,
+      "step": 12595
+    },
+    {
+      "epoch": 9.481369966127211,
+      "grad_norm": 0.15681879222393036,
+      "learning_rate": 5e-06,
+      "loss": 0.1235,
+      "step": 12596
+    },
+    {
+      "epoch": 9.482122694768536,
+      "grad_norm": 0.18112421035766602,
+      "learning_rate": 5e-06,
+      "loss": 0.1796,
+      "step": 12597
+    },
+    {
+      "epoch": 9.48287542340986,
+      "grad_norm": 0.17677482962608337,
+      "learning_rate": 5e-06,
+      "loss": 0.1409,
+      "step": 12598
+    },
+    {
+      "epoch": 9.483628152051185,
+      "grad_norm": 0.18905535340309143,
+      "learning_rate": 5e-06,
+      "loss": 0.1368,
+      "step": 12599
+    },
+    {
+      "epoch": 9.48438088069251,
+      "grad_norm": 0.17207416892051697,
+      "learning_rate": 5e-06,
+      "loss": 0.1571,
+      "step": 12600
+    },
+    {
+      "epoch": 9.48438088069251,
+      "eval_loss": 0.19207799434661865,
+      "eval_runtime": 452.7725,
+      "eval_samples_per_second": 21.262,
+      "eval_steps_per_second": 0.665,
+      "step": 12600
+    },
+    {
+      "epoch": 9.485133609333834,
+      "grad_norm": 0.18096798658370972,
+      "learning_rate": 5e-06,
+      "loss": 0.1949,
+      "step": 12601
+    },
+    {
+      "epoch": 9.48588633797516,
+      "grad_norm": 0.18510697782039642,
+      "learning_rate": 5e-06,
+      "loss": 0.2012,
+      "step": 12602
+    },
+    {
+      "epoch": 9.486639066616485,
+      "grad_norm": 0.16158144176006317,
+      "learning_rate": 5e-06,
+      "loss": 0.0658,
+      "step": 12603
+    },
+    {
+      "epoch": 9.48739179525781,
+      "grad_norm": 0.1714627891778946,
+      "learning_rate": 5e-06,
+      "loss": 0.1995,
+      "step": 12604
+    },
+    {
+      "epoch": 9.488144523899134,
+      "grad_norm": 0.19897203147411346,
+      "learning_rate": 5e-06,
+      "loss": 0.1864,
+      "step": 12605
+    },
+    {
+      "epoch": 9.48889725254046,
+      "grad_norm": 0.20276780426502228,
+      "learning_rate": 5e-06,
+      "loss": 0.2132,
+      "step": 12606
+    },
+    {
+      "epoch": 9.489649981181785,
+      "grad_norm": 0.16723528504371643,
+      "learning_rate": 5e-06,
+      "loss": 0.1624,
+      "step": 12607
+    },
+    {
+      "epoch": 9.490402709823108,
+      "grad_norm": 0.18356211483478546,
+      "learning_rate": 5e-06,
+      "loss": 0.1443,
+      "step": 12608
+    },
+    {
+      "epoch": 9.491155438464434,
+      "grad_norm": 0.17787978053092957,
+      "learning_rate": 5e-06,
+      "loss": 0.1162,
+      "step": 12609
+    },
+    {
+      "epoch": 9.491908167105759,
+      "grad_norm": 0.17915809154510498,
+      "learning_rate": 5e-06,
+      "loss": 0.1229,
+      "step": 12610
+    },
+    {
+      "epoch": 9.492660895747083,
+      "grad_norm": 0.1848391890525818,
+      "learning_rate": 5e-06,
+      "loss": 0.1471,
+      "step": 12611
+    },
+    {
+      "epoch": 9.493413624388408,
+      "grad_norm": 0.17871622741222382,
+      "learning_rate": 5e-06,
+      "loss": 0.1327,
+      "step": 12612
+    },
+    {
+      "epoch": 9.494166353029733,
+      "grad_norm": 0.15791793167591095,
+      "learning_rate": 5e-06,
+      "loss": 0.1346,
+      "step": 12613
+    },
+    {
+      "epoch": 9.494919081671057,
+      "grad_norm": 0.18940559029579163,
+      "learning_rate": 5e-06,
+      "loss": 0.106,
+      "step": 12614
+    },
+    {
+      "epoch": 9.495671810312382,
+      "grad_norm": 0.15352404117584229,
+      "learning_rate": 5e-06,
+      "loss": 0.0677,
+      "step": 12615
+    },
+    {
+      "epoch": 9.496424538953708,
+      "grad_norm": 0.19264662265777588,
+      "learning_rate": 5e-06,
+      "loss": 0.1693,
+      "step": 12616
+    },
+    {
+      "epoch": 9.497177267595031,
+      "grad_norm": 0.17859819531440735,
+      "learning_rate": 5e-06,
+      "loss": 0.1756,
+      "step": 12617
+    },
+    {
+      "epoch": 9.497929996236357,
+      "grad_norm": 0.1915789246559143,
+      "learning_rate": 5e-06,
+      "loss": 0.1027,
+      "step": 12618
+    },
+    {
+      "epoch": 9.498682724877682,
+      "grad_norm": 0.18018777668476105,
+      "learning_rate": 5e-06,
+      "loss": 0.1189,
+      "step": 12619
+    },
+    {
+      "epoch": 9.499435453519006,
+      "grad_norm": 0.1811092048883438,
+      "learning_rate": 5e-06,
+      "loss": 0.1673,
+      "step": 12620
+    },
+    {
+      "epoch": 9.500188182160331,
+      "grad_norm": 0.17693562805652618,
+      "learning_rate": 5e-06,
+      "loss": 0.1594,
+      "step": 12621
+    },
+    {
+      "epoch": 9.500940910801656,
+      "grad_norm": 0.18777406215667725,
+      "learning_rate": 5e-06,
+      "loss": 0.1798,
+      "step": 12622
+    },
+    {
+      "epoch": 9.501693639442982,
+      "grad_norm": 0.1709575057029724,
+      "learning_rate": 5e-06,
+      "loss": 0.0506,
+      "step": 12623
+    },
+    {
+      "epoch": 9.502446368084305,
+      "grad_norm": 0.1832125037908554,
+      "learning_rate": 5e-06,
+      "loss": 0.193,
+      "step": 12624
+    },
+    {
+      "epoch": 9.50319909672563,
+      "grad_norm": 0.19015517830848694,
+      "learning_rate": 5e-06,
+      "loss": 0.1165,
+      "step": 12625
+    },
+    {
+      "epoch": 9.503951825366956,
+      "grad_norm": 0.16246795654296875,
+      "learning_rate": 5e-06,
+      "loss": 0.1871,
+      "step": 12626
+    },
+    {
+      "epoch": 9.50470455400828,
+      "grad_norm": 0.187358558177948,
+      "learning_rate": 5e-06,
+      "loss": 0.1367,
+      "step": 12627
+    },
+    {
+      "epoch": 9.505457282649605,
+      "grad_norm": 0.1718154400587082,
+      "learning_rate": 5e-06,
+      "loss": 0.1843,
+      "step": 12628
+    },
+    {
+      "epoch": 9.50621001129093,
+      "grad_norm": 0.15886187553405762,
+      "learning_rate": 5e-06,
+      "loss": 0.2037,
+      "step": 12629
+    },
+    {
+      "epoch": 9.506962739932254,
+      "grad_norm": 0.16300109028816223,
+      "learning_rate": 5e-06,
+      "loss": 0.1831,
+      "step": 12630
+    },
+    {
+      "epoch": 9.50771546857358,
+      "grad_norm": 0.1598212569952011,
+      "learning_rate": 5e-06,
+      "loss": 0.1139,
+      "step": 12631
+    },
+    {
+      "epoch": 9.508468197214905,
+      "grad_norm": 0.18298669159412384,
+      "learning_rate": 5e-06,
+      "loss": 0.1272,
+      "step": 12632
+    },
+    {
+      "epoch": 9.509220925856228,
+      "grad_norm": 0.17444010078907013,
+      "learning_rate": 5e-06,
+      "loss": 0.0771,
+      "step": 12633
+    },
+    {
+      "epoch": 9.509973654497553,
+      "grad_norm": 0.19244246184825897,
+      "learning_rate": 5e-06,
+      "loss": 0.228,
+      "step": 12634
+    },
+    {
+      "epoch": 9.510726383138879,
+      "grad_norm": 0.16983583569526672,
+      "learning_rate": 5e-06,
+      "loss": 0.1716,
+      "step": 12635
+    },
+    {
+      "epoch": 9.511479111780202,
+      "grad_norm": 0.19027431309223175,
+      "learning_rate": 5e-06,
+      "loss": 0.175,
+      "step": 12636
+    },
+    {
+      "epoch": 9.512231840421528,
+      "grad_norm": 0.18975171446800232,
+      "learning_rate": 5e-06,
+      "loss": 0.1657,
+      "step": 12637
+    },
+    {
+      "epoch": 9.512984569062853,
+      "grad_norm": 0.1587967723608017,
+      "learning_rate": 5e-06,
+      "loss": 0.107,
+      "step": 12638
+    },
+    {
+      "epoch": 9.513737297704179,
+      "grad_norm": 0.19070805609226227,
+      "learning_rate": 5e-06,
+      "loss": 0.1348,
+      "step": 12639
+    },
+    {
+      "epoch": 9.514490026345502,
+      "grad_norm": 0.198325052857399,
+      "learning_rate": 5e-06,
+      "loss": 0.1713,
+      "step": 12640
+    },
+    {
+      "epoch": 9.515242754986827,
+      "grad_norm": 0.16371159255504608,
+      "learning_rate": 5e-06,
+      "loss": 0.1283,
+      "step": 12641
+    },
+    {
+      "epoch": 9.515995483628153,
+      "grad_norm": 0.1560315638780594,
+      "learning_rate": 5e-06,
+      "loss": 0.1453,
+      "step": 12642
+    },
+    {
+      "epoch": 9.516748212269476,
+      "grad_norm": 0.18498322367668152,
+      "learning_rate": 5e-06,
+      "loss": 0.1212,
+      "step": 12643
+    },
+    {
+      "epoch": 9.517500940910802,
+      "grad_norm": 0.16538071632385254,
+      "learning_rate": 5e-06,
+      "loss": 0.1441,
+      "step": 12644
+    },
+    {
+      "epoch": 9.518253669552127,
+      "grad_norm": 0.15429827570915222,
+      "learning_rate": 5e-06,
+      "loss": 0.1062,
+      "step": 12645
+    },
+    {
+      "epoch": 9.51900639819345,
+      "grad_norm": 0.1772065907716751,
+      "learning_rate": 5e-06,
+      "loss": 0.1264,
+      "step": 12646
+    },
+    {
+      "epoch": 9.519759126834776,
+      "grad_norm": 0.1888885498046875,
+      "learning_rate": 5e-06,
+      "loss": 0.2484,
+      "step": 12647
+    },
+    {
+      "epoch": 9.520511855476101,
+      "grad_norm": 0.17571060359477997,
+      "learning_rate": 5e-06,
+      "loss": 0.0699,
+      "step": 12648
+    },
+    {
+      "epoch": 9.521264584117425,
+      "grad_norm": 0.17800135910511017,
+      "learning_rate": 5e-06,
+      "loss": 0.1164,
+      "step": 12649
+    },
+    {
+      "epoch": 9.52201731275875,
+      "grad_norm": 0.19583550095558167,
+      "learning_rate": 5e-06,
+      "loss": 0.1596,
+      "step": 12650
+    },
+    {
+      "epoch": 9.522770041400076,
+      "grad_norm": 0.187790185213089,
+      "learning_rate": 5e-06,
+      "loss": 0.1428,
+      "step": 12651
+    },
+    {
+      "epoch": 9.5235227700414,
+      "grad_norm": 0.19137470424175262,
+      "learning_rate": 5e-06,
+      "loss": 0.1496,
+      "step": 12652
+    },
+    {
+      "epoch": 9.524275498682725,
+      "grad_norm": 0.18141303956508636,
+      "learning_rate": 5e-06,
+      "loss": 0.1803,
+      "step": 12653
+    },
+    {
+      "epoch": 9.52502822732405,
+      "grad_norm": 0.1810612827539444,
+      "learning_rate": 5e-06,
+      "loss": 0.1921,
+      "step": 12654
+    },
+    {
+      "epoch": 9.525780955965374,
+      "grad_norm": 0.17893944680690765,
+      "learning_rate": 5e-06,
+      "loss": 0.1422,
+      "step": 12655
+    },
+    {
+      "epoch": 9.526533684606699,
+      "grad_norm": 0.18012046813964844,
+      "learning_rate": 5e-06,
+      "loss": 0.0882,
+      "step": 12656
+    },
+    {
+      "epoch": 9.527286413248024,
+      "grad_norm": 0.17549705505371094,
+      "learning_rate": 5e-06,
+      "loss": 0.1726,
+      "step": 12657
+    },
+    {
+      "epoch": 9.528039141889348,
+      "grad_norm": 0.19193421304225922,
+      "learning_rate": 5e-06,
+      "loss": 0.1745,
+      "step": 12658
+    },
+    {
+      "epoch": 9.528791870530673,
+      "grad_norm": 0.1631852239370346,
+      "learning_rate": 5e-06,
+      "loss": 0.1032,
+      "step": 12659
+    },
+    {
+      "epoch": 9.529544599171999,
+      "grad_norm": 0.17168989777565002,
+      "learning_rate": 5e-06,
+      "loss": 0.1103,
+      "step": 12660
+    },
+    {
+      "epoch": 9.530297327813324,
+      "grad_norm": 0.18796871602535248,
+      "learning_rate": 5e-06,
+      "loss": 0.154,
+      "step": 12661
+    },
+    {
+      "epoch": 9.531050056454648,
+      "grad_norm": 0.17126572132110596,
+      "learning_rate": 5e-06,
+      "loss": 0.1454,
+      "step": 12662
+    },
+    {
+      "epoch": 9.531802785095973,
+      "grad_norm": 0.19784051179885864,
+      "learning_rate": 5e-06,
+      "loss": 0.14,
+      "step": 12663
+    },
+    {
+      "epoch": 9.532555513737298,
+      "grad_norm": 0.18331386148929596,
+      "learning_rate": 5e-06,
+      "loss": 0.0943,
+      "step": 12664
+    },
+    {
+      "epoch": 9.533308242378622,
+      "grad_norm": 0.1488504856824875,
+      "learning_rate": 5e-06,
+      "loss": 0.0571,
+      "step": 12665
+    },
+    {
+      "epoch": 9.534060971019947,
+      "grad_norm": 0.1705537736415863,
+      "learning_rate": 5e-06,
+      "loss": 0.1402,
+      "step": 12666
+    },
+    {
+      "epoch": 9.534813699661273,
+      "grad_norm": 0.19176989793777466,
+      "learning_rate": 5e-06,
+      "loss": 0.148,
+      "step": 12667
+    },
+    {
+      "epoch": 9.535566428302596,
+      "grad_norm": 0.1686851680278778,
+      "learning_rate": 5e-06,
+      "loss": 0.1247,
+      "step": 12668
+    },
+    {
+      "epoch": 9.536319156943922,
+      "grad_norm": 0.1853509545326233,
+      "learning_rate": 5e-06,
+      "loss": 0.1422,
+      "step": 12669
+    },
+    {
+      "epoch": 9.537071885585247,
+      "grad_norm": 0.2007792890071869,
+      "learning_rate": 5e-06,
+      "loss": 0.193,
+      "step": 12670
+    },
+    {
+      "epoch": 9.53782461422657,
+      "grad_norm": 0.1601753830909729,
+      "learning_rate": 5e-06,
+      "loss": 0.1389,
+      "step": 12671
+    },
+    {
+      "epoch": 9.538577342867896,
+      "grad_norm": 0.20173467695713043,
+      "learning_rate": 5e-06,
+      "loss": 0.1958,
+      "step": 12672
+    },
+    {
+      "epoch": 9.539330071509221,
+      "grad_norm": 0.1565696746110916,
+      "learning_rate": 5e-06,
+      "loss": 0.0996,
+      "step": 12673
+    },
+    {
+      "epoch": 9.540082800150547,
+      "grad_norm": 0.18365636467933655,
+      "learning_rate": 5e-06,
+      "loss": 0.184,
+      "step": 12674
+    },
+    {
+      "epoch": 9.54083552879187,
+      "grad_norm": 0.20166254043579102,
+      "learning_rate": 5e-06,
+      "loss": 0.1555,
+      "step": 12675
+    },
+    {
+      "epoch": 9.541588257433196,
+      "grad_norm": 0.20229597389698029,
+      "learning_rate": 5e-06,
+      "loss": 0.192,
+      "step": 12676
+    },
+    {
+      "epoch": 9.542340986074521,
+      "grad_norm": 0.19659778475761414,
+      "learning_rate": 5e-06,
+      "loss": 0.1222,
+      "step": 12677
+    },
+    {
+      "epoch": 9.543093714715845,
+      "grad_norm": 0.17945948243141174,
+      "learning_rate": 5e-06,
+      "loss": 0.1358,
+      "step": 12678
+    },
+    {
+      "epoch": 9.54384644335717,
+      "grad_norm": 0.1736861616373062,
+      "learning_rate": 5e-06,
+      "loss": 0.1026,
+      "step": 12679
+    },
+    {
+      "epoch": 9.544599171998495,
+      "grad_norm": 0.22726453840732574,
+      "learning_rate": 5e-06,
+      "loss": 0.1954,
+      "step": 12680
+    },
+    {
+      "epoch": 9.545351900639819,
+      "grad_norm": 0.18931329250335693,
+      "learning_rate": 5e-06,
+      "loss": 0.1847,
+      "step": 12681
+    },
+    {
+      "epoch": 9.546104629281144,
+      "grad_norm": 0.18578998744487762,
+      "learning_rate": 5e-06,
+      "loss": 0.0843,
+      "step": 12682
+    },
+    {
+      "epoch": 9.54685735792247,
+      "grad_norm": 0.17868737876415253,
+      "learning_rate": 5e-06,
+      "loss": 0.1705,
+      "step": 12683
+    },
+    {
+      "epoch": 9.547610086563793,
+      "grad_norm": 0.17408478260040283,
+      "learning_rate": 5e-06,
+      "loss": 0.1049,
+      "step": 12684
+    },
+    {
+      "epoch": 9.548362815205119,
+      "grad_norm": 0.18304981291294098,
+      "learning_rate": 5e-06,
+      "loss": 0.1214,
+      "step": 12685
+    },
+    {
+      "epoch": 9.549115543846444,
+      "grad_norm": 0.20909567177295685,
+      "learning_rate": 5e-06,
+      "loss": 0.1874,
+      "step": 12686
+    },
+    {
+      "epoch": 9.549868272487767,
+      "grad_norm": 0.19168438017368317,
+      "learning_rate": 5e-06,
+      "loss": 0.2836,
+      "step": 12687
+    },
+    {
+      "epoch": 9.550621001129093,
+      "grad_norm": 0.19112582504749298,
+      "learning_rate": 5e-06,
+      "loss": 0.1357,
+      "step": 12688
+    },
+    {
+      "epoch": 9.551373729770418,
+      "grad_norm": 0.18414270877838135,
+      "learning_rate": 5e-06,
+      "loss": 0.1454,
+      "step": 12689
+    },
+    {
+      "epoch": 9.552126458411742,
+      "grad_norm": 0.20064914226531982,
+      "learning_rate": 5e-06,
+      "loss": 0.1537,
+      "step": 12690
+    },
+    {
+      "epoch": 9.552879187053067,
+      "grad_norm": 0.1999475061893463,
+      "learning_rate": 5e-06,
+      "loss": 0.1689,
+      "step": 12691
+    },
+    {
+      "epoch": 9.553631915694393,
+      "grad_norm": 0.1813231110572815,
+      "learning_rate": 5e-06,
+      "loss": 0.0727,
+      "step": 12692
+    },
+    {
+      "epoch": 9.554384644335716,
+      "grad_norm": 0.18661507964134216,
+      "learning_rate": 5e-06,
+      "loss": 0.1571,
+      "step": 12693
+    },
+    {
+      "epoch": 9.555137372977041,
+      "grad_norm": 0.179462268948555,
+      "learning_rate": 5e-06,
+      "loss": 0.1042,
+      "step": 12694
+    },
+    {
+      "epoch": 9.555890101618367,
+      "grad_norm": 0.16433210670948029,
+      "learning_rate": 5e-06,
+      "loss": 0.2039,
+      "step": 12695
+    },
+    {
+      "epoch": 9.556642830259692,
+      "grad_norm": 0.1902100145816803,
+      "learning_rate": 5e-06,
+      "loss": 0.1456,
+      "step": 12696
+    },
+    {
+      "epoch": 9.557395558901016,
+      "grad_norm": 0.2605251669883728,
+      "learning_rate": 5e-06,
+      "loss": 0.2039,
+      "step": 12697
+    },
+    {
+      "epoch": 9.558148287542341,
+      "grad_norm": 0.189018115401268,
+      "learning_rate": 5e-06,
+      "loss": 0.143,
+      "step": 12698
+    },
+    {
+      "epoch": 9.558901016183666,
+      "grad_norm": 0.20691576600074768,
+      "learning_rate": 5e-06,
+      "loss": 0.1537,
+      "step": 12699
+    },
+    {
+      "epoch": 9.55965374482499,
+      "grad_norm": 0.18962804973125458,
+      "learning_rate": 5e-06,
+      "loss": 0.1746,
+      "step": 12700
+    },
+    {
+      "epoch": 9.560406473466315,
+      "grad_norm": 0.1867157518863678,
+      "learning_rate": 5e-06,
+      "loss": 0.0882,
+      "step": 12701
+    },
+    {
+      "epoch": 9.56115920210764,
+      "grad_norm": 0.17726914584636688,
+      "learning_rate": 5e-06,
+      "loss": 0.122,
+      "step": 12702
+    },
+    {
+      "epoch": 9.561911930748964,
+      "grad_norm": 0.1595357060432434,
+      "learning_rate": 5e-06,
+      "loss": 0.1338,
+      "step": 12703
+    },
+    {
+      "epoch": 9.56266465939029,
+      "grad_norm": 0.19325436651706696,
+      "learning_rate": 5e-06,
+      "loss": 0.1084,
+      "step": 12704
+    },
+    {
+      "epoch": 9.563417388031615,
+      "grad_norm": 0.17637848854064941,
+      "learning_rate": 5e-06,
+      "loss": 0.1541,
+      "step": 12705
+    },
+    {
+      "epoch": 9.564170116672939,
+      "grad_norm": 0.140973299741745,
+      "learning_rate": 5e-06,
+      "loss": 0.1154,
+      "step": 12706
+    },
+    {
+      "epoch": 9.564922845314264,
+      "grad_norm": 0.17898184061050415,
+      "learning_rate": 5e-06,
+      "loss": 0.1074,
+      "step": 12707
+    },
+    {
+      "epoch": 9.56567557395559,
+      "grad_norm": 0.19284410774707794,
+      "learning_rate": 5e-06,
+      "loss": 0.1298,
+      "step": 12708
+    },
+    {
+      "epoch": 9.566428302596913,
+      "grad_norm": 0.16026563942432404,
+      "learning_rate": 5e-06,
+      "loss": 0.1064,
+      "step": 12709
+    },
+    {
+      "epoch": 9.567181031238238,
+      "grad_norm": 0.19388669729232788,
+      "learning_rate": 5e-06,
+      "loss": 0.1385,
+      "step": 12710
+    },
+    {
+      "epoch": 9.567933759879564,
+      "grad_norm": 0.18271908164024353,
+      "learning_rate": 5e-06,
+      "loss": 0.1559,
+      "step": 12711
+    },
+    {
+      "epoch": 9.568686488520889,
+      "grad_norm": 0.1856573075056076,
+      "learning_rate": 5e-06,
+      "loss": 0.1763,
+      "step": 12712
+    },
+    {
+      "epoch": 9.569439217162213,
+      "grad_norm": 0.17689405381679535,
+      "learning_rate": 5e-06,
+      "loss": 0.1776,
+      "step": 12713
+    },
+    {
+      "epoch": 9.570191945803538,
+      "grad_norm": 0.16489508748054504,
+      "learning_rate": 5e-06,
+      "loss": 0.1829,
+      "step": 12714
+    },
+    {
+      "epoch": 9.570944674444863,
+      "grad_norm": 0.15739701688289642,
+      "learning_rate": 5e-06,
+      "loss": 0.1237,
+      "step": 12715
+    },
+    {
+      "epoch": 9.571697403086187,
+      "grad_norm": 0.16427300870418549,
+      "learning_rate": 5e-06,
+      "loss": 0.1294,
+      "step": 12716
+    },
+    {
+      "epoch": 9.572450131727512,
+      "grad_norm": 0.19150443375110626,
+      "learning_rate": 5e-06,
+      "loss": 0.1958,
+      "step": 12717
+    },
+    {
+      "epoch": 9.573202860368838,
+      "grad_norm": 0.1675471067428589,
+      "learning_rate": 5e-06,
+      "loss": 0.1675,
+      "step": 12718
+    },
+    {
+      "epoch": 9.573955589010161,
+      "grad_norm": 0.1614656001329422,
+      "learning_rate": 5e-06,
+      "loss": 0.145,
+      "step": 12719
+    },
+    {
+      "epoch": 9.574708317651487,
+      "grad_norm": 0.1804269403219223,
+      "learning_rate": 5e-06,
+      "loss": 0.0843,
+      "step": 12720
+    },
+    {
+      "epoch": 9.575461046292812,
+      "grad_norm": 0.1697445511817932,
+      "learning_rate": 5e-06,
+      "loss": 0.187,
+      "step": 12721
+    },
+    {
+      "epoch": 9.576213774934136,
+      "grad_norm": 0.20426605641841888,
+      "learning_rate": 5e-06,
+      "loss": 0.1736,
+      "step": 12722
+    },
+    {
+      "epoch": 9.576966503575461,
+      "grad_norm": 0.1865149438381195,
+      "learning_rate": 5e-06,
+      "loss": 0.2133,
+      "step": 12723
+    },
+    {
+      "epoch": 9.577719232216786,
+      "grad_norm": 0.18441757559776306,
+      "learning_rate": 5e-06,
+      "loss": 0.1916,
+      "step": 12724
+    },
+    {
+      "epoch": 9.57847196085811,
+      "grad_norm": 0.1914454996585846,
+      "learning_rate": 5e-06,
+      "loss": 0.1332,
+      "step": 12725
+    },
+    {
+      "epoch": 9.579224689499435,
+      "grad_norm": 0.4907441735267639,
+      "learning_rate": 5e-06,
+      "loss": 0.1323,
+      "step": 12726
+    },
+    {
+      "epoch": 9.57997741814076,
+      "grad_norm": 0.17360268533229828,
+      "learning_rate": 5e-06,
+      "loss": 0.0993,
+      "step": 12727
+    },
+    {
+      "epoch": 9.580730146782084,
+      "grad_norm": 0.17904236912727356,
+      "learning_rate": 5e-06,
+      "loss": 0.1514,
+      "step": 12728
+    },
+    {
+      "epoch": 9.58148287542341,
+      "grad_norm": 0.20431683957576752,
+      "learning_rate": 5e-06,
+      "loss": 0.1423,
+      "step": 12729
+    },
+    {
+      "epoch": 9.582235604064735,
+      "grad_norm": 0.20095643401145935,
+      "learning_rate": 5e-06,
+      "loss": 0.1669,
+      "step": 12730
+    },
+    {
+      "epoch": 9.582988332706059,
+      "grad_norm": 0.19447919726371765,
+      "learning_rate": 5e-06,
+      "loss": 0.1197,
+      "step": 12731
+    },
+    {
+      "epoch": 9.583741061347384,
+      "grad_norm": 0.1838419884443283,
+      "learning_rate": 5e-06,
+      "loss": 0.1938,
+      "step": 12732
+    },
+    {
+      "epoch": 9.58449378998871,
+      "grad_norm": 0.1976281851530075,
+      "learning_rate": 5e-06,
+      "loss": 0.1633,
+      "step": 12733
+    },
+    {
+      "epoch": 9.585246518630035,
+      "grad_norm": 0.16935133934020996,
+      "learning_rate": 5e-06,
+      "loss": 0.1882,
+      "step": 12734
+    },
+    {
+      "epoch": 9.585999247271358,
+      "grad_norm": 0.16525065898895264,
+      "learning_rate": 5e-06,
+      "loss": 0.0918,
+      "step": 12735
+    },
+    {
+      "epoch": 9.586751975912684,
+      "grad_norm": 0.17744076251983643,
+      "learning_rate": 5e-06,
+      "loss": 0.143,
+      "step": 12736
+    },
+    {
+      "epoch": 9.587504704554009,
+      "grad_norm": 0.16681940853595734,
+      "learning_rate": 5e-06,
+      "loss": 0.1239,
+      "step": 12737
+    },
+    {
+      "epoch": 9.588257433195333,
+      "grad_norm": 0.16389480233192444,
+      "learning_rate": 5e-06,
+      "loss": 0.1314,
+      "step": 12738
+    },
+    {
+      "epoch": 9.589010161836658,
+      "grad_norm": 0.15617085993289948,
+      "learning_rate": 5e-06,
+      "loss": 0.1422,
+      "step": 12739
+    },
+    {
+      "epoch": 9.589762890477983,
+      "grad_norm": 0.1812976449728012,
+      "learning_rate": 5e-06,
+      "loss": 0.1547,
+      "step": 12740
+    },
+    {
+      "epoch": 9.590515619119307,
+      "grad_norm": 0.15461869537830353,
+      "learning_rate": 5e-06,
+      "loss": 0.0995,
+      "step": 12741
+    },
+    {
+      "epoch": 9.591268347760632,
+      "grad_norm": 0.18644297122955322,
+      "learning_rate": 5e-06,
+      "loss": 0.1184,
+      "step": 12742
+    },
+    {
+      "epoch": 9.592021076401958,
+      "grad_norm": 0.15615804493427277,
+      "learning_rate": 5e-06,
+      "loss": 0.1446,
+      "step": 12743
+    },
+    {
+      "epoch": 9.592773805043281,
+      "grad_norm": 0.17230083048343658,
+      "learning_rate": 5e-06,
+      "loss": 0.1506,
+      "step": 12744
+    },
+    {
+      "epoch": 9.593526533684607,
+      "grad_norm": 0.15895739197731018,
+      "learning_rate": 5e-06,
+      "loss": 0.1188,
+      "step": 12745
+    },
+    {
+      "epoch": 9.594279262325932,
+      "grad_norm": 0.18380138278007507,
+      "learning_rate": 5e-06,
+      "loss": 0.1464,
+      "step": 12746
+    },
+    {
+      "epoch": 9.595031990967257,
+      "grad_norm": 0.18512246012687683,
+      "learning_rate": 5e-06,
+      "loss": 0.1246,
+      "step": 12747
+    },
+    {
+      "epoch": 9.59578471960858,
+      "grad_norm": 0.18226245045661926,
+      "learning_rate": 5e-06,
+      "loss": 0.1584,
+      "step": 12748
+    },
+    {
+      "epoch": 9.596537448249906,
+      "grad_norm": 0.21060572564601898,
+      "learning_rate": 5e-06,
+      "loss": 0.2155,
+      "step": 12749
+    },
+    {
+      "epoch": 9.597290176891232,
+      "grad_norm": 0.17634856700897217,
+      "learning_rate": 5e-06,
+      "loss": 0.0928,
+      "step": 12750
+    },
+    {
+      "epoch": 9.598042905532555,
+      "grad_norm": 0.1667959839105606,
+      "learning_rate": 5e-06,
+      "loss": 0.1048,
+      "step": 12751
+    },
+    {
+      "epoch": 9.59879563417388,
+      "grad_norm": 0.15292681753635406,
+      "learning_rate": 5e-06,
+      "loss": 0.1023,
+      "step": 12752
+    },
+    {
+      "epoch": 9.599548362815206,
+      "grad_norm": 0.2023695558309555,
+      "learning_rate": 5e-06,
+      "loss": 0.1171,
+      "step": 12753
+    },
+    {
+      "epoch": 9.60030109145653,
+      "grad_norm": 0.18187546730041504,
+      "learning_rate": 5e-06,
+      "loss": 0.1481,
+      "step": 12754
+    },
+    {
+      "epoch": 9.601053820097855,
+      "grad_norm": 0.1782281994819641,
+      "learning_rate": 5e-06,
+      "loss": 0.1529,
+      "step": 12755
+    },
+    {
+      "epoch": 9.60180654873918,
+      "grad_norm": 0.1899934858083725,
+      "learning_rate": 5e-06,
+      "loss": 0.2153,
+      "step": 12756
+    },
+    {
+      "epoch": 9.602559277380504,
+      "grad_norm": 0.15494771301746368,
+      "learning_rate": 5e-06,
+      "loss": 0.1538,
+      "step": 12757
+    },
+    {
+      "epoch": 9.603312006021829,
+      "grad_norm": 0.20244154334068298,
+      "learning_rate": 5e-06,
+      "loss": 0.2208,
+      "step": 12758
+    },
+    {
+      "epoch": 9.604064734663154,
+      "grad_norm": 0.1654874086380005,
+      "learning_rate": 5e-06,
+      "loss": 0.1293,
+      "step": 12759
+    },
+    {
+      "epoch": 9.604817463304478,
+      "grad_norm": 0.14945417642593384,
+      "learning_rate": 5e-06,
+      "loss": 0.111,
+      "step": 12760
+    },
+    {
+      "epoch": 9.605570191945803,
+      "grad_norm": 0.1846141815185547,
+      "learning_rate": 5e-06,
+      "loss": 0.1535,
+      "step": 12761
+    },
+    {
+      "epoch": 9.606322920587129,
+      "grad_norm": 0.17822545766830444,
+      "learning_rate": 5e-06,
+      "loss": 0.1322,
+      "step": 12762
+    },
+    {
+      "epoch": 9.607075649228452,
+      "grad_norm": 0.19410310685634613,
+      "learning_rate": 5e-06,
+      "loss": 0.1131,
+      "step": 12763
+    },
+    {
+      "epoch": 9.607828377869778,
+      "grad_norm": 0.17865192890167236,
+      "learning_rate": 5e-06,
+      "loss": 0.1349,
+      "step": 12764
+    },
+    {
+      "epoch": 9.608581106511103,
+      "grad_norm": 0.18313173949718475,
+      "learning_rate": 5e-06,
+      "loss": 0.152,
+      "step": 12765
+    },
+    {
+      "epoch": 9.609333835152427,
+      "grad_norm": 0.1463337242603302,
+      "learning_rate": 5e-06,
+      "loss": 0.1026,
+      "step": 12766
+    },
+    {
+      "epoch": 9.610086563793752,
+      "grad_norm": 0.20081964135169983,
+      "learning_rate": 5e-06,
+      "loss": 0.1766,
+      "step": 12767
+    },
+    {
+      "epoch": 9.610839292435077,
+      "grad_norm": 0.17481087148189545,
+      "learning_rate": 5e-06,
+      "loss": 0.188,
+      "step": 12768
+    },
+    {
+      "epoch": 9.611592021076403,
+      "grad_norm": 0.15313100814819336,
+      "learning_rate": 5e-06,
+      "loss": 0.0978,
+      "step": 12769
+    },
+    {
+      "epoch": 9.612344749717726,
+      "grad_norm": 0.16688622534275055,
+      "learning_rate": 5e-06,
+      "loss": 0.1268,
+      "step": 12770
+    },
+    {
+      "epoch": 9.613097478359052,
+      "grad_norm": 0.17885011434555054,
+      "learning_rate": 5e-06,
+      "loss": 0.1946,
+      "step": 12771
+    },
+    {
+      "epoch": 9.613850207000377,
+      "grad_norm": 0.1738988608121872,
+      "learning_rate": 5e-06,
+      "loss": 0.0794,
+      "step": 12772
+    },
+    {
+      "epoch": 9.6146029356417,
+      "grad_norm": 0.16009436547756195,
+      "learning_rate": 5e-06,
+      "loss": 0.2133,
+      "step": 12773
+    },
+    {
+      "epoch": 9.615355664283026,
+      "grad_norm": 0.1820458173751831,
+      "learning_rate": 5e-06,
+      "loss": 0.114,
+      "step": 12774
+    },
+    {
+      "epoch": 9.616108392924351,
+      "grad_norm": 0.1744653731584549,
+      "learning_rate": 5e-06,
+      "loss": 0.1356,
+      "step": 12775
+    },
+    {
+      "epoch": 9.616861121565675,
+      "grad_norm": 0.1885727047920227,
+      "learning_rate": 5e-06,
+      "loss": 0.1078,
+      "step": 12776
+    },
+    {
+      "epoch": 9.617613850207,
+      "grad_norm": 0.16814641654491425,
+      "learning_rate": 5e-06,
+      "loss": 0.0901,
+      "step": 12777
+    },
+    {
+      "epoch": 9.618366578848326,
+      "grad_norm": 0.1854034811258316,
+      "learning_rate": 5e-06,
+      "loss": 0.1435,
+      "step": 12778
+    },
+    {
+      "epoch": 9.61911930748965,
+      "grad_norm": 0.1890738159418106,
+      "learning_rate": 5e-06,
+      "loss": 0.219,
+      "step": 12779
+    },
+    {
+      "epoch": 9.619872036130975,
+      "grad_norm": 0.17539863288402557,
+      "learning_rate": 5e-06,
+      "loss": 0.1519,
+      "step": 12780
+    },
+    {
+      "epoch": 9.6206247647723,
+      "grad_norm": 0.20416124165058136,
+      "learning_rate": 5e-06,
+      "loss": 0.2007,
+      "step": 12781
+    },
+    {
+      "epoch": 9.621377493413624,
+      "grad_norm": 0.19222280383110046,
+      "learning_rate": 5e-06,
+      "loss": 0.1753,
+      "step": 12782
+    },
+    {
+      "epoch": 9.622130222054949,
+      "grad_norm": 0.18646059930324554,
+      "learning_rate": 5e-06,
+      "loss": 0.1589,
+      "step": 12783
+    },
+    {
+      "epoch": 9.622882950696274,
+      "grad_norm": 0.17620918154716492,
+      "learning_rate": 5e-06,
+      "loss": 0.1961,
+      "step": 12784
+    },
+    {
+      "epoch": 9.6236356793376,
+      "grad_norm": 0.1511250138282776,
+      "learning_rate": 5e-06,
+      "loss": 0.1502,
+      "step": 12785
+    },
+    {
+      "epoch": 9.624388407978923,
+      "grad_norm": 0.1593305468559265,
+      "learning_rate": 5e-06,
+      "loss": 0.1478,
+      "step": 12786
+    },
+    {
+      "epoch": 9.625141136620249,
+      "grad_norm": 0.17176076769828796,
+      "learning_rate": 5e-06,
+      "loss": 0.1034,
+      "step": 12787
+    },
+    {
+      "epoch": 9.625893865261574,
+      "grad_norm": 0.19234530627727509,
+      "learning_rate": 5e-06,
+      "loss": 0.1269,
+      "step": 12788
+    },
+    {
+      "epoch": 9.626646593902898,
+      "grad_norm": 0.16257408261299133,
+      "learning_rate": 5e-06,
+      "loss": 0.1419,
+      "step": 12789
+    },
+    {
+      "epoch": 9.627399322544223,
+      "grad_norm": 0.16373685002326965,
+      "learning_rate": 5e-06,
+      "loss": 0.1585,
+      "step": 12790
+    },
+    {
+      "epoch": 9.628152051185548,
+      "grad_norm": 0.1779770851135254,
+      "learning_rate": 5e-06,
+      "loss": 0.1563,
+      "step": 12791
+    },
+    {
+      "epoch": 9.628904779826872,
+      "grad_norm": 0.1900164633989334,
+      "learning_rate": 5e-06,
+      "loss": 0.1662,
+      "step": 12792
+    },
+    {
+      "epoch": 9.629657508468197,
+      "grad_norm": 0.18099670112133026,
+      "learning_rate": 5e-06,
+      "loss": 0.2198,
+      "step": 12793
+    },
+    {
+      "epoch": 9.630410237109523,
+      "grad_norm": 0.19814717769622803,
+      "learning_rate": 5e-06,
+      "loss": 0.141,
+      "step": 12794
+    },
+    {
+      "epoch": 9.631162965750846,
+      "grad_norm": 0.18567705154418945,
+      "learning_rate": 5e-06,
+      "loss": 0.2277,
+      "step": 12795
+    },
+    {
+      "epoch": 9.631915694392172,
+      "grad_norm": 0.1584610491991043,
+      "learning_rate": 5e-06,
+      "loss": 0.1432,
+      "step": 12796
+    },
+    {
+      "epoch": 9.632668423033497,
+      "grad_norm": 0.19070769846439362,
+      "learning_rate": 5e-06,
+      "loss": 0.1422,
+      "step": 12797
+    },
+    {
+      "epoch": 9.63342115167482,
+      "grad_norm": 0.1619374305009842,
+      "learning_rate": 5e-06,
+      "loss": 0.0666,
+      "step": 12798
+    },
+    {
+      "epoch": 9.634173880316146,
+      "grad_norm": 0.21334603428840637,
+      "learning_rate": 5e-06,
+      "loss": 0.1754,
+      "step": 12799
+    },
+    {
+      "epoch": 9.634926608957471,
+      "grad_norm": 0.17233029007911682,
+      "learning_rate": 5e-06,
+      "loss": 0.1381,
+      "step": 12800
+    },
+    {
+      "epoch": 9.634926608957471,
+      "eval_loss": 0.19214271008968353,
+      "eval_runtime": 452.9509,
+      "eval_samples_per_second": 21.254,
+      "eval_steps_per_second": 0.665,
+      "step": 12800
+    },
+    {
+      "epoch": 9.635679337598795,
+      "grad_norm": 0.16918744146823883,
+      "learning_rate": 5e-06,
+      "loss": 0.117,
+      "step": 12801
+    },
+    {
+      "epoch": 9.63643206624012,
+      "grad_norm": 0.18484044075012207,
+      "learning_rate": 5e-06,
+      "loss": 0.1134,
+      "step": 12802
+    },
+    {
+      "epoch": 9.637184794881446,
+      "grad_norm": 0.17575015127658844,
+      "learning_rate": 5e-06,
+      "loss": 0.0902,
+      "step": 12803
+    },
+    {
+      "epoch": 9.63793752352277,
+      "grad_norm": 0.1990097016096115,
+      "learning_rate": 5e-06,
+      "loss": 0.2604,
+      "step": 12804
+    },
+    {
+      "epoch": 9.638690252164094,
+      "grad_norm": 0.15917131304740906,
+      "learning_rate": 5e-06,
+      "loss": 0.1145,
+      "step": 12805
+    },
+    {
+      "epoch": 9.63944298080542,
+      "grad_norm": 0.1754232794046402,
+      "learning_rate": 5e-06,
+      "loss": 0.1416,
+      "step": 12806
+    },
+    {
+      "epoch": 9.640195709446745,
+      "grad_norm": 0.1921617090702057,
+      "learning_rate": 5e-06,
+      "loss": 0.2344,
+      "step": 12807
+    },
+    {
+      "epoch": 9.640948438088069,
+      "grad_norm": 0.18518279492855072,
+      "learning_rate": 5e-06,
+      "loss": 0.1619,
+      "step": 12808
+    },
+    {
+      "epoch": 9.641701166729394,
+      "grad_norm": 0.19433027505874634,
+      "learning_rate": 5e-06,
+      "loss": 0.1877,
+      "step": 12809
+    },
+    {
+      "epoch": 9.64245389537072,
+      "grad_norm": 0.1694916933774948,
+      "learning_rate": 5e-06,
+      "loss": 0.1363,
+      "step": 12810
+    },
+    {
+      "epoch": 9.643206624012043,
+      "grad_norm": 0.21958111226558685,
+      "learning_rate": 5e-06,
+      "loss": 0.1587,
+      "step": 12811
+    },
+    {
+      "epoch": 9.643959352653368,
+      "grad_norm": 0.19633512198925018,
+      "learning_rate": 5e-06,
+      "loss": 0.1308,
+      "step": 12812
+    },
+    {
+      "epoch": 9.644712081294694,
+      "grad_norm": 0.189519003033638,
+      "learning_rate": 5e-06,
+      "loss": 0.1185,
+      "step": 12813
+    },
+    {
+      "epoch": 9.645464809936017,
+      "grad_norm": 0.15937137603759766,
+      "learning_rate": 5e-06,
+      "loss": 0.1573,
+      "step": 12814
+    },
+    {
+      "epoch": 9.646217538577343,
+      "grad_norm": 0.18432556092739105,
+      "learning_rate": 5e-06,
+      "loss": 0.1616,
+      "step": 12815
+    },
+    {
+      "epoch": 9.646970267218668,
+      "grad_norm": 0.1693231463432312,
+      "learning_rate": 5e-06,
+      "loss": 0.1403,
+      "step": 12816
+    },
+    {
+      "epoch": 9.647722995859992,
+      "grad_norm": 0.1633654683828354,
+      "learning_rate": 5e-06,
+      "loss": 0.1548,
+      "step": 12817
+    },
+    {
+      "epoch": 9.648475724501317,
+      "grad_norm": 0.20555944740772247,
+      "learning_rate": 5e-06,
+      "loss": 0.1373,
+      "step": 12818
+    },
+    {
+      "epoch": 9.649228453142642,
+      "grad_norm": 0.18934260308742523,
+      "learning_rate": 5e-06,
+      "loss": 0.1385,
+      "step": 12819
+    },
+    {
+      "epoch": 9.649981181783968,
+      "grad_norm": 0.15839047729969025,
+      "learning_rate": 5e-06,
+      "loss": 0.1296,
+      "step": 12820
+    },
+    {
+      "epoch": 9.650733910425291,
+      "grad_norm": 0.18912792205810547,
+      "learning_rate": 5e-06,
+      "loss": 0.1612,
+      "step": 12821
+    },
+    {
+      "epoch": 9.651486639066617,
+      "grad_norm": 0.19857855141162872,
+      "learning_rate": 5e-06,
+      "loss": 0.285,
+      "step": 12822
+    },
+    {
+      "epoch": 9.652239367707942,
+      "grad_norm": 0.1985931247472763,
+      "learning_rate": 5e-06,
+      "loss": 0.1511,
+      "step": 12823
+    },
+    {
+      "epoch": 9.652992096349266,
+      "grad_norm": 0.17747944593429565,
+      "learning_rate": 5e-06,
+      "loss": 0.1606,
+      "step": 12824
+    },
+    {
+      "epoch": 9.653744824990591,
+      "grad_norm": 0.1757776439189911,
+      "learning_rate": 5e-06,
+      "loss": 0.0988,
+      "step": 12825
+    },
+    {
+      "epoch": 9.654497553631916,
+      "grad_norm": 0.17090503871440887,
+      "learning_rate": 5e-06,
+      "loss": 0.1065,
+      "step": 12826
+    },
+    {
+      "epoch": 9.65525028227324,
+      "grad_norm": 0.16273270547389984,
+      "learning_rate": 5e-06,
+      "loss": 0.1234,
+      "step": 12827
+    },
+    {
+      "epoch": 9.656003010914565,
+      "grad_norm": 0.19802404940128326,
+      "learning_rate": 5e-06,
+      "loss": 0.1376,
+      "step": 12828
+    },
+    {
+      "epoch": 9.65675573955589,
+      "grad_norm": 0.19213177263736725,
+      "learning_rate": 5e-06,
+      "loss": 0.1176,
+      "step": 12829
+    },
+    {
+      "epoch": 9.657508468197214,
+      "grad_norm": 0.17917035520076752,
+      "learning_rate": 5e-06,
+      "loss": 0.1252,
+      "step": 12830
+    },
+    {
+      "epoch": 9.65826119683854,
+      "grad_norm": 0.1816496104001999,
+      "learning_rate": 5e-06,
+      "loss": 0.0958,
+      "step": 12831
+    },
+    {
+      "epoch": 9.659013925479865,
+      "grad_norm": 0.19498799741268158,
+      "learning_rate": 5e-06,
+      "loss": 0.2239,
+      "step": 12832
+    },
+    {
+      "epoch": 9.659766654121189,
+      "grad_norm": 0.19271479547023773,
+      "learning_rate": 5e-06,
+      "loss": 0.1789,
+      "step": 12833
+    },
+    {
+      "epoch": 9.660519382762514,
+      "grad_norm": 0.2133963704109192,
+      "learning_rate": 5e-06,
+      "loss": 0.2282,
+      "step": 12834
+    },
+    {
+      "epoch": 9.66127211140384,
+      "grad_norm": 0.1579626500606537,
+      "learning_rate": 5e-06,
+      "loss": 0.1177,
+      "step": 12835
+    },
+    {
+      "epoch": 9.662024840045163,
+      "grad_norm": 0.1629030853509903,
+      "learning_rate": 5e-06,
+      "loss": 0.1878,
+      "step": 12836
+    },
+    {
+      "epoch": 9.662777568686488,
+      "grad_norm": 0.1796843260526657,
+      "learning_rate": 5e-06,
+      "loss": 0.137,
+      "step": 12837
+    },
+    {
+      "epoch": 9.663530297327814,
+      "grad_norm": 0.19710126519203186,
+      "learning_rate": 5e-06,
+      "loss": 0.1823,
+      "step": 12838
+    },
+    {
+      "epoch": 9.664283025969137,
+      "grad_norm": 0.1781919002532959,
+      "learning_rate": 5e-06,
+      "loss": 0.1283,
+      "step": 12839
+    },
+    {
+      "epoch": 9.665035754610463,
+      "grad_norm": 0.174128457903862,
+      "learning_rate": 5e-06,
+      "loss": 0.168,
+      "step": 12840
+    },
+    {
+      "epoch": 9.665788483251788,
+      "grad_norm": 0.19240052998065948,
+      "learning_rate": 5e-06,
+      "loss": 0.1216,
+      "step": 12841
+    },
+    {
+      "epoch": 9.666541211893113,
+      "grad_norm": 0.18607968091964722,
+      "learning_rate": 5e-06,
+      "loss": 0.1389,
+      "step": 12842
+    },
+    {
+      "epoch": 9.667293940534437,
+      "grad_norm": 0.16240906715393066,
+      "learning_rate": 5e-06,
+      "loss": 0.0972,
+      "step": 12843
+    },
+    {
+      "epoch": 9.668046669175762,
+      "grad_norm": 0.1668807864189148,
+      "learning_rate": 5e-06,
+      "loss": 0.1335,
+      "step": 12844
+    },
+    {
+      "epoch": 9.668799397817088,
+      "grad_norm": 0.17793434858322144,
+      "learning_rate": 5e-06,
+      "loss": 0.1185,
+      "step": 12845
+    },
+    {
+      "epoch": 9.669552126458411,
+      "grad_norm": 0.20403030514717102,
+      "learning_rate": 5e-06,
+      "loss": 0.1252,
+      "step": 12846
+    },
+    {
+      "epoch": 9.670304855099737,
+      "grad_norm": 0.18947359919548035,
+      "learning_rate": 5e-06,
+      "loss": 0.1682,
+      "step": 12847
+    },
+    {
+      "epoch": 9.671057583741062,
+      "grad_norm": 0.1648491770029068,
+      "learning_rate": 5e-06,
+      "loss": 0.138,
+      "step": 12848
+    },
+    {
+      "epoch": 9.671810312382386,
+      "grad_norm": 0.16282905638217926,
+      "learning_rate": 5e-06,
+      "loss": 0.1717,
+      "step": 12849
+    },
+    {
+      "epoch": 9.672563041023711,
+      "grad_norm": 0.17440618574619293,
+      "learning_rate": 5e-06,
+      "loss": 0.1391,
+      "step": 12850
+    },
+    {
+      "epoch": 9.673315769665036,
+      "grad_norm": 0.1800837218761444,
+      "learning_rate": 5e-06,
+      "loss": 0.1702,
+      "step": 12851
+    },
+    {
+      "epoch": 9.67406849830636,
+      "grad_norm": 0.17088447511196136,
+      "learning_rate": 5e-06,
+      "loss": 0.0926,
+      "step": 12852
+    },
+    {
+      "epoch": 9.674821226947685,
+      "grad_norm": 0.17384029924869537,
+      "learning_rate": 5e-06,
+      "loss": 0.0928,
+      "step": 12853
+    },
+    {
+      "epoch": 9.67557395558901,
+      "grad_norm": 0.17563927173614502,
+      "learning_rate": 5e-06,
+      "loss": 0.1428,
+      "step": 12854
+    },
+    {
+      "epoch": 9.676326684230334,
+      "grad_norm": 0.17575684189796448,
+      "learning_rate": 5e-06,
+      "loss": 0.1793,
+      "step": 12855
+    },
+    {
+      "epoch": 9.67707941287166,
+      "grad_norm": 0.18490754067897797,
+      "learning_rate": 5e-06,
+      "loss": 0.1777,
+      "step": 12856
+    },
+    {
+      "epoch": 9.677832141512985,
+      "grad_norm": 0.16546204686164856,
+      "learning_rate": 5e-06,
+      "loss": 0.1528,
+      "step": 12857
+    },
+    {
+      "epoch": 9.67858487015431,
+      "grad_norm": 0.17704583704471588,
+      "learning_rate": 5e-06,
+      "loss": 0.135,
+      "step": 12858
+    },
+    {
+      "epoch": 9.679337598795634,
+      "grad_norm": 0.16940315067768097,
+      "learning_rate": 5e-06,
+      "loss": 0.0909,
+      "step": 12859
+    },
+    {
+      "epoch": 9.68009032743696,
+      "grad_norm": 0.18852819502353668,
+      "learning_rate": 5e-06,
+      "loss": 0.1148,
+      "step": 12860
+    },
+    {
+      "epoch": 9.680843056078285,
+      "grad_norm": 0.1646193414926529,
+      "learning_rate": 5e-06,
+      "loss": 0.2911,
+      "step": 12861
+    },
+    {
+      "epoch": 9.681595784719608,
+      "grad_norm": 0.1575043797492981,
+      "learning_rate": 5e-06,
+      "loss": 0.0902,
+      "step": 12862
+    },
+    {
+      "epoch": 9.682348513360933,
+      "grad_norm": 0.18218286335468292,
+      "learning_rate": 5e-06,
+      "loss": 0.0973,
+      "step": 12863
+    },
+    {
+      "epoch": 9.683101242002259,
+      "grad_norm": 0.1924348622560501,
+      "learning_rate": 5e-06,
+      "loss": 0.2015,
+      "step": 12864
+    },
+    {
+      "epoch": 9.683853970643582,
+      "grad_norm": 0.16287124156951904,
+      "learning_rate": 5e-06,
+      "loss": 0.1563,
+      "step": 12865
+    },
+    {
+      "epoch": 9.684606699284908,
+      "grad_norm": 0.1640198677778244,
+      "learning_rate": 5e-06,
+      "loss": 0.1069,
+      "step": 12866
+    },
+    {
+      "epoch": 9.685359427926233,
+      "grad_norm": 0.17921936511993408,
+      "learning_rate": 5e-06,
+      "loss": 0.1437,
+      "step": 12867
+    },
+    {
+      "epoch": 9.686112156567557,
+      "grad_norm": 0.18912896513938904,
+      "learning_rate": 5e-06,
+      "loss": 0.1408,
+      "step": 12868
+    },
+    {
+      "epoch": 9.686864885208882,
+      "grad_norm": 0.1608870029449463,
+      "learning_rate": 5e-06,
+      "loss": 0.0813,
+      "step": 12869
+    },
+    {
+      "epoch": 9.687617613850207,
+      "grad_norm": 0.18341757357120514,
+      "learning_rate": 5e-06,
+      "loss": 0.1614,
+      "step": 12870
+    },
+    {
+      "epoch": 9.688370342491531,
+      "grad_norm": 0.15620313584804535,
+      "learning_rate": 5e-06,
+      "loss": 0.1431,
+      "step": 12871
+    },
+    {
+      "epoch": 9.689123071132856,
+      "grad_norm": 0.17755566537380219,
+      "learning_rate": 5e-06,
+      "loss": 0.1448,
+      "step": 12872
+    },
+    {
+      "epoch": 9.689875799774182,
+      "grad_norm": 0.17166562378406525,
+      "learning_rate": 5e-06,
+      "loss": 0.1735,
+      "step": 12873
+    },
+    {
+      "epoch": 9.690628528415505,
+      "grad_norm": 0.18860924243927002,
+      "learning_rate": 5e-06,
+      "loss": 0.1452,
+      "step": 12874
+    },
+    {
+      "epoch": 9.69138125705683,
+      "grad_norm": 0.19398002326488495,
+      "learning_rate": 5e-06,
+      "loss": 0.1346,
+      "step": 12875
+    },
+    {
+      "epoch": 9.692133985698156,
+      "grad_norm": 0.1810147613286972,
+      "learning_rate": 5e-06,
+      "loss": 0.1702,
+      "step": 12876
+    },
+    {
+      "epoch": 9.69288671433948,
+      "grad_norm": 0.1670236885547638,
+      "learning_rate": 5e-06,
+      "loss": 0.1385,
+      "step": 12877
+    },
+    {
+      "epoch": 9.693639442980805,
+      "grad_norm": 0.1740531176328659,
+      "learning_rate": 5e-06,
+      "loss": 0.0888,
+      "step": 12878
+    },
+    {
+      "epoch": 9.69439217162213,
+      "grad_norm": 0.14712101221084595,
+      "learning_rate": 5e-06,
+      "loss": 0.0909,
+      "step": 12879
+    },
+    {
+      "epoch": 9.695144900263456,
+      "grad_norm": 0.16453778743743896,
+      "learning_rate": 5e-06,
+      "loss": 0.0909,
+      "step": 12880
+    },
+    {
+      "epoch": 9.69589762890478,
+      "grad_norm": 0.1899816244840622,
+      "learning_rate": 5e-06,
+      "loss": 0.1724,
+      "step": 12881
+    },
+    {
+      "epoch": 9.696650357546105,
+      "grad_norm": 0.1744036227464676,
+      "learning_rate": 5e-06,
+      "loss": 0.1311,
+      "step": 12882
+    },
+    {
+      "epoch": 9.69740308618743,
+      "grad_norm": 0.16456253826618195,
+      "learning_rate": 5e-06,
+      "loss": 0.1548,
+      "step": 12883
+    },
+    {
+      "epoch": 9.698155814828754,
+      "grad_norm": 0.16617469489574432,
+      "learning_rate": 5e-06,
+      "loss": 0.1649,
+      "step": 12884
+    },
+    {
+      "epoch": 9.698908543470079,
+      "grad_norm": 0.16721615195274353,
+      "learning_rate": 5e-06,
+      "loss": 0.1077,
+      "step": 12885
+    },
+    {
+      "epoch": 9.699661272111404,
+      "grad_norm": 0.15114960074424744,
+      "learning_rate": 5e-06,
+      "loss": 0.1169,
+      "step": 12886
+    },
+    {
+      "epoch": 9.700414000752728,
+      "grad_norm": 0.20747165381908417,
+      "learning_rate": 5e-06,
+      "loss": 0.126,
+      "step": 12887
+    },
+    {
+      "epoch": 9.701166729394053,
+      "grad_norm": 0.18471115827560425,
+      "learning_rate": 5e-06,
+      "loss": 0.117,
+      "step": 12888
+    },
+    {
+      "epoch": 9.701919458035379,
+      "grad_norm": 0.15783466398715973,
+      "learning_rate": 5e-06,
+      "loss": 0.1573,
+      "step": 12889
+    },
+    {
+      "epoch": 9.702672186676702,
+      "grad_norm": 0.17182229459285736,
+      "learning_rate": 5e-06,
+      "loss": 0.1122,
+      "step": 12890
+    },
+    {
+      "epoch": 9.703424915318028,
+      "grad_norm": 0.18396730720996857,
+      "learning_rate": 5e-06,
+      "loss": 0.1311,
+      "step": 12891
+    },
+    {
+      "epoch": 9.704177643959353,
+      "grad_norm": 0.18857333064079285,
+      "learning_rate": 5e-06,
+      "loss": 0.1324,
+      "step": 12892
+    },
+    {
+      "epoch": 9.704930372600678,
+      "grad_norm": 0.17848534882068634,
+      "learning_rate": 5e-06,
+      "loss": 0.1021,
+      "step": 12893
+    },
+    {
+      "epoch": 9.705683101242002,
+      "grad_norm": 0.17953473329544067,
+      "learning_rate": 5e-06,
+      "loss": 0.1493,
+      "step": 12894
+    },
+    {
+      "epoch": 9.706435829883327,
+      "grad_norm": 0.17229846119880676,
+      "learning_rate": 5e-06,
+      "loss": 0.1533,
+      "step": 12895
+    },
+    {
+      "epoch": 9.707188558524653,
+      "grad_norm": 0.17079812288284302,
+      "learning_rate": 5e-06,
+      "loss": 0.1553,
+      "step": 12896
+    },
+    {
+      "epoch": 9.707941287165976,
+      "grad_norm": 0.18614287674427032,
+      "learning_rate": 5e-06,
+      "loss": 0.1938,
+      "step": 12897
+    },
+    {
+      "epoch": 9.708694015807302,
+      "grad_norm": 0.17835387587547302,
+      "learning_rate": 5e-06,
+      "loss": 0.1835,
+      "step": 12898
+    },
+    {
+      "epoch": 9.709446744448627,
+      "grad_norm": 0.1816953420639038,
+      "learning_rate": 5e-06,
+      "loss": 0.0898,
+      "step": 12899
+    },
+    {
+      "epoch": 9.71019947308995,
+      "grad_norm": 0.17752660810947418,
+      "learning_rate": 5e-06,
+      "loss": 0.1353,
+      "step": 12900
+    },
+    {
+      "epoch": 9.710952201731276,
+      "grad_norm": 0.1727837324142456,
+      "learning_rate": 5e-06,
+      "loss": 0.0882,
+      "step": 12901
+    },
+    {
+      "epoch": 9.711704930372601,
+      "grad_norm": 0.17249111831188202,
+      "learning_rate": 5e-06,
+      "loss": 0.1377,
+      "step": 12902
+    },
+    {
+      "epoch": 9.712457659013925,
+      "grad_norm": 0.17353224754333496,
+      "learning_rate": 5e-06,
+      "loss": 0.2532,
+      "step": 12903
+    },
+    {
+      "epoch": 9.71321038765525,
+      "grad_norm": 0.17165154218673706,
+      "learning_rate": 5e-06,
+      "loss": 0.1209,
+      "step": 12904
+    },
+    {
+      "epoch": 9.713963116296576,
+      "grad_norm": 0.16794277727603912,
+      "learning_rate": 5e-06,
+      "loss": 0.11,
+      "step": 12905
+    },
+    {
+      "epoch": 9.7147158449379,
+      "grad_norm": 0.17770390212535858,
+      "learning_rate": 5e-06,
+      "loss": 0.1046,
+      "step": 12906
+    },
+    {
+      "epoch": 9.715468573579225,
+      "grad_norm": 0.20371058583259583,
+      "learning_rate": 5e-06,
+      "loss": 0.2224,
+      "step": 12907
+    },
+    {
+      "epoch": 9.71622130222055,
+      "grad_norm": 0.18293359875679016,
+      "learning_rate": 5e-06,
+      "loss": 0.1488,
+      "step": 12908
+    },
+    {
+      "epoch": 9.716974030861874,
+      "grad_norm": 0.15916109085083008,
+      "learning_rate": 5e-06,
+      "loss": 0.0672,
+      "step": 12909
+    },
+    {
+      "epoch": 9.717726759503199,
+      "grad_norm": 0.18277142941951752,
+      "learning_rate": 5e-06,
+      "loss": 0.1156,
+      "step": 12910
+    },
+    {
+      "epoch": 9.718479488144524,
+      "grad_norm": 0.20804579555988312,
+      "learning_rate": 5e-06,
+      "loss": 0.1236,
+      "step": 12911
+    },
+    {
+      "epoch": 9.719232216785848,
+      "grad_norm": 0.18047448992729187,
+      "learning_rate": 5e-06,
+      "loss": 0.1558,
+      "step": 12912
+    },
+    {
+      "epoch": 9.719984945427173,
+      "grad_norm": 0.18360650539398193,
+      "learning_rate": 5e-06,
+      "loss": 0.1884,
+      "step": 12913
+    },
+    {
+      "epoch": 9.720737674068499,
+      "grad_norm": 0.1967216283082962,
+      "learning_rate": 5e-06,
+      "loss": 0.1991,
+      "step": 12914
+    },
+    {
+      "epoch": 9.721490402709824,
+      "grad_norm": 0.1783003807067871,
+      "learning_rate": 5e-06,
+      "loss": 0.1881,
+      "step": 12915
+    },
+    {
+      "epoch": 9.722243131351147,
+      "grad_norm": 0.15685400366783142,
+      "learning_rate": 5e-06,
+      "loss": 0.1045,
+      "step": 12916
+    },
+    {
+      "epoch": 9.722995859992473,
+      "grad_norm": 0.18301057815551758,
+      "learning_rate": 5e-06,
+      "loss": 0.1231,
+      "step": 12917
+    },
+    {
+      "epoch": 9.723748588633798,
+      "grad_norm": 0.18617641925811768,
+      "learning_rate": 5e-06,
+      "loss": 0.1593,
+      "step": 12918
+    },
+    {
+      "epoch": 9.724501317275122,
+      "grad_norm": 0.16652098298072815,
+      "learning_rate": 5e-06,
+      "loss": 0.1401,
+      "step": 12919
+    },
+    {
+      "epoch": 9.725254045916447,
+      "grad_norm": 0.18909236788749695,
+      "learning_rate": 5e-06,
+      "loss": 0.1112,
+      "step": 12920
+    },
+    {
+      "epoch": 9.726006774557773,
+      "grad_norm": 0.19311127066612244,
+      "learning_rate": 5e-06,
+      "loss": 0.1685,
+      "step": 12921
+    },
+    {
+      "epoch": 9.726759503199096,
+      "grad_norm": 0.20133988559246063,
+      "learning_rate": 5e-06,
+      "loss": 0.1743,
+      "step": 12922
+    },
+    {
+      "epoch": 9.727512231840421,
+      "grad_norm": 0.18554478883743286,
+      "learning_rate": 5e-06,
+      "loss": 0.1586,
+      "step": 12923
+    },
+    {
+      "epoch": 9.728264960481747,
+      "grad_norm": 0.2116854041814804,
+      "learning_rate": 5e-06,
+      "loss": 0.1717,
+      "step": 12924
+    },
+    {
+      "epoch": 9.72901768912307,
+      "grad_norm": 0.185796320438385,
+      "learning_rate": 5e-06,
+      "loss": 0.1291,
+      "step": 12925
+    },
+    {
+      "epoch": 9.729770417764396,
+      "grad_norm": 0.16991423070430756,
+      "learning_rate": 5e-06,
+      "loss": 0.0969,
+      "step": 12926
+    },
+    {
+      "epoch": 9.730523146405721,
+      "grad_norm": 0.17379307746887207,
+      "learning_rate": 5e-06,
+      "loss": 0.1779,
+      "step": 12927
+    },
+    {
+      "epoch": 9.731275875047045,
+      "grad_norm": 0.16067840158939362,
+      "learning_rate": 5e-06,
+      "loss": 0.1317,
+      "step": 12928
+    },
+    {
+      "epoch": 9.73202860368837,
+      "grad_norm": 0.213572695851326,
+      "learning_rate": 5e-06,
+      "loss": 0.1722,
+      "step": 12929
+    },
+    {
+      "epoch": 9.732781332329695,
+      "grad_norm": 0.17275460064411163,
+      "learning_rate": 5e-06,
+      "loss": 0.1416,
+      "step": 12930
+    },
+    {
+      "epoch": 9.73353406097102,
+      "grad_norm": 0.1780078411102295,
+      "learning_rate": 5e-06,
+      "loss": 0.1794,
+      "step": 12931
+    },
+    {
+      "epoch": 9.734286789612344,
+      "grad_norm": 0.18421430885791779,
+      "learning_rate": 5e-06,
+      "loss": 0.1344,
+      "step": 12932
+    },
+    {
+      "epoch": 9.73503951825367,
+      "grad_norm": 0.22255919873714447,
+      "learning_rate": 5e-06,
+      "loss": 0.1833,
+      "step": 12933
+    },
+    {
+      "epoch": 9.735792246894995,
+      "grad_norm": 0.1610349714756012,
+      "learning_rate": 5e-06,
+      "loss": 0.1747,
+      "step": 12934
+    },
+    {
+      "epoch": 9.736544975536319,
+      "grad_norm": 0.1775798499584198,
+      "learning_rate": 5e-06,
+      "loss": 0.2186,
+      "step": 12935
+    },
+    {
+      "epoch": 9.737297704177644,
+      "grad_norm": 0.1905764937400818,
+      "learning_rate": 5e-06,
+      "loss": 0.21,
+      "step": 12936
+    },
+    {
+      "epoch": 9.73805043281897,
+      "grad_norm": 0.194075345993042,
+      "learning_rate": 5e-06,
+      "loss": 0.1999,
+      "step": 12937
+    },
+    {
+      "epoch": 9.738803161460293,
+      "grad_norm": 0.1700761616230011,
+      "learning_rate": 5e-06,
+      "loss": 0.1566,
+      "step": 12938
+    },
+    {
+      "epoch": 9.739555890101618,
+      "grad_norm": 0.1686561405658722,
+      "learning_rate": 5e-06,
+      "loss": 0.1213,
+      "step": 12939
+    },
+    {
+      "epoch": 9.740308618742944,
+      "grad_norm": 0.1802045851945877,
+      "learning_rate": 5e-06,
+      "loss": 0.1622,
+      "step": 12940
+    },
+    {
+      "epoch": 9.741061347384267,
+      "grad_norm": 0.16900691390037537,
+      "learning_rate": 5e-06,
+      "loss": 0.1295,
+      "step": 12941
+    },
+    {
+      "epoch": 9.741814076025593,
+      "grad_norm": 0.19935362040996552,
+      "learning_rate": 5e-06,
+      "loss": 0.1695,
+      "step": 12942
+    },
+    {
+      "epoch": 9.742566804666918,
+      "grad_norm": 0.2047579288482666,
+      "learning_rate": 5e-06,
+      "loss": 0.1586,
+      "step": 12943
+    },
+    {
+      "epoch": 9.743319533308242,
+      "grad_norm": 0.1820802241563797,
+      "learning_rate": 5e-06,
+      "loss": 0.133,
+      "step": 12944
+    },
+    {
+      "epoch": 9.744072261949567,
+      "grad_norm": 0.17195993661880493,
+      "learning_rate": 5e-06,
+      "loss": 0.1534,
+      "step": 12945
+    },
+    {
+      "epoch": 9.744824990590892,
+      "grad_norm": 0.18094001710414886,
+      "learning_rate": 5e-06,
+      "loss": 0.1154,
+      "step": 12946
+    },
+    {
+      "epoch": 9.745577719232216,
+      "grad_norm": 0.19106519222259521,
+      "learning_rate": 5e-06,
+      "loss": 0.0855,
+      "step": 12947
+    },
+    {
+      "epoch": 9.746330447873541,
+      "grad_norm": 0.1974828690290451,
+      "learning_rate": 5e-06,
+      "loss": 0.1539,
+      "step": 12948
+    },
+    {
+      "epoch": 9.747083176514867,
+      "grad_norm": 0.1865929216146469,
+      "learning_rate": 5e-06,
+      "loss": 0.1407,
+      "step": 12949
+    },
+    {
+      "epoch": 9.74783590515619,
+      "grad_norm": 0.18630895018577576,
+      "learning_rate": 5e-06,
+      "loss": 0.1325,
+      "step": 12950
+    },
+    {
+      "epoch": 9.748588633797516,
+      "grad_norm": 0.16390369832515717,
+      "learning_rate": 5e-06,
+      "loss": 0.1289,
+      "step": 12951
+    },
+    {
+      "epoch": 9.749341362438841,
+      "grad_norm": 0.17925423383712769,
+      "learning_rate": 5e-06,
+      "loss": 0.147,
+      "step": 12952
+    },
+    {
+      "epoch": 9.750094091080166,
+      "grad_norm": 0.1612803339958191,
+      "learning_rate": 5e-06,
+      "loss": 0.0974,
+      "step": 12953
+    },
+    {
+      "epoch": 9.75084681972149,
+      "grad_norm": 0.15816466510295868,
+      "learning_rate": 5e-06,
+      "loss": 0.0773,
+      "step": 12954
+    },
+    {
+      "epoch": 9.751599548362815,
+      "grad_norm": 0.19712874293327332,
+      "learning_rate": 5e-06,
+      "loss": 0.2254,
+      "step": 12955
+    },
+    {
+      "epoch": 9.75235227700414,
+      "grad_norm": 0.18403805792331696,
+      "learning_rate": 5e-06,
+      "loss": 0.1843,
+      "step": 12956
+    },
+    {
+      "epoch": 9.753105005645464,
+      "grad_norm": 0.19091729819774628,
+      "learning_rate": 5e-06,
+      "loss": 0.1943,
+      "step": 12957
+    },
+    {
+      "epoch": 9.75385773428679,
+      "grad_norm": 0.17691680788993835,
+      "learning_rate": 5e-06,
+      "loss": 0.1616,
+      "step": 12958
+    },
+    {
+      "epoch": 9.754610462928115,
+      "grad_norm": 0.16509433090686798,
+      "learning_rate": 5e-06,
+      "loss": 0.0835,
+      "step": 12959
+    },
+    {
+      "epoch": 9.755363191569439,
+      "grad_norm": 0.1876877248287201,
+      "learning_rate": 5e-06,
+      "loss": 0.1585,
+      "step": 12960
+    },
+    {
+      "epoch": 9.756115920210764,
+      "grad_norm": 0.1822490096092224,
+      "learning_rate": 5e-06,
+      "loss": 0.0618,
+      "step": 12961
+    },
+    {
+      "epoch": 9.75686864885209,
+      "grad_norm": 0.19136658310890198,
+      "learning_rate": 5e-06,
+      "loss": 0.1425,
+      "step": 12962
+    },
+    {
+      "epoch": 9.757621377493413,
+      "grad_norm": 0.16651438176631927,
+      "learning_rate": 5e-06,
+      "loss": 0.1121,
+      "step": 12963
+    },
+    {
+      "epoch": 9.758374106134738,
+      "grad_norm": 0.14649762213230133,
+      "learning_rate": 5e-06,
+      "loss": 0.1025,
+      "step": 12964
+    },
+    {
+      "epoch": 9.759126834776064,
+      "grad_norm": 0.16370604932308197,
+      "learning_rate": 5e-06,
+      "loss": 0.1336,
+      "step": 12965
+    },
+    {
+      "epoch": 9.759879563417389,
+      "grad_norm": 0.1756216138601303,
+      "learning_rate": 5e-06,
+      "loss": 0.1317,
+      "step": 12966
+    },
+    {
+      "epoch": 9.760632292058713,
+      "grad_norm": 0.1910608857870102,
+      "learning_rate": 5e-06,
+      "loss": 0.1999,
+      "step": 12967
+    },
+    {
+      "epoch": 9.761385020700038,
+      "grad_norm": 0.1705760359764099,
+      "learning_rate": 5e-06,
+      "loss": 0.1807,
+      "step": 12968
+    },
+    {
+      "epoch": 9.762137749341363,
+      "grad_norm": 0.15822021663188934,
+      "learning_rate": 5e-06,
+      "loss": 0.1056,
+      "step": 12969
+    },
+    {
+      "epoch": 9.762890477982687,
+      "grad_norm": 0.18654118478298187,
+      "learning_rate": 5e-06,
+      "loss": 0.196,
+      "step": 12970
+    },
+    {
+      "epoch": 9.763643206624012,
+      "grad_norm": 0.18175870180130005,
+      "learning_rate": 5e-06,
+      "loss": 0.1055,
+      "step": 12971
+    },
+    {
+      "epoch": 9.764395935265338,
+      "grad_norm": 0.17290668189525604,
+      "learning_rate": 5e-06,
+      "loss": 0.2208,
+      "step": 12972
+    },
+    {
+      "epoch": 9.765148663906661,
+      "grad_norm": 0.18000903725624084,
+      "learning_rate": 5e-06,
+      "loss": 0.1322,
+      "step": 12973
+    },
+    {
+      "epoch": 9.765901392547987,
+      "grad_norm": 0.17735563218593597,
+      "learning_rate": 5e-06,
+      "loss": 0.102,
+      "step": 12974
+    },
+    {
+      "epoch": 9.766654121189312,
+      "grad_norm": 0.1905563473701477,
+      "learning_rate": 5e-06,
+      "loss": 0.167,
+      "step": 12975
+    },
+    {
+      "epoch": 9.767406849830635,
+      "grad_norm": 0.1574086844921112,
+      "learning_rate": 5e-06,
+      "loss": 0.1244,
+      "step": 12976
+    },
+    {
+      "epoch": 9.76815957847196,
+      "grad_norm": 0.15428391098976135,
+      "learning_rate": 5e-06,
+      "loss": 0.1057,
+      "step": 12977
+    },
+    {
+      "epoch": 9.768912307113286,
+      "grad_norm": 0.19069026410579681,
+      "learning_rate": 5e-06,
+      "loss": 0.1559,
+      "step": 12978
+    },
+    {
+      "epoch": 9.76966503575461,
+      "grad_norm": 0.191801056265831,
+      "learning_rate": 5e-06,
+      "loss": 0.1615,
+      "step": 12979
+    },
+    {
+      "epoch": 9.770417764395935,
+      "grad_norm": 0.19298721849918365,
+      "learning_rate": 5e-06,
+      "loss": 0.1897,
+      "step": 12980
+    },
+    {
+      "epoch": 9.77117049303726,
+      "grad_norm": 0.1972801685333252,
+      "learning_rate": 5e-06,
+      "loss": 0.1582,
+      "step": 12981
+    },
+    {
+      "epoch": 9.771923221678584,
+      "grad_norm": 0.17947833240032196,
+      "learning_rate": 5e-06,
+      "loss": 0.1636,
+      "step": 12982
+    },
+    {
+      "epoch": 9.77267595031991,
+      "grad_norm": 0.1839820146560669,
+      "learning_rate": 5e-06,
+      "loss": 0.1964,
+      "step": 12983
+    },
+    {
+      "epoch": 9.773428678961235,
+      "grad_norm": 0.1689162701368332,
+      "learning_rate": 5e-06,
+      "loss": 0.1437,
+      "step": 12984
+    },
+    {
+      "epoch": 9.774181407602558,
+      "grad_norm": 0.18555085361003876,
+      "learning_rate": 5e-06,
+      "loss": 0.1467,
+      "step": 12985
+    },
+    {
+      "epoch": 9.774934136243884,
+      "grad_norm": 0.17190003395080566,
+      "learning_rate": 5e-06,
+      "loss": 0.142,
+      "step": 12986
+    },
+    {
+      "epoch": 9.775686864885209,
+      "grad_norm": 0.18584580719470978,
+      "learning_rate": 5e-06,
+      "loss": 0.1886,
+      "step": 12987
+    },
+    {
+      "epoch": 9.776439593526534,
+      "grad_norm": 0.16829974949359894,
+      "learning_rate": 5e-06,
+      "loss": 0.1194,
+      "step": 12988
+    },
+    {
+      "epoch": 9.777192322167858,
+      "grad_norm": 0.1519288867712021,
+      "learning_rate": 5e-06,
+      "loss": 0.1318,
+      "step": 12989
+    },
+    {
+      "epoch": 9.777945050809183,
+      "grad_norm": 0.1516488492488861,
+      "learning_rate": 5e-06,
+      "loss": 0.0584,
+      "step": 12990
+    },
+    {
+      "epoch": 9.778697779450509,
+      "grad_norm": 0.16280300915241241,
+      "learning_rate": 5e-06,
+      "loss": 0.1445,
+      "step": 12991
+    },
+    {
+      "epoch": 9.779450508091832,
+      "grad_norm": 0.19075460731983185,
+      "learning_rate": 5e-06,
+      "loss": 0.1945,
+      "step": 12992
+    },
+    {
+      "epoch": 9.780203236733158,
+      "grad_norm": 0.18143600225448608,
+      "learning_rate": 5e-06,
+      "loss": 0.1278,
+      "step": 12993
+    },
+    {
+      "epoch": 9.780955965374483,
+      "grad_norm": 0.17078472673892975,
+      "learning_rate": 5e-06,
+      "loss": 0.1186,
+      "step": 12994
+    },
+    {
+      "epoch": 9.781708694015807,
+      "grad_norm": 0.1724635362625122,
+      "learning_rate": 5e-06,
+      "loss": 0.0938,
+      "step": 12995
+    },
+    {
+      "epoch": 9.782461422657132,
+      "grad_norm": 0.1678427904844284,
+      "learning_rate": 5e-06,
+      "loss": 0.094,
+      "step": 12996
+    },
+    {
+      "epoch": 9.783214151298457,
+      "grad_norm": 0.18439151346683502,
+      "learning_rate": 5e-06,
+      "loss": 0.1646,
+      "step": 12997
+    },
+    {
+      "epoch": 9.783966879939781,
+      "grad_norm": 0.14453095197677612,
+      "learning_rate": 5e-06,
+      "loss": 0.0984,
+      "step": 12998
+    },
+    {
+      "epoch": 9.784719608581106,
+      "grad_norm": 0.14745287597179413,
+      "learning_rate": 5e-06,
+      "loss": 0.0945,
+      "step": 12999
+    },
+    {
+      "epoch": 9.785472337222432,
+      "grad_norm": 0.18517597019672394,
+      "learning_rate": 5e-06,
+      "loss": 0.153,
+      "step": 13000
+    },
+    {
+      "epoch": 9.785472337222432,
+      "eval_loss": 0.19013871252536774,
+      "eval_runtime": 453.0011,
+      "eval_samples_per_second": 21.252,
+      "eval_steps_per_second": 0.664,
+      "step": 13000
+    },
+    {
+      "epoch": 9.786225065863755,
+      "grad_norm": 0.18439596891403198,
+      "learning_rate": 5e-06,
+      "loss": 0.1363,
+      "step": 13001
+    },
+    {
+      "epoch": 9.78697779450508,
+      "grad_norm": 0.19769099354743958,
+      "learning_rate": 5e-06,
+      "loss": 0.1441,
+      "step": 13002
+    },
+    {
+      "epoch": 9.787730523146406,
+      "grad_norm": 0.178147554397583,
+      "learning_rate": 5e-06,
+      "loss": 0.1637,
+      "step": 13003
+    },
+    {
+      "epoch": 9.788483251787731,
+      "grad_norm": 0.16063354909420013,
+      "learning_rate": 5e-06,
+      "loss": 0.0951,
+      "step": 13004
+    },
+    {
+      "epoch": 9.789235980429055,
+      "grad_norm": 0.20681080222129822,
+      "learning_rate": 5e-06,
+      "loss": 0.1585,
+      "step": 13005
+    },
+    {
+      "epoch": 9.78998870907038,
+      "grad_norm": 0.1950402706861496,
+      "learning_rate": 5e-06,
+      "loss": 0.1198,
+      "step": 13006
+    },
+    {
+      "epoch": 9.790741437711706,
+      "grad_norm": 0.15427392721176147,
+      "learning_rate": 5e-06,
+      "loss": 0.0755,
+      "step": 13007
+    },
+    {
+      "epoch": 9.79149416635303,
+      "grad_norm": 0.1806425303220749,
+      "learning_rate": 5e-06,
+      "loss": 0.1229,
+      "step": 13008
+    },
+    {
+      "epoch": 9.792246894994355,
+      "grad_norm": 0.18484003841876984,
+      "learning_rate": 5e-06,
+      "loss": 0.2022,
+      "step": 13009
+    },
+    {
+      "epoch": 9.79299962363568,
+      "grad_norm": 0.21827439963817596,
+      "learning_rate": 5e-06,
+      "loss": 0.1894,
+      "step": 13010
+    },
+    {
+      "epoch": 9.793752352277004,
+      "grad_norm": 0.16445790231227875,
+      "learning_rate": 5e-06,
+      "loss": 0.1481,
+      "step": 13011
+    },
+    {
+      "epoch": 9.794505080918329,
+      "grad_norm": 0.1890716254711151,
+      "learning_rate": 5e-06,
+      "loss": 0.1544,
+      "step": 13012
+    },
+    {
+      "epoch": 9.795257809559654,
+      "grad_norm": 0.1625918298959732,
+      "learning_rate": 5e-06,
+      "loss": 0.1283,
+      "step": 13013
+    },
+    {
+      "epoch": 9.796010538200978,
+      "grad_norm": 0.17508673667907715,
+      "learning_rate": 5e-06,
+      "loss": 0.1282,
+      "step": 13014
+    },
+    {
+      "epoch": 9.796763266842303,
+      "grad_norm": 0.1597030907869339,
+      "learning_rate": 5e-06,
+      "loss": 0.1073,
+      "step": 13015
+    },
+    {
+      "epoch": 9.797515995483629,
+      "grad_norm": 0.15383581817150116,
+      "learning_rate": 5e-06,
+      "loss": 0.1263,
+      "step": 13016
+    },
+    {
+      "epoch": 9.798268724124952,
+      "grad_norm": 0.1812647581100464,
+      "learning_rate": 5e-06,
+      "loss": 0.1785,
+      "step": 13017
+    },
+    {
+      "epoch": 9.799021452766278,
+      "grad_norm": 0.1940285563468933,
+      "learning_rate": 5e-06,
+      "loss": 0.2208,
+      "step": 13018
+    },
+    {
+      "epoch": 9.799774181407603,
+      "grad_norm": 0.18803390860557556,
+      "learning_rate": 5e-06,
+      "loss": 0.1431,
+      "step": 13019
+    },
+    {
+      "epoch": 9.800526910048927,
+      "grad_norm": 0.18864025175571442,
+      "learning_rate": 5e-06,
+      "loss": 0.1707,
+      "step": 13020
+    },
+    {
+      "epoch": 9.801279638690252,
+      "grad_norm": 0.17000697553157806,
+      "learning_rate": 5e-06,
+      "loss": 0.1632,
+      "step": 13021
+    },
+    {
+      "epoch": 9.802032367331577,
+      "grad_norm": 0.17463241517543793,
+      "learning_rate": 5e-06,
+      "loss": 0.1797,
+      "step": 13022
+    },
+    {
+      "epoch": 9.802785095972903,
+      "grad_norm": 0.18049584329128265,
+      "learning_rate": 5e-06,
+      "loss": 0.1927,
+      "step": 13023
+    },
+    {
+      "epoch": 9.803537824614226,
+      "grad_norm": 0.19072973728179932,
+      "learning_rate": 5e-06,
+      "loss": 0.1891,
+      "step": 13024
+    },
+    {
+      "epoch": 9.804290553255552,
+      "grad_norm": 0.18648578226566315,
+      "learning_rate": 5e-06,
+      "loss": 0.2147,
+      "step": 13025
+    },
+    {
+      "epoch": 9.805043281896877,
+      "grad_norm": 0.16708241403102875,
+      "learning_rate": 5e-06,
+      "loss": 0.1593,
+      "step": 13026
+    },
+    {
+      "epoch": 9.8057960105382,
+      "grad_norm": 0.1715615838766098,
+      "learning_rate": 5e-06,
+      "loss": 0.1185,
+      "step": 13027
+    },
+    {
+      "epoch": 9.806548739179526,
+      "grad_norm": 0.18442568182945251,
+      "learning_rate": 5e-06,
+      "loss": 0.1493,
+      "step": 13028
+    },
+    {
+      "epoch": 9.807301467820851,
+      "grad_norm": 0.20089299976825714,
+      "learning_rate": 5e-06,
+      "loss": 0.2006,
+      "step": 13029
+    },
+    {
+      "epoch": 9.808054196462175,
+      "grad_norm": 0.18451951444149017,
+      "learning_rate": 5e-06,
+      "loss": 0.1686,
+      "step": 13030
+    },
+    {
+      "epoch": 9.8088069251035,
+      "grad_norm": 0.1885918527841568,
+      "learning_rate": 5e-06,
+      "loss": 0.2261,
+      "step": 13031
+    },
+    {
+      "epoch": 9.809559653744826,
+      "grad_norm": 0.18974648416042328,
+      "learning_rate": 5e-06,
+      "loss": 0.2662,
+      "step": 13032
+    },
+    {
+      "epoch": 9.810312382386149,
+      "grad_norm": 0.18265730142593384,
+      "learning_rate": 5e-06,
+      "loss": 0.1821,
+      "step": 13033
+    },
+    {
+      "epoch": 9.811065111027474,
+      "grad_norm": 0.18647035956382751,
+      "learning_rate": 5e-06,
+      "loss": 0.1419,
+      "step": 13034
+    },
+    {
+      "epoch": 9.8118178396688,
+      "grad_norm": 0.16562069952487946,
+      "learning_rate": 5e-06,
+      "loss": 0.1643,
+      "step": 13035
+    },
+    {
+      "epoch": 9.812570568310123,
+      "grad_norm": 0.17054402828216553,
+      "learning_rate": 5e-06,
+      "loss": 0.178,
+      "step": 13036
+    },
+    {
+      "epoch": 9.813323296951449,
+      "grad_norm": 0.17453713715076447,
+      "learning_rate": 5e-06,
+      "loss": 0.1597,
+      "step": 13037
+    },
+    {
+      "epoch": 9.814076025592774,
+      "grad_norm": 0.1853579878807068,
+      "learning_rate": 5e-06,
+      "loss": 0.1223,
+      "step": 13038
+    },
+    {
+      "epoch": 9.8148287542341,
+      "grad_norm": 0.203618124127388,
+      "learning_rate": 5e-06,
+      "loss": 0.1919,
+      "step": 13039
+    },
+    {
+      "epoch": 9.815581482875423,
+      "grad_norm": 0.1798381805419922,
+      "learning_rate": 5e-06,
+      "loss": 0.173,
+      "step": 13040
+    },
+    {
+      "epoch": 9.816334211516748,
+      "grad_norm": 0.15775558352470398,
+      "learning_rate": 5e-06,
+      "loss": 0.1257,
+      "step": 13041
+    },
+    {
+      "epoch": 9.817086940158074,
+      "grad_norm": 0.17074400186538696,
+      "learning_rate": 5e-06,
+      "loss": 0.1045,
+      "step": 13042
+    },
+    {
+      "epoch": 9.817839668799397,
+      "grad_norm": 0.17092691361904144,
+      "learning_rate": 5e-06,
+      "loss": 0.0973,
+      "step": 13043
+    },
+    {
+      "epoch": 9.818592397440723,
+      "grad_norm": 0.20677627623081207,
+      "learning_rate": 5e-06,
+      "loss": 0.2343,
+      "step": 13044
+    },
+    {
+      "epoch": 9.819345126082048,
+      "grad_norm": 0.18550391495227814,
+      "learning_rate": 5e-06,
+      "loss": 0.1596,
+      "step": 13045
+    },
+    {
+      "epoch": 9.820097854723372,
+      "grad_norm": 0.19329960644245148,
+      "learning_rate": 5e-06,
+      "loss": 0.1472,
+      "step": 13046
+    },
+    {
+      "epoch": 9.820850583364697,
+      "grad_norm": 0.1760677695274353,
+      "learning_rate": 5e-06,
+      "loss": 0.1712,
+      "step": 13047
+    },
+    {
+      "epoch": 9.821603312006022,
+      "grad_norm": 0.20074446499347687,
+      "learning_rate": 5e-06,
+      "loss": 0.1687,
+      "step": 13048
+    },
+    {
+      "epoch": 9.822356040647346,
+      "grad_norm": 0.19178897142410278,
+      "learning_rate": 5e-06,
+      "loss": 0.1086,
+      "step": 13049
+    },
+    {
+      "epoch": 9.823108769288671,
+      "grad_norm": 0.19814960658550262,
+      "learning_rate": 5e-06,
+      "loss": 0.2246,
+      "step": 13050
+    },
+    {
+      "epoch": 9.823861497929997,
+      "grad_norm": 0.17061062157154083,
+      "learning_rate": 5e-06,
+      "loss": 0.0996,
+      "step": 13051
+    },
+    {
+      "epoch": 9.82461422657132,
+      "grad_norm": 0.17724579572677612,
+      "learning_rate": 5e-06,
+      "loss": 0.1219,
+      "step": 13052
+    },
+    {
+      "epoch": 9.825366955212646,
+      "grad_norm": 0.1818351149559021,
+      "learning_rate": 5e-06,
+      "loss": 0.2791,
+      "step": 13053
+    },
+    {
+      "epoch": 9.826119683853971,
+      "grad_norm": 0.1774625927209854,
+      "learning_rate": 5e-06,
+      "loss": 0.1427,
+      "step": 13054
+    },
+    {
+      "epoch": 9.826872412495295,
+      "grad_norm": 0.1839015632867813,
+      "learning_rate": 5e-06,
+      "loss": 0.1788,
+      "step": 13055
+    },
+    {
+      "epoch": 9.82762514113662,
+      "grad_norm": 0.16980767250061035,
+      "learning_rate": 5e-06,
+      "loss": 0.147,
+      "step": 13056
+    },
+    {
+      "epoch": 9.828377869777945,
+      "grad_norm": 0.18469330668449402,
+      "learning_rate": 5e-06,
+      "loss": 0.1622,
+      "step": 13057
+    },
+    {
+      "epoch": 9.829130598419269,
+      "grad_norm": 0.18140670657157898,
+      "learning_rate": 5e-06,
+      "loss": 0.1108,
+      "step": 13058
+    },
+    {
+      "epoch": 9.829883327060594,
+      "grad_norm": 0.220688596367836,
+      "learning_rate": 5e-06,
+      "loss": 0.1757,
+      "step": 13059
+    },
+    {
+      "epoch": 9.83063605570192,
+      "grad_norm": 0.1553303301334381,
+      "learning_rate": 5e-06,
+      "loss": 0.1332,
+      "step": 13060
+    },
+    {
+      "epoch": 9.831388784343245,
+      "grad_norm": 0.17457233369350433,
+      "learning_rate": 5e-06,
+      "loss": 0.1443,
+      "step": 13061
+    },
+    {
+      "epoch": 9.832141512984569,
+      "grad_norm": 0.16880889236927032,
+      "learning_rate": 5e-06,
+      "loss": 0.1535,
+      "step": 13062
+    },
+    {
+      "epoch": 9.832894241625894,
+      "grad_norm": 0.19132821261882782,
+      "learning_rate": 5e-06,
+      "loss": 0.096,
+      "step": 13063
+    },
+    {
+      "epoch": 9.83364697026722,
+      "grad_norm": 0.20524577796459198,
+      "learning_rate": 5e-06,
+      "loss": 0.1494,
+      "step": 13064
+    },
+    {
+      "epoch": 9.834399698908543,
+      "grad_norm": 0.2080898880958557,
+      "learning_rate": 5e-06,
+      "loss": 0.1849,
+      "step": 13065
+    },
+    {
+      "epoch": 9.835152427549868,
+      "grad_norm": 0.19964265823364258,
+      "learning_rate": 5e-06,
+      "loss": 0.1935,
+      "step": 13066
+    },
+    {
+      "epoch": 9.835905156191194,
+      "grad_norm": 0.19685962796211243,
+      "learning_rate": 5e-06,
+      "loss": 0.215,
+      "step": 13067
+    },
+    {
+      "epoch": 9.836657884832517,
+      "grad_norm": 0.2146751433610916,
+      "learning_rate": 5e-06,
+      "loss": 0.1036,
+      "step": 13068
+    },
+    {
+      "epoch": 9.837410613473843,
+      "grad_norm": 0.1963040679693222,
+      "learning_rate": 5e-06,
+      "loss": 0.1005,
+      "step": 13069
+    },
+    {
+      "epoch": 9.838163342115168,
+      "grad_norm": 0.1472076028585434,
+      "learning_rate": 5e-06,
+      "loss": 0.1592,
+      "step": 13070
+    },
+    {
+      "epoch": 9.838916070756492,
+      "grad_norm": 0.1930358111858368,
+      "learning_rate": 5e-06,
+      "loss": 0.1262,
+      "step": 13071
+    },
+    {
+      "epoch": 9.839668799397817,
+      "grad_norm": 0.1940043717622757,
+      "learning_rate": 5e-06,
+      "loss": 0.146,
+      "step": 13072
+    },
+    {
+      "epoch": 9.840421528039142,
+      "grad_norm": 0.17529131472110748,
+      "learning_rate": 5e-06,
+      "loss": 0.1608,
+      "step": 13073
+    },
+    {
+      "epoch": 9.841174256680466,
+      "grad_norm": 0.1669771522283554,
+      "learning_rate": 5e-06,
+      "loss": 0.3053,
+      "step": 13074
+    },
+    {
+      "epoch": 9.841926985321791,
+      "grad_norm": 0.1696605384349823,
+      "learning_rate": 5e-06,
+      "loss": 0.1399,
+      "step": 13075
+    },
+    {
+      "epoch": 9.842679713963117,
+      "grad_norm": 0.18285028636455536,
+      "learning_rate": 5e-06,
+      "loss": 0.1452,
+      "step": 13076
+    },
+    {
+      "epoch": 9.843432442604442,
+      "grad_norm": 0.1629471480846405,
+      "learning_rate": 5e-06,
+      "loss": 0.1073,
+      "step": 13077
+    },
+    {
+      "epoch": 9.844185171245766,
+      "grad_norm": 0.15098312497138977,
+      "learning_rate": 5e-06,
+      "loss": 0.0848,
+      "step": 13078
+    },
+    {
+      "epoch": 9.844937899887091,
+      "grad_norm": 0.20874592661857605,
+      "learning_rate": 5e-06,
+      "loss": 0.1889,
+      "step": 13079
+    },
+    {
+      "epoch": 9.845690628528416,
+      "grad_norm": 0.1476861536502838,
+      "learning_rate": 5e-06,
+      "loss": 0.1281,
+      "step": 13080
+    },
+    {
+      "epoch": 9.84644335716974,
+      "grad_norm": 0.20148268342018127,
+      "learning_rate": 5e-06,
+      "loss": 0.1332,
+      "step": 13081
+    },
+    {
+      "epoch": 9.847196085811065,
+      "grad_norm": 0.1991550624370575,
+      "learning_rate": 5e-06,
+      "loss": 0.2066,
+      "step": 13082
+    },
+    {
+      "epoch": 9.84794881445239,
+      "grad_norm": 0.17707090079784393,
+      "learning_rate": 5e-06,
+      "loss": 0.1417,
+      "step": 13083
+    },
+    {
+      "epoch": 9.848701543093714,
+      "grad_norm": 0.19121575355529785,
+      "learning_rate": 5e-06,
+      "loss": 0.1772,
+      "step": 13084
+    },
+    {
+      "epoch": 9.84945427173504,
+      "grad_norm": 0.16578271985054016,
+      "learning_rate": 5e-06,
+      "loss": 0.1471,
+      "step": 13085
+    },
+    {
+      "epoch": 9.850207000376365,
+      "grad_norm": 0.1745569109916687,
+      "learning_rate": 5e-06,
+      "loss": 0.1003,
+      "step": 13086
+    },
+    {
+      "epoch": 9.850959729017688,
+      "grad_norm": 0.19379064440727234,
+      "learning_rate": 5e-06,
+      "loss": 0.1818,
+      "step": 13087
+    },
+    {
+      "epoch": 9.851712457659014,
+      "grad_norm": 0.1853293627500534,
+      "learning_rate": 5e-06,
+      "loss": 0.1751,
+      "step": 13088
+    },
+    {
+      "epoch": 9.85246518630034,
+      "grad_norm": 0.17179875075817108,
+      "learning_rate": 5e-06,
+      "loss": 0.0938,
+      "step": 13089
+    },
+    {
+      "epoch": 9.853217914941663,
+      "grad_norm": 0.19949907064437866,
+      "learning_rate": 5e-06,
+      "loss": 0.1514,
+      "step": 13090
+    },
+    {
+      "epoch": 9.853970643582988,
+      "grad_norm": 0.1740541160106659,
+      "learning_rate": 5e-06,
+      "loss": 0.1246,
+      "step": 13091
+    },
+    {
+      "epoch": 9.854723372224313,
+      "grad_norm": 0.1830870509147644,
+      "learning_rate": 5e-06,
+      "loss": 0.116,
+      "step": 13092
+    },
+    {
+      "epoch": 9.855476100865637,
+      "grad_norm": 0.19778142869472504,
+      "learning_rate": 5e-06,
+      "loss": 0.1553,
+      "step": 13093
+    },
+    {
+      "epoch": 9.856228829506962,
+      "grad_norm": 0.20653893053531647,
+      "learning_rate": 5e-06,
+      "loss": 0.1686,
+      "step": 13094
+    },
+    {
+      "epoch": 9.856981558148288,
+      "grad_norm": 0.18024888634681702,
+      "learning_rate": 5e-06,
+      "loss": 0.1749,
+      "step": 13095
+    },
+    {
+      "epoch": 9.857734286789613,
+      "grad_norm": 0.160894975066185,
+      "learning_rate": 5e-06,
+      "loss": 0.1195,
+      "step": 13096
+    },
+    {
+      "epoch": 9.858487015430937,
+      "grad_norm": 0.16707362234592438,
+      "learning_rate": 5e-06,
+      "loss": 0.1603,
+      "step": 13097
+    },
+    {
+      "epoch": 9.859239744072262,
+      "grad_norm": 0.16300541162490845,
+      "learning_rate": 5e-06,
+      "loss": 0.1427,
+      "step": 13098
+    },
+    {
+      "epoch": 9.859992472713587,
+      "grad_norm": 0.164672389626503,
+      "learning_rate": 5e-06,
+      "loss": 0.1269,
+      "step": 13099
+    },
+    {
+      "epoch": 9.860745201354911,
+      "grad_norm": 0.17853295803070068,
+      "learning_rate": 5e-06,
+      "loss": 0.1022,
+      "step": 13100
+    },
+    {
+      "epoch": 9.861497929996236,
+      "grad_norm": 0.1810889095067978,
+      "learning_rate": 5e-06,
+      "loss": 0.1472,
+      "step": 13101
+    },
+    {
+      "epoch": 9.862250658637562,
+      "grad_norm": 0.15842390060424805,
+      "learning_rate": 5e-06,
+      "loss": 0.1524,
+      "step": 13102
+    },
+    {
+      "epoch": 9.863003387278885,
+      "grad_norm": 0.17747312784194946,
+      "learning_rate": 5e-06,
+      "loss": 0.1417,
+      "step": 13103
+    },
+    {
+      "epoch": 9.86375611592021,
+      "grad_norm": 0.18648935854434967,
+      "learning_rate": 5e-06,
+      "loss": 0.1123,
+      "step": 13104
+    },
+    {
+      "epoch": 9.864508844561536,
+      "grad_norm": 0.20258265733718872,
+      "learning_rate": 5e-06,
+      "loss": 0.1564,
+      "step": 13105
+    },
+    {
+      "epoch": 9.86526157320286,
+      "grad_norm": 0.20961563289165497,
+      "learning_rate": 5e-06,
+      "loss": 0.1573,
+      "step": 13106
+    },
+    {
+      "epoch": 9.866014301844185,
+      "grad_norm": 0.2050522416830063,
+      "learning_rate": 5e-06,
+      "loss": 0.2396,
+      "step": 13107
+    },
+    {
+      "epoch": 9.86676703048551,
+      "grad_norm": 0.15205201506614685,
+      "learning_rate": 5e-06,
+      "loss": 0.1076,
+      "step": 13108
+    },
+    {
+      "epoch": 9.867519759126834,
+      "grad_norm": 0.18398895859718323,
+      "learning_rate": 5e-06,
+      "loss": 0.1269,
+      "step": 13109
+    },
+    {
+      "epoch": 9.86827248776816,
+      "grad_norm": 0.18638549745082855,
+      "learning_rate": 5e-06,
+      "loss": 0.16,
+      "step": 13110
+    },
+    {
+      "epoch": 9.869025216409485,
+      "grad_norm": 0.19149313867092133,
+      "learning_rate": 5e-06,
+      "loss": 0.1722,
+      "step": 13111
+    },
+    {
+      "epoch": 9.86977794505081,
+      "grad_norm": 0.2080046534538269,
+      "learning_rate": 5e-06,
+      "loss": 0.1389,
+      "step": 13112
+    },
+    {
+      "epoch": 9.870530673692134,
+      "grad_norm": 0.18834596872329712,
+      "learning_rate": 5e-06,
+      "loss": 0.2022,
+      "step": 13113
+    },
+    {
+      "epoch": 9.871283402333459,
+      "grad_norm": 0.18313132226467133,
+      "learning_rate": 5e-06,
+      "loss": 0.1606,
+      "step": 13114
+    },
+    {
+      "epoch": 9.872036130974784,
+      "grad_norm": 0.18476319313049316,
+      "learning_rate": 5e-06,
+      "loss": 0.1262,
+      "step": 13115
+    },
+    {
+      "epoch": 9.872788859616108,
+      "grad_norm": 0.16123732924461365,
+      "learning_rate": 5e-06,
+      "loss": 0.1559,
+      "step": 13116
+    },
+    {
+      "epoch": 9.873541588257433,
+      "grad_norm": 0.17517024278640747,
+      "learning_rate": 5e-06,
+      "loss": 0.2559,
+      "step": 13117
+    },
+    {
+      "epoch": 9.874294316898759,
+      "grad_norm": 0.16684705018997192,
+      "learning_rate": 5e-06,
+      "loss": 0.1438,
+      "step": 13118
+    },
+    {
+      "epoch": 9.875047045540082,
+      "grad_norm": 0.17549771070480347,
+      "learning_rate": 5e-06,
+      "loss": 0.1668,
+      "step": 13119
+    },
+    {
+      "epoch": 9.875799774181408,
+      "grad_norm": 0.15255427360534668,
+      "learning_rate": 5e-06,
+      "loss": 0.138,
+      "step": 13120
+    },
+    {
+      "epoch": 9.876552502822733,
+      "grad_norm": 0.16653083264827728,
+      "learning_rate": 5e-06,
+      "loss": 0.1372,
+      "step": 13121
+    },
+    {
+      "epoch": 9.877305231464057,
+      "grad_norm": 0.1893136352300644,
+      "learning_rate": 5e-06,
+      "loss": 0.1422,
+      "step": 13122
+    },
+    {
+      "epoch": 9.878057960105382,
+      "grad_norm": 0.19786055386066437,
+      "learning_rate": 5e-06,
+      "loss": 0.1693,
+      "step": 13123
+    },
+    {
+      "epoch": 9.878810688746707,
+      "grad_norm": 0.18498317897319794,
+      "learning_rate": 5e-06,
+      "loss": 0.1187,
+      "step": 13124
+    },
+    {
+      "epoch": 9.879563417388031,
+      "grad_norm": 0.16777165234088898,
+      "learning_rate": 5e-06,
+      "loss": 0.0925,
+      "step": 13125
+    },
+    {
+      "epoch": 9.880316146029356,
+      "grad_norm": 0.17761950194835663,
+      "learning_rate": 5e-06,
+      "loss": 0.1693,
+      "step": 13126
+    },
+    {
+      "epoch": 9.881068874670682,
+      "grad_norm": 0.1971915364265442,
+      "learning_rate": 5e-06,
+      "loss": 0.1518,
+      "step": 13127
+    },
+    {
+      "epoch": 9.881821603312005,
+      "grad_norm": 0.1862911581993103,
+      "learning_rate": 5e-06,
+      "loss": 0.2127,
+      "step": 13128
+    },
+    {
+      "epoch": 9.88257433195333,
+      "grad_norm": 0.20134767889976501,
+      "learning_rate": 5e-06,
+      "loss": 0.163,
+      "step": 13129
+    },
+    {
+      "epoch": 9.883327060594656,
+      "grad_norm": 0.19430607557296753,
+      "learning_rate": 5e-06,
+      "loss": 0.2036,
+      "step": 13130
+    },
+    {
+      "epoch": 9.88407978923598,
+      "grad_norm": 0.17391417920589447,
+      "learning_rate": 5e-06,
+      "loss": 0.1394,
+      "step": 13131
+    },
+    {
+      "epoch": 9.884832517877305,
+      "grad_norm": 0.18942858278751373,
+      "learning_rate": 5e-06,
+      "loss": 0.18,
+      "step": 13132
+    },
+    {
+      "epoch": 9.88558524651863,
+      "grad_norm": 0.16535617411136627,
+      "learning_rate": 5e-06,
+      "loss": 0.1187,
+      "step": 13133
+    },
+    {
+      "epoch": 9.886337975159956,
+      "grad_norm": 0.16850443184375763,
+      "learning_rate": 5e-06,
+      "loss": 0.0749,
+      "step": 13134
+    },
+    {
+      "epoch": 9.88709070380128,
+      "grad_norm": 0.18188421428203583,
+      "learning_rate": 5e-06,
+      "loss": 0.1007,
+      "step": 13135
+    },
+    {
+      "epoch": 9.887843432442605,
+      "grad_norm": 0.18257999420166016,
+      "learning_rate": 5e-06,
+      "loss": 0.102,
+      "step": 13136
+    },
+    {
+      "epoch": 9.88859616108393,
+      "grad_norm": 0.19866353273391724,
+      "learning_rate": 5e-06,
+      "loss": 0.1618,
+      "step": 13137
+    },
+    {
+      "epoch": 9.889348889725253,
+      "grad_norm": 0.17160765826702118,
+      "learning_rate": 5e-06,
+      "loss": 0.1767,
+      "step": 13138
+    },
+    {
+      "epoch": 9.890101618366579,
+      "grad_norm": 0.1477920562028885,
+      "learning_rate": 5e-06,
+      "loss": 0.1081,
+      "step": 13139
+    },
+    {
+      "epoch": 9.890854347007904,
+      "grad_norm": 0.18782898783683777,
+      "learning_rate": 5e-06,
+      "loss": 0.1206,
+      "step": 13140
+    },
+    {
+      "epoch": 9.891607075649228,
+      "grad_norm": 0.19161760807037354,
+      "learning_rate": 5e-06,
+      "loss": 0.1617,
+      "step": 13141
+    },
+    {
+      "epoch": 9.892359804290553,
+      "grad_norm": 0.19939476251602173,
+      "learning_rate": 5e-06,
+      "loss": 0.1504,
+      "step": 13142
+    },
+    {
+      "epoch": 9.893112532931879,
+      "grad_norm": 0.17918521165847778,
+      "learning_rate": 5e-06,
+      "loss": 0.1812,
+      "step": 13143
+    },
+    {
+      "epoch": 9.893865261573202,
+      "grad_norm": 0.16205079853534698,
+      "learning_rate": 5e-06,
+      "loss": 0.1174,
+      "step": 13144
+    },
+    {
+      "epoch": 9.894617990214527,
+      "grad_norm": 0.1811121702194214,
+      "learning_rate": 5e-06,
+      "loss": 0.116,
+      "step": 13145
+    },
+    {
+      "epoch": 9.895370718855853,
+      "grad_norm": 0.14130350947380066,
+      "learning_rate": 5e-06,
+      "loss": 0.144,
+      "step": 13146
+    },
+    {
+      "epoch": 9.896123447497178,
+      "grad_norm": 0.17137108743190765,
+      "learning_rate": 5e-06,
+      "loss": 0.1879,
+      "step": 13147
+    },
+    {
+      "epoch": 9.896876176138502,
+      "grad_norm": 0.1606842428445816,
+      "learning_rate": 5e-06,
+      "loss": 0.0869,
+      "step": 13148
+    },
+    {
+      "epoch": 9.897628904779827,
+      "grad_norm": 0.2081773430109024,
+      "learning_rate": 5e-06,
+      "loss": 0.1946,
+      "step": 13149
+    },
+    {
+      "epoch": 9.898381633421153,
+      "grad_norm": 0.1812097579240799,
+      "learning_rate": 5e-06,
+      "loss": 0.1533,
+      "step": 13150
+    },
+    {
+      "epoch": 9.899134362062476,
+      "grad_norm": 0.17405976355075836,
+      "learning_rate": 5e-06,
+      "loss": 0.2105,
+      "step": 13151
+    },
+    {
+      "epoch": 9.899887090703801,
+      "grad_norm": 0.18167071044445038,
+      "learning_rate": 5e-06,
+      "loss": 0.1646,
+      "step": 13152
+    },
+    {
+      "epoch": 9.900639819345127,
+      "grad_norm": 0.1751294732093811,
+      "learning_rate": 5e-06,
+      "loss": 0.1491,
+      "step": 13153
+    },
+    {
+      "epoch": 9.90139254798645,
+      "grad_norm": 0.19107474386692047,
+      "learning_rate": 5e-06,
+      "loss": 0.1575,
+      "step": 13154
+    },
+    {
+      "epoch": 9.902145276627776,
+      "grad_norm": 0.1857837438583374,
+      "learning_rate": 5e-06,
+      "loss": 0.0733,
+      "step": 13155
+    },
+    {
+      "epoch": 9.902898005269101,
+      "grad_norm": 0.17589974403381348,
+      "learning_rate": 5e-06,
+      "loss": 0.0867,
+      "step": 13156
+    },
+    {
+      "epoch": 9.903650733910425,
+      "grad_norm": 0.18829944729804993,
+      "learning_rate": 5e-06,
+      "loss": 0.1813,
+      "step": 13157
+    },
+    {
+      "epoch": 9.90440346255175,
+      "grad_norm": 0.1688997894525528,
+      "learning_rate": 5e-06,
+      "loss": 0.1454,
+      "step": 13158
+    },
+    {
+      "epoch": 9.905156191193075,
+      "grad_norm": 0.1713249683380127,
+      "learning_rate": 5e-06,
+      "loss": 0.1553,
+      "step": 13159
+    },
+    {
+      "epoch": 9.905908919834399,
+      "grad_norm": 0.19595468044281006,
+      "learning_rate": 5e-06,
+      "loss": 0.1399,
+      "step": 13160
+    },
+    {
+      "epoch": 9.906661648475724,
+      "grad_norm": 0.16014748811721802,
+      "learning_rate": 5e-06,
+      "loss": 0.1761,
+      "step": 13161
+    },
+    {
+      "epoch": 9.90741437711705,
+      "grad_norm": 0.19557519257068634,
+      "learning_rate": 5e-06,
+      "loss": 0.0965,
+      "step": 13162
+    },
+    {
+      "epoch": 9.908167105758373,
+      "grad_norm": 0.18579073250293732,
+      "learning_rate": 5e-06,
+      "loss": 0.1282,
+      "step": 13163
+    },
+    {
+      "epoch": 9.908919834399699,
+      "grad_norm": 0.15706419944763184,
+      "learning_rate": 5e-06,
+      "loss": 0.1351,
+      "step": 13164
+    },
+    {
+      "epoch": 9.909672563041024,
+      "grad_norm": 0.15537413954734802,
+      "learning_rate": 5e-06,
+      "loss": 0.1608,
+      "step": 13165
+    },
+    {
+      "epoch": 9.910425291682348,
+      "grad_norm": 0.20437577366828918,
+      "learning_rate": 5e-06,
+      "loss": 0.1706,
+      "step": 13166
+    },
+    {
+      "epoch": 9.911178020323673,
+      "grad_norm": 0.2060416340827942,
+      "learning_rate": 5e-06,
+      "loss": 0.2065,
+      "step": 13167
+    },
+    {
+      "epoch": 9.911930748964998,
+      "grad_norm": 0.16628435254096985,
+      "learning_rate": 5e-06,
+      "loss": 0.1401,
+      "step": 13168
+    },
+    {
+      "epoch": 9.912683477606324,
+      "grad_norm": 0.16121414303779602,
+      "learning_rate": 5e-06,
+      "loss": 0.1019,
+      "step": 13169
+    },
+    {
+      "epoch": 9.913436206247647,
+      "grad_norm": 0.21109077334403992,
+      "learning_rate": 5e-06,
+      "loss": 0.2005,
+      "step": 13170
+    },
+    {
+      "epoch": 9.914188934888973,
+      "grad_norm": 0.1684553325176239,
+      "learning_rate": 5e-06,
+      "loss": 0.1423,
+      "step": 13171
+    },
+    {
+      "epoch": 9.914941663530298,
+      "grad_norm": 0.16461175680160522,
+      "learning_rate": 5e-06,
+      "loss": 0.1163,
+      "step": 13172
+    },
+    {
+      "epoch": 9.915694392171622,
+      "grad_norm": 0.17973697185516357,
+      "learning_rate": 5e-06,
+      "loss": 0.1222,
+      "step": 13173
+    },
+    {
+      "epoch": 9.916447120812947,
+      "grad_norm": 0.21265247464179993,
+      "learning_rate": 5e-06,
+      "loss": 0.2376,
+      "step": 13174
+    },
+    {
+      "epoch": 9.917199849454272,
+      "grad_norm": 0.18687206506729126,
+      "learning_rate": 5e-06,
+      "loss": 0.1444,
+      "step": 13175
+    },
+    {
+      "epoch": 9.917952578095596,
+      "grad_norm": 0.17012222111225128,
+      "learning_rate": 5e-06,
+      "loss": 0.192,
+      "step": 13176
+    },
+    {
+      "epoch": 9.918705306736921,
+      "grad_norm": 0.1952119767665863,
+      "learning_rate": 5e-06,
+      "loss": 0.1307,
+      "step": 13177
+    },
+    {
+      "epoch": 9.919458035378247,
+      "grad_norm": 0.2124917358160019,
+      "learning_rate": 5e-06,
+      "loss": 0.1998,
+      "step": 13178
+    },
+    {
+      "epoch": 9.92021076401957,
+      "grad_norm": 0.15669015049934387,
+      "learning_rate": 5e-06,
+      "loss": 0.1104,
+      "step": 13179
+    },
+    {
+      "epoch": 9.920963492660896,
+      "grad_norm": 0.17778058350086212,
+      "learning_rate": 5e-06,
+      "loss": 0.1649,
+      "step": 13180
+    },
+    {
+      "epoch": 9.921716221302221,
+      "grad_norm": 0.15483829379081726,
+      "learning_rate": 5e-06,
+      "loss": 0.1258,
+      "step": 13181
+    },
+    {
+      "epoch": 9.922468949943545,
+      "grad_norm": 0.18250714242458344,
+      "learning_rate": 5e-06,
+      "loss": 0.1821,
+      "step": 13182
+    },
+    {
+      "epoch": 9.92322167858487,
+      "grad_norm": 0.16034138202667236,
+      "learning_rate": 5e-06,
+      "loss": 0.1732,
+      "step": 13183
+    },
+    {
+      "epoch": 9.923974407226195,
+      "grad_norm": 0.1671404391527176,
+      "learning_rate": 5e-06,
+      "loss": 0.1537,
+      "step": 13184
+    },
+    {
+      "epoch": 9.92472713586752,
+      "grad_norm": 0.1842643916606903,
+      "learning_rate": 5e-06,
+      "loss": 0.1842,
+      "step": 13185
+    },
+    {
+      "epoch": 9.925479864508844,
+      "grad_norm": 0.18441720306873322,
+      "learning_rate": 5e-06,
+      "loss": 0.1296,
+      "step": 13186
+    },
+    {
+      "epoch": 9.92623259315017,
+      "grad_norm": 0.16181635856628418,
+      "learning_rate": 5e-06,
+      "loss": 0.1156,
+      "step": 13187
+    },
+    {
+      "epoch": 9.926985321791495,
+      "grad_norm": 0.1874782294034958,
+      "learning_rate": 5e-06,
+      "loss": 0.1271,
+      "step": 13188
+    },
+    {
+      "epoch": 9.927738050432819,
+      "grad_norm": 0.2060391902923584,
+      "learning_rate": 5e-06,
+      "loss": 0.1812,
+      "step": 13189
+    },
+    {
+      "epoch": 9.928490779074144,
+      "grad_norm": 0.2137683629989624,
+      "learning_rate": 5e-06,
+      "loss": 0.1326,
+      "step": 13190
+    },
+    {
+      "epoch": 9.92924350771547,
+      "grad_norm": 0.20070479810237885,
+      "learning_rate": 5e-06,
+      "loss": 0.205,
+      "step": 13191
+    },
+    {
+      "epoch": 9.929996236356793,
+      "grad_norm": 0.21401512622833252,
+      "learning_rate": 5e-06,
+      "loss": 0.1419,
+      "step": 13192
+    },
+    {
+      "epoch": 9.930748964998118,
+      "grad_norm": 0.17347174882888794,
+      "learning_rate": 5e-06,
+      "loss": 0.1477,
+      "step": 13193
+    },
+    {
+      "epoch": 9.931501693639444,
+      "grad_norm": 0.17406409978866577,
+      "learning_rate": 5e-06,
+      "loss": 0.1673,
+      "step": 13194
+    },
+    {
+      "epoch": 9.932254422280767,
+      "grad_norm": 0.19315020740032196,
+      "learning_rate": 5e-06,
+      "loss": 0.1551,
+      "step": 13195
+    },
+    {
+      "epoch": 9.933007150922093,
+      "grad_norm": 0.21726100146770477,
+      "learning_rate": 5e-06,
+      "loss": 0.221,
+      "step": 13196
+    },
+    {
+      "epoch": 9.933759879563418,
+      "grad_norm": 0.1767137199640274,
+      "learning_rate": 5e-06,
+      "loss": 0.1697,
+      "step": 13197
+    },
+    {
+      "epoch": 9.934512608204741,
+      "grad_norm": 0.1779627501964569,
+      "learning_rate": 5e-06,
+      "loss": 0.1494,
+      "step": 13198
+    },
+    {
+      "epoch": 9.935265336846067,
+      "grad_norm": 0.17452524602413177,
+      "learning_rate": 5e-06,
+      "loss": 0.1188,
+      "step": 13199
+    },
+    {
+      "epoch": 9.936018065487392,
+      "grad_norm": 0.20532368123531342,
+      "learning_rate": 5e-06,
+      "loss": 0.1797,
+      "step": 13200
+    },
+    {
+      "epoch": 9.936018065487392,
+      "eval_loss": 0.1906425505876541,
+      "eval_runtime": 452.6953,
+      "eval_samples_per_second": 21.266,
+      "eval_steps_per_second": 0.665,
+      "step": 13200
+    },
+    {
+      "epoch": 9.936770794128716,
+      "grad_norm": 0.188282772898674,
+      "learning_rate": 5e-06,
+      "loss": 0.1492,
+      "step": 13201
+    },
+    {
+      "epoch": 9.937523522770041,
+      "grad_norm": 0.1831858605146408,
+      "learning_rate": 5e-06,
+      "loss": 0.1499,
+      "step": 13202
+    },
+    {
+      "epoch": 9.938276251411367,
+      "grad_norm": 0.16965636610984802,
+      "learning_rate": 5e-06,
+      "loss": 0.1356,
+      "step": 13203
+    },
+    {
+      "epoch": 9.93902898005269,
+      "grad_norm": 0.17721587419509888,
+      "learning_rate": 5e-06,
+      "loss": 0.1172,
+      "step": 13204
+    },
+    {
+      "epoch": 9.939781708694015,
+      "grad_norm": 0.17273761332035065,
+      "learning_rate": 5e-06,
+      "loss": 0.1091,
+      "step": 13205
+    },
+    {
+      "epoch": 9.94053443733534,
+      "grad_norm": 0.1736735701560974,
+      "learning_rate": 5e-06,
+      "loss": 0.0979,
+      "step": 13206
+    },
+    {
+      "epoch": 9.941287165976666,
+      "grad_norm": 0.16636990010738373,
+      "learning_rate": 5e-06,
+      "loss": 0.1228,
+      "step": 13207
+    },
+    {
+      "epoch": 9.94203989461799,
+      "grad_norm": 0.17415516078472137,
+      "learning_rate": 5e-06,
+      "loss": 0.1688,
+      "step": 13208
+    },
+    {
+      "epoch": 9.942792623259315,
+      "grad_norm": 0.20134113729000092,
+      "learning_rate": 5e-06,
+      "loss": 0.1627,
+      "step": 13209
+    },
+    {
+      "epoch": 9.94354535190064,
+      "grad_norm": 0.18516255915164948,
+      "learning_rate": 5e-06,
+      "loss": 0.161,
+      "step": 13210
+    },
+    {
+      "epoch": 9.944298080541964,
+      "grad_norm": 0.1961136907339096,
+      "learning_rate": 5e-06,
+      "loss": 0.1862,
+      "step": 13211
+    },
+    {
+      "epoch": 9.94505080918329,
+      "grad_norm": 0.17654292285442352,
+      "learning_rate": 5e-06,
+      "loss": 0.1288,
+      "step": 13212
+    },
+    {
+      "epoch": 9.945803537824615,
+      "grad_norm": 0.18269269168376923,
+      "learning_rate": 5e-06,
+      "loss": 0.1711,
+      "step": 13213
+    },
+    {
+      "epoch": 9.946556266465938,
+      "grad_norm": 0.1599319577217102,
+      "learning_rate": 5e-06,
+      "loss": 0.0889,
+      "step": 13214
+    },
+    {
+      "epoch": 9.947308995107264,
+      "grad_norm": 0.1676236391067505,
+      "learning_rate": 5e-06,
+      "loss": 0.187,
+      "step": 13215
+    },
+    {
+      "epoch": 9.948061723748589,
+      "grad_norm": 0.18220452964305878,
+      "learning_rate": 5e-06,
+      "loss": 0.1427,
+      "step": 13216
+    },
+    {
+      "epoch": 9.948814452389913,
+      "grad_norm": 0.16156457364559174,
+      "learning_rate": 5e-06,
+      "loss": 0.161,
+      "step": 13217
+    },
+    {
+      "epoch": 9.949567181031238,
+      "grad_norm": 0.2132469266653061,
+      "learning_rate": 5e-06,
+      "loss": 0.1479,
+      "step": 13218
+    },
+    {
+      "epoch": 9.950319909672563,
+      "grad_norm": 0.18438996374607086,
+      "learning_rate": 5e-06,
+      "loss": 0.0854,
+      "step": 13219
+    },
+    {
+      "epoch": 9.951072638313889,
+      "grad_norm": 0.17849402129650116,
+      "learning_rate": 5e-06,
+      "loss": 0.1261,
+      "step": 13220
+    },
+    {
+      "epoch": 9.951825366955212,
+      "grad_norm": 0.21250759065151215,
+      "learning_rate": 5e-06,
+      "loss": 0.2528,
+      "step": 13221
+    },
+    {
+      "epoch": 9.952578095596538,
+      "grad_norm": 0.1855168640613556,
+      "learning_rate": 5e-06,
+      "loss": 0.1479,
+      "step": 13222
+    },
+    {
+      "epoch": 9.953330824237863,
+      "grad_norm": 0.17901460826396942,
+      "learning_rate": 5e-06,
+      "loss": 0.1012,
+      "step": 13223
+    },
+    {
+      "epoch": 9.954083552879187,
+      "grad_norm": 0.192086324095726,
+      "learning_rate": 5e-06,
+      "loss": 0.1392,
+      "step": 13224
+    },
+    {
+      "epoch": 9.954836281520512,
+      "grad_norm": 0.18458107113838196,
+      "learning_rate": 5e-06,
+      "loss": 0.176,
+      "step": 13225
+    },
+    {
+      "epoch": 9.955589010161837,
+      "grad_norm": 0.17509213089942932,
+      "learning_rate": 5e-06,
+      "loss": 0.1146,
+      "step": 13226
+    },
+    {
+      "epoch": 9.956341738803161,
+      "grad_norm": 0.16141128540039062,
+      "learning_rate": 5e-06,
+      "loss": 0.1781,
+      "step": 13227
+    },
+    {
+      "epoch": 9.957094467444486,
+      "grad_norm": 0.19031067192554474,
+      "learning_rate": 5e-06,
+      "loss": 0.2135,
+      "step": 13228
+    },
+    {
+      "epoch": 9.957847196085812,
+      "grad_norm": 0.19958311319351196,
+      "learning_rate": 5e-06,
+      "loss": 0.134,
+      "step": 13229
+    },
+    {
+      "epoch": 9.958599924727135,
+      "grad_norm": 0.2096472829580307,
+      "learning_rate": 5e-06,
+      "loss": 0.1582,
+      "step": 13230
+    },
+    {
+      "epoch": 9.95935265336846,
+      "grad_norm": 0.19276346266269684,
+      "learning_rate": 5e-06,
+      "loss": 0.1405,
+      "step": 13231
+    },
+    {
+      "epoch": 9.960105382009786,
+      "grad_norm": 0.1921595185995102,
+      "learning_rate": 5e-06,
+      "loss": 0.2186,
+      "step": 13232
+    },
+    {
+      "epoch": 9.96085811065111,
+      "grad_norm": 0.2079499065876007,
+      "learning_rate": 5e-06,
+      "loss": 0.185,
+      "step": 13233
+    },
+    {
+      "epoch": 9.961610839292435,
+      "grad_norm": 0.1782999485731125,
+      "learning_rate": 5e-06,
+      "loss": 0.163,
+      "step": 13234
+    },
+    {
+      "epoch": 9.96236356793376,
+      "grad_norm": 0.18412652611732483,
+      "learning_rate": 5e-06,
+      "loss": 0.1045,
+      "step": 13235
+    },
+    {
+      "epoch": 9.963116296575084,
+      "grad_norm": 0.18981777131557465,
+      "learning_rate": 5e-06,
+      "loss": 0.1996,
+      "step": 13236
+    },
+    {
+      "epoch": 9.96386902521641,
+      "grad_norm": 0.17610085010528564,
+      "learning_rate": 5e-06,
+      "loss": 0.1653,
+      "step": 13237
+    },
+    {
+      "epoch": 9.964621753857735,
+      "grad_norm": 0.16295285522937775,
+      "learning_rate": 5e-06,
+      "loss": 0.1091,
+      "step": 13238
+    },
+    {
+      "epoch": 9.965374482499058,
+      "grad_norm": 0.14490404725074768,
+      "learning_rate": 5e-06,
+      "loss": 0.0813,
+      "step": 13239
+    },
+    {
+      "epoch": 9.966127211140384,
+      "grad_norm": 0.19015353918075562,
+      "learning_rate": 5e-06,
+      "loss": 0.1801,
+      "step": 13240
+    },
+    {
+      "epoch": 9.966879939781709,
+      "grad_norm": 0.17203941941261292,
+      "learning_rate": 5e-06,
+      "loss": 0.0865,
+      "step": 13241
+    },
+    {
+      "epoch": 9.967632668423034,
+      "grad_norm": 0.23017963767051697,
+      "learning_rate": 5e-06,
+      "loss": 0.1522,
+      "step": 13242
+    },
+    {
+      "epoch": 9.968385397064358,
+      "grad_norm": 0.2094835638999939,
+      "learning_rate": 5e-06,
+      "loss": 0.2053,
+      "step": 13243
+    },
+    {
+      "epoch": 9.969138125705683,
+      "grad_norm": 0.16221795976161957,
+      "learning_rate": 5e-06,
+      "loss": 0.1599,
+      "step": 13244
+    },
+    {
+      "epoch": 9.969890854347009,
+      "grad_norm": 0.18544414639472961,
+      "learning_rate": 5e-06,
+      "loss": 0.19,
+      "step": 13245
+    },
+    {
+      "epoch": 9.970643582988332,
+      "grad_norm": 0.17010058462619781,
+      "learning_rate": 5e-06,
+      "loss": 0.1493,
+      "step": 13246
+    },
+    {
+      "epoch": 9.971396311629658,
+      "grad_norm": 0.1808137446641922,
+      "learning_rate": 5e-06,
+      "loss": 0.0787,
+      "step": 13247
+    },
+    {
+      "epoch": 9.972149040270983,
+      "grad_norm": 0.1614324152469635,
+      "learning_rate": 5e-06,
+      "loss": 0.179,
+      "step": 13248
+    },
+    {
+      "epoch": 9.972901768912307,
+      "grad_norm": 0.18464291095733643,
+      "learning_rate": 5e-06,
+      "loss": 0.1738,
+      "step": 13249
+    },
+    {
+      "epoch": 9.973654497553632,
+      "grad_norm": 0.17591208219528198,
+      "learning_rate": 5e-06,
+      "loss": 0.074,
+      "step": 13250
+    },
+    {
+      "epoch": 9.974407226194957,
+      "grad_norm": 0.15751425921916962,
+      "learning_rate": 5e-06,
+      "loss": 0.1065,
+      "step": 13251
+    },
+    {
+      "epoch": 9.97515995483628,
+      "grad_norm": 0.16656021773815155,
+      "learning_rate": 5e-06,
+      "loss": 0.1608,
+      "step": 13252
+    },
+    {
+      "epoch": 9.975912683477606,
+      "grad_norm": 0.16117669641971588,
+      "learning_rate": 5e-06,
+      "loss": 0.1236,
+      "step": 13253
+    },
+    {
+      "epoch": 9.976665412118932,
+      "grad_norm": 0.15195521712303162,
+      "learning_rate": 5e-06,
+      "loss": 0.0999,
+      "step": 13254
+    },
+    {
+      "epoch": 9.977418140760255,
+      "grad_norm": 0.16343232989311218,
+      "learning_rate": 5e-06,
+      "loss": 0.1055,
+      "step": 13255
+    },
+    {
+      "epoch": 9.97817086940158,
+      "grad_norm": 0.18440751731395721,
+      "learning_rate": 5e-06,
+      "loss": 0.1903,
+      "step": 13256
+    },
+    {
+      "epoch": 9.978923598042906,
+      "grad_norm": 0.18059292435646057,
+      "learning_rate": 5e-06,
+      "loss": 0.1809,
+      "step": 13257
+    },
+    {
+      "epoch": 9.979676326684231,
+      "grad_norm": 0.16207091510295868,
+      "learning_rate": 5e-06,
+      "loss": 0.1064,
+      "step": 13258
+    },
+    {
+      "epoch": 9.980429055325555,
+      "grad_norm": 0.16767513751983643,
+      "learning_rate": 5e-06,
+      "loss": 0.1202,
+      "step": 13259
+    },
+    {
+      "epoch": 9.98118178396688,
+      "grad_norm": 0.1647658497095108,
+      "learning_rate": 5e-06,
+      "loss": 0.1091,
+      "step": 13260
+    },
+    {
+      "epoch": 9.981934512608206,
+      "grad_norm": 0.15847192704677582,
+      "learning_rate": 5e-06,
+      "loss": 0.1164,
+      "step": 13261
+    },
+    {
+      "epoch": 9.982687241249529,
+      "grad_norm": 0.17955957353115082,
+      "learning_rate": 5e-06,
+      "loss": 0.2289,
+      "step": 13262
+    },
+    {
+      "epoch": 9.983439969890854,
+      "grad_norm": 0.17003706097602844,
+      "learning_rate": 5e-06,
+      "loss": 0.1405,
+      "step": 13263
+    },
+    {
+      "epoch": 9.98419269853218,
+      "grad_norm": 0.20044048130512238,
+      "learning_rate": 5e-06,
+      "loss": 0.1649,
+      "step": 13264
+    },
+    {
+      "epoch": 9.984945427173503,
+      "grad_norm": 0.15484821796417236,
+      "learning_rate": 5e-06,
+      "loss": 0.0515,
+      "step": 13265
+    },
+    {
+      "epoch": 9.985698155814829,
+      "grad_norm": 0.15390059351921082,
+      "learning_rate": 5e-06,
+      "loss": 0.1063,
+      "step": 13266
+    },
+    {
+      "epoch": 9.986450884456154,
+      "grad_norm": 0.23536691069602966,
+      "learning_rate": 5e-06,
+      "loss": 0.1766,
+      "step": 13267
+    },
+    {
+      "epoch": 9.987203613097478,
+      "grad_norm": 0.20348384976387024,
+      "learning_rate": 5e-06,
+      "loss": 0.2089,
+      "step": 13268
+    },
+    {
+      "epoch": 9.987956341738803,
+      "grad_norm": 0.19151480495929718,
+      "learning_rate": 5e-06,
+      "loss": 0.1322,
+      "step": 13269
+    },
+    {
+      "epoch": 9.988709070380128,
+      "grad_norm": 0.17966780066490173,
+      "learning_rate": 5e-06,
+      "loss": 0.154,
+      "step": 13270
+    },
+    {
+      "epoch": 9.989461799021452,
+      "grad_norm": 0.15124988555908203,
+      "learning_rate": 5e-06,
+      "loss": 0.1489,
+      "step": 13271
+    },
+    {
+      "epoch": 9.990214527662777,
+      "grad_norm": 0.16950839757919312,
+      "learning_rate": 5e-06,
+      "loss": 0.179,
+      "step": 13272
+    },
+    {
+      "epoch": 9.990967256304103,
+      "grad_norm": 0.1830974966287613,
+      "learning_rate": 5e-06,
+      "loss": 0.178,
+      "step": 13273
+    },
+    {
+      "epoch": 9.991719984945426,
+      "grad_norm": 0.1647062450647354,
+      "learning_rate": 5e-06,
+      "loss": 0.0964,
+      "step": 13274
+    },
+    {
+      "epoch": 9.992472713586752,
+      "grad_norm": 0.20445416867733002,
+      "learning_rate": 5e-06,
+      "loss": 0.17,
+      "step": 13275
+    },
+    {
+      "epoch": 9.993225442228077,
+      "grad_norm": 0.18197977542877197,
+      "learning_rate": 5e-06,
+      "loss": 0.1451,
+      "step": 13276
+    },
+    {
+      "epoch": 9.9939781708694,
+      "grad_norm": 0.16522568464279175,
+      "learning_rate": 5e-06,
+      "loss": 0.1291,
+      "step": 13277
+    },
+    {
+      "epoch": 9.994730899510726,
+      "grad_norm": 0.18524326384067535,
+      "learning_rate": 5e-06,
+      "loss": 0.1467,
+      "step": 13278
+    },
+    {
+      "epoch": 9.995483628152051,
+      "grad_norm": 0.1654081493616104,
+      "learning_rate": 5e-06,
+      "loss": 0.1293,
+      "step": 13279
+    },
+    {
+      "epoch": 9.996236356793377,
+      "grad_norm": 0.20250733196735382,
+      "learning_rate": 5e-06,
+      "loss": 0.1332,
+      "step": 13280
+    },
+    {
+      "epoch": 9.9969890854347,
+      "grad_norm": 0.1637529730796814,
+      "learning_rate": 5e-06,
+      "loss": 0.172,
+      "step": 13281
+    },
+    {
+      "epoch": 9.997741814076026,
+      "grad_norm": 0.18635669350624084,
+      "learning_rate": 5e-06,
+      "loss": 0.1874,
+      "step": 13282
+    },
+    {
+      "epoch": 9.998494542717351,
+      "grad_norm": 0.1652553379535675,
+      "learning_rate": 5e-06,
+      "loss": 0.1197,
+      "step": 13283
+    },
+    {
+      "epoch": 9.999247271358675,
+      "grad_norm": 0.18186993896961212,
+      "learning_rate": 5e-06,
+      "loss": 0.1592,
+      "step": 13284
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 0.15946529805660248,
+      "learning_rate": 5e-06,
+      "loss": 0.0862,
+      "step": 13285
+    },
+    {
+      "epoch": 10.000752728641325,
+      "grad_norm": 0.17813819646835327,
+      "learning_rate": 5e-06,
+      "loss": 0.1902,
+      "step": 13286
+    },
+    {
+      "epoch": 10.001505457282649,
+      "grad_norm": 0.17515838146209717,
+      "learning_rate": 5e-06,
+      "loss": 0.1315,
+      "step": 13287
+    },
+    {
+      "epoch": 10.002258185923974,
+      "grad_norm": 0.21367764472961426,
+      "learning_rate": 5e-06,
+      "loss": 0.1888,
+      "step": 13288
+    },
+    {
+      "epoch": 10.0030109145653,
+      "grad_norm": 0.16627123951911926,
+      "learning_rate": 5e-06,
+      "loss": 0.1234,
+      "step": 13289
+    },
+    {
+      "epoch": 10.003763643206623,
+      "grad_norm": 0.17159375548362732,
+      "learning_rate": 5e-06,
+      "loss": 0.0826,
+      "step": 13290
+    },
+    {
+      "epoch": 10.004516371847949,
+      "grad_norm": 0.16260169446468353,
+      "learning_rate": 5e-06,
+      "loss": 0.1316,
+      "step": 13291
+    },
+    {
+      "epoch": 10.005269100489274,
+      "grad_norm": 0.17795690894126892,
+      "learning_rate": 5e-06,
+      "loss": 0.1544,
+      "step": 13292
+    },
+    {
+      "epoch": 10.006021829130598,
+      "grad_norm": 0.172810360789299,
+      "learning_rate": 5e-06,
+      "loss": 0.1489,
+      "step": 13293
+    },
+    {
+      "epoch": 10.006774557771923,
+      "grad_norm": 0.16840335726737976,
+      "learning_rate": 5e-06,
+      "loss": 0.1879,
+      "step": 13294
+    },
+    {
+      "epoch": 10.007527286413248,
+      "grad_norm": 0.2135988473892212,
+      "learning_rate": 5e-06,
+      "loss": 0.2072,
+      "step": 13295
+    },
+    {
+      "epoch": 10.008280015054574,
+      "grad_norm": 0.15978781878948212,
+      "learning_rate": 5e-06,
+      "loss": 0.191,
+      "step": 13296
+    },
+    {
+      "epoch": 10.009032743695897,
+      "grad_norm": 0.1679542362689972,
+      "learning_rate": 5e-06,
+      "loss": 0.1476,
+      "step": 13297
+    },
+    {
+      "epoch": 10.009785472337223,
+      "grad_norm": 0.18868935108184814,
+      "learning_rate": 5e-06,
+      "loss": 0.1337,
+      "step": 13298
+    },
+    {
+      "epoch": 10.010538200978548,
+      "grad_norm": 0.18513014912605286,
+      "learning_rate": 5e-06,
+      "loss": 0.1547,
+      "step": 13299
+    },
+    {
+      "epoch": 10.011290929619872,
+      "grad_norm": 0.15383101999759674,
+      "learning_rate": 5e-06,
+      "loss": 0.1116,
+      "step": 13300
+    },
+    {
+      "epoch": 10.012043658261197,
+      "grad_norm": 0.17448903620243073,
+      "learning_rate": 5e-06,
+      "loss": 0.1438,
+      "step": 13301
+    },
+    {
+      "epoch": 10.012796386902522,
+      "grad_norm": 0.20291167497634888,
+      "learning_rate": 5e-06,
+      "loss": 0.1845,
+      "step": 13302
+    },
+    {
+      "epoch": 10.013549115543846,
+      "grad_norm": 0.2216302901506424,
+      "learning_rate": 5e-06,
+      "loss": 0.1119,
+      "step": 13303
+    },
+    {
+      "epoch": 10.014301844185171,
+      "grad_norm": 0.1888408213853836,
+      "learning_rate": 5e-06,
+      "loss": 0.1527,
+      "step": 13304
+    },
+    {
+      "epoch": 10.015054572826497,
+      "grad_norm": 0.16176101565361023,
+      "learning_rate": 5e-06,
+      "loss": 0.1621,
+      "step": 13305
+    },
+    {
+      "epoch": 10.01580730146782,
+      "grad_norm": 0.16853588819503784,
+      "learning_rate": 5e-06,
+      "loss": 0.0923,
+      "step": 13306
+    },
+    {
+      "epoch": 10.016560030109146,
+      "grad_norm": 0.19796158373355865,
+      "learning_rate": 5e-06,
+      "loss": 0.1181,
+      "step": 13307
+    },
+    {
+      "epoch": 10.017312758750471,
+      "grad_norm": 0.1799767017364502,
+      "learning_rate": 5e-06,
+      "loss": 0.0912,
+      "step": 13308
+    },
+    {
+      "epoch": 10.018065487391794,
+      "grad_norm": 0.20834419131278992,
+      "learning_rate": 5e-06,
+      "loss": 0.1721,
+      "step": 13309
+    },
+    {
+      "epoch": 10.01881821603312,
+      "grad_norm": 0.18610140681266785,
+      "learning_rate": 5e-06,
+      "loss": 0.1087,
+      "step": 13310
+    },
+    {
+      "epoch": 10.019570944674445,
+      "grad_norm": 0.1930844783782959,
+      "learning_rate": 5e-06,
+      "loss": 0.1969,
+      "step": 13311
+    },
+    {
+      "epoch": 10.020323673315769,
+      "grad_norm": 0.17048919200897217,
+      "learning_rate": 5e-06,
+      "loss": 0.1232,
+      "step": 13312
+    },
+    {
+      "epoch": 10.021076401957094,
+      "grad_norm": 0.18271692097187042,
+      "learning_rate": 5e-06,
+      "loss": 0.2968,
+      "step": 13313
+    },
+    {
+      "epoch": 10.02182913059842,
+      "grad_norm": 0.16376011073589325,
+      "learning_rate": 5e-06,
+      "loss": 0.0917,
+      "step": 13314
+    },
+    {
+      "epoch": 10.022581859239745,
+      "grad_norm": 0.2022957056760788,
+      "learning_rate": 5e-06,
+      "loss": 0.1743,
+      "step": 13315
+    },
+    {
+      "epoch": 10.023334587881068,
+      "grad_norm": 0.18002913892269135,
+      "learning_rate": 5e-06,
+      "loss": 0.14,
+      "step": 13316
+    },
+    {
+      "epoch": 10.024087316522394,
+      "grad_norm": 0.20825988054275513,
+      "learning_rate": 5e-06,
+      "loss": 0.2424,
+      "step": 13317
+    },
+    {
+      "epoch": 10.02484004516372,
+      "grad_norm": 0.18146324157714844,
+      "learning_rate": 5e-06,
+      "loss": 0.1324,
+      "step": 13318
+    },
+    {
+      "epoch": 10.025592773805043,
+      "grad_norm": 0.17388606071472168,
+      "learning_rate": 5e-06,
+      "loss": 0.1063,
+      "step": 13319
+    },
+    {
+      "epoch": 10.026345502446368,
+      "grad_norm": 0.1744130253791809,
+      "learning_rate": 5e-06,
+      "loss": 0.1312,
+      "step": 13320
+    },
+    {
+      "epoch": 10.027098231087693,
+      "grad_norm": 0.1549263596534729,
+      "learning_rate": 5e-06,
+      "loss": 0.0982,
+      "step": 13321
+    },
+    {
+      "epoch": 10.027850959729017,
+      "grad_norm": 0.2447318285703659,
+      "learning_rate": 5e-06,
+      "loss": 0.1872,
+      "step": 13322
+    },
+    {
+      "epoch": 10.028603688370342,
+      "grad_norm": 0.1710658073425293,
+      "learning_rate": 5e-06,
+      "loss": 0.177,
+      "step": 13323
+    },
+    {
+      "epoch": 10.029356417011668,
+      "grad_norm": 0.1555674821138382,
+      "learning_rate": 5e-06,
+      "loss": 0.1609,
+      "step": 13324
+    },
+    {
+      "epoch": 10.030109145652991,
+      "grad_norm": 0.17252323031425476,
+      "learning_rate": 5e-06,
+      "loss": 0.1261,
+      "step": 13325
+    },
+    {
+      "epoch": 10.030861874294317,
+      "grad_norm": 0.17927300930023193,
+      "learning_rate": 5e-06,
+      "loss": 0.1533,
+      "step": 13326
+    },
+    {
+      "epoch": 10.031614602935642,
+      "grad_norm": 0.1791871339082718,
+      "learning_rate": 5e-06,
+      "loss": 0.1201,
+      "step": 13327
+    },
+    {
+      "epoch": 10.032367331576966,
+      "grad_norm": 0.18067562580108643,
+      "learning_rate": 5e-06,
+      "loss": 0.1713,
+      "step": 13328
+    },
+    {
+      "epoch": 10.033120060218291,
+      "grad_norm": 0.17387865483760834,
+      "learning_rate": 5e-06,
+      "loss": 0.1416,
+      "step": 13329
+    },
+    {
+      "epoch": 10.033872788859616,
+      "grad_norm": 0.15912921726703644,
+      "learning_rate": 5e-06,
+      "loss": 0.1997,
+      "step": 13330
+    },
+    {
+      "epoch": 10.034625517500942,
+      "grad_norm": 0.17847341299057007,
+      "learning_rate": 5e-06,
+      "loss": 0.1066,
+      "step": 13331
+    },
+    {
+      "epoch": 10.035378246142265,
+      "grad_norm": 0.1768241822719574,
+      "learning_rate": 5e-06,
+      "loss": 0.1732,
+      "step": 13332
+    },
+    {
+      "epoch": 10.03613097478359,
+      "grad_norm": 0.2000025361776352,
+      "learning_rate": 5e-06,
+      "loss": 0.1498,
+      "step": 13333
+    },
+    {
+      "epoch": 10.036883703424916,
+      "grad_norm": 0.1863272488117218,
+      "learning_rate": 5e-06,
+      "loss": 0.2206,
+      "step": 13334
+    },
+    {
+      "epoch": 10.03763643206624,
+      "grad_norm": 0.16446220874786377,
+      "learning_rate": 5e-06,
+      "loss": 0.1373,
+      "step": 13335
+    },
+    {
+      "epoch": 10.038389160707565,
+      "grad_norm": 0.18446534872055054,
+      "learning_rate": 5e-06,
+      "loss": 0.1711,
+      "step": 13336
+    },
+    {
+      "epoch": 10.03914188934889,
+      "grad_norm": 0.19477668404579163,
+      "learning_rate": 5e-06,
+      "loss": 0.1975,
+      "step": 13337
+    },
+    {
+      "epoch": 10.039894617990214,
+      "grad_norm": 0.16548694670200348,
+      "learning_rate": 5e-06,
+      "loss": 0.1401,
+      "step": 13338
+    },
+    {
+      "epoch": 10.04064734663154,
+      "grad_norm": 0.18498829007148743,
+      "learning_rate": 5e-06,
+      "loss": 0.1327,
+      "step": 13339
+    },
+    {
+      "epoch": 10.041400075272865,
+      "grad_norm": 0.16595807671546936,
+      "learning_rate": 5e-06,
+      "loss": 0.0853,
+      "step": 13340
+    },
+    {
+      "epoch": 10.042152803914188,
+      "grad_norm": 0.20179292559623718,
+      "learning_rate": 5e-06,
+      "loss": 0.1883,
+      "step": 13341
+    },
+    {
+      "epoch": 10.042905532555514,
+      "grad_norm": 0.15533055365085602,
+      "learning_rate": 5e-06,
+      "loss": 0.1216,
+      "step": 13342
+    },
+    {
+      "epoch": 10.043658261196839,
+      "grad_norm": 0.18227483332157135,
+      "learning_rate": 5e-06,
+      "loss": 0.1145,
+      "step": 13343
+    },
+    {
+      "epoch": 10.044410989838163,
+      "grad_norm": 0.17248858511447906,
+      "learning_rate": 5e-06,
+      "loss": 0.1456,
+      "step": 13344
+    },
+    {
+      "epoch": 10.045163718479488,
+      "grad_norm": 0.1674109399318695,
+      "learning_rate": 5e-06,
+      "loss": 0.2164,
+      "step": 13345
+    },
+    {
+      "epoch": 10.045916447120813,
+      "grad_norm": 0.17371731996536255,
+      "learning_rate": 5e-06,
+      "loss": 0.1695,
+      "step": 13346
+    },
+    {
+      "epoch": 10.046669175762137,
+      "grad_norm": 0.1729731410741806,
+      "learning_rate": 5e-06,
+      "loss": 0.0894,
+      "step": 13347
+    },
+    {
+      "epoch": 10.047421904403462,
+      "grad_norm": 0.16022877395153046,
+      "learning_rate": 5e-06,
+      "loss": 0.1495,
+      "step": 13348
+    },
+    {
+      "epoch": 10.048174633044788,
+      "grad_norm": 0.15691393613815308,
+      "learning_rate": 5e-06,
+      "loss": 0.0988,
+      "step": 13349
+    },
+    {
+      "epoch": 10.048927361686113,
+      "grad_norm": 0.1840100884437561,
+      "learning_rate": 5e-06,
+      "loss": 0.2074,
+      "step": 13350
+    },
+    {
+      "epoch": 10.049680090327437,
+      "grad_norm": 0.17508062720298767,
+      "learning_rate": 5e-06,
+      "loss": 0.1444,
+      "step": 13351
+    },
+    {
+      "epoch": 10.050432818968762,
+      "grad_norm": 0.16296595335006714,
+      "learning_rate": 5e-06,
+      "loss": 0.1646,
+      "step": 13352
+    },
+    {
+      "epoch": 10.051185547610087,
+      "grad_norm": 0.1835949420928955,
+      "learning_rate": 5e-06,
+      "loss": 0.1199,
+      "step": 13353
+    },
+    {
+      "epoch": 10.051938276251411,
+      "grad_norm": 0.17333820462226868,
+      "learning_rate": 5e-06,
+      "loss": 0.0904,
+      "step": 13354
+    },
+    {
+      "epoch": 10.052691004892736,
+      "grad_norm": 0.19023378193378448,
+      "learning_rate": 5e-06,
+      "loss": 0.2457,
+      "step": 13355
+    },
+    {
+      "epoch": 10.053443733534062,
+      "grad_norm": 0.21891699731349945,
+      "learning_rate": 5e-06,
+      "loss": 0.2476,
+      "step": 13356
+    },
+    {
+      "epoch": 10.054196462175385,
+      "grad_norm": 0.18235404789447784,
+      "learning_rate": 5e-06,
+      "loss": 0.1452,
+      "step": 13357
+    },
+    {
+      "epoch": 10.05494919081671,
+      "grad_norm": 0.17057518661022186,
+      "learning_rate": 5e-06,
+      "loss": 0.1036,
+      "step": 13358
+    },
+    {
+      "epoch": 10.055701919458036,
+      "grad_norm": 0.19287918508052826,
+      "learning_rate": 5e-06,
+      "loss": 0.1684,
+      "step": 13359
+    },
+    {
+      "epoch": 10.05645464809936,
+      "grad_norm": 0.16211755573749542,
+      "learning_rate": 5e-06,
+      "loss": 0.1238,
+      "step": 13360
+    },
+    {
+      "epoch": 10.057207376740685,
+      "grad_norm": 0.21664847433567047,
+      "learning_rate": 5e-06,
+      "loss": 0.1382,
+      "step": 13361
+    },
+    {
+      "epoch": 10.05796010538201,
+      "grad_norm": 0.22008918225765228,
+      "learning_rate": 5e-06,
+      "loss": 0.1375,
+      "step": 13362
+    },
+    {
+      "epoch": 10.058712834023334,
+      "grad_norm": 0.189519003033638,
+      "learning_rate": 5e-06,
+      "loss": 0.1464,
+      "step": 13363
+    },
+    {
+      "epoch": 10.05946556266466,
+      "grad_norm": 0.16838723421096802,
+      "learning_rate": 5e-06,
+      "loss": 0.1011,
+      "step": 13364
+    },
+    {
+      "epoch": 10.060218291305985,
+      "grad_norm": 0.1890249401330948,
+      "learning_rate": 5e-06,
+      "loss": 0.1502,
+      "step": 13365
+    },
+    {
+      "epoch": 10.060971019947308,
+      "grad_norm": 0.17568589746952057,
+      "learning_rate": 5e-06,
+      "loss": 0.1623,
+      "step": 13366
+    },
+    {
+      "epoch": 10.061723748588633,
+      "grad_norm": 0.1647765338420868,
+      "learning_rate": 5e-06,
+      "loss": 0.1316,
+      "step": 13367
+    },
+    {
+      "epoch": 10.062476477229959,
+      "grad_norm": 0.1711231768131256,
+      "learning_rate": 5e-06,
+      "loss": 0.1365,
+      "step": 13368
+    },
+    {
+      "epoch": 10.063229205871284,
+      "grad_norm": 0.20490558445453644,
+      "learning_rate": 5e-06,
+      "loss": 0.2004,
+      "step": 13369
+    },
+    {
+      "epoch": 10.063981934512608,
+      "grad_norm": 0.18445922434329987,
+      "learning_rate": 5e-06,
+      "loss": 0.1599,
+      "step": 13370
+    },
+    {
+      "epoch": 10.064734663153933,
+      "grad_norm": 0.1722644865512848,
+      "learning_rate": 5e-06,
+      "loss": 0.1555,
+      "step": 13371
+    },
+    {
+      "epoch": 10.065487391795259,
+      "grad_norm": 0.1988976150751114,
+      "learning_rate": 5e-06,
+      "loss": 0.1684,
+      "step": 13372
+    },
+    {
+      "epoch": 10.066240120436582,
+      "grad_norm": 0.18801555037498474,
+      "learning_rate": 5e-06,
+      "loss": 0.2027,
+      "step": 13373
+    },
+    {
+      "epoch": 10.066992849077907,
+      "grad_norm": 0.18760254979133606,
+      "learning_rate": 5e-06,
+      "loss": 0.1783,
+      "step": 13374
+    },
+    {
+      "epoch": 10.067745577719233,
+      "grad_norm": 0.23059377074241638,
+      "learning_rate": 5e-06,
+      "loss": 0.1809,
+      "step": 13375
+    },
+    {
+      "epoch": 10.068498306360556,
+      "grad_norm": 0.17979548871517181,
+      "learning_rate": 5e-06,
+      "loss": 0.094,
+      "step": 13376
+    },
+    {
+      "epoch": 10.069251035001882,
+      "grad_norm": 0.16571983695030212,
+      "learning_rate": 5e-06,
+      "loss": 0.125,
+      "step": 13377
+    },
+    {
+      "epoch": 10.070003763643207,
+      "grad_norm": 0.16291284561157227,
+      "learning_rate": 5e-06,
+      "loss": 0.1026,
+      "step": 13378
+    },
+    {
+      "epoch": 10.07075649228453,
+      "grad_norm": 0.18285438418388367,
+      "learning_rate": 5e-06,
+      "loss": 0.1529,
+      "step": 13379
+    },
+    {
+      "epoch": 10.071509220925856,
+      "grad_norm": 0.18302959203720093,
+      "learning_rate": 5e-06,
+      "loss": 0.1324,
+      "step": 13380
+    },
+    {
+      "epoch": 10.072261949567181,
+      "grad_norm": 0.18335990607738495,
+      "learning_rate": 5e-06,
+      "loss": 0.2102,
+      "step": 13381
+    },
+    {
+      "epoch": 10.073014678208505,
+      "grad_norm": 0.18344181776046753,
+      "learning_rate": 5e-06,
+      "loss": 0.1645,
+      "step": 13382
+    },
+    {
+      "epoch": 10.07376740684983,
+      "grad_norm": 0.18027175962924957,
+      "learning_rate": 5e-06,
+      "loss": 0.1707,
+      "step": 13383
+    },
+    {
+      "epoch": 10.074520135491156,
+      "grad_norm": 0.18206429481506348,
+      "learning_rate": 5e-06,
+      "loss": 0.1088,
+      "step": 13384
+    },
+    {
+      "epoch": 10.07527286413248,
+      "grad_norm": 0.18567712604999542,
+      "learning_rate": 5e-06,
+      "loss": 0.1369,
+      "step": 13385
+    },
+    {
+      "epoch": 10.076025592773805,
+      "grad_norm": 0.21432185173034668,
+      "learning_rate": 5e-06,
+      "loss": 0.2603,
+      "step": 13386
+    },
+    {
+      "epoch": 10.07677832141513,
+      "grad_norm": 0.1630665361881256,
+      "learning_rate": 5e-06,
+      "loss": 0.2213,
+      "step": 13387
+    },
+    {
+      "epoch": 10.077531050056455,
+      "grad_norm": 0.19565607607364655,
+      "learning_rate": 5e-06,
+      "loss": 0.1366,
+      "step": 13388
+    },
+    {
+      "epoch": 10.078283778697779,
+      "grad_norm": 0.14493544399738312,
+      "learning_rate": 5e-06,
+      "loss": 0.1265,
+      "step": 13389
+    },
+    {
+      "epoch": 10.079036507339104,
+      "grad_norm": 0.17690901458263397,
+      "learning_rate": 5e-06,
+      "loss": 0.0768,
+      "step": 13390
+    },
+    {
+      "epoch": 10.07978923598043,
+      "grad_norm": 0.19138772785663605,
+      "learning_rate": 5e-06,
+      "loss": 0.2193,
+      "step": 13391
+    },
+    {
+      "epoch": 10.080541964621753,
+      "grad_norm": 0.19313272833824158,
+      "learning_rate": 5e-06,
+      "loss": 0.1525,
+      "step": 13392
+    },
+    {
+      "epoch": 10.081294693263079,
+      "grad_norm": 0.17137731611728668,
+      "learning_rate": 5e-06,
+      "loss": 0.1161,
+      "step": 13393
+    },
+    {
+      "epoch": 10.082047421904404,
+      "grad_norm": 0.18495433032512665,
+      "learning_rate": 5e-06,
+      "loss": 0.1756,
+      "step": 13394
+    },
+    {
+      "epoch": 10.082800150545728,
+      "grad_norm": 0.19693398475646973,
+      "learning_rate": 5e-06,
+      "loss": 0.1675,
+      "step": 13395
+    },
+    {
+      "epoch": 10.083552879187053,
+      "grad_norm": 0.1908663958311081,
+      "learning_rate": 5e-06,
+      "loss": 0.1649,
+      "step": 13396
+    },
+    {
+      "epoch": 10.084305607828378,
+      "grad_norm": 0.18470512330532074,
+      "learning_rate": 5e-06,
+      "loss": 0.1877,
+      "step": 13397
+    },
+    {
+      "epoch": 10.085058336469702,
+      "grad_norm": 0.18690401315689087,
+      "learning_rate": 5e-06,
+      "loss": 0.1592,
+      "step": 13398
+    },
+    {
+      "epoch": 10.085811065111027,
+      "grad_norm": 0.22137537598609924,
+      "learning_rate": 5e-06,
+      "loss": 0.1375,
+      "step": 13399
+    },
+    {
+      "epoch": 10.086563793752353,
+      "grad_norm": 0.1965940147638321,
+      "learning_rate": 5e-06,
+      "loss": 0.1391,
+      "step": 13400
+    },
+    {
+      "epoch": 10.086563793752353,
+      "eval_loss": 0.18935222923755646,
+      "eval_runtime": 452.9855,
+      "eval_samples_per_second": 21.252,
+      "eval_steps_per_second": 0.664,
+      "step": 13400
+    },
+    {
+      "epoch": 10.087316522393676,
+      "grad_norm": 0.1752120852470398,
+      "learning_rate": 5e-06,
+      "loss": 0.2067,
+      "step": 13401
+    },
+    {
+      "epoch": 10.088069251035002,
+      "grad_norm": 0.18396688997745514,
+      "learning_rate": 5e-06,
+      "loss": 0.1557,
+      "step": 13402
+    },
+    {
+      "epoch": 10.088821979676327,
+      "grad_norm": 0.15846364200115204,
+      "learning_rate": 5e-06,
+      "loss": 0.1499,
+      "step": 13403
+    },
+    {
+      "epoch": 10.089574708317652,
+      "grad_norm": 0.18345606327056885,
+      "learning_rate": 5e-06,
+      "loss": 0.1338,
+      "step": 13404
+    },
+    {
+      "epoch": 10.090327436958976,
+      "grad_norm": 0.19596479833126068,
+      "learning_rate": 5e-06,
+      "loss": 0.1382,
+      "step": 13405
+    },
+    {
+      "epoch": 10.091080165600301,
+      "grad_norm": 0.16617736220359802,
+      "learning_rate": 5e-06,
+      "loss": 0.1538,
+      "step": 13406
+    },
+    {
+      "epoch": 10.091832894241627,
+      "grad_norm": 0.22337210178375244,
+      "learning_rate": 5e-06,
+      "loss": 0.155,
+      "step": 13407
+    },
+    {
+      "epoch": 10.09258562288295,
+      "grad_norm": 0.16521969437599182,
+      "learning_rate": 5e-06,
+      "loss": 0.1242,
+      "step": 13408
+    },
+    {
+      "epoch": 10.093338351524276,
+      "grad_norm": 0.1631004810333252,
+      "learning_rate": 5e-06,
+      "loss": 0.1419,
+      "step": 13409
+    },
+    {
+      "epoch": 10.094091080165601,
+      "grad_norm": 0.17572999000549316,
+      "learning_rate": 5e-06,
+      "loss": 0.1323,
+      "step": 13410
+    },
+    {
+      "epoch": 10.094843808806925,
+      "grad_norm": 0.18866799771785736,
+      "learning_rate": 5e-06,
+      "loss": 0.1673,
+      "step": 13411
+    },
+    {
+      "epoch": 10.09559653744825,
+      "grad_norm": 0.17116175591945648,
+      "learning_rate": 5e-06,
+      "loss": 0.1075,
+      "step": 13412
+    },
+    {
+      "epoch": 10.096349266089575,
+      "grad_norm": 0.15502329170703888,
+      "learning_rate": 5e-06,
+      "loss": 0.1795,
+      "step": 13413
+    },
+    {
+      "epoch": 10.097101994730899,
+      "grad_norm": 0.21933329105377197,
+      "learning_rate": 5e-06,
+      "loss": 0.1769,
+      "step": 13414
+    },
+    {
+      "epoch": 10.097854723372224,
+      "grad_norm": 0.1816486269235611,
+      "learning_rate": 5e-06,
+      "loss": 0.1371,
+      "step": 13415
+    },
+    {
+      "epoch": 10.09860745201355,
+      "grad_norm": 0.17694181203842163,
+      "learning_rate": 5e-06,
+      "loss": 0.1385,
+      "step": 13416
+    },
+    {
+      "epoch": 10.099360180654873,
+      "grad_norm": 0.1910516917705536,
+      "learning_rate": 5e-06,
+      "loss": 0.136,
+      "step": 13417
+    },
+    {
+      "epoch": 10.100112909296199,
+      "grad_norm": 0.18203186988830566,
+      "learning_rate": 5e-06,
+      "loss": 0.1516,
+      "step": 13418
+    },
+    {
+      "epoch": 10.100865637937524,
+      "grad_norm": 0.1958986073732376,
+      "learning_rate": 5e-06,
+      "loss": 0.2191,
+      "step": 13419
+    },
+    {
+      "epoch": 10.101618366578847,
+      "grad_norm": 0.19179221987724304,
+      "learning_rate": 5e-06,
+      "loss": 0.1307,
+      "step": 13420
+    },
+    {
+      "epoch": 10.102371095220173,
+      "grad_norm": 0.1412205547094345,
+      "learning_rate": 5e-06,
+      "loss": 0.1088,
+      "step": 13421
+    },
+    {
+      "epoch": 10.103123823861498,
+      "grad_norm": 0.18192753195762634,
+      "learning_rate": 5e-06,
+      "loss": 0.1441,
+      "step": 13422
+    },
+    {
+      "epoch": 10.103876552502824,
+      "grad_norm": 0.16633756458759308,
+      "learning_rate": 5e-06,
+      "loss": 0.1266,
+      "step": 13423
+    },
+    {
+      "epoch": 10.104629281144147,
+      "grad_norm": 0.1839495301246643,
+      "learning_rate": 5e-06,
+      "loss": 0.1001,
+      "step": 13424
+    },
+    {
+      "epoch": 10.105382009785473,
+      "grad_norm": 0.16399145126342773,
+      "learning_rate": 5e-06,
+      "loss": 0.0994,
+      "step": 13425
+    },
+    {
+      "epoch": 10.106134738426798,
+      "grad_norm": 0.1684437245130539,
+      "learning_rate": 5e-06,
+      "loss": 0.0839,
+      "step": 13426
+    },
+    {
+      "epoch": 10.106887467068121,
+      "grad_norm": 0.17962613701820374,
+      "learning_rate": 5e-06,
+      "loss": 0.1673,
+      "step": 13427
+    },
+    {
+      "epoch": 10.107640195709447,
+      "grad_norm": 0.16240304708480835,
+      "learning_rate": 5e-06,
+      "loss": 0.098,
+      "step": 13428
+    },
+    {
+      "epoch": 10.108392924350772,
+      "grad_norm": 0.14051850140094757,
+      "learning_rate": 5e-06,
+      "loss": 0.134,
+      "step": 13429
+    },
+    {
+      "epoch": 10.109145652992096,
+      "grad_norm": 0.18486922979354858,
+      "learning_rate": 5e-06,
+      "loss": 0.0974,
+      "step": 13430
+    },
+    {
+      "epoch": 10.109898381633421,
+      "grad_norm": 0.17544005811214447,
+      "learning_rate": 5e-06,
+      "loss": 0.1592,
+      "step": 13431
+    },
+    {
+      "epoch": 10.110651110274747,
+      "grad_norm": 0.17313814163208008,
+      "learning_rate": 5e-06,
+      "loss": 0.2182,
+      "step": 13432
+    },
+    {
+      "epoch": 10.11140383891607,
+      "grad_norm": 0.20014476776123047,
+      "learning_rate": 5e-06,
+      "loss": 0.1472,
+      "step": 13433
+    },
+    {
+      "epoch": 10.112156567557395,
+      "grad_norm": 0.1948850154876709,
+      "learning_rate": 5e-06,
+      "loss": 0.1742,
+      "step": 13434
+    },
+    {
+      "epoch": 10.11290929619872,
+      "grad_norm": 0.17303477227687836,
+      "learning_rate": 5e-06,
+      "loss": 0.1713,
+      "step": 13435
+    },
+    {
+      "epoch": 10.113662024840044,
+      "grad_norm": 0.1800646036863327,
+      "learning_rate": 5e-06,
+      "loss": 0.1435,
+      "step": 13436
+    },
+    {
+      "epoch": 10.11441475348137,
+      "grad_norm": 0.16412827372550964,
+      "learning_rate": 5e-06,
+      "loss": 0.0873,
+      "step": 13437
+    },
+    {
+      "epoch": 10.115167482122695,
+      "grad_norm": 0.17218531668186188,
+      "learning_rate": 5e-06,
+      "loss": 0.1734,
+      "step": 13438
+    },
+    {
+      "epoch": 10.115920210764019,
+      "grad_norm": 0.15459176898002625,
+      "learning_rate": 5e-06,
+      "loss": 0.1069,
+      "step": 13439
+    },
+    {
+      "epoch": 10.116672939405344,
+      "grad_norm": 0.16335561871528625,
+      "learning_rate": 5e-06,
+      "loss": 0.1329,
+      "step": 13440
+    },
+    {
+      "epoch": 10.11742566804667,
+      "grad_norm": 0.18411682546138763,
+      "learning_rate": 5e-06,
+      "loss": 0.1776,
+      "step": 13441
+    },
+    {
+      "epoch": 10.118178396687995,
+      "grad_norm": 0.18121518194675446,
+      "learning_rate": 5e-06,
+      "loss": 0.1552,
+      "step": 13442
+    },
+    {
+      "epoch": 10.118931125329318,
+      "grad_norm": 0.16242609918117523,
+      "learning_rate": 5e-06,
+      "loss": 0.0905,
+      "step": 13443
+    },
+    {
+      "epoch": 10.119683853970644,
+      "grad_norm": 0.1711452603340149,
+      "learning_rate": 5e-06,
+      "loss": 0.1325,
+      "step": 13444
+    },
+    {
+      "epoch": 10.120436582611969,
+      "grad_norm": 0.19923828542232513,
+      "learning_rate": 5e-06,
+      "loss": 0.0885,
+      "step": 13445
+    },
+    {
+      "epoch": 10.121189311253293,
+      "grad_norm": 0.18581627309322357,
+      "learning_rate": 5e-06,
+      "loss": 0.1323,
+      "step": 13446
+    },
+    {
+      "epoch": 10.121942039894618,
+      "grad_norm": 0.17776243388652802,
+      "learning_rate": 5e-06,
+      "loss": 0.1542,
+      "step": 13447
+    },
+    {
+      "epoch": 10.122694768535943,
+      "grad_norm": 0.19290786981582642,
+      "learning_rate": 5e-06,
+      "loss": 0.1305,
+      "step": 13448
+    },
+    {
+      "epoch": 10.123447497177267,
+      "grad_norm": 0.1975339651107788,
+      "learning_rate": 5e-06,
+      "loss": 0.1964,
+      "step": 13449
+    },
+    {
+      "epoch": 10.124200225818592,
+      "grad_norm": 0.18885862827301025,
+      "learning_rate": 5e-06,
+      "loss": 0.178,
+      "step": 13450
+    },
+    {
+      "epoch": 10.124952954459918,
+      "grad_norm": 0.1758224368095398,
+      "learning_rate": 5e-06,
+      "loss": 0.1702,
+      "step": 13451
+    },
+    {
+      "epoch": 10.125705683101241,
+      "grad_norm": 0.1946447193622589,
+      "learning_rate": 5e-06,
+      "loss": 0.1462,
+      "step": 13452
+    },
+    {
+      "epoch": 10.126458411742567,
+      "grad_norm": 0.20298954844474792,
+      "learning_rate": 5e-06,
+      "loss": 0.1454,
+      "step": 13453
+    },
+    {
+      "epoch": 10.127211140383892,
+      "grad_norm": 0.1768442541360855,
+      "learning_rate": 5e-06,
+      "loss": 0.1335,
+      "step": 13454
+    },
+    {
+      "epoch": 10.127963869025216,
+      "grad_norm": 0.20647135376930237,
+      "learning_rate": 5e-06,
+      "loss": 0.1425,
+      "step": 13455
+    },
+    {
+      "epoch": 10.128716597666541,
+      "grad_norm": 0.17303314805030823,
+      "learning_rate": 5e-06,
+      "loss": 0.1182,
+      "step": 13456
+    },
+    {
+      "epoch": 10.129469326307866,
+      "grad_norm": 0.17151041328907013,
+      "learning_rate": 5e-06,
+      "loss": 0.1424,
+      "step": 13457
+    },
+    {
+      "epoch": 10.13022205494919,
+      "grad_norm": 0.16388683021068573,
+      "learning_rate": 5e-06,
+      "loss": 0.1505,
+      "step": 13458
+    },
+    {
+      "epoch": 10.130974783590515,
+      "grad_norm": 0.18761706352233887,
+      "learning_rate": 5e-06,
+      "loss": 0.1809,
+      "step": 13459
+    },
+    {
+      "epoch": 10.13172751223184,
+      "grad_norm": 0.17856405675411224,
+      "learning_rate": 5e-06,
+      "loss": 0.1639,
+      "step": 13460
+    },
+    {
+      "epoch": 10.132480240873166,
+      "grad_norm": 0.17128774523735046,
+      "learning_rate": 5e-06,
+      "loss": 0.1978,
+      "step": 13461
+    },
+    {
+      "epoch": 10.13323296951449,
+      "grad_norm": 0.18795214593410492,
+      "learning_rate": 5e-06,
+      "loss": 0.1,
+      "step": 13462
+    },
+    {
+      "epoch": 10.133985698155815,
+      "grad_norm": 0.18282687664031982,
+      "learning_rate": 5e-06,
+      "loss": 0.1347,
+      "step": 13463
+    },
+    {
+      "epoch": 10.13473842679714,
+      "grad_norm": 0.1709667146205902,
+      "learning_rate": 5e-06,
+      "loss": 0.0891,
+      "step": 13464
+    },
+    {
+      "epoch": 10.135491155438464,
+      "grad_norm": 0.1683989316225052,
+      "learning_rate": 5e-06,
+      "loss": 0.1901,
+      "step": 13465
+    },
+    {
+      "epoch": 10.13624388407979,
+      "grad_norm": 0.20951680839061737,
+      "learning_rate": 5e-06,
+      "loss": 0.1498,
+      "step": 13466
+    },
+    {
+      "epoch": 10.136996612721115,
+      "grad_norm": 0.18151874840259552,
+      "learning_rate": 5e-06,
+      "loss": 0.1926,
+      "step": 13467
+    },
+    {
+      "epoch": 10.137749341362438,
+      "grad_norm": 0.18392528593540192,
+      "learning_rate": 5e-06,
+      "loss": 0.1861,
+      "step": 13468
+    },
+    {
+      "epoch": 10.138502070003764,
+      "grad_norm": 0.1803014874458313,
+      "learning_rate": 5e-06,
+      "loss": 0.1732,
+      "step": 13469
+    },
+    {
+      "epoch": 10.139254798645089,
+      "grad_norm": 0.18637758493423462,
+      "learning_rate": 5e-06,
+      "loss": 0.1623,
+      "step": 13470
+    },
+    {
+      "epoch": 10.140007527286413,
+      "grad_norm": 0.17773319780826569,
+      "learning_rate": 5e-06,
+      "loss": 0.104,
+      "step": 13471
+    },
+    {
+      "epoch": 10.140760255927738,
+      "grad_norm": 0.1729484647512436,
+      "learning_rate": 5e-06,
+      "loss": 0.1574,
+      "step": 13472
+    },
+    {
+      "epoch": 10.141512984569063,
+      "grad_norm": 0.1714978963136673,
+      "learning_rate": 5e-06,
+      "loss": 0.1327,
+      "step": 13473
+    },
+    {
+      "epoch": 10.142265713210387,
+      "grad_norm": 0.2066894918680191,
+      "learning_rate": 5e-06,
+      "loss": 0.1497,
+      "step": 13474
+    },
+    {
+      "epoch": 10.143018441851712,
+      "grad_norm": 0.17545504868030548,
+      "learning_rate": 5e-06,
+      "loss": 0.1439,
+      "step": 13475
+    },
+    {
+      "epoch": 10.143771170493038,
+      "grad_norm": 0.16790008544921875,
+      "learning_rate": 5e-06,
+      "loss": 0.1631,
+      "step": 13476
+    },
+    {
+      "epoch": 10.144523899134363,
+      "grad_norm": 0.16385814547538757,
+      "learning_rate": 5e-06,
+      "loss": 0.1009,
+      "step": 13477
+    },
+    {
+      "epoch": 10.145276627775687,
+      "grad_norm": 0.153341144323349,
+      "learning_rate": 5e-06,
+      "loss": 0.123,
+      "step": 13478
+    },
+    {
+      "epoch": 10.146029356417012,
+      "grad_norm": 0.20665913820266724,
+      "learning_rate": 5e-06,
+      "loss": 0.1187,
+      "step": 13479
+    },
+    {
+      "epoch": 10.146782085058337,
+      "grad_norm": 0.18895433843135834,
+      "learning_rate": 5e-06,
+      "loss": 0.1474,
+      "step": 13480
+    },
+    {
+      "epoch": 10.14753481369966,
+      "grad_norm": 0.19189704954624176,
+      "learning_rate": 5e-06,
+      "loss": 0.1779,
+      "step": 13481
+    },
+    {
+      "epoch": 10.148287542340986,
+      "grad_norm": 0.15350666642189026,
+      "learning_rate": 5e-06,
+      "loss": 0.0862,
+      "step": 13482
+    },
+    {
+      "epoch": 10.149040270982312,
+      "grad_norm": 0.18721739947795868,
+      "learning_rate": 5e-06,
+      "loss": 0.1367,
+      "step": 13483
+    },
+    {
+      "epoch": 10.149792999623635,
+      "grad_norm": 0.1744260936975479,
+      "learning_rate": 5e-06,
+      "loss": 0.1323,
+      "step": 13484
+    },
+    {
+      "epoch": 10.15054572826496,
+      "grad_norm": 0.1888989508152008,
+      "learning_rate": 5e-06,
+      "loss": 0.1396,
+      "step": 13485
+    },
+    {
+      "epoch": 10.151298456906286,
+      "grad_norm": 0.1733049601316452,
+      "learning_rate": 5e-06,
+      "loss": 0.0939,
+      "step": 13486
+    },
+    {
+      "epoch": 10.15205118554761,
+      "grad_norm": 0.17272010445594788,
+      "learning_rate": 5e-06,
+      "loss": 0.1835,
+      "step": 13487
+    },
+    {
+      "epoch": 10.152803914188935,
+      "grad_norm": 0.17803676426410675,
+      "learning_rate": 5e-06,
+      "loss": 0.1805,
+      "step": 13488
+    },
+    {
+      "epoch": 10.15355664283026,
+      "grad_norm": 0.17459408938884735,
+      "learning_rate": 5e-06,
+      "loss": 0.1796,
+      "step": 13489
+    },
+    {
+      "epoch": 10.154309371471584,
+      "grad_norm": 0.156485453248024,
+      "learning_rate": 5e-06,
+      "loss": 0.1623,
+      "step": 13490
+    },
+    {
+      "epoch": 10.155062100112909,
+      "grad_norm": 0.1700584441423416,
+      "learning_rate": 5e-06,
+      "loss": 0.0681,
+      "step": 13491
+    },
+    {
+      "epoch": 10.155814828754234,
+      "grad_norm": 0.16317011415958405,
+      "learning_rate": 5e-06,
+      "loss": 0.1042,
+      "step": 13492
+    },
+    {
+      "epoch": 10.156567557395558,
+      "grad_norm": 0.16132202744483948,
+      "learning_rate": 5e-06,
+      "loss": 0.2263,
+      "step": 13493
+    },
+    {
+      "epoch": 10.157320286036883,
+      "grad_norm": 0.16036805510520935,
+      "learning_rate": 5e-06,
+      "loss": 0.1364,
+      "step": 13494
+    },
+    {
+      "epoch": 10.158073014678209,
+      "grad_norm": 0.1951083540916443,
+      "learning_rate": 5e-06,
+      "loss": 0.1116,
+      "step": 13495
+    },
+    {
+      "epoch": 10.158825743319534,
+      "grad_norm": 0.1791706383228302,
+      "learning_rate": 5e-06,
+      "loss": 0.1289,
+      "step": 13496
+    },
+    {
+      "epoch": 10.159578471960858,
+      "grad_norm": 0.19414520263671875,
+      "learning_rate": 5e-06,
+      "loss": 0.1156,
+      "step": 13497
+    },
+    {
+      "epoch": 10.160331200602183,
+      "grad_norm": 0.1767050325870514,
+      "learning_rate": 5e-06,
+      "loss": 0.1246,
+      "step": 13498
+    },
+    {
+      "epoch": 10.161083929243508,
+      "grad_norm": 0.202445387840271,
+      "learning_rate": 5e-06,
+      "loss": 0.1529,
+      "step": 13499
+    },
+    {
+      "epoch": 10.161836657884832,
+      "grad_norm": 0.19281326234340668,
+      "learning_rate": 5e-06,
+      "loss": 0.094,
+      "step": 13500
+    },
+    {
+      "epoch": 10.162589386526157,
+      "grad_norm": 0.16162113845348358,
+      "learning_rate": 5e-06,
+      "loss": 0.1709,
+      "step": 13501
+    },
+    {
+      "epoch": 10.163342115167483,
+      "grad_norm": 0.17779184877872467,
+      "learning_rate": 5e-06,
+      "loss": 0.1504,
+      "step": 13502
+    },
+    {
+      "epoch": 10.164094843808806,
+      "grad_norm": 0.14437392354011536,
+      "learning_rate": 5e-06,
+      "loss": 0.1863,
+      "step": 13503
+    },
+    {
+      "epoch": 10.164847572450132,
+      "grad_norm": 0.15762975811958313,
+      "learning_rate": 5e-06,
+      "loss": 0.1577,
+      "step": 13504
+    },
+    {
+      "epoch": 10.165600301091457,
+      "grad_norm": 0.17454484105110168,
+      "learning_rate": 5e-06,
+      "loss": 0.1726,
+      "step": 13505
+    },
+    {
+      "epoch": 10.16635302973278,
+      "grad_norm": 0.18612244725227356,
+      "learning_rate": 5e-06,
+      "loss": 0.144,
+      "step": 13506
+    },
+    {
+      "epoch": 10.167105758374106,
+      "grad_norm": 0.15561960637569427,
+      "learning_rate": 5e-06,
+      "loss": 0.1393,
+      "step": 13507
+    },
+    {
+      "epoch": 10.167858487015431,
+      "grad_norm": 0.18564850091934204,
+      "learning_rate": 5e-06,
+      "loss": 0.169,
+      "step": 13508
+    },
+    {
+      "epoch": 10.168611215656755,
+      "grad_norm": 0.1791832149028778,
+      "learning_rate": 5e-06,
+      "loss": 0.1894,
+      "step": 13509
+    },
+    {
+      "epoch": 10.16936394429808,
+      "grad_norm": 0.1797928512096405,
+      "learning_rate": 5e-06,
+      "loss": 0.1288,
+      "step": 13510
+    },
+    {
+      "epoch": 10.170116672939406,
+      "grad_norm": 0.19475817680358887,
+      "learning_rate": 5e-06,
+      "loss": 0.1037,
+      "step": 13511
+    },
+    {
+      "epoch": 10.17086940158073,
+      "grad_norm": 0.16883298754692078,
+      "learning_rate": 5e-06,
+      "loss": 0.1055,
+      "step": 13512
+    },
+    {
+      "epoch": 10.171622130222055,
+      "grad_norm": 0.19085349142551422,
+      "learning_rate": 5e-06,
+      "loss": 0.1338,
+      "step": 13513
+    },
+    {
+      "epoch": 10.17237485886338,
+      "grad_norm": 0.17843428254127502,
+      "learning_rate": 5e-06,
+      "loss": 0.1618,
+      "step": 13514
+    },
+    {
+      "epoch": 10.173127587504705,
+      "grad_norm": 0.18013125658035278,
+      "learning_rate": 5e-06,
+      "loss": 0.1455,
+      "step": 13515
+    },
+    {
+      "epoch": 10.173880316146029,
+      "grad_norm": 0.16357475519180298,
+      "learning_rate": 5e-06,
+      "loss": 0.1416,
+      "step": 13516
+    },
+    {
+      "epoch": 10.174633044787354,
+      "grad_norm": 0.14091342687606812,
+      "learning_rate": 5e-06,
+      "loss": 0.1245,
+      "step": 13517
+    },
+    {
+      "epoch": 10.17538577342868,
+      "grad_norm": 0.181215301156044,
+      "learning_rate": 5e-06,
+      "loss": 0.139,
+      "step": 13518
+    },
+    {
+      "epoch": 10.176138502070003,
+      "grad_norm": 0.17975811660289764,
+      "learning_rate": 5e-06,
+      "loss": 0.1436,
+      "step": 13519
+    },
+    {
+      "epoch": 10.176891230711329,
+      "grad_norm": 0.16583740711212158,
+      "learning_rate": 5e-06,
+      "loss": 0.1757,
+      "step": 13520
+    },
+    {
+      "epoch": 10.177643959352654,
+      "grad_norm": 0.16871362924575806,
+      "learning_rate": 5e-06,
+      "loss": 0.0813,
+      "step": 13521
+    },
+    {
+      "epoch": 10.178396687993978,
+      "grad_norm": 0.17332476377487183,
+      "learning_rate": 5e-06,
+      "loss": 0.1073,
+      "step": 13522
+    },
+    {
+      "epoch": 10.179149416635303,
+      "grad_norm": 0.16204170882701874,
+      "learning_rate": 5e-06,
+      "loss": 0.1692,
+      "step": 13523
+    },
+    {
+      "epoch": 10.179902145276628,
+      "grad_norm": 0.2033138871192932,
+      "learning_rate": 5e-06,
+      "loss": 0.1298,
+      "step": 13524
+    },
+    {
+      "epoch": 10.180654873917952,
+      "grad_norm": 0.17736993730068207,
+      "learning_rate": 5e-06,
+      "loss": 0.1181,
+      "step": 13525
+    },
+    {
+      "epoch": 10.181407602559277,
+      "grad_norm": 0.16700199246406555,
+      "learning_rate": 5e-06,
+      "loss": 0.1213,
+      "step": 13526
+    },
+    {
+      "epoch": 10.182160331200603,
+      "grad_norm": 0.15763452649116516,
+      "learning_rate": 5e-06,
+      "loss": 0.0951,
+      "step": 13527
+    },
+    {
+      "epoch": 10.182913059841926,
+      "grad_norm": 0.18177810311317444,
+      "learning_rate": 5e-06,
+      "loss": 0.1309,
+      "step": 13528
+    },
+    {
+      "epoch": 10.183665788483252,
+      "grad_norm": 0.18409818410873413,
+      "learning_rate": 5e-06,
+      "loss": 0.184,
+      "step": 13529
+    },
+    {
+      "epoch": 10.184418517124577,
+      "grad_norm": 0.1857498735189438,
+      "learning_rate": 5e-06,
+      "loss": 0.1565,
+      "step": 13530
+    },
+    {
+      "epoch": 10.1851712457659,
+      "grad_norm": 0.1880352646112442,
+      "learning_rate": 5e-06,
+      "loss": 0.1978,
+      "step": 13531
+    },
+    {
+      "epoch": 10.185923974407226,
+      "grad_norm": 0.19455505907535553,
+      "learning_rate": 5e-06,
+      "loss": 0.1563,
+      "step": 13532
+    },
+    {
+      "epoch": 10.186676703048551,
+      "grad_norm": 0.17987264692783356,
+      "learning_rate": 5e-06,
+      "loss": 0.1435,
+      "step": 13533
+    },
+    {
+      "epoch": 10.187429431689877,
+      "grad_norm": 0.17249268293380737,
+      "learning_rate": 5e-06,
+      "loss": 0.0843,
+      "step": 13534
+    },
+    {
+      "epoch": 10.1881821603312,
+      "grad_norm": 0.1891711801290512,
+      "learning_rate": 5e-06,
+      "loss": 0.1791,
+      "step": 13535
+    },
+    {
+      "epoch": 10.188934888972526,
+      "grad_norm": 0.18733493983745575,
+      "learning_rate": 5e-06,
+      "loss": 0.202,
+      "step": 13536
+    },
+    {
+      "epoch": 10.189687617613851,
+      "grad_norm": 0.18435949087142944,
+      "learning_rate": 5e-06,
+      "loss": 0.1616,
+      "step": 13537
+    },
+    {
+      "epoch": 10.190440346255174,
+      "grad_norm": 0.16372114419937134,
+      "learning_rate": 5e-06,
+      "loss": 0.1249,
+      "step": 13538
+    },
+    {
+      "epoch": 10.1911930748965,
+      "grad_norm": 0.18070736527442932,
+      "learning_rate": 5e-06,
+      "loss": 0.1178,
+      "step": 13539
+    },
+    {
+      "epoch": 10.191945803537825,
+      "grad_norm": 0.18156808614730835,
+      "learning_rate": 5e-06,
+      "loss": 0.2377,
+      "step": 13540
+    },
+    {
+      "epoch": 10.192698532179149,
+      "grad_norm": 0.18919357657432556,
+      "learning_rate": 5e-06,
+      "loss": 0.1472,
+      "step": 13541
+    },
+    {
+      "epoch": 10.193451260820474,
+      "grad_norm": 0.16905304789543152,
+      "learning_rate": 5e-06,
+      "loss": 0.1201,
+      "step": 13542
+    },
+    {
+      "epoch": 10.1942039894618,
+      "grad_norm": 0.17141494154930115,
+      "learning_rate": 5e-06,
+      "loss": 0.1879,
+      "step": 13543
+    },
+    {
+      "epoch": 10.194956718103123,
+      "grad_norm": 0.17107391357421875,
+      "learning_rate": 5e-06,
+      "loss": 0.1993,
+      "step": 13544
+    },
+    {
+      "epoch": 10.195709446744448,
+      "grad_norm": 0.21142441034317017,
+      "learning_rate": 5e-06,
+      "loss": 0.1139,
+      "step": 13545
+    },
+    {
+      "epoch": 10.196462175385774,
+      "grad_norm": 0.1924089640378952,
+      "learning_rate": 5e-06,
+      "loss": 0.1976,
+      "step": 13546
+    },
+    {
+      "epoch": 10.197214904027097,
+      "grad_norm": 0.1745571345090866,
+      "learning_rate": 5e-06,
+      "loss": 0.1002,
+      "step": 13547
+    },
+    {
+      "epoch": 10.197967632668423,
+      "grad_norm": 0.17307831346988678,
+      "learning_rate": 5e-06,
+      "loss": 0.1306,
+      "step": 13548
+    },
+    {
+      "epoch": 10.198720361309748,
+      "grad_norm": 0.1602180451154709,
+      "learning_rate": 5e-06,
+      "loss": 0.1122,
+      "step": 13549
+    },
+    {
+      "epoch": 10.199473089951073,
+      "grad_norm": 0.1788467913866043,
+      "learning_rate": 5e-06,
+      "loss": 0.12,
+      "step": 13550
+    },
+    {
+      "epoch": 10.200225818592397,
+      "grad_norm": 0.18091291189193726,
+      "learning_rate": 5e-06,
+      "loss": 0.1196,
+      "step": 13551
+    },
+    {
+      "epoch": 10.200978547233722,
+      "grad_norm": 0.18924270570278168,
+      "learning_rate": 5e-06,
+      "loss": 0.1434,
+      "step": 13552
+    },
+    {
+      "epoch": 10.201731275875048,
+      "grad_norm": 0.17827972769737244,
+      "learning_rate": 5e-06,
+      "loss": 0.1504,
+      "step": 13553
+    },
+    {
+      "epoch": 10.202484004516371,
+      "grad_norm": 0.17990967631340027,
+      "learning_rate": 5e-06,
+      "loss": 0.1206,
+      "step": 13554
+    },
+    {
+      "epoch": 10.203236733157697,
+      "grad_norm": 0.16086047887802124,
+      "learning_rate": 5e-06,
+      "loss": 0.165,
+      "step": 13555
+    },
+    {
+      "epoch": 10.203989461799022,
+      "grad_norm": 0.17751586437225342,
+      "learning_rate": 5e-06,
+      "loss": 0.2038,
+      "step": 13556
+    },
+    {
+      "epoch": 10.204742190440346,
+      "grad_norm": 0.15505722165107727,
+      "learning_rate": 5e-06,
+      "loss": 0.0969,
+      "step": 13557
+    },
+    {
+      "epoch": 10.205494919081671,
+      "grad_norm": 0.17500866949558258,
+      "learning_rate": 5e-06,
+      "loss": 0.1481,
+      "step": 13558
+    },
+    {
+      "epoch": 10.206247647722996,
+      "grad_norm": 0.17931094765663147,
+      "learning_rate": 5e-06,
+      "loss": 0.1678,
+      "step": 13559
+    },
+    {
+      "epoch": 10.20700037636432,
+      "grad_norm": 0.19209548830986023,
+      "learning_rate": 5e-06,
+      "loss": 0.1832,
+      "step": 13560
+    },
+    {
+      "epoch": 10.207753105005645,
+      "grad_norm": 0.1636185646057129,
+      "learning_rate": 5e-06,
+      "loss": 0.0928,
+      "step": 13561
+    },
+    {
+      "epoch": 10.20850583364697,
+      "grad_norm": 0.15814295411109924,
+      "learning_rate": 5e-06,
+      "loss": 0.1884,
+      "step": 13562
+    },
+    {
+      "epoch": 10.209258562288294,
+      "grad_norm": 0.1464461088180542,
+      "learning_rate": 5e-06,
+      "loss": 0.1003,
+      "step": 13563
+    },
+    {
+      "epoch": 10.21001129092962,
+      "grad_norm": 0.1786775290966034,
+      "learning_rate": 5e-06,
+      "loss": 0.0709,
+      "step": 13564
+    },
+    {
+      "epoch": 10.210764019570945,
+      "grad_norm": 0.1671760529279709,
+      "learning_rate": 5e-06,
+      "loss": 0.1263,
+      "step": 13565
+    },
+    {
+      "epoch": 10.211516748212269,
+      "grad_norm": 0.19761665165424347,
+      "learning_rate": 5e-06,
+      "loss": 0.1393,
+      "step": 13566
+    },
+    {
+      "epoch": 10.212269476853594,
+      "grad_norm": 0.16367368400096893,
+      "learning_rate": 5e-06,
+      "loss": 0.0925,
+      "step": 13567
+    },
+    {
+      "epoch": 10.21302220549492,
+      "grad_norm": 0.17140308022499084,
+      "learning_rate": 5e-06,
+      "loss": 0.1526,
+      "step": 13568
+    },
+    {
+      "epoch": 10.213774934136245,
+      "grad_norm": 0.16091111302375793,
+      "learning_rate": 5e-06,
+      "loss": 0.1265,
+      "step": 13569
+    },
+    {
+      "epoch": 10.214527662777568,
+      "grad_norm": 0.17034697532653809,
+      "learning_rate": 5e-06,
+      "loss": 0.1155,
+      "step": 13570
+    },
+    {
+      "epoch": 10.215280391418894,
+      "grad_norm": 0.17702971398830414,
+      "learning_rate": 5e-06,
+      "loss": 0.1695,
+      "step": 13571
+    },
+    {
+      "epoch": 10.216033120060219,
+      "grad_norm": 0.18973861634731293,
+      "learning_rate": 5e-06,
+      "loss": 0.1746,
+      "step": 13572
+    },
+    {
+      "epoch": 10.216785848701543,
+      "grad_norm": 0.17669852077960968,
+      "learning_rate": 5e-06,
+      "loss": 0.116,
+      "step": 13573
+    },
+    {
+      "epoch": 10.217538577342868,
+      "grad_norm": 0.16207504272460938,
+      "learning_rate": 5e-06,
+      "loss": 0.1014,
+      "step": 13574
+    },
+    {
+      "epoch": 10.218291305984193,
+      "grad_norm": 0.1993863731622696,
+      "learning_rate": 5e-06,
+      "loss": 0.1349,
+      "step": 13575
+    },
+    {
+      "epoch": 10.219044034625517,
+      "grad_norm": 0.17054437100887299,
+      "learning_rate": 5e-06,
+      "loss": 0.1234,
+      "step": 13576
+    },
+    {
+      "epoch": 10.219796763266842,
+      "grad_norm": 0.15590517222881317,
+      "learning_rate": 5e-06,
+      "loss": 0.1247,
+      "step": 13577
+    },
+    {
+      "epoch": 10.220549491908168,
+      "grad_norm": 0.18715418875217438,
+      "learning_rate": 5e-06,
+      "loss": 0.164,
+      "step": 13578
+    },
+    {
+      "epoch": 10.221302220549491,
+      "grad_norm": 0.2083241045475006,
+      "learning_rate": 5e-06,
+      "loss": 0.1454,
+      "step": 13579
+    },
+    {
+      "epoch": 10.222054949190817,
+      "grad_norm": 0.1800081431865692,
+      "learning_rate": 5e-06,
+      "loss": 0.1382,
+      "step": 13580
+    },
+    {
+      "epoch": 10.222807677832142,
+      "grad_norm": 0.16057439148426056,
+      "learning_rate": 5e-06,
+      "loss": 0.122,
+      "step": 13581
+    },
+    {
+      "epoch": 10.223560406473466,
+      "grad_norm": 0.17814487218856812,
+      "learning_rate": 5e-06,
+      "loss": 0.123,
+      "step": 13582
+    },
+    {
+      "epoch": 10.224313135114791,
+      "grad_norm": 0.1659562587738037,
+      "learning_rate": 5e-06,
+      "loss": 0.1099,
+      "step": 13583
+    },
+    {
+      "epoch": 10.225065863756116,
+      "grad_norm": 0.17105498909950256,
+      "learning_rate": 5e-06,
+      "loss": 0.1394,
+      "step": 13584
+    },
+    {
+      "epoch": 10.225818592397442,
+      "grad_norm": 0.15430980920791626,
+      "learning_rate": 5e-06,
+      "loss": 0.0978,
+      "step": 13585
+    },
+    {
+      "epoch": 10.226571321038765,
+      "grad_norm": 0.18354086577892303,
+      "learning_rate": 5e-06,
+      "loss": 0.148,
+      "step": 13586
+    },
+    {
+      "epoch": 10.22732404968009,
+      "grad_norm": 0.20183736085891724,
+      "learning_rate": 5e-06,
+      "loss": 0.1565,
+      "step": 13587
+    },
+    {
+      "epoch": 10.228076778321416,
+      "grad_norm": 0.17656537890434265,
+      "learning_rate": 5e-06,
+      "loss": 0.1632,
+      "step": 13588
+    },
+    {
+      "epoch": 10.22882950696274,
+      "grad_norm": 0.20881876349449158,
+      "learning_rate": 5e-06,
+      "loss": 0.1252,
+      "step": 13589
+    },
+    {
+      "epoch": 10.229582235604065,
+      "grad_norm": 0.15785349905490875,
+      "learning_rate": 5e-06,
+      "loss": 0.1015,
+      "step": 13590
+    },
+    {
+      "epoch": 10.23033496424539,
+      "grad_norm": 0.19451308250427246,
+      "learning_rate": 5e-06,
+      "loss": 0.1419,
+      "step": 13591
+    },
+    {
+      "epoch": 10.231087692886714,
+      "grad_norm": 0.18701809644699097,
+      "learning_rate": 5e-06,
+      "loss": 0.2166,
+      "step": 13592
+    },
+    {
+      "epoch": 10.23184042152804,
+      "grad_norm": 0.173284113407135,
+      "learning_rate": 5e-06,
+      "loss": 0.1459,
+      "step": 13593
+    },
+    {
+      "epoch": 10.232593150169365,
+      "grad_norm": 0.17978470027446747,
+      "learning_rate": 5e-06,
+      "loss": 0.1788,
+      "step": 13594
+    },
+    {
+      "epoch": 10.233345878810688,
+      "grad_norm": 0.18608778715133667,
+      "learning_rate": 5e-06,
+      "loss": 0.1219,
+      "step": 13595
+    },
+    {
+      "epoch": 10.234098607452013,
+      "grad_norm": 0.15408669412136078,
+      "learning_rate": 5e-06,
+      "loss": 0.1346,
+      "step": 13596
+    },
+    {
+      "epoch": 10.234851336093339,
+      "grad_norm": 0.18519197404384613,
+      "learning_rate": 5e-06,
+      "loss": 0.156,
+      "step": 13597
+    },
+    {
+      "epoch": 10.235604064734662,
+      "grad_norm": 0.21463923156261444,
+      "learning_rate": 5e-06,
+      "loss": 0.175,
+      "step": 13598
+    },
+    {
+      "epoch": 10.236356793375988,
+      "grad_norm": 0.18304230272769928,
+      "learning_rate": 5e-06,
+      "loss": 0.113,
+      "step": 13599
+    },
+    {
+      "epoch": 10.237109522017313,
+      "grad_norm": 0.14667505025863647,
+      "learning_rate": 5e-06,
+      "loss": 0.0756,
+      "step": 13600
+    },
+    {
+      "epoch": 10.237109522017313,
+      "eval_loss": 0.1887189894914627,
+      "eval_runtime": 452.9005,
+      "eval_samples_per_second": 21.256,
+      "eval_steps_per_second": 0.665,
+      "step": 13600
+    },
+    {
+      "epoch": 10.237862250658637,
+      "grad_norm": 0.15892234444618225,
+      "learning_rate": 5e-06,
+      "loss": 0.1162,
+      "step": 13601
+    },
+    {
+      "epoch": 10.238614979299962,
+      "grad_norm": 0.15701767802238464,
+      "learning_rate": 5e-06,
+      "loss": 0.1124,
+      "step": 13602
+    },
+    {
+      "epoch": 10.239367707941287,
+      "grad_norm": 0.1931447684764862,
+      "learning_rate": 5e-06,
+      "loss": 0.1147,
+      "step": 13603
+    },
+    {
+      "epoch": 10.240120436582613,
+      "grad_norm": 0.17082850635051727,
+      "learning_rate": 5e-06,
+      "loss": 0.1365,
+      "step": 13604
+    },
+    {
+      "epoch": 10.240873165223936,
+      "grad_norm": 0.18531611561775208,
+      "learning_rate": 5e-06,
+      "loss": 0.1342,
+      "step": 13605
+    },
+    {
+      "epoch": 10.241625893865262,
+      "grad_norm": 0.17647919058799744,
+      "learning_rate": 5e-06,
+      "loss": 0.1201,
+      "step": 13606
+    },
+    {
+      "epoch": 10.242378622506587,
+      "grad_norm": 0.1740003079175949,
+      "learning_rate": 5e-06,
+      "loss": 0.1305,
+      "step": 13607
+    },
+    {
+      "epoch": 10.24313135114791,
+      "grad_norm": 0.1797844022512436,
+      "learning_rate": 5e-06,
+      "loss": 0.1281,
+      "step": 13608
+    },
+    {
+      "epoch": 10.243884079789236,
+      "grad_norm": 0.1587996631860733,
+      "learning_rate": 5e-06,
+      "loss": 0.1198,
+      "step": 13609
+    },
+    {
+      "epoch": 10.244636808430561,
+      "grad_norm": 0.1994905173778534,
+      "learning_rate": 5e-06,
+      "loss": 0.1317,
+      "step": 13610
+    },
+    {
+      "epoch": 10.245389537071885,
+      "grad_norm": 0.15402504801750183,
+      "learning_rate": 5e-06,
+      "loss": 0.1385,
+      "step": 13611
+    },
+    {
+      "epoch": 10.24614226571321,
+      "grad_norm": 0.20545795559883118,
+      "learning_rate": 5e-06,
+      "loss": 0.197,
+      "step": 13612
+    },
+    {
+      "epoch": 10.246894994354536,
+      "grad_norm": 0.18453702330589294,
+      "learning_rate": 5e-06,
+      "loss": 0.1542,
+      "step": 13613
+    },
+    {
+      "epoch": 10.24764772299586,
+      "grad_norm": 0.16818802058696747,
+      "learning_rate": 5e-06,
+      "loss": 0.0954,
+      "step": 13614
+    },
+    {
+      "epoch": 10.248400451637185,
+      "grad_norm": 0.16506828367710114,
+      "learning_rate": 5e-06,
+      "loss": 0.1556,
+      "step": 13615
+    },
+    {
+      "epoch": 10.24915318027851,
+      "grad_norm": 0.2104124277830124,
+      "learning_rate": 5e-06,
+      "loss": 0.0835,
+      "step": 13616
+    },
+    {
+      "epoch": 10.249905908919834,
+      "grad_norm": 0.20042537152767181,
+      "learning_rate": 5e-06,
+      "loss": 0.2117,
+      "step": 13617
+    },
+    {
+      "epoch": 10.250658637561159,
+      "grad_norm": 0.15572653710842133,
+      "learning_rate": 5e-06,
+      "loss": 0.068,
+      "step": 13618
+    },
+    {
+      "epoch": 10.251411366202484,
+      "grad_norm": 0.1647142618894577,
+      "learning_rate": 5e-06,
+      "loss": 0.1411,
+      "step": 13619
+    },
+    {
+      "epoch": 10.252164094843808,
+      "grad_norm": 0.19408254325389862,
+      "learning_rate": 5e-06,
+      "loss": 0.1582,
+      "step": 13620
+    },
+    {
+      "epoch": 10.252916823485133,
+      "grad_norm": 0.1723993867635727,
+      "learning_rate": 5e-06,
+      "loss": 0.142,
+      "step": 13621
+    },
+    {
+      "epoch": 10.253669552126459,
+      "grad_norm": 0.17653045058250427,
+      "learning_rate": 5e-06,
+      "loss": 0.1872,
+      "step": 13622
+    },
+    {
+      "epoch": 10.254422280767784,
+      "grad_norm": 0.21062378585338593,
+      "learning_rate": 5e-06,
+      "loss": 0.1457,
+      "step": 13623
+    },
+    {
+      "epoch": 10.255175009409108,
+      "grad_norm": 0.2027885615825653,
+      "learning_rate": 5e-06,
+      "loss": 0.1672,
+      "step": 13624
+    },
+    {
+      "epoch": 10.255927738050433,
+      "grad_norm": 0.17974066734313965,
+      "learning_rate": 5e-06,
+      "loss": 0.125,
+      "step": 13625
+    },
+    {
+      "epoch": 10.256680466691758,
+      "grad_norm": 0.18490250408649445,
+      "learning_rate": 5e-06,
+      "loss": 0.1307,
+      "step": 13626
+    },
+    {
+      "epoch": 10.257433195333082,
+      "grad_norm": 0.17601127922534943,
+      "learning_rate": 5e-06,
+      "loss": 0.1434,
+      "step": 13627
+    },
+    {
+      "epoch": 10.258185923974407,
+      "grad_norm": 0.17937391996383667,
+      "learning_rate": 5e-06,
+      "loss": 0.2035,
+      "step": 13628
+    },
+    {
+      "epoch": 10.258938652615733,
+      "grad_norm": 0.17672067880630493,
+      "learning_rate": 5e-06,
+      "loss": 0.1661,
+      "step": 13629
+    },
+    {
+      "epoch": 10.259691381257056,
+      "grad_norm": 0.1529737412929535,
+      "learning_rate": 5e-06,
+      "loss": 0.0997,
+      "step": 13630
+    },
+    {
+      "epoch": 10.260444109898382,
+      "grad_norm": 0.2172139286994934,
+      "learning_rate": 5e-06,
+      "loss": 0.2008,
+      "step": 13631
+    },
+    {
+      "epoch": 10.261196838539707,
+      "grad_norm": 0.2173309326171875,
+      "learning_rate": 5e-06,
+      "loss": 0.226,
+      "step": 13632
+    },
+    {
+      "epoch": 10.26194956718103,
+      "grad_norm": 0.1953049749135971,
+      "learning_rate": 5e-06,
+      "loss": 0.1728,
+      "step": 13633
+    },
+    {
+      "epoch": 10.262702295822356,
+      "grad_norm": 0.18471509218215942,
+      "learning_rate": 5e-06,
+      "loss": 0.1553,
+      "step": 13634
+    },
+    {
+      "epoch": 10.263455024463681,
+      "grad_norm": 0.2048618495464325,
+      "learning_rate": 5e-06,
+      "loss": 0.159,
+      "step": 13635
+    },
+    {
+      "epoch": 10.264207753105005,
+      "grad_norm": 0.16436007618904114,
+      "learning_rate": 5e-06,
+      "loss": 0.1353,
+      "step": 13636
+    },
+    {
+      "epoch": 10.26496048174633,
+      "grad_norm": 0.1801706701517105,
+      "learning_rate": 5e-06,
+      "loss": 0.117,
+      "step": 13637
+    },
+    {
+      "epoch": 10.265713210387656,
+      "grad_norm": 0.15504255890846252,
+      "learning_rate": 5e-06,
+      "loss": 0.0934,
+      "step": 13638
+    },
+    {
+      "epoch": 10.26646593902898,
+      "grad_norm": 0.1769474595785141,
+      "learning_rate": 5e-06,
+      "loss": 0.1467,
+      "step": 13639
+    },
+    {
+      "epoch": 10.267218667670305,
+      "grad_norm": 0.19995015859603882,
+      "learning_rate": 5e-06,
+      "loss": 0.1859,
+      "step": 13640
+    },
+    {
+      "epoch": 10.26797139631163,
+      "grad_norm": 0.16132456064224243,
+      "learning_rate": 5e-06,
+      "loss": 0.1023,
+      "step": 13641
+    },
+    {
+      "epoch": 10.268724124952955,
+      "grad_norm": 0.19472189247608185,
+      "learning_rate": 5e-06,
+      "loss": 0.1285,
+      "step": 13642
+    },
+    {
+      "epoch": 10.269476853594279,
+      "grad_norm": 0.197132870554924,
+      "learning_rate": 5e-06,
+      "loss": 0.1712,
+      "step": 13643
+    },
+    {
+      "epoch": 10.270229582235604,
+      "grad_norm": 0.19674497842788696,
+      "learning_rate": 5e-06,
+      "loss": 0.1241,
+      "step": 13644
+    },
+    {
+      "epoch": 10.27098231087693,
+      "grad_norm": 0.1666429340839386,
+      "learning_rate": 5e-06,
+      "loss": 0.1134,
+      "step": 13645
+    },
+    {
+      "epoch": 10.271735039518253,
+      "grad_norm": 0.18407467007637024,
+      "learning_rate": 5e-06,
+      "loss": 0.1245,
+      "step": 13646
+    },
+    {
+      "epoch": 10.272487768159579,
+      "grad_norm": 0.195142924785614,
+      "learning_rate": 5e-06,
+      "loss": 0.1449,
+      "step": 13647
+    },
+    {
+      "epoch": 10.273240496800904,
+      "grad_norm": 0.1931609809398651,
+      "learning_rate": 5e-06,
+      "loss": 0.1361,
+      "step": 13648
+    },
+    {
+      "epoch": 10.273993225442227,
+      "grad_norm": 0.18550065159797668,
+      "learning_rate": 5e-06,
+      "loss": 0.2383,
+      "step": 13649
+    },
+    {
+      "epoch": 10.274745954083553,
+      "grad_norm": 0.1951633095741272,
+      "learning_rate": 5e-06,
+      "loss": 0.1223,
+      "step": 13650
+    },
+    {
+      "epoch": 10.275498682724878,
+      "grad_norm": 0.1652405858039856,
+      "learning_rate": 5e-06,
+      "loss": 0.1627,
+      "step": 13651
+    },
+    {
+      "epoch": 10.276251411366202,
+      "grad_norm": 0.16642087697982788,
+      "learning_rate": 5e-06,
+      "loss": 0.1355,
+      "step": 13652
+    },
+    {
+      "epoch": 10.277004140007527,
+      "grad_norm": 0.20107176899909973,
+      "learning_rate": 5e-06,
+      "loss": 0.2504,
+      "step": 13653
+    },
+    {
+      "epoch": 10.277756868648853,
+      "grad_norm": 0.18402056396007538,
+      "learning_rate": 5e-06,
+      "loss": 0.1159,
+      "step": 13654
+    },
+    {
+      "epoch": 10.278509597290176,
+      "grad_norm": 0.15913575887680054,
+      "learning_rate": 5e-06,
+      "loss": 0.1163,
+      "step": 13655
+    },
+    {
+      "epoch": 10.279262325931501,
+      "grad_norm": 0.16181601583957672,
+      "learning_rate": 5e-06,
+      "loss": 0.1215,
+      "step": 13656
+    },
+    {
+      "epoch": 10.280015054572827,
+      "grad_norm": 0.20131321251392365,
+      "learning_rate": 5e-06,
+      "loss": 0.172,
+      "step": 13657
+    },
+    {
+      "epoch": 10.280767783214152,
+      "grad_norm": 0.16748152673244476,
+      "learning_rate": 5e-06,
+      "loss": 0.1221,
+      "step": 13658
+    },
+    {
+      "epoch": 10.281520511855476,
+      "grad_norm": 0.1573709100484848,
+      "learning_rate": 5e-06,
+      "loss": 0.0734,
+      "step": 13659
+    },
+    {
+      "epoch": 10.282273240496801,
+      "grad_norm": 0.17850913107395172,
+      "learning_rate": 5e-06,
+      "loss": 0.1361,
+      "step": 13660
+    },
+    {
+      "epoch": 10.283025969138126,
+      "grad_norm": 0.20926818251609802,
+      "learning_rate": 5e-06,
+      "loss": 0.1839,
+      "step": 13661
+    },
+    {
+      "epoch": 10.28377869777945,
+      "grad_norm": 0.21419960260391235,
+      "learning_rate": 5e-06,
+      "loss": 0.1907,
+      "step": 13662
+    },
+    {
+      "epoch": 10.284531426420775,
+      "grad_norm": 0.19092710316181183,
+      "learning_rate": 5e-06,
+      "loss": 0.1431,
+      "step": 13663
+    },
+    {
+      "epoch": 10.2852841550621,
+      "grad_norm": 0.19113978743553162,
+      "learning_rate": 5e-06,
+      "loss": 0.1786,
+      "step": 13664
+    },
+    {
+      "epoch": 10.286036883703424,
+      "grad_norm": 0.17939932644367218,
+      "learning_rate": 5e-06,
+      "loss": 0.1133,
+      "step": 13665
+    },
+    {
+      "epoch": 10.28678961234475,
+      "grad_norm": 0.16359542310237885,
+      "learning_rate": 5e-06,
+      "loss": 0.1027,
+      "step": 13666
+    },
+    {
+      "epoch": 10.287542340986075,
+      "grad_norm": 0.1964401751756668,
+      "learning_rate": 5e-06,
+      "loss": 0.1692,
+      "step": 13667
+    },
+    {
+      "epoch": 10.288295069627399,
+      "grad_norm": 0.17568950355052948,
+      "learning_rate": 5e-06,
+      "loss": 0.13,
+      "step": 13668
+    },
+    {
+      "epoch": 10.289047798268724,
+      "grad_norm": 0.1916183978319168,
+      "learning_rate": 5e-06,
+      "loss": 0.1716,
+      "step": 13669
+    },
+    {
+      "epoch": 10.28980052691005,
+      "grad_norm": 0.26769086718559265,
+      "learning_rate": 5e-06,
+      "loss": 0.1282,
+      "step": 13670
+    },
+    {
+      "epoch": 10.290553255551373,
+      "grad_norm": 0.172886922955513,
+      "learning_rate": 5e-06,
+      "loss": 0.1063,
+      "step": 13671
+    },
+    {
+      "epoch": 10.291305984192698,
+      "grad_norm": 0.164017915725708,
+      "learning_rate": 5e-06,
+      "loss": 0.1804,
+      "step": 13672
+    },
+    {
+      "epoch": 10.292058712834024,
+      "grad_norm": 0.15348687767982483,
+      "learning_rate": 5e-06,
+      "loss": 0.092,
+      "step": 13673
+    },
+    {
+      "epoch": 10.292811441475347,
+      "grad_norm": 0.17192499339580536,
+      "learning_rate": 5e-06,
+      "loss": 0.1315,
+      "step": 13674
+    },
+    {
+      "epoch": 10.293564170116673,
+      "grad_norm": 0.17225417494773865,
+      "learning_rate": 5e-06,
+      "loss": 0.1205,
+      "step": 13675
+    },
+    {
+      "epoch": 10.294316898757998,
+      "grad_norm": 0.19607549905776978,
+      "learning_rate": 5e-06,
+      "loss": 0.2133,
+      "step": 13676
+    },
+    {
+      "epoch": 10.295069627399322,
+      "grad_norm": 0.16977499425411224,
+      "learning_rate": 5e-06,
+      "loss": 0.1356,
+      "step": 13677
+    },
+    {
+      "epoch": 10.295822356040647,
+      "grad_norm": 0.18839313089847565,
+      "learning_rate": 5e-06,
+      "loss": 0.1831,
+      "step": 13678
+    },
+    {
+      "epoch": 10.296575084681972,
+      "grad_norm": 0.1763528734445572,
+      "learning_rate": 5e-06,
+      "loss": 0.1768,
+      "step": 13679
+    },
+    {
+      "epoch": 10.297327813323298,
+      "grad_norm": 0.1757877618074417,
+      "learning_rate": 5e-06,
+      "loss": 0.1186,
+      "step": 13680
+    },
+    {
+      "epoch": 10.298080541964621,
+      "grad_norm": 0.18969357013702393,
+      "learning_rate": 5e-06,
+      "loss": 0.1074,
+      "step": 13681
+    },
+    {
+      "epoch": 10.298833270605947,
+      "grad_norm": 0.1673484891653061,
+      "learning_rate": 5e-06,
+      "loss": 0.1211,
+      "step": 13682
+    },
+    {
+      "epoch": 10.299585999247272,
+      "grad_norm": 0.1774778664112091,
+      "learning_rate": 5e-06,
+      "loss": 0.1483,
+      "step": 13683
+    },
+    {
+      "epoch": 10.300338727888596,
+      "grad_norm": 0.1896854192018509,
+      "learning_rate": 5e-06,
+      "loss": 0.0896,
+      "step": 13684
+    },
+    {
+      "epoch": 10.301091456529921,
+      "grad_norm": 0.17790132761001587,
+      "learning_rate": 5e-06,
+      "loss": 0.1477,
+      "step": 13685
+    },
+    {
+      "epoch": 10.301844185171246,
+      "grad_norm": 0.1691836714744568,
+      "learning_rate": 5e-06,
+      "loss": 0.1054,
+      "step": 13686
+    },
+    {
+      "epoch": 10.30259691381257,
+      "grad_norm": 0.18003258109092712,
+      "learning_rate": 5e-06,
+      "loss": 0.1381,
+      "step": 13687
+    },
+    {
+      "epoch": 10.303349642453895,
+      "grad_norm": 0.17041845619678497,
+      "learning_rate": 5e-06,
+      "loss": 0.1092,
+      "step": 13688
+    },
+    {
+      "epoch": 10.30410237109522,
+      "grad_norm": 0.21277682483196259,
+      "learning_rate": 5e-06,
+      "loss": 0.1093,
+      "step": 13689
+    },
+    {
+      "epoch": 10.304855099736544,
+      "grad_norm": 0.16456879675388336,
+      "learning_rate": 5e-06,
+      "loss": 0.1359,
+      "step": 13690
+    },
+    {
+      "epoch": 10.30560782837787,
+      "grad_norm": 0.17934173345565796,
+      "learning_rate": 5e-06,
+      "loss": 0.2107,
+      "step": 13691
+    },
+    {
+      "epoch": 10.306360557019195,
+      "grad_norm": 0.1774134486913681,
+      "learning_rate": 5e-06,
+      "loss": 0.1411,
+      "step": 13692
+    },
+    {
+      "epoch": 10.307113285660519,
+      "grad_norm": 0.1957666277885437,
+      "learning_rate": 5e-06,
+      "loss": 0.1586,
+      "step": 13693
+    },
+    {
+      "epoch": 10.307866014301844,
+      "grad_norm": 0.16260769963264465,
+      "learning_rate": 5e-06,
+      "loss": 0.1194,
+      "step": 13694
+    },
+    {
+      "epoch": 10.30861874294317,
+      "grad_norm": 0.18351006507873535,
+      "learning_rate": 5e-06,
+      "loss": 0.1476,
+      "step": 13695
+    },
+    {
+      "epoch": 10.309371471584495,
+      "grad_norm": 0.17777115106582642,
+      "learning_rate": 5e-06,
+      "loss": 0.1381,
+      "step": 13696
+    },
+    {
+      "epoch": 10.310124200225818,
+      "grad_norm": 0.18470628559589386,
+      "learning_rate": 5e-06,
+      "loss": 0.1399,
+      "step": 13697
+    },
+    {
+      "epoch": 10.310876928867144,
+      "grad_norm": 0.19656343758106232,
+      "learning_rate": 5e-06,
+      "loss": 0.1579,
+      "step": 13698
+    },
+    {
+      "epoch": 10.311629657508469,
+      "grad_norm": 0.17416022717952728,
+      "learning_rate": 5e-06,
+      "loss": 0.1026,
+      "step": 13699
+    },
+    {
+      "epoch": 10.312382386149793,
+      "grad_norm": 0.17178447544574738,
+      "learning_rate": 5e-06,
+      "loss": 0.2021,
+      "step": 13700
+    },
+    {
+      "epoch": 10.313135114791118,
+      "grad_norm": 0.18755660951137543,
+      "learning_rate": 5e-06,
+      "loss": 0.2226,
+      "step": 13701
+    },
+    {
+      "epoch": 10.313887843432443,
+      "grad_norm": 0.17469875514507294,
+      "learning_rate": 5e-06,
+      "loss": 0.1369,
+      "step": 13702
+    },
+    {
+      "epoch": 10.314640572073767,
+      "grad_norm": 0.17099644243717194,
+      "learning_rate": 5e-06,
+      "loss": 0.0931,
+      "step": 13703
+    },
+    {
+      "epoch": 10.315393300715092,
+      "grad_norm": 0.16658101975917816,
+      "learning_rate": 5e-06,
+      "loss": 0.1324,
+      "step": 13704
+    },
+    {
+      "epoch": 10.316146029356418,
+      "grad_norm": 0.1596565991640091,
+      "learning_rate": 5e-06,
+      "loss": 0.1143,
+      "step": 13705
+    },
+    {
+      "epoch": 10.316898757997741,
+      "grad_norm": 0.18693257868289948,
+      "learning_rate": 5e-06,
+      "loss": 0.1834,
+      "step": 13706
+    },
+    {
+      "epoch": 10.317651486639067,
+      "grad_norm": 0.17902247607707977,
+      "learning_rate": 5e-06,
+      "loss": 0.1974,
+      "step": 13707
+    },
+    {
+      "epoch": 10.318404215280392,
+      "grad_norm": 0.1513330638408661,
+      "learning_rate": 5e-06,
+      "loss": 0.0836,
+      "step": 13708
+    },
+    {
+      "epoch": 10.319156943921715,
+      "grad_norm": 0.151978000998497,
+      "learning_rate": 5e-06,
+      "loss": 0.0717,
+      "step": 13709
+    },
+    {
+      "epoch": 10.31990967256304,
+      "grad_norm": 0.19107265770435333,
+      "learning_rate": 5e-06,
+      "loss": 0.0792,
+      "step": 13710
+    },
+    {
+      "epoch": 10.320662401204366,
+      "grad_norm": 0.21085287630558014,
+      "learning_rate": 5e-06,
+      "loss": 0.1797,
+      "step": 13711
+    },
+    {
+      "epoch": 10.32141512984569,
+      "grad_norm": 0.19239497184753418,
+      "learning_rate": 5e-06,
+      "loss": 0.1187,
+      "step": 13712
+    },
+    {
+      "epoch": 10.322167858487015,
+      "grad_norm": 0.19394826889038086,
+      "learning_rate": 5e-06,
+      "loss": 0.1628,
+      "step": 13713
+    },
+    {
+      "epoch": 10.32292058712834,
+      "grad_norm": 0.16878145933151245,
+      "learning_rate": 5e-06,
+      "loss": 0.1348,
+      "step": 13714
+    },
+    {
+      "epoch": 10.323673315769666,
+      "grad_norm": 0.16706852614879608,
+      "learning_rate": 5e-06,
+      "loss": 0.2015,
+      "step": 13715
+    },
+    {
+      "epoch": 10.32442604441099,
+      "grad_norm": 0.17475947737693787,
+      "learning_rate": 5e-06,
+      "loss": 0.1445,
+      "step": 13716
+    },
+    {
+      "epoch": 10.325178773052315,
+      "grad_norm": 0.16471979022026062,
+      "learning_rate": 5e-06,
+      "loss": 0.1306,
+      "step": 13717
+    },
+    {
+      "epoch": 10.32593150169364,
+      "grad_norm": 0.17630517482757568,
+      "learning_rate": 5e-06,
+      "loss": 0.1283,
+      "step": 13718
+    },
+    {
+      "epoch": 10.326684230334964,
+      "grad_norm": 0.19168245792388916,
+      "learning_rate": 5e-06,
+      "loss": 0.1507,
+      "step": 13719
+    },
+    {
+      "epoch": 10.327436958976289,
+      "grad_norm": 0.17395564913749695,
+      "learning_rate": 5e-06,
+      "loss": 0.138,
+      "step": 13720
+    },
+    {
+      "epoch": 10.328189687617614,
+      "grad_norm": 0.15886126458644867,
+      "learning_rate": 5e-06,
+      "loss": 0.1187,
+      "step": 13721
+    },
+    {
+      "epoch": 10.328942416258938,
+      "grad_norm": 0.19389811158180237,
+      "learning_rate": 5e-06,
+      "loss": 0.1917,
+      "step": 13722
+    },
+    {
+      "epoch": 10.329695144900263,
+      "grad_norm": 0.19200550019741058,
+      "learning_rate": 5e-06,
+      "loss": 0.1647,
+      "step": 13723
+    },
+    {
+      "epoch": 10.330447873541589,
+      "grad_norm": 0.18049244582653046,
+      "learning_rate": 5e-06,
+      "loss": 0.1196,
+      "step": 13724
+    },
+    {
+      "epoch": 10.331200602182912,
+      "grad_norm": 0.17951038479804993,
+      "learning_rate": 5e-06,
+      "loss": 0.1287,
+      "step": 13725
+    },
+    {
+      "epoch": 10.331953330824238,
+      "grad_norm": 0.19739095866680145,
+      "learning_rate": 5e-06,
+      "loss": 0.2107,
+      "step": 13726
+    },
+    {
+      "epoch": 10.332706059465563,
+      "grad_norm": 0.17144371569156647,
+      "learning_rate": 5e-06,
+      "loss": 0.0638,
+      "step": 13727
+    },
+    {
+      "epoch": 10.333458788106887,
+      "grad_norm": 0.16666048765182495,
+      "learning_rate": 5e-06,
+      "loss": 0.0965,
+      "step": 13728
+    },
+    {
+      "epoch": 10.334211516748212,
+      "grad_norm": 0.1864050328731537,
+      "learning_rate": 5e-06,
+      "loss": 0.1891,
+      "step": 13729
+    },
+    {
+      "epoch": 10.334964245389537,
+      "grad_norm": 0.16632674634456635,
+      "learning_rate": 5e-06,
+      "loss": 0.1085,
+      "step": 13730
+    },
+    {
+      "epoch": 10.335716974030863,
+      "grad_norm": 0.18737977743148804,
+      "learning_rate": 5e-06,
+      "loss": 0.1996,
+      "step": 13731
+    },
+    {
+      "epoch": 10.336469702672186,
+      "grad_norm": 0.17277589440345764,
+      "learning_rate": 5e-06,
+      "loss": 0.1005,
+      "step": 13732
+    },
+    {
+      "epoch": 10.337222431313512,
+      "grad_norm": 0.1572033315896988,
+      "learning_rate": 5e-06,
+      "loss": 0.1084,
+      "step": 13733
+    },
+    {
+      "epoch": 10.337975159954837,
+      "grad_norm": 0.17556984722614288,
+      "learning_rate": 5e-06,
+      "loss": 0.1182,
+      "step": 13734
+    },
+    {
+      "epoch": 10.33872788859616,
+      "grad_norm": 0.15721231698989868,
+      "learning_rate": 5e-06,
+      "loss": 0.1593,
+      "step": 13735
+    },
+    {
+      "epoch": 10.339480617237486,
+      "grad_norm": 0.1801440566778183,
+      "learning_rate": 5e-06,
+      "loss": 0.1007,
+      "step": 13736
+    },
+    {
+      "epoch": 10.340233345878811,
+      "grad_norm": 0.18506158888339996,
+      "learning_rate": 5e-06,
+      "loss": 0.1371,
+      "step": 13737
+    },
+    {
+      "epoch": 10.340986074520135,
+      "grad_norm": 0.19228726625442505,
+      "learning_rate": 5e-06,
+      "loss": 0.1595,
+      "step": 13738
+    },
+    {
+      "epoch": 10.34173880316146,
+      "grad_norm": 0.19283533096313477,
+      "learning_rate": 5e-06,
+      "loss": 0.2004,
+      "step": 13739
+    },
+    {
+      "epoch": 10.342491531802786,
+      "grad_norm": 0.15503861010074615,
+      "learning_rate": 5e-06,
+      "loss": 0.1069,
+      "step": 13740
+    },
+    {
+      "epoch": 10.34324426044411,
+      "grad_norm": 0.18406113982200623,
+      "learning_rate": 5e-06,
+      "loss": 0.2167,
+      "step": 13741
+    },
+    {
+      "epoch": 10.343996989085435,
+      "grad_norm": 0.1798771619796753,
+      "learning_rate": 5e-06,
+      "loss": 0.1164,
+      "step": 13742
+    },
+    {
+      "epoch": 10.34474971772676,
+      "grad_norm": 0.16215820610523224,
+      "learning_rate": 5e-06,
+      "loss": 0.1421,
+      "step": 13743
+    },
+    {
+      "epoch": 10.345502446368084,
+      "grad_norm": 0.21586789190769196,
+      "learning_rate": 5e-06,
+      "loss": 0.157,
+      "step": 13744
+    },
+    {
+      "epoch": 10.346255175009409,
+      "grad_norm": 0.14901861548423767,
+      "learning_rate": 5e-06,
+      "loss": 0.0834,
+      "step": 13745
+    },
+    {
+      "epoch": 10.347007903650734,
+      "grad_norm": 0.17851601541042328,
+      "learning_rate": 5e-06,
+      "loss": 0.152,
+      "step": 13746
+    },
+    {
+      "epoch": 10.347760632292058,
+      "grad_norm": 0.15481311082839966,
+      "learning_rate": 5e-06,
+      "loss": 0.1634,
+      "step": 13747
+    },
+    {
+      "epoch": 10.348513360933383,
+      "grad_norm": 0.20771166682243347,
+      "learning_rate": 5e-06,
+      "loss": 0.1552,
+      "step": 13748
+    },
+    {
+      "epoch": 10.349266089574709,
+      "grad_norm": 0.1679760068655014,
+      "learning_rate": 5e-06,
+      "loss": 0.1186,
+      "step": 13749
+    },
+    {
+      "epoch": 10.350018818216034,
+      "grad_norm": 0.1849009096622467,
+      "learning_rate": 5e-06,
+      "loss": 0.1958,
+      "step": 13750
+    },
+    {
+      "epoch": 10.350771546857358,
+      "grad_norm": 0.18012219667434692,
+      "learning_rate": 5e-06,
+      "loss": 0.1411,
+      "step": 13751
+    },
+    {
+      "epoch": 10.351524275498683,
+      "grad_norm": 0.17978528141975403,
+      "learning_rate": 5e-06,
+      "loss": 0.1378,
+      "step": 13752
+    },
+    {
+      "epoch": 10.352277004140008,
+      "grad_norm": 0.1509474366903305,
+      "learning_rate": 5e-06,
+      "loss": 0.0926,
+      "step": 13753
+    },
+    {
+      "epoch": 10.353029732781332,
+      "grad_norm": 0.22622202336788177,
+      "learning_rate": 5e-06,
+      "loss": 0.1889,
+      "step": 13754
+    },
+    {
+      "epoch": 10.353782461422657,
+      "grad_norm": 0.1828690767288208,
+      "learning_rate": 5e-06,
+      "loss": 0.1879,
+      "step": 13755
+    },
+    {
+      "epoch": 10.354535190063983,
+      "grad_norm": 0.20414677262306213,
+      "learning_rate": 5e-06,
+      "loss": 0.1492,
+      "step": 13756
+    },
+    {
+      "epoch": 10.355287918705306,
+      "grad_norm": 0.16781790554523468,
+      "learning_rate": 5e-06,
+      "loss": 0.1878,
+      "step": 13757
+    },
+    {
+      "epoch": 10.356040647346632,
+      "grad_norm": 0.1541563719511032,
+      "learning_rate": 5e-06,
+      "loss": 0.0764,
+      "step": 13758
+    },
+    {
+      "epoch": 10.356793375987957,
+      "grad_norm": 0.19637265801429749,
+      "learning_rate": 5e-06,
+      "loss": 0.1258,
+      "step": 13759
+    },
+    {
+      "epoch": 10.35754610462928,
+      "grad_norm": 0.16712969541549683,
+      "learning_rate": 5e-06,
+      "loss": 0.087,
+      "step": 13760
+    },
+    {
+      "epoch": 10.358298833270606,
+      "grad_norm": 0.1720980852842331,
+      "learning_rate": 5e-06,
+      "loss": 0.1648,
+      "step": 13761
+    },
+    {
+      "epoch": 10.359051561911931,
+      "grad_norm": 0.19633305072784424,
+      "learning_rate": 5e-06,
+      "loss": 0.1328,
+      "step": 13762
+    },
+    {
+      "epoch": 10.359804290553255,
+      "grad_norm": 0.1685507446527481,
+      "learning_rate": 5e-06,
+      "loss": 0.0692,
+      "step": 13763
+    },
+    {
+      "epoch": 10.36055701919458,
+      "grad_norm": 0.1985330730676651,
+      "learning_rate": 5e-06,
+      "loss": 0.1719,
+      "step": 13764
+    },
+    {
+      "epoch": 10.361309747835906,
+      "grad_norm": 0.18168684840202332,
+      "learning_rate": 5e-06,
+      "loss": 0.1046,
+      "step": 13765
+    },
+    {
+      "epoch": 10.362062476477229,
+      "grad_norm": 0.18961592018604279,
+      "learning_rate": 5e-06,
+      "loss": 0.1362,
+      "step": 13766
+    },
+    {
+      "epoch": 10.362815205118554,
+      "grad_norm": 0.1964840441942215,
+      "learning_rate": 5e-06,
+      "loss": 0.1679,
+      "step": 13767
+    },
+    {
+      "epoch": 10.36356793375988,
+      "grad_norm": 0.15504372119903564,
+      "learning_rate": 5e-06,
+      "loss": 0.0803,
+      "step": 13768
+    },
+    {
+      "epoch": 10.364320662401205,
+      "grad_norm": 0.1749049574136734,
+      "learning_rate": 5e-06,
+      "loss": 0.1423,
+      "step": 13769
+    },
+    {
+      "epoch": 10.365073391042529,
+      "grad_norm": 0.19958548247814178,
+      "learning_rate": 5e-06,
+      "loss": 0.147,
+      "step": 13770
+    },
+    {
+      "epoch": 10.365826119683854,
+      "grad_norm": 0.16655012965202332,
+      "learning_rate": 5e-06,
+      "loss": 0.1741,
+      "step": 13771
+    },
+    {
+      "epoch": 10.36657884832518,
+      "grad_norm": 0.17277652025222778,
+      "learning_rate": 5e-06,
+      "loss": 0.083,
+      "step": 13772
+    },
+    {
+      "epoch": 10.367331576966503,
+      "grad_norm": 0.22771212458610535,
+      "learning_rate": 5e-06,
+      "loss": 0.2621,
+      "step": 13773
+    },
+    {
+      "epoch": 10.368084305607828,
+      "grad_norm": 0.20825353264808655,
+      "learning_rate": 5e-06,
+      "loss": 0.1617,
+      "step": 13774
+    },
+    {
+      "epoch": 10.368837034249154,
+      "grad_norm": 0.17803612351417542,
+      "learning_rate": 5e-06,
+      "loss": 0.105,
+      "step": 13775
+    },
+    {
+      "epoch": 10.369589762890477,
+      "grad_norm": 0.14493779838085175,
+      "learning_rate": 5e-06,
+      "loss": 0.1171,
+      "step": 13776
+    },
+    {
+      "epoch": 10.370342491531803,
+      "grad_norm": 0.19996123015880585,
+      "learning_rate": 5e-06,
+      "loss": 0.2578,
+      "step": 13777
+    },
+    {
+      "epoch": 10.371095220173128,
+      "grad_norm": 0.16644686460494995,
+      "learning_rate": 5e-06,
+      "loss": 0.1292,
+      "step": 13778
+    },
+    {
+      "epoch": 10.371847948814452,
+      "grad_norm": 0.19850820302963257,
+      "learning_rate": 5e-06,
+      "loss": 0.129,
+      "step": 13779
+    },
+    {
+      "epoch": 10.372600677455777,
+      "grad_norm": 0.1988259106874466,
+      "learning_rate": 5e-06,
+      "loss": 0.2008,
+      "step": 13780
+    },
+    {
+      "epoch": 10.373353406097102,
+      "grad_norm": 0.17538997530937195,
+      "learning_rate": 5e-06,
+      "loss": 0.121,
+      "step": 13781
+    },
+    {
+      "epoch": 10.374106134738426,
+      "grad_norm": 0.18526875972747803,
+      "learning_rate": 5e-06,
+      "loss": 0.2056,
+      "step": 13782
+    },
+    {
+      "epoch": 10.374858863379751,
+      "grad_norm": 0.18702735006809235,
+      "learning_rate": 5e-06,
+      "loss": 0.1795,
+      "step": 13783
+    },
+    {
+      "epoch": 10.375611592021077,
+      "grad_norm": 0.20966288447380066,
+      "learning_rate": 5e-06,
+      "loss": 0.1598,
+      "step": 13784
+    },
+    {
+      "epoch": 10.3763643206624,
+      "grad_norm": 0.18672606348991394,
+      "learning_rate": 5e-06,
+      "loss": 0.1644,
+      "step": 13785
+    },
+    {
+      "epoch": 10.377117049303726,
+      "grad_norm": 0.16906419396400452,
+      "learning_rate": 5e-06,
+      "loss": 0.1302,
+      "step": 13786
+    },
+    {
+      "epoch": 10.377869777945051,
+      "grad_norm": 0.20002026855945587,
+      "learning_rate": 5e-06,
+      "loss": 0.1155,
+      "step": 13787
+    },
+    {
+      "epoch": 10.378622506586376,
+      "grad_norm": 0.19242942333221436,
+      "learning_rate": 5e-06,
+      "loss": 0.1076,
+      "step": 13788
+    },
+    {
+      "epoch": 10.3793752352277,
+      "grad_norm": 0.16745704412460327,
+      "learning_rate": 5e-06,
+      "loss": 0.1134,
+      "step": 13789
+    },
+    {
+      "epoch": 10.380127963869025,
+      "grad_norm": 0.17296844720840454,
+      "learning_rate": 5e-06,
+      "loss": 0.1242,
+      "step": 13790
+    },
+    {
+      "epoch": 10.38088069251035,
+      "grad_norm": 0.1773504614830017,
+      "learning_rate": 5e-06,
+      "loss": 0.1266,
+      "step": 13791
+    },
+    {
+      "epoch": 10.381633421151674,
+      "grad_norm": 0.1615578979253769,
+      "learning_rate": 5e-06,
+      "loss": 0.12,
+      "step": 13792
+    },
+    {
+      "epoch": 10.382386149793,
+      "grad_norm": 0.17756174504756927,
+      "learning_rate": 5e-06,
+      "loss": 0.221,
+      "step": 13793
+    },
+    {
+      "epoch": 10.383138878434325,
+      "grad_norm": 0.18712468445301056,
+      "learning_rate": 5e-06,
+      "loss": 0.1142,
+      "step": 13794
+    },
+    {
+      "epoch": 10.383891607075649,
+      "grad_norm": 0.1792829930782318,
+      "learning_rate": 5e-06,
+      "loss": 0.1266,
+      "step": 13795
+    },
+    {
+      "epoch": 10.384644335716974,
+      "grad_norm": 0.18612414598464966,
+      "learning_rate": 5e-06,
+      "loss": 0.1507,
+      "step": 13796
+    },
+    {
+      "epoch": 10.3853970643583,
+      "grad_norm": 0.17802274227142334,
+      "learning_rate": 5e-06,
+      "loss": 0.1287,
+      "step": 13797
+    },
+    {
+      "epoch": 10.386149792999623,
+      "grad_norm": 0.18434323370456696,
+      "learning_rate": 5e-06,
+      "loss": 0.2041,
+      "step": 13798
+    },
+    {
+      "epoch": 10.386902521640948,
+      "grad_norm": 0.18774239718914032,
+      "learning_rate": 5e-06,
+      "loss": 0.1263,
+      "step": 13799
+    },
+    {
+      "epoch": 10.387655250282274,
+      "grad_norm": 0.18326039612293243,
+      "learning_rate": 5e-06,
+      "loss": 0.1437,
+      "step": 13800
+    },
+    {
+      "epoch": 10.387655250282274,
+      "eval_loss": 0.18928825855255127,
+      "eval_runtime": 452.8863,
+      "eval_samples_per_second": 21.257,
+      "eval_steps_per_second": 0.665,
+      "step": 13800
+    },
+    {
+      "epoch": 10.388407978923597,
+      "grad_norm": 0.17706847190856934,
+      "learning_rate": 5e-06,
+      "loss": 0.1372,
+      "step": 13801
+    },
+    {
+      "epoch": 10.389160707564923,
+      "grad_norm": 0.18085399270057678,
+      "learning_rate": 5e-06,
+      "loss": 0.1017,
+      "step": 13802
+    },
+    {
+      "epoch": 10.389913436206248,
+      "grad_norm": 0.199459969997406,
+      "learning_rate": 5e-06,
+      "loss": 0.284,
+      "step": 13803
+    },
+    {
+      "epoch": 10.390666164847573,
+      "grad_norm": 0.15579335391521454,
+      "learning_rate": 5e-06,
+      "loss": 0.1274,
+      "step": 13804
+    },
+    {
+      "epoch": 10.391418893488897,
+      "grad_norm": 0.1747254878282547,
+      "learning_rate": 5e-06,
+      "loss": 0.1941,
+      "step": 13805
+    },
+    {
+      "epoch": 10.392171622130222,
+      "grad_norm": 0.18603728711605072,
+      "learning_rate": 5e-06,
+      "loss": 0.1629,
+      "step": 13806
+    },
+    {
+      "epoch": 10.392924350771548,
+      "grad_norm": 0.16262482106685638,
+      "learning_rate": 5e-06,
+      "loss": 0.0735,
+      "step": 13807
+    },
+    {
+      "epoch": 10.393677079412871,
+      "grad_norm": 0.19441105425357819,
+      "learning_rate": 5e-06,
+      "loss": 0.2011,
+      "step": 13808
+    },
+    {
+      "epoch": 10.394429808054197,
+      "grad_norm": 0.16450929641723633,
+      "learning_rate": 5e-06,
+      "loss": 0.1624,
+      "step": 13809
+    },
+    {
+      "epoch": 10.395182536695522,
+      "grad_norm": 0.17936575412750244,
+      "learning_rate": 5e-06,
+      "loss": 0.1767,
+      "step": 13810
+    },
+    {
+      "epoch": 10.395935265336846,
+      "grad_norm": 0.18944692611694336,
+      "learning_rate": 5e-06,
+      "loss": 0.1185,
+      "step": 13811
+    },
+    {
+      "epoch": 10.396687993978171,
+      "grad_norm": 0.17411114275455475,
+      "learning_rate": 5e-06,
+      "loss": 0.0955,
+      "step": 13812
+    },
+    {
+      "epoch": 10.397440722619496,
+      "grad_norm": 0.17029127478599548,
+      "learning_rate": 5e-06,
+      "loss": 0.1421,
+      "step": 13813
+    },
+    {
+      "epoch": 10.39819345126082,
+      "grad_norm": 0.1873115599155426,
+      "learning_rate": 5e-06,
+      "loss": 0.1756,
+      "step": 13814
+    },
+    {
+      "epoch": 10.398946179902145,
+      "grad_norm": 0.1747685968875885,
+      "learning_rate": 5e-06,
+      "loss": 0.1619,
+      "step": 13815
+    },
+    {
+      "epoch": 10.39969890854347,
+      "grad_norm": 0.16723118722438812,
+      "learning_rate": 5e-06,
+      "loss": 0.179,
+      "step": 13816
+    },
+    {
+      "epoch": 10.400451637184794,
+      "grad_norm": 0.18674418330192566,
+      "learning_rate": 5e-06,
+      "loss": 0.2101,
+      "step": 13817
+    },
+    {
+      "epoch": 10.40120436582612,
+      "grad_norm": 0.17239591479301453,
+      "learning_rate": 5e-06,
+      "loss": 0.118,
+      "step": 13818
+    },
+    {
+      "epoch": 10.401957094467445,
+      "grad_norm": 0.19828060269355774,
+      "learning_rate": 5e-06,
+      "loss": 0.2026,
+      "step": 13819
+    },
+    {
+      "epoch": 10.402709823108768,
+      "grad_norm": 0.1760793924331665,
+      "learning_rate": 5e-06,
+      "loss": 0.0805,
+      "step": 13820
+    },
+    {
+      "epoch": 10.403462551750094,
+      "grad_norm": 0.17240439355373383,
+      "learning_rate": 5e-06,
+      "loss": 0.1291,
+      "step": 13821
+    },
+    {
+      "epoch": 10.40421528039142,
+      "grad_norm": 0.17587728798389435,
+      "learning_rate": 5e-06,
+      "loss": 0.1033,
+      "step": 13822
+    },
+    {
+      "epoch": 10.404968009032745,
+      "grad_norm": 0.16722631454467773,
+      "learning_rate": 5e-06,
+      "loss": 0.0851,
+      "step": 13823
+    },
+    {
+      "epoch": 10.405720737674068,
+      "grad_norm": 0.1704685539007187,
+      "learning_rate": 5e-06,
+      "loss": 0.194,
+      "step": 13824
+    },
+    {
+      "epoch": 10.406473466315393,
+      "grad_norm": 0.1667814403772354,
+      "learning_rate": 5e-06,
+      "loss": 0.1164,
+      "step": 13825
+    },
+    {
+      "epoch": 10.407226194956719,
+      "grad_norm": 0.20368890464305878,
+      "learning_rate": 5e-06,
+      "loss": 0.2158,
+      "step": 13826
+    },
+    {
+      "epoch": 10.407978923598042,
+      "grad_norm": 0.16178376972675323,
+      "learning_rate": 5e-06,
+      "loss": 0.1682,
+      "step": 13827
+    },
+    {
+      "epoch": 10.408731652239368,
+      "grad_norm": 0.18553951382637024,
+      "learning_rate": 5e-06,
+      "loss": 0.2393,
+      "step": 13828
+    },
+    {
+      "epoch": 10.409484380880693,
+      "grad_norm": 0.16436487436294556,
+      "learning_rate": 5e-06,
+      "loss": 0.1733,
+      "step": 13829
+    },
+    {
+      "epoch": 10.410237109522017,
+      "grad_norm": 0.19176329672336578,
+      "learning_rate": 5e-06,
+      "loss": 0.1649,
+      "step": 13830
+    },
+    {
+      "epoch": 10.410989838163342,
+      "grad_norm": 0.21981145441532135,
+      "learning_rate": 5e-06,
+      "loss": 0.1748,
+      "step": 13831
+    },
+    {
+      "epoch": 10.411742566804667,
+      "grad_norm": 0.19966985285282135,
+      "learning_rate": 5e-06,
+      "loss": 0.1335,
+      "step": 13832
+    },
+    {
+      "epoch": 10.412495295445991,
+      "grad_norm": 0.1565265655517578,
+      "learning_rate": 5e-06,
+      "loss": 0.102,
+      "step": 13833
+    },
+    {
+      "epoch": 10.413248024087316,
+      "grad_norm": 0.19706565141677856,
+      "learning_rate": 5e-06,
+      "loss": 0.1127,
+      "step": 13834
+    },
+    {
+      "epoch": 10.414000752728642,
+      "grad_norm": 0.1795915812253952,
+      "learning_rate": 5e-06,
+      "loss": 0.2482,
+      "step": 13835
+    },
+    {
+      "epoch": 10.414753481369965,
+      "grad_norm": 0.1767095923423767,
+      "learning_rate": 5e-06,
+      "loss": 0.2105,
+      "step": 13836
+    },
+    {
+      "epoch": 10.41550621001129,
+      "grad_norm": 0.16204999387264252,
+      "learning_rate": 5e-06,
+      "loss": 0.1758,
+      "step": 13837
+    },
+    {
+      "epoch": 10.416258938652616,
+      "grad_norm": 0.20445075631141663,
+      "learning_rate": 5e-06,
+      "loss": 0.1644,
+      "step": 13838
+    },
+    {
+      "epoch": 10.41701166729394,
+      "grad_norm": 0.19239653646945953,
+      "learning_rate": 5e-06,
+      "loss": 0.1182,
+      "step": 13839
+    },
+    {
+      "epoch": 10.417764395935265,
+      "grad_norm": 0.18736033141613007,
+      "learning_rate": 5e-06,
+      "loss": 0.2085,
+      "step": 13840
+    },
+    {
+      "epoch": 10.41851712457659,
+      "grad_norm": 0.2138427495956421,
+      "learning_rate": 5e-06,
+      "loss": 0.1501,
+      "step": 13841
+    },
+    {
+      "epoch": 10.419269853217916,
+      "grad_norm": 0.1891966164112091,
+      "learning_rate": 5e-06,
+      "loss": 0.1144,
+      "step": 13842
+    },
+    {
+      "epoch": 10.42002258185924,
+      "grad_norm": 0.1830829381942749,
+      "learning_rate": 5e-06,
+      "loss": 0.1341,
+      "step": 13843
+    },
+    {
+      "epoch": 10.420775310500565,
+      "grad_norm": 0.17573603987693787,
+      "learning_rate": 5e-06,
+      "loss": 0.1293,
+      "step": 13844
+    },
+    {
+      "epoch": 10.42152803914189,
+      "grad_norm": 0.19117578864097595,
+      "learning_rate": 5e-06,
+      "loss": 0.1477,
+      "step": 13845
+    },
+    {
+      "epoch": 10.422280767783214,
+      "grad_norm": 0.1838386356830597,
+      "learning_rate": 5e-06,
+      "loss": 0.0907,
+      "step": 13846
+    },
+    {
+      "epoch": 10.423033496424539,
+      "grad_norm": 0.16478762030601501,
+      "learning_rate": 5e-06,
+      "loss": 0.1448,
+      "step": 13847
+    },
+    {
+      "epoch": 10.423786225065864,
+      "grad_norm": 0.18704140186309814,
+      "learning_rate": 5e-06,
+      "loss": 0.1843,
+      "step": 13848
+    },
+    {
+      "epoch": 10.424538953707188,
+      "grad_norm": 0.18129530549049377,
+      "learning_rate": 5e-06,
+      "loss": 0.1605,
+      "step": 13849
+    },
+    {
+      "epoch": 10.425291682348513,
+      "grad_norm": 0.1564822494983673,
+      "learning_rate": 5e-06,
+      "loss": 0.1176,
+      "step": 13850
+    },
+    {
+      "epoch": 10.426044410989839,
+      "grad_norm": 0.1804022639989853,
+      "learning_rate": 5e-06,
+      "loss": 0.1558,
+      "step": 13851
+    },
+    {
+      "epoch": 10.426797139631162,
+      "grad_norm": 0.1616307944059372,
+      "learning_rate": 5e-06,
+      "loss": 0.1261,
+      "step": 13852
+    },
+    {
+      "epoch": 10.427549868272488,
+      "grad_norm": 0.18065133690834045,
+      "learning_rate": 5e-06,
+      "loss": 0.2434,
+      "step": 13853
+    },
+    {
+      "epoch": 10.428302596913813,
+      "grad_norm": 0.1687004417181015,
+      "learning_rate": 5e-06,
+      "loss": 0.1274,
+      "step": 13854
+    },
+    {
+      "epoch": 10.429055325555137,
+      "grad_norm": 0.17577625811100006,
+      "learning_rate": 5e-06,
+      "loss": 0.1443,
+      "step": 13855
+    },
+    {
+      "epoch": 10.429808054196462,
+      "grad_norm": 0.15284843742847443,
+      "learning_rate": 5e-06,
+      "loss": 0.1333,
+      "step": 13856
+    },
+    {
+      "epoch": 10.430560782837787,
+      "grad_norm": 0.18746018409729004,
+      "learning_rate": 5e-06,
+      "loss": 0.1785,
+      "step": 13857
+    },
+    {
+      "epoch": 10.431313511479111,
+      "grad_norm": 0.19236283004283905,
+      "learning_rate": 5e-06,
+      "loss": 0.1767,
+      "step": 13858
+    },
+    {
+      "epoch": 10.432066240120436,
+      "grad_norm": 0.21406203508377075,
+      "learning_rate": 5e-06,
+      "loss": 0.2266,
+      "step": 13859
+    },
+    {
+      "epoch": 10.432818968761762,
+      "grad_norm": 0.16577060520648956,
+      "learning_rate": 5e-06,
+      "loss": 0.1849,
+      "step": 13860
+    },
+    {
+      "epoch": 10.433571697403087,
+      "grad_norm": 0.18349969387054443,
+      "learning_rate": 5e-06,
+      "loss": 0.1126,
+      "step": 13861
+    },
+    {
+      "epoch": 10.43432442604441,
+      "grad_norm": 0.16272716224193573,
+      "learning_rate": 5e-06,
+      "loss": 0.1747,
+      "step": 13862
+    },
+    {
+      "epoch": 10.435077154685736,
+      "grad_norm": 0.1536000370979309,
+      "learning_rate": 5e-06,
+      "loss": 0.1467,
+      "step": 13863
+    },
+    {
+      "epoch": 10.435829883327061,
+      "grad_norm": 0.16819386184215546,
+      "learning_rate": 5e-06,
+      "loss": 0.1532,
+      "step": 13864
+    },
+    {
+      "epoch": 10.436582611968385,
+      "grad_norm": 0.18286624550819397,
+      "learning_rate": 5e-06,
+      "loss": 0.1697,
+      "step": 13865
+    },
+    {
+      "epoch": 10.43733534060971,
+      "grad_norm": 0.17704561352729797,
+      "learning_rate": 5e-06,
+      "loss": 0.13,
+      "step": 13866
+    },
+    {
+      "epoch": 10.438088069251036,
+      "grad_norm": 0.19871672987937927,
+      "learning_rate": 5e-06,
+      "loss": 0.1677,
+      "step": 13867
+    },
+    {
+      "epoch": 10.43884079789236,
+      "grad_norm": 0.2155175805091858,
+      "learning_rate": 5e-06,
+      "loss": 0.2094,
+      "step": 13868
+    },
+    {
+      "epoch": 10.439593526533685,
+      "grad_norm": 0.1733430176973343,
+      "learning_rate": 5e-06,
+      "loss": 0.1404,
+      "step": 13869
+    },
+    {
+      "epoch": 10.44034625517501,
+      "grad_norm": 0.16154079139232635,
+      "learning_rate": 5e-06,
+      "loss": 0.1761,
+      "step": 13870
+    },
+    {
+      "epoch": 10.441098983816334,
+      "grad_norm": 0.20070935785770416,
+      "learning_rate": 5e-06,
+      "loss": 0.26,
+      "step": 13871
+    },
+    {
+      "epoch": 10.441851712457659,
+      "grad_norm": 0.16560639441013336,
+      "learning_rate": 5e-06,
+      "loss": 0.1634,
+      "step": 13872
+    },
+    {
+      "epoch": 10.442604441098984,
+      "grad_norm": 0.16525150835514069,
+      "learning_rate": 5e-06,
+      "loss": 0.1407,
+      "step": 13873
+    },
+    {
+      "epoch": 10.443357169740308,
+      "grad_norm": 0.17089803516864777,
+      "learning_rate": 5e-06,
+      "loss": 0.0938,
+      "step": 13874
+    },
+    {
+      "epoch": 10.444109898381633,
+      "grad_norm": 0.18190458416938782,
+      "learning_rate": 5e-06,
+      "loss": 0.13,
+      "step": 13875
+    },
+    {
+      "epoch": 10.444862627022959,
+      "grad_norm": 0.1659877747297287,
+      "learning_rate": 5e-06,
+      "loss": 0.1141,
+      "step": 13876
+    },
+    {
+      "epoch": 10.445615355664284,
+      "grad_norm": 0.18328312039375305,
+      "learning_rate": 5e-06,
+      "loss": 0.1076,
+      "step": 13877
+    },
+    {
+      "epoch": 10.446368084305607,
+      "grad_norm": 0.18140466511249542,
+      "learning_rate": 5e-06,
+      "loss": 0.1359,
+      "step": 13878
+    },
+    {
+      "epoch": 10.447120812946933,
+      "grad_norm": 0.2048158049583435,
+      "learning_rate": 5e-06,
+      "loss": 0.1483,
+      "step": 13879
+    },
+    {
+      "epoch": 10.447873541588258,
+      "grad_norm": 0.16788329184055328,
+      "learning_rate": 5e-06,
+      "loss": 0.136,
+      "step": 13880
+    },
+    {
+      "epoch": 10.448626270229582,
+      "grad_norm": 0.18223388493061066,
+      "learning_rate": 5e-06,
+      "loss": 0.165,
+      "step": 13881
+    },
+    {
+      "epoch": 10.449378998870907,
+      "grad_norm": 0.17753984034061432,
+      "learning_rate": 5e-06,
+      "loss": 0.1986,
+      "step": 13882
+    },
+    {
+      "epoch": 10.450131727512233,
+      "grad_norm": 0.1694202870130539,
+      "learning_rate": 5e-06,
+      "loss": 0.138,
+      "step": 13883
+    },
+    {
+      "epoch": 10.450884456153556,
+      "grad_norm": 0.17649707198143005,
+      "learning_rate": 5e-06,
+      "loss": 0.1349,
+      "step": 13884
+    },
+    {
+      "epoch": 10.451637184794881,
+      "grad_norm": 0.17099778354167938,
+      "learning_rate": 5e-06,
+      "loss": 0.1826,
+      "step": 13885
+    },
+    {
+      "epoch": 10.452389913436207,
+      "grad_norm": 0.18053025007247925,
+      "learning_rate": 5e-06,
+      "loss": 0.1263,
+      "step": 13886
+    },
+    {
+      "epoch": 10.45314264207753,
+      "grad_norm": 0.17486608028411865,
+      "learning_rate": 5e-06,
+      "loss": 0.1444,
+      "step": 13887
+    },
+    {
+      "epoch": 10.453895370718856,
+      "grad_norm": 0.1787915974855423,
+      "learning_rate": 5e-06,
+      "loss": 0.1711,
+      "step": 13888
+    },
+    {
+      "epoch": 10.454648099360181,
+      "grad_norm": 0.1644836813211441,
+      "learning_rate": 5e-06,
+      "loss": 0.1034,
+      "step": 13889
+    },
+    {
+      "epoch": 10.455400828001505,
+      "grad_norm": 0.19404391944408417,
+      "learning_rate": 5e-06,
+      "loss": 0.1611,
+      "step": 13890
+    },
+    {
+      "epoch": 10.45615355664283,
+      "grad_norm": 0.1865273118019104,
+      "learning_rate": 5e-06,
+      "loss": 0.1129,
+      "step": 13891
+    },
+    {
+      "epoch": 10.456906285284155,
+      "grad_norm": 0.18517424166202545,
+      "learning_rate": 5e-06,
+      "loss": 0.141,
+      "step": 13892
+    },
+    {
+      "epoch": 10.457659013925479,
+      "grad_norm": 0.1833188682794571,
+      "learning_rate": 5e-06,
+      "loss": 0.2613,
+      "step": 13893
+    },
+    {
+      "epoch": 10.458411742566804,
+      "grad_norm": 0.16104811429977417,
+      "learning_rate": 5e-06,
+      "loss": 0.1264,
+      "step": 13894
+    },
+    {
+      "epoch": 10.45916447120813,
+      "grad_norm": 0.17391008138656616,
+      "learning_rate": 5e-06,
+      "loss": 0.1713,
+      "step": 13895
+    },
+    {
+      "epoch": 10.459917199849455,
+      "grad_norm": 0.1806344985961914,
+      "learning_rate": 5e-06,
+      "loss": 0.1087,
+      "step": 13896
+    },
+    {
+      "epoch": 10.460669928490779,
+      "grad_norm": 0.1697186678647995,
+      "learning_rate": 5e-06,
+      "loss": 0.1361,
+      "step": 13897
+    },
+    {
+      "epoch": 10.461422657132104,
+      "grad_norm": 0.16349005699157715,
+      "learning_rate": 5e-06,
+      "loss": 0.1417,
+      "step": 13898
+    },
+    {
+      "epoch": 10.46217538577343,
+      "grad_norm": 0.16171559691429138,
+      "learning_rate": 5e-06,
+      "loss": 0.1291,
+      "step": 13899
+    },
+    {
+      "epoch": 10.462928114414753,
+      "grad_norm": 0.18444345891475677,
+      "learning_rate": 5e-06,
+      "loss": 0.1377,
+      "step": 13900
+    },
+    {
+      "epoch": 10.463680843056078,
+      "grad_norm": 0.19253993034362793,
+      "learning_rate": 5e-06,
+      "loss": 0.1466,
+      "step": 13901
+    },
+    {
+      "epoch": 10.464433571697404,
+      "grad_norm": 0.18635189533233643,
+      "learning_rate": 5e-06,
+      "loss": 0.2089,
+      "step": 13902
+    },
+    {
+      "epoch": 10.465186300338727,
+      "grad_norm": 0.170416459441185,
+      "learning_rate": 5e-06,
+      "loss": 0.1254,
+      "step": 13903
+    },
+    {
+      "epoch": 10.465939028980053,
+      "grad_norm": 0.19407759606838226,
+      "learning_rate": 5e-06,
+      "loss": 0.1972,
+      "step": 13904
+    },
+    {
+      "epoch": 10.466691757621378,
+      "grad_norm": 0.1794780194759369,
+      "learning_rate": 5e-06,
+      "loss": 0.1477,
+      "step": 13905
+    },
+    {
+      "epoch": 10.467444486262702,
+      "grad_norm": 0.16401360929012299,
+      "learning_rate": 5e-06,
+      "loss": 0.0617,
+      "step": 13906
+    },
+    {
+      "epoch": 10.468197214904027,
+      "grad_norm": 0.18694210052490234,
+      "learning_rate": 5e-06,
+      "loss": 0.0966,
+      "step": 13907
+    },
+    {
+      "epoch": 10.468949943545352,
+      "grad_norm": 0.1980971246957779,
+      "learning_rate": 5e-06,
+      "loss": 0.171,
+      "step": 13908
+    },
+    {
+      "epoch": 10.469702672186676,
+      "grad_norm": 0.17992958426475525,
+      "learning_rate": 5e-06,
+      "loss": 0.1544,
+      "step": 13909
+    },
+    {
+      "epoch": 10.470455400828001,
+      "grad_norm": 0.1738530993461609,
+      "learning_rate": 5e-06,
+      "loss": 0.1528,
+      "step": 13910
+    },
+    {
+      "epoch": 10.471208129469327,
+      "grad_norm": 0.2159535139799118,
+      "learning_rate": 5e-06,
+      "loss": 0.2177,
+      "step": 13911
+    },
+    {
+      "epoch": 10.471960858110652,
+      "grad_norm": 0.18553157150745392,
+      "learning_rate": 5e-06,
+      "loss": 0.1159,
+      "step": 13912
+    },
+    {
+      "epoch": 10.472713586751976,
+      "grad_norm": 0.18144840002059937,
+      "learning_rate": 5e-06,
+      "loss": 0.1287,
+      "step": 13913
+    },
+    {
+      "epoch": 10.473466315393301,
+      "grad_norm": 0.1614321768283844,
+      "learning_rate": 5e-06,
+      "loss": 0.1727,
+      "step": 13914
+    },
+    {
+      "epoch": 10.474219044034626,
+      "grad_norm": 0.2030748724937439,
+      "learning_rate": 5e-06,
+      "loss": 0.1845,
+      "step": 13915
+    },
+    {
+      "epoch": 10.47497177267595,
+      "grad_norm": 0.19099003076553345,
+      "learning_rate": 5e-06,
+      "loss": 0.2124,
+      "step": 13916
+    },
+    {
+      "epoch": 10.475724501317275,
+      "grad_norm": 0.1976771056652069,
+      "learning_rate": 5e-06,
+      "loss": 0.1862,
+      "step": 13917
+    },
+    {
+      "epoch": 10.4764772299586,
+      "grad_norm": 0.160480797290802,
+      "learning_rate": 5e-06,
+      "loss": 0.0976,
+      "step": 13918
+    },
+    {
+      "epoch": 10.477229958599924,
+      "grad_norm": 0.2054896354675293,
+      "learning_rate": 5e-06,
+      "loss": 0.1909,
+      "step": 13919
+    },
+    {
+      "epoch": 10.47798268724125,
+      "grad_norm": 0.16600976884365082,
+      "learning_rate": 5e-06,
+      "loss": 0.1839,
+      "step": 13920
+    },
+    {
+      "epoch": 10.478735415882575,
+      "grad_norm": 0.1834578961133957,
+      "learning_rate": 5e-06,
+      "loss": 0.1375,
+      "step": 13921
+    },
+    {
+      "epoch": 10.479488144523899,
+      "grad_norm": 0.19513866305351257,
+      "learning_rate": 5e-06,
+      "loss": 0.1497,
+      "step": 13922
+    },
+    {
+      "epoch": 10.480240873165224,
+      "grad_norm": 0.17468346655368805,
+      "learning_rate": 5e-06,
+      "loss": 0.0696,
+      "step": 13923
+    },
+    {
+      "epoch": 10.48099360180655,
+      "grad_norm": 0.186568945646286,
+      "learning_rate": 5e-06,
+      "loss": 0.122,
+      "step": 13924
+    },
+    {
+      "epoch": 10.481746330447873,
+      "grad_norm": 0.19248563051223755,
+      "learning_rate": 5e-06,
+      "loss": 0.1681,
+      "step": 13925
+    },
+    {
+      "epoch": 10.482499059089198,
+      "grad_norm": 0.1657383143901825,
+      "learning_rate": 5e-06,
+      "loss": 0.0944,
+      "step": 13926
+    },
+    {
+      "epoch": 10.483251787730524,
+      "grad_norm": 0.17990809679031372,
+      "learning_rate": 5e-06,
+      "loss": 0.1106,
+      "step": 13927
+    },
+    {
+      "epoch": 10.484004516371847,
+      "grad_norm": 0.18659695982933044,
+      "learning_rate": 5e-06,
+      "loss": 0.1507,
+      "step": 13928
+    },
+    {
+      "epoch": 10.484757245013173,
+      "grad_norm": 0.1827491670846939,
+      "learning_rate": 5e-06,
+      "loss": 0.1822,
+      "step": 13929
+    },
+    {
+      "epoch": 10.485509973654498,
+      "grad_norm": 0.17310214042663574,
+      "learning_rate": 5e-06,
+      "loss": 0.1269,
+      "step": 13930
+    },
+    {
+      "epoch": 10.486262702295821,
+      "grad_norm": 0.16758321225643158,
+      "learning_rate": 5e-06,
+      "loss": 0.0836,
+      "step": 13931
+    },
+    {
+      "epoch": 10.487015430937147,
+      "grad_norm": 0.22328265011310577,
+      "learning_rate": 5e-06,
+      "loss": 0.1944,
+      "step": 13932
+    },
+    {
+      "epoch": 10.487768159578472,
+      "grad_norm": 0.1971476674079895,
+      "learning_rate": 5e-06,
+      "loss": 0.1863,
+      "step": 13933
+    },
+    {
+      "epoch": 10.488520888219798,
+      "grad_norm": 0.1759151965379715,
+      "learning_rate": 5e-06,
+      "loss": 0.1659,
+      "step": 13934
+    },
+    {
+      "epoch": 10.489273616861121,
+      "grad_norm": 0.17573729157447815,
+      "learning_rate": 5e-06,
+      "loss": 0.179,
+      "step": 13935
+    },
+    {
+      "epoch": 10.490026345502447,
+      "grad_norm": 0.17546993494033813,
+      "learning_rate": 5e-06,
+      "loss": 0.1154,
+      "step": 13936
+    },
+    {
+      "epoch": 10.490779074143772,
+      "grad_norm": 0.18473537266254425,
+      "learning_rate": 5e-06,
+      "loss": 0.2085,
+      "step": 13937
+    },
+    {
+      "epoch": 10.491531802785095,
+      "grad_norm": 0.1739046275615692,
+      "learning_rate": 5e-06,
+      "loss": 0.0807,
+      "step": 13938
+    },
+    {
+      "epoch": 10.49228453142642,
+      "grad_norm": 0.16043664515018463,
+      "learning_rate": 5e-06,
+      "loss": 0.1387,
+      "step": 13939
+    },
+    {
+      "epoch": 10.493037260067746,
+      "grad_norm": 0.20544937252998352,
+      "learning_rate": 5e-06,
+      "loss": 0.1524,
+      "step": 13940
+    },
+    {
+      "epoch": 10.49378998870907,
+      "grad_norm": 0.18785439431667328,
+      "learning_rate": 5e-06,
+      "loss": 0.1885,
+      "step": 13941
+    },
+    {
+      "epoch": 10.494542717350395,
+      "grad_norm": 0.1904473453760147,
+      "learning_rate": 5e-06,
+      "loss": 0.1885,
+      "step": 13942
+    },
+    {
+      "epoch": 10.49529544599172,
+      "grad_norm": 0.16386203467845917,
+      "learning_rate": 5e-06,
+      "loss": 0.1442,
+      "step": 13943
+    },
+    {
+      "epoch": 10.496048174633044,
+      "grad_norm": 0.17387621104717255,
+      "learning_rate": 5e-06,
+      "loss": 0.1209,
+      "step": 13944
+    },
+    {
+      "epoch": 10.49680090327437,
+      "grad_norm": 0.18398644030094147,
+      "learning_rate": 5e-06,
+      "loss": 0.1444,
+      "step": 13945
+    },
+    {
+      "epoch": 10.497553631915695,
+      "grad_norm": 0.14533385634422302,
+      "learning_rate": 5e-06,
+      "loss": 0.1041,
+      "step": 13946
+    },
+    {
+      "epoch": 10.498306360557018,
+      "grad_norm": 0.17817825078964233,
+      "learning_rate": 5e-06,
+      "loss": 0.1511,
+      "step": 13947
+    },
+    {
+      "epoch": 10.499059089198344,
+      "grad_norm": 0.15942783653736115,
+      "learning_rate": 5e-06,
+      "loss": 0.1215,
+      "step": 13948
+    },
+    {
+      "epoch": 10.499811817839669,
+      "grad_norm": 0.16559860110282898,
+      "learning_rate": 5e-06,
+      "loss": 0.1932,
+      "step": 13949
+    },
+    {
+      "epoch": 10.500564546480994,
+      "grad_norm": 0.16932722926139832,
+      "learning_rate": 5e-06,
+      "loss": 0.1114,
+      "step": 13950
+    },
+    {
+      "epoch": 10.501317275122318,
+      "grad_norm": 0.15609434247016907,
+      "learning_rate": 5e-06,
+      "loss": 0.0981,
+      "step": 13951
+    },
+    {
+      "epoch": 10.502070003763643,
+      "grad_norm": 0.20415371656417847,
+      "learning_rate": 5e-06,
+      "loss": 0.1662,
+      "step": 13952
+    },
+    {
+      "epoch": 10.502822732404969,
+      "grad_norm": 0.18594299256801605,
+      "learning_rate": 5e-06,
+      "loss": 0.1172,
+      "step": 13953
+    },
+    {
+      "epoch": 10.503575461046292,
+      "grad_norm": 0.17119763791561127,
+      "learning_rate": 5e-06,
+      "loss": 0.1307,
+      "step": 13954
+    },
+    {
+      "epoch": 10.504328189687618,
+      "grad_norm": 0.1617899239063263,
+      "learning_rate": 5e-06,
+      "loss": 0.0992,
+      "step": 13955
+    },
+    {
+      "epoch": 10.505080918328943,
+      "grad_norm": 0.1876956671476364,
+      "learning_rate": 5e-06,
+      "loss": 0.2131,
+      "step": 13956
+    },
+    {
+      "epoch": 10.505833646970267,
+      "grad_norm": 0.2383471429347992,
+      "learning_rate": 5e-06,
+      "loss": 0.1642,
+      "step": 13957
+    },
+    {
+      "epoch": 10.506586375611592,
+      "grad_norm": 0.18146145343780518,
+      "learning_rate": 5e-06,
+      "loss": 0.2088,
+      "step": 13958
+    },
+    {
+      "epoch": 10.507339104252917,
+      "grad_norm": 0.1620083600282669,
+      "learning_rate": 5e-06,
+      "loss": 0.1003,
+      "step": 13959
+    },
+    {
+      "epoch": 10.508091832894241,
+      "grad_norm": 0.19214917719364166,
+      "learning_rate": 5e-06,
+      "loss": 0.1381,
+      "step": 13960
+    },
+    {
+      "epoch": 10.508844561535566,
+      "grad_norm": 0.15470482409000397,
+      "learning_rate": 5e-06,
+      "loss": 0.1918,
+      "step": 13961
+    },
+    {
+      "epoch": 10.509597290176892,
+      "grad_norm": 0.17410266399383545,
+      "learning_rate": 5e-06,
+      "loss": 0.0892,
+      "step": 13962
+    },
+    {
+      "epoch": 10.510350018818215,
+      "grad_norm": 0.1746608465909958,
+      "learning_rate": 5e-06,
+      "loss": 0.1086,
+      "step": 13963
+    },
+    {
+      "epoch": 10.51110274745954,
+      "grad_norm": 0.1821085512638092,
+      "learning_rate": 5e-06,
+      "loss": 0.1701,
+      "step": 13964
+    },
+    {
+      "epoch": 10.511855476100866,
+      "grad_norm": 0.1798097789287567,
+      "learning_rate": 5e-06,
+      "loss": 0.1689,
+      "step": 13965
+    },
+    {
+      "epoch": 10.51260820474219,
+      "grad_norm": 0.1802084892988205,
+      "learning_rate": 5e-06,
+      "loss": 0.2003,
+      "step": 13966
+    },
+    {
+      "epoch": 10.513360933383515,
+      "grad_norm": 0.15938805043697357,
+      "learning_rate": 5e-06,
+      "loss": 0.0615,
+      "step": 13967
+    },
+    {
+      "epoch": 10.51411366202484,
+      "grad_norm": 0.1958893984556198,
+      "learning_rate": 5e-06,
+      "loss": 0.1543,
+      "step": 13968
+    },
+    {
+      "epoch": 10.514866390666164,
+      "grad_norm": 0.1838836371898651,
+      "learning_rate": 5e-06,
+      "loss": 0.1138,
+      "step": 13969
+    },
+    {
+      "epoch": 10.51561911930749,
+      "grad_norm": 0.17272794246673584,
+      "learning_rate": 5e-06,
+      "loss": 0.1508,
+      "step": 13970
+    },
+    {
+      "epoch": 10.516371847948815,
+      "grad_norm": 0.18438977003097534,
+      "learning_rate": 5e-06,
+      "loss": 0.151,
+      "step": 13971
+    },
+    {
+      "epoch": 10.51712457659014,
+      "grad_norm": 0.19499963521957397,
+      "learning_rate": 5e-06,
+      "loss": 0.1543,
+      "step": 13972
+    },
+    {
+      "epoch": 10.517877305231464,
+      "grad_norm": 0.18701262772083282,
+      "learning_rate": 5e-06,
+      "loss": 0.1684,
+      "step": 13973
+    },
+    {
+      "epoch": 10.518630033872789,
+      "grad_norm": 0.18727551400661469,
+      "learning_rate": 5e-06,
+      "loss": 0.0895,
+      "step": 13974
+    },
+    {
+      "epoch": 10.519382762514114,
+      "grad_norm": 0.18017466366291046,
+      "learning_rate": 5e-06,
+      "loss": 0.1491,
+      "step": 13975
+    },
+    {
+      "epoch": 10.520135491155438,
+      "grad_norm": 0.166222482919693,
+      "learning_rate": 5e-06,
+      "loss": 0.1429,
+      "step": 13976
+    },
+    {
+      "epoch": 10.520888219796763,
+      "grad_norm": 0.1605900675058365,
+      "learning_rate": 5e-06,
+      "loss": 0.1246,
+      "step": 13977
+    },
+    {
+      "epoch": 10.521640948438089,
+      "grad_norm": 0.17087534070014954,
+      "learning_rate": 5e-06,
+      "loss": 0.1859,
+      "step": 13978
+    },
+    {
+      "epoch": 10.522393677079412,
+      "grad_norm": 0.18731878697872162,
+      "learning_rate": 5e-06,
+      "loss": 0.1318,
+      "step": 13979
+    },
+    {
+      "epoch": 10.523146405720738,
+      "grad_norm": 0.15776140987873077,
+      "learning_rate": 5e-06,
+      "loss": 0.0987,
+      "step": 13980
+    },
+    {
+      "epoch": 10.523899134362063,
+      "grad_norm": 0.20482608675956726,
+      "learning_rate": 5e-06,
+      "loss": 0.1812,
+      "step": 13981
+    },
+    {
+      "epoch": 10.524651863003387,
+      "grad_norm": 0.21540547907352448,
+      "learning_rate": 5e-06,
+      "loss": 0.149,
+      "step": 13982
+    },
+    {
+      "epoch": 10.525404591644712,
+      "grad_norm": 0.18606510758399963,
+      "learning_rate": 5e-06,
+      "loss": 0.1853,
+      "step": 13983
+    },
+    {
+      "epoch": 10.526157320286037,
+      "grad_norm": 0.17241600155830383,
+      "learning_rate": 5e-06,
+      "loss": 0.0854,
+      "step": 13984
+    },
+    {
+      "epoch": 10.526910048927363,
+      "grad_norm": 0.17950654029846191,
+      "learning_rate": 5e-06,
+      "loss": 0.2235,
+      "step": 13985
+    },
+    {
+      "epoch": 10.527662777568686,
+      "grad_norm": 0.17184481024742126,
+      "learning_rate": 5e-06,
+      "loss": 0.0488,
+      "step": 13986
+    },
+    {
+      "epoch": 10.528415506210012,
+      "grad_norm": 0.2100927233695984,
+      "learning_rate": 5e-06,
+      "loss": 0.171,
+      "step": 13987
+    },
+    {
+      "epoch": 10.529168234851337,
+      "grad_norm": 0.192730113863945,
+      "learning_rate": 5e-06,
+      "loss": 0.1813,
+      "step": 13988
+    },
+    {
+      "epoch": 10.52992096349266,
+      "grad_norm": 0.19934329390525818,
+      "learning_rate": 5e-06,
+      "loss": 0.1697,
+      "step": 13989
+    },
+    {
+      "epoch": 10.530673692133986,
+      "grad_norm": 0.16482369601726532,
+      "learning_rate": 5e-06,
+      "loss": 0.1172,
+      "step": 13990
+    },
+    {
+      "epoch": 10.531426420775311,
+      "grad_norm": 0.1792038083076477,
+      "learning_rate": 5e-06,
+      "loss": 0.1582,
+      "step": 13991
+    },
+    {
+      "epoch": 10.532179149416635,
+      "grad_norm": 0.1876499503850937,
+      "learning_rate": 5e-06,
+      "loss": 0.1655,
+      "step": 13992
+    },
+    {
+      "epoch": 10.53293187805796,
+      "grad_norm": 0.1953495442867279,
+      "learning_rate": 5e-06,
+      "loss": 0.1656,
+      "step": 13993
+    },
+    {
+      "epoch": 10.533684606699286,
+      "grad_norm": 0.15327592194080353,
+      "learning_rate": 5e-06,
+      "loss": 0.1068,
+      "step": 13994
+    },
+    {
+      "epoch": 10.534437335340609,
+      "grad_norm": 0.17996014654636383,
+      "learning_rate": 5e-06,
+      "loss": 0.1286,
+      "step": 13995
+    },
+    {
+      "epoch": 10.535190063981934,
+      "grad_norm": 0.17785459756851196,
+      "learning_rate": 5e-06,
+      "loss": 0.1713,
+      "step": 13996
+    },
+    {
+      "epoch": 10.53594279262326,
+      "grad_norm": 0.1852606236934662,
+      "learning_rate": 5e-06,
+      "loss": 0.2099,
+      "step": 13997
+    },
+    {
+      "epoch": 10.536695521264583,
+      "grad_norm": 0.15857484936714172,
+      "learning_rate": 5e-06,
+      "loss": 0.1287,
+      "step": 13998
+    },
+    {
+      "epoch": 10.537448249905909,
+      "grad_norm": 0.1787622570991516,
+      "learning_rate": 5e-06,
+      "loss": 0.1776,
+      "step": 13999
+    },
+    {
+      "epoch": 10.538200978547234,
+      "grad_norm": 0.18230322003364563,
+      "learning_rate": 5e-06,
+      "loss": 0.0957,
+      "step": 14000
+    },
+    {
+      "epoch": 10.538200978547234,
+      "eval_loss": 0.1883184164762497,
+      "eval_runtime": 453.1528,
+      "eval_samples_per_second": 21.244,
+      "eval_steps_per_second": 0.664,
+      "step": 14000
+    },
+    {
+      "epoch": 10.538953707188558,
+      "grad_norm": 0.17378751933574677,
+      "learning_rate": 5e-06,
+      "loss": 0.1382,
+      "step": 14001
+    },
+    {
+      "epoch": 10.539706435829883,
+      "grad_norm": 0.22039742767810822,
+      "learning_rate": 5e-06,
+      "loss": 0.1229,
+      "step": 14002
+    },
+    {
+      "epoch": 10.540459164471208,
+      "grad_norm": 0.21054911613464355,
+      "learning_rate": 5e-06,
+      "loss": 0.221,
+      "step": 14003
+    },
+    {
+      "epoch": 10.541211893112532,
+      "grad_norm": 0.16920337080955505,
+      "learning_rate": 5e-06,
+      "loss": 0.1319,
+      "step": 14004
+    },
+    {
+      "epoch": 10.541964621753857,
+      "grad_norm": 0.20208284258842468,
+      "learning_rate": 5e-06,
+      "loss": 0.1299,
+      "step": 14005
+    },
+    {
+      "epoch": 10.542717350395183,
+      "grad_norm": 0.16755008697509766,
+      "learning_rate": 5e-06,
+      "loss": 0.1093,
+      "step": 14006
+    },
+    {
+      "epoch": 10.543470079036508,
+      "grad_norm": 0.20114322006702423,
+      "learning_rate": 5e-06,
+      "loss": 0.1218,
+      "step": 14007
+    },
+    {
+      "epoch": 10.544222807677832,
+      "grad_norm": 0.17308463156223297,
+      "learning_rate": 5e-06,
+      "loss": 0.1342,
+      "step": 14008
+    },
+    {
+      "epoch": 10.544975536319157,
+      "grad_norm": 0.21862806379795074,
+      "learning_rate": 5e-06,
+      "loss": 0.1836,
+      "step": 14009
+    },
+    {
+      "epoch": 10.545728264960482,
+      "grad_norm": 0.17627635598182678,
+      "learning_rate": 5e-06,
+      "loss": 0.1926,
+      "step": 14010
+    },
+    {
+      "epoch": 10.546480993601806,
+      "grad_norm": 0.18063002824783325,
+      "learning_rate": 5e-06,
+      "loss": 0.144,
+      "step": 14011
+    },
+    {
+      "epoch": 10.547233722243131,
+      "grad_norm": 0.17180253565311432,
+      "learning_rate": 5e-06,
+      "loss": 0.1278,
+      "step": 14012
+    },
+    {
+      "epoch": 10.547986450884457,
+      "grad_norm": 0.16173627972602844,
+      "learning_rate": 5e-06,
+      "loss": 0.1554,
+      "step": 14013
+    },
+    {
+      "epoch": 10.54873917952578,
+      "grad_norm": 0.17061835527420044,
+      "learning_rate": 5e-06,
+      "loss": 0.1716,
+      "step": 14014
+    },
+    {
+      "epoch": 10.549491908167106,
+      "grad_norm": 0.18615300953388214,
+      "learning_rate": 5e-06,
+      "loss": 0.1332,
+      "step": 14015
+    },
+    {
+      "epoch": 10.550244636808431,
+      "grad_norm": 0.15071134269237518,
+      "learning_rate": 5e-06,
+      "loss": 0.158,
+      "step": 14016
+    },
+    {
+      "epoch": 10.550997365449755,
+      "grad_norm": 0.19409069418907166,
+      "learning_rate": 5e-06,
+      "loss": 0.1997,
+      "step": 14017
+    },
+    {
+      "epoch": 10.55175009409108,
+      "grad_norm": 0.18683332204818726,
+      "learning_rate": 5e-06,
+      "loss": 0.1314,
+      "step": 14018
+    },
+    {
+      "epoch": 10.552502822732405,
+      "grad_norm": 0.19399793446063995,
+      "learning_rate": 5e-06,
+      "loss": 0.199,
+      "step": 14019
+    },
+    {
+      "epoch": 10.553255551373729,
+      "grad_norm": 0.17413467168807983,
+      "learning_rate": 5e-06,
+      "loss": 0.0995,
+      "step": 14020
+    },
+    {
+      "epoch": 10.554008280015054,
+      "grad_norm": 0.16443774104118347,
+      "learning_rate": 5e-06,
+      "loss": 0.0996,
+      "step": 14021
+    },
+    {
+      "epoch": 10.55476100865638,
+      "grad_norm": 0.14917735755443573,
+      "learning_rate": 5e-06,
+      "loss": 0.1228,
+      "step": 14022
+    },
+    {
+      "epoch": 10.555513737297705,
+      "grad_norm": 0.20055173337459564,
+      "learning_rate": 5e-06,
+      "loss": 0.2369,
+      "step": 14023
+    },
+    {
+      "epoch": 10.556266465939029,
+      "grad_norm": 0.19651977717876434,
+      "learning_rate": 5e-06,
+      "loss": 0.1125,
+      "step": 14024
+    },
+    {
+      "epoch": 10.557019194580354,
+      "grad_norm": 0.16233760118484497,
+      "learning_rate": 5e-06,
+      "loss": 0.098,
+      "step": 14025
+    },
+    {
+      "epoch": 10.55777192322168,
+      "grad_norm": 0.20312751829624176,
+      "learning_rate": 5e-06,
+      "loss": 0.2158,
+      "step": 14026
+    },
+    {
+      "epoch": 10.558524651863003,
+      "grad_norm": 0.1950967162847519,
+      "learning_rate": 5e-06,
+      "loss": 0.1921,
+      "step": 14027
+    },
+    {
+      "epoch": 10.559277380504328,
+      "grad_norm": 0.20000331103801727,
+      "learning_rate": 5e-06,
+      "loss": 0.2068,
+      "step": 14028
+    },
+    {
+      "epoch": 10.560030109145654,
+      "grad_norm": 0.16762392222881317,
+      "learning_rate": 5e-06,
+      "loss": 0.1126,
+      "step": 14029
+    },
+    {
+      "epoch": 10.560782837786977,
+      "grad_norm": 0.16104844212532043,
+      "learning_rate": 5e-06,
+      "loss": 0.1104,
+      "step": 14030
+    },
+    {
+      "epoch": 10.561535566428303,
+      "grad_norm": 0.18622560799121857,
+      "learning_rate": 5e-06,
+      "loss": 0.1388,
+      "step": 14031
+    },
+    {
+      "epoch": 10.562288295069628,
+      "grad_norm": 0.1748524308204651,
+      "learning_rate": 5e-06,
+      "loss": 0.0961,
+      "step": 14032
+    },
+    {
+      "epoch": 10.563041023710952,
+      "grad_norm": 0.17698046565055847,
+      "learning_rate": 5e-06,
+      "loss": 0.0999,
+      "step": 14033
+    },
+    {
+      "epoch": 10.563793752352277,
+      "grad_norm": 0.17362529039382935,
+      "learning_rate": 5e-06,
+      "loss": 0.15,
+      "step": 14034
+    },
+    {
+      "epoch": 10.564546480993602,
+      "grad_norm": 0.1811431497335434,
+      "learning_rate": 5e-06,
+      "loss": 0.1261,
+      "step": 14035
+    },
+    {
+      "epoch": 10.565299209634926,
+      "grad_norm": 0.18709689378738403,
+      "learning_rate": 5e-06,
+      "loss": 0.1508,
+      "step": 14036
+    },
+    {
+      "epoch": 10.566051938276251,
+      "grad_norm": 0.15602658689022064,
+      "learning_rate": 5e-06,
+      "loss": 0.1026,
+      "step": 14037
+    },
+    {
+      "epoch": 10.566804666917577,
+      "grad_norm": 0.19074255228042603,
+      "learning_rate": 5e-06,
+      "loss": 0.1727,
+      "step": 14038
+    },
+    {
+      "epoch": 10.5675573955589,
+      "grad_norm": 0.20333629846572876,
+      "learning_rate": 5e-06,
+      "loss": 0.134,
+      "step": 14039
+    },
+    {
+      "epoch": 10.568310124200226,
+      "grad_norm": 0.18859919905662537,
+      "learning_rate": 5e-06,
+      "loss": 0.1574,
+      "step": 14040
+    },
+    {
+      "epoch": 10.569062852841551,
+      "grad_norm": 0.18672558665275574,
+      "learning_rate": 5e-06,
+      "loss": 0.1696,
+      "step": 14041
+    },
+    {
+      "epoch": 10.569815581482876,
+      "grad_norm": 0.18933798372745514,
+      "learning_rate": 5e-06,
+      "loss": 0.1671,
+      "step": 14042
+    },
+    {
+      "epoch": 10.5705683101242,
+      "grad_norm": 0.17085662484169006,
+      "learning_rate": 5e-06,
+      "loss": 0.1188,
+      "step": 14043
+    },
+    {
+      "epoch": 10.571321038765525,
+      "grad_norm": 0.2173951417207718,
+      "learning_rate": 5e-06,
+      "loss": 0.1963,
+      "step": 14044
+    },
+    {
+      "epoch": 10.57207376740685,
+      "grad_norm": 0.20224033296108246,
+      "learning_rate": 5e-06,
+      "loss": 0.1118,
+      "step": 14045
+    },
+    {
+      "epoch": 10.572826496048174,
+      "grad_norm": 0.18545152246952057,
+      "learning_rate": 5e-06,
+      "loss": 0.1201,
+      "step": 14046
+    },
+    {
+      "epoch": 10.5735792246895,
+      "grad_norm": 0.21842239797115326,
+      "learning_rate": 5e-06,
+      "loss": 0.1818,
+      "step": 14047
+    },
+    {
+      "epoch": 10.574331953330825,
+      "grad_norm": 0.19590376317501068,
+      "learning_rate": 5e-06,
+      "loss": 0.2236,
+      "step": 14048
+    },
+    {
+      "epoch": 10.575084681972148,
+      "grad_norm": 0.17275962233543396,
+      "learning_rate": 5e-06,
+      "loss": 0.0738,
+      "step": 14049
+    },
+    {
+      "epoch": 10.575837410613474,
+      "grad_norm": 0.1953277438879013,
+      "learning_rate": 5e-06,
+      "loss": 0.1083,
+      "step": 14050
+    },
+    {
+      "epoch": 10.5765901392548,
+      "grad_norm": 0.1993320882320404,
+      "learning_rate": 5e-06,
+      "loss": 0.1552,
+      "step": 14051
+    },
+    {
+      "epoch": 10.577342867896123,
+      "grad_norm": 0.16615980863571167,
+      "learning_rate": 5e-06,
+      "loss": 0.1041,
+      "step": 14052
+    },
+    {
+      "epoch": 10.578095596537448,
+      "grad_norm": 0.18153150379657745,
+      "learning_rate": 5e-06,
+      "loss": 0.1662,
+      "step": 14053
+    },
+    {
+      "epoch": 10.578848325178773,
+      "grad_norm": 0.15693792700767517,
+      "learning_rate": 5e-06,
+      "loss": 0.1362,
+      "step": 14054
+    },
+    {
+      "epoch": 10.579601053820097,
+      "grad_norm": 0.17989078164100647,
+      "learning_rate": 5e-06,
+      "loss": 0.1623,
+      "step": 14055
+    },
+    {
+      "epoch": 10.580353782461422,
+      "grad_norm": 0.18124215304851532,
+      "learning_rate": 5e-06,
+      "loss": 0.1352,
+      "step": 14056
+    },
+    {
+      "epoch": 10.581106511102748,
+      "grad_norm": 0.18067847192287445,
+      "learning_rate": 5e-06,
+      "loss": 0.1596,
+      "step": 14057
+    },
+    {
+      "epoch": 10.581859239744073,
+      "grad_norm": 0.16385602951049805,
+      "learning_rate": 5e-06,
+      "loss": 0.1132,
+      "step": 14058
+    },
+    {
+      "epoch": 10.582611968385397,
+      "grad_norm": 0.1696121096611023,
+      "learning_rate": 5e-06,
+      "loss": 0.117,
+      "step": 14059
+    },
+    {
+      "epoch": 10.583364697026722,
+      "grad_norm": 0.2133602350950241,
+      "learning_rate": 5e-06,
+      "loss": 0.1458,
+      "step": 14060
+    },
+    {
+      "epoch": 10.584117425668047,
+      "grad_norm": 0.17964491248130798,
+      "learning_rate": 5e-06,
+      "loss": 0.0591,
+      "step": 14061
+    },
+    {
+      "epoch": 10.584870154309371,
+      "grad_norm": 0.1744619607925415,
+      "learning_rate": 5e-06,
+      "loss": 0.072,
+      "step": 14062
+    },
+    {
+      "epoch": 10.585622882950696,
+      "grad_norm": 0.18360535800457,
+      "learning_rate": 5e-06,
+      "loss": 0.1418,
+      "step": 14063
+    },
+    {
+      "epoch": 10.586375611592022,
+      "grad_norm": 0.16674326360225677,
+      "learning_rate": 5e-06,
+      "loss": 0.1217,
+      "step": 14064
+    },
+    {
+      "epoch": 10.587128340233345,
+      "grad_norm": 0.19152094423770905,
+      "learning_rate": 5e-06,
+      "loss": 0.1588,
+      "step": 14065
+    },
+    {
+      "epoch": 10.58788106887467,
+      "grad_norm": 0.161716490983963,
+      "learning_rate": 5e-06,
+      "loss": 0.1241,
+      "step": 14066
+    },
+    {
+      "epoch": 10.588633797515996,
+      "grad_norm": 0.2210443764925003,
+      "learning_rate": 5e-06,
+      "loss": 0.2163,
+      "step": 14067
+    },
+    {
+      "epoch": 10.58938652615732,
+      "grad_norm": 0.16889317333698273,
+      "learning_rate": 5e-06,
+      "loss": 0.1185,
+      "step": 14068
+    },
+    {
+      "epoch": 10.590139254798645,
+      "grad_norm": 0.19773152470588684,
+      "learning_rate": 5e-06,
+      "loss": 0.2272,
+      "step": 14069
+    },
+    {
+      "epoch": 10.59089198343997,
+      "grad_norm": 0.18528322875499725,
+      "learning_rate": 5e-06,
+      "loss": 0.1545,
+      "step": 14070
+    },
+    {
+      "epoch": 10.591644712081294,
+      "grad_norm": 0.1815340667963028,
+      "learning_rate": 5e-06,
+      "loss": 0.1139,
+      "step": 14071
+    },
+    {
+      "epoch": 10.59239744072262,
+      "grad_norm": 0.16636212170124054,
+      "learning_rate": 5e-06,
+      "loss": 0.1365,
+      "step": 14072
+    },
+    {
+      "epoch": 10.593150169363945,
+      "grad_norm": 0.2202916443347931,
+      "learning_rate": 5e-06,
+      "loss": 0.1141,
+      "step": 14073
+    },
+    {
+      "epoch": 10.593902898005268,
+      "grad_norm": 0.1790512204170227,
+      "learning_rate": 5e-06,
+      "loss": 0.0967,
+      "step": 14074
+    },
+    {
+      "epoch": 10.594655626646594,
+      "grad_norm": 0.19719789922237396,
+      "learning_rate": 5e-06,
+      "loss": 0.184,
+      "step": 14075
+    },
+    {
+      "epoch": 10.595408355287919,
+      "grad_norm": 0.1870943158864975,
+      "learning_rate": 5e-06,
+      "loss": 0.1508,
+      "step": 14076
+    },
+    {
+      "epoch": 10.596161083929243,
+      "grad_norm": 0.17368581891059875,
+      "learning_rate": 5e-06,
+      "loss": 0.1881,
+      "step": 14077
+    },
+    {
+      "epoch": 10.596913812570568,
+      "grad_norm": 0.19527573883533478,
+      "learning_rate": 5e-06,
+      "loss": 0.2047,
+      "step": 14078
+    },
+    {
+      "epoch": 10.597666541211893,
+      "grad_norm": 0.19786716997623444,
+      "learning_rate": 5e-06,
+      "loss": 0.1854,
+      "step": 14079
+    },
+    {
+      "epoch": 10.598419269853219,
+      "grad_norm": 0.17155130207538605,
+      "learning_rate": 5e-06,
+      "loss": 0.1003,
+      "step": 14080
+    },
+    {
+      "epoch": 10.599171998494542,
+      "grad_norm": 0.17208550870418549,
+      "learning_rate": 5e-06,
+      "loss": 0.1447,
+      "step": 14081
+    },
+    {
+      "epoch": 10.599924727135868,
+      "grad_norm": 0.17500820755958557,
+      "learning_rate": 5e-06,
+      "loss": 0.1884,
+      "step": 14082
+    },
+    {
+      "epoch": 10.600677455777193,
+      "grad_norm": 0.19354018568992615,
+      "learning_rate": 5e-06,
+      "loss": 0.1915,
+      "step": 14083
+    },
+    {
+      "epoch": 10.601430184418517,
+      "grad_norm": 0.19854825735092163,
+      "learning_rate": 5e-06,
+      "loss": 0.222,
+      "step": 14084
+    },
+    {
+      "epoch": 10.602182913059842,
+      "grad_norm": 0.1881929486989975,
+      "learning_rate": 5e-06,
+      "loss": 0.1569,
+      "step": 14085
+    },
+    {
+      "epoch": 10.602935641701167,
+      "grad_norm": 0.16965532302856445,
+      "learning_rate": 5e-06,
+      "loss": 0.1229,
+      "step": 14086
+    },
+    {
+      "epoch": 10.603688370342491,
+      "grad_norm": 0.1729069948196411,
+      "learning_rate": 5e-06,
+      "loss": 0.1515,
+      "step": 14087
+    },
+    {
+      "epoch": 10.604441098983816,
+      "grad_norm": 0.17938067018985748,
+      "learning_rate": 5e-06,
+      "loss": 0.1296,
+      "step": 14088
+    },
+    {
+      "epoch": 10.605193827625142,
+      "grad_norm": 0.15545877814292908,
+      "learning_rate": 5e-06,
+      "loss": 0.107,
+      "step": 14089
+    },
+    {
+      "epoch": 10.605946556266465,
+      "grad_norm": 0.17570017278194427,
+      "learning_rate": 5e-06,
+      "loss": 0.1642,
+      "step": 14090
+    },
+    {
+      "epoch": 10.60669928490779,
+      "grad_norm": 0.18339082598686218,
+      "learning_rate": 5e-06,
+      "loss": 0.163,
+      "step": 14091
+    },
+    {
+      "epoch": 10.607452013549116,
+      "grad_norm": 0.16545113921165466,
+      "learning_rate": 5e-06,
+      "loss": 0.1212,
+      "step": 14092
+    },
+    {
+      "epoch": 10.60820474219044,
+      "grad_norm": 0.17430195212364197,
+      "learning_rate": 5e-06,
+      "loss": 0.1326,
+      "step": 14093
+    },
+    {
+      "epoch": 10.608957470831765,
+      "grad_norm": 0.20193861424922943,
+      "learning_rate": 5e-06,
+      "loss": 0.1067,
+      "step": 14094
+    },
+    {
+      "epoch": 10.60971019947309,
+      "grad_norm": 0.17527875304222107,
+      "learning_rate": 5e-06,
+      "loss": 0.12,
+      "step": 14095
+    },
+    {
+      "epoch": 10.610462928114416,
+      "grad_norm": 0.17821556329727173,
+      "learning_rate": 5e-06,
+      "loss": 0.1501,
+      "step": 14096
+    },
+    {
+      "epoch": 10.61121565675574,
+      "grad_norm": 0.16338536143302917,
+      "learning_rate": 5e-06,
+      "loss": 0.1307,
+      "step": 14097
+    },
+    {
+      "epoch": 10.611968385397065,
+      "grad_norm": 0.1899128407239914,
+      "learning_rate": 5e-06,
+      "loss": 0.1095,
+      "step": 14098
+    },
+    {
+      "epoch": 10.61272111403839,
+      "grad_norm": 0.18102794885635376,
+      "learning_rate": 5e-06,
+      "loss": 0.1389,
+      "step": 14099
+    },
+    {
+      "epoch": 10.613473842679713,
+      "grad_norm": 0.20105966925621033,
+      "learning_rate": 5e-06,
+      "loss": 0.2631,
+      "step": 14100
+    },
+    {
+      "epoch": 10.614226571321039,
+      "grad_norm": 0.1848464459180832,
+      "learning_rate": 5e-06,
+      "loss": 0.2124,
+      "step": 14101
+    },
+    {
+      "epoch": 10.614979299962364,
+      "grad_norm": 0.1718348264694214,
+      "learning_rate": 5e-06,
+      "loss": 0.1534,
+      "step": 14102
+    },
+    {
+      "epoch": 10.615732028603688,
+      "grad_norm": 0.1569463461637497,
+      "learning_rate": 5e-06,
+      "loss": 0.1324,
+      "step": 14103
+    },
+    {
+      "epoch": 10.616484757245013,
+      "grad_norm": 0.17359201610088348,
+      "learning_rate": 5e-06,
+      "loss": 0.121,
+      "step": 14104
+    },
+    {
+      "epoch": 10.617237485886339,
+      "grad_norm": 0.17818842828273773,
+      "learning_rate": 5e-06,
+      "loss": 0.1769,
+      "step": 14105
+    },
+    {
+      "epoch": 10.617990214527662,
+      "grad_norm": 0.15221494436264038,
+      "learning_rate": 5e-06,
+      "loss": 0.0798,
+      "step": 14106
+    },
+    {
+      "epoch": 10.618742943168987,
+      "grad_norm": 0.20418907701969147,
+      "learning_rate": 5e-06,
+      "loss": 0.1195,
+      "step": 14107
+    },
+    {
+      "epoch": 10.619495671810313,
+      "grad_norm": 0.1695852279663086,
+      "learning_rate": 5e-06,
+      "loss": 0.1582,
+      "step": 14108
+    },
+    {
+      "epoch": 10.620248400451636,
+      "grad_norm": 0.17736101150512695,
+      "learning_rate": 5e-06,
+      "loss": 0.1778,
+      "step": 14109
+    },
+    {
+      "epoch": 10.621001129092962,
+      "grad_norm": 0.17781518399715424,
+      "learning_rate": 5e-06,
+      "loss": 0.1176,
+      "step": 14110
+    },
+    {
+      "epoch": 10.621753857734287,
+      "grad_norm": 0.17341256141662598,
+      "learning_rate": 5e-06,
+      "loss": 0.137,
+      "step": 14111
+    },
+    {
+      "epoch": 10.62250658637561,
+      "grad_norm": 0.1845449060201645,
+      "learning_rate": 5e-06,
+      "loss": 0.0829,
+      "step": 14112
+    },
+    {
+      "epoch": 10.623259315016936,
+      "grad_norm": 0.18692374229431152,
+      "learning_rate": 5e-06,
+      "loss": 0.1148,
+      "step": 14113
+    },
+    {
+      "epoch": 10.624012043658261,
+      "grad_norm": 0.17043359577655792,
+      "learning_rate": 5e-06,
+      "loss": 0.1478,
+      "step": 14114
+    },
+    {
+      "epoch": 10.624764772299587,
+      "grad_norm": 0.176884725689888,
+      "learning_rate": 5e-06,
+      "loss": 0.1266,
+      "step": 14115
+    },
+    {
+      "epoch": 10.62551750094091,
+      "grad_norm": 0.17479874193668365,
+      "learning_rate": 5e-06,
+      "loss": 0.1008,
+      "step": 14116
+    },
+    {
+      "epoch": 10.626270229582236,
+      "grad_norm": 0.16793736815452576,
+      "learning_rate": 5e-06,
+      "loss": 0.13,
+      "step": 14117
+    },
+    {
+      "epoch": 10.627022958223561,
+      "grad_norm": 0.14567799866199493,
+      "learning_rate": 5e-06,
+      "loss": 0.0777,
+      "step": 14118
+    },
+    {
+      "epoch": 10.627775686864885,
+      "grad_norm": 0.1714901477098465,
+      "learning_rate": 5e-06,
+      "loss": 0.1849,
+      "step": 14119
+    },
+    {
+      "epoch": 10.62852841550621,
+      "grad_norm": 0.19993983209133148,
+      "learning_rate": 5e-06,
+      "loss": 0.1413,
+      "step": 14120
+    },
+    {
+      "epoch": 10.629281144147535,
+      "grad_norm": 0.15315857529640198,
+      "learning_rate": 5e-06,
+      "loss": 0.1312,
+      "step": 14121
+    },
+    {
+      "epoch": 10.630033872788859,
+      "grad_norm": 0.144384503364563,
+      "learning_rate": 5e-06,
+      "loss": 0.0812,
+      "step": 14122
+    },
+    {
+      "epoch": 10.630786601430184,
+      "grad_norm": 0.1857648342847824,
+      "learning_rate": 5e-06,
+      "loss": 0.1516,
+      "step": 14123
+    },
+    {
+      "epoch": 10.63153933007151,
+      "grad_norm": 0.18286332488059998,
+      "learning_rate": 5e-06,
+      "loss": 0.1071,
+      "step": 14124
+    },
+    {
+      "epoch": 10.632292058712833,
+      "grad_norm": 0.19617398083209991,
+      "learning_rate": 5e-06,
+      "loss": 0.1691,
+      "step": 14125
+    },
+    {
+      "epoch": 10.633044787354159,
+      "grad_norm": 0.18648973107337952,
+      "learning_rate": 5e-06,
+      "loss": 0.1184,
+      "step": 14126
+    },
+    {
+      "epoch": 10.633797515995484,
+      "grad_norm": 0.17805127799510956,
+      "learning_rate": 5e-06,
+      "loss": 0.1509,
+      "step": 14127
+    },
+    {
+      "epoch": 10.634550244636808,
+      "grad_norm": 0.18279209733009338,
+      "learning_rate": 5e-06,
+      "loss": 0.1807,
+      "step": 14128
+    },
+    {
+      "epoch": 10.635302973278133,
+      "grad_norm": 0.19440364837646484,
+      "learning_rate": 5e-06,
+      "loss": 0.1607,
+      "step": 14129
+    },
+    {
+      "epoch": 10.636055701919458,
+      "grad_norm": 0.168212890625,
+      "learning_rate": 5e-06,
+      "loss": 0.1987,
+      "step": 14130
+    },
+    {
+      "epoch": 10.636808430560784,
+      "grad_norm": 0.18154560029506683,
+      "learning_rate": 5e-06,
+      "loss": 0.0795,
+      "step": 14131
+    },
+    {
+      "epoch": 10.637561159202107,
+      "grad_norm": 0.16531644761562347,
+      "learning_rate": 5e-06,
+      "loss": 0.106,
+      "step": 14132
+    },
+    {
+      "epoch": 10.638313887843433,
+      "grad_norm": 0.20095092058181763,
+      "learning_rate": 5e-06,
+      "loss": 0.2195,
+      "step": 14133
+    },
+    {
+      "epoch": 10.639066616484758,
+      "grad_norm": 0.2290249913930893,
+      "learning_rate": 5e-06,
+      "loss": 0.1956,
+      "step": 14134
+    },
+    {
+      "epoch": 10.639819345126082,
+      "grad_norm": 0.18053996562957764,
+      "learning_rate": 5e-06,
+      "loss": 0.1541,
+      "step": 14135
+    },
+    {
+      "epoch": 10.640572073767407,
+      "grad_norm": 0.16842041909694672,
+      "learning_rate": 5e-06,
+      "loss": 0.133,
+      "step": 14136
+    },
+    {
+      "epoch": 10.641324802408732,
+      "grad_norm": 0.1981242150068283,
+      "learning_rate": 5e-06,
+      "loss": 0.0891,
+      "step": 14137
+    },
+    {
+      "epoch": 10.642077531050056,
+      "grad_norm": 0.1842576116323471,
+      "learning_rate": 5e-06,
+      "loss": 0.1814,
+      "step": 14138
+    },
+    {
+      "epoch": 10.642830259691381,
+      "grad_norm": 0.17265890538692474,
+      "learning_rate": 5e-06,
+      "loss": 0.1544,
+      "step": 14139
+    },
+    {
+      "epoch": 10.643582988332707,
+      "grad_norm": 0.1935025304555893,
+      "learning_rate": 5e-06,
+      "loss": 0.1676,
+      "step": 14140
+    },
+    {
+      "epoch": 10.64433571697403,
+      "grad_norm": 0.17867420613765717,
+      "learning_rate": 5e-06,
+      "loss": 0.1143,
+      "step": 14141
+    },
+    {
+      "epoch": 10.645088445615356,
+      "grad_norm": 0.18369537591934204,
+      "learning_rate": 5e-06,
+      "loss": 0.1366,
+      "step": 14142
+    },
+    {
+      "epoch": 10.645841174256681,
+      "grad_norm": 0.1758817434310913,
+      "learning_rate": 5e-06,
+      "loss": 0.0822,
+      "step": 14143
+    },
+    {
+      "epoch": 10.646593902898005,
+      "grad_norm": 0.17176593840122223,
+      "learning_rate": 5e-06,
+      "loss": 0.1868,
+      "step": 14144
+    },
+    {
+      "epoch": 10.64734663153933,
+      "grad_norm": 0.1923086941242218,
+      "learning_rate": 5e-06,
+      "loss": 0.232,
+      "step": 14145
+    },
+    {
+      "epoch": 10.648099360180655,
+      "grad_norm": 0.1931944489479065,
+      "learning_rate": 5e-06,
+      "loss": 0.215,
+      "step": 14146
+    },
+    {
+      "epoch": 10.648852088821979,
+      "grad_norm": 0.20107628405094147,
+      "learning_rate": 5e-06,
+      "loss": 0.1903,
+      "step": 14147
+    },
+    {
+      "epoch": 10.649604817463304,
+      "grad_norm": 0.17417779564857483,
+      "learning_rate": 5e-06,
+      "loss": 0.1201,
+      "step": 14148
+    },
+    {
+      "epoch": 10.65035754610463,
+      "grad_norm": 0.18210744857788086,
+      "learning_rate": 5e-06,
+      "loss": 0.1111,
+      "step": 14149
+    },
+    {
+      "epoch": 10.651110274745953,
+      "grad_norm": 0.19191163778305054,
+      "learning_rate": 5e-06,
+      "loss": 0.1409,
+      "step": 14150
+    },
+    {
+      "epoch": 10.651863003387279,
+      "grad_norm": 0.17590861022472382,
+      "learning_rate": 5e-06,
+      "loss": 0.1533,
+      "step": 14151
+    },
+    {
+      "epoch": 10.652615732028604,
+      "grad_norm": 0.22969642281532288,
+      "learning_rate": 5e-06,
+      "loss": 0.1489,
+      "step": 14152
+    },
+    {
+      "epoch": 10.65336846066993,
+      "grad_norm": 0.20536181330680847,
+      "learning_rate": 5e-06,
+      "loss": 0.1208,
+      "step": 14153
+    },
+    {
+      "epoch": 10.654121189311253,
+      "grad_norm": 0.19049963355064392,
+      "learning_rate": 5e-06,
+      "loss": 0.2016,
+      "step": 14154
+    },
+    {
+      "epoch": 10.654873917952578,
+      "grad_norm": 0.1965624988079071,
+      "learning_rate": 5e-06,
+      "loss": 0.1756,
+      "step": 14155
+    },
+    {
+      "epoch": 10.655626646593904,
+      "grad_norm": 0.1595124751329422,
+      "learning_rate": 5e-06,
+      "loss": 0.0709,
+      "step": 14156
+    },
+    {
+      "epoch": 10.656379375235227,
+      "grad_norm": 0.1828056275844574,
+      "learning_rate": 5e-06,
+      "loss": 0.1061,
+      "step": 14157
+    },
+    {
+      "epoch": 10.657132103876553,
+      "grad_norm": 0.1977677345275879,
+      "learning_rate": 5e-06,
+      "loss": 0.0673,
+      "step": 14158
+    },
+    {
+      "epoch": 10.657884832517878,
+      "grad_norm": 0.18634426593780518,
+      "learning_rate": 5e-06,
+      "loss": 0.1256,
+      "step": 14159
+    },
+    {
+      "epoch": 10.658637561159201,
+      "grad_norm": 0.1827855408191681,
+      "learning_rate": 5e-06,
+      "loss": 0.0969,
+      "step": 14160
+    },
+    {
+      "epoch": 10.659390289800527,
+      "grad_norm": 0.17190538346767426,
+      "learning_rate": 5e-06,
+      "loss": 0.1303,
+      "step": 14161
+    },
+    {
+      "epoch": 10.660143018441852,
+      "grad_norm": 0.21105384826660156,
+      "learning_rate": 5e-06,
+      "loss": 0.1725,
+      "step": 14162
+    },
+    {
+      "epoch": 10.660895747083176,
+      "grad_norm": 0.22822622954845428,
+      "learning_rate": 5e-06,
+      "loss": 0.2611,
+      "step": 14163
+    },
+    {
+      "epoch": 10.661648475724501,
+      "grad_norm": 0.18824443221092224,
+      "learning_rate": 5e-06,
+      "loss": 0.1717,
+      "step": 14164
+    },
+    {
+      "epoch": 10.662401204365827,
+      "grad_norm": 0.16759951412677765,
+      "learning_rate": 5e-06,
+      "loss": 0.1705,
+      "step": 14165
+    },
+    {
+      "epoch": 10.663153933007152,
+      "grad_norm": 0.17268957197666168,
+      "learning_rate": 5e-06,
+      "loss": 0.1238,
+      "step": 14166
+    },
+    {
+      "epoch": 10.663906661648475,
+      "grad_norm": 0.18026062846183777,
+      "learning_rate": 5e-06,
+      "loss": 0.0928,
+      "step": 14167
+    },
+    {
+      "epoch": 10.6646593902898,
+      "grad_norm": 0.17766423523426056,
+      "learning_rate": 5e-06,
+      "loss": 0.101,
+      "step": 14168
+    },
+    {
+      "epoch": 10.665412118931126,
+      "grad_norm": 0.18136750161647797,
+      "learning_rate": 5e-06,
+      "loss": 0.1544,
+      "step": 14169
+    },
+    {
+      "epoch": 10.66616484757245,
+      "grad_norm": 0.20394344627857208,
+      "learning_rate": 5e-06,
+      "loss": 0.0903,
+      "step": 14170
+    },
+    {
+      "epoch": 10.666917576213775,
+      "grad_norm": 0.16325587034225464,
+      "learning_rate": 5e-06,
+      "loss": 0.1348,
+      "step": 14171
+    },
+    {
+      "epoch": 10.6676703048551,
+      "grad_norm": 0.17852406203746796,
+      "learning_rate": 5e-06,
+      "loss": 0.1665,
+      "step": 14172
+    },
+    {
+      "epoch": 10.668423033496424,
+      "grad_norm": 0.2004937380552292,
+      "learning_rate": 5e-06,
+      "loss": 0.1851,
+      "step": 14173
+    },
+    {
+      "epoch": 10.66917576213775,
+      "grad_norm": 0.1670045703649521,
+      "learning_rate": 5e-06,
+      "loss": 0.1031,
+      "step": 14174
+    },
+    {
+      "epoch": 10.669928490779075,
+      "grad_norm": 0.18446142971515656,
+      "learning_rate": 5e-06,
+      "loss": 0.241,
+      "step": 14175
+    },
+    {
+      "epoch": 10.670681219420398,
+      "grad_norm": 0.1864936798810959,
+      "learning_rate": 5e-06,
+      "loss": 0.1019,
+      "step": 14176
+    },
+    {
+      "epoch": 10.671433948061724,
+      "grad_norm": 0.1811881959438324,
+      "learning_rate": 5e-06,
+      "loss": 0.16,
+      "step": 14177
+    },
+    {
+      "epoch": 10.672186676703049,
+      "grad_norm": 0.1744629442691803,
+      "learning_rate": 5e-06,
+      "loss": 0.1175,
+      "step": 14178
+    },
+    {
+      "epoch": 10.672939405344373,
+      "grad_norm": 0.1956765055656433,
+      "learning_rate": 5e-06,
+      "loss": 0.1282,
+      "step": 14179
+    },
+    {
+      "epoch": 10.673692133985698,
+      "grad_norm": 0.21257953345775604,
+      "learning_rate": 5e-06,
+      "loss": 0.1218,
+      "step": 14180
+    },
+    {
+      "epoch": 10.674444862627023,
+      "grad_norm": 0.15615709125995636,
+      "learning_rate": 5e-06,
+      "loss": 0.0752,
+      "step": 14181
+    },
+    {
+      "epoch": 10.675197591268347,
+      "grad_norm": 0.17187552154064178,
+      "learning_rate": 5e-06,
+      "loss": 0.2023,
+      "step": 14182
+    },
+    {
+      "epoch": 10.675950319909672,
+      "grad_norm": 0.1718648225069046,
+      "learning_rate": 5e-06,
+      "loss": 0.1357,
+      "step": 14183
+    },
+    {
+      "epoch": 10.676703048550998,
+      "grad_norm": 0.19882267713546753,
+      "learning_rate": 5e-06,
+      "loss": 0.1357,
+      "step": 14184
+    },
+    {
+      "epoch": 10.677455777192321,
+      "grad_norm": 0.16063126921653748,
+      "learning_rate": 5e-06,
+      "loss": 0.1736,
+      "step": 14185
+    },
+    {
+      "epoch": 10.678208505833647,
+      "grad_norm": 0.1775631159543991,
+      "learning_rate": 5e-06,
+      "loss": 0.1182,
+      "step": 14186
+    },
+    {
+      "epoch": 10.678961234474972,
+      "grad_norm": 0.18709872663021088,
+      "learning_rate": 5e-06,
+      "loss": 0.0956,
+      "step": 14187
+    },
+    {
+      "epoch": 10.679713963116297,
+      "grad_norm": 0.16665463149547577,
+      "learning_rate": 5e-06,
+      "loss": 0.2008,
+      "step": 14188
+    },
+    {
+      "epoch": 10.680466691757621,
+      "grad_norm": 0.19401930272579193,
+      "learning_rate": 5e-06,
+      "loss": 0.1469,
+      "step": 14189
+    },
+    {
+      "epoch": 10.681219420398946,
+      "grad_norm": 0.18648721277713776,
+      "learning_rate": 5e-06,
+      "loss": 0.1545,
+      "step": 14190
+    },
+    {
+      "epoch": 10.681972149040272,
+      "grad_norm": 0.16623848676681519,
+      "learning_rate": 5e-06,
+      "loss": 0.0993,
+      "step": 14191
+    },
+    {
+      "epoch": 10.682724877681595,
+      "grad_norm": 0.19178946316242218,
+      "learning_rate": 5e-06,
+      "loss": 0.1776,
+      "step": 14192
+    },
+    {
+      "epoch": 10.68347760632292,
+      "grad_norm": 0.1788882613182068,
+      "learning_rate": 5e-06,
+      "loss": 0.1249,
+      "step": 14193
+    },
+    {
+      "epoch": 10.684230334964246,
+      "grad_norm": 0.1718546748161316,
+      "learning_rate": 5e-06,
+      "loss": 0.1676,
+      "step": 14194
+    },
+    {
+      "epoch": 10.68498306360557,
+      "grad_norm": 0.1933562308549881,
+      "learning_rate": 5e-06,
+      "loss": 0.2088,
+      "step": 14195
+    },
+    {
+      "epoch": 10.685735792246895,
+      "grad_norm": 0.21555154025554657,
+      "learning_rate": 5e-06,
+      "loss": 0.2199,
+      "step": 14196
+    },
+    {
+      "epoch": 10.68648852088822,
+      "grad_norm": 0.17848455905914307,
+      "learning_rate": 5e-06,
+      "loss": 0.2031,
+      "step": 14197
+    },
+    {
+      "epoch": 10.687241249529544,
+      "grad_norm": 0.19473129510879517,
+      "learning_rate": 5e-06,
+      "loss": 0.182,
+      "step": 14198
+    },
+    {
+      "epoch": 10.68799397817087,
+      "grad_norm": 0.1799483299255371,
+      "learning_rate": 5e-06,
+      "loss": 0.1261,
+      "step": 14199
+    },
+    {
+      "epoch": 10.688746706812195,
+      "grad_norm": 0.16298842430114746,
+      "learning_rate": 5e-06,
+      "loss": 0.0942,
+      "step": 14200
+    },
+    {
+      "epoch": 10.688746706812195,
+      "eval_loss": 0.18904021382331848,
+      "eval_runtime": 452.7276,
+      "eval_samples_per_second": 21.264,
+      "eval_steps_per_second": 0.665,
+      "step": 14200
+    },
+    {
+      "epoch": 10.689499435453518,
+      "grad_norm": 0.21136699616909027,
+      "learning_rate": 5e-06,
+      "loss": 0.1219,
+      "step": 14201
+    },
+    {
+      "epoch": 10.690252164094844,
+      "grad_norm": 0.18229041993618011,
+      "learning_rate": 5e-06,
+      "loss": 0.187,
+      "step": 14202
+    },
+    {
+      "epoch": 10.691004892736169,
+      "grad_norm": 0.1581963151693344,
+      "learning_rate": 5e-06,
+      "loss": 0.1171,
+      "step": 14203
+    },
+    {
+      "epoch": 10.691757621377494,
+      "grad_norm": 0.19341634213924408,
+      "learning_rate": 5e-06,
+      "loss": 0.2075,
+      "step": 14204
+    },
+    {
+      "epoch": 10.692510350018818,
+      "grad_norm": 0.19413398206233978,
+      "learning_rate": 5e-06,
+      "loss": 0.1479,
+      "step": 14205
+    },
+    {
+      "epoch": 10.693263078660143,
+      "grad_norm": 0.19371551275253296,
+      "learning_rate": 5e-06,
+      "loss": 0.2256,
+      "step": 14206
+    },
+    {
+      "epoch": 10.694015807301469,
+      "grad_norm": 0.17630566656589508,
+      "learning_rate": 5e-06,
+      "loss": 0.1257,
+      "step": 14207
+    },
+    {
+      "epoch": 10.694768535942792,
+      "grad_norm": 0.20503418147563934,
+      "learning_rate": 5e-06,
+      "loss": 0.1882,
+      "step": 14208
+    },
+    {
+      "epoch": 10.695521264584118,
+      "grad_norm": 0.19183139503002167,
+      "learning_rate": 5e-06,
+      "loss": 0.1912,
+      "step": 14209
+    },
+    {
+      "epoch": 10.696273993225443,
+      "grad_norm": 0.2090710997581482,
+      "learning_rate": 5e-06,
+      "loss": 0.1612,
+      "step": 14210
+    },
+    {
+      "epoch": 10.697026721866767,
+      "grad_norm": 0.19281622767448425,
+      "learning_rate": 5e-06,
+      "loss": 0.202,
+      "step": 14211
+    },
+    {
+      "epoch": 10.697779450508092,
+      "grad_norm": 0.17207348346710205,
+      "learning_rate": 5e-06,
+      "loss": 0.1153,
+      "step": 14212
+    },
+    {
+      "epoch": 10.698532179149417,
+      "grad_norm": 0.19428320229053497,
+      "learning_rate": 5e-06,
+      "loss": 0.0814,
+      "step": 14213
+    },
+    {
+      "epoch": 10.69928490779074,
+      "grad_norm": 0.18575014173984528,
+      "learning_rate": 5e-06,
+      "loss": 0.1778,
+      "step": 14214
+    },
+    {
+      "epoch": 10.700037636432066,
+      "grad_norm": 0.18575656414031982,
+      "learning_rate": 5e-06,
+      "loss": 0.1456,
+      "step": 14215
+    },
+    {
+      "epoch": 10.700790365073392,
+      "grad_norm": 0.18791240453720093,
+      "learning_rate": 5e-06,
+      "loss": 0.1582,
+      "step": 14216
+    },
+    {
+      "epoch": 10.701543093714715,
+      "grad_norm": 0.17241305112838745,
+      "learning_rate": 5e-06,
+      "loss": 0.1551,
+      "step": 14217
+    },
+    {
+      "epoch": 10.70229582235604,
+      "grad_norm": 0.14885956048965454,
+      "learning_rate": 5e-06,
+      "loss": 0.1027,
+      "step": 14218
+    },
+    {
+      "epoch": 10.703048550997366,
+      "grad_norm": 0.1796952784061432,
+      "learning_rate": 5e-06,
+      "loss": 0.1567,
+      "step": 14219
+    },
+    {
+      "epoch": 10.70380127963869,
+      "grad_norm": 0.16960673034191132,
+      "learning_rate": 5e-06,
+      "loss": 0.0923,
+      "step": 14220
+    },
+    {
+      "epoch": 10.704554008280015,
+      "grad_norm": 0.17820776998996735,
+      "learning_rate": 5e-06,
+      "loss": 0.1543,
+      "step": 14221
+    },
+    {
+      "epoch": 10.70530673692134,
+      "grad_norm": 0.18459270894527435,
+      "learning_rate": 5e-06,
+      "loss": 0.1249,
+      "step": 14222
+    },
+    {
+      "epoch": 10.706059465562664,
+      "grad_norm": 0.2050110548734665,
+      "learning_rate": 5e-06,
+      "loss": 0.1871,
+      "step": 14223
+    },
+    {
+      "epoch": 10.706812194203989,
+      "grad_norm": 0.19486911594867706,
+      "learning_rate": 5e-06,
+      "loss": 0.1816,
+      "step": 14224
+    },
+    {
+      "epoch": 10.707564922845314,
+      "grad_norm": 0.18358857929706573,
+      "learning_rate": 5e-06,
+      "loss": 0.1176,
+      "step": 14225
+    },
+    {
+      "epoch": 10.70831765148664,
+      "grad_norm": 0.2017323225736618,
+      "learning_rate": 5e-06,
+      "loss": 0.1428,
+      "step": 14226
+    },
+    {
+      "epoch": 10.709070380127963,
+      "grad_norm": 0.21479013562202454,
+      "learning_rate": 5e-06,
+      "loss": 0.2483,
+      "step": 14227
+    },
+    {
+      "epoch": 10.709823108769289,
+      "grad_norm": 0.16832071542739868,
+      "learning_rate": 5e-06,
+      "loss": 0.1811,
+      "step": 14228
+    },
+    {
+      "epoch": 10.710575837410614,
+      "grad_norm": 0.1756107360124588,
+      "learning_rate": 5e-06,
+      "loss": 0.1766,
+      "step": 14229
+    },
+    {
+      "epoch": 10.711328566051938,
+      "grad_norm": 0.1583676040172577,
+      "learning_rate": 5e-06,
+      "loss": 0.107,
+      "step": 14230
+    },
+    {
+      "epoch": 10.712081294693263,
+      "grad_norm": 0.16699270904064178,
+      "learning_rate": 5e-06,
+      "loss": 0.1245,
+      "step": 14231
+    },
+    {
+      "epoch": 10.712834023334588,
+      "grad_norm": 0.17899957299232483,
+      "learning_rate": 5e-06,
+      "loss": 0.1624,
+      "step": 14232
+    },
+    {
+      "epoch": 10.713586751975912,
+      "grad_norm": 0.18752402067184448,
+      "learning_rate": 5e-06,
+      "loss": 0.1697,
+      "step": 14233
+    },
+    {
+      "epoch": 10.714339480617237,
+      "grad_norm": 0.18102386593818665,
+      "learning_rate": 5e-06,
+      "loss": 0.1157,
+      "step": 14234
+    },
+    {
+      "epoch": 10.715092209258563,
+      "grad_norm": 0.19886547327041626,
+      "learning_rate": 5e-06,
+      "loss": 0.1225,
+      "step": 14235
+    },
+    {
+      "epoch": 10.715844937899886,
+      "grad_norm": 0.18530632555484772,
+      "learning_rate": 5e-06,
+      "loss": 0.0821,
+      "step": 14236
+    },
+    {
+      "epoch": 10.716597666541212,
+      "grad_norm": 0.19013622403144836,
+      "learning_rate": 5e-06,
+      "loss": 0.1778,
+      "step": 14237
+    },
+    {
+      "epoch": 10.717350395182537,
+      "grad_norm": 0.1814160794019699,
+      "learning_rate": 5e-06,
+      "loss": 0.1209,
+      "step": 14238
+    },
+    {
+      "epoch": 10.718103123823862,
+      "grad_norm": 0.16173163056373596,
+      "learning_rate": 5e-06,
+      "loss": 0.1115,
+      "step": 14239
+    },
+    {
+      "epoch": 10.718855852465186,
+      "grad_norm": 0.16646137833595276,
+      "learning_rate": 5e-06,
+      "loss": 0.2083,
+      "step": 14240
+    },
+    {
+      "epoch": 10.719608581106511,
+      "grad_norm": 0.20016776025295258,
+      "learning_rate": 5e-06,
+      "loss": 0.15,
+      "step": 14241
+    },
+    {
+      "epoch": 10.720361309747837,
+      "grad_norm": 0.2063586413860321,
+      "learning_rate": 5e-06,
+      "loss": 0.1723,
+      "step": 14242
+    },
+    {
+      "epoch": 10.72111403838916,
+      "grad_norm": 0.1779755800962448,
+      "learning_rate": 5e-06,
+      "loss": 0.1823,
+      "step": 14243
+    },
+    {
+      "epoch": 10.721866767030486,
+      "grad_norm": 0.17475219070911407,
+      "learning_rate": 5e-06,
+      "loss": 0.1027,
+      "step": 14244
+    },
+    {
+      "epoch": 10.722619495671811,
+      "grad_norm": 0.19524870812892914,
+      "learning_rate": 5e-06,
+      "loss": 0.1775,
+      "step": 14245
+    },
+    {
+      "epoch": 10.723372224313135,
+      "grad_norm": 0.1625683754682541,
+      "learning_rate": 5e-06,
+      "loss": 0.0787,
+      "step": 14246
+    },
+    {
+      "epoch": 10.72412495295446,
+      "grad_norm": 0.2321576029062271,
+      "learning_rate": 5e-06,
+      "loss": 0.1077,
+      "step": 14247
+    },
+    {
+      "epoch": 10.724877681595785,
+      "grad_norm": 0.16647788882255554,
+      "learning_rate": 5e-06,
+      "loss": 0.0981,
+      "step": 14248
+    },
+    {
+      "epoch": 10.725630410237109,
+      "grad_norm": 0.18819843232631683,
+      "learning_rate": 5e-06,
+      "loss": 0.1739,
+      "step": 14249
+    },
+    {
+      "epoch": 10.726383138878434,
+      "grad_norm": 0.16981413960456848,
+      "learning_rate": 5e-06,
+      "loss": 0.0963,
+      "step": 14250
+    },
+    {
+      "epoch": 10.72713586751976,
+      "grad_norm": 0.18027541041374207,
+      "learning_rate": 5e-06,
+      "loss": 0.113,
+      "step": 14251
+    },
+    {
+      "epoch": 10.727888596161083,
+      "grad_norm": 0.18491001427173615,
+      "learning_rate": 5e-06,
+      "loss": 0.1348,
+      "step": 14252
+    },
+    {
+      "epoch": 10.728641324802409,
+      "grad_norm": 0.1744569092988968,
+      "learning_rate": 5e-06,
+      "loss": 0.0979,
+      "step": 14253
+    },
+    {
+      "epoch": 10.729394053443734,
+      "grad_norm": 0.16700303554534912,
+      "learning_rate": 5e-06,
+      "loss": 0.1843,
+      "step": 14254
+    },
+    {
+      "epoch": 10.730146782085058,
+      "grad_norm": 0.18042497336864471,
+      "learning_rate": 5e-06,
+      "loss": 0.1639,
+      "step": 14255
+    },
+    {
+      "epoch": 10.730899510726383,
+      "grad_norm": 0.17015276849269867,
+      "learning_rate": 5e-06,
+      "loss": 0.1667,
+      "step": 14256
+    },
+    {
+      "epoch": 10.731652239367708,
+      "grad_norm": 0.1716407984495163,
+      "learning_rate": 5e-06,
+      "loss": 0.1234,
+      "step": 14257
+    },
+    {
+      "epoch": 10.732404968009032,
+      "grad_norm": 0.18838664889335632,
+      "learning_rate": 5e-06,
+      "loss": 0.1635,
+      "step": 14258
+    },
+    {
+      "epoch": 10.733157696650357,
+      "grad_norm": 0.17517231404781342,
+      "learning_rate": 5e-06,
+      "loss": 0.1188,
+      "step": 14259
+    },
+    {
+      "epoch": 10.733910425291683,
+      "grad_norm": 0.16941092908382416,
+      "learning_rate": 5e-06,
+      "loss": 0.1268,
+      "step": 14260
+    },
+    {
+      "epoch": 10.734663153933008,
+      "grad_norm": 0.16080307960510254,
+      "learning_rate": 5e-06,
+      "loss": 0.1805,
+      "step": 14261
+    },
+    {
+      "epoch": 10.735415882574332,
+      "grad_norm": 0.1700969934463501,
+      "learning_rate": 5e-06,
+      "loss": 0.1072,
+      "step": 14262
+    },
+    {
+      "epoch": 10.736168611215657,
+      "grad_norm": 0.1903751790523529,
+      "learning_rate": 5e-06,
+      "loss": 0.1541,
+      "step": 14263
+    },
+    {
+      "epoch": 10.736921339856982,
+      "grad_norm": 0.1906186044216156,
+      "learning_rate": 5e-06,
+      "loss": 0.1653,
+      "step": 14264
+    },
+    {
+      "epoch": 10.737674068498306,
+      "grad_norm": 0.19306690990924835,
+      "learning_rate": 5e-06,
+      "loss": 0.209,
+      "step": 14265
+    },
+    {
+      "epoch": 10.738426797139631,
+      "grad_norm": 0.18937371671199799,
+      "learning_rate": 5e-06,
+      "loss": 0.1934,
+      "step": 14266
+    },
+    {
+      "epoch": 10.739179525780957,
+      "grad_norm": 0.17740172147750854,
+      "learning_rate": 5e-06,
+      "loss": 0.1732,
+      "step": 14267
+    },
+    {
+      "epoch": 10.73993225442228,
+      "grad_norm": 0.1941055953502655,
+      "learning_rate": 5e-06,
+      "loss": 0.2176,
+      "step": 14268
+    },
+    {
+      "epoch": 10.740684983063606,
+      "grad_norm": 0.18336471915245056,
+      "learning_rate": 5e-06,
+      "loss": 0.2154,
+      "step": 14269
+    },
+    {
+      "epoch": 10.741437711704931,
+      "grad_norm": 0.17336612939834595,
+      "learning_rate": 5e-06,
+      "loss": 0.1512,
+      "step": 14270
+    },
+    {
+      "epoch": 10.742190440346254,
+      "grad_norm": 0.17474253475666046,
+      "learning_rate": 5e-06,
+      "loss": 0.1145,
+      "step": 14271
+    },
+    {
+      "epoch": 10.74294316898758,
+      "grad_norm": 0.1772441416978836,
+      "learning_rate": 5e-06,
+      "loss": 0.1388,
+      "step": 14272
+    },
+    {
+      "epoch": 10.743695897628905,
+      "grad_norm": 0.15980173647403717,
+      "learning_rate": 5e-06,
+      "loss": 0.1097,
+      "step": 14273
+    },
+    {
+      "epoch": 10.744448626270229,
+      "grad_norm": 0.18341164290905,
+      "learning_rate": 5e-06,
+      "loss": 0.136,
+      "step": 14274
+    },
+    {
+      "epoch": 10.745201354911554,
+      "grad_norm": 0.19923943281173706,
+      "learning_rate": 5e-06,
+      "loss": 0.1924,
+      "step": 14275
+    },
+    {
+      "epoch": 10.74595408355288,
+      "grad_norm": 0.14827021956443787,
+      "learning_rate": 5e-06,
+      "loss": 0.0995,
+      "step": 14276
+    },
+    {
+      "epoch": 10.746706812194205,
+      "grad_norm": 0.18148590624332428,
+      "learning_rate": 5e-06,
+      "loss": 0.134,
+      "step": 14277
+    },
+    {
+      "epoch": 10.747459540835528,
+      "grad_norm": 0.17166344821453094,
+      "learning_rate": 5e-06,
+      "loss": 0.1893,
+      "step": 14278
+    },
+    {
+      "epoch": 10.748212269476854,
+      "grad_norm": 0.190485879778862,
+      "learning_rate": 5e-06,
+      "loss": 0.1285,
+      "step": 14279
+    },
+    {
+      "epoch": 10.74896499811818,
+      "grad_norm": 0.15847855806350708,
+      "learning_rate": 5e-06,
+      "loss": 0.1176,
+      "step": 14280
+    },
+    {
+      "epoch": 10.749717726759503,
+      "grad_norm": 0.18720708787441254,
+      "learning_rate": 5e-06,
+      "loss": 0.1906,
+      "step": 14281
+    },
+    {
+      "epoch": 10.750470455400828,
+      "grad_norm": 0.17465345561504364,
+      "learning_rate": 5e-06,
+      "loss": 0.1866,
+      "step": 14282
+    },
+    {
+      "epoch": 10.751223184042153,
+      "grad_norm": 0.19241197407245636,
+      "learning_rate": 5e-06,
+      "loss": 0.2271,
+      "step": 14283
+    },
+    {
+      "epoch": 10.751975912683477,
+      "grad_norm": 0.18984541296958923,
+      "learning_rate": 5e-06,
+      "loss": 0.1549,
+      "step": 14284
+    },
+    {
+      "epoch": 10.752728641324802,
+      "grad_norm": 0.18669393658638,
+      "learning_rate": 5e-06,
+      "loss": 0.1238,
+      "step": 14285
+    },
+    {
+      "epoch": 10.753481369966128,
+      "grad_norm": 0.19135376811027527,
+      "learning_rate": 5e-06,
+      "loss": 0.1478,
+      "step": 14286
+    },
+    {
+      "epoch": 10.754234098607451,
+      "grad_norm": 0.18011362850666046,
+      "learning_rate": 5e-06,
+      "loss": 0.1539,
+      "step": 14287
+    },
+    {
+      "epoch": 10.754986827248777,
+      "grad_norm": 0.16716057062149048,
+      "learning_rate": 5e-06,
+      "loss": 0.1087,
+      "step": 14288
+    },
+    {
+      "epoch": 10.755739555890102,
+      "grad_norm": 0.17798861861228943,
+      "learning_rate": 5e-06,
+      "loss": 0.1391,
+      "step": 14289
+    },
+    {
+      "epoch": 10.756492284531426,
+      "grad_norm": 0.17154191434383392,
+      "learning_rate": 5e-06,
+      "loss": 0.1109,
+      "step": 14290
+    },
+    {
+      "epoch": 10.757245013172751,
+      "grad_norm": 0.19872383773326874,
+      "learning_rate": 5e-06,
+      "loss": 0.1577,
+      "step": 14291
+    },
+    {
+      "epoch": 10.757997741814076,
+      "grad_norm": 0.18913504481315613,
+      "learning_rate": 5e-06,
+      "loss": 0.1817,
+      "step": 14292
+    },
+    {
+      "epoch": 10.7587504704554,
+      "grad_norm": 0.16796855628490448,
+      "learning_rate": 5e-06,
+      "loss": 0.1084,
+      "step": 14293
+    },
+    {
+      "epoch": 10.759503199096725,
+      "grad_norm": 0.20493605732917786,
+      "learning_rate": 5e-06,
+      "loss": 0.1513,
+      "step": 14294
+    },
+    {
+      "epoch": 10.76025592773805,
+      "grad_norm": 0.18375861644744873,
+      "learning_rate": 5e-06,
+      "loss": 0.128,
+      "step": 14295
+    },
+    {
+      "epoch": 10.761008656379374,
+      "grad_norm": 0.18258127570152283,
+      "learning_rate": 5e-06,
+      "loss": 0.1346,
+      "step": 14296
+    },
+    {
+      "epoch": 10.7617613850207,
+      "grad_norm": 0.1887219399213791,
+      "learning_rate": 5e-06,
+      "loss": 0.1862,
+      "step": 14297
+    },
+    {
+      "epoch": 10.762514113662025,
+      "grad_norm": 0.2059863954782486,
+      "learning_rate": 5e-06,
+      "loss": 0.2273,
+      "step": 14298
+    },
+    {
+      "epoch": 10.76326684230335,
+      "grad_norm": 0.174458846449852,
+      "learning_rate": 5e-06,
+      "loss": 0.0921,
+      "step": 14299
+    },
+    {
+      "epoch": 10.764019570944674,
+      "grad_norm": 0.19078442454338074,
+      "learning_rate": 5e-06,
+      "loss": 0.1842,
+      "step": 14300
+    },
+    {
+      "epoch": 10.764772299586,
+      "grad_norm": 0.19728179275989532,
+      "learning_rate": 5e-06,
+      "loss": 0.1887,
+      "step": 14301
+    },
+    {
+      "epoch": 10.765525028227325,
+      "grad_norm": 0.17573677003383636,
+      "learning_rate": 5e-06,
+      "loss": 0.1726,
+      "step": 14302
+    },
+    {
+      "epoch": 10.766277756868648,
+      "grad_norm": 0.17526525259017944,
+      "learning_rate": 5e-06,
+      "loss": 0.1496,
+      "step": 14303
+    },
+    {
+      "epoch": 10.767030485509974,
+      "grad_norm": 0.16668865084648132,
+      "learning_rate": 5e-06,
+      "loss": 0.1112,
+      "step": 14304
+    },
+    {
+      "epoch": 10.767783214151299,
+      "grad_norm": 0.16608810424804688,
+      "learning_rate": 5e-06,
+      "loss": 0.1391,
+      "step": 14305
+    },
+    {
+      "epoch": 10.768535942792623,
+      "grad_norm": 0.17057017982006073,
+      "learning_rate": 5e-06,
+      "loss": 0.1233,
+      "step": 14306
+    },
+    {
+      "epoch": 10.769288671433948,
+      "grad_norm": 0.1812014877796173,
+      "learning_rate": 5e-06,
+      "loss": 0.2546,
+      "step": 14307
+    },
+    {
+      "epoch": 10.770041400075273,
+      "grad_norm": 0.20330265164375305,
+      "learning_rate": 5e-06,
+      "loss": 0.206,
+      "step": 14308
+    },
+    {
+      "epoch": 10.770794128716597,
+      "grad_norm": 0.19667254388332367,
+      "learning_rate": 5e-06,
+      "loss": 0.1549,
+      "step": 14309
+    },
+    {
+      "epoch": 10.771546857357922,
+      "grad_norm": 0.18861545622348785,
+      "learning_rate": 5e-06,
+      "loss": 0.1245,
+      "step": 14310
+    },
+    {
+      "epoch": 10.772299585999248,
+      "grad_norm": 0.1506573110818863,
+      "learning_rate": 5e-06,
+      "loss": 0.108,
+      "step": 14311
+    },
+    {
+      "epoch": 10.773052314640573,
+      "grad_norm": 0.17775166034698486,
+      "learning_rate": 5e-06,
+      "loss": 0.1603,
+      "step": 14312
+    },
+    {
+      "epoch": 10.773805043281897,
+      "grad_norm": 0.17200922966003418,
+      "learning_rate": 5e-06,
+      "loss": 0.1192,
+      "step": 14313
+    },
+    {
+      "epoch": 10.774557771923222,
+      "grad_norm": 0.16886352002620697,
+      "learning_rate": 5e-06,
+      "loss": 0.086,
+      "step": 14314
+    },
+    {
+      "epoch": 10.775310500564547,
+      "grad_norm": 0.15908409655094147,
+      "learning_rate": 5e-06,
+      "loss": 0.1445,
+      "step": 14315
+    },
+    {
+      "epoch": 10.776063229205871,
+      "grad_norm": 0.17372994124889374,
+      "learning_rate": 5e-06,
+      "loss": 0.1678,
+      "step": 14316
+    },
+    {
+      "epoch": 10.776815957847196,
+      "grad_norm": 0.16043832898139954,
+      "learning_rate": 5e-06,
+      "loss": 0.0912,
+      "step": 14317
+    },
+    {
+      "epoch": 10.777568686488522,
+      "grad_norm": 0.18514734506607056,
+      "learning_rate": 5e-06,
+      "loss": 0.1231,
+      "step": 14318
+    },
+    {
+      "epoch": 10.778321415129845,
+      "grad_norm": 0.16953116655349731,
+      "learning_rate": 5e-06,
+      "loss": 0.1196,
+      "step": 14319
+    },
+    {
+      "epoch": 10.77907414377117,
+      "grad_norm": 0.18054795265197754,
+      "learning_rate": 5e-06,
+      "loss": 0.1593,
+      "step": 14320
+    },
+    {
+      "epoch": 10.779826872412496,
+      "grad_norm": 0.16372032463550568,
+      "learning_rate": 5e-06,
+      "loss": 0.1064,
+      "step": 14321
+    },
+    {
+      "epoch": 10.78057960105382,
+      "grad_norm": 0.16017001867294312,
+      "learning_rate": 5e-06,
+      "loss": 0.1214,
+      "step": 14322
+    },
+    {
+      "epoch": 10.781332329695145,
+      "grad_norm": 0.17691095173358917,
+      "learning_rate": 5e-06,
+      "loss": 0.2087,
+      "step": 14323
+    },
+    {
+      "epoch": 10.78208505833647,
+      "grad_norm": 0.17341923713684082,
+      "learning_rate": 5e-06,
+      "loss": 0.1481,
+      "step": 14324
+    },
+    {
+      "epoch": 10.782837786977794,
+      "grad_norm": 0.15879607200622559,
+      "learning_rate": 5e-06,
+      "loss": 0.1007,
+      "step": 14325
+    },
+    {
+      "epoch": 10.78359051561912,
+      "grad_norm": 0.20140665769577026,
+      "learning_rate": 5e-06,
+      "loss": 0.1248,
+      "step": 14326
+    },
+    {
+      "epoch": 10.784343244260445,
+      "grad_norm": 0.20892249047756195,
+      "learning_rate": 5e-06,
+      "loss": 0.2251,
+      "step": 14327
+    },
+    {
+      "epoch": 10.785095972901768,
+      "grad_norm": 0.16252590715885162,
+      "learning_rate": 5e-06,
+      "loss": 0.1183,
+      "step": 14328
+    },
+    {
+      "epoch": 10.785848701543093,
+      "grad_norm": 0.19594071805477142,
+      "learning_rate": 5e-06,
+      "loss": 0.1425,
+      "step": 14329
+    },
+    {
+      "epoch": 10.786601430184419,
+      "grad_norm": 0.1694783866405487,
+      "learning_rate": 5e-06,
+      "loss": 0.1326,
+      "step": 14330
+    },
+    {
+      "epoch": 10.787354158825742,
+      "grad_norm": 0.17838139832019806,
+      "learning_rate": 5e-06,
+      "loss": 0.2864,
+      "step": 14331
+    },
+    {
+      "epoch": 10.788106887467068,
+      "grad_norm": 0.20569156110286713,
+      "learning_rate": 5e-06,
+      "loss": 0.129,
+      "step": 14332
+    },
+    {
+      "epoch": 10.788859616108393,
+      "grad_norm": 0.1694013923406601,
+      "learning_rate": 5e-06,
+      "loss": 0.0817,
+      "step": 14333
+    },
+    {
+      "epoch": 10.789612344749719,
+      "grad_norm": 0.2258712649345398,
+      "learning_rate": 5e-06,
+      "loss": 0.1617,
+      "step": 14334
+    },
+    {
+      "epoch": 10.790365073391042,
+      "grad_norm": 0.18872800469398499,
+      "learning_rate": 5e-06,
+      "loss": 0.1421,
+      "step": 14335
+    },
+    {
+      "epoch": 10.791117802032367,
+      "grad_norm": 0.17513225972652435,
+      "learning_rate": 5e-06,
+      "loss": 0.0947,
+      "step": 14336
+    },
+    {
+      "epoch": 10.791870530673693,
+      "grad_norm": 0.20942072570323944,
+      "learning_rate": 5e-06,
+      "loss": 0.1077,
+      "step": 14337
+    },
+    {
+      "epoch": 10.792623259315016,
+      "grad_norm": 0.1680697202682495,
+      "learning_rate": 5e-06,
+      "loss": 0.0993,
+      "step": 14338
+    },
+    {
+      "epoch": 10.793375987956342,
+      "grad_norm": 0.16651295125484467,
+      "learning_rate": 5e-06,
+      "loss": 0.13,
+      "step": 14339
+    },
+    {
+      "epoch": 10.794128716597667,
+      "grad_norm": 0.19678440690040588,
+      "learning_rate": 5e-06,
+      "loss": 0.1884,
+      "step": 14340
+    },
+    {
+      "epoch": 10.79488144523899,
+      "grad_norm": 0.19529303908348083,
+      "learning_rate": 5e-06,
+      "loss": 0.1118,
+      "step": 14341
+    },
+    {
+      "epoch": 10.795634173880316,
+      "grad_norm": 0.1746082454919815,
+      "learning_rate": 5e-06,
+      "loss": 0.1434,
+      "step": 14342
+    },
+    {
+      "epoch": 10.796386902521641,
+      "grad_norm": 0.16689176857471466,
+      "learning_rate": 5e-06,
+      "loss": 0.1064,
+      "step": 14343
+    },
+    {
+      "epoch": 10.797139631162965,
+      "grad_norm": 0.17295730113983154,
+      "learning_rate": 5e-06,
+      "loss": 0.0913,
+      "step": 14344
+    },
+    {
+      "epoch": 10.79789235980429,
+      "grad_norm": 0.2117011547088623,
+      "learning_rate": 5e-06,
+      "loss": 0.2193,
+      "step": 14345
+    },
+    {
+      "epoch": 10.798645088445616,
+      "grad_norm": 0.1839875429868698,
+      "learning_rate": 5e-06,
+      "loss": 0.1575,
+      "step": 14346
+    },
+    {
+      "epoch": 10.79939781708694,
+      "grad_norm": 0.17947563529014587,
+      "learning_rate": 5e-06,
+      "loss": 0.1202,
+      "step": 14347
+    },
+    {
+      "epoch": 10.800150545728265,
+      "grad_norm": 0.16612295806407928,
+      "learning_rate": 5e-06,
+      "loss": 0.1437,
+      "step": 14348
+    },
+    {
+      "epoch": 10.80090327436959,
+      "grad_norm": 0.18644589185714722,
+      "learning_rate": 5e-06,
+      "loss": 0.1758,
+      "step": 14349
+    },
+    {
+      "epoch": 10.801656003010915,
+      "grad_norm": 0.20253746211528778,
+      "learning_rate": 5e-06,
+      "loss": 0.1325,
+      "step": 14350
+    },
+    {
+      "epoch": 10.802408731652239,
+      "grad_norm": 0.21048317849636078,
+      "learning_rate": 5e-06,
+      "loss": 0.2065,
+      "step": 14351
+    },
+    {
+      "epoch": 10.803161460293564,
+      "grad_norm": 0.17799779772758484,
+      "learning_rate": 5e-06,
+      "loss": 0.1627,
+      "step": 14352
+    },
+    {
+      "epoch": 10.80391418893489,
+      "grad_norm": 0.15666459500789642,
+      "learning_rate": 5e-06,
+      "loss": 0.0991,
+      "step": 14353
+    },
+    {
+      "epoch": 10.804666917576213,
+      "grad_norm": 0.17331592738628387,
+      "learning_rate": 5e-06,
+      "loss": 0.1431,
+      "step": 14354
+    },
+    {
+      "epoch": 10.805419646217539,
+      "grad_norm": 0.19864878058433533,
+      "learning_rate": 5e-06,
+      "loss": 0.1224,
+      "step": 14355
+    },
+    {
+      "epoch": 10.806172374858864,
+      "grad_norm": 0.1582428514957428,
+      "learning_rate": 5e-06,
+      "loss": 0.1281,
+      "step": 14356
+    },
+    {
+      "epoch": 10.806925103500188,
+      "grad_norm": 0.15551209449768066,
+      "learning_rate": 5e-06,
+      "loss": 0.1539,
+      "step": 14357
+    },
+    {
+      "epoch": 10.807677832141513,
+      "grad_norm": 0.19819554686546326,
+      "learning_rate": 5e-06,
+      "loss": 0.0905,
+      "step": 14358
+    },
+    {
+      "epoch": 10.808430560782838,
+      "grad_norm": 0.17021606862545013,
+      "learning_rate": 5e-06,
+      "loss": 0.0691,
+      "step": 14359
+    },
+    {
+      "epoch": 10.809183289424162,
+      "grad_norm": 0.17668762803077698,
+      "learning_rate": 5e-06,
+      "loss": 0.1362,
+      "step": 14360
+    },
+    {
+      "epoch": 10.809936018065487,
+      "grad_norm": 0.18722005188465118,
+      "learning_rate": 5e-06,
+      "loss": 0.1463,
+      "step": 14361
+    },
+    {
+      "epoch": 10.810688746706813,
+      "grad_norm": 0.17914147675037384,
+      "learning_rate": 5e-06,
+      "loss": 0.0888,
+      "step": 14362
+    },
+    {
+      "epoch": 10.811441475348136,
+      "grad_norm": 0.1860346794128418,
+      "learning_rate": 5e-06,
+      "loss": 0.1183,
+      "step": 14363
+    },
+    {
+      "epoch": 10.812194203989462,
+      "grad_norm": 0.17533567547798157,
+      "learning_rate": 5e-06,
+      "loss": 0.0715,
+      "step": 14364
+    },
+    {
+      "epoch": 10.812946932630787,
+      "grad_norm": 0.2010514885187149,
+      "learning_rate": 5e-06,
+      "loss": 0.1544,
+      "step": 14365
+    },
+    {
+      "epoch": 10.81369966127211,
+      "grad_norm": 0.19825440645217896,
+      "learning_rate": 5e-06,
+      "loss": 0.1636,
+      "step": 14366
+    },
+    {
+      "epoch": 10.814452389913436,
+      "grad_norm": 0.16717754304409027,
+      "learning_rate": 5e-06,
+      "loss": 0.1757,
+      "step": 14367
+    },
+    {
+      "epoch": 10.815205118554761,
+      "grad_norm": 0.17620573937892914,
+      "learning_rate": 5e-06,
+      "loss": 0.1362,
+      "step": 14368
+    },
+    {
+      "epoch": 10.815957847196085,
+      "grad_norm": 0.17616774141788483,
+      "learning_rate": 5e-06,
+      "loss": 0.1029,
+      "step": 14369
+    },
+    {
+      "epoch": 10.81671057583741,
+      "grad_norm": 0.19668516516685486,
+      "learning_rate": 5e-06,
+      "loss": 0.1709,
+      "step": 14370
+    },
+    {
+      "epoch": 10.817463304478736,
+      "grad_norm": 0.23510922491550446,
+      "learning_rate": 5e-06,
+      "loss": 0.1842,
+      "step": 14371
+    },
+    {
+      "epoch": 10.818216033120061,
+      "grad_norm": 0.1686299592256546,
+      "learning_rate": 5e-06,
+      "loss": 0.0938,
+      "step": 14372
+    },
+    {
+      "epoch": 10.818968761761385,
+      "grad_norm": 0.1851203888654709,
+      "learning_rate": 5e-06,
+      "loss": 0.1503,
+      "step": 14373
+    },
+    {
+      "epoch": 10.81972149040271,
+      "grad_norm": 0.19933705031871796,
+      "learning_rate": 5e-06,
+      "loss": 0.1961,
+      "step": 14374
+    },
+    {
+      "epoch": 10.820474219044035,
+      "grad_norm": 0.17796222865581512,
+      "learning_rate": 5e-06,
+      "loss": 0.0617,
+      "step": 14375
+    },
+    {
+      "epoch": 10.821226947685359,
+      "grad_norm": 0.1772078424692154,
+      "learning_rate": 5e-06,
+      "loss": 0.1275,
+      "step": 14376
+    },
+    {
+      "epoch": 10.821979676326684,
+      "grad_norm": 0.17279092967510223,
+      "learning_rate": 5e-06,
+      "loss": 0.1442,
+      "step": 14377
+    },
+    {
+      "epoch": 10.82273240496801,
+      "grad_norm": 0.18674275279045105,
+      "learning_rate": 5e-06,
+      "loss": 0.1376,
+      "step": 14378
+    },
+    {
+      "epoch": 10.823485133609333,
+      "grad_norm": 0.20066037774085999,
+      "learning_rate": 5e-06,
+      "loss": 0.164,
+      "step": 14379
+    },
+    {
+      "epoch": 10.824237862250659,
+      "grad_norm": 0.15604010224342346,
+      "learning_rate": 5e-06,
+      "loss": 0.0932,
+      "step": 14380
+    },
+    {
+      "epoch": 10.824990590891984,
+      "grad_norm": 0.19690877199172974,
+      "learning_rate": 5e-06,
+      "loss": 0.116,
+      "step": 14381
+    },
+    {
+      "epoch": 10.825743319533307,
+      "grad_norm": 0.18184177577495575,
+      "learning_rate": 5e-06,
+      "loss": 0.1353,
+      "step": 14382
+    },
+    {
+      "epoch": 10.826496048174633,
+      "grad_norm": 0.18252037465572357,
+      "learning_rate": 5e-06,
+      "loss": 0.168,
+      "step": 14383
+    },
+    {
+      "epoch": 10.827248776815958,
+      "grad_norm": 0.17027324438095093,
+      "learning_rate": 5e-06,
+      "loss": 0.1381,
+      "step": 14384
+    },
+    {
+      "epoch": 10.828001505457284,
+      "grad_norm": 0.18737775087356567,
+      "learning_rate": 5e-06,
+      "loss": 0.1482,
+      "step": 14385
+    },
+    {
+      "epoch": 10.828754234098607,
+      "grad_norm": 0.17817124724388123,
+      "learning_rate": 5e-06,
+      "loss": 0.2125,
+      "step": 14386
+    },
+    {
+      "epoch": 10.829506962739933,
+      "grad_norm": 0.2023610770702362,
+      "learning_rate": 5e-06,
+      "loss": 0.2595,
+      "step": 14387
+    },
+    {
+      "epoch": 10.830259691381258,
+      "grad_norm": 0.16786988079547882,
+      "learning_rate": 5e-06,
+      "loss": 0.1709,
+      "step": 14388
+    },
+    {
+      "epoch": 10.831012420022581,
+      "grad_norm": 0.18353241682052612,
+      "learning_rate": 5e-06,
+      "loss": 0.1512,
+      "step": 14389
+    },
+    {
+      "epoch": 10.831765148663907,
+      "grad_norm": 0.15544822812080383,
+      "learning_rate": 5e-06,
+      "loss": 0.1273,
+      "step": 14390
+    },
+    {
+      "epoch": 10.832517877305232,
+      "grad_norm": 0.17652922868728638,
+      "learning_rate": 5e-06,
+      "loss": 0.1027,
+      "step": 14391
+    },
+    {
+      "epoch": 10.833270605946556,
+      "grad_norm": 0.18970225751399994,
+      "learning_rate": 5e-06,
+      "loss": 0.2677,
+      "step": 14392
+    },
+    {
+      "epoch": 10.834023334587881,
+      "grad_norm": 0.17192205786705017,
+      "learning_rate": 5e-06,
+      "loss": 0.0928,
+      "step": 14393
+    },
+    {
+      "epoch": 10.834776063229206,
+      "grad_norm": 0.19697120785713196,
+      "learning_rate": 5e-06,
+      "loss": 0.1231,
+      "step": 14394
+    },
+    {
+      "epoch": 10.83552879187053,
+      "grad_norm": 0.1893579661846161,
+      "learning_rate": 5e-06,
+      "loss": 0.1276,
+      "step": 14395
+    },
+    {
+      "epoch": 10.836281520511855,
+      "grad_norm": 0.17557284235954285,
+      "learning_rate": 5e-06,
+      "loss": 0.104,
+      "step": 14396
+    },
+    {
+      "epoch": 10.83703424915318,
+      "grad_norm": 0.18502259254455566,
+      "learning_rate": 5e-06,
+      "loss": 0.1029,
+      "step": 14397
+    },
+    {
+      "epoch": 10.837786977794504,
+      "grad_norm": 0.16586610674858093,
+      "learning_rate": 5e-06,
+      "loss": 0.1232,
+      "step": 14398
+    },
+    {
+      "epoch": 10.83853970643583,
+      "grad_norm": 0.17571398615837097,
+      "learning_rate": 5e-06,
+      "loss": 0.1278,
+      "step": 14399
+    },
+    {
+      "epoch": 10.839292435077155,
+      "grad_norm": 0.17180366814136505,
+      "learning_rate": 5e-06,
+      "loss": 0.1412,
+      "step": 14400
+    },
+    {
+      "epoch": 10.839292435077155,
+      "eval_loss": 0.1875005066394806,
+      "eval_runtime": 452.6867,
+      "eval_samples_per_second": 21.266,
+      "eval_steps_per_second": 0.665,
+      "step": 14400
+    },
+    {
+      "epoch": 10.840045163718479,
+      "grad_norm": 0.17679090797901154,
+      "learning_rate": 5e-06,
+      "loss": 0.0774,
+      "step": 14401
+    },
+    {
+      "epoch": 10.840797892359804,
+      "grad_norm": 0.15261703729629517,
+      "learning_rate": 5e-06,
+      "loss": 0.117,
+      "step": 14402
+    },
+    {
+      "epoch": 10.84155062100113,
+      "grad_norm": 0.176604762673378,
+      "learning_rate": 5e-06,
+      "loss": 0.1539,
+      "step": 14403
+    },
+    {
+      "epoch": 10.842303349642453,
+      "grad_norm": 0.18435628712177277,
+      "learning_rate": 5e-06,
+      "loss": 0.1116,
+      "step": 14404
+    },
+    {
+      "epoch": 10.843056078283778,
+      "grad_norm": 0.1937635987997055,
+      "learning_rate": 5e-06,
+      "loss": 0.1539,
+      "step": 14405
+    },
+    {
+      "epoch": 10.843808806925104,
+      "grad_norm": 0.17779093980789185,
+      "learning_rate": 5e-06,
+      "loss": 0.1242,
+      "step": 14406
+    },
+    {
+      "epoch": 10.844561535566429,
+      "grad_norm": 0.19423089921474457,
+      "learning_rate": 5e-06,
+      "loss": 0.2132,
+      "step": 14407
+    },
+    {
+      "epoch": 10.845314264207753,
+      "grad_norm": 0.19310519099235535,
+      "learning_rate": 5e-06,
+      "loss": 0.1587,
+      "step": 14408
+    },
+    {
+      "epoch": 10.846066992849078,
+      "grad_norm": 0.19234599173069,
+      "learning_rate": 5e-06,
+      "loss": 0.1832,
+      "step": 14409
+    },
+    {
+      "epoch": 10.846819721490403,
+      "grad_norm": 0.16328619420528412,
+      "learning_rate": 5e-06,
+      "loss": 0.1368,
+      "step": 14410
+    },
+    {
+      "epoch": 10.847572450131727,
+      "grad_norm": 0.1626119464635849,
+      "learning_rate": 5e-06,
+      "loss": 0.125,
+      "step": 14411
+    },
+    {
+      "epoch": 10.848325178773052,
+      "grad_norm": 0.17169396579265594,
+      "learning_rate": 5e-06,
+      "loss": 0.1728,
+      "step": 14412
+    },
+    {
+      "epoch": 10.849077907414378,
+      "grad_norm": 0.19420276582241058,
+      "learning_rate": 5e-06,
+      "loss": 0.1542,
+      "step": 14413
+    },
+    {
+      "epoch": 10.849830636055701,
+      "grad_norm": 0.16269351541996002,
+      "learning_rate": 5e-06,
+      "loss": 0.081,
+      "step": 14414
+    },
+    {
+      "epoch": 10.850583364697027,
+      "grad_norm": 0.18392062187194824,
+      "learning_rate": 5e-06,
+      "loss": 0.1234,
+      "step": 14415
+    },
+    {
+      "epoch": 10.851336093338352,
+      "grad_norm": 0.2189609855413437,
+      "learning_rate": 5e-06,
+      "loss": 0.202,
+      "step": 14416
+    },
+    {
+      "epoch": 10.852088821979676,
+      "grad_norm": 0.19184283912181854,
+      "learning_rate": 5e-06,
+      "loss": 0.1657,
+      "step": 14417
+    },
+    {
+      "epoch": 10.852841550621001,
+      "grad_norm": 0.16782882809638977,
+      "learning_rate": 5e-06,
+      "loss": 0.0427,
+      "step": 14418
+    },
+    {
+      "epoch": 10.853594279262326,
+      "grad_norm": 0.19292935729026794,
+      "learning_rate": 5e-06,
+      "loss": 0.1852,
+      "step": 14419
+    },
+    {
+      "epoch": 10.85434700790365,
+      "grad_norm": 0.15753129124641418,
+      "learning_rate": 5e-06,
+      "loss": 0.1329,
+      "step": 14420
+    },
+    {
+      "epoch": 10.855099736544975,
+      "grad_norm": 0.15779812633991241,
+      "learning_rate": 5e-06,
+      "loss": 0.064,
+      "step": 14421
+    },
+    {
+      "epoch": 10.8558524651863,
+      "grad_norm": 0.17206668853759766,
+      "learning_rate": 5e-06,
+      "loss": 0.1055,
+      "step": 14422
+    },
+    {
+      "epoch": 10.856605193827626,
+      "grad_norm": 0.16541898250579834,
+      "learning_rate": 5e-06,
+      "loss": 0.1392,
+      "step": 14423
+    },
+    {
+      "epoch": 10.85735792246895,
+      "grad_norm": 0.1643025130033493,
+      "learning_rate": 5e-06,
+      "loss": 0.1477,
+      "step": 14424
+    },
+    {
+      "epoch": 10.858110651110275,
+      "grad_norm": 0.18590056896209717,
+      "learning_rate": 5e-06,
+      "loss": 0.1302,
+      "step": 14425
+    },
+    {
+      "epoch": 10.8588633797516,
+      "grad_norm": 0.17240487039089203,
+      "learning_rate": 5e-06,
+      "loss": 0.1033,
+      "step": 14426
+    },
+    {
+      "epoch": 10.859616108392924,
+      "grad_norm": 0.19312851130962372,
+      "learning_rate": 5e-06,
+      "loss": 0.2165,
+      "step": 14427
+    },
+    {
+      "epoch": 10.86036883703425,
+      "grad_norm": 0.16515874862670898,
+      "learning_rate": 5e-06,
+      "loss": 0.119,
+      "step": 14428
+    },
+    {
+      "epoch": 10.861121565675575,
+      "grad_norm": 0.20391803979873657,
+      "learning_rate": 5e-06,
+      "loss": 0.1274,
+      "step": 14429
+    },
+    {
+      "epoch": 10.861874294316898,
+      "grad_norm": 0.17936846613883972,
+      "learning_rate": 5e-06,
+      "loss": 0.192,
+      "step": 14430
+    },
+    {
+      "epoch": 10.862627022958224,
+      "grad_norm": 0.1776500642299652,
+      "learning_rate": 5e-06,
+      "loss": 0.1246,
+      "step": 14431
+    },
+    {
+      "epoch": 10.863379751599549,
+      "grad_norm": 0.15618112683296204,
+      "learning_rate": 5e-06,
+      "loss": 0.1208,
+      "step": 14432
+    },
+    {
+      "epoch": 10.864132480240873,
+      "grad_norm": 0.17758338153362274,
+      "learning_rate": 5e-06,
+      "loss": 0.1157,
+      "step": 14433
+    },
+    {
+      "epoch": 10.864885208882198,
+      "grad_norm": 0.22091355919837952,
+      "learning_rate": 5e-06,
+      "loss": 0.1678,
+      "step": 14434
+    },
+    {
+      "epoch": 10.865637937523523,
+      "grad_norm": 0.18907350301742554,
+      "learning_rate": 5e-06,
+      "loss": 0.1747,
+      "step": 14435
+    },
+    {
+      "epoch": 10.866390666164847,
+      "grad_norm": 0.1686103790998459,
+      "learning_rate": 5e-06,
+      "loss": 0.1256,
+      "step": 14436
+    },
+    {
+      "epoch": 10.867143394806172,
+      "grad_norm": 0.17415010929107666,
+      "learning_rate": 5e-06,
+      "loss": 0.1804,
+      "step": 14437
+    },
+    {
+      "epoch": 10.867896123447498,
+      "grad_norm": 0.19265878200531006,
+      "learning_rate": 5e-06,
+      "loss": 0.1461,
+      "step": 14438
+    },
+    {
+      "epoch": 10.868648852088821,
+      "grad_norm": 0.19795207679271698,
+      "learning_rate": 5e-06,
+      "loss": 0.1409,
+      "step": 14439
+    },
+    {
+      "epoch": 10.869401580730147,
+      "grad_norm": 0.17445167899131775,
+      "learning_rate": 5e-06,
+      "loss": 0.1667,
+      "step": 14440
+    },
+    {
+      "epoch": 10.870154309371472,
+      "grad_norm": 0.21594029664993286,
+      "learning_rate": 5e-06,
+      "loss": 0.1357,
+      "step": 14441
+    },
+    {
+      "epoch": 10.870907038012795,
+      "grad_norm": 0.19213345646858215,
+      "learning_rate": 5e-06,
+      "loss": 0.1466,
+      "step": 14442
+    },
+    {
+      "epoch": 10.87165976665412,
+      "grad_norm": 0.20386762917041779,
+      "learning_rate": 5e-06,
+      "loss": 0.1233,
+      "step": 14443
+    },
+    {
+      "epoch": 10.872412495295446,
+      "grad_norm": 0.188279926776886,
+      "learning_rate": 5e-06,
+      "loss": 0.1032,
+      "step": 14444
+    },
+    {
+      "epoch": 10.873165223936772,
+      "grad_norm": 0.2051478922367096,
+      "learning_rate": 5e-06,
+      "loss": 0.1314,
+      "step": 14445
+    },
+    {
+      "epoch": 10.873917952578095,
+      "grad_norm": 0.18950045108795166,
+      "learning_rate": 5e-06,
+      "loss": 0.0934,
+      "step": 14446
+    },
+    {
+      "epoch": 10.87467068121942,
+      "grad_norm": 0.17323528230190277,
+      "learning_rate": 5e-06,
+      "loss": 0.1366,
+      "step": 14447
+    },
+    {
+      "epoch": 10.875423409860746,
+      "grad_norm": 0.19274885952472687,
+      "learning_rate": 5e-06,
+      "loss": 0.1759,
+      "step": 14448
+    },
+    {
+      "epoch": 10.87617613850207,
+      "grad_norm": 0.16187983751296997,
+      "learning_rate": 5e-06,
+      "loss": 0.1103,
+      "step": 14449
+    },
+    {
+      "epoch": 10.876928867143395,
+      "grad_norm": 0.17115923762321472,
+      "learning_rate": 5e-06,
+      "loss": 0.0969,
+      "step": 14450
+    },
+    {
+      "epoch": 10.87768159578472,
+      "grad_norm": 0.18409976363182068,
+      "learning_rate": 5e-06,
+      "loss": 0.2429,
+      "step": 14451
+    },
+    {
+      "epoch": 10.878434324426044,
+      "grad_norm": 0.17831316590309143,
+      "learning_rate": 5e-06,
+      "loss": 0.0879,
+      "step": 14452
+    },
+    {
+      "epoch": 10.879187053067369,
+      "grad_norm": 0.15594884753227234,
+      "learning_rate": 5e-06,
+      "loss": 0.0957,
+      "step": 14453
+    },
+    {
+      "epoch": 10.879939781708694,
+      "grad_norm": 0.16918326914310455,
+      "learning_rate": 5e-06,
+      "loss": 0.1488,
+      "step": 14454
+    },
+    {
+      "epoch": 10.880692510350018,
+      "grad_norm": 0.16503487527370453,
+      "learning_rate": 5e-06,
+      "loss": 0.1113,
+      "step": 14455
+    },
+    {
+      "epoch": 10.881445238991343,
+      "grad_norm": 0.16256968677043915,
+      "learning_rate": 5e-06,
+      "loss": 0.1128,
+      "step": 14456
+    },
+    {
+      "epoch": 10.882197967632669,
+      "grad_norm": 0.19881927967071533,
+      "learning_rate": 5e-06,
+      "loss": 0.1179,
+      "step": 14457
+    },
+    {
+      "epoch": 10.882950696273994,
+      "grad_norm": 0.20111778378486633,
+      "learning_rate": 5e-06,
+      "loss": 0.2273,
+      "step": 14458
+    },
+    {
+      "epoch": 10.883703424915318,
+      "grad_norm": 0.20303568243980408,
+      "learning_rate": 5e-06,
+      "loss": 0.2044,
+      "step": 14459
+    },
+    {
+      "epoch": 10.884456153556643,
+      "grad_norm": 0.16953891515731812,
+      "learning_rate": 5e-06,
+      "loss": 0.1399,
+      "step": 14460
+    },
+    {
+      "epoch": 10.885208882197968,
+      "grad_norm": 0.19798409938812256,
+      "learning_rate": 5e-06,
+      "loss": 0.1789,
+      "step": 14461
+    },
+    {
+      "epoch": 10.885961610839292,
+      "grad_norm": 0.1918734908103943,
+      "learning_rate": 5e-06,
+      "loss": 0.1158,
+      "step": 14462
+    },
+    {
+      "epoch": 10.886714339480617,
+      "grad_norm": 0.19538530707359314,
+      "learning_rate": 5e-06,
+      "loss": 0.1192,
+      "step": 14463
+    },
+    {
+      "epoch": 10.887467068121943,
+      "grad_norm": 0.160261869430542,
+      "learning_rate": 5e-06,
+      "loss": 0.1222,
+      "step": 14464
+    },
+    {
+      "epoch": 10.888219796763266,
+      "grad_norm": 0.15909035503864288,
+      "learning_rate": 5e-06,
+      "loss": 0.0877,
+      "step": 14465
+    },
+    {
+      "epoch": 10.888972525404592,
+      "grad_norm": 0.16888928413391113,
+      "learning_rate": 5e-06,
+      "loss": 0.1485,
+      "step": 14466
+    },
+    {
+      "epoch": 10.889725254045917,
+      "grad_norm": 0.2658569812774658,
+      "learning_rate": 5e-06,
+      "loss": 0.2029,
+      "step": 14467
+    },
+    {
+      "epoch": 10.89047798268724,
+      "grad_norm": 0.21241529285907745,
+      "learning_rate": 5e-06,
+      "loss": 0.1284,
+      "step": 14468
+    },
+    {
+      "epoch": 10.891230711328566,
+      "grad_norm": 0.18445399403572083,
+      "learning_rate": 5e-06,
+      "loss": 0.136,
+      "step": 14469
+    },
+    {
+      "epoch": 10.891983439969891,
+      "grad_norm": 0.18414829671382904,
+      "learning_rate": 5e-06,
+      "loss": 0.1658,
+      "step": 14470
+    },
+    {
+      "epoch": 10.892736168611215,
+      "grad_norm": 0.15159176290035248,
+      "learning_rate": 5e-06,
+      "loss": 0.064,
+      "step": 14471
+    },
+    {
+      "epoch": 10.89348889725254,
+      "grad_norm": 0.16723041236400604,
+      "learning_rate": 5e-06,
+      "loss": 0.1249,
+      "step": 14472
+    },
+    {
+      "epoch": 10.894241625893866,
+      "grad_norm": 0.1801624298095703,
+      "learning_rate": 5e-06,
+      "loss": 0.1383,
+      "step": 14473
+    },
+    {
+      "epoch": 10.89499435453519,
+      "grad_norm": 0.17099615931510925,
+      "learning_rate": 5e-06,
+      "loss": 0.0585,
+      "step": 14474
+    },
+    {
+      "epoch": 10.895747083176515,
+      "grad_norm": 0.18378224968910217,
+      "learning_rate": 5e-06,
+      "loss": 0.1323,
+      "step": 14475
+    },
+    {
+      "epoch": 10.89649981181784,
+      "grad_norm": 0.18388044834136963,
+      "learning_rate": 5e-06,
+      "loss": 0.1487,
+      "step": 14476
+    },
+    {
+      "epoch": 10.897252540459164,
+      "grad_norm": 0.17475338280200958,
+      "learning_rate": 5e-06,
+      "loss": 0.1369,
+      "step": 14477
+    },
+    {
+      "epoch": 10.898005269100489,
+      "grad_norm": 0.18437829613685608,
+      "learning_rate": 5e-06,
+      "loss": 0.161,
+      "step": 14478
+    },
+    {
+      "epoch": 10.898757997741814,
+      "grad_norm": 0.179927259683609,
+      "learning_rate": 5e-06,
+      "loss": 0.1415,
+      "step": 14479
+    },
+    {
+      "epoch": 10.89951072638314,
+      "grad_norm": 0.19400222599506378,
+      "learning_rate": 5e-06,
+      "loss": 0.1244,
+      "step": 14480
+    },
+    {
+      "epoch": 10.900263455024463,
+      "grad_norm": 0.18417693674564362,
+      "learning_rate": 5e-06,
+      "loss": 0.117,
+      "step": 14481
+    },
+    {
+      "epoch": 10.901016183665789,
+      "grad_norm": 0.17708376049995422,
+      "learning_rate": 5e-06,
+      "loss": 0.1705,
+      "step": 14482
+    },
+    {
+      "epoch": 10.901768912307114,
+      "grad_norm": 0.1867242306470871,
+      "learning_rate": 5e-06,
+      "loss": 0.1475,
+      "step": 14483
+    },
+    {
+      "epoch": 10.902521640948438,
+      "grad_norm": 0.17413339018821716,
+      "learning_rate": 5e-06,
+      "loss": 0.1253,
+      "step": 14484
+    },
+    {
+      "epoch": 10.903274369589763,
+      "grad_norm": 0.17600390315055847,
+      "learning_rate": 5e-06,
+      "loss": 0.1452,
+      "step": 14485
+    },
+    {
+      "epoch": 10.904027098231088,
+      "grad_norm": 0.18776944279670715,
+      "learning_rate": 5e-06,
+      "loss": 0.2288,
+      "step": 14486
+    },
+    {
+      "epoch": 10.904779826872412,
+      "grad_norm": 0.19129787385463715,
+      "learning_rate": 5e-06,
+      "loss": 0.0945,
+      "step": 14487
+    },
+    {
+      "epoch": 10.905532555513737,
+      "grad_norm": 0.18680816888809204,
+      "learning_rate": 5e-06,
+      "loss": 0.1499,
+      "step": 14488
+    },
+    {
+      "epoch": 10.906285284155063,
+      "grad_norm": 0.18206357955932617,
+      "learning_rate": 5e-06,
+      "loss": 0.1135,
+      "step": 14489
+    },
+    {
+      "epoch": 10.907038012796386,
+      "grad_norm": 0.1868438422679901,
+      "learning_rate": 5e-06,
+      "loss": 0.1165,
+      "step": 14490
+    },
+    {
+      "epoch": 10.907790741437712,
+      "grad_norm": 0.15509755909442902,
+      "learning_rate": 5e-06,
+      "loss": 0.1567,
+      "step": 14491
+    },
+    {
+      "epoch": 10.908543470079037,
+      "grad_norm": 0.17212609946727753,
+      "learning_rate": 5e-06,
+      "loss": 0.2008,
+      "step": 14492
+    },
+    {
+      "epoch": 10.90929619872036,
+      "grad_norm": 0.2029128223657608,
+      "learning_rate": 5e-06,
+      "loss": 0.1376,
+      "step": 14493
+    },
+    {
+      "epoch": 10.910048927361686,
+      "grad_norm": 0.1640879213809967,
+      "learning_rate": 5e-06,
+      "loss": 0.0319,
+      "step": 14494
+    },
+    {
+      "epoch": 10.910801656003011,
+      "grad_norm": 0.18052977323532104,
+      "learning_rate": 5e-06,
+      "loss": 0.1327,
+      "step": 14495
+    },
+    {
+      "epoch": 10.911554384644337,
+      "grad_norm": 0.18353308737277985,
+      "learning_rate": 5e-06,
+      "loss": 0.1162,
+      "step": 14496
+    },
+    {
+      "epoch": 10.91230711328566,
+      "grad_norm": 0.16812293231487274,
+      "learning_rate": 5e-06,
+      "loss": 0.1862,
+      "step": 14497
+    },
+    {
+      "epoch": 10.913059841926986,
+      "grad_norm": 0.15592479705810547,
+      "learning_rate": 5e-06,
+      "loss": 0.1301,
+      "step": 14498
+    },
+    {
+      "epoch": 10.913812570568311,
+      "grad_norm": 0.19364427030086517,
+      "learning_rate": 5e-06,
+      "loss": 0.1267,
+      "step": 14499
+    },
+    {
+      "epoch": 10.914565299209634,
+      "grad_norm": 0.18513745069503784,
+      "learning_rate": 5e-06,
+      "loss": 0.1114,
+      "step": 14500
+    },
+    {
+      "epoch": 10.91531802785096,
+      "grad_norm": 0.17392191290855408,
+      "learning_rate": 5e-06,
+      "loss": 0.1066,
+      "step": 14501
+    },
+    {
+      "epoch": 10.916070756492285,
+      "grad_norm": 0.19423949718475342,
+      "learning_rate": 5e-06,
+      "loss": 0.1105,
+      "step": 14502
+    },
+    {
+      "epoch": 10.916823485133609,
+      "grad_norm": 0.1787460595369339,
+      "learning_rate": 5e-06,
+      "loss": 0.1256,
+      "step": 14503
+    },
+    {
+      "epoch": 10.917576213774934,
+      "grad_norm": 0.18669021129608154,
+      "learning_rate": 5e-06,
+      "loss": 0.1239,
+      "step": 14504
+    },
+    {
+      "epoch": 10.91832894241626,
+      "grad_norm": 0.184555321931839,
+      "learning_rate": 5e-06,
+      "loss": 0.1391,
+      "step": 14505
+    },
+    {
+      "epoch": 10.919081671057583,
+      "grad_norm": 0.17834840714931488,
+      "learning_rate": 5e-06,
+      "loss": 0.1055,
+      "step": 14506
+    },
+    {
+      "epoch": 10.919834399698908,
+      "grad_norm": 0.196051225066185,
+      "learning_rate": 5e-06,
+      "loss": 0.1571,
+      "step": 14507
+    },
+    {
+      "epoch": 10.920587128340234,
+      "grad_norm": 0.17661558091640472,
+      "learning_rate": 5e-06,
+      "loss": 0.0814,
+      "step": 14508
+    },
+    {
+      "epoch": 10.921339856981557,
+      "grad_norm": 0.1737830936908722,
+      "learning_rate": 5e-06,
+      "loss": 0.1294,
+      "step": 14509
+    },
+    {
+      "epoch": 10.922092585622883,
+      "grad_norm": 0.19062799215316772,
+      "learning_rate": 5e-06,
+      "loss": 0.203,
+      "step": 14510
+    },
+    {
+      "epoch": 10.922845314264208,
+      "grad_norm": 0.1897672712802887,
+      "learning_rate": 5e-06,
+      "loss": 0.1896,
+      "step": 14511
+    },
+    {
+      "epoch": 10.923598042905532,
+      "grad_norm": 0.17189782857894897,
+      "learning_rate": 5e-06,
+      "loss": 0.1034,
+      "step": 14512
+    },
+    {
+      "epoch": 10.924350771546857,
+      "grad_norm": 0.18902002274990082,
+      "learning_rate": 5e-06,
+      "loss": 0.1742,
+      "step": 14513
+    },
+    {
+      "epoch": 10.925103500188182,
+      "grad_norm": 0.17301428318023682,
+      "learning_rate": 5e-06,
+      "loss": 0.1441,
+      "step": 14514
+    },
+    {
+      "epoch": 10.925856228829506,
+      "grad_norm": 0.18107399344444275,
+      "learning_rate": 5e-06,
+      "loss": 0.1117,
+      "step": 14515
+    },
+    {
+      "epoch": 10.926608957470831,
+      "grad_norm": 0.1732046753168106,
+      "learning_rate": 5e-06,
+      "loss": 0.2224,
+      "step": 14516
+    },
+    {
+      "epoch": 10.927361686112157,
+      "grad_norm": 0.20941248536109924,
+      "learning_rate": 5e-06,
+      "loss": 0.1637,
+      "step": 14517
+    },
+    {
+      "epoch": 10.928114414753482,
+      "grad_norm": 0.19306537508964539,
+      "learning_rate": 5e-06,
+      "loss": 0.1841,
+      "step": 14518
+    },
+    {
+      "epoch": 10.928867143394806,
+      "grad_norm": 0.18534554541110992,
+      "learning_rate": 5e-06,
+      "loss": 0.1775,
+      "step": 14519
+    },
+    {
+      "epoch": 10.929619872036131,
+      "grad_norm": 0.18514972925186157,
+      "learning_rate": 5e-06,
+      "loss": 0.1607,
+      "step": 14520
+    },
+    {
+      "epoch": 10.930372600677456,
+      "grad_norm": 0.1630241870880127,
+      "learning_rate": 5e-06,
+      "loss": 0.1149,
+      "step": 14521
+    },
+    {
+      "epoch": 10.93112532931878,
+      "grad_norm": 0.1748550534248352,
+      "learning_rate": 5e-06,
+      "loss": 0.1,
+      "step": 14522
+    },
+    {
+      "epoch": 10.931878057960105,
+      "grad_norm": 0.17178484797477722,
+      "learning_rate": 5e-06,
+      "loss": 0.153,
+      "step": 14523
+    },
+    {
+      "epoch": 10.93263078660143,
+      "grad_norm": 0.20789429545402527,
+      "learning_rate": 5e-06,
+      "loss": 0.1516,
+      "step": 14524
+    },
+    {
+      "epoch": 10.933383515242754,
+      "grad_norm": 0.20264428853988647,
+      "learning_rate": 5e-06,
+      "loss": 0.1712,
+      "step": 14525
+    },
+    {
+      "epoch": 10.93413624388408,
+      "grad_norm": 0.1997961550951004,
+      "learning_rate": 5e-06,
+      "loss": 0.2412,
+      "step": 14526
+    },
+    {
+      "epoch": 10.934888972525405,
+      "grad_norm": 0.18399746716022491,
+      "learning_rate": 5e-06,
+      "loss": 0.11,
+      "step": 14527
+    },
+    {
+      "epoch": 10.935641701166729,
+      "grad_norm": 0.1858602613210678,
+      "learning_rate": 5e-06,
+      "loss": 0.1597,
+      "step": 14528
+    },
+    {
+      "epoch": 10.936394429808054,
+      "grad_norm": 0.19239255785942078,
+      "learning_rate": 5e-06,
+      "loss": 0.1845,
+      "step": 14529
+    },
+    {
+      "epoch": 10.93714715844938,
+      "grad_norm": 0.18755514919757843,
+      "learning_rate": 5e-06,
+      "loss": 0.1309,
+      "step": 14530
+    },
+    {
+      "epoch": 10.937899887090705,
+      "grad_norm": 0.1690790057182312,
+      "learning_rate": 5e-06,
+      "loss": 0.2017,
+      "step": 14531
+    },
+    {
+      "epoch": 10.938652615732028,
+      "grad_norm": 0.1901036947965622,
+      "learning_rate": 5e-06,
+      "loss": 0.1411,
+      "step": 14532
+    },
+    {
+      "epoch": 10.939405344373354,
+      "grad_norm": 0.16954828798770905,
+      "learning_rate": 5e-06,
+      "loss": 0.1793,
+      "step": 14533
+    },
+    {
+      "epoch": 10.940158073014679,
+      "grad_norm": 0.17251832783222198,
+      "learning_rate": 5e-06,
+      "loss": 0.1186,
+      "step": 14534
+    },
+    {
+      "epoch": 10.940910801656003,
+      "grad_norm": 0.18618883192539215,
+      "learning_rate": 5e-06,
+      "loss": 0.2137,
+      "step": 14535
+    },
+    {
+      "epoch": 10.941663530297328,
+      "grad_norm": 0.19863897562026978,
+      "learning_rate": 5e-06,
+      "loss": 0.185,
+      "step": 14536
+    },
+    {
+      "epoch": 10.942416258938653,
+      "grad_norm": 0.14545801281929016,
+      "learning_rate": 5e-06,
+      "loss": 0.0576,
+      "step": 14537
+    },
+    {
+      "epoch": 10.943168987579977,
+      "grad_norm": 0.17706479132175446,
+      "learning_rate": 5e-06,
+      "loss": 0.0878,
+      "step": 14538
+    },
+    {
+      "epoch": 10.943921716221302,
+      "grad_norm": 0.16406860947608948,
+      "learning_rate": 5e-06,
+      "loss": 0.1127,
+      "step": 14539
+    },
+    {
+      "epoch": 10.944674444862628,
+      "grad_norm": 0.18813830614089966,
+      "learning_rate": 5e-06,
+      "loss": 0.1093,
+      "step": 14540
+    },
+    {
+      "epoch": 10.945427173503951,
+      "grad_norm": 0.16243976354599,
+      "learning_rate": 5e-06,
+      "loss": 0.1395,
+      "step": 14541
+    },
+    {
+      "epoch": 10.946179902145277,
+      "grad_norm": 0.1937618851661682,
+      "learning_rate": 5e-06,
+      "loss": 0.1112,
+      "step": 14542
+    },
+    {
+      "epoch": 10.946932630786602,
+      "grad_norm": 0.18322493135929108,
+      "learning_rate": 5e-06,
+      "loss": 0.1949,
+      "step": 14543
+    },
+    {
+      "epoch": 10.947685359427926,
+      "grad_norm": 0.19510184228420258,
+      "learning_rate": 5e-06,
+      "loss": 0.1376,
+      "step": 14544
+    },
+    {
+      "epoch": 10.948438088069251,
+      "grad_norm": 0.19516046345233917,
+      "learning_rate": 5e-06,
+      "loss": 0.1454,
+      "step": 14545
+    },
+    {
+      "epoch": 10.949190816710576,
+      "grad_norm": 0.17431484162807465,
+      "learning_rate": 5e-06,
+      "loss": 0.1591,
+      "step": 14546
+    },
+    {
+      "epoch": 10.9499435453519,
+      "grad_norm": 0.16362273693084717,
+      "learning_rate": 5e-06,
+      "loss": 0.1366,
+      "step": 14547
+    },
+    {
+      "epoch": 10.950696273993225,
+      "grad_norm": 0.17138685286045074,
+      "learning_rate": 5e-06,
+      "loss": 0.1387,
+      "step": 14548
+    },
+    {
+      "epoch": 10.95144900263455,
+      "grad_norm": 0.17667296528816223,
+      "learning_rate": 5e-06,
+      "loss": 0.1704,
+      "step": 14549
+    },
+    {
+      "epoch": 10.952201731275874,
+      "grad_norm": 0.18257054686546326,
+      "learning_rate": 5e-06,
+      "loss": 0.1043,
+      "step": 14550
+    },
+    {
+      "epoch": 10.9529544599172,
+      "grad_norm": 0.1853748857975006,
+      "learning_rate": 5e-06,
+      "loss": 0.1482,
+      "step": 14551
+    },
+    {
+      "epoch": 10.953707188558525,
+      "grad_norm": 0.1490965485572815,
+      "learning_rate": 5e-06,
+      "loss": 0.0877,
+      "step": 14552
+    },
+    {
+      "epoch": 10.95445991719985,
+      "grad_norm": 0.18952269852161407,
+      "learning_rate": 5e-06,
+      "loss": 0.1341,
+      "step": 14553
+    },
+    {
+      "epoch": 10.955212645841174,
+      "grad_norm": 0.19085443019866943,
+      "learning_rate": 5e-06,
+      "loss": 0.1728,
+      "step": 14554
+    },
+    {
+      "epoch": 10.9559653744825,
+      "grad_norm": 0.20801852643489838,
+      "learning_rate": 5e-06,
+      "loss": 0.1905,
+      "step": 14555
+    },
+    {
+      "epoch": 10.956718103123825,
+      "grad_norm": 0.18569931387901306,
+      "learning_rate": 5e-06,
+      "loss": 0.1403,
+      "step": 14556
+    },
+    {
+      "epoch": 10.957470831765148,
+      "grad_norm": 0.17013420164585114,
+      "learning_rate": 5e-06,
+      "loss": 0.1103,
+      "step": 14557
+    },
+    {
+      "epoch": 10.958223560406473,
+      "grad_norm": 0.2063283920288086,
+      "learning_rate": 5e-06,
+      "loss": 0.1524,
+      "step": 14558
+    },
+    {
+      "epoch": 10.958976289047799,
+      "grad_norm": 0.18229448795318604,
+      "learning_rate": 5e-06,
+      "loss": 0.1582,
+      "step": 14559
+    },
+    {
+      "epoch": 10.959729017689122,
+      "grad_norm": 0.18356893956661224,
+      "learning_rate": 5e-06,
+      "loss": 0.1051,
+      "step": 14560
+    },
+    {
+      "epoch": 10.960481746330448,
+      "grad_norm": 0.17167475819587708,
+      "learning_rate": 5e-06,
+      "loss": 0.1183,
+      "step": 14561
+    },
+    {
+      "epoch": 10.961234474971773,
+      "grad_norm": 0.2133016139268875,
+      "learning_rate": 5e-06,
+      "loss": 0.191,
+      "step": 14562
+    },
+    {
+      "epoch": 10.961987203613097,
+      "grad_norm": 0.1866026520729065,
+      "learning_rate": 5e-06,
+      "loss": 0.2043,
+      "step": 14563
+    },
+    {
+      "epoch": 10.962739932254422,
+      "grad_norm": 0.1707158386707306,
+      "learning_rate": 5e-06,
+      "loss": 0.1494,
+      "step": 14564
+    },
+    {
+      "epoch": 10.963492660895747,
+      "grad_norm": 0.21988674998283386,
+      "learning_rate": 5e-06,
+      "loss": 0.1729,
+      "step": 14565
+    },
+    {
+      "epoch": 10.964245389537071,
+      "grad_norm": 0.17288357019424438,
+      "learning_rate": 5e-06,
+      "loss": 0.0925,
+      "step": 14566
+    },
+    {
+      "epoch": 10.964998118178396,
+      "grad_norm": 0.17297038435935974,
+      "learning_rate": 5e-06,
+      "loss": 0.1303,
+      "step": 14567
+    },
+    {
+      "epoch": 10.965750846819722,
+      "grad_norm": 0.18855631351470947,
+      "learning_rate": 5e-06,
+      "loss": 0.157,
+      "step": 14568
+    },
+    {
+      "epoch": 10.966503575461047,
+      "grad_norm": 0.15907378494739532,
+      "learning_rate": 5e-06,
+      "loss": 0.0906,
+      "step": 14569
+    },
+    {
+      "epoch": 10.96725630410237,
+      "grad_norm": 0.2034999132156372,
+      "learning_rate": 5e-06,
+      "loss": 0.1349,
+      "step": 14570
+    },
+    {
+      "epoch": 10.968009032743696,
+      "grad_norm": 0.2241806983947754,
+      "learning_rate": 5e-06,
+      "loss": 0.1418,
+      "step": 14571
+    },
+    {
+      "epoch": 10.968761761385021,
+      "grad_norm": 0.21242570877075195,
+      "learning_rate": 5e-06,
+      "loss": 0.1769,
+      "step": 14572
+    },
+    {
+      "epoch": 10.969514490026345,
+      "grad_norm": 0.19177573919296265,
+      "learning_rate": 5e-06,
+      "loss": 0.1853,
+      "step": 14573
+    },
+    {
+      "epoch": 10.97026721866767,
+      "grad_norm": 0.19091948866844177,
+      "learning_rate": 5e-06,
+      "loss": 0.1742,
+      "step": 14574
+    },
+    {
+      "epoch": 10.971019947308996,
+      "grad_norm": 0.18640902638435364,
+      "learning_rate": 5e-06,
+      "loss": 0.1138,
+      "step": 14575
+    },
+    {
+      "epoch": 10.97177267595032,
+      "grad_norm": 0.1774032562971115,
+      "learning_rate": 5e-06,
+      "loss": 0.1312,
+      "step": 14576
+    },
+    {
+      "epoch": 10.972525404591645,
+      "grad_norm": 0.17123891413211823,
+      "learning_rate": 5e-06,
+      "loss": 0.1116,
+      "step": 14577
+    },
+    {
+      "epoch": 10.97327813323297,
+      "grad_norm": 0.17326074838638306,
+      "learning_rate": 5e-06,
+      "loss": 0.1015,
+      "step": 14578
+    },
+    {
+      "epoch": 10.974030861874294,
+      "grad_norm": 0.17148756980895996,
+      "learning_rate": 5e-06,
+      "loss": 0.1536,
+      "step": 14579
+    },
+    {
+      "epoch": 10.974783590515619,
+      "grad_norm": 0.174369215965271,
+      "learning_rate": 5e-06,
+      "loss": 0.1185,
+      "step": 14580
+    },
+    {
+      "epoch": 10.975536319156944,
+      "grad_norm": 0.17184221744537354,
+      "learning_rate": 5e-06,
+      "loss": 0.1738,
+      "step": 14581
+    },
+    {
+      "epoch": 10.976289047798268,
+      "grad_norm": 0.17655467987060547,
+      "learning_rate": 5e-06,
+      "loss": 0.1386,
+      "step": 14582
+    },
+    {
+      "epoch": 10.977041776439593,
+      "grad_norm": 0.17361880838871002,
+      "learning_rate": 5e-06,
+      "loss": 0.141,
+      "step": 14583
+    },
+    {
+      "epoch": 10.977794505080919,
+      "grad_norm": 0.18356451392173767,
+      "learning_rate": 5e-06,
+      "loss": 0.1452,
+      "step": 14584
+    },
+    {
+      "epoch": 10.978547233722242,
+      "grad_norm": 0.1861501932144165,
+      "learning_rate": 5e-06,
+      "loss": 0.1533,
+      "step": 14585
+    },
+    {
+      "epoch": 10.979299962363568,
+      "grad_norm": 0.1861690729856491,
+      "learning_rate": 5e-06,
+      "loss": 0.1723,
+      "step": 14586
+    },
+    {
+      "epoch": 10.980052691004893,
+      "grad_norm": 0.18990853428840637,
+      "learning_rate": 5e-06,
+      "loss": 0.1963,
+      "step": 14587
+    },
+    {
+      "epoch": 10.980805419646218,
+      "grad_norm": 0.1540593057870865,
+      "learning_rate": 5e-06,
+      "loss": 0.0844,
+      "step": 14588
+    },
+    {
+      "epoch": 10.981558148287542,
+      "grad_norm": 0.1899046152830124,
+      "learning_rate": 5e-06,
+      "loss": 0.185,
+      "step": 14589
+    },
+    {
+      "epoch": 10.982310876928867,
+      "grad_norm": 0.184747114777565,
+      "learning_rate": 5e-06,
+      "loss": 0.2018,
+      "step": 14590
+    },
+    {
+      "epoch": 10.983063605570193,
+      "grad_norm": 0.19149674475193024,
+      "learning_rate": 5e-06,
+      "loss": 0.155,
+      "step": 14591
+    },
+    {
+      "epoch": 10.983816334211516,
+      "grad_norm": 0.17518118023872375,
+      "learning_rate": 5e-06,
+      "loss": 0.0737,
+      "step": 14592
+    },
+    {
+      "epoch": 10.984569062852842,
+      "grad_norm": 0.19127151370048523,
+      "learning_rate": 5e-06,
+      "loss": 0.2026,
+      "step": 14593
+    },
+    {
+      "epoch": 10.985321791494167,
+      "grad_norm": 0.19062469899654388,
+      "learning_rate": 5e-06,
+      "loss": 0.1797,
+      "step": 14594
+    },
+    {
+      "epoch": 10.98607452013549,
+      "grad_norm": 0.17960210144519806,
+      "learning_rate": 5e-06,
+      "loss": 0.1732,
+      "step": 14595
+    },
+    {
+      "epoch": 10.986827248776816,
+      "grad_norm": 0.18233072757720947,
+      "learning_rate": 5e-06,
+      "loss": 0.1462,
+      "step": 14596
+    },
+    {
+      "epoch": 10.987579977418141,
+      "grad_norm": 0.1867019087076187,
+      "learning_rate": 5e-06,
+      "loss": 0.2426,
+      "step": 14597
+    },
+    {
+      "epoch": 10.988332706059465,
+      "grad_norm": 0.15228162705898285,
+      "learning_rate": 5e-06,
+      "loss": 0.1335,
+      "step": 14598
+    },
+    {
+      "epoch": 10.98908543470079,
+      "grad_norm": 0.17986992001533508,
+      "learning_rate": 5e-06,
+      "loss": 0.1935,
+      "step": 14599
+    },
+    {
+      "epoch": 10.989838163342116,
+      "grad_norm": 0.17509298026561737,
+      "learning_rate": 5e-06,
+      "loss": 0.1454,
+      "step": 14600
+    },
+    {
+      "epoch": 10.989838163342116,
+      "eval_loss": 0.18785539269447327,
+      "eval_runtime": 452.8283,
+      "eval_samples_per_second": 21.26,
+      "eval_steps_per_second": 0.665,
+      "step": 14600
+    },
+    {
+      "epoch": 10.99059089198344,
+      "grad_norm": 0.15446244180202484,
+      "learning_rate": 5e-06,
+      "loss": 0.0916,
+      "step": 14601
+    },
+    {
+      "epoch": 10.991343620624765,
+      "grad_norm": 0.20100699365139008,
+      "learning_rate": 5e-06,
+      "loss": 0.1826,
+      "step": 14602
+    },
+    {
+      "epoch": 10.99209634926609,
+      "grad_norm": 0.1754789650440216,
+      "learning_rate": 5e-06,
+      "loss": 0.1529,
+      "step": 14603
+    },
+    {
+      "epoch": 10.992849077907415,
+      "grad_norm": 0.1833295375108719,
+      "learning_rate": 5e-06,
+      "loss": 0.1902,
+      "step": 14604
+    },
+    {
+      "epoch": 10.993601806548739,
+      "grad_norm": 0.16832277178764343,
+      "learning_rate": 5e-06,
+      "loss": 0.0543,
+      "step": 14605
+    },
+    {
+      "epoch": 10.994354535190064,
+      "grad_norm": 0.19486556947231293,
+      "learning_rate": 5e-06,
+      "loss": 0.1111,
+      "step": 14606
+    },
+    {
+      "epoch": 10.99510726383139,
+      "grad_norm": 0.22401298582553864,
+      "learning_rate": 5e-06,
+      "loss": 0.16,
+      "step": 14607
+    },
+    {
+      "epoch": 10.995859992472713,
+      "grad_norm": 0.1837022602558136,
+      "learning_rate": 5e-06,
+      "loss": 0.0865,
+      "step": 14608
+    },
+    {
+      "epoch": 10.996612721114039,
+      "grad_norm": 0.17388911545276642,
+      "learning_rate": 5e-06,
+      "loss": 0.1681,
+      "step": 14609
+    },
+    {
+      "epoch": 10.997365449755364,
+      "grad_norm": 0.1982296109199524,
+      "learning_rate": 5e-06,
+      "loss": 0.1608,
+      "step": 14610
+    },
+    {
+      "epoch": 10.998118178396687,
+      "grad_norm": 0.2271118313074112,
+      "learning_rate": 5e-06,
+      "loss": 0.2033,
+      "step": 14611
+    },
+    {
+      "epoch": 10.998870907038013,
+      "grad_norm": 0.19808587431907654,
+      "learning_rate": 5e-06,
+      "loss": 0.188,
+      "step": 14612
+    },
+    {
+      "epoch": 10.999623635679338,
+      "grad_norm": 0.1949656903743744,
+      "learning_rate": 5e-06,
+      "loss": 0.1101,
+      "step": 14613
+    },
+    {
+      "epoch": 11.000376364320662,
+      "grad_norm": 0.17800050973892212,
+      "learning_rate": 5e-06,
+      "loss": 0.1592,
+      "step": 14614
+    },
+    {
+      "epoch": 11.001129092961987,
+      "grad_norm": 0.17555353045463562,
+      "learning_rate": 5e-06,
+      "loss": 0.1229,
+      "step": 14615
+    },
+    {
+      "epoch": 11.001881821603313,
+      "grad_norm": 0.2209366261959076,
+      "learning_rate": 5e-06,
+      "loss": 0.1291,
+      "step": 14616
+    },
+    {
+      "epoch": 11.002634550244636,
+      "grad_norm": 0.17533238232135773,
+      "learning_rate": 5e-06,
+      "loss": 0.1341,
+      "step": 14617
+    },
+    {
+      "epoch": 11.003387278885961,
+      "grad_norm": 0.19003073871135712,
+      "learning_rate": 5e-06,
+      "loss": 0.1656,
+      "step": 14618
+    },
+    {
+      "epoch": 11.004140007527287,
+      "grad_norm": 0.16040267050266266,
+      "learning_rate": 5e-06,
+      "loss": 0.226,
+      "step": 14619
+    },
+    {
+      "epoch": 11.00489273616861,
+      "grad_norm": 0.1951071172952652,
+      "learning_rate": 5e-06,
+      "loss": 0.1171,
+      "step": 14620
+    },
+    {
+      "epoch": 11.005645464809936,
+      "grad_norm": 0.1801379919052124,
+      "learning_rate": 5e-06,
+      "loss": 0.1604,
+      "step": 14621
+    },
+    {
+      "epoch": 11.006398193451261,
+      "grad_norm": 0.18252646923065186,
+      "learning_rate": 5e-06,
+      "loss": 0.1016,
+      "step": 14622
+    },
+    {
+      "epoch": 11.007150922092586,
+      "grad_norm": 0.16301710903644562,
+      "learning_rate": 5e-06,
+      "loss": 0.1386,
+      "step": 14623
+    },
+    {
+      "epoch": 11.00790365073391,
+      "grad_norm": 0.16673840582370758,
+      "learning_rate": 5e-06,
+      "loss": 0.1463,
+      "step": 14624
+    },
+    {
+      "epoch": 11.008656379375235,
+      "grad_norm": 0.17697566747665405,
+      "learning_rate": 5e-06,
+      "loss": 0.16,
+      "step": 14625
+    },
+    {
+      "epoch": 11.00940910801656,
+      "grad_norm": 0.16294562816619873,
+      "learning_rate": 5e-06,
+      "loss": 0.0785,
+      "step": 14626
+    },
+    {
+      "epoch": 11.010161836657884,
+      "grad_norm": 0.1581495851278305,
+      "learning_rate": 5e-06,
+      "loss": 0.114,
+      "step": 14627
+    },
+    {
+      "epoch": 11.01091456529921,
+      "grad_norm": 0.1780480146408081,
+      "learning_rate": 5e-06,
+      "loss": 0.1898,
+      "step": 14628
+    },
+    {
+      "epoch": 11.011667293940535,
+      "grad_norm": 0.21075528860092163,
+      "learning_rate": 5e-06,
+      "loss": 0.1161,
+      "step": 14629
+    },
+    {
+      "epoch": 11.012420022581859,
+      "grad_norm": 0.177546426653862,
+      "learning_rate": 5e-06,
+      "loss": 0.1289,
+      "step": 14630
+    },
+    {
+      "epoch": 11.013172751223184,
+      "grad_norm": 0.16668257117271423,
+      "learning_rate": 5e-06,
+      "loss": 0.1513,
+      "step": 14631
+    },
+    {
+      "epoch": 11.01392547986451,
+      "grad_norm": 0.17436404526233673,
+      "learning_rate": 5e-06,
+      "loss": 0.1,
+      "step": 14632
+    },
+    {
+      "epoch": 11.014678208505833,
+      "grad_norm": 0.1704719364643097,
+      "learning_rate": 5e-06,
+      "loss": 0.104,
+      "step": 14633
+    },
+    {
+      "epoch": 11.015430937147158,
+      "grad_norm": 0.19254834949970245,
+      "learning_rate": 5e-06,
+      "loss": 0.1802,
+      "step": 14634
+    },
+    {
+      "epoch": 11.016183665788484,
+      "grad_norm": 0.17145316302776337,
+      "learning_rate": 5e-06,
+      "loss": 0.1148,
+      "step": 14635
+    },
+    {
+      "epoch": 11.016936394429807,
+      "grad_norm": 0.19799017906188965,
+      "learning_rate": 5e-06,
+      "loss": 0.127,
+      "step": 14636
+    },
+    {
+      "epoch": 11.017689123071133,
+      "grad_norm": 0.16377666592597961,
+      "learning_rate": 5e-06,
+      "loss": 0.1121,
+      "step": 14637
+    },
+    {
+      "epoch": 11.018441851712458,
+      "grad_norm": 0.20926587283611298,
+      "learning_rate": 5e-06,
+      "loss": 0.1811,
+      "step": 14638
+    },
+    {
+      "epoch": 11.019194580353782,
+      "grad_norm": 0.17118307948112488,
+      "learning_rate": 5e-06,
+      "loss": 0.1156,
+      "step": 14639
+    },
+    {
+      "epoch": 11.019947308995107,
+      "grad_norm": 0.18117175996303558,
+      "learning_rate": 5e-06,
+      "loss": 0.1927,
+      "step": 14640
+    },
+    {
+      "epoch": 11.020700037636432,
+      "grad_norm": 0.19120080769062042,
+      "learning_rate": 5e-06,
+      "loss": 0.1616,
+      "step": 14641
+    },
+    {
+      "epoch": 11.021452766277758,
+      "grad_norm": 0.16580727696418762,
+      "learning_rate": 5e-06,
+      "loss": 0.1395,
+      "step": 14642
+    },
+    {
+      "epoch": 11.022205494919081,
+      "grad_norm": 0.16193577647209167,
+      "learning_rate": 5e-06,
+      "loss": 0.1231,
+      "step": 14643
+    },
+    {
+      "epoch": 11.022958223560407,
+      "grad_norm": 0.17863531410694122,
+      "learning_rate": 5e-06,
+      "loss": 0.1763,
+      "step": 14644
+    },
+    {
+      "epoch": 11.023710952201732,
+      "grad_norm": 0.15640822052955627,
+      "learning_rate": 5e-06,
+      "loss": 0.1229,
+      "step": 14645
+    },
+    {
+      "epoch": 11.024463680843056,
+      "grad_norm": 0.1923549473285675,
+      "learning_rate": 5e-06,
+      "loss": 0.1957,
+      "step": 14646
+    },
+    {
+      "epoch": 11.025216409484381,
+      "grad_norm": 0.19152502715587616,
+      "learning_rate": 5e-06,
+      "loss": 0.1319,
+      "step": 14647
+    },
+    {
+      "epoch": 11.025969138125706,
+      "grad_norm": 0.20825353264808655,
+      "learning_rate": 5e-06,
+      "loss": 0.1787,
+      "step": 14648
+    },
+    {
+      "epoch": 11.02672186676703,
+      "grad_norm": 0.2057691216468811,
+      "learning_rate": 5e-06,
+      "loss": 0.2175,
+      "step": 14649
+    },
+    {
+      "epoch": 11.027474595408355,
+      "grad_norm": 0.15750661492347717,
+      "learning_rate": 5e-06,
+      "loss": 0.0737,
+      "step": 14650
+    },
+    {
+      "epoch": 11.02822732404968,
+      "grad_norm": 0.17078514397144318,
+      "learning_rate": 5e-06,
+      "loss": 0.081,
+      "step": 14651
+    },
+    {
+      "epoch": 11.028980052691004,
+      "grad_norm": 0.1784096360206604,
+      "learning_rate": 5e-06,
+      "loss": 0.1501,
+      "step": 14652
+    },
+    {
+      "epoch": 11.02973278133233,
+      "grad_norm": 0.20559518039226532,
+      "learning_rate": 5e-06,
+      "loss": 0.1844,
+      "step": 14653
+    },
+    {
+      "epoch": 11.030485509973655,
+      "grad_norm": 0.18815261125564575,
+      "learning_rate": 5e-06,
+      "loss": 0.2015,
+      "step": 14654
+    },
+    {
+      "epoch": 11.031238238614979,
+      "grad_norm": 0.18219226598739624,
+      "learning_rate": 5e-06,
+      "loss": 0.1533,
+      "step": 14655
+    },
+    {
+      "epoch": 11.031990967256304,
+      "grad_norm": 0.20162031054496765,
+      "learning_rate": 5e-06,
+      "loss": 0.166,
+      "step": 14656
+    },
+    {
+      "epoch": 11.03274369589763,
+      "grad_norm": 0.1781560331583023,
+      "learning_rate": 5e-06,
+      "loss": 0.111,
+      "step": 14657
+    },
+    {
+      "epoch": 11.033496424538953,
+      "grad_norm": 0.185746431350708,
+      "learning_rate": 5e-06,
+      "loss": 0.2318,
+      "step": 14658
+    },
+    {
+      "epoch": 11.034249153180278,
+      "grad_norm": 0.1878928542137146,
+      "learning_rate": 5e-06,
+      "loss": 0.1206,
+      "step": 14659
+    },
+    {
+      "epoch": 11.035001881821604,
+      "grad_norm": 0.19155162572860718,
+      "learning_rate": 5e-06,
+      "loss": 0.1356,
+      "step": 14660
+    },
+    {
+      "epoch": 11.035754610462929,
+      "grad_norm": 0.17263276875019073,
+      "learning_rate": 5e-06,
+      "loss": 0.2381,
+      "step": 14661
+    },
+    {
+      "epoch": 11.036507339104253,
+      "grad_norm": 0.16645848751068115,
+      "learning_rate": 5e-06,
+      "loss": 0.0846,
+      "step": 14662
+    },
+    {
+      "epoch": 11.037260067745578,
+      "grad_norm": 0.17840608954429626,
+      "learning_rate": 5e-06,
+      "loss": 0.1511,
+      "step": 14663
+    },
+    {
+      "epoch": 11.038012796386903,
+      "grad_norm": 0.15970781445503235,
+      "learning_rate": 5e-06,
+      "loss": 0.0895,
+      "step": 14664
+    },
+    {
+      "epoch": 11.038765525028227,
+      "grad_norm": 0.17444734275341034,
+      "learning_rate": 5e-06,
+      "loss": 0.1467,
+      "step": 14665
+    },
+    {
+      "epoch": 11.039518253669552,
+      "grad_norm": 0.20410889387130737,
+      "learning_rate": 5e-06,
+      "loss": 0.1857,
+      "step": 14666
+    },
+    {
+      "epoch": 11.040270982310878,
+      "grad_norm": 0.1810118854045868,
+      "learning_rate": 5e-06,
+      "loss": 0.1317,
+      "step": 14667
+    },
+    {
+      "epoch": 11.041023710952201,
+      "grad_norm": 0.18635880947113037,
+      "learning_rate": 5e-06,
+      "loss": 0.19,
+      "step": 14668
+    },
+    {
+      "epoch": 11.041776439593527,
+      "grad_norm": 0.15728391706943512,
+      "learning_rate": 5e-06,
+      "loss": 0.0982,
+      "step": 14669
+    },
+    {
+      "epoch": 11.042529168234852,
+      "grad_norm": 0.17787154018878937,
+      "learning_rate": 5e-06,
+      "loss": 0.1744,
+      "step": 14670
+    },
+    {
+      "epoch": 11.043281896876175,
+      "grad_norm": 0.19107775390148163,
+      "learning_rate": 5e-06,
+      "loss": 0.1376,
+      "step": 14671
+    },
+    {
+      "epoch": 11.0440346255175,
+      "grad_norm": 0.18829824030399323,
+      "learning_rate": 5e-06,
+      "loss": 0.1649,
+      "step": 14672
+    },
+    {
+      "epoch": 11.044787354158826,
+      "grad_norm": 0.18130157887935638,
+      "learning_rate": 5e-06,
+      "loss": 0.2532,
+      "step": 14673
+    },
+    {
+      "epoch": 11.04554008280015,
+      "grad_norm": 0.18140524625778198,
+      "learning_rate": 5e-06,
+      "loss": 0.1511,
+      "step": 14674
+    },
+    {
+      "epoch": 11.046292811441475,
+      "grad_norm": 0.17893467843532562,
+      "learning_rate": 5e-06,
+      "loss": 0.1452,
+      "step": 14675
+    },
+    {
+      "epoch": 11.0470455400828,
+      "grad_norm": 0.16050854325294495,
+      "learning_rate": 5e-06,
+      "loss": 0.1105,
+      "step": 14676
+    },
+    {
+      "epoch": 11.047798268724124,
+      "grad_norm": 0.18894128501415253,
+      "learning_rate": 5e-06,
+      "loss": 0.224,
+      "step": 14677
+    },
+    {
+      "epoch": 11.04855099736545,
+      "grad_norm": 0.20536762475967407,
+      "learning_rate": 5e-06,
+      "loss": 0.1772,
+      "step": 14678
+    },
+    {
+      "epoch": 11.049303726006775,
+      "grad_norm": 0.19032415747642517,
+      "learning_rate": 5e-06,
+      "loss": 0.166,
+      "step": 14679
+    },
+    {
+      "epoch": 11.0500564546481,
+      "grad_norm": 0.16132314503192902,
+      "learning_rate": 5e-06,
+      "loss": 0.0989,
+      "step": 14680
+    },
+    {
+      "epoch": 11.050809183289424,
+      "grad_norm": 0.1861380934715271,
+      "learning_rate": 5e-06,
+      "loss": 0.1511,
+      "step": 14681
+    },
+    {
+      "epoch": 11.051561911930749,
+      "grad_norm": 0.18455281853675842,
+      "learning_rate": 5e-06,
+      "loss": 0.201,
+      "step": 14682
+    },
+    {
+      "epoch": 11.052314640572074,
+      "grad_norm": 0.18069776892662048,
+      "learning_rate": 5e-06,
+      "loss": 0.0854,
+      "step": 14683
+    },
+    {
+      "epoch": 11.053067369213398,
+      "grad_norm": 0.2255975306034088,
+      "learning_rate": 5e-06,
+      "loss": 0.1781,
+      "step": 14684
+    },
+    {
+      "epoch": 11.053820097854723,
+      "grad_norm": 0.1816260665655136,
+      "learning_rate": 5e-06,
+      "loss": 0.1386,
+      "step": 14685
+    },
+    {
+      "epoch": 11.054572826496049,
+      "grad_norm": 0.16985705494880676,
+      "learning_rate": 5e-06,
+      "loss": 0.1658,
+      "step": 14686
+    },
+    {
+      "epoch": 11.055325555137372,
+      "grad_norm": 0.19096645712852478,
+      "learning_rate": 5e-06,
+      "loss": 0.1208,
+      "step": 14687
+    },
+    {
+      "epoch": 11.056078283778698,
+      "grad_norm": 0.17356303334236145,
+      "learning_rate": 5e-06,
+      "loss": 0.1714,
+      "step": 14688
+    },
+    {
+      "epoch": 11.056831012420023,
+      "grad_norm": 0.19218802452087402,
+      "learning_rate": 5e-06,
+      "loss": 0.1692,
+      "step": 14689
+    },
+    {
+      "epoch": 11.057583741061347,
+      "grad_norm": 0.20756231248378754,
+      "learning_rate": 5e-06,
+      "loss": 0.168,
+      "step": 14690
+    },
+    {
+      "epoch": 11.058336469702672,
+      "grad_norm": 0.18469764292240143,
+      "learning_rate": 5e-06,
+      "loss": 0.2465,
+      "step": 14691
+    },
+    {
+      "epoch": 11.059089198343997,
+      "grad_norm": 0.16362226009368896,
+      "learning_rate": 5e-06,
+      "loss": 0.1677,
+      "step": 14692
+    },
+    {
+      "epoch": 11.059841926985321,
+      "grad_norm": 0.19953377544879913,
+      "learning_rate": 5e-06,
+      "loss": 0.1453,
+      "step": 14693
+    },
+    {
+      "epoch": 11.060594655626646,
+      "grad_norm": 0.16419611871242523,
+      "learning_rate": 5e-06,
+      "loss": 0.1229,
+      "step": 14694
+    },
+    {
+      "epoch": 11.061347384267972,
+      "grad_norm": 0.1504286676645279,
+      "learning_rate": 5e-06,
+      "loss": 0.0658,
+      "step": 14695
+    },
+    {
+      "epoch": 11.062100112909297,
+      "grad_norm": 0.17702628672122955,
+      "learning_rate": 5e-06,
+      "loss": 0.1181,
+      "step": 14696
+    },
+    {
+      "epoch": 11.06285284155062,
+      "grad_norm": 0.2017195224761963,
+      "learning_rate": 5e-06,
+      "loss": 0.1254,
+      "step": 14697
+    },
+    {
+      "epoch": 11.063605570191946,
+      "grad_norm": 0.18961618840694427,
+      "learning_rate": 5e-06,
+      "loss": 0.1188,
+      "step": 14698
+    },
+    {
+      "epoch": 11.064358298833271,
+      "grad_norm": 0.17185597121715546,
+      "learning_rate": 5e-06,
+      "loss": 0.1696,
+      "step": 14699
+    },
+    {
+      "epoch": 11.065111027474595,
+      "grad_norm": 0.18562600016593933,
+      "learning_rate": 5e-06,
+      "loss": 0.156,
+      "step": 14700
+    },
+    {
+      "epoch": 11.06586375611592,
+      "grad_norm": 0.2268870621919632,
+      "learning_rate": 5e-06,
+      "loss": 0.1787,
+      "step": 14701
+    },
+    {
+      "epoch": 11.066616484757246,
+      "grad_norm": 0.18318764865398407,
+      "learning_rate": 5e-06,
+      "loss": 0.1406,
+      "step": 14702
+    },
+    {
+      "epoch": 11.06736921339857,
+      "grad_norm": 0.19623912870883942,
+      "learning_rate": 5e-06,
+      "loss": 0.1795,
+      "step": 14703
+    },
+    {
+      "epoch": 11.068121942039895,
+      "grad_norm": 0.2138264775276184,
+      "learning_rate": 5e-06,
+      "loss": 0.1972,
+      "step": 14704
+    },
+    {
+      "epoch": 11.06887467068122,
+      "grad_norm": 0.15390260517597198,
+      "learning_rate": 5e-06,
+      "loss": 0.0909,
+      "step": 14705
+    },
+    {
+      "epoch": 11.069627399322544,
+      "grad_norm": 0.17041060328483582,
+      "learning_rate": 5e-06,
+      "loss": 0.2038,
+      "step": 14706
+    },
+    {
+      "epoch": 11.070380127963869,
+      "grad_norm": 0.19361667335033417,
+      "learning_rate": 5e-06,
+      "loss": 0.1219,
+      "step": 14707
+    },
+    {
+      "epoch": 11.071132856605194,
+      "grad_norm": 0.1814035326242447,
+      "learning_rate": 5e-06,
+      "loss": 0.1395,
+      "step": 14708
+    },
+    {
+      "epoch": 11.071885585246518,
+      "grad_norm": 0.19861356914043427,
+      "learning_rate": 5e-06,
+      "loss": 0.2393,
+      "step": 14709
+    },
+    {
+      "epoch": 11.072638313887843,
+      "grad_norm": 0.1893034130334854,
+      "learning_rate": 5e-06,
+      "loss": 0.1495,
+      "step": 14710
+    },
+    {
+      "epoch": 11.073391042529169,
+      "grad_norm": 0.19729742407798767,
+      "learning_rate": 5e-06,
+      "loss": 0.1439,
+      "step": 14711
+    },
+    {
+      "epoch": 11.074143771170492,
+      "grad_norm": 0.17948681116104126,
+      "learning_rate": 5e-06,
+      "loss": 0.2019,
+      "step": 14712
+    },
+    {
+      "epoch": 11.074896499811818,
+      "grad_norm": 0.1914609968662262,
+      "learning_rate": 5e-06,
+      "loss": 0.1902,
+      "step": 14713
+    },
+    {
+      "epoch": 11.075649228453143,
+      "grad_norm": 0.17920680344104767,
+      "learning_rate": 5e-06,
+      "loss": 0.1235,
+      "step": 14714
+    },
+    {
+      "epoch": 11.076401957094468,
+      "grad_norm": 0.1966412514448166,
+      "learning_rate": 5e-06,
+      "loss": 0.1577,
+      "step": 14715
+    },
+    {
+      "epoch": 11.077154685735792,
+      "grad_norm": 0.16337646543979645,
+      "learning_rate": 5e-06,
+      "loss": 0.0934,
+      "step": 14716
+    },
+    {
+      "epoch": 11.077907414377117,
+      "grad_norm": 0.1912173479795456,
+      "learning_rate": 5e-06,
+      "loss": 0.1876,
+      "step": 14717
+    },
+    {
+      "epoch": 11.078660143018443,
+      "grad_norm": 0.1460835486650467,
+      "learning_rate": 5e-06,
+      "loss": 0.154,
+      "step": 14718
+    },
+    {
+      "epoch": 11.079412871659766,
+      "grad_norm": 0.1819196194410324,
+      "learning_rate": 5e-06,
+      "loss": 0.1845,
+      "step": 14719
+    },
+    {
+      "epoch": 11.080165600301092,
+      "grad_norm": 0.17125923931598663,
+      "learning_rate": 5e-06,
+      "loss": 0.2153,
+      "step": 14720
+    },
+    {
+      "epoch": 11.080918328942417,
+      "grad_norm": 0.17012956738471985,
+      "learning_rate": 5e-06,
+      "loss": 0.1482,
+      "step": 14721
+    },
+    {
+      "epoch": 11.08167105758374,
+      "grad_norm": 0.19119347631931305,
+      "learning_rate": 5e-06,
+      "loss": 0.0991,
+      "step": 14722
+    },
+    {
+      "epoch": 11.082423786225066,
+      "grad_norm": 0.18069575726985931,
+      "learning_rate": 5e-06,
+      "loss": 0.1717,
+      "step": 14723
+    },
+    {
+      "epoch": 11.083176514866391,
+      "grad_norm": 0.20700375735759735,
+      "learning_rate": 5e-06,
+      "loss": 0.1867,
+      "step": 14724
+    },
+    {
+      "epoch": 11.083929243507715,
+      "grad_norm": 0.1638200581073761,
+      "learning_rate": 5e-06,
+      "loss": 0.1171,
+      "step": 14725
+    },
+    {
+      "epoch": 11.08468197214904,
+      "grad_norm": 0.15568751096725464,
+      "learning_rate": 5e-06,
+      "loss": 0.1226,
+      "step": 14726
+    },
+    {
+      "epoch": 11.085434700790366,
+      "grad_norm": 0.18224786221981049,
+      "learning_rate": 5e-06,
+      "loss": 0.1297,
+      "step": 14727
+    },
+    {
+      "epoch": 11.086187429431689,
+      "grad_norm": 0.18718686699867249,
+      "learning_rate": 5e-06,
+      "loss": 0.1174,
+      "step": 14728
+    },
+    {
+      "epoch": 11.086940158073014,
+      "grad_norm": 0.1795811802148819,
+      "learning_rate": 5e-06,
+      "loss": 0.1419,
+      "step": 14729
+    },
+    {
+      "epoch": 11.08769288671434,
+      "grad_norm": 0.19356897473335266,
+      "learning_rate": 5e-06,
+      "loss": 0.1828,
+      "step": 14730
+    },
+    {
+      "epoch": 11.088445615355663,
+      "grad_norm": 0.18179042637348175,
+      "learning_rate": 5e-06,
+      "loss": 0.1578,
+      "step": 14731
+    },
+    {
+      "epoch": 11.089198343996989,
+      "grad_norm": 0.17637276649475098,
+      "learning_rate": 5e-06,
+      "loss": 0.1393,
+      "step": 14732
+    },
+    {
+      "epoch": 11.089951072638314,
+      "grad_norm": 0.16941088438034058,
+      "learning_rate": 5e-06,
+      "loss": 0.1489,
+      "step": 14733
+    },
+    {
+      "epoch": 11.09070380127964,
+      "grad_norm": 0.16449816524982452,
+      "learning_rate": 5e-06,
+      "loss": 0.1296,
+      "step": 14734
+    },
+    {
+      "epoch": 11.091456529920963,
+      "grad_norm": 0.1860218197107315,
+      "learning_rate": 5e-06,
+      "loss": 0.1315,
+      "step": 14735
+    },
+    {
+      "epoch": 11.092209258562288,
+      "grad_norm": 0.15493252873420715,
+      "learning_rate": 5e-06,
+      "loss": 0.0982,
+      "step": 14736
+    },
+    {
+      "epoch": 11.092961987203614,
+      "grad_norm": 0.1607757806777954,
+      "learning_rate": 5e-06,
+      "loss": 0.1633,
+      "step": 14737
+    },
+    {
+      "epoch": 11.093714715844937,
+      "grad_norm": 0.20988336205482483,
+      "learning_rate": 5e-06,
+      "loss": 0.1412,
+      "step": 14738
+    },
+    {
+      "epoch": 11.094467444486263,
+      "grad_norm": 0.1983179748058319,
+      "learning_rate": 5e-06,
+      "loss": 0.1869,
+      "step": 14739
+    },
+    {
+      "epoch": 11.095220173127588,
+      "grad_norm": 0.16584739089012146,
+      "learning_rate": 5e-06,
+      "loss": 0.1096,
+      "step": 14740
+    },
+    {
+      "epoch": 11.095972901768912,
+      "grad_norm": 0.18882928788661957,
+      "learning_rate": 5e-06,
+      "loss": 0.1832,
+      "step": 14741
+    },
+    {
+      "epoch": 11.096725630410237,
+      "grad_norm": 0.19600312411785126,
+      "learning_rate": 5e-06,
+      "loss": 0.1653,
+      "step": 14742
+    },
+    {
+      "epoch": 11.097478359051562,
+      "grad_norm": 0.18170911073684692,
+      "learning_rate": 5e-06,
+      "loss": 0.1202,
+      "step": 14743
+    },
+    {
+      "epoch": 11.098231087692886,
+      "grad_norm": 0.17362695932388306,
+      "learning_rate": 5e-06,
+      "loss": 0.0874,
+      "step": 14744
+    },
+    {
+      "epoch": 11.098983816334211,
+      "grad_norm": 0.16749189794063568,
+      "learning_rate": 5e-06,
+      "loss": 0.1334,
+      "step": 14745
+    },
+    {
+      "epoch": 11.099736544975537,
+      "grad_norm": 0.1839837282896042,
+      "learning_rate": 5e-06,
+      "loss": 0.2183,
+      "step": 14746
+    },
+    {
+      "epoch": 11.10048927361686,
+      "grad_norm": 0.1657716929912567,
+      "learning_rate": 5e-06,
+      "loss": 0.0971,
+      "step": 14747
+    },
+    {
+      "epoch": 11.101242002258186,
+      "grad_norm": 0.17359136044979095,
+      "learning_rate": 5e-06,
+      "loss": 0.071,
+      "step": 14748
+    },
+    {
+      "epoch": 11.101994730899511,
+      "grad_norm": 0.1876412034034729,
+      "learning_rate": 5e-06,
+      "loss": 0.2053,
+      "step": 14749
+    },
+    {
+      "epoch": 11.102747459540836,
+      "grad_norm": 0.20401234924793243,
+      "learning_rate": 5e-06,
+      "loss": 0.1982,
+      "step": 14750
+    },
+    {
+      "epoch": 11.10350018818216,
+      "grad_norm": 0.19486217200756073,
+      "learning_rate": 5e-06,
+      "loss": 0.162,
+      "step": 14751
+    },
+    {
+      "epoch": 11.104252916823485,
+      "grad_norm": 0.1855892688035965,
+      "learning_rate": 5e-06,
+      "loss": 0.0753,
+      "step": 14752
+    },
+    {
+      "epoch": 11.10500564546481,
+      "grad_norm": 0.18040655553340912,
+      "learning_rate": 5e-06,
+      "loss": 0.1556,
+      "step": 14753
+    },
+    {
+      "epoch": 11.105758374106134,
+      "grad_norm": 0.17353899776935577,
+      "learning_rate": 5e-06,
+      "loss": 0.0675,
+      "step": 14754
+    },
+    {
+      "epoch": 11.10651110274746,
+      "grad_norm": 0.18809862434864044,
+      "learning_rate": 5e-06,
+      "loss": 0.0943,
+      "step": 14755
+    },
+    {
+      "epoch": 11.107263831388785,
+      "grad_norm": 0.19124239683151245,
+      "learning_rate": 5e-06,
+      "loss": 0.1725,
+      "step": 14756
+    },
+    {
+      "epoch": 11.108016560030109,
+      "grad_norm": 0.19827789068222046,
+      "learning_rate": 5e-06,
+      "loss": 0.1822,
+      "step": 14757
+    },
+    {
+      "epoch": 11.108769288671434,
+      "grad_norm": 0.19077639281749725,
+      "learning_rate": 5e-06,
+      "loss": 0.116,
+      "step": 14758
+    },
+    {
+      "epoch": 11.10952201731276,
+      "grad_norm": 0.17967155575752258,
+      "learning_rate": 5e-06,
+      "loss": 0.1158,
+      "step": 14759
+    },
+    {
+      "epoch": 11.110274745954083,
+      "grad_norm": 0.1769682615995407,
+      "learning_rate": 5e-06,
+      "loss": 0.1606,
+      "step": 14760
+    },
+    {
+      "epoch": 11.111027474595408,
+      "grad_norm": 0.19750061631202698,
+      "learning_rate": 5e-06,
+      "loss": 0.1482,
+      "step": 14761
+    },
+    {
+      "epoch": 11.111780203236734,
+      "grad_norm": 0.20273610949516296,
+      "learning_rate": 5e-06,
+      "loss": 0.2287,
+      "step": 14762
+    },
+    {
+      "epoch": 11.112532931878057,
+      "grad_norm": 0.1719943881034851,
+      "learning_rate": 5e-06,
+      "loss": 0.0777,
+      "step": 14763
+    },
+    {
+      "epoch": 11.113285660519383,
+      "grad_norm": 0.20314615964889526,
+      "learning_rate": 5e-06,
+      "loss": 0.1764,
+      "step": 14764
+    },
+    {
+      "epoch": 11.114038389160708,
+      "grad_norm": 0.1756068915128708,
+      "learning_rate": 5e-06,
+      "loss": 0.1019,
+      "step": 14765
+    },
+    {
+      "epoch": 11.114791117802032,
+      "grad_norm": 0.17703302204608917,
+      "learning_rate": 5e-06,
+      "loss": 0.0699,
+      "step": 14766
+    },
+    {
+      "epoch": 11.115543846443357,
+      "grad_norm": 0.15416060388088226,
+      "learning_rate": 5e-06,
+      "loss": 0.0946,
+      "step": 14767
+    },
+    {
+      "epoch": 11.116296575084682,
+      "grad_norm": 0.1605541706085205,
+      "learning_rate": 5e-06,
+      "loss": 0.0834,
+      "step": 14768
+    },
+    {
+      "epoch": 11.117049303726008,
+      "grad_norm": 0.1745283603668213,
+      "learning_rate": 5e-06,
+      "loss": 0.1773,
+      "step": 14769
+    },
+    {
+      "epoch": 11.117802032367331,
+      "grad_norm": 0.1909770965576172,
+      "learning_rate": 5e-06,
+      "loss": 0.1631,
+      "step": 14770
+    },
+    {
+      "epoch": 11.118554761008657,
+      "grad_norm": 0.1672181636095047,
+      "learning_rate": 5e-06,
+      "loss": 0.1096,
+      "step": 14771
+    },
+    {
+      "epoch": 11.119307489649982,
+      "grad_norm": 0.17458105087280273,
+      "learning_rate": 5e-06,
+      "loss": 0.1514,
+      "step": 14772
+    },
+    {
+      "epoch": 11.120060218291306,
+      "grad_norm": 0.1918034553527832,
+      "learning_rate": 5e-06,
+      "loss": 0.0681,
+      "step": 14773
+    },
+    {
+      "epoch": 11.120812946932631,
+      "grad_norm": 0.17855818569660187,
+      "learning_rate": 5e-06,
+      "loss": 0.1553,
+      "step": 14774
+    },
+    {
+      "epoch": 11.121565675573956,
+      "grad_norm": 0.17067041993141174,
+      "learning_rate": 5e-06,
+      "loss": 0.1206,
+      "step": 14775
+    },
+    {
+      "epoch": 11.12231840421528,
+      "grad_norm": 0.17931771278381348,
+      "learning_rate": 5e-06,
+      "loss": 0.1094,
+      "step": 14776
+    },
+    {
+      "epoch": 11.123071132856605,
+      "grad_norm": 0.1559484750032425,
+      "learning_rate": 5e-06,
+      "loss": 0.1151,
+      "step": 14777
+    },
+    {
+      "epoch": 11.12382386149793,
+      "grad_norm": 0.18524815142154694,
+      "learning_rate": 5e-06,
+      "loss": 0.1699,
+      "step": 14778
+    },
+    {
+      "epoch": 11.124576590139254,
+      "grad_norm": 0.17065903544425964,
+      "learning_rate": 5e-06,
+      "loss": 0.1431,
+      "step": 14779
+    },
+    {
+      "epoch": 11.12532931878058,
+      "grad_norm": 0.19535885751247406,
+      "learning_rate": 5e-06,
+      "loss": 0.1779,
+      "step": 14780
+    },
+    {
+      "epoch": 11.126082047421905,
+      "grad_norm": 0.190781831741333,
+      "learning_rate": 5e-06,
+      "loss": 0.13,
+      "step": 14781
+    },
+    {
+      "epoch": 11.126834776063228,
+      "grad_norm": 0.18039818108081818,
+      "learning_rate": 5e-06,
+      "loss": 0.1886,
+      "step": 14782
+    },
+    {
+      "epoch": 11.127587504704554,
+      "grad_norm": 0.17439720034599304,
+      "learning_rate": 5e-06,
+      "loss": 0.0989,
+      "step": 14783
+    },
+    {
+      "epoch": 11.12834023334588,
+      "grad_norm": 0.18992416560649872,
+      "learning_rate": 5e-06,
+      "loss": 0.1232,
+      "step": 14784
+    },
+    {
+      "epoch": 11.129092961987203,
+      "grad_norm": 0.17920558154582977,
+      "learning_rate": 5e-06,
+      "loss": 0.0906,
+      "step": 14785
+    },
+    {
+      "epoch": 11.129845690628528,
+      "grad_norm": 0.18104496598243713,
+      "learning_rate": 5e-06,
+      "loss": 0.1379,
+      "step": 14786
+    },
+    {
+      "epoch": 11.130598419269853,
+      "grad_norm": 0.1791551262140274,
+      "learning_rate": 5e-06,
+      "loss": 0.1935,
+      "step": 14787
+    },
+    {
+      "epoch": 11.131351147911179,
+      "grad_norm": 0.17741553485393524,
+      "learning_rate": 5e-06,
+      "loss": 0.1868,
+      "step": 14788
+    },
+    {
+      "epoch": 11.132103876552502,
+      "grad_norm": 0.17449674010276794,
+      "learning_rate": 5e-06,
+      "loss": 0.1653,
+      "step": 14789
+    },
+    {
+      "epoch": 11.132856605193828,
+      "grad_norm": 0.18294234573841095,
+      "learning_rate": 5e-06,
+      "loss": 0.1933,
+      "step": 14790
+    },
+    {
+      "epoch": 11.133609333835153,
+      "grad_norm": 0.16397930681705475,
+      "learning_rate": 5e-06,
+      "loss": 0.0726,
+      "step": 14791
+    },
+    {
+      "epoch": 11.134362062476477,
+      "grad_norm": 0.19565501809120178,
+      "learning_rate": 5e-06,
+      "loss": 0.1247,
+      "step": 14792
+    },
+    {
+      "epoch": 11.135114791117802,
+      "grad_norm": 0.16637705266475677,
+      "learning_rate": 5e-06,
+      "loss": 0.1721,
+      "step": 14793
+    },
+    {
+      "epoch": 11.135867519759127,
+      "grad_norm": 0.17564120888710022,
+      "learning_rate": 5e-06,
+      "loss": 0.1715,
+      "step": 14794
+    },
+    {
+      "epoch": 11.136620248400451,
+      "grad_norm": 0.16244864463806152,
+      "learning_rate": 5e-06,
+      "loss": 0.1183,
+      "step": 14795
+    },
+    {
+      "epoch": 11.137372977041776,
+      "grad_norm": 0.1945173740386963,
+      "learning_rate": 5e-06,
+      "loss": 0.1325,
+      "step": 14796
+    },
+    {
+      "epoch": 11.138125705683102,
+      "grad_norm": 0.19735059142112732,
+      "learning_rate": 5e-06,
+      "loss": 0.2103,
+      "step": 14797
+    },
+    {
+      "epoch": 11.138878434324425,
+      "grad_norm": 0.17041146755218506,
+      "learning_rate": 5e-06,
+      "loss": 0.097,
+      "step": 14798
+    },
+    {
+      "epoch": 11.13963116296575,
+      "grad_norm": 0.15573334693908691,
+      "learning_rate": 5e-06,
+      "loss": 0.1069,
+      "step": 14799
+    },
+    {
+      "epoch": 11.140383891607076,
+      "grad_norm": 0.21030430495738983,
+      "learning_rate": 5e-06,
+      "loss": 0.0955,
+      "step": 14800
+    },
+    {
+      "epoch": 11.140383891607076,
+      "eval_loss": 0.18728302419185638,
+      "eval_runtime": 452.682,
+      "eval_samples_per_second": 21.267,
+      "eval_steps_per_second": 0.665,
+      "step": 14800
+    },
+    {
+      "epoch": 11.1411366202484,
+      "grad_norm": 0.19011299312114716,
+      "learning_rate": 5e-06,
+      "loss": 0.1102,
+      "step": 14801
+    },
+    {
+      "epoch": 11.141889348889725,
+      "grad_norm": 0.20328572392463684,
+      "learning_rate": 5e-06,
+      "loss": 0.1378,
+      "step": 14802
+    },
+    {
+      "epoch": 11.14264207753105,
+      "grad_norm": 0.1999550461769104,
+      "learning_rate": 5e-06,
+      "loss": 0.151,
+      "step": 14803
+    },
+    {
+      "epoch": 11.143394806172374,
+      "grad_norm": 0.18123337626457214,
+      "learning_rate": 5e-06,
+      "loss": 0.0851,
+      "step": 14804
+    },
+    {
+      "epoch": 11.1441475348137,
+      "grad_norm": 0.16264620423316956,
+      "learning_rate": 5e-06,
+      "loss": 0.1083,
+      "step": 14805
+    },
+    {
+      "epoch": 11.144900263455025,
+      "grad_norm": 0.16072243452072144,
+      "learning_rate": 5e-06,
+      "loss": 0.163,
+      "step": 14806
+    },
+    {
+      "epoch": 11.14565299209635,
+      "grad_norm": 0.16674530506134033,
+      "learning_rate": 5e-06,
+      "loss": 0.1025,
+      "step": 14807
+    },
+    {
+      "epoch": 11.146405720737674,
+      "grad_norm": 0.1806487888097763,
+      "learning_rate": 5e-06,
+      "loss": 0.166,
+      "step": 14808
+    },
+    {
+      "epoch": 11.147158449378999,
+      "grad_norm": 0.20102480053901672,
+      "learning_rate": 5e-06,
+      "loss": 0.1747,
+      "step": 14809
+    },
+    {
+      "epoch": 11.147911178020324,
+      "grad_norm": 0.19214217364788055,
+      "learning_rate": 5e-06,
+      "loss": 0.1281,
+      "step": 14810
+    },
+    {
+      "epoch": 11.148663906661648,
+      "grad_norm": 0.1806429922580719,
+      "learning_rate": 5e-06,
+      "loss": 0.1453,
+      "step": 14811
+    },
+    {
+      "epoch": 11.149416635302973,
+      "grad_norm": 0.18962691724300385,
+      "learning_rate": 5e-06,
+      "loss": 0.1985,
+      "step": 14812
+    },
+    {
+      "epoch": 11.150169363944299,
+      "grad_norm": 0.16180600225925446,
+      "learning_rate": 5e-06,
+      "loss": 0.1458,
+      "step": 14813
+    },
+    {
+      "epoch": 11.150922092585622,
+      "grad_norm": 0.19710668921470642,
+      "learning_rate": 5e-06,
+      "loss": 0.1548,
+      "step": 14814
+    },
+    {
+      "epoch": 11.151674821226948,
+      "grad_norm": 0.1644078940153122,
+      "learning_rate": 5e-06,
+      "loss": 0.1059,
+      "step": 14815
+    },
+    {
+      "epoch": 11.152427549868273,
+      "grad_norm": 0.16627205908298492,
+      "learning_rate": 5e-06,
+      "loss": 0.1288,
+      "step": 14816
+    },
+    {
+      "epoch": 11.153180278509597,
+      "grad_norm": 0.19510747492313385,
+      "learning_rate": 5e-06,
+      "loss": 0.1729,
+      "step": 14817
+    },
+    {
+      "epoch": 11.153933007150922,
+      "grad_norm": 0.17169563472270966,
+      "learning_rate": 5e-06,
+      "loss": 0.1028,
+      "step": 14818
+    },
+    {
+      "epoch": 11.154685735792247,
+      "grad_norm": 0.19315610826015472,
+      "learning_rate": 5e-06,
+      "loss": 0.1272,
+      "step": 14819
+    },
+    {
+      "epoch": 11.155438464433571,
+      "grad_norm": 0.16556361317634583,
+      "learning_rate": 5e-06,
+      "loss": 0.1232,
+      "step": 14820
+    },
+    {
+      "epoch": 11.156191193074896,
+      "grad_norm": 0.18950004875659943,
+      "learning_rate": 5e-06,
+      "loss": 0.1439,
+      "step": 14821
+    },
+    {
+      "epoch": 11.156943921716222,
+      "grad_norm": 0.15688778460025787,
+      "learning_rate": 5e-06,
+      "loss": 0.119,
+      "step": 14822
+    },
+    {
+      "epoch": 11.157696650357547,
+      "grad_norm": 0.2220509797334671,
+      "learning_rate": 5e-06,
+      "loss": 0.1689,
+      "step": 14823
+    },
+    {
+      "epoch": 11.15844937899887,
+      "grad_norm": 0.17872020602226257,
+      "learning_rate": 5e-06,
+      "loss": 0.123,
+      "step": 14824
+    },
+    {
+      "epoch": 11.159202107640196,
+      "grad_norm": 0.18923814594745636,
+      "learning_rate": 5e-06,
+      "loss": 0.219,
+      "step": 14825
+    },
+    {
+      "epoch": 11.159954836281521,
+      "grad_norm": 0.1820460855960846,
+      "learning_rate": 5e-06,
+      "loss": 0.1529,
+      "step": 14826
+    },
+    {
+      "epoch": 11.160707564922845,
+      "grad_norm": 0.16604363918304443,
+      "learning_rate": 5e-06,
+      "loss": 0.0964,
+      "step": 14827
+    },
+    {
+      "epoch": 11.16146029356417,
+      "grad_norm": 0.1968279927968979,
+      "learning_rate": 5e-06,
+      "loss": 0.1714,
+      "step": 14828
+    },
+    {
+      "epoch": 11.162213022205496,
+      "grad_norm": 0.18304622173309326,
+      "learning_rate": 5e-06,
+      "loss": 0.1375,
+      "step": 14829
+    },
+    {
+      "epoch": 11.16296575084682,
+      "grad_norm": 0.16282790899276733,
+      "learning_rate": 5e-06,
+      "loss": 0.1038,
+      "step": 14830
+    },
+    {
+      "epoch": 11.163718479488145,
+      "grad_norm": 0.1804710328578949,
+      "learning_rate": 5e-06,
+      "loss": 0.1233,
+      "step": 14831
+    },
+    {
+      "epoch": 11.16447120812947,
+      "grad_norm": 0.17502360045909882,
+      "learning_rate": 5e-06,
+      "loss": 0.1653,
+      "step": 14832
+    },
+    {
+      "epoch": 11.165223936770794,
+      "grad_norm": 0.1709318906068802,
+      "learning_rate": 5e-06,
+      "loss": 0.1058,
+      "step": 14833
+    },
+    {
+      "epoch": 11.165976665412119,
+      "grad_norm": 0.1622878462076187,
+      "learning_rate": 5e-06,
+      "loss": 0.0929,
+      "step": 14834
+    },
+    {
+      "epoch": 11.166729394053444,
+      "grad_norm": 0.22366048395633698,
+      "learning_rate": 5e-06,
+      "loss": 0.1768,
+      "step": 14835
+    },
+    {
+      "epoch": 11.167482122694768,
+      "grad_norm": 0.1902773678302765,
+      "learning_rate": 5e-06,
+      "loss": 0.1025,
+      "step": 14836
+    },
+    {
+      "epoch": 11.168234851336093,
+      "grad_norm": 0.17565560340881348,
+      "learning_rate": 5e-06,
+      "loss": 0.1098,
+      "step": 14837
+    },
+    {
+      "epoch": 11.168987579977419,
+      "grad_norm": 0.20084530115127563,
+      "learning_rate": 5e-06,
+      "loss": 0.1749,
+      "step": 14838
+    },
+    {
+      "epoch": 11.169740308618742,
+      "grad_norm": 0.17309603095054626,
+      "learning_rate": 5e-06,
+      "loss": 0.1746,
+      "step": 14839
+    },
+    {
+      "epoch": 11.170493037260067,
+      "grad_norm": 0.16925394535064697,
+      "learning_rate": 5e-06,
+      "loss": 0.1109,
+      "step": 14840
+    },
+    {
+      "epoch": 11.171245765901393,
+      "grad_norm": 0.18390141427516937,
+      "learning_rate": 5e-06,
+      "loss": 0.1043,
+      "step": 14841
+    },
+    {
+      "epoch": 11.171998494542718,
+      "grad_norm": 0.21691913902759552,
+      "learning_rate": 5e-06,
+      "loss": 0.1582,
+      "step": 14842
+    },
+    {
+      "epoch": 11.172751223184042,
+      "grad_norm": 0.19992879033088684,
+      "learning_rate": 5e-06,
+      "loss": 0.1079,
+      "step": 14843
+    },
+    {
+      "epoch": 11.173503951825367,
+      "grad_norm": 0.17334602773189545,
+      "learning_rate": 5e-06,
+      "loss": 0.1762,
+      "step": 14844
+    },
+    {
+      "epoch": 11.174256680466693,
+      "grad_norm": 0.19325172901153564,
+      "learning_rate": 5e-06,
+      "loss": 0.1656,
+      "step": 14845
+    },
+    {
+      "epoch": 11.175009409108016,
+      "grad_norm": 0.15736718475818634,
+      "learning_rate": 5e-06,
+      "loss": 0.1014,
+      "step": 14846
+    },
+    {
+      "epoch": 11.175762137749341,
+      "grad_norm": 0.16770227253437042,
+      "learning_rate": 5e-06,
+      "loss": 0.0534,
+      "step": 14847
+    },
+    {
+      "epoch": 11.176514866390667,
+      "grad_norm": 0.18042854964733124,
+      "learning_rate": 5e-06,
+      "loss": 0.0714,
+      "step": 14848
+    },
+    {
+      "epoch": 11.17726759503199,
+      "grad_norm": 0.18823575973510742,
+      "learning_rate": 5e-06,
+      "loss": 0.2005,
+      "step": 14849
+    },
+    {
+      "epoch": 11.178020323673316,
+      "grad_norm": 0.1906793713569641,
+      "learning_rate": 5e-06,
+      "loss": 0.1551,
+      "step": 14850
+    },
+    {
+      "epoch": 11.178773052314641,
+      "grad_norm": 0.17282900214195251,
+      "learning_rate": 5e-06,
+      "loss": 0.139,
+      "step": 14851
+    },
+    {
+      "epoch": 11.179525780955965,
+      "grad_norm": 0.18799087405204773,
+      "learning_rate": 5e-06,
+      "loss": 0.1248,
+      "step": 14852
+    },
+    {
+      "epoch": 11.18027850959729,
+      "grad_norm": 0.16222429275512695,
+      "learning_rate": 5e-06,
+      "loss": 0.0899,
+      "step": 14853
+    },
+    {
+      "epoch": 11.181031238238615,
+      "grad_norm": 0.15515996515750885,
+      "learning_rate": 5e-06,
+      "loss": 0.1111,
+      "step": 14854
+    },
+    {
+      "epoch": 11.181783966879939,
+      "grad_norm": 0.1811126172542572,
+      "learning_rate": 5e-06,
+      "loss": 0.1186,
+      "step": 14855
+    },
+    {
+      "epoch": 11.182536695521264,
+      "grad_norm": 0.1988145112991333,
+      "learning_rate": 5e-06,
+      "loss": 0.1734,
+      "step": 14856
+    },
+    {
+      "epoch": 11.18328942416259,
+      "grad_norm": 0.1749912053346634,
+      "learning_rate": 5e-06,
+      "loss": 0.1139,
+      "step": 14857
+    },
+    {
+      "epoch": 11.184042152803913,
+      "grad_norm": 0.17767830193042755,
+      "learning_rate": 5e-06,
+      "loss": 0.1467,
+      "step": 14858
+    },
+    {
+      "epoch": 11.184794881445239,
+      "grad_norm": 0.1639220267534256,
+      "learning_rate": 5e-06,
+      "loss": 0.1701,
+      "step": 14859
+    },
+    {
+      "epoch": 11.185547610086564,
+      "grad_norm": 0.1878834217786789,
+      "learning_rate": 5e-06,
+      "loss": 0.1375,
+      "step": 14860
+    },
+    {
+      "epoch": 11.18630033872789,
+      "grad_norm": 0.17276020348072052,
+      "learning_rate": 5e-06,
+      "loss": 0.1122,
+      "step": 14861
+    },
+    {
+      "epoch": 11.187053067369213,
+      "grad_norm": 0.17373442649841309,
+      "learning_rate": 5e-06,
+      "loss": 0.1515,
+      "step": 14862
+    },
+    {
+      "epoch": 11.187805796010538,
+      "grad_norm": 0.17643272876739502,
+      "learning_rate": 5e-06,
+      "loss": 0.1256,
+      "step": 14863
+    },
+    {
+      "epoch": 11.188558524651864,
+      "grad_norm": 0.15693356096744537,
+      "learning_rate": 5e-06,
+      "loss": 0.1491,
+      "step": 14864
+    },
+    {
+      "epoch": 11.189311253293187,
+      "grad_norm": 0.1609104871749878,
+      "learning_rate": 5e-06,
+      "loss": 0.0657,
+      "step": 14865
+    },
+    {
+      "epoch": 11.190063981934513,
+      "grad_norm": 0.1815064400434494,
+      "learning_rate": 5e-06,
+      "loss": 0.1082,
+      "step": 14866
+    },
+    {
+      "epoch": 11.190816710575838,
+      "grad_norm": 0.20165814459323883,
+      "learning_rate": 5e-06,
+      "loss": 0.1782,
+      "step": 14867
+    },
+    {
+      "epoch": 11.191569439217162,
+      "grad_norm": 0.19169381260871887,
+      "learning_rate": 5e-06,
+      "loss": 0.1933,
+      "step": 14868
+    },
+    {
+      "epoch": 11.192322167858487,
+      "grad_norm": 0.23829735815525055,
+      "learning_rate": 5e-06,
+      "loss": 0.1255,
+      "step": 14869
+    },
+    {
+      "epoch": 11.193074896499812,
+      "grad_norm": 0.176339790225029,
+      "learning_rate": 5e-06,
+      "loss": 0.2597,
+      "step": 14870
+    },
+    {
+      "epoch": 11.193827625141136,
+      "grad_norm": 0.20489776134490967,
+      "learning_rate": 5e-06,
+      "loss": 0.2538,
+      "step": 14871
+    },
+    {
+      "epoch": 11.194580353782461,
+      "grad_norm": 0.16357511281967163,
+      "learning_rate": 5e-06,
+      "loss": 0.1299,
+      "step": 14872
+    },
+    {
+      "epoch": 11.195333082423787,
+      "grad_norm": 0.18961802124977112,
+      "learning_rate": 5e-06,
+      "loss": 0.1035,
+      "step": 14873
+    },
+    {
+      "epoch": 11.19608581106511,
+      "grad_norm": 0.2074664980173111,
+      "learning_rate": 5e-06,
+      "loss": 0.1788,
+      "step": 14874
+    },
+    {
+      "epoch": 11.196838539706436,
+      "grad_norm": 0.21492555737495422,
+      "learning_rate": 5e-06,
+      "loss": 0.1409,
+      "step": 14875
+    },
+    {
+      "epoch": 11.197591268347761,
+      "grad_norm": 0.18928620219230652,
+      "learning_rate": 5e-06,
+      "loss": 0.1379,
+      "step": 14876
+    },
+    {
+      "epoch": 11.198343996989085,
+      "grad_norm": 0.24098791182041168,
+      "learning_rate": 5e-06,
+      "loss": 0.1422,
+      "step": 14877
+    },
+    {
+      "epoch": 11.19909672563041,
+      "grad_norm": 0.17554932832717896,
+      "learning_rate": 5e-06,
+      "loss": 0.1556,
+      "step": 14878
+    },
+    {
+      "epoch": 11.199849454271735,
+      "grad_norm": 0.20842105150222778,
+      "learning_rate": 5e-06,
+      "loss": 0.2287,
+      "step": 14879
+    },
+    {
+      "epoch": 11.20060218291306,
+      "grad_norm": 0.19213423132896423,
+      "learning_rate": 5e-06,
+      "loss": 0.1389,
+      "step": 14880
+    },
+    {
+      "epoch": 11.201354911554384,
+      "grad_norm": 0.1945149302482605,
+      "learning_rate": 5e-06,
+      "loss": 0.1511,
+      "step": 14881
+    },
+    {
+      "epoch": 11.20210764019571,
+      "grad_norm": 0.18626970052719116,
+      "learning_rate": 5e-06,
+      "loss": 0.1899,
+      "step": 14882
+    },
+    {
+      "epoch": 11.202860368837035,
+      "grad_norm": 0.19046948850154877,
+      "learning_rate": 5e-06,
+      "loss": 0.1525,
+      "step": 14883
+    },
+    {
+      "epoch": 11.203613097478359,
+      "grad_norm": 0.2190067172050476,
+      "learning_rate": 5e-06,
+      "loss": 0.1746,
+      "step": 14884
+    },
+    {
+      "epoch": 11.204365826119684,
+      "grad_norm": 0.17182794213294983,
+      "learning_rate": 5e-06,
+      "loss": 0.0809,
+      "step": 14885
+    },
+    {
+      "epoch": 11.20511855476101,
+      "grad_norm": 0.2097056657075882,
+      "learning_rate": 5e-06,
+      "loss": 0.1719,
+      "step": 14886
+    },
+    {
+      "epoch": 11.205871283402333,
+      "grad_norm": 0.1861467808485031,
+      "learning_rate": 5e-06,
+      "loss": 0.0993,
+      "step": 14887
+    },
+    {
+      "epoch": 11.206624012043658,
+      "grad_norm": 0.20271892845630646,
+      "learning_rate": 5e-06,
+      "loss": 0.1258,
+      "step": 14888
+    },
+    {
+      "epoch": 11.207376740684984,
+      "grad_norm": 0.1721133440732956,
+      "learning_rate": 5e-06,
+      "loss": 0.0915,
+      "step": 14889
+    },
+    {
+      "epoch": 11.208129469326307,
+      "grad_norm": 0.1607070118188858,
+      "learning_rate": 5e-06,
+      "loss": 0.1216,
+      "step": 14890
+    },
+    {
+      "epoch": 11.208882197967633,
+      "grad_norm": 0.19416400790214539,
+      "learning_rate": 5e-06,
+      "loss": 0.1436,
+      "step": 14891
+    },
+    {
+      "epoch": 11.209634926608958,
+      "grad_norm": 0.16832183301448822,
+      "learning_rate": 5e-06,
+      "loss": 0.1474,
+      "step": 14892
+    },
+    {
+      "epoch": 11.210387655250281,
+      "grad_norm": 0.1784304827451706,
+      "learning_rate": 5e-06,
+      "loss": 0.0941,
+      "step": 14893
+    },
+    {
+      "epoch": 11.211140383891607,
+      "grad_norm": 0.1979239284992218,
+      "learning_rate": 5e-06,
+      "loss": 0.1618,
+      "step": 14894
+    },
+    {
+      "epoch": 11.211893112532932,
+      "grad_norm": 0.18177509307861328,
+      "learning_rate": 5e-06,
+      "loss": 0.1398,
+      "step": 14895
+    },
+    {
+      "epoch": 11.212645841174258,
+      "grad_norm": 0.2160494327545166,
+      "learning_rate": 5e-06,
+      "loss": 0.1135,
+      "step": 14896
+    },
+    {
+      "epoch": 11.213398569815581,
+      "grad_norm": 0.17427273094654083,
+      "learning_rate": 5e-06,
+      "loss": 0.1301,
+      "step": 14897
+    },
+    {
+      "epoch": 11.214151298456907,
+      "grad_norm": 0.18128998577594757,
+      "learning_rate": 5e-06,
+      "loss": 0.1147,
+      "step": 14898
+    },
+    {
+      "epoch": 11.214904027098232,
+      "grad_norm": 0.18142755329608917,
+      "learning_rate": 5e-06,
+      "loss": 0.1599,
+      "step": 14899
+    },
+    {
+      "epoch": 11.215656755739555,
+      "grad_norm": 0.20231647789478302,
+      "learning_rate": 5e-06,
+      "loss": 0.1243,
+      "step": 14900
+    },
+    {
+      "epoch": 11.21640948438088,
+      "grad_norm": 0.17667371034622192,
+      "learning_rate": 5e-06,
+      "loss": 0.1501,
+      "step": 14901
+    },
+    {
+      "epoch": 11.217162213022206,
+      "grad_norm": 0.1699059009552002,
+      "learning_rate": 5e-06,
+      "loss": 0.1148,
+      "step": 14902
+    },
+    {
+      "epoch": 11.21791494166353,
+      "grad_norm": 0.19872219860553741,
+      "learning_rate": 5e-06,
+      "loss": 0.1309,
+      "step": 14903
+    },
+    {
+      "epoch": 11.218667670304855,
+      "grad_norm": 0.19650769233703613,
+      "learning_rate": 5e-06,
+      "loss": 0.2078,
+      "step": 14904
+    },
+    {
+      "epoch": 11.21942039894618,
+      "grad_norm": 0.16727162897586823,
+      "learning_rate": 5e-06,
+      "loss": 0.1105,
+      "step": 14905
+    },
+    {
+      "epoch": 11.220173127587504,
+      "grad_norm": 0.18666818737983704,
+      "learning_rate": 5e-06,
+      "loss": 0.1302,
+      "step": 14906
+    },
+    {
+      "epoch": 11.22092585622883,
+      "grad_norm": 0.18365789949893951,
+      "learning_rate": 5e-06,
+      "loss": 0.1728,
+      "step": 14907
+    },
+    {
+      "epoch": 11.221678584870155,
+      "grad_norm": 0.17393150925636292,
+      "learning_rate": 5e-06,
+      "loss": 0.1559,
+      "step": 14908
+    },
+    {
+      "epoch": 11.222431313511478,
+      "grad_norm": 0.15818485617637634,
+      "learning_rate": 5e-06,
+      "loss": 0.1025,
+      "step": 14909
+    },
+    {
+      "epoch": 11.223184042152804,
+      "grad_norm": 0.1699003279209137,
+      "learning_rate": 5e-06,
+      "loss": 0.146,
+      "step": 14910
+    },
+    {
+      "epoch": 11.223936770794129,
+      "grad_norm": 0.15623238682746887,
+      "learning_rate": 5e-06,
+      "loss": 0.1159,
+      "step": 14911
+    },
+    {
+      "epoch": 11.224689499435453,
+      "grad_norm": 0.1669670045375824,
+      "learning_rate": 5e-06,
+      "loss": 0.1023,
+      "step": 14912
+    },
+    {
+      "epoch": 11.225442228076778,
+      "grad_norm": 0.16255192458629608,
+      "learning_rate": 5e-06,
+      "loss": 0.1156,
+      "step": 14913
+    },
+    {
+      "epoch": 11.226194956718103,
+      "grad_norm": 0.17024260759353638,
+      "learning_rate": 5e-06,
+      "loss": 0.1347,
+      "step": 14914
+    },
+    {
+      "epoch": 11.226947685359429,
+      "grad_norm": 0.18645474314689636,
+      "learning_rate": 5e-06,
+      "loss": 0.1205,
+      "step": 14915
+    },
+    {
+      "epoch": 11.227700414000752,
+      "grad_norm": 0.17720314860343933,
+      "learning_rate": 5e-06,
+      "loss": 0.2098,
+      "step": 14916
+    },
+    {
+      "epoch": 11.228453142642078,
+      "grad_norm": 0.19273458421230316,
+      "learning_rate": 5e-06,
+      "loss": 0.168,
+      "step": 14917
+    },
+    {
+      "epoch": 11.229205871283403,
+      "grad_norm": 0.2240995615720749,
+      "learning_rate": 5e-06,
+      "loss": 0.1888,
+      "step": 14918
+    },
+    {
+      "epoch": 11.229958599924727,
+      "grad_norm": 0.15409056842327118,
+      "learning_rate": 5e-06,
+      "loss": 0.0826,
+      "step": 14919
+    },
+    {
+      "epoch": 11.230711328566052,
+      "grad_norm": 0.19192633032798767,
+      "learning_rate": 5e-06,
+      "loss": 0.1422,
+      "step": 14920
+    },
+    {
+      "epoch": 11.231464057207377,
+      "grad_norm": 0.17521069943904877,
+      "learning_rate": 5e-06,
+      "loss": 0.1368,
+      "step": 14921
+    },
+    {
+      "epoch": 11.232216785848701,
+      "grad_norm": 0.2080090194940567,
+      "learning_rate": 5e-06,
+      "loss": 0.2087,
+      "step": 14922
+    },
+    {
+      "epoch": 11.232969514490026,
+      "grad_norm": 0.15513311326503754,
+      "learning_rate": 5e-06,
+      "loss": 0.0994,
+      "step": 14923
+    },
+    {
+      "epoch": 11.233722243131352,
+      "grad_norm": 0.16970084607601166,
+      "learning_rate": 5e-06,
+      "loss": 0.0933,
+      "step": 14924
+    },
+    {
+      "epoch": 11.234474971772675,
+      "grad_norm": 0.1898103803396225,
+      "learning_rate": 5e-06,
+      "loss": 0.1415,
+      "step": 14925
+    },
+    {
+      "epoch": 11.235227700414,
+      "grad_norm": 0.22986997663974762,
+      "learning_rate": 5e-06,
+      "loss": 0.2094,
+      "step": 14926
+    },
+    {
+      "epoch": 11.235980429055326,
+      "grad_norm": 0.18323007225990295,
+      "learning_rate": 5e-06,
+      "loss": 0.1532,
+      "step": 14927
+    },
+    {
+      "epoch": 11.23673315769665,
+      "grad_norm": 0.1593945324420929,
+      "learning_rate": 5e-06,
+      "loss": 0.1152,
+      "step": 14928
+    },
+    {
+      "epoch": 11.237485886337975,
+      "grad_norm": 0.18428365886211395,
+      "learning_rate": 5e-06,
+      "loss": 0.1727,
+      "step": 14929
+    },
+    {
+      "epoch": 11.2382386149793,
+      "grad_norm": 0.2174096554517746,
+      "learning_rate": 5e-06,
+      "loss": 0.2055,
+      "step": 14930
+    },
+    {
+      "epoch": 11.238991343620624,
+      "grad_norm": 0.18964442610740662,
+      "learning_rate": 5e-06,
+      "loss": 0.1456,
+      "step": 14931
+    },
+    {
+      "epoch": 11.23974407226195,
+      "grad_norm": 0.18949134647846222,
+      "learning_rate": 5e-06,
+      "loss": 0.1331,
+      "step": 14932
+    },
+    {
+      "epoch": 11.240496800903275,
+      "grad_norm": 0.1851973980665207,
+      "learning_rate": 5e-06,
+      "loss": 0.1564,
+      "step": 14933
+    },
+    {
+      "epoch": 11.2412495295446,
+      "grad_norm": 0.2113349288702011,
+      "learning_rate": 5e-06,
+      "loss": 0.1675,
+      "step": 14934
+    },
+    {
+      "epoch": 11.242002258185924,
+      "grad_norm": 0.1796308308839798,
+      "learning_rate": 5e-06,
+      "loss": 0.1078,
+      "step": 14935
+    },
+    {
+      "epoch": 11.242754986827249,
+      "grad_norm": 0.1922968029975891,
+      "learning_rate": 5e-06,
+      "loss": 0.1802,
+      "step": 14936
+    },
+    {
+      "epoch": 11.243507715468574,
+      "grad_norm": 0.205651193857193,
+      "learning_rate": 5e-06,
+      "loss": 0.1921,
+      "step": 14937
+    },
+    {
+      "epoch": 11.244260444109898,
+      "grad_norm": 0.17744402587413788,
+      "learning_rate": 5e-06,
+      "loss": 0.16,
+      "step": 14938
+    },
+    {
+      "epoch": 11.245013172751223,
+      "grad_norm": 0.1870199292898178,
+      "learning_rate": 5e-06,
+      "loss": 0.1435,
+      "step": 14939
+    },
+    {
+      "epoch": 11.245765901392549,
+      "grad_norm": 0.16941922903060913,
+      "learning_rate": 5e-06,
+      "loss": 0.0966,
+      "step": 14940
+    },
+    {
+      "epoch": 11.246518630033872,
+      "grad_norm": 0.20027337968349457,
+      "learning_rate": 5e-06,
+      "loss": 0.1182,
+      "step": 14941
+    },
+    {
+      "epoch": 11.247271358675198,
+      "grad_norm": 0.17359653115272522,
+      "learning_rate": 5e-06,
+      "loss": 0.1466,
+      "step": 14942
+    },
+    {
+      "epoch": 11.248024087316523,
+      "grad_norm": 0.19150570034980774,
+      "learning_rate": 5e-06,
+      "loss": 0.1091,
+      "step": 14943
+    },
+    {
+      "epoch": 11.248776815957847,
+      "grad_norm": 0.19175341725349426,
+      "learning_rate": 5e-06,
+      "loss": 0.1756,
+      "step": 14944
+    },
+    {
+      "epoch": 11.249529544599172,
+      "grad_norm": 0.20801687240600586,
+      "learning_rate": 5e-06,
+      "loss": 0.166,
+      "step": 14945
+    },
+    {
+      "epoch": 11.250282273240497,
+      "grad_norm": 0.1724909245967865,
+      "learning_rate": 5e-06,
+      "loss": 0.1755,
+      "step": 14946
+    },
+    {
+      "epoch": 11.25103500188182,
+      "grad_norm": 0.19873906672000885,
+      "learning_rate": 5e-06,
+      "loss": 0.1287,
+      "step": 14947
+    },
+    {
+      "epoch": 11.251787730523146,
+      "grad_norm": 0.19745267927646637,
+      "learning_rate": 5e-06,
+      "loss": 0.1889,
+      "step": 14948
+    },
+    {
+      "epoch": 11.252540459164472,
+      "grad_norm": 0.1749572455883026,
+      "learning_rate": 5e-06,
+      "loss": 0.0752,
+      "step": 14949
+    },
+    {
+      "epoch": 11.253293187805795,
+      "grad_norm": 0.1785125732421875,
+      "learning_rate": 5e-06,
+      "loss": 0.1576,
+      "step": 14950
+    },
+    {
+      "epoch": 11.25404591644712,
+      "grad_norm": 0.2056531459093094,
+      "learning_rate": 5e-06,
+      "loss": 0.2214,
+      "step": 14951
+    },
+    {
+      "epoch": 11.254798645088446,
+      "grad_norm": 0.20328469574451447,
+      "learning_rate": 5e-06,
+      "loss": 0.1301,
+      "step": 14952
+    },
+    {
+      "epoch": 11.255551373729771,
+      "grad_norm": 0.16046443581581116,
+      "learning_rate": 5e-06,
+      "loss": 0.0946,
+      "step": 14953
+    },
+    {
+      "epoch": 11.256304102371095,
+      "grad_norm": 0.1748947650194168,
+      "learning_rate": 5e-06,
+      "loss": 0.1506,
+      "step": 14954
+    },
+    {
+      "epoch": 11.25705683101242,
+      "grad_norm": 0.20890913903713226,
+      "learning_rate": 5e-06,
+      "loss": 0.1141,
+      "step": 14955
+    },
+    {
+      "epoch": 11.257809559653746,
+      "grad_norm": 0.1695561707019806,
+      "learning_rate": 5e-06,
+      "loss": 0.0788,
+      "step": 14956
+    },
+    {
+      "epoch": 11.258562288295069,
+      "grad_norm": 0.1707061231136322,
+      "learning_rate": 5e-06,
+      "loss": 0.1532,
+      "step": 14957
+    },
+    {
+      "epoch": 11.259315016936394,
+      "grad_norm": 0.17715346813201904,
+      "learning_rate": 5e-06,
+      "loss": 0.1875,
+      "step": 14958
+    },
+    {
+      "epoch": 11.26006774557772,
+      "grad_norm": 0.1668746918439865,
+      "learning_rate": 5e-06,
+      "loss": 0.1533,
+      "step": 14959
+    },
+    {
+      "epoch": 11.260820474219043,
+      "grad_norm": 0.16586287319660187,
+      "learning_rate": 5e-06,
+      "loss": 0.076,
+      "step": 14960
+    },
+    {
+      "epoch": 11.261573202860369,
+      "grad_norm": 0.21081513166427612,
+      "learning_rate": 5e-06,
+      "loss": 0.1139,
+      "step": 14961
+    },
+    {
+      "epoch": 11.262325931501694,
+      "grad_norm": 0.19864341616630554,
+      "learning_rate": 5e-06,
+      "loss": 0.1102,
+      "step": 14962
+    },
+    {
+      "epoch": 11.263078660143018,
+      "grad_norm": 0.17366968095302582,
+      "learning_rate": 5e-06,
+      "loss": 0.1414,
+      "step": 14963
+    },
+    {
+      "epoch": 11.263831388784343,
+      "grad_norm": 0.1947343349456787,
+      "learning_rate": 5e-06,
+      "loss": 0.1603,
+      "step": 14964
+    },
+    {
+      "epoch": 11.264584117425668,
+      "grad_norm": 0.21239379048347473,
+      "learning_rate": 5e-06,
+      "loss": 0.1629,
+      "step": 14965
+    },
+    {
+      "epoch": 11.265336846066992,
+      "grad_norm": 0.21441568434238434,
+      "learning_rate": 5e-06,
+      "loss": 0.1536,
+      "step": 14966
+    },
+    {
+      "epoch": 11.266089574708317,
+      "grad_norm": 0.18264716863632202,
+      "learning_rate": 5e-06,
+      "loss": 0.1189,
+      "step": 14967
+    },
+    {
+      "epoch": 11.266842303349643,
+      "grad_norm": 0.17720608413219452,
+      "learning_rate": 5e-06,
+      "loss": 0.182,
+      "step": 14968
+    },
+    {
+      "epoch": 11.267595031990968,
+      "grad_norm": 0.21148650348186493,
+      "learning_rate": 5e-06,
+      "loss": 0.164,
+      "step": 14969
+    },
+    {
+      "epoch": 11.268347760632292,
+      "grad_norm": 0.17817366123199463,
+      "learning_rate": 5e-06,
+      "loss": 0.1206,
+      "step": 14970
+    },
+    {
+      "epoch": 11.269100489273617,
+      "grad_norm": 0.18789704144001007,
+      "learning_rate": 5e-06,
+      "loss": 0.1283,
+      "step": 14971
+    },
+    {
+      "epoch": 11.269853217914942,
+      "grad_norm": 0.17486625909805298,
+      "learning_rate": 5e-06,
+      "loss": 0.0972,
+      "step": 14972
+    },
+    {
+      "epoch": 11.270605946556266,
+      "grad_norm": 0.18096482753753662,
+      "learning_rate": 5e-06,
+      "loss": 0.1517,
+      "step": 14973
+    },
+    {
+      "epoch": 11.271358675197591,
+      "grad_norm": 0.17482225596904755,
+      "learning_rate": 5e-06,
+      "loss": 0.1111,
+      "step": 14974
+    },
+    {
+      "epoch": 11.272111403838917,
+      "grad_norm": 0.19433853030204773,
+      "learning_rate": 5e-06,
+      "loss": 0.1863,
+      "step": 14975
+    },
+    {
+      "epoch": 11.27286413248024,
+      "grad_norm": 0.17639221251010895,
+      "learning_rate": 5e-06,
+      "loss": 0.1048,
+      "step": 14976
+    },
+    {
+      "epoch": 11.273616861121566,
+      "grad_norm": 0.18675243854522705,
+      "learning_rate": 5e-06,
+      "loss": 0.1887,
+      "step": 14977
+    },
+    {
+      "epoch": 11.274369589762891,
+      "grad_norm": 0.19055995345115662,
+      "learning_rate": 5e-06,
+      "loss": 0.1815,
+      "step": 14978
+    },
+    {
+      "epoch": 11.275122318404215,
+      "grad_norm": 0.16495707631111145,
+      "learning_rate": 5e-06,
+      "loss": 0.1073,
+      "step": 14979
+    },
+    {
+      "epoch": 11.27587504704554,
+      "grad_norm": 0.17271405458450317,
+      "learning_rate": 5e-06,
+      "loss": 0.0949,
+      "step": 14980
+    },
+    {
+      "epoch": 11.276627775686865,
+      "grad_norm": 0.2195117324590683,
+      "learning_rate": 5e-06,
+      "loss": 0.1375,
+      "step": 14981
+    },
+    {
+      "epoch": 11.277380504328189,
+      "grad_norm": 0.18514885008335114,
+      "learning_rate": 5e-06,
+      "loss": 0.1682,
+      "step": 14982
+    },
+    {
+      "epoch": 11.278133232969514,
+      "grad_norm": 0.1788724958896637,
+      "learning_rate": 5e-06,
+      "loss": 0.1138,
+      "step": 14983
+    },
+    {
+      "epoch": 11.27888596161084,
+      "grad_norm": 0.18887434899806976,
+      "learning_rate": 5e-06,
+      "loss": 0.2142,
+      "step": 14984
+    },
+    {
+      "epoch": 11.279638690252163,
+      "grad_norm": 0.1876523494720459,
+      "learning_rate": 5e-06,
+      "loss": 0.1879,
+      "step": 14985
+    },
+    {
+      "epoch": 11.280391418893489,
+      "grad_norm": 0.20100463926792145,
+      "learning_rate": 5e-06,
+      "loss": 0.1872,
+      "step": 14986
+    },
+    {
+      "epoch": 11.281144147534814,
+      "grad_norm": 0.15760323405265808,
+      "learning_rate": 5e-06,
+      "loss": 0.179,
+      "step": 14987
+    },
+    {
+      "epoch": 11.28189687617614,
+      "grad_norm": 0.20327183604240417,
+      "learning_rate": 5e-06,
+      "loss": 0.1575,
+      "step": 14988
+    },
+    {
+      "epoch": 11.282649604817463,
+      "grad_norm": 0.2047901451587677,
+      "learning_rate": 5e-06,
+      "loss": 0.1168,
+      "step": 14989
+    },
+    {
+      "epoch": 11.283402333458788,
+      "grad_norm": 0.20007163286209106,
+      "learning_rate": 5e-06,
+      "loss": 0.1739,
+      "step": 14990
+    },
+    {
+      "epoch": 11.284155062100114,
+      "grad_norm": 0.2349494993686676,
+      "learning_rate": 5e-06,
+      "loss": 0.1752,
+      "step": 14991
+    },
+    {
+      "epoch": 11.284907790741437,
+      "grad_norm": 0.18426214158535004,
+      "learning_rate": 5e-06,
+      "loss": 0.1949,
+      "step": 14992
+    },
+    {
+      "epoch": 11.285660519382763,
+      "grad_norm": 0.16646084189414978,
+      "learning_rate": 5e-06,
+      "loss": 0.1116,
+      "step": 14993
+    },
+    {
+      "epoch": 11.286413248024088,
+      "grad_norm": 0.1696571409702301,
+      "learning_rate": 5e-06,
+      "loss": 0.1458,
+      "step": 14994
+    },
+    {
+      "epoch": 11.287165976665412,
+      "grad_norm": 0.18156984448432922,
+      "learning_rate": 5e-06,
+      "loss": 0.1184,
+      "step": 14995
+    },
+    {
+      "epoch": 11.287918705306737,
+      "grad_norm": 0.16824743151664734,
+      "learning_rate": 5e-06,
+      "loss": 0.0841,
+      "step": 14996
+    },
+    {
+      "epoch": 11.288671433948062,
+      "grad_norm": 0.18768930435180664,
+      "learning_rate": 5e-06,
+      "loss": 0.2001,
+      "step": 14997
+    },
+    {
+      "epoch": 11.289424162589386,
+      "grad_norm": 0.16364774107933044,
+      "learning_rate": 5e-06,
+      "loss": 0.1684,
+      "step": 14998
+    },
+    {
+      "epoch": 11.290176891230711,
+      "grad_norm": 0.19122710824012756,
+      "learning_rate": 5e-06,
+      "loss": 0.128,
+      "step": 14999
+    },
+    {
+      "epoch": 11.290929619872037,
+      "grad_norm": 0.19520621001720428,
+      "learning_rate": 5e-06,
+      "loss": 0.1139,
+      "step": 15000
+    },
+    {
+      "epoch": 11.290929619872037,
+      "eval_loss": 0.18596623837947845,
+      "eval_runtime": 452.7025,
+      "eval_samples_per_second": 21.266,
+      "eval_steps_per_second": 0.665,
+      "step": 15000
+    },
+    {
+      "epoch": 11.29168234851336,
+      "grad_norm": 0.16837935149669647,
+      "learning_rate": 5e-06,
+      "loss": 0.1275,
+      "step": 15001
+    },
+    {
+      "epoch": 11.292435077154686,
+      "grad_norm": 0.28820323944091797,
+      "learning_rate": 5e-06,
+      "loss": 0.2027,
+      "step": 15002
+    },
+    {
+      "epoch": 11.293187805796011,
+      "grad_norm": 0.17704153060913086,
+      "learning_rate": 5e-06,
+      "loss": 0.1362,
+      "step": 15003
+    },
+    {
+      "epoch": 11.293940534437336,
+      "grad_norm": 0.18403461575508118,
+      "learning_rate": 5e-06,
+      "loss": 0.1277,
+      "step": 15004
+    },
+    {
+      "epoch": 11.29469326307866,
+      "grad_norm": 0.20933157205581665,
+      "learning_rate": 5e-06,
+      "loss": 0.1417,
+      "step": 15005
+    },
+    {
+      "epoch": 11.295445991719985,
+      "grad_norm": 0.17760062217712402,
+      "learning_rate": 5e-06,
+      "loss": 0.0907,
+      "step": 15006
+    },
+    {
+      "epoch": 11.29619872036131,
+      "grad_norm": 0.1881420761346817,
+      "learning_rate": 5e-06,
+      "loss": 0.0887,
+      "step": 15007
+    },
+    {
+      "epoch": 11.296951449002634,
+      "grad_norm": 0.17742136120796204,
+      "learning_rate": 5e-06,
+      "loss": 0.1124,
+      "step": 15008
+    },
+    {
+      "epoch": 11.29770417764396,
+      "grad_norm": 0.1597941815853119,
+      "learning_rate": 5e-06,
+      "loss": 0.1644,
+      "step": 15009
+    },
+    {
+      "epoch": 11.298456906285285,
+      "grad_norm": 0.20276004076004028,
+      "learning_rate": 5e-06,
+      "loss": 0.2147,
+      "step": 15010
+    },
+    {
+      "epoch": 11.299209634926608,
+      "grad_norm": 0.18984192609786987,
+      "learning_rate": 5e-06,
+      "loss": 0.2658,
+      "step": 15011
+    },
+    {
+      "epoch": 11.299962363567934,
+      "grad_norm": 0.18264015018939972,
+      "learning_rate": 5e-06,
+      "loss": 0.1182,
+      "step": 15012
+    },
+    {
+      "epoch": 11.30071509220926,
+      "grad_norm": 0.1967144012451172,
+      "learning_rate": 5e-06,
+      "loss": 0.1385,
+      "step": 15013
+    },
+    {
+      "epoch": 11.301467820850583,
+      "grad_norm": 0.16850410401821136,
+      "learning_rate": 5e-06,
+      "loss": 0.1467,
+      "step": 15014
+    },
+    {
+      "epoch": 11.302220549491908,
+      "grad_norm": 0.19638095796108246,
+      "learning_rate": 5e-06,
+      "loss": 0.1793,
+      "step": 15015
+    },
+    {
+      "epoch": 11.302973278133233,
+      "grad_norm": 0.17581182718276978,
+      "learning_rate": 5e-06,
+      "loss": 0.1497,
+      "step": 15016
+    },
+    {
+      "epoch": 11.303726006774557,
+      "grad_norm": 0.1778106689453125,
+      "learning_rate": 5e-06,
+      "loss": 0.2053,
+      "step": 15017
+    },
+    {
+      "epoch": 11.304478735415882,
+      "grad_norm": 0.1706477254629135,
+      "learning_rate": 5e-06,
+      "loss": 0.1414,
+      "step": 15018
+    },
+    {
+      "epoch": 11.305231464057208,
+      "grad_norm": 0.19102172553539276,
+      "learning_rate": 5e-06,
+      "loss": 0.1748,
+      "step": 15019
+    },
+    {
+      "epoch": 11.305984192698531,
+      "grad_norm": 0.17424669861793518,
+      "learning_rate": 5e-06,
+      "loss": 0.0976,
+      "step": 15020
+    },
+    {
+      "epoch": 11.306736921339857,
+      "grad_norm": 0.19761112332344055,
+      "learning_rate": 5e-06,
+      "loss": 0.1781,
+      "step": 15021
+    },
+    {
+      "epoch": 11.307489649981182,
+      "grad_norm": 0.18848994374275208,
+      "learning_rate": 5e-06,
+      "loss": 0.1298,
+      "step": 15022
+    },
+    {
+      "epoch": 11.308242378622506,
+      "grad_norm": 0.18608836829662323,
+      "learning_rate": 5e-06,
+      "loss": 0.148,
+      "step": 15023
+    },
+    {
+      "epoch": 11.308995107263831,
+      "grad_norm": 0.17123989760875702,
+      "learning_rate": 5e-06,
+      "loss": 0.1117,
+      "step": 15024
+    },
+    {
+      "epoch": 11.309747835905156,
+      "grad_norm": 0.17374210059642792,
+      "learning_rate": 5e-06,
+      "loss": 0.1484,
+      "step": 15025
+    },
+    {
+      "epoch": 11.310500564546482,
+      "grad_norm": 0.18598882853984833,
+      "learning_rate": 5e-06,
+      "loss": 0.2317,
+      "step": 15026
+    },
+    {
+      "epoch": 11.311253293187805,
+      "grad_norm": 0.1987432837486267,
+      "learning_rate": 5e-06,
+      "loss": 0.1606,
+      "step": 15027
+    },
+    {
+      "epoch": 11.31200602182913,
+      "grad_norm": 0.20942755043506622,
+      "learning_rate": 5e-06,
+      "loss": 0.1579,
+      "step": 15028
+    },
+    {
+      "epoch": 11.312758750470456,
+      "grad_norm": 0.16522537171840668,
+      "learning_rate": 5e-06,
+      "loss": 0.1376,
+      "step": 15029
+    },
+    {
+      "epoch": 11.31351147911178,
+      "grad_norm": 0.1471552848815918,
+      "learning_rate": 5e-06,
+      "loss": 0.0507,
+      "step": 15030
+    },
+    {
+      "epoch": 11.314264207753105,
+      "grad_norm": 0.18673619627952576,
+      "learning_rate": 5e-06,
+      "loss": 0.1282,
+      "step": 15031
+    },
+    {
+      "epoch": 11.31501693639443,
+      "grad_norm": 0.18051984906196594,
+      "learning_rate": 5e-06,
+      "loss": 0.1438,
+      "step": 15032
+    },
+    {
+      "epoch": 11.315769665035754,
+      "grad_norm": 0.17510734498500824,
+      "learning_rate": 5e-06,
+      "loss": 0.1312,
+      "step": 15033
+    },
+    {
+      "epoch": 11.31652239367708,
+      "grad_norm": 0.18032442033290863,
+      "learning_rate": 5e-06,
+      "loss": 0.1409,
+      "step": 15034
+    },
+    {
+      "epoch": 11.317275122318405,
+      "grad_norm": 0.17971666157245636,
+      "learning_rate": 5e-06,
+      "loss": 0.0815,
+      "step": 15035
+    },
+    {
+      "epoch": 11.318027850959728,
+      "grad_norm": 0.16304822266101837,
+      "learning_rate": 5e-06,
+      "loss": 0.1349,
+      "step": 15036
+    },
+    {
+      "epoch": 11.318780579601054,
+      "grad_norm": 0.1777772605419159,
+      "learning_rate": 5e-06,
+      "loss": 0.1646,
+      "step": 15037
+    },
+    {
+      "epoch": 11.319533308242379,
+      "grad_norm": 0.1969400942325592,
+      "learning_rate": 5e-06,
+      "loss": 0.137,
+      "step": 15038
+    },
+    {
+      "epoch": 11.320286036883703,
+      "grad_norm": 0.18909378349781036,
+      "learning_rate": 5e-06,
+      "loss": 0.1331,
+      "step": 15039
+    },
+    {
+      "epoch": 11.321038765525028,
+      "grad_norm": 0.25971004366874695,
+      "learning_rate": 5e-06,
+      "loss": 0.1454,
+      "step": 15040
+    },
+    {
+      "epoch": 11.321791494166353,
+      "grad_norm": 0.1839953362941742,
+      "learning_rate": 5e-06,
+      "loss": 0.1406,
+      "step": 15041
+    },
+    {
+      "epoch": 11.322544222807679,
+      "grad_norm": 0.18701794743537903,
+      "learning_rate": 5e-06,
+      "loss": 0.1284,
+      "step": 15042
+    },
+    {
+      "epoch": 11.323296951449002,
+      "grad_norm": 0.2126777619123459,
+      "learning_rate": 5e-06,
+      "loss": 0.1278,
+      "step": 15043
+    },
+    {
+      "epoch": 11.324049680090328,
+      "grad_norm": 0.1685115098953247,
+      "learning_rate": 5e-06,
+      "loss": 0.1338,
+      "step": 15044
+    },
+    {
+      "epoch": 11.324802408731653,
+      "grad_norm": 0.1956368386745453,
+      "learning_rate": 5e-06,
+      "loss": 0.1578,
+      "step": 15045
+    },
+    {
+      "epoch": 11.325555137372977,
+      "grad_norm": 0.16715551912784576,
+      "learning_rate": 5e-06,
+      "loss": 0.107,
+      "step": 15046
+    },
+    {
+      "epoch": 11.326307866014302,
+      "grad_norm": 0.19278118014335632,
+      "learning_rate": 5e-06,
+      "loss": 0.1824,
+      "step": 15047
+    },
+    {
+      "epoch": 11.327060594655627,
+      "grad_norm": 0.18524974584579468,
+      "learning_rate": 5e-06,
+      "loss": 0.0877,
+      "step": 15048
+    },
+    {
+      "epoch": 11.327813323296951,
+      "grad_norm": 0.2038186639547348,
+      "learning_rate": 5e-06,
+      "loss": 0.1628,
+      "step": 15049
+    },
+    {
+      "epoch": 11.328566051938276,
+      "grad_norm": 0.17597299814224243,
+      "learning_rate": 5e-06,
+      "loss": 0.0697,
+      "step": 15050
+    },
+    {
+      "epoch": 11.329318780579602,
+      "grad_norm": 0.18205031752586365,
+      "learning_rate": 5e-06,
+      "loss": 0.143,
+      "step": 15051
+    },
+    {
+      "epoch": 11.330071509220925,
+      "grad_norm": 0.19399188458919525,
+      "learning_rate": 5e-06,
+      "loss": 0.12,
+      "step": 15052
+    },
+    {
+      "epoch": 11.33082423786225,
+      "grad_norm": 0.16592323780059814,
+      "learning_rate": 5e-06,
+      "loss": 0.0776,
+      "step": 15053
+    },
+    {
+      "epoch": 11.331576966503576,
+      "grad_norm": 0.1655736267566681,
+      "learning_rate": 5e-06,
+      "loss": 0.1534,
+      "step": 15054
+    },
+    {
+      "epoch": 11.3323296951449,
+      "grad_norm": 0.1851305216550827,
+      "learning_rate": 5e-06,
+      "loss": 0.1647,
+      "step": 15055
+    },
+    {
+      "epoch": 11.333082423786225,
+      "grad_norm": 0.15674932301044464,
+      "learning_rate": 5e-06,
+      "loss": 0.16,
+      "step": 15056
+    },
+    {
+      "epoch": 11.33383515242755,
+      "grad_norm": 0.1676856130361557,
+      "learning_rate": 5e-06,
+      "loss": 0.1203,
+      "step": 15057
+    },
+    {
+      "epoch": 11.334587881068874,
+      "grad_norm": 0.17766782641410828,
+      "learning_rate": 5e-06,
+      "loss": 0.1732,
+      "step": 15058
+    },
+    {
+      "epoch": 11.3353406097102,
+      "grad_norm": 0.15851709246635437,
+      "learning_rate": 5e-06,
+      "loss": 0.1809,
+      "step": 15059
+    },
+    {
+      "epoch": 11.336093338351525,
+      "grad_norm": 0.174681156873703,
+      "learning_rate": 5e-06,
+      "loss": 0.1364,
+      "step": 15060
+    },
+    {
+      "epoch": 11.33684606699285,
+      "grad_norm": 0.2067001909017563,
+      "learning_rate": 5e-06,
+      "loss": 0.184,
+      "step": 15061
+    },
+    {
+      "epoch": 11.337598795634173,
+      "grad_norm": 0.18242330849170685,
+      "learning_rate": 5e-06,
+      "loss": 0.1547,
+      "step": 15062
+    },
+    {
+      "epoch": 11.338351524275499,
+      "grad_norm": 0.19805799424648285,
+      "learning_rate": 5e-06,
+      "loss": 0.203,
+      "step": 15063
+    },
+    {
+      "epoch": 11.339104252916824,
+      "grad_norm": 0.17603465914726257,
+      "learning_rate": 5e-06,
+      "loss": 0.1386,
+      "step": 15064
+    },
+    {
+      "epoch": 11.339856981558148,
+      "grad_norm": 0.19910475611686707,
+      "learning_rate": 5e-06,
+      "loss": 0.1292,
+      "step": 15065
+    },
+    {
+      "epoch": 11.340609710199473,
+      "grad_norm": 0.18487416207790375,
+      "learning_rate": 5e-06,
+      "loss": 0.139,
+      "step": 15066
+    },
+    {
+      "epoch": 11.341362438840799,
+      "grad_norm": 0.22432343661785126,
+      "learning_rate": 5e-06,
+      "loss": 0.206,
+      "step": 15067
+    },
+    {
+      "epoch": 11.342115167482122,
+      "grad_norm": 0.1903543323278427,
+      "learning_rate": 5e-06,
+      "loss": 0.155,
+      "step": 15068
+    },
+    {
+      "epoch": 11.342867896123447,
+      "grad_norm": 0.16379080712795258,
+      "learning_rate": 5e-06,
+      "loss": 0.1702,
+      "step": 15069
+    },
+    {
+      "epoch": 11.343620624764773,
+      "grad_norm": 0.17029471695423126,
+      "learning_rate": 5e-06,
+      "loss": 0.2152,
+      "step": 15070
+    },
+    {
+      "epoch": 11.344373353406096,
+      "grad_norm": 0.1914007067680359,
+      "learning_rate": 5e-06,
+      "loss": 0.1761,
+      "step": 15071
+    },
+    {
+      "epoch": 11.345126082047422,
+      "grad_norm": 0.15714678168296814,
+      "learning_rate": 5e-06,
+      "loss": 0.0974,
+      "step": 15072
+    },
+    {
+      "epoch": 11.345878810688747,
+      "grad_norm": 0.1823018193244934,
+      "learning_rate": 5e-06,
+      "loss": 0.1717,
+      "step": 15073
+    },
+    {
+      "epoch": 11.34663153933007,
+      "grad_norm": 0.17209266126155853,
+      "learning_rate": 5e-06,
+      "loss": 0.1409,
+      "step": 15074
+    },
+    {
+      "epoch": 11.347384267971396,
+      "grad_norm": 0.17569929361343384,
+      "learning_rate": 5e-06,
+      "loss": 0.2037,
+      "step": 15075
+    },
+    {
+      "epoch": 11.348136996612721,
+      "grad_norm": 0.18091587722301483,
+      "learning_rate": 5e-06,
+      "loss": 0.0936,
+      "step": 15076
+    },
+    {
+      "epoch": 11.348889725254047,
+      "grad_norm": 0.16457347571849823,
+      "learning_rate": 5e-06,
+      "loss": 0.1405,
+      "step": 15077
+    },
+    {
+      "epoch": 11.34964245389537,
+      "grad_norm": 0.20130757987499237,
+      "learning_rate": 5e-06,
+      "loss": 0.1249,
+      "step": 15078
+    },
+    {
+      "epoch": 11.350395182536696,
+      "grad_norm": 0.19006864726543427,
+      "learning_rate": 5e-06,
+      "loss": 0.1726,
+      "step": 15079
+    },
+    {
+      "epoch": 11.351147911178021,
+      "grad_norm": 0.19786526262760162,
+      "learning_rate": 5e-06,
+      "loss": 0.1622,
+      "step": 15080
+    },
+    {
+      "epoch": 11.351900639819345,
+      "grad_norm": 0.17777101695537567,
+      "learning_rate": 5e-06,
+      "loss": 0.1627,
+      "step": 15081
+    },
+    {
+      "epoch": 11.35265336846067,
+      "grad_norm": 0.19773589074611664,
+      "learning_rate": 5e-06,
+      "loss": 0.2198,
+      "step": 15082
+    },
+    {
+      "epoch": 11.353406097101995,
+      "grad_norm": 0.1853761523962021,
+      "learning_rate": 5e-06,
+      "loss": 0.1651,
+      "step": 15083
+    },
+    {
+      "epoch": 11.354158825743319,
+      "grad_norm": 0.16510865092277527,
+      "learning_rate": 5e-06,
+      "loss": 0.1344,
+      "step": 15084
+    },
+    {
+      "epoch": 11.354911554384644,
+      "grad_norm": 0.17784161865711212,
+      "learning_rate": 5e-06,
+      "loss": 0.0865,
+      "step": 15085
+    },
+    {
+      "epoch": 11.35566428302597,
+      "grad_norm": 0.20078599452972412,
+      "learning_rate": 5e-06,
+      "loss": 0.2007,
+      "step": 15086
+    },
+    {
+      "epoch": 11.356417011667293,
+      "grad_norm": 0.1767657995223999,
+      "learning_rate": 5e-06,
+      "loss": 0.1671,
+      "step": 15087
+    },
+    {
+      "epoch": 11.357169740308619,
+      "grad_norm": 0.19230806827545166,
+      "learning_rate": 5e-06,
+      "loss": 0.142,
+      "step": 15088
+    },
+    {
+      "epoch": 11.357922468949944,
+      "grad_norm": 0.17567181587219238,
+      "learning_rate": 5e-06,
+      "loss": 0.1336,
+      "step": 15089
+    },
+    {
+      "epoch": 11.358675197591268,
+      "grad_norm": 0.1692219078540802,
+      "learning_rate": 5e-06,
+      "loss": 0.1479,
+      "step": 15090
+    },
+    {
+      "epoch": 11.359427926232593,
+      "grad_norm": 0.19515249133110046,
+      "learning_rate": 5e-06,
+      "loss": 0.1725,
+      "step": 15091
+    },
+    {
+      "epoch": 11.360180654873918,
+      "grad_norm": 0.210249125957489,
+      "learning_rate": 5e-06,
+      "loss": 0.2002,
+      "step": 15092
+    },
+    {
+      "epoch": 11.360933383515242,
+      "grad_norm": 0.19336655735969543,
+      "learning_rate": 5e-06,
+      "loss": 0.1828,
+      "step": 15093
+    },
+    {
+      "epoch": 11.361686112156567,
+      "grad_norm": 0.17939543724060059,
+      "learning_rate": 5e-06,
+      "loss": 0.1219,
+      "step": 15094
+    },
+    {
+      "epoch": 11.362438840797893,
+      "grad_norm": 0.19464604556560516,
+      "learning_rate": 5e-06,
+      "loss": 0.1751,
+      "step": 15095
+    },
+    {
+      "epoch": 11.363191569439216,
+      "grad_norm": 0.1851925551891327,
+      "learning_rate": 5e-06,
+      "loss": 0.15,
+      "step": 15096
+    },
+    {
+      "epoch": 11.363944298080542,
+      "grad_norm": 0.23305077850818634,
+      "learning_rate": 5e-06,
+      "loss": 0.204,
+      "step": 15097
+    },
+    {
+      "epoch": 11.364697026721867,
+      "grad_norm": 0.1858210414648056,
+      "learning_rate": 5e-06,
+      "loss": 0.139,
+      "step": 15098
+    },
+    {
+      "epoch": 11.365449755363192,
+      "grad_norm": 0.29778534173965454,
+      "learning_rate": 5e-06,
+      "loss": 0.1103,
+      "step": 15099
+    },
+    {
+      "epoch": 11.366202484004516,
+      "grad_norm": 0.20623186230659485,
+      "learning_rate": 5e-06,
+      "loss": 0.1304,
+      "step": 15100
+    },
+    {
+      "epoch": 11.366955212645841,
+      "grad_norm": 0.21118873357772827,
+      "learning_rate": 5e-06,
+      "loss": 0.1359,
+      "step": 15101
+    },
+    {
+      "epoch": 11.367707941287167,
+      "grad_norm": 0.21927309036254883,
+      "learning_rate": 5e-06,
+      "loss": 0.1962,
+      "step": 15102
+    },
+    {
+      "epoch": 11.36846066992849,
+      "grad_norm": 0.1795198619365692,
+      "learning_rate": 5e-06,
+      "loss": 0.1258,
+      "step": 15103
+    },
+    {
+      "epoch": 11.369213398569816,
+      "grad_norm": 0.17400717735290527,
+      "learning_rate": 5e-06,
+      "loss": 0.1345,
+      "step": 15104
+    },
+    {
+      "epoch": 11.369966127211141,
+      "grad_norm": 0.18235079944133759,
+      "learning_rate": 5e-06,
+      "loss": 0.1778,
+      "step": 15105
+    },
+    {
+      "epoch": 11.370718855852465,
+      "grad_norm": 0.18174009025096893,
+      "learning_rate": 5e-06,
+      "loss": 0.1765,
+      "step": 15106
+    },
+    {
+      "epoch": 11.37147158449379,
+      "grad_norm": 0.18080852925777435,
+      "learning_rate": 5e-06,
+      "loss": 0.129,
+      "step": 15107
+    },
+    {
+      "epoch": 11.372224313135115,
+      "grad_norm": 0.17598354816436768,
+      "learning_rate": 5e-06,
+      "loss": 0.1483,
+      "step": 15108
+    },
+    {
+      "epoch": 11.372977041776439,
+      "grad_norm": 0.1811891347169876,
+      "learning_rate": 5e-06,
+      "loss": 0.1236,
+      "step": 15109
+    },
+    {
+      "epoch": 11.373729770417764,
+      "grad_norm": 0.17206525802612305,
+      "learning_rate": 5e-06,
+      "loss": 0.1576,
+      "step": 15110
+    },
+    {
+      "epoch": 11.37448249905909,
+      "grad_norm": 0.18468911945819855,
+      "learning_rate": 5e-06,
+      "loss": 0.1135,
+      "step": 15111
+    },
+    {
+      "epoch": 11.375235227700413,
+      "grad_norm": 0.2046259641647339,
+      "learning_rate": 5e-06,
+      "loss": 0.1251,
+      "step": 15112
+    },
+    {
+      "epoch": 11.375987956341739,
+      "grad_norm": 0.1777375191450119,
+      "learning_rate": 5e-06,
+      "loss": 0.2025,
+      "step": 15113
+    },
+    {
+      "epoch": 11.376740684983064,
+      "grad_norm": 0.19654178619384766,
+      "learning_rate": 5e-06,
+      "loss": 0.1104,
+      "step": 15114
+    },
+    {
+      "epoch": 11.37749341362439,
+      "grad_norm": 0.19765779376029968,
+      "learning_rate": 5e-06,
+      "loss": 0.1752,
+      "step": 15115
+    },
+    {
+      "epoch": 11.378246142265713,
+      "grad_norm": 0.17084521055221558,
+      "learning_rate": 5e-06,
+      "loss": 0.1198,
+      "step": 15116
+    },
+    {
+      "epoch": 11.378998870907038,
+      "grad_norm": 0.18602770566940308,
+      "learning_rate": 5e-06,
+      "loss": 0.2578,
+      "step": 15117
+    },
+    {
+      "epoch": 11.379751599548364,
+      "grad_norm": 0.18878178298473358,
+      "learning_rate": 5e-06,
+      "loss": 0.1041,
+      "step": 15118
+    },
+    {
+      "epoch": 11.380504328189687,
+      "grad_norm": 0.16515788435935974,
+      "learning_rate": 5e-06,
+      "loss": 0.1481,
+      "step": 15119
+    },
+    {
+      "epoch": 11.381257056831013,
+      "grad_norm": 0.18691888451576233,
+      "learning_rate": 5e-06,
+      "loss": 0.1529,
+      "step": 15120
+    },
+    {
+      "epoch": 11.382009785472338,
+      "grad_norm": 0.1960705667734146,
+      "learning_rate": 5e-06,
+      "loss": 0.1192,
+      "step": 15121
+    },
+    {
+      "epoch": 11.382762514113661,
+      "grad_norm": 0.21251945197582245,
+      "learning_rate": 5e-06,
+      "loss": 0.1029,
+      "step": 15122
+    },
+    {
+      "epoch": 11.383515242754987,
+      "grad_norm": 0.18491223454475403,
+      "learning_rate": 5e-06,
+      "loss": 0.1083,
+      "step": 15123
+    },
+    {
+      "epoch": 11.384267971396312,
+      "grad_norm": 0.1702854037284851,
+      "learning_rate": 5e-06,
+      "loss": 0.0966,
+      "step": 15124
+    },
+    {
+      "epoch": 11.385020700037636,
+      "grad_norm": 0.19505473971366882,
+      "learning_rate": 5e-06,
+      "loss": 0.2931,
+      "step": 15125
+    },
+    {
+      "epoch": 11.385773428678961,
+      "grad_norm": 0.17792576551437378,
+      "learning_rate": 5e-06,
+      "loss": 0.1454,
+      "step": 15126
+    },
+    {
+      "epoch": 11.386526157320287,
+      "grad_norm": 0.15717937052249908,
+      "learning_rate": 5e-06,
+      "loss": 0.0838,
+      "step": 15127
+    },
+    {
+      "epoch": 11.38727888596161,
+      "grad_norm": 0.19245149195194244,
+      "learning_rate": 5e-06,
+      "loss": 0.2183,
+      "step": 15128
+    },
+    {
+      "epoch": 11.388031614602935,
+      "grad_norm": 0.1922212839126587,
+      "learning_rate": 5e-06,
+      "loss": 0.1459,
+      "step": 15129
+    },
+    {
+      "epoch": 11.38878434324426,
+      "grad_norm": 0.17715758085250854,
+      "learning_rate": 5e-06,
+      "loss": 0.1362,
+      "step": 15130
+    },
+    {
+      "epoch": 11.389537071885584,
+      "grad_norm": 0.18386803567409515,
+      "learning_rate": 5e-06,
+      "loss": 0.1044,
+      "step": 15131
+    },
+    {
+      "epoch": 11.39028980052691,
+      "grad_norm": 0.2365635484457016,
+      "learning_rate": 5e-06,
+      "loss": 0.2032,
+      "step": 15132
+    },
+    {
+      "epoch": 11.391042529168235,
+      "grad_norm": 0.19914063811302185,
+      "learning_rate": 5e-06,
+      "loss": 0.123,
+      "step": 15133
+    },
+    {
+      "epoch": 11.39179525780956,
+      "grad_norm": 0.1584664285182953,
+      "learning_rate": 5e-06,
+      "loss": 0.0621,
+      "step": 15134
+    },
+    {
+      "epoch": 11.392547986450884,
+      "grad_norm": 0.19291576743125916,
+      "learning_rate": 5e-06,
+      "loss": 0.195,
+      "step": 15135
+    },
+    {
+      "epoch": 11.39330071509221,
+      "grad_norm": 0.19710887968540192,
+      "learning_rate": 5e-06,
+      "loss": 0.1995,
+      "step": 15136
+    },
+    {
+      "epoch": 11.394053443733535,
+      "grad_norm": 0.18053682148456573,
+      "learning_rate": 5e-06,
+      "loss": 0.1666,
+      "step": 15137
+    },
+    {
+      "epoch": 11.394806172374858,
+      "grad_norm": 0.1647602766752243,
+      "learning_rate": 5e-06,
+      "loss": 0.1042,
+      "step": 15138
+    },
+    {
+      "epoch": 11.395558901016184,
+      "grad_norm": 0.15513139963150024,
+      "learning_rate": 5e-06,
+      "loss": 0.0983,
+      "step": 15139
+    },
+    {
+      "epoch": 11.396311629657509,
+      "grad_norm": 0.18820323050022125,
+      "learning_rate": 5e-06,
+      "loss": 0.23,
+      "step": 15140
+    },
+    {
+      "epoch": 11.397064358298833,
+      "grad_norm": 0.1933794468641281,
+      "learning_rate": 5e-06,
+      "loss": 0.1211,
+      "step": 15141
+    },
+    {
+      "epoch": 11.397817086940158,
+      "grad_norm": 0.2104528844356537,
+      "learning_rate": 5e-06,
+      "loss": 0.1995,
+      "step": 15142
+    },
+    {
+      "epoch": 11.398569815581483,
+      "grad_norm": 0.21750643849372864,
+      "learning_rate": 5e-06,
+      "loss": 0.2611,
+      "step": 15143
+    },
+    {
+      "epoch": 11.399322544222807,
+      "grad_norm": 0.1625976860523224,
+      "learning_rate": 5e-06,
+      "loss": 0.0811,
+      "step": 15144
+    },
+    {
+      "epoch": 11.400075272864132,
+      "grad_norm": 0.17491459846496582,
+      "learning_rate": 5e-06,
+      "loss": 0.1084,
+      "step": 15145
+    },
+    {
+      "epoch": 11.400828001505458,
+      "grad_norm": 0.1605367511510849,
+      "learning_rate": 5e-06,
+      "loss": 0.1427,
+      "step": 15146
+    },
+    {
+      "epoch": 11.401580730146781,
+      "grad_norm": 0.20755258202552795,
+      "learning_rate": 5e-06,
+      "loss": 0.1465,
+      "step": 15147
+    },
+    {
+      "epoch": 11.402333458788107,
+      "grad_norm": 0.17295962572097778,
+      "learning_rate": 5e-06,
+      "loss": 0.1197,
+      "step": 15148
+    },
+    {
+      "epoch": 11.403086187429432,
+      "grad_norm": 0.1828027069568634,
+      "learning_rate": 5e-06,
+      "loss": 0.1717,
+      "step": 15149
+    },
+    {
+      "epoch": 11.403838916070757,
+      "grad_norm": 0.24568435549736023,
+      "learning_rate": 5e-06,
+      "loss": 0.1022,
+      "step": 15150
+    },
+    {
+      "epoch": 11.404591644712081,
+      "grad_norm": 0.2093690037727356,
+      "learning_rate": 5e-06,
+      "loss": 0.1588,
+      "step": 15151
+    },
+    {
+      "epoch": 11.405344373353406,
+      "grad_norm": 0.17838068306446075,
+      "learning_rate": 5e-06,
+      "loss": 0.1906,
+      "step": 15152
+    },
+    {
+      "epoch": 11.406097101994732,
+      "grad_norm": 0.19784137606620789,
+      "learning_rate": 5e-06,
+      "loss": 0.1232,
+      "step": 15153
+    },
+    {
+      "epoch": 11.406849830636055,
+      "grad_norm": 0.19735285639762878,
+      "learning_rate": 5e-06,
+      "loss": 0.1561,
+      "step": 15154
+    },
+    {
+      "epoch": 11.40760255927738,
+      "grad_norm": 0.17813006043434143,
+      "learning_rate": 5e-06,
+      "loss": 0.1819,
+      "step": 15155
+    },
+    {
+      "epoch": 11.408355287918706,
+      "grad_norm": 0.17899352312088013,
+      "learning_rate": 5e-06,
+      "loss": 0.2082,
+      "step": 15156
+    },
+    {
+      "epoch": 11.40910801656003,
+      "grad_norm": 0.15754787623882294,
+      "learning_rate": 5e-06,
+      "loss": 0.1131,
+      "step": 15157
+    },
+    {
+      "epoch": 11.409860745201355,
+      "grad_norm": 0.17342181503772736,
+      "learning_rate": 5e-06,
+      "loss": 0.149,
+      "step": 15158
+    },
+    {
+      "epoch": 11.41061347384268,
+      "grad_norm": 0.1702650934457779,
+      "learning_rate": 5e-06,
+      "loss": 0.1786,
+      "step": 15159
+    },
+    {
+      "epoch": 11.411366202484004,
+      "grad_norm": 0.2151508778333664,
+      "learning_rate": 5e-06,
+      "loss": 0.1266,
+      "step": 15160
+    },
+    {
+      "epoch": 11.41211893112533,
+      "grad_norm": 0.21592502295970917,
+      "learning_rate": 5e-06,
+      "loss": 0.1542,
+      "step": 15161
+    },
+    {
+      "epoch": 11.412871659766655,
+      "grad_norm": 0.18461471796035767,
+      "learning_rate": 5e-06,
+      "loss": 0.1559,
+      "step": 15162
+    },
+    {
+      "epoch": 11.413624388407978,
+      "grad_norm": 0.17324113845825195,
+      "learning_rate": 5e-06,
+      "loss": 0.0929,
+      "step": 15163
+    },
+    {
+      "epoch": 11.414377117049304,
+      "grad_norm": 0.21373796463012695,
+      "learning_rate": 5e-06,
+      "loss": 0.1275,
+      "step": 15164
+    },
+    {
+      "epoch": 11.415129845690629,
+      "grad_norm": 0.1663697063922882,
+      "learning_rate": 5e-06,
+      "loss": 0.1219,
+      "step": 15165
+    },
+    {
+      "epoch": 11.415882574331953,
+      "grad_norm": 0.192871555685997,
+      "learning_rate": 5e-06,
+      "loss": 0.2068,
+      "step": 15166
+    },
+    {
+      "epoch": 11.416635302973278,
+      "grad_norm": 0.18082834780216217,
+      "learning_rate": 5e-06,
+      "loss": 0.1638,
+      "step": 15167
+    },
+    {
+      "epoch": 11.417388031614603,
+      "grad_norm": 0.16938826441764832,
+      "learning_rate": 5e-06,
+      "loss": 0.1142,
+      "step": 15168
+    },
+    {
+      "epoch": 11.418140760255927,
+      "grad_norm": 0.17011189460754395,
+      "learning_rate": 5e-06,
+      "loss": 0.1332,
+      "step": 15169
+    },
+    {
+      "epoch": 11.418893488897252,
+      "grad_norm": 0.21336239576339722,
+      "learning_rate": 5e-06,
+      "loss": 0.1361,
+      "step": 15170
+    },
+    {
+      "epoch": 11.419646217538578,
+      "grad_norm": 0.1884782463312149,
+      "learning_rate": 5e-06,
+      "loss": 0.1507,
+      "step": 15171
+    },
+    {
+      "epoch": 11.420398946179903,
+      "grad_norm": 0.16033443808555603,
+      "learning_rate": 5e-06,
+      "loss": 0.0952,
+      "step": 15172
+    },
+    {
+      "epoch": 11.421151674821227,
+      "grad_norm": 0.16788893938064575,
+      "learning_rate": 5e-06,
+      "loss": 0.099,
+      "step": 15173
+    },
+    {
+      "epoch": 11.421904403462552,
+      "grad_norm": 0.17133073508739471,
+      "learning_rate": 5e-06,
+      "loss": 0.1735,
+      "step": 15174
+    },
+    {
+      "epoch": 11.422657132103877,
+      "grad_norm": 0.1553279608488083,
+      "learning_rate": 5e-06,
+      "loss": 0.1116,
+      "step": 15175
+    },
+    {
+      "epoch": 11.4234098607452,
+      "grad_norm": 0.1936197578907013,
+      "learning_rate": 5e-06,
+      "loss": 0.1447,
+      "step": 15176
+    },
+    {
+      "epoch": 11.424162589386526,
+      "grad_norm": 0.16084672510623932,
+      "learning_rate": 5e-06,
+      "loss": 0.0878,
+      "step": 15177
+    },
+    {
+      "epoch": 11.424915318027852,
+      "grad_norm": 0.19920530915260315,
+      "learning_rate": 5e-06,
+      "loss": 0.1843,
+      "step": 15178
+    },
+    {
+      "epoch": 11.425668046669175,
+      "grad_norm": 0.19088681042194366,
+      "learning_rate": 5e-06,
+      "loss": 0.0953,
+      "step": 15179
+    },
+    {
+      "epoch": 11.4264207753105,
+      "grad_norm": 0.1958848237991333,
+      "learning_rate": 5e-06,
+      "loss": 0.1871,
+      "step": 15180
+    },
+    {
+      "epoch": 11.427173503951826,
+      "grad_norm": 0.17622967064380646,
+      "learning_rate": 5e-06,
+      "loss": 0.1175,
+      "step": 15181
+    },
+    {
+      "epoch": 11.42792623259315,
+      "grad_norm": 0.16004544496536255,
+      "learning_rate": 5e-06,
+      "loss": 0.0819,
+      "step": 15182
+    },
+    {
+      "epoch": 11.428678961234475,
+      "grad_norm": 0.1906105875968933,
+      "learning_rate": 5e-06,
+      "loss": 0.1644,
+      "step": 15183
+    },
+    {
+      "epoch": 11.4294316898758,
+      "grad_norm": 0.1869257688522339,
+      "learning_rate": 5e-06,
+      "loss": 0.1133,
+      "step": 15184
+    },
+    {
+      "epoch": 11.430184418517124,
+      "grad_norm": 0.19233009219169617,
+      "learning_rate": 5e-06,
+      "loss": 0.1284,
+      "step": 15185
+    },
+    {
+      "epoch": 11.430937147158449,
+      "grad_norm": 0.19359175860881805,
+      "learning_rate": 5e-06,
+      "loss": 0.1264,
+      "step": 15186
+    },
+    {
+      "epoch": 11.431689875799774,
+      "grad_norm": 0.17826181650161743,
+      "learning_rate": 5e-06,
+      "loss": 0.1367,
+      "step": 15187
+    },
+    {
+      "epoch": 11.4324426044411,
+      "grad_norm": 0.1941363364458084,
+      "learning_rate": 5e-06,
+      "loss": 0.1525,
+      "step": 15188
+    },
+    {
+      "epoch": 11.433195333082423,
+      "grad_norm": 0.19571298360824585,
+      "learning_rate": 5e-06,
+      "loss": 0.1894,
+      "step": 15189
+    },
+    {
+      "epoch": 11.433948061723749,
+      "grad_norm": 0.15247461199760437,
+      "learning_rate": 5e-06,
+      "loss": 0.1027,
+      "step": 15190
+    },
+    {
+      "epoch": 11.434700790365074,
+      "grad_norm": 0.17351771891117096,
+      "learning_rate": 5e-06,
+      "loss": 0.0443,
+      "step": 15191
+    },
+    {
+      "epoch": 11.435453519006398,
+      "grad_norm": 0.21165116131305695,
+      "learning_rate": 5e-06,
+      "loss": 0.1924,
+      "step": 15192
+    },
+    {
+      "epoch": 11.436206247647723,
+      "grad_norm": 0.1919352412223816,
+      "learning_rate": 5e-06,
+      "loss": 0.1218,
+      "step": 15193
+    },
+    {
+      "epoch": 11.436958976289048,
+      "grad_norm": 0.16468125581741333,
+      "learning_rate": 5e-06,
+      "loss": 0.1415,
+      "step": 15194
+    },
+    {
+      "epoch": 11.437711704930372,
+      "grad_norm": 0.18994835019111633,
+      "learning_rate": 5e-06,
+      "loss": 0.1785,
+      "step": 15195
+    },
+    {
+      "epoch": 11.438464433571697,
+      "grad_norm": 0.18362566828727722,
+      "learning_rate": 5e-06,
+      "loss": 0.1269,
+      "step": 15196
+    },
+    {
+      "epoch": 11.439217162213023,
+      "grad_norm": 0.1499590277671814,
+      "learning_rate": 5e-06,
+      "loss": 0.0604,
+      "step": 15197
+    },
+    {
+      "epoch": 11.439969890854346,
+      "grad_norm": 0.1677277684211731,
+      "learning_rate": 5e-06,
+      "loss": 0.1079,
+      "step": 15198
+    },
+    {
+      "epoch": 11.440722619495672,
+      "grad_norm": 0.18121521174907684,
+      "learning_rate": 5e-06,
+      "loss": 0.135,
+      "step": 15199
+    },
+    {
+      "epoch": 11.441475348136997,
+      "grad_norm": 0.17216306924819946,
+      "learning_rate": 5e-06,
+      "loss": 0.0947,
+      "step": 15200
+    },
+    {
+      "epoch": 11.441475348136997,
+      "eval_loss": 0.18579921126365662,
+      "eval_runtime": 453.0954,
+      "eval_samples_per_second": 21.247,
+      "eval_steps_per_second": 0.664,
+      "step": 15200
+    },
+    {
+      "epoch": 11.44222807677832,
+      "grad_norm": 0.1855546534061432,
+      "learning_rate": 5e-06,
+      "loss": 0.1293,
+      "step": 15201
+    },
+    {
+      "epoch": 11.442980805419646,
+      "grad_norm": 0.17222627997398376,
+      "learning_rate": 5e-06,
+      "loss": 0.1224,
+      "step": 15202
+    },
+    {
+      "epoch": 11.443733534060971,
+      "grad_norm": 0.17391161620616913,
+      "learning_rate": 5e-06,
+      "loss": 0.1124,
+      "step": 15203
+    },
+    {
+      "epoch": 11.444486262702295,
+      "grad_norm": 0.1891934871673584,
+      "learning_rate": 5e-06,
+      "loss": 0.1271,
+      "step": 15204
+    },
+    {
+      "epoch": 11.44523899134362,
+      "grad_norm": 0.18221856653690338,
+      "learning_rate": 5e-06,
+      "loss": 0.0783,
+      "step": 15205
+    },
+    {
+      "epoch": 11.445991719984946,
+      "grad_norm": 0.1920732706785202,
+      "learning_rate": 5e-06,
+      "loss": 0.2042,
+      "step": 15206
+    },
+    {
+      "epoch": 11.446744448626271,
+      "grad_norm": 0.19082894921302795,
+      "learning_rate": 5e-06,
+      "loss": 0.1357,
+      "step": 15207
+    },
+    {
+      "epoch": 11.447497177267595,
+      "grad_norm": 0.2158365249633789,
+      "learning_rate": 5e-06,
+      "loss": 0.1276,
+      "step": 15208
+    },
+    {
+      "epoch": 11.44824990590892,
+      "grad_norm": 0.1725281924009323,
+      "learning_rate": 5e-06,
+      "loss": 0.1849,
+      "step": 15209
+    },
+    {
+      "epoch": 11.449002634550245,
+      "grad_norm": 0.1691107302904129,
+      "learning_rate": 5e-06,
+      "loss": 0.1289,
+      "step": 15210
+    },
+    {
+      "epoch": 11.449755363191569,
+      "grad_norm": 0.1998799443244934,
+      "learning_rate": 5e-06,
+      "loss": 0.1007,
+      "step": 15211
+    },
+    {
+      "epoch": 11.450508091832894,
+      "grad_norm": 0.16592298448085785,
+      "learning_rate": 5e-06,
+      "loss": 0.1275,
+      "step": 15212
+    },
+    {
+      "epoch": 11.45126082047422,
+      "grad_norm": 0.18168780207633972,
+      "learning_rate": 5e-06,
+      "loss": 0.1947,
+      "step": 15213
+    },
+    {
+      "epoch": 11.452013549115543,
+      "grad_norm": 0.19634820520877838,
+      "learning_rate": 5e-06,
+      "loss": 0.1686,
+      "step": 15214
+    },
+    {
+      "epoch": 11.452766277756869,
+      "grad_norm": 0.18831753730773926,
+      "learning_rate": 5e-06,
+      "loss": 0.1506,
+      "step": 15215
+    },
+    {
+      "epoch": 11.453519006398194,
+      "grad_norm": 0.17521189153194427,
+      "learning_rate": 5e-06,
+      "loss": 0.1223,
+      "step": 15216
+    },
+    {
+      "epoch": 11.454271735039518,
+      "grad_norm": 0.17607524991035461,
+      "learning_rate": 5e-06,
+      "loss": 0.0765,
+      "step": 15217
+    },
+    {
+      "epoch": 11.455024463680843,
+      "grad_norm": 0.1953452080488205,
+      "learning_rate": 5e-06,
+      "loss": 0.1601,
+      "step": 15218
+    },
+    {
+      "epoch": 11.455777192322168,
+      "grad_norm": 0.20226749777793884,
+      "learning_rate": 5e-06,
+      "loss": 0.2113,
+      "step": 15219
+    },
+    {
+      "epoch": 11.456529920963492,
+      "grad_norm": 0.16415488719940186,
+      "learning_rate": 5e-06,
+      "loss": 0.1115,
+      "step": 15220
+    },
+    {
+      "epoch": 11.457282649604817,
+      "grad_norm": 0.1691894233226776,
+      "learning_rate": 5e-06,
+      "loss": 0.0642,
+      "step": 15221
+    },
+    {
+      "epoch": 11.458035378246143,
+      "grad_norm": 0.1612430065870285,
+      "learning_rate": 5e-06,
+      "loss": 0.0956,
+      "step": 15222
+    },
+    {
+      "epoch": 11.458788106887468,
+      "grad_norm": 0.17577609419822693,
+      "learning_rate": 5e-06,
+      "loss": 0.0977,
+      "step": 15223
+    },
+    {
+      "epoch": 11.459540835528792,
+      "grad_norm": 0.23150087893009186,
+      "learning_rate": 5e-06,
+      "loss": 0.1885,
+      "step": 15224
+    },
+    {
+      "epoch": 11.460293564170117,
+      "grad_norm": 0.18573850393295288,
+      "learning_rate": 5e-06,
+      "loss": 0.1677,
+      "step": 15225
+    },
+    {
+      "epoch": 11.461046292811442,
+      "grad_norm": 0.18642368912696838,
+      "learning_rate": 5e-06,
+      "loss": 0.196,
+      "step": 15226
+    },
+    {
+      "epoch": 11.461799021452766,
+      "grad_norm": 0.16857869923114777,
+      "learning_rate": 5e-06,
+      "loss": 0.1854,
+      "step": 15227
+    },
+    {
+      "epoch": 11.462551750094091,
+      "grad_norm": 0.19637766480445862,
+      "learning_rate": 5e-06,
+      "loss": 0.2242,
+      "step": 15228
+    },
+    {
+      "epoch": 11.463304478735417,
+      "grad_norm": 0.17782887816429138,
+      "learning_rate": 5e-06,
+      "loss": 0.1068,
+      "step": 15229
+    },
+    {
+      "epoch": 11.46405720737674,
+      "grad_norm": 0.16832217574119568,
+      "learning_rate": 5e-06,
+      "loss": 0.2067,
+      "step": 15230
+    },
+    {
+      "epoch": 11.464809936018066,
+      "grad_norm": 0.1893640160560608,
+      "learning_rate": 5e-06,
+      "loss": 0.1395,
+      "step": 15231
+    },
+    {
+      "epoch": 11.465562664659391,
+      "grad_norm": 0.17671120166778564,
+      "learning_rate": 5e-06,
+      "loss": 0.1286,
+      "step": 15232
+    },
+    {
+      "epoch": 11.466315393300714,
+      "grad_norm": 0.20142492651939392,
+      "learning_rate": 5e-06,
+      "loss": 0.1976,
+      "step": 15233
+    },
+    {
+      "epoch": 11.46706812194204,
+      "grad_norm": 0.2079554796218872,
+      "learning_rate": 5e-06,
+      "loss": 0.1148,
+      "step": 15234
+    },
+    {
+      "epoch": 11.467820850583365,
+      "grad_norm": 0.18313248455524445,
+      "learning_rate": 5e-06,
+      "loss": 0.1323,
+      "step": 15235
+    },
+    {
+      "epoch": 11.468573579224689,
+      "grad_norm": 0.18477286398410797,
+      "learning_rate": 5e-06,
+      "loss": 0.2285,
+      "step": 15236
+    },
+    {
+      "epoch": 11.469326307866014,
+      "grad_norm": 0.17938154935836792,
+      "learning_rate": 5e-06,
+      "loss": 0.1692,
+      "step": 15237
+    },
+    {
+      "epoch": 11.47007903650734,
+      "grad_norm": 0.19527281820774078,
+      "learning_rate": 5e-06,
+      "loss": 0.1453,
+      "step": 15238
+    },
+    {
+      "epoch": 11.470831765148663,
+      "grad_norm": 0.1751263439655304,
+      "learning_rate": 5e-06,
+      "loss": 0.1248,
+      "step": 15239
+    },
+    {
+      "epoch": 11.471584493789988,
+      "grad_norm": 0.1672459989786148,
+      "learning_rate": 5e-06,
+      "loss": 0.1334,
+      "step": 15240
+    },
+    {
+      "epoch": 11.472337222431314,
+      "grad_norm": 0.19905880093574524,
+      "learning_rate": 5e-06,
+      "loss": 0.1301,
+      "step": 15241
+    },
+    {
+      "epoch": 11.473089951072637,
+      "grad_norm": 0.1728474497795105,
+      "learning_rate": 5e-06,
+      "loss": 0.1399,
+      "step": 15242
+    },
+    {
+      "epoch": 11.473842679713963,
+      "grad_norm": 0.18407557904720306,
+      "learning_rate": 5e-06,
+      "loss": 0.1295,
+      "step": 15243
+    },
+    {
+      "epoch": 11.474595408355288,
+      "grad_norm": 0.18678194284439087,
+      "learning_rate": 5e-06,
+      "loss": 0.1325,
+      "step": 15244
+    },
+    {
+      "epoch": 11.475348136996613,
+      "grad_norm": 0.1978190690279007,
+      "learning_rate": 5e-06,
+      "loss": 0.1078,
+      "step": 15245
+    },
+    {
+      "epoch": 11.476100865637937,
+      "grad_norm": 0.1867973953485489,
+      "learning_rate": 5e-06,
+      "loss": 0.1439,
+      "step": 15246
+    },
+    {
+      "epoch": 11.476853594279262,
+      "grad_norm": 0.1891653686761856,
+      "learning_rate": 5e-06,
+      "loss": 0.109,
+      "step": 15247
+    },
+    {
+      "epoch": 11.477606322920588,
+      "grad_norm": 0.20922620594501495,
+      "learning_rate": 5e-06,
+      "loss": 0.1867,
+      "step": 15248
+    },
+    {
+      "epoch": 11.478359051561911,
+      "grad_norm": 0.17750991880893707,
+      "learning_rate": 5e-06,
+      "loss": 0.0578,
+      "step": 15249
+    },
+    {
+      "epoch": 11.479111780203237,
+      "grad_norm": 0.17142005264759064,
+      "learning_rate": 5e-06,
+      "loss": 0.1472,
+      "step": 15250
+    },
+    {
+      "epoch": 11.479864508844562,
+      "grad_norm": 0.17947959899902344,
+      "learning_rate": 5e-06,
+      "loss": 0.0914,
+      "step": 15251
+    },
+    {
+      "epoch": 11.480617237485886,
+      "grad_norm": 0.18580587208271027,
+      "learning_rate": 5e-06,
+      "loss": 0.2079,
+      "step": 15252
+    },
+    {
+      "epoch": 11.481369966127211,
+      "grad_norm": 0.19834010303020477,
+      "learning_rate": 5e-06,
+      "loss": 0.1712,
+      "step": 15253
+    },
+    {
+      "epoch": 11.482122694768536,
+      "grad_norm": 0.17350070178508759,
+      "learning_rate": 5e-06,
+      "loss": 0.1655,
+      "step": 15254
+    },
+    {
+      "epoch": 11.48287542340986,
+      "grad_norm": 0.1775667518377304,
+      "learning_rate": 5e-06,
+      "loss": 0.118,
+      "step": 15255
+    },
+    {
+      "epoch": 11.483628152051185,
+      "grad_norm": 0.1936013400554657,
+      "learning_rate": 5e-06,
+      "loss": 0.1233,
+      "step": 15256
+    },
+    {
+      "epoch": 11.48438088069251,
+      "grad_norm": 0.1720784306526184,
+      "learning_rate": 5e-06,
+      "loss": 0.1725,
+      "step": 15257
+    },
+    {
+      "epoch": 11.485133609333834,
+      "grad_norm": 0.17432133853435516,
+      "learning_rate": 5e-06,
+      "loss": 0.0577,
+      "step": 15258
+    },
+    {
+      "epoch": 11.48588633797516,
+      "grad_norm": 0.16736242175102234,
+      "learning_rate": 5e-06,
+      "loss": 0.0921,
+      "step": 15259
+    },
+    {
+      "epoch": 11.486639066616485,
+      "grad_norm": 0.18062728643417358,
+      "learning_rate": 5e-06,
+      "loss": 0.1272,
+      "step": 15260
+    },
+    {
+      "epoch": 11.48739179525781,
+      "grad_norm": 0.16933098435401917,
+      "learning_rate": 5e-06,
+      "loss": 0.13,
+      "step": 15261
+    },
+    {
+      "epoch": 11.488144523899134,
+      "grad_norm": 0.18125362694263458,
+      "learning_rate": 5e-06,
+      "loss": 0.1348,
+      "step": 15262
+    },
+    {
+      "epoch": 11.48889725254046,
+      "grad_norm": 0.2109428495168686,
+      "learning_rate": 5e-06,
+      "loss": 0.1859,
+      "step": 15263
+    },
+    {
+      "epoch": 11.489649981181785,
+      "grad_norm": 0.18194769322872162,
+      "learning_rate": 5e-06,
+      "loss": 0.1653,
+      "step": 15264
+    },
+    {
+      "epoch": 11.490402709823108,
+      "grad_norm": 0.17470067739486694,
+      "learning_rate": 5e-06,
+      "loss": 0.1446,
+      "step": 15265
+    },
+    {
+      "epoch": 11.491155438464434,
+      "grad_norm": 0.1864662617444992,
+      "learning_rate": 5e-06,
+      "loss": 0.1644,
+      "step": 15266
+    },
+    {
+      "epoch": 11.491908167105759,
+      "grad_norm": 0.18832701444625854,
+      "learning_rate": 5e-06,
+      "loss": 0.1663,
+      "step": 15267
+    },
+    {
+      "epoch": 11.492660895747083,
+      "grad_norm": 0.19269758462905884,
+      "learning_rate": 5e-06,
+      "loss": 0.1157,
+      "step": 15268
+    },
+    {
+      "epoch": 11.493413624388408,
+      "grad_norm": 0.19565002620220184,
+      "learning_rate": 5e-06,
+      "loss": 0.1438,
+      "step": 15269
+    },
+    {
+      "epoch": 11.494166353029733,
+      "grad_norm": 0.19609390199184418,
+      "learning_rate": 5e-06,
+      "loss": 0.1523,
+      "step": 15270
+    },
+    {
+      "epoch": 11.494919081671057,
+      "grad_norm": 0.16823747754096985,
+      "learning_rate": 5e-06,
+      "loss": 0.1978,
+      "step": 15271
+    },
+    {
+      "epoch": 11.495671810312382,
+      "grad_norm": 0.18173758685588837,
+      "learning_rate": 5e-06,
+      "loss": 0.1471,
+      "step": 15272
+    },
+    {
+      "epoch": 11.496424538953708,
+      "grad_norm": 0.186704620718956,
+      "learning_rate": 5e-06,
+      "loss": 0.1497,
+      "step": 15273
+    },
+    {
+      "epoch": 11.497177267595031,
+      "grad_norm": 0.1720120906829834,
+      "learning_rate": 5e-06,
+      "loss": 0.1129,
+      "step": 15274
+    },
+    {
+      "epoch": 11.497929996236357,
+      "grad_norm": 0.17433257400989532,
+      "learning_rate": 5e-06,
+      "loss": 0.178,
+      "step": 15275
+    },
+    {
+      "epoch": 11.498682724877682,
+      "grad_norm": 0.1656228005886078,
+      "learning_rate": 5e-06,
+      "loss": 0.1267,
+      "step": 15276
+    },
+    {
+      "epoch": 11.499435453519006,
+      "grad_norm": 0.185775026679039,
+      "learning_rate": 5e-06,
+      "loss": 0.1047,
+      "step": 15277
+    },
+    {
+      "epoch": 11.500188182160331,
+      "grad_norm": 0.19019873440265656,
+      "learning_rate": 5e-06,
+      "loss": 0.1271,
+      "step": 15278
+    },
+    {
+      "epoch": 11.500940910801656,
+      "grad_norm": 0.19401653110980988,
+      "learning_rate": 5e-06,
+      "loss": 0.1165,
+      "step": 15279
+    },
+    {
+      "epoch": 11.501693639442982,
+      "grad_norm": 0.20986121892929077,
+      "learning_rate": 5e-06,
+      "loss": 0.1535,
+      "step": 15280
+    },
+    {
+      "epoch": 11.502446368084305,
+      "grad_norm": 0.1791776567697525,
+      "learning_rate": 5e-06,
+      "loss": 0.1814,
+      "step": 15281
+    },
+    {
+      "epoch": 11.50319909672563,
+      "grad_norm": 0.18205681443214417,
+      "learning_rate": 5e-06,
+      "loss": 0.1405,
+      "step": 15282
+    },
+    {
+      "epoch": 11.503951825366956,
+      "grad_norm": 0.18308709561824799,
+      "learning_rate": 5e-06,
+      "loss": 0.1338,
+      "step": 15283
+    },
+    {
+      "epoch": 11.50470455400828,
+      "grad_norm": 0.17722970247268677,
+      "learning_rate": 5e-06,
+      "loss": 0.189,
+      "step": 15284
+    },
+    {
+      "epoch": 11.505457282649605,
+      "grad_norm": 0.1815502792596817,
+      "learning_rate": 5e-06,
+      "loss": 0.158,
+      "step": 15285
+    },
+    {
+      "epoch": 11.50621001129093,
+      "grad_norm": 0.20682214200496674,
+      "learning_rate": 5e-06,
+      "loss": 0.2406,
+      "step": 15286
+    },
+    {
+      "epoch": 11.506962739932254,
+      "grad_norm": 0.1648474484682083,
+      "learning_rate": 5e-06,
+      "loss": 0.1547,
+      "step": 15287
+    },
+    {
+      "epoch": 11.50771546857358,
+      "grad_norm": 0.16658447682857513,
+      "learning_rate": 5e-06,
+      "loss": 0.1197,
+      "step": 15288
+    },
+    {
+      "epoch": 11.508468197214905,
+      "grad_norm": 0.18997609615325928,
+      "learning_rate": 5e-06,
+      "loss": 0.1041,
+      "step": 15289
+    },
+    {
+      "epoch": 11.509220925856228,
+      "grad_norm": 0.19123826920986176,
+      "learning_rate": 5e-06,
+      "loss": 0.1341,
+      "step": 15290
+    },
+    {
+      "epoch": 11.509973654497553,
+      "grad_norm": 0.17054803669452667,
+      "learning_rate": 5e-06,
+      "loss": 0.0389,
+      "step": 15291
+    },
+    {
+      "epoch": 11.510726383138879,
+      "grad_norm": 0.20527349412441254,
+      "learning_rate": 5e-06,
+      "loss": 0.1023,
+      "step": 15292
+    },
+    {
+      "epoch": 11.511479111780202,
+      "grad_norm": 0.17310602962970734,
+      "learning_rate": 5e-06,
+      "loss": 0.1693,
+      "step": 15293
+    },
+    {
+      "epoch": 11.512231840421528,
+      "grad_norm": 0.19257670640945435,
+      "learning_rate": 5e-06,
+      "loss": 0.1163,
+      "step": 15294
+    },
+    {
+      "epoch": 11.512984569062853,
+      "grad_norm": 0.163971945643425,
+      "learning_rate": 5e-06,
+      "loss": 0.1353,
+      "step": 15295
+    },
+    {
+      "epoch": 11.513737297704179,
+      "grad_norm": 0.17143027484416962,
+      "learning_rate": 5e-06,
+      "loss": 0.1222,
+      "step": 15296
+    },
+    {
+      "epoch": 11.514490026345502,
+      "grad_norm": 0.17749537527561188,
+      "learning_rate": 5e-06,
+      "loss": 0.102,
+      "step": 15297
+    },
+    {
+      "epoch": 11.515242754986827,
+      "grad_norm": 0.2014029175043106,
+      "learning_rate": 5e-06,
+      "loss": 0.277,
+      "step": 15298
+    },
+    {
+      "epoch": 11.515995483628153,
+      "grad_norm": 0.20297354459762573,
+      "learning_rate": 5e-06,
+      "loss": 0.1355,
+      "step": 15299
+    },
+    {
+      "epoch": 11.516748212269476,
+      "grad_norm": 0.1833321899175644,
+      "learning_rate": 5e-06,
+      "loss": 0.1696,
+      "step": 15300
+    },
+    {
+      "epoch": 11.517500940910802,
+      "grad_norm": 0.20239543914794922,
+      "learning_rate": 5e-06,
+      "loss": 0.1454,
+      "step": 15301
+    },
+    {
+      "epoch": 11.518253669552127,
+      "grad_norm": 0.19776159524917603,
+      "learning_rate": 5e-06,
+      "loss": 0.1466,
+      "step": 15302
+    },
+    {
+      "epoch": 11.51900639819345,
+      "grad_norm": 0.19654856622219086,
+      "learning_rate": 5e-06,
+      "loss": 0.1701,
+      "step": 15303
+    },
+    {
+      "epoch": 11.519759126834776,
+      "grad_norm": 0.18491004407405853,
+      "learning_rate": 5e-06,
+      "loss": 0.1795,
+      "step": 15304
+    },
+    {
+      "epoch": 11.520511855476101,
+      "grad_norm": 0.1773645132780075,
+      "learning_rate": 5e-06,
+      "loss": 0.1355,
+      "step": 15305
+    },
+    {
+      "epoch": 11.521264584117425,
+      "grad_norm": 0.15620282292366028,
+      "learning_rate": 5e-06,
+      "loss": 0.1837,
+      "step": 15306
+    },
+    {
+      "epoch": 11.52201731275875,
+      "grad_norm": 0.20085002481937408,
+      "learning_rate": 5e-06,
+      "loss": 0.1917,
+      "step": 15307
+    },
+    {
+      "epoch": 11.522770041400076,
+      "grad_norm": 0.17189159989356995,
+      "learning_rate": 5e-06,
+      "loss": 0.0788,
+      "step": 15308
+    },
+    {
+      "epoch": 11.5235227700414,
+      "grad_norm": 0.1890510618686676,
+      "learning_rate": 5e-06,
+      "loss": 0.1627,
+      "step": 15309
+    },
+    {
+      "epoch": 11.524275498682725,
+      "grad_norm": 0.18564119935035706,
+      "learning_rate": 5e-06,
+      "loss": 0.1351,
+      "step": 15310
+    },
+    {
+      "epoch": 11.52502822732405,
+      "grad_norm": 0.21036291122436523,
+      "learning_rate": 5e-06,
+      "loss": 0.1285,
+      "step": 15311
+    },
+    {
+      "epoch": 11.525780955965374,
+      "grad_norm": 0.18161579966545105,
+      "learning_rate": 5e-06,
+      "loss": 0.11,
+      "step": 15312
+    },
+    {
+      "epoch": 11.526533684606699,
+      "grad_norm": 0.16365350782871246,
+      "learning_rate": 5e-06,
+      "loss": 0.1099,
+      "step": 15313
+    },
+    {
+      "epoch": 11.527286413248024,
+      "grad_norm": 0.1784616857767105,
+      "learning_rate": 5e-06,
+      "loss": 0.1263,
+      "step": 15314
+    },
+    {
+      "epoch": 11.528039141889348,
+      "grad_norm": 0.18272723257541656,
+      "learning_rate": 5e-06,
+      "loss": 0.1549,
+      "step": 15315
+    },
+    {
+      "epoch": 11.528791870530673,
+      "grad_norm": 0.19497260451316833,
+      "learning_rate": 5e-06,
+      "loss": 0.1538,
+      "step": 15316
+    },
+    {
+      "epoch": 11.529544599171999,
+      "grad_norm": 0.1693209558725357,
+      "learning_rate": 5e-06,
+      "loss": 0.1301,
+      "step": 15317
+    },
+    {
+      "epoch": 11.530297327813324,
+      "grad_norm": 0.17413431406021118,
+      "learning_rate": 5e-06,
+      "loss": 0.1224,
+      "step": 15318
+    },
+    {
+      "epoch": 11.531050056454648,
+      "grad_norm": 0.1715451329946518,
+      "learning_rate": 5e-06,
+      "loss": 0.1935,
+      "step": 15319
+    },
+    {
+      "epoch": 11.531802785095973,
+      "grad_norm": 0.16009017825126648,
+      "learning_rate": 5e-06,
+      "loss": 0.0843,
+      "step": 15320
+    },
+    {
+      "epoch": 11.532555513737298,
+      "grad_norm": 0.20076726377010345,
+      "learning_rate": 5e-06,
+      "loss": 0.1671,
+      "step": 15321
+    },
+    {
+      "epoch": 11.533308242378622,
+      "grad_norm": 0.20761187374591827,
+      "learning_rate": 5e-06,
+      "loss": 0.1497,
+      "step": 15322
+    },
+    {
+      "epoch": 11.534060971019947,
+      "grad_norm": 0.18851982057094574,
+      "learning_rate": 5e-06,
+      "loss": 0.2415,
+      "step": 15323
+    },
+    {
+      "epoch": 11.534813699661273,
+      "grad_norm": 0.1719018667936325,
+      "learning_rate": 5e-06,
+      "loss": 0.1334,
+      "step": 15324
+    },
+    {
+      "epoch": 11.535566428302596,
+      "grad_norm": 0.16416878998279572,
+      "learning_rate": 5e-06,
+      "loss": 0.139,
+      "step": 15325
+    },
+    {
+      "epoch": 11.536319156943922,
+      "grad_norm": 0.20325927436351776,
+      "learning_rate": 5e-06,
+      "loss": 0.1657,
+      "step": 15326
+    },
+    {
+      "epoch": 11.537071885585247,
+      "grad_norm": 0.1676267683506012,
+      "learning_rate": 5e-06,
+      "loss": 0.0944,
+      "step": 15327
+    },
+    {
+      "epoch": 11.53782461422657,
+      "grad_norm": 0.18753260374069214,
+      "learning_rate": 5e-06,
+      "loss": 0.119,
+      "step": 15328
+    },
+    {
+      "epoch": 11.538577342867896,
+      "grad_norm": 0.15155306458473206,
+      "learning_rate": 5e-06,
+      "loss": 0.062,
+      "step": 15329
+    },
+    {
+      "epoch": 11.539330071509221,
+      "grad_norm": 0.1979656219482422,
+      "learning_rate": 5e-06,
+      "loss": 0.173,
+      "step": 15330
+    },
+    {
+      "epoch": 11.540082800150547,
+      "grad_norm": 0.1735519915819168,
+      "learning_rate": 5e-06,
+      "loss": 0.0548,
+      "step": 15331
+    },
+    {
+      "epoch": 11.54083552879187,
+      "grad_norm": 0.19605492055416107,
+      "learning_rate": 5e-06,
+      "loss": 0.1429,
+      "step": 15332
+    },
+    {
+      "epoch": 11.541588257433196,
+      "grad_norm": 0.187017023563385,
+      "learning_rate": 5e-06,
+      "loss": 0.1652,
+      "step": 15333
+    },
+    {
+      "epoch": 11.542340986074521,
+      "grad_norm": 0.19398434460163116,
+      "learning_rate": 5e-06,
+      "loss": 0.223,
+      "step": 15334
+    },
+    {
+      "epoch": 11.543093714715845,
+      "grad_norm": 0.1958903968334198,
+      "learning_rate": 5e-06,
+      "loss": 0.1311,
+      "step": 15335
+    },
+    {
+      "epoch": 11.54384644335717,
+      "grad_norm": 0.1798068732023239,
+      "learning_rate": 5e-06,
+      "loss": 0.1003,
+      "step": 15336
+    },
+    {
+      "epoch": 11.544599171998495,
+      "grad_norm": 0.17604784667491913,
+      "learning_rate": 5e-06,
+      "loss": 0.1594,
+      "step": 15337
+    },
+    {
+      "epoch": 11.545351900639819,
+      "grad_norm": 0.18487808108329773,
+      "learning_rate": 5e-06,
+      "loss": 0.1553,
+      "step": 15338
+    },
+    {
+      "epoch": 11.546104629281144,
+      "grad_norm": 0.16895359754562378,
+      "learning_rate": 5e-06,
+      "loss": 0.1176,
+      "step": 15339
+    },
+    {
+      "epoch": 11.54685735792247,
+      "grad_norm": 0.19894883036613464,
+      "learning_rate": 5e-06,
+      "loss": 0.202,
+      "step": 15340
+    },
+    {
+      "epoch": 11.547610086563793,
+      "grad_norm": 0.17687857151031494,
+      "learning_rate": 5e-06,
+      "loss": 0.1507,
+      "step": 15341
+    },
+    {
+      "epoch": 11.548362815205119,
+      "grad_norm": 0.1587030440568924,
+      "learning_rate": 5e-06,
+      "loss": 0.1183,
+      "step": 15342
+    },
+    {
+      "epoch": 11.549115543846444,
+      "grad_norm": 0.18111848831176758,
+      "learning_rate": 5e-06,
+      "loss": 0.1658,
+      "step": 15343
+    },
+    {
+      "epoch": 11.549868272487767,
+      "grad_norm": 0.19798271358013153,
+      "learning_rate": 5e-06,
+      "loss": 0.1338,
+      "step": 15344
+    },
+    {
+      "epoch": 11.550621001129093,
+      "grad_norm": 0.1947207897901535,
+      "learning_rate": 5e-06,
+      "loss": 0.2118,
+      "step": 15345
+    },
+    {
+      "epoch": 11.551373729770418,
+      "grad_norm": 0.21385855972766876,
+      "learning_rate": 5e-06,
+      "loss": 0.1754,
+      "step": 15346
+    },
+    {
+      "epoch": 11.552126458411742,
+      "grad_norm": 0.15864886343479156,
+      "learning_rate": 5e-06,
+      "loss": 0.1414,
+      "step": 15347
+    },
+    {
+      "epoch": 11.552879187053067,
+      "grad_norm": 0.1817694455385208,
+      "learning_rate": 5e-06,
+      "loss": 0.1586,
+      "step": 15348
+    },
+    {
+      "epoch": 11.553631915694393,
+      "grad_norm": 0.17727315425872803,
+      "learning_rate": 5e-06,
+      "loss": 0.0924,
+      "step": 15349
+    },
+    {
+      "epoch": 11.554384644335716,
+      "grad_norm": 0.1696443110704422,
+      "learning_rate": 5e-06,
+      "loss": 0.1117,
+      "step": 15350
+    },
+    {
+      "epoch": 11.555137372977041,
+      "grad_norm": 0.17270034551620483,
+      "learning_rate": 5e-06,
+      "loss": 0.1162,
+      "step": 15351
+    },
+    {
+      "epoch": 11.555890101618367,
+      "grad_norm": 0.18591921031475067,
+      "learning_rate": 5e-06,
+      "loss": 0.1685,
+      "step": 15352
+    },
+    {
+      "epoch": 11.556642830259692,
+      "grad_norm": 0.1628810316324234,
+      "learning_rate": 5e-06,
+      "loss": 0.1086,
+      "step": 15353
+    },
+    {
+      "epoch": 11.557395558901016,
+      "grad_norm": 0.1825563907623291,
+      "learning_rate": 5e-06,
+      "loss": 0.1643,
+      "step": 15354
+    },
+    {
+      "epoch": 11.558148287542341,
+      "grad_norm": 0.1999026983976364,
+      "learning_rate": 5e-06,
+      "loss": 0.1298,
+      "step": 15355
+    },
+    {
+      "epoch": 11.558901016183666,
+      "grad_norm": 0.20441579818725586,
+      "learning_rate": 5e-06,
+      "loss": 0.1852,
+      "step": 15356
+    },
+    {
+      "epoch": 11.55965374482499,
+      "grad_norm": 0.18090513348579407,
+      "learning_rate": 5e-06,
+      "loss": 0.1462,
+      "step": 15357
+    },
+    {
+      "epoch": 11.560406473466315,
+      "grad_norm": 0.20428521931171417,
+      "learning_rate": 5e-06,
+      "loss": 0.2108,
+      "step": 15358
+    },
+    {
+      "epoch": 11.56115920210764,
+      "grad_norm": 0.17408300936222076,
+      "learning_rate": 5e-06,
+      "loss": 0.1391,
+      "step": 15359
+    },
+    {
+      "epoch": 11.561911930748964,
+      "grad_norm": 0.15470005571842194,
+      "learning_rate": 5e-06,
+      "loss": 0.1929,
+      "step": 15360
+    },
+    {
+      "epoch": 11.56266465939029,
+      "grad_norm": 0.1684541404247284,
+      "learning_rate": 5e-06,
+      "loss": 0.0633,
+      "step": 15361
+    },
+    {
+      "epoch": 11.563417388031615,
+      "grad_norm": 0.1720270812511444,
+      "learning_rate": 5e-06,
+      "loss": 0.0972,
+      "step": 15362
+    },
+    {
+      "epoch": 11.564170116672939,
+      "grad_norm": 0.17911987006664276,
+      "learning_rate": 5e-06,
+      "loss": 0.1503,
+      "step": 15363
+    },
+    {
+      "epoch": 11.564922845314264,
+      "grad_norm": 0.190771222114563,
+      "learning_rate": 5e-06,
+      "loss": 0.1506,
+      "step": 15364
+    },
+    {
+      "epoch": 11.56567557395559,
+      "grad_norm": 0.1691066026687622,
+      "learning_rate": 5e-06,
+      "loss": 0.0984,
+      "step": 15365
+    },
+    {
+      "epoch": 11.566428302596913,
+      "grad_norm": 0.1722022145986557,
+      "learning_rate": 5e-06,
+      "loss": 0.0853,
+      "step": 15366
+    },
+    {
+      "epoch": 11.567181031238238,
+      "grad_norm": 0.16261523962020874,
+      "learning_rate": 5e-06,
+      "loss": 0.0812,
+      "step": 15367
+    },
+    {
+      "epoch": 11.567933759879564,
+      "grad_norm": 0.18080390989780426,
+      "learning_rate": 5e-06,
+      "loss": 0.0826,
+      "step": 15368
+    },
+    {
+      "epoch": 11.568686488520889,
+      "grad_norm": 0.1521061807870865,
+      "learning_rate": 5e-06,
+      "loss": 0.0919,
+      "step": 15369
+    },
+    {
+      "epoch": 11.569439217162213,
+      "grad_norm": 0.17070066928863525,
+      "learning_rate": 5e-06,
+      "loss": 0.0934,
+      "step": 15370
+    },
+    {
+      "epoch": 11.570191945803538,
+      "grad_norm": 0.1662646383047104,
+      "learning_rate": 5e-06,
+      "loss": 0.1636,
+      "step": 15371
+    },
+    {
+      "epoch": 11.570944674444863,
+      "grad_norm": 0.19213058054447174,
+      "learning_rate": 5e-06,
+      "loss": 0.1573,
+      "step": 15372
+    },
+    {
+      "epoch": 11.571697403086187,
+      "grad_norm": 0.175612673163414,
+      "learning_rate": 5e-06,
+      "loss": 0.0912,
+      "step": 15373
+    },
+    {
+      "epoch": 11.572450131727512,
+      "grad_norm": 0.1671852022409439,
+      "learning_rate": 5e-06,
+      "loss": 0.11,
+      "step": 15374
+    },
+    {
+      "epoch": 11.573202860368838,
+      "grad_norm": 0.16577516496181488,
+      "learning_rate": 5e-06,
+      "loss": 0.1397,
+      "step": 15375
+    },
+    {
+      "epoch": 11.573955589010161,
+      "grad_norm": 0.18235290050506592,
+      "learning_rate": 5e-06,
+      "loss": 0.1402,
+      "step": 15376
+    },
+    {
+      "epoch": 11.574708317651487,
+      "grad_norm": 0.18592019379138947,
+      "learning_rate": 5e-06,
+      "loss": 0.1221,
+      "step": 15377
+    },
+    {
+      "epoch": 11.575461046292812,
+      "grad_norm": 0.1742488294839859,
+      "learning_rate": 5e-06,
+      "loss": 0.1881,
+      "step": 15378
+    },
+    {
+      "epoch": 11.576213774934136,
+      "grad_norm": 0.17706164717674255,
+      "learning_rate": 5e-06,
+      "loss": 0.1586,
+      "step": 15379
+    },
+    {
+      "epoch": 11.576966503575461,
+      "grad_norm": 0.18152602016925812,
+      "learning_rate": 5e-06,
+      "loss": 0.1271,
+      "step": 15380
+    },
+    {
+      "epoch": 11.577719232216786,
+      "grad_norm": 0.18097375333309174,
+      "learning_rate": 5e-06,
+      "loss": 0.1508,
+      "step": 15381
+    },
+    {
+      "epoch": 11.57847196085811,
+      "grad_norm": 0.19472715258598328,
+      "learning_rate": 5e-06,
+      "loss": 0.2362,
+      "step": 15382
+    },
+    {
+      "epoch": 11.579224689499435,
+      "grad_norm": 0.18323251605033875,
+      "learning_rate": 5e-06,
+      "loss": 0.0985,
+      "step": 15383
+    },
+    {
+      "epoch": 11.57997741814076,
+      "grad_norm": 0.20813509821891785,
+      "learning_rate": 5e-06,
+      "loss": 0.1194,
+      "step": 15384
+    },
+    {
+      "epoch": 11.580730146782084,
+      "grad_norm": 0.17980512976646423,
+      "learning_rate": 5e-06,
+      "loss": 0.1369,
+      "step": 15385
+    },
+    {
+      "epoch": 11.58148287542341,
+      "grad_norm": 0.15845392644405365,
+      "learning_rate": 5e-06,
+      "loss": 0.0613,
+      "step": 15386
+    },
+    {
+      "epoch": 11.582235604064735,
+      "grad_norm": 0.15887489914894104,
+      "learning_rate": 5e-06,
+      "loss": 0.087,
+      "step": 15387
+    },
+    {
+      "epoch": 11.582988332706059,
+      "grad_norm": 0.16917990148067474,
+      "learning_rate": 5e-06,
+      "loss": 0.1149,
+      "step": 15388
+    },
+    {
+      "epoch": 11.583741061347384,
+      "grad_norm": 0.16532613337039948,
+      "learning_rate": 5e-06,
+      "loss": 0.0935,
+      "step": 15389
+    },
+    {
+      "epoch": 11.58449378998871,
+      "grad_norm": 0.18434423208236694,
+      "learning_rate": 5e-06,
+      "loss": 0.1711,
+      "step": 15390
+    },
+    {
+      "epoch": 11.585246518630035,
+      "grad_norm": 0.16000044345855713,
+      "learning_rate": 5e-06,
+      "loss": 0.1465,
+      "step": 15391
+    },
+    {
+      "epoch": 11.585999247271358,
+      "grad_norm": 0.18806087970733643,
+      "learning_rate": 5e-06,
+      "loss": 0.1362,
+      "step": 15392
+    },
+    {
+      "epoch": 11.586751975912684,
+      "grad_norm": 0.19311489164829254,
+      "learning_rate": 5e-06,
+      "loss": 0.1443,
+      "step": 15393
+    },
+    {
+      "epoch": 11.587504704554009,
+      "grad_norm": 0.1718987226486206,
+      "learning_rate": 5e-06,
+      "loss": 0.1998,
+      "step": 15394
+    },
+    {
+      "epoch": 11.588257433195333,
+      "grad_norm": 0.1915034055709839,
+      "learning_rate": 5e-06,
+      "loss": 0.1458,
+      "step": 15395
+    },
+    {
+      "epoch": 11.589010161836658,
+      "grad_norm": 0.18203730881214142,
+      "learning_rate": 5e-06,
+      "loss": 0.0907,
+      "step": 15396
+    },
+    {
+      "epoch": 11.589762890477983,
+      "grad_norm": 0.19063138961791992,
+      "learning_rate": 5e-06,
+      "loss": 0.0894,
+      "step": 15397
+    },
+    {
+      "epoch": 11.590515619119307,
+      "grad_norm": 0.20591261982917786,
+      "learning_rate": 5e-06,
+      "loss": 0.1736,
+      "step": 15398
+    },
+    {
+      "epoch": 11.591268347760632,
+      "grad_norm": 0.19725321233272552,
+      "learning_rate": 5e-06,
+      "loss": 0.1461,
+      "step": 15399
+    },
+    {
+      "epoch": 11.592021076401958,
+      "grad_norm": 0.17848868668079376,
+      "learning_rate": 5e-06,
+      "loss": 0.1302,
+      "step": 15400
+    },
+    {
+      "epoch": 11.592021076401958,
+      "eval_loss": 0.184519425034523,
+      "eval_runtime": 452.84,
+      "eval_samples_per_second": 21.259,
+      "eval_steps_per_second": 0.665,
+      "step": 15400
+    },
+    {
+      "epoch": 11.592773805043281,
+      "grad_norm": 0.1770467758178711,
+      "learning_rate": 5e-06,
+      "loss": 0.1382,
+      "step": 15401
+    },
+    {
+      "epoch": 11.593526533684607,
+      "grad_norm": 0.1885230988264084,
+      "learning_rate": 5e-06,
+      "loss": 0.1759,
+      "step": 15402
+    },
+    {
+      "epoch": 11.594279262325932,
+      "grad_norm": 0.18475979566574097,
+      "learning_rate": 5e-06,
+      "loss": 0.1762,
+      "step": 15403
+    },
+    {
+      "epoch": 11.595031990967257,
+      "grad_norm": 0.18281026184558868,
+      "learning_rate": 5e-06,
+      "loss": 0.1833,
+      "step": 15404
+    },
+    {
+      "epoch": 11.59578471960858,
+      "grad_norm": 0.17484179139137268,
+      "learning_rate": 5e-06,
+      "loss": 0.0944,
+      "step": 15405
+    },
+    {
+      "epoch": 11.596537448249906,
+      "grad_norm": 0.17451901733875275,
+      "learning_rate": 5e-06,
+      "loss": 0.1674,
+      "step": 15406
+    },
+    {
+      "epoch": 11.597290176891232,
+      "grad_norm": 0.19784753024578094,
+      "learning_rate": 5e-06,
+      "loss": 0.1572,
+      "step": 15407
+    },
+    {
+      "epoch": 11.598042905532555,
+      "grad_norm": 0.17086459696292877,
+      "learning_rate": 5e-06,
+      "loss": 0.122,
+      "step": 15408
+    },
+    {
+      "epoch": 11.59879563417388,
+      "grad_norm": 0.1870698183774948,
+      "learning_rate": 5e-06,
+      "loss": 0.1578,
+      "step": 15409
+    },
+    {
+      "epoch": 11.599548362815206,
+      "grad_norm": 0.20618075132369995,
+      "learning_rate": 5e-06,
+      "loss": 0.1478,
+      "step": 15410
+    },
+    {
+      "epoch": 11.60030109145653,
+      "grad_norm": 0.17969143390655518,
+      "learning_rate": 5e-06,
+      "loss": 0.1326,
+      "step": 15411
+    },
+    {
+      "epoch": 11.601053820097855,
+      "grad_norm": 0.1959327906370163,
+      "learning_rate": 5e-06,
+      "loss": 0.1555,
+      "step": 15412
+    },
+    {
+      "epoch": 11.60180654873918,
+      "grad_norm": 0.16972585022449493,
+      "learning_rate": 5e-06,
+      "loss": 0.1698,
+      "step": 15413
+    },
+    {
+      "epoch": 11.602559277380504,
+      "grad_norm": 0.18458513915538788,
+      "learning_rate": 5e-06,
+      "loss": 0.159,
+      "step": 15414
+    },
+    {
+      "epoch": 11.603312006021829,
+      "grad_norm": 0.17035417258739471,
+      "learning_rate": 5e-06,
+      "loss": 0.1352,
+      "step": 15415
+    },
+    {
+      "epoch": 11.604064734663154,
+      "grad_norm": 0.17494098842144012,
+      "learning_rate": 5e-06,
+      "loss": 0.1827,
+      "step": 15416
+    },
+    {
+      "epoch": 11.604817463304478,
+      "grad_norm": 0.1963125765323639,
+      "learning_rate": 5e-06,
+      "loss": 0.1529,
+      "step": 15417
+    },
+    {
+      "epoch": 11.605570191945803,
+      "grad_norm": 0.1552955061197281,
+      "learning_rate": 5e-06,
+      "loss": 0.0935,
+      "step": 15418
+    },
+    {
+      "epoch": 11.606322920587129,
+      "grad_norm": 0.1876605600118637,
+      "learning_rate": 5e-06,
+      "loss": 0.1372,
+      "step": 15419
+    },
+    {
+      "epoch": 11.607075649228452,
+      "grad_norm": 0.1812496930360794,
+      "learning_rate": 5e-06,
+      "loss": 0.1066,
+      "step": 15420
+    },
+    {
+      "epoch": 11.607828377869778,
+      "grad_norm": 0.20558001101016998,
+      "learning_rate": 5e-06,
+      "loss": 0.0874,
+      "step": 15421
+    },
+    {
+      "epoch": 11.608581106511103,
+      "grad_norm": 0.20599739253520966,
+      "learning_rate": 5e-06,
+      "loss": 0.178,
+      "step": 15422
+    },
+    {
+      "epoch": 11.609333835152427,
+      "grad_norm": 0.15205317735671997,
+      "learning_rate": 5e-06,
+      "loss": 0.1767,
+      "step": 15423
+    },
+    {
+      "epoch": 11.610086563793752,
+      "grad_norm": 0.18490557372570038,
+      "learning_rate": 5e-06,
+      "loss": 0.1938,
+      "step": 15424
+    },
+    {
+      "epoch": 11.610839292435077,
+      "grad_norm": 0.2729277014732361,
+      "learning_rate": 5e-06,
+      "loss": 0.1562,
+      "step": 15425
+    },
+    {
+      "epoch": 11.611592021076403,
+      "grad_norm": 0.1720006763935089,
+      "learning_rate": 5e-06,
+      "loss": 0.0754,
+      "step": 15426
+    },
+    {
+      "epoch": 11.612344749717726,
+      "grad_norm": 0.15232069790363312,
+      "learning_rate": 5e-06,
+      "loss": 0.1077,
+      "step": 15427
+    },
+    {
+      "epoch": 11.613097478359052,
+      "grad_norm": 0.20620551705360413,
+      "learning_rate": 5e-06,
+      "loss": 0.1584,
+      "step": 15428
+    },
+    {
+      "epoch": 11.613850207000377,
+      "grad_norm": 0.18897350132465363,
+      "learning_rate": 5e-06,
+      "loss": 0.1451,
+      "step": 15429
+    },
+    {
+      "epoch": 11.6146029356417,
+      "grad_norm": 0.18916961550712585,
+      "learning_rate": 5e-06,
+      "loss": 0.1048,
+      "step": 15430
+    },
+    {
+      "epoch": 11.615355664283026,
+      "grad_norm": 0.18242789804935455,
+      "learning_rate": 5e-06,
+      "loss": 0.1756,
+      "step": 15431
+    },
+    {
+      "epoch": 11.616108392924351,
+      "grad_norm": 0.16444994509220123,
+      "learning_rate": 5e-06,
+      "loss": 0.1785,
+      "step": 15432
+    },
+    {
+      "epoch": 11.616861121565675,
+      "grad_norm": 0.20764142274856567,
+      "learning_rate": 5e-06,
+      "loss": 0.1356,
+      "step": 15433
+    },
+    {
+      "epoch": 11.617613850207,
+      "grad_norm": 0.18824665248394012,
+      "learning_rate": 5e-06,
+      "loss": 0.1632,
+      "step": 15434
+    },
+    {
+      "epoch": 11.618366578848326,
+      "grad_norm": 0.1928657442331314,
+      "learning_rate": 5e-06,
+      "loss": 0.1322,
+      "step": 15435
+    },
+    {
+      "epoch": 11.61911930748965,
+      "grad_norm": 0.17871585488319397,
+      "learning_rate": 5e-06,
+      "loss": 0.088,
+      "step": 15436
+    },
+    {
+      "epoch": 11.619872036130975,
+      "grad_norm": 0.18969765305519104,
+      "learning_rate": 5e-06,
+      "loss": 0.1143,
+      "step": 15437
+    },
+    {
+      "epoch": 11.6206247647723,
+      "grad_norm": 0.2159905582666397,
+      "learning_rate": 5e-06,
+      "loss": 0.1504,
+      "step": 15438
+    },
+    {
+      "epoch": 11.621377493413624,
+      "grad_norm": 0.186123326420784,
+      "learning_rate": 5e-06,
+      "loss": 0.0848,
+      "step": 15439
+    },
+    {
+      "epoch": 11.622130222054949,
+      "grad_norm": 0.2077920138835907,
+      "learning_rate": 5e-06,
+      "loss": 0.1766,
+      "step": 15440
+    },
+    {
+      "epoch": 11.622882950696274,
+      "grad_norm": 0.17837201058864594,
+      "learning_rate": 5e-06,
+      "loss": 0.1869,
+      "step": 15441
+    },
+    {
+      "epoch": 11.6236356793376,
+      "grad_norm": 0.15508364140987396,
+      "learning_rate": 5e-06,
+      "loss": 0.0581,
+      "step": 15442
+    },
+    {
+      "epoch": 11.624388407978923,
+      "grad_norm": 0.20882906019687653,
+      "learning_rate": 5e-06,
+      "loss": 0.1194,
+      "step": 15443
+    },
+    {
+      "epoch": 11.625141136620249,
+      "grad_norm": 0.18338294327259064,
+      "learning_rate": 5e-06,
+      "loss": 0.1877,
+      "step": 15444
+    },
+    {
+      "epoch": 11.625893865261574,
+      "grad_norm": 0.15343894064426422,
+      "learning_rate": 5e-06,
+      "loss": 0.1728,
+      "step": 15445
+    },
+    {
+      "epoch": 11.626646593902898,
+      "grad_norm": 0.18334418535232544,
+      "learning_rate": 5e-06,
+      "loss": 0.174,
+      "step": 15446
+    },
+    {
+      "epoch": 11.627399322544223,
+      "grad_norm": 0.16809000074863434,
+      "learning_rate": 5e-06,
+      "loss": 0.13,
+      "step": 15447
+    },
+    {
+      "epoch": 11.628152051185548,
+      "grad_norm": 0.16948620975017548,
+      "learning_rate": 5e-06,
+      "loss": 0.1372,
+      "step": 15448
+    },
+    {
+      "epoch": 11.628904779826872,
+      "grad_norm": 0.19818821549415588,
+      "learning_rate": 5e-06,
+      "loss": 0.219,
+      "step": 15449
+    },
+    {
+      "epoch": 11.629657508468197,
+      "grad_norm": 0.19659791886806488,
+      "learning_rate": 5e-06,
+      "loss": 0.1046,
+      "step": 15450
+    },
+    {
+      "epoch": 11.630410237109523,
+      "grad_norm": 0.1701478511095047,
+      "learning_rate": 5e-06,
+      "loss": 0.1086,
+      "step": 15451
+    },
+    {
+      "epoch": 11.631162965750846,
+      "grad_norm": 0.18667230010032654,
+      "learning_rate": 5e-06,
+      "loss": 0.1329,
+      "step": 15452
+    },
+    {
+      "epoch": 11.631915694392172,
+      "grad_norm": 0.1586209088563919,
+      "learning_rate": 5e-06,
+      "loss": 0.1399,
+      "step": 15453
+    },
+    {
+      "epoch": 11.632668423033497,
+      "grad_norm": 0.1644863784313202,
+      "learning_rate": 5e-06,
+      "loss": 0.1289,
+      "step": 15454
+    },
+    {
+      "epoch": 11.63342115167482,
+      "grad_norm": 0.188917875289917,
+      "learning_rate": 5e-06,
+      "loss": 0.2154,
+      "step": 15455
+    },
+    {
+      "epoch": 11.634173880316146,
+      "grad_norm": 0.17432498931884766,
+      "learning_rate": 5e-06,
+      "loss": 0.1711,
+      "step": 15456
+    },
+    {
+      "epoch": 11.634926608957471,
+      "grad_norm": 0.20095373690128326,
+      "learning_rate": 5e-06,
+      "loss": 0.1284,
+      "step": 15457
+    },
+    {
+      "epoch": 11.635679337598795,
+      "grad_norm": 0.1630779206752777,
+      "learning_rate": 5e-06,
+      "loss": 0.0663,
+      "step": 15458
+    },
+    {
+      "epoch": 11.63643206624012,
+      "grad_norm": 0.17537862062454224,
+      "learning_rate": 5e-06,
+      "loss": 0.1038,
+      "step": 15459
+    },
+    {
+      "epoch": 11.637184794881446,
+      "grad_norm": 0.20606105029582977,
+      "learning_rate": 5e-06,
+      "loss": 0.0989,
+      "step": 15460
+    },
+    {
+      "epoch": 11.63793752352277,
+      "grad_norm": 0.2007083147764206,
+      "learning_rate": 5e-06,
+      "loss": 0.1435,
+      "step": 15461
+    },
+    {
+      "epoch": 11.638690252164094,
+      "grad_norm": 0.19006043672561646,
+      "learning_rate": 5e-06,
+      "loss": 0.144,
+      "step": 15462
+    },
+    {
+      "epoch": 11.63944298080542,
+      "grad_norm": 0.15554435551166534,
+      "learning_rate": 5e-06,
+      "loss": 0.1147,
+      "step": 15463
+    },
+    {
+      "epoch": 11.640195709446745,
+      "grad_norm": 0.19429460167884827,
+      "learning_rate": 5e-06,
+      "loss": 0.1596,
+      "step": 15464
+    },
+    {
+      "epoch": 11.640948438088069,
+      "grad_norm": 0.15658637881278992,
+      "learning_rate": 5e-06,
+      "loss": 0.1035,
+      "step": 15465
+    },
+    {
+      "epoch": 11.641701166729394,
+      "grad_norm": 0.15614710748195648,
+      "learning_rate": 5e-06,
+      "loss": 0.1312,
+      "step": 15466
+    },
+    {
+      "epoch": 11.64245389537072,
+      "grad_norm": 0.18431805074214935,
+      "learning_rate": 5e-06,
+      "loss": 0.1362,
+      "step": 15467
+    },
+    {
+      "epoch": 11.643206624012043,
+      "grad_norm": 0.16690020263195038,
+      "learning_rate": 5e-06,
+      "loss": 0.1104,
+      "step": 15468
+    },
+    {
+      "epoch": 11.643959352653368,
+      "grad_norm": 0.14547297358512878,
+      "learning_rate": 5e-06,
+      "loss": 0.1385,
+      "step": 15469
+    },
+    {
+      "epoch": 11.644712081294694,
+      "grad_norm": 0.1923530548810959,
+      "learning_rate": 5e-06,
+      "loss": 0.1485,
+      "step": 15470
+    },
+    {
+      "epoch": 11.645464809936017,
+      "grad_norm": 0.17798587679862976,
+      "learning_rate": 5e-06,
+      "loss": 0.1145,
+      "step": 15471
+    },
+    {
+      "epoch": 11.646217538577343,
+      "grad_norm": 0.17260752618312836,
+      "learning_rate": 5e-06,
+      "loss": 0.1276,
+      "step": 15472
+    },
+    {
+      "epoch": 11.646970267218668,
+      "grad_norm": 0.19197729229927063,
+      "learning_rate": 5e-06,
+      "loss": 0.2239,
+      "step": 15473
+    },
+    {
+      "epoch": 11.647722995859992,
+      "grad_norm": 0.16841478645801544,
+      "learning_rate": 5e-06,
+      "loss": 0.1199,
+      "step": 15474
+    },
+    {
+      "epoch": 11.648475724501317,
+      "grad_norm": 0.18434715270996094,
+      "learning_rate": 5e-06,
+      "loss": 0.1476,
+      "step": 15475
+    },
+    {
+      "epoch": 11.649228453142642,
+      "grad_norm": 0.1724986881017685,
+      "learning_rate": 5e-06,
+      "loss": 0.1247,
+      "step": 15476
+    },
+    {
+      "epoch": 11.649981181783968,
+      "grad_norm": 0.19501636922359467,
+      "learning_rate": 5e-06,
+      "loss": 0.1892,
+      "step": 15477
+    },
+    {
+      "epoch": 11.650733910425291,
+      "grad_norm": 0.18086151778697968,
+      "learning_rate": 5e-06,
+      "loss": 0.1681,
+      "step": 15478
+    },
+    {
+      "epoch": 11.651486639066617,
+      "grad_norm": 0.2069239467382431,
+      "learning_rate": 5e-06,
+      "loss": 0.1322,
+      "step": 15479
+    },
+    {
+      "epoch": 11.652239367707942,
+      "grad_norm": 0.1832161396741867,
+      "learning_rate": 5e-06,
+      "loss": 0.1724,
+      "step": 15480
+    },
+    {
+      "epoch": 11.652992096349266,
+      "grad_norm": 0.19423627853393555,
+      "learning_rate": 5e-06,
+      "loss": 0.1676,
+      "step": 15481
+    },
+    {
+      "epoch": 11.653744824990591,
+      "grad_norm": 0.1789165586233139,
+      "learning_rate": 5e-06,
+      "loss": 0.0849,
+      "step": 15482
+    },
+    {
+      "epoch": 11.654497553631916,
+      "grad_norm": 0.198493093252182,
+      "learning_rate": 5e-06,
+      "loss": 0.1444,
+      "step": 15483
+    },
+    {
+      "epoch": 11.65525028227324,
+      "grad_norm": 0.18653394281864166,
+      "learning_rate": 5e-06,
+      "loss": 0.1294,
+      "step": 15484
+    },
+    {
+      "epoch": 11.656003010914565,
+      "grad_norm": 0.16049379110336304,
+      "learning_rate": 5e-06,
+      "loss": 0.0777,
+      "step": 15485
+    },
+    {
+      "epoch": 11.65675573955589,
+      "grad_norm": 0.1787497103214264,
+      "learning_rate": 5e-06,
+      "loss": 0.1325,
+      "step": 15486
+    },
+    {
+      "epoch": 11.657508468197214,
+      "grad_norm": 0.19684284925460815,
+      "learning_rate": 5e-06,
+      "loss": 0.1746,
+      "step": 15487
+    },
+    {
+      "epoch": 11.65826119683854,
+      "grad_norm": 0.2038959264755249,
+      "learning_rate": 5e-06,
+      "loss": 0.1477,
+      "step": 15488
+    },
+    {
+      "epoch": 11.659013925479865,
+      "grad_norm": 0.17235489189624786,
+      "learning_rate": 5e-06,
+      "loss": 0.2136,
+      "step": 15489
+    },
+    {
+      "epoch": 11.659766654121189,
+      "grad_norm": 0.15359064936637878,
+      "learning_rate": 5e-06,
+      "loss": 0.1192,
+      "step": 15490
+    },
+    {
+      "epoch": 11.660519382762514,
+      "grad_norm": 0.16667526960372925,
+      "learning_rate": 5e-06,
+      "loss": 0.0914,
+      "step": 15491
+    },
+    {
+      "epoch": 11.66127211140384,
+      "grad_norm": 0.1708340346813202,
+      "learning_rate": 5e-06,
+      "loss": 0.158,
+      "step": 15492
+    },
+    {
+      "epoch": 11.662024840045163,
+      "grad_norm": 0.20591013133525848,
+      "learning_rate": 5e-06,
+      "loss": 0.1411,
+      "step": 15493
+    },
+    {
+      "epoch": 11.662777568686488,
+      "grad_norm": 0.1732543557882309,
+      "learning_rate": 5e-06,
+      "loss": 0.1045,
+      "step": 15494
+    },
+    {
+      "epoch": 11.663530297327814,
+      "grad_norm": 0.18148860335350037,
+      "learning_rate": 5e-06,
+      "loss": 0.1176,
+      "step": 15495
+    },
+    {
+      "epoch": 11.664283025969137,
+      "grad_norm": 0.23467113077640533,
+      "learning_rate": 5e-06,
+      "loss": 0.1911,
+      "step": 15496
+    },
+    {
+      "epoch": 11.665035754610463,
+      "grad_norm": 0.18429775536060333,
+      "learning_rate": 5e-06,
+      "loss": 0.1085,
+      "step": 15497
+    },
+    {
+      "epoch": 11.665788483251788,
+      "grad_norm": 0.18102730810642242,
+      "learning_rate": 5e-06,
+      "loss": 0.1459,
+      "step": 15498
+    },
+    {
+      "epoch": 11.666541211893113,
+      "grad_norm": 0.17655785381793976,
+      "learning_rate": 5e-06,
+      "loss": 0.1975,
+      "step": 15499
+    },
+    {
+      "epoch": 11.667293940534437,
+      "grad_norm": 0.16414274275302887,
+      "learning_rate": 5e-06,
+      "loss": 0.0967,
+      "step": 15500
+    },
+    {
+      "epoch": 11.668046669175762,
+      "grad_norm": 0.16870880126953125,
+      "learning_rate": 5e-06,
+      "loss": 0.1081,
+      "step": 15501
+    },
+    {
+      "epoch": 11.668799397817088,
+      "grad_norm": 0.1731245070695877,
+      "learning_rate": 5e-06,
+      "loss": 0.1218,
+      "step": 15502
+    },
+    {
+      "epoch": 11.669552126458411,
+      "grad_norm": 0.18523825705051422,
+      "learning_rate": 5e-06,
+      "loss": 0.1238,
+      "step": 15503
+    },
+    {
+      "epoch": 11.670304855099737,
+      "grad_norm": 0.223882794380188,
+      "learning_rate": 5e-06,
+      "loss": 0.137,
+      "step": 15504
+    },
+    {
+      "epoch": 11.671057583741062,
+      "grad_norm": 0.17244453728199005,
+      "learning_rate": 5e-06,
+      "loss": 0.1277,
+      "step": 15505
+    },
+    {
+      "epoch": 11.671810312382386,
+      "grad_norm": 0.19342315196990967,
+      "learning_rate": 5e-06,
+      "loss": 0.1283,
+      "step": 15506
+    },
+    {
+      "epoch": 11.672563041023711,
+      "grad_norm": 0.194451704621315,
+      "learning_rate": 5e-06,
+      "loss": 0.1159,
+      "step": 15507
+    },
+    {
+      "epoch": 11.673315769665036,
+      "grad_norm": 0.1711118221282959,
+      "learning_rate": 5e-06,
+      "loss": 0.1874,
+      "step": 15508
+    },
+    {
+      "epoch": 11.67406849830636,
+      "grad_norm": 0.1676807999610901,
+      "learning_rate": 5e-06,
+      "loss": 0.1054,
+      "step": 15509
+    },
+    {
+      "epoch": 11.674821226947685,
+      "grad_norm": 0.20052528381347656,
+      "learning_rate": 5e-06,
+      "loss": 0.1409,
+      "step": 15510
+    },
+    {
+      "epoch": 11.67557395558901,
+      "grad_norm": 0.20358166098594666,
+      "learning_rate": 5e-06,
+      "loss": 0.1002,
+      "step": 15511
+    },
+    {
+      "epoch": 11.676326684230334,
+      "grad_norm": 0.15990808606147766,
+      "learning_rate": 5e-06,
+      "loss": 0.154,
+      "step": 15512
+    },
+    {
+      "epoch": 11.67707941287166,
+      "grad_norm": 0.19407868385314941,
+      "learning_rate": 5e-06,
+      "loss": 0.1612,
+      "step": 15513
+    },
+    {
+      "epoch": 11.677832141512985,
+      "grad_norm": 0.19694986939430237,
+      "learning_rate": 5e-06,
+      "loss": 0.1302,
+      "step": 15514
+    },
+    {
+      "epoch": 11.67858487015431,
+      "grad_norm": 0.1957172453403473,
+      "learning_rate": 5e-06,
+      "loss": 0.2224,
+      "step": 15515
+    },
+    {
+      "epoch": 11.679337598795634,
+      "grad_norm": 0.18297751247882843,
+      "learning_rate": 5e-06,
+      "loss": 0.1721,
+      "step": 15516
+    },
+    {
+      "epoch": 11.68009032743696,
+      "grad_norm": 0.17224961519241333,
+      "learning_rate": 5e-06,
+      "loss": 0.1488,
+      "step": 15517
+    },
+    {
+      "epoch": 11.680843056078285,
+      "grad_norm": 0.17011846601963043,
+      "learning_rate": 5e-06,
+      "loss": 0.0959,
+      "step": 15518
+    },
+    {
+      "epoch": 11.681595784719608,
+      "grad_norm": 0.1765829622745514,
+      "learning_rate": 5e-06,
+      "loss": 0.1934,
+      "step": 15519
+    },
+    {
+      "epoch": 11.682348513360933,
+      "grad_norm": 0.1903749257326126,
+      "learning_rate": 5e-06,
+      "loss": 0.1297,
+      "step": 15520
+    },
+    {
+      "epoch": 11.683101242002259,
+      "grad_norm": 0.1730547547340393,
+      "learning_rate": 5e-06,
+      "loss": 0.09,
+      "step": 15521
+    },
+    {
+      "epoch": 11.683853970643582,
+      "grad_norm": 0.19153964519500732,
+      "learning_rate": 5e-06,
+      "loss": 0.213,
+      "step": 15522
+    },
+    {
+      "epoch": 11.684606699284908,
+      "grad_norm": 0.16519446671009064,
+      "learning_rate": 5e-06,
+      "loss": 0.1397,
+      "step": 15523
+    },
+    {
+      "epoch": 11.685359427926233,
+      "grad_norm": 0.16788940131664276,
+      "learning_rate": 5e-06,
+      "loss": 0.131,
+      "step": 15524
+    },
+    {
+      "epoch": 11.686112156567557,
+      "grad_norm": 0.22099533677101135,
+      "learning_rate": 5e-06,
+      "loss": 0.1584,
+      "step": 15525
+    },
+    {
+      "epoch": 11.686864885208882,
+      "grad_norm": 0.18912473320960999,
+      "learning_rate": 5e-06,
+      "loss": 0.2039,
+      "step": 15526
+    },
+    {
+      "epoch": 11.687617613850207,
+      "grad_norm": 0.17561458051204681,
+      "learning_rate": 5e-06,
+      "loss": 0.0968,
+      "step": 15527
+    },
+    {
+      "epoch": 11.688370342491531,
+      "grad_norm": 0.17670504748821259,
+      "learning_rate": 5e-06,
+      "loss": 0.1263,
+      "step": 15528
+    },
+    {
+      "epoch": 11.689123071132856,
+      "grad_norm": 0.1823635846376419,
+      "learning_rate": 5e-06,
+      "loss": 0.0979,
+      "step": 15529
+    },
+    {
+      "epoch": 11.689875799774182,
+      "grad_norm": 0.21820251643657684,
+      "learning_rate": 5e-06,
+      "loss": 0.1496,
+      "step": 15530
+    },
+    {
+      "epoch": 11.690628528415505,
+      "grad_norm": 0.18003162741661072,
+      "learning_rate": 5e-06,
+      "loss": 0.1482,
+      "step": 15531
+    },
+    {
+      "epoch": 11.69138125705683,
+      "grad_norm": 0.18650594353675842,
+      "learning_rate": 5e-06,
+      "loss": 0.1688,
+      "step": 15532
+    },
+    {
+      "epoch": 11.692133985698156,
+      "grad_norm": 0.2237236499786377,
+      "learning_rate": 5e-06,
+      "loss": 0.1444,
+      "step": 15533
+    },
+    {
+      "epoch": 11.69288671433948,
+      "grad_norm": 0.1900700181722641,
+      "learning_rate": 5e-06,
+      "loss": 0.1878,
+      "step": 15534
+    },
+    {
+      "epoch": 11.693639442980805,
+      "grad_norm": 0.1752176582813263,
+      "learning_rate": 5e-06,
+      "loss": 0.0789,
+      "step": 15535
+    },
+    {
+      "epoch": 11.69439217162213,
+      "grad_norm": 0.17529338598251343,
+      "learning_rate": 5e-06,
+      "loss": 0.1069,
+      "step": 15536
+    },
+    {
+      "epoch": 11.695144900263456,
+      "grad_norm": 0.15317043662071228,
+      "learning_rate": 5e-06,
+      "loss": 0.0955,
+      "step": 15537
+    },
+    {
+      "epoch": 11.69589762890478,
+      "grad_norm": 0.2130090594291687,
+      "learning_rate": 5e-06,
+      "loss": 0.1403,
+      "step": 15538
+    },
+    {
+      "epoch": 11.696650357546105,
+      "grad_norm": 0.16958199441432953,
+      "learning_rate": 5e-06,
+      "loss": 0.0692,
+      "step": 15539
+    },
+    {
+      "epoch": 11.69740308618743,
+      "grad_norm": 0.18320316076278687,
+      "learning_rate": 5e-06,
+      "loss": 0.1287,
+      "step": 15540
+    },
+    {
+      "epoch": 11.698155814828754,
+      "grad_norm": 0.1916675567626953,
+      "learning_rate": 5e-06,
+      "loss": 0.1617,
+      "step": 15541
+    },
+    {
+      "epoch": 11.698908543470079,
+      "grad_norm": 0.21558815240859985,
+      "learning_rate": 5e-06,
+      "loss": 0.1401,
+      "step": 15542
+    },
+    {
+      "epoch": 11.699661272111404,
+      "grad_norm": 0.1854674369096756,
+      "learning_rate": 5e-06,
+      "loss": 0.1382,
+      "step": 15543
+    },
+    {
+      "epoch": 11.700414000752728,
+      "grad_norm": 0.21162179112434387,
+      "learning_rate": 5e-06,
+      "loss": 0.1775,
+      "step": 15544
+    },
+    {
+      "epoch": 11.701166729394053,
+      "grad_norm": 0.173762246966362,
+      "learning_rate": 5e-06,
+      "loss": 0.1035,
+      "step": 15545
+    },
+    {
+      "epoch": 11.701919458035379,
+      "grad_norm": 0.19828084111213684,
+      "learning_rate": 5e-06,
+      "loss": 0.1854,
+      "step": 15546
+    },
+    {
+      "epoch": 11.702672186676702,
+      "grad_norm": 0.18815584480762482,
+      "learning_rate": 5e-06,
+      "loss": 0.073,
+      "step": 15547
+    },
+    {
+      "epoch": 11.703424915318028,
+      "grad_norm": 0.18528494238853455,
+      "learning_rate": 5e-06,
+      "loss": 0.1567,
+      "step": 15548
+    },
+    {
+      "epoch": 11.704177643959353,
+      "grad_norm": 0.17168088257312775,
+      "learning_rate": 5e-06,
+      "loss": 0.1143,
+      "step": 15549
+    },
+    {
+      "epoch": 11.704930372600678,
+      "grad_norm": 0.23440171778202057,
+      "learning_rate": 5e-06,
+      "loss": 0.1792,
+      "step": 15550
+    },
+    {
+      "epoch": 11.705683101242002,
+      "grad_norm": 0.17481622099876404,
+      "learning_rate": 5e-06,
+      "loss": 0.0941,
+      "step": 15551
+    },
+    {
+      "epoch": 11.706435829883327,
+      "grad_norm": 0.19086851179599762,
+      "learning_rate": 5e-06,
+      "loss": 0.1139,
+      "step": 15552
+    },
+    {
+      "epoch": 11.707188558524653,
+      "grad_norm": 0.21640782058238983,
+      "learning_rate": 5e-06,
+      "loss": 0.146,
+      "step": 15553
+    },
+    {
+      "epoch": 11.707941287165976,
+      "grad_norm": 0.19290539622306824,
+      "learning_rate": 5e-06,
+      "loss": 0.1639,
+      "step": 15554
+    },
+    {
+      "epoch": 11.708694015807302,
+      "grad_norm": 0.18196424841880798,
+      "learning_rate": 5e-06,
+      "loss": 0.1278,
+      "step": 15555
+    },
+    {
+      "epoch": 11.709446744448627,
+      "grad_norm": 0.16895972192287445,
+      "learning_rate": 5e-06,
+      "loss": 0.1516,
+      "step": 15556
+    },
+    {
+      "epoch": 11.71019947308995,
+      "grad_norm": 0.1935516744852066,
+      "learning_rate": 5e-06,
+      "loss": 0.1514,
+      "step": 15557
+    },
+    {
+      "epoch": 11.710952201731276,
+      "grad_norm": 0.18644635379314423,
+      "learning_rate": 5e-06,
+      "loss": 0.2182,
+      "step": 15558
+    },
+    {
+      "epoch": 11.711704930372601,
+      "grad_norm": 0.16882053017616272,
+      "learning_rate": 5e-06,
+      "loss": 0.1098,
+      "step": 15559
+    },
+    {
+      "epoch": 11.712457659013925,
+      "grad_norm": 0.1929585188627243,
+      "learning_rate": 5e-06,
+      "loss": 0.1238,
+      "step": 15560
+    },
+    {
+      "epoch": 11.71321038765525,
+      "grad_norm": 0.2196321189403534,
+      "learning_rate": 5e-06,
+      "loss": 0.1865,
+      "step": 15561
+    },
+    {
+      "epoch": 11.713963116296576,
+      "grad_norm": 0.18753093481063843,
+      "learning_rate": 5e-06,
+      "loss": 0.2335,
+      "step": 15562
+    },
+    {
+      "epoch": 11.7147158449379,
+      "grad_norm": 0.17857849597930908,
+      "learning_rate": 5e-06,
+      "loss": 0.1313,
+      "step": 15563
+    },
+    {
+      "epoch": 11.715468573579225,
+      "grad_norm": 0.17243574559688568,
+      "learning_rate": 5e-06,
+      "loss": 0.1059,
+      "step": 15564
+    },
+    {
+      "epoch": 11.71622130222055,
+      "grad_norm": 0.1716749519109726,
+      "learning_rate": 5e-06,
+      "loss": 0.1022,
+      "step": 15565
+    },
+    {
+      "epoch": 11.716974030861874,
+      "grad_norm": 0.18808797001838684,
+      "learning_rate": 5e-06,
+      "loss": 0.1593,
+      "step": 15566
+    },
+    {
+      "epoch": 11.717726759503199,
+      "grad_norm": 0.17482790350914001,
+      "learning_rate": 5e-06,
+      "loss": 0.1605,
+      "step": 15567
+    },
+    {
+      "epoch": 11.718479488144524,
+      "grad_norm": 0.1997700035572052,
+      "learning_rate": 5e-06,
+      "loss": 0.1523,
+      "step": 15568
+    },
+    {
+      "epoch": 11.719232216785848,
+      "grad_norm": 0.18126720190048218,
+      "learning_rate": 5e-06,
+      "loss": 0.1566,
+      "step": 15569
+    },
+    {
+      "epoch": 11.719984945427173,
+      "grad_norm": 0.1888725310564041,
+      "learning_rate": 5e-06,
+      "loss": 0.1269,
+      "step": 15570
+    },
+    {
+      "epoch": 11.720737674068499,
+      "grad_norm": 0.18215471506118774,
+      "learning_rate": 5e-06,
+      "loss": 0.1298,
+      "step": 15571
+    },
+    {
+      "epoch": 11.721490402709824,
+      "grad_norm": 0.17272886633872986,
+      "learning_rate": 5e-06,
+      "loss": 0.1561,
+      "step": 15572
+    },
+    {
+      "epoch": 11.722243131351147,
+      "grad_norm": 0.21529695391654968,
+      "learning_rate": 5e-06,
+      "loss": 0.158,
+      "step": 15573
+    },
+    {
+      "epoch": 11.722995859992473,
+      "grad_norm": 0.19245275855064392,
+      "learning_rate": 5e-06,
+      "loss": 0.0952,
+      "step": 15574
+    },
+    {
+      "epoch": 11.723748588633798,
+      "grad_norm": 0.16165533661842346,
+      "learning_rate": 5e-06,
+      "loss": 0.1924,
+      "step": 15575
+    },
+    {
+      "epoch": 11.724501317275122,
+      "grad_norm": 0.2007397562265396,
+      "learning_rate": 5e-06,
+      "loss": 0.1192,
+      "step": 15576
+    },
+    {
+      "epoch": 11.725254045916447,
+      "grad_norm": 0.18169689178466797,
+      "learning_rate": 5e-06,
+      "loss": 0.1827,
+      "step": 15577
+    },
+    {
+      "epoch": 11.726006774557773,
+      "grad_norm": 0.16563443839550018,
+      "learning_rate": 5e-06,
+      "loss": 0.1663,
+      "step": 15578
+    },
+    {
+      "epoch": 11.726759503199096,
+      "grad_norm": 0.20146793127059937,
+      "learning_rate": 5e-06,
+      "loss": 0.1617,
+      "step": 15579
+    },
+    {
+      "epoch": 11.727512231840421,
+      "grad_norm": 0.2126997411251068,
+      "learning_rate": 5e-06,
+      "loss": 0.234,
+      "step": 15580
+    },
+    {
+      "epoch": 11.728264960481747,
+      "grad_norm": 0.170682892203331,
+      "learning_rate": 5e-06,
+      "loss": 0.0794,
+      "step": 15581
+    },
+    {
+      "epoch": 11.72901768912307,
+      "grad_norm": 0.1665184050798416,
+      "learning_rate": 5e-06,
+      "loss": 0.17,
+      "step": 15582
+    },
+    {
+      "epoch": 11.729770417764396,
+      "grad_norm": 0.1951136589050293,
+      "learning_rate": 5e-06,
+      "loss": 0.14,
+      "step": 15583
+    },
+    {
+      "epoch": 11.730523146405721,
+      "grad_norm": 0.18755675852298737,
+      "learning_rate": 5e-06,
+      "loss": 0.1787,
+      "step": 15584
+    },
+    {
+      "epoch": 11.731275875047045,
+      "grad_norm": 0.180768683552742,
+      "learning_rate": 5e-06,
+      "loss": 0.143,
+      "step": 15585
+    },
+    {
+      "epoch": 11.73202860368837,
+      "grad_norm": 0.17257797718048096,
+      "learning_rate": 5e-06,
+      "loss": 0.1175,
+      "step": 15586
+    },
+    {
+      "epoch": 11.732781332329695,
+      "grad_norm": 0.16294975578784943,
+      "learning_rate": 5e-06,
+      "loss": 0.1179,
+      "step": 15587
+    },
+    {
+      "epoch": 11.73353406097102,
+      "grad_norm": 0.19438450038433075,
+      "learning_rate": 5e-06,
+      "loss": 0.1283,
+      "step": 15588
+    },
+    {
+      "epoch": 11.734286789612344,
+      "grad_norm": 0.21745319664478302,
+      "learning_rate": 5e-06,
+      "loss": 0.1553,
+      "step": 15589
+    },
+    {
+      "epoch": 11.73503951825367,
+      "grad_norm": 0.16183842718601227,
+      "learning_rate": 5e-06,
+      "loss": 0.0783,
+      "step": 15590
+    },
+    {
+      "epoch": 11.735792246894995,
+      "grad_norm": 0.17426908016204834,
+      "learning_rate": 5e-06,
+      "loss": 0.1264,
+      "step": 15591
+    },
+    {
+      "epoch": 11.736544975536319,
+      "grad_norm": 0.20654408633708954,
+      "learning_rate": 5e-06,
+      "loss": 0.1582,
+      "step": 15592
+    },
+    {
+      "epoch": 11.737297704177644,
+      "grad_norm": 0.19315728545188904,
+      "learning_rate": 5e-06,
+      "loss": 0.1776,
+      "step": 15593
+    },
+    {
+      "epoch": 11.73805043281897,
+      "grad_norm": 0.18763332068920135,
+      "learning_rate": 5e-06,
+      "loss": 0.153,
+      "step": 15594
+    },
+    {
+      "epoch": 11.738803161460293,
+      "grad_norm": 0.16840243339538574,
+      "learning_rate": 5e-06,
+      "loss": 0.0869,
+      "step": 15595
+    },
+    {
+      "epoch": 11.739555890101618,
+      "grad_norm": 0.18071962893009186,
+      "learning_rate": 5e-06,
+      "loss": 0.1342,
+      "step": 15596
+    },
+    {
+      "epoch": 11.740308618742944,
+      "grad_norm": 0.14329023659229279,
+      "learning_rate": 5e-06,
+      "loss": 0.0966,
+      "step": 15597
+    },
+    {
+      "epoch": 11.741061347384267,
+      "grad_norm": 0.16430258750915527,
+      "learning_rate": 5e-06,
+      "loss": 0.1256,
+      "step": 15598
+    },
+    {
+      "epoch": 11.741814076025593,
+      "grad_norm": 0.1880432814359665,
+      "learning_rate": 5e-06,
+      "loss": 0.1232,
+      "step": 15599
+    },
+    {
+      "epoch": 11.742566804666918,
+      "grad_norm": 0.19535793364048004,
+      "learning_rate": 5e-06,
+      "loss": 0.2372,
+      "step": 15600
+    },
+    {
+      "epoch": 11.742566804666918,
+      "eval_loss": 0.18495379388332367,
+      "eval_runtime": 452.8834,
+      "eval_samples_per_second": 21.257,
+      "eval_steps_per_second": 0.665,
+      "step": 15600
+    },
+    {
+      "epoch": 11.743319533308242,
+      "grad_norm": 0.18971046805381775,
+      "learning_rate": 5e-06,
+      "loss": 0.1526,
+      "step": 15601
+    },
+    {
+      "epoch": 11.744072261949567,
+      "grad_norm": 0.1849757730960846,
+      "learning_rate": 5e-06,
+      "loss": 0.1452,
+      "step": 15602
+    },
+    {
+      "epoch": 11.744824990590892,
+      "grad_norm": 0.16755801439285278,
+      "learning_rate": 5e-06,
+      "loss": 0.0932,
+      "step": 15603
+    },
+    {
+      "epoch": 11.745577719232216,
+      "grad_norm": 0.1864762306213379,
+      "learning_rate": 5e-06,
+      "loss": 0.1256,
+      "step": 15604
+    },
+    {
+      "epoch": 11.746330447873541,
+      "grad_norm": 0.21144597232341766,
+      "learning_rate": 5e-06,
+      "loss": 0.1155,
+      "step": 15605
+    },
+    {
+      "epoch": 11.747083176514867,
+      "grad_norm": 0.187644362449646,
+      "learning_rate": 5e-06,
+      "loss": 0.1465,
+      "step": 15606
+    },
+    {
+      "epoch": 11.74783590515619,
+      "grad_norm": 0.18690164387226105,
+      "learning_rate": 5e-06,
+      "loss": 0.097,
+      "step": 15607
+    },
+    {
+      "epoch": 11.748588633797516,
+      "grad_norm": 0.1786900758743286,
+      "learning_rate": 5e-06,
+      "loss": 0.0696,
+      "step": 15608
+    },
+    {
+      "epoch": 11.749341362438841,
+      "grad_norm": 0.18185625970363617,
+      "learning_rate": 5e-06,
+      "loss": 0.162,
+      "step": 15609
+    },
+    {
+      "epoch": 11.750094091080166,
+      "grad_norm": 0.1942054033279419,
+      "learning_rate": 5e-06,
+      "loss": 0.172,
+      "step": 15610
+    },
+    {
+      "epoch": 11.75084681972149,
+      "grad_norm": 0.1844143122434616,
+      "learning_rate": 5e-06,
+      "loss": 0.2223,
+      "step": 15611
+    },
+    {
+      "epoch": 11.751599548362815,
+      "grad_norm": 0.19657015800476074,
+      "learning_rate": 5e-06,
+      "loss": 0.1318,
+      "step": 15612
+    },
+    {
+      "epoch": 11.75235227700414,
+      "grad_norm": 0.179269477725029,
+      "learning_rate": 5e-06,
+      "loss": 0.0685,
+      "step": 15613
+    },
+    {
+      "epoch": 11.753105005645464,
+      "grad_norm": 0.17993168532848358,
+      "learning_rate": 5e-06,
+      "loss": 0.1018,
+      "step": 15614
+    },
+    {
+      "epoch": 11.75385773428679,
+      "grad_norm": 0.18037398159503937,
+      "learning_rate": 5e-06,
+      "loss": 0.0981,
+      "step": 15615
+    },
+    {
+      "epoch": 11.754610462928115,
+      "grad_norm": 0.17822374403476715,
+      "learning_rate": 5e-06,
+      "loss": 0.169,
+      "step": 15616
+    },
+    {
+      "epoch": 11.755363191569439,
+      "grad_norm": 0.17944598197937012,
+      "learning_rate": 5e-06,
+      "loss": 0.1812,
+      "step": 15617
+    },
+    {
+      "epoch": 11.756115920210764,
+      "grad_norm": 0.18213459849357605,
+      "learning_rate": 5e-06,
+      "loss": 0.1166,
+      "step": 15618
+    },
+    {
+      "epoch": 11.75686864885209,
+      "grad_norm": 0.16632820665836334,
+      "learning_rate": 5e-06,
+      "loss": 0.1309,
+      "step": 15619
+    },
+    {
+      "epoch": 11.757621377493413,
+      "grad_norm": 0.1740647256374359,
+      "learning_rate": 5e-06,
+      "loss": 0.2267,
+      "step": 15620
+    },
+    {
+      "epoch": 11.758374106134738,
+      "grad_norm": 0.17106883227825165,
+      "learning_rate": 5e-06,
+      "loss": 0.1561,
+      "step": 15621
+    },
+    {
+      "epoch": 11.759126834776064,
+      "grad_norm": 0.1898563951253891,
+      "learning_rate": 5e-06,
+      "loss": 0.1254,
+      "step": 15622
+    },
+    {
+      "epoch": 11.759879563417389,
+      "grad_norm": 0.19887635111808777,
+      "learning_rate": 5e-06,
+      "loss": 0.1068,
+      "step": 15623
+    },
+    {
+      "epoch": 11.760632292058713,
+      "grad_norm": 0.19003926217556,
+      "learning_rate": 5e-06,
+      "loss": 0.1296,
+      "step": 15624
+    },
+    {
+      "epoch": 11.761385020700038,
+      "grad_norm": 0.16996324062347412,
+      "learning_rate": 5e-06,
+      "loss": 0.0729,
+      "step": 15625
+    },
+    {
+      "epoch": 11.762137749341363,
+      "grad_norm": 0.1996210515499115,
+      "learning_rate": 5e-06,
+      "loss": 0.1677,
+      "step": 15626
+    },
+    {
+      "epoch": 11.762890477982687,
+      "grad_norm": 0.19581466913223267,
+      "learning_rate": 5e-06,
+      "loss": 0.1038,
+      "step": 15627
+    },
+    {
+      "epoch": 11.763643206624012,
+      "grad_norm": 0.17972643673419952,
+      "learning_rate": 5e-06,
+      "loss": 0.1148,
+      "step": 15628
+    },
+    {
+      "epoch": 11.764395935265338,
+      "grad_norm": 0.17396783828735352,
+      "learning_rate": 5e-06,
+      "loss": 0.193,
+      "step": 15629
+    },
+    {
+      "epoch": 11.765148663906661,
+      "grad_norm": 0.15227089822292328,
+      "learning_rate": 5e-06,
+      "loss": 0.0801,
+      "step": 15630
+    },
+    {
+      "epoch": 11.765901392547987,
+      "grad_norm": 0.1668006330728531,
+      "learning_rate": 5e-06,
+      "loss": 0.0904,
+      "step": 15631
+    },
+    {
+      "epoch": 11.766654121189312,
+      "grad_norm": 0.18117259442806244,
+      "learning_rate": 5e-06,
+      "loss": 0.2505,
+      "step": 15632
+    },
+    {
+      "epoch": 11.767406849830635,
+      "grad_norm": 0.18638642132282257,
+      "learning_rate": 5e-06,
+      "loss": 0.1468,
+      "step": 15633
+    },
+    {
+      "epoch": 11.76815957847196,
+      "grad_norm": 0.18147650361061096,
+      "learning_rate": 5e-06,
+      "loss": 0.1517,
+      "step": 15634
+    },
+    {
+      "epoch": 11.768912307113286,
+      "grad_norm": 0.17144958674907684,
+      "learning_rate": 5e-06,
+      "loss": 0.1191,
+      "step": 15635
+    },
+    {
+      "epoch": 11.76966503575461,
+      "grad_norm": 0.1852876991033554,
+      "learning_rate": 5e-06,
+      "loss": 0.1545,
+      "step": 15636
+    },
+    {
+      "epoch": 11.770417764395935,
+      "grad_norm": 0.19837062060832977,
+      "learning_rate": 5e-06,
+      "loss": 0.2079,
+      "step": 15637
+    },
+    {
+      "epoch": 11.77117049303726,
+      "grad_norm": 0.19130414724349976,
+      "learning_rate": 5e-06,
+      "loss": 0.1373,
+      "step": 15638
+    },
+    {
+      "epoch": 11.771923221678584,
+      "grad_norm": 0.20306096971035004,
+      "learning_rate": 5e-06,
+      "loss": 0.1776,
+      "step": 15639
+    },
+    {
+      "epoch": 11.77267595031991,
+      "grad_norm": 0.1780768483877182,
+      "learning_rate": 5e-06,
+      "loss": 0.1276,
+      "step": 15640
+    },
+    {
+      "epoch": 11.773428678961235,
+      "grad_norm": 0.18872255086898804,
+      "learning_rate": 5e-06,
+      "loss": 0.0918,
+      "step": 15641
+    },
+    {
+      "epoch": 11.774181407602558,
+      "grad_norm": 0.1544257253408432,
+      "learning_rate": 5e-06,
+      "loss": 0.078,
+      "step": 15642
+    },
+    {
+      "epoch": 11.774934136243884,
+      "grad_norm": 0.18284186720848083,
+      "learning_rate": 5e-06,
+      "loss": 0.1908,
+      "step": 15643
+    },
+    {
+      "epoch": 11.775686864885209,
+      "grad_norm": 0.2082328349351883,
+      "learning_rate": 5e-06,
+      "loss": 0.1686,
+      "step": 15644
+    },
+    {
+      "epoch": 11.776439593526534,
+      "grad_norm": 0.19120915234088898,
+      "learning_rate": 5e-06,
+      "loss": 0.1443,
+      "step": 15645
+    },
+    {
+      "epoch": 11.777192322167858,
+      "grad_norm": 0.19163239002227783,
+      "learning_rate": 5e-06,
+      "loss": 0.1037,
+      "step": 15646
+    },
+    {
+      "epoch": 11.777945050809183,
+      "grad_norm": 0.1900513768196106,
+      "learning_rate": 5e-06,
+      "loss": 0.1586,
+      "step": 15647
+    },
+    {
+      "epoch": 11.778697779450509,
+      "grad_norm": 0.15142889320850372,
+      "learning_rate": 5e-06,
+      "loss": 0.1231,
+      "step": 15648
+    },
+    {
+      "epoch": 11.779450508091832,
+      "grad_norm": 0.1667884737253189,
+      "learning_rate": 5e-06,
+      "loss": 0.1772,
+      "step": 15649
+    },
+    {
+      "epoch": 11.780203236733158,
+      "grad_norm": 0.1939195692539215,
+      "learning_rate": 5e-06,
+      "loss": 0.1751,
+      "step": 15650
+    },
+    {
+      "epoch": 11.780955965374483,
+      "grad_norm": 0.17214007675647736,
+      "learning_rate": 5e-06,
+      "loss": 0.1147,
+      "step": 15651
+    },
+    {
+      "epoch": 11.781708694015807,
+      "grad_norm": 0.17732001841068268,
+      "learning_rate": 5e-06,
+      "loss": 0.1045,
+      "step": 15652
+    },
+    {
+      "epoch": 11.782461422657132,
+      "grad_norm": 0.19626228511333466,
+      "learning_rate": 5e-06,
+      "loss": 0.1739,
+      "step": 15653
+    },
+    {
+      "epoch": 11.783214151298457,
+      "grad_norm": 0.1603565365076065,
+      "learning_rate": 5e-06,
+      "loss": 0.1377,
+      "step": 15654
+    },
+    {
+      "epoch": 11.783966879939781,
+      "grad_norm": 0.1840067207813263,
+      "learning_rate": 5e-06,
+      "loss": 0.1214,
+      "step": 15655
+    },
+    {
+      "epoch": 11.784719608581106,
+      "grad_norm": 0.1712629646062851,
+      "learning_rate": 5e-06,
+      "loss": 0.1722,
+      "step": 15656
+    },
+    {
+      "epoch": 11.785472337222432,
+      "grad_norm": 0.18209724128246307,
+      "learning_rate": 5e-06,
+      "loss": 0.1211,
+      "step": 15657
+    },
+    {
+      "epoch": 11.786225065863755,
+      "grad_norm": 0.17203226685523987,
+      "learning_rate": 5e-06,
+      "loss": 0.1052,
+      "step": 15658
+    },
+    {
+      "epoch": 11.78697779450508,
+      "grad_norm": 0.19347186386585236,
+      "learning_rate": 5e-06,
+      "loss": 0.172,
+      "step": 15659
+    },
+    {
+      "epoch": 11.787730523146406,
+      "grad_norm": 0.17743784189224243,
+      "learning_rate": 5e-06,
+      "loss": 0.1109,
+      "step": 15660
+    },
+    {
+      "epoch": 11.788483251787731,
+      "grad_norm": 0.1757591813802719,
+      "learning_rate": 5e-06,
+      "loss": 0.1744,
+      "step": 15661
+    },
+    {
+      "epoch": 11.789235980429055,
+      "grad_norm": 0.1775028556585312,
+      "learning_rate": 5e-06,
+      "loss": 0.1703,
+      "step": 15662
+    },
+    {
+      "epoch": 11.78998870907038,
+      "grad_norm": 0.18446853756904602,
+      "learning_rate": 5e-06,
+      "loss": 0.1288,
+      "step": 15663
+    },
+    {
+      "epoch": 11.790741437711706,
+      "grad_norm": 0.18576259911060333,
+      "learning_rate": 5e-06,
+      "loss": 0.1168,
+      "step": 15664
+    },
+    {
+      "epoch": 11.79149416635303,
+      "grad_norm": 0.19750584661960602,
+      "learning_rate": 5e-06,
+      "loss": 0.1557,
+      "step": 15665
+    },
+    {
+      "epoch": 11.792246894994355,
+      "grad_norm": 0.16093584895133972,
+      "learning_rate": 5e-06,
+      "loss": 0.1234,
+      "step": 15666
+    },
+    {
+      "epoch": 11.79299962363568,
+      "grad_norm": 0.17439262568950653,
+      "learning_rate": 5e-06,
+      "loss": 0.1014,
+      "step": 15667
+    },
+    {
+      "epoch": 11.793752352277004,
+      "grad_norm": 0.21149423718452454,
+      "learning_rate": 5e-06,
+      "loss": 0.1278,
+      "step": 15668
+    },
+    {
+      "epoch": 11.794505080918329,
+      "grad_norm": 0.1760322004556656,
+      "learning_rate": 5e-06,
+      "loss": 0.156,
+      "step": 15669
+    },
+    {
+      "epoch": 11.795257809559654,
+      "grad_norm": 0.1845463216304779,
+      "learning_rate": 5e-06,
+      "loss": 0.13,
+      "step": 15670
+    },
+    {
+      "epoch": 11.796010538200978,
+      "grad_norm": 0.1993756741285324,
+      "learning_rate": 5e-06,
+      "loss": 0.1636,
+      "step": 15671
+    },
+    {
+      "epoch": 11.796763266842303,
+      "grad_norm": 0.1837548464536667,
+      "learning_rate": 5e-06,
+      "loss": 0.1171,
+      "step": 15672
+    },
+    {
+      "epoch": 11.797515995483629,
+      "grad_norm": 0.17716729640960693,
+      "learning_rate": 5e-06,
+      "loss": 0.1246,
+      "step": 15673
+    },
+    {
+      "epoch": 11.798268724124952,
+      "grad_norm": 0.18407510221004486,
+      "learning_rate": 5e-06,
+      "loss": 0.125,
+      "step": 15674
+    },
+    {
+      "epoch": 11.799021452766278,
+      "grad_norm": 0.1717749834060669,
+      "learning_rate": 5e-06,
+      "loss": 0.1552,
+      "step": 15675
+    },
+    {
+      "epoch": 11.799774181407603,
+      "grad_norm": 0.184727743268013,
+      "learning_rate": 5e-06,
+      "loss": 0.1317,
+      "step": 15676
+    },
+    {
+      "epoch": 11.800526910048927,
+      "grad_norm": 0.1847883015871048,
+      "learning_rate": 5e-06,
+      "loss": 0.1402,
+      "step": 15677
+    },
+    {
+      "epoch": 11.801279638690252,
+      "grad_norm": 0.19496563076972961,
+      "learning_rate": 5e-06,
+      "loss": 0.1403,
+      "step": 15678
+    },
+    {
+      "epoch": 11.802032367331577,
+      "grad_norm": 0.2011791169643402,
+      "learning_rate": 5e-06,
+      "loss": 0.2227,
+      "step": 15679
+    },
+    {
+      "epoch": 11.802785095972903,
+      "grad_norm": 0.18246051669120789,
+      "learning_rate": 5e-06,
+      "loss": 0.1661,
+      "step": 15680
+    },
+    {
+      "epoch": 11.803537824614226,
+      "grad_norm": 0.19058220088481903,
+      "learning_rate": 5e-06,
+      "loss": 0.1173,
+      "step": 15681
+    },
+    {
+      "epoch": 11.804290553255552,
+      "grad_norm": 0.17637240886688232,
+      "learning_rate": 5e-06,
+      "loss": 0.1052,
+      "step": 15682
+    },
+    {
+      "epoch": 11.805043281896877,
+      "grad_norm": 0.19151857495307922,
+      "learning_rate": 5e-06,
+      "loss": 0.1802,
+      "step": 15683
+    },
+    {
+      "epoch": 11.8057960105382,
+      "grad_norm": 0.19008225202560425,
+      "learning_rate": 5e-06,
+      "loss": 0.1426,
+      "step": 15684
+    },
+    {
+      "epoch": 11.806548739179526,
+      "grad_norm": 0.16167806088924408,
+      "learning_rate": 5e-06,
+      "loss": 0.1148,
+      "step": 15685
+    },
+    {
+      "epoch": 11.807301467820851,
+      "grad_norm": 0.19759856164455414,
+      "learning_rate": 5e-06,
+      "loss": 0.0902,
+      "step": 15686
+    },
+    {
+      "epoch": 11.808054196462175,
+      "grad_norm": 0.16456711292266846,
+      "learning_rate": 5e-06,
+      "loss": 0.1277,
+      "step": 15687
+    },
+    {
+      "epoch": 11.8088069251035,
+      "grad_norm": 0.1859598010778427,
+      "learning_rate": 5e-06,
+      "loss": 0.0817,
+      "step": 15688
+    },
+    {
+      "epoch": 11.809559653744826,
+      "grad_norm": 0.1839214563369751,
+      "learning_rate": 5e-06,
+      "loss": 0.1182,
+      "step": 15689
+    },
+    {
+      "epoch": 11.810312382386149,
+      "grad_norm": 0.16062679886817932,
+      "learning_rate": 5e-06,
+      "loss": 0.1371,
+      "step": 15690
+    },
+    {
+      "epoch": 11.811065111027474,
+      "grad_norm": 0.18374855816364288,
+      "learning_rate": 5e-06,
+      "loss": 0.0854,
+      "step": 15691
+    },
+    {
+      "epoch": 11.8118178396688,
+      "grad_norm": 0.17090214788913727,
+      "learning_rate": 5e-06,
+      "loss": 0.1044,
+      "step": 15692
+    },
+    {
+      "epoch": 11.812570568310123,
+      "grad_norm": 0.18784955143928528,
+      "learning_rate": 5e-06,
+      "loss": 0.1335,
+      "step": 15693
+    },
+    {
+      "epoch": 11.813323296951449,
+      "grad_norm": 0.16272753477096558,
+      "learning_rate": 5e-06,
+      "loss": 0.1558,
+      "step": 15694
+    },
+    {
+      "epoch": 11.814076025592774,
+      "grad_norm": 0.19660595059394836,
+      "learning_rate": 5e-06,
+      "loss": 0.1706,
+      "step": 15695
+    },
+    {
+      "epoch": 11.8148287542341,
+      "grad_norm": 0.18279695510864258,
+      "learning_rate": 5e-06,
+      "loss": 0.1268,
+      "step": 15696
+    },
+    {
+      "epoch": 11.815581482875423,
+      "grad_norm": 0.1655539721250534,
+      "learning_rate": 5e-06,
+      "loss": 0.0979,
+      "step": 15697
+    },
+    {
+      "epoch": 11.816334211516748,
+      "grad_norm": 0.1973138153553009,
+      "learning_rate": 5e-06,
+      "loss": 0.1846,
+      "step": 15698
+    },
+    {
+      "epoch": 11.817086940158074,
+      "grad_norm": 0.25773316621780396,
+      "learning_rate": 5e-06,
+      "loss": 0.1018,
+      "step": 15699
+    },
+    {
+      "epoch": 11.817839668799397,
+      "grad_norm": 0.20374614000320435,
+      "learning_rate": 5e-06,
+      "loss": 0.1289,
+      "step": 15700
+    },
+    {
+      "epoch": 11.818592397440723,
+      "grad_norm": 0.18270859122276306,
+      "learning_rate": 5e-06,
+      "loss": 0.1133,
+      "step": 15701
+    },
+    {
+      "epoch": 11.819345126082048,
+      "grad_norm": 0.1655222773551941,
+      "learning_rate": 5e-06,
+      "loss": 0.0827,
+      "step": 15702
+    },
+    {
+      "epoch": 11.820097854723372,
+      "grad_norm": 0.16954734921455383,
+      "learning_rate": 5e-06,
+      "loss": 0.1488,
+      "step": 15703
+    },
+    {
+      "epoch": 11.820850583364697,
+      "grad_norm": 0.174703449010849,
+      "learning_rate": 5e-06,
+      "loss": 0.1163,
+      "step": 15704
+    },
+    {
+      "epoch": 11.821603312006022,
+      "grad_norm": 0.1715126633644104,
+      "learning_rate": 5e-06,
+      "loss": 0.1641,
+      "step": 15705
+    },
+    {
+      "epoch": 11.822356040647346,
+      "grad_norm": 0.19084671139717102,
+      "learning_rate": 5e-06,
+      "loss": 0.2284,
+      "step": 15706
+    },
+    {
+      "epoch": 11.823108769288671,
+      "grad_norm": 0.1744319200515747,
+      "learning_rate": 5e-06,
+      "loss": 0.1499,
+      "step": 15707
+    },
+    {
+      "epoch": 11.823861497929997,
+      "grad_norm": 0.1927090734243393,
+      "learning_rate": 5e-06,
+      "loss": 0.1653,
+      "step": 15708
+    },
+    {
+      "epoch": 11.82461422657132,
+      "grad_norm": 0.16150464117527008,
+      "learning_rate": 5e-06,
+      "loss": 0.1525,
+      "step": 15709
+    },
+    {
+      "epoch": 11.825366955212646,
+      "grad_norm": 0.17831271886825562,
+      "learning_rate": 5e-06,
+      "loss": 0.0901,
+      "step": 15710
+    },
+    {
+      "epoch": 11.826119683853971,
+      "grad_norm": 0.15653443336486816,
+      "learning_rate": 5e-06,
+      "loss": 0.1109,
+      "step": 15711
+    },
+    {
+      "epoch": 11.826872412495295,
+      "grad_norm": 0.18990804255008698,
+      "learning_rate": 5e-06,
+      "loss": 0.1598,
+      "step": 15712
+    },
+    {
+      "epoch": 11.82762514113662,
+      "grad_norm": 0.1993468850851059,
+      "learning_rate": 5e-06,
+      "loss": 0.1341,
+      "step": 15713
+    },
+    {
+      "epoch": 11.828377869777945,
+      "grad_norm": 0.188460573554039,
+      "learning_rate": 5e-06,
+      "loss": 0.1438,
+      "step": 15714
+    },
+    {
+      "epoch": 11.829130598419269,
+      "grad_norm": 0.20010726153850555,
+      "learning_rate": 5e-06,
+      "loss": 0.1866,
+      "step": 15715
+    },
+    {
+      "epoch": 11.829883327060594,
+      "grad_norm": 0.16943952441215515,
+      "learning_rate": 5e-06,
+      "loss": 0.079,
+      "step": 15716
+    },
+    {
+      "epoch": 11.83063605570192,
+      "grad_norm": 0.17264162003993988,
+      "learning_rate": 5e-06,
+      "loss": 0.1765,
+      "step": 15717
+    },
+    {
+      "epoch": 11.831388784343245,
+      "grad_norm": 0.21276628971099854,
+      "learning_rate": 5e-06,
+      "loss": 0.1726,
+      "step": 15718
+    },
+    {
+      "epoch": 11.832141512984569,
+      "grad_norm": 0.18445229530334473,
+      "learning_rate": 5e-06,
+      "loss": 0.1125,
+      "step": 15719
+    },
+    {
+      "epoch": 11.832894241625894,
+      "grad_norm": 0.18044574558734894,
+      "learning_rate": 5e-06,
+      "loss": 0.1801,
+      "step": 15720
+    },
+    {
+      "epoch": 11.83364697026722,
+      "grad_norm": 0.17399144172668457,
+      "learning_rate": 5e-06,
+      "loss": 0.1305,
+      "step": 15721
+    },
+    {
+      "epoch": 11.834399698908543,
+      "grad_norm": 0.16621161997318268,
+      "learning_rate": 5e-06,
+      "loss": 0.0601,
+      "step": 15722
+    },
+    {
+      "epoch": 11.835152427549868,
+      "grad_norm": 0.17258892953395844,
+      "learning_rate": 5e-06,
+      "loss": 0.1543,
+      "step": 15723
+    },
+    {
+      "epoch": 11.835905156191194,
+      "grad_norm": 0.19957834482192993,
+      "learning_rate": 5e-06,
+      "loss": 0.2298,
+      "step": 15724
+    },
+    {
+      "epoch": 11.836657884832517,
+      "grad_norm": 0.18198524415493011,
+      "learning_rate": 5e-06,
+      "loss": 0.2123,
+      "step": 15725
+    },
+    {
+      "epoch": 11.837410613473843,
+      "grad_norm": 0.150855153799057,
+      "learning_rate": 5e-06,
+      "loss": 0.0953,
+      "step": 15726
+    },
+    {
+      "epoch": 11.838163342115168,
+      "grad_norm": 0.18126976490020752,
+      "learning_rate": 5e-06,
+      "loss": 0.154,
+      "step": 15727
+    },
+    {
+      "epoch": 11.838916070756492,
+      "grad_norm": 0.19186779856681824,
+      "learning_rate": 5e-06,
+      "loss": 0.213,
+      "step": 15728
+    },
+    {
+      "epoch": 11.839668799397817,
+      "grad_norm": 0.21605657041072845,
+      "learning_rate": 5e-06,
+      "loss": 0.1481,
+      "step": 15729
+    },
+    {
+      "epoch": 11.840421528039142,
+      "grad_norm": 0.19137068092823029,
+      "learning_rate": 5e-06,
+      "loss": 0.138,
+      "step": 15730
+    },
+    {
+      "epoch": 11.841174256680466,
+      "grad_norm": 0.19247199594974518,
+      "learning_rate": 5e-06,
+      "loss": 0.1433,
+      "step": 15731
+    },
+    {
+      "epoch": 11.841926985321791,
+      "grad_norm": 0.15977050364017487,
+      "learning_rate": 5e-06,
+      "loss": 0.0963,
+      "step": 15732
+    },
+    {
+      "epoch": 11.842679713963117,
+      "grad_norm": 0.2071814090013504,
+      "learning_rate": 5e-06,
+      "loss": 0.1424,
+      "step": 15733
+    },
+    {
+      "epoch": 11.843432442604442,
+      "grad_norm": 0.16342192888259888,
+      "learning_rate": 5e-06,
+      "loss": 0.0771,
+      "step": 15734
+    },
+    {
+      "epoch": 11.844185171245766,
+      "grad_norm": 0.1598941534757614,
+      "learning_rate": 5e-06,
+      "loss": 0.1399,
+      "step": 15735
+    },
+    {
+      "epoch": 11.844937899887091,
+      "grad_norm": 0.1663612574338913,
+      "learning_rate": 5e-06,
+      "loss": 0.135,
+      "step": 15736
+    },
+    {
+      "epoch": 11.845690628528416,
+      "grad_norm": 0.17159925401210785,
+      "learning_rate": 5e-06,
+      "loss": 0.1378,
+      "step": 15737
+    },
+    {
+      "epoch": 11.84644335716974,
+      "grad_norm": 0.2007126361131668,
+      "learning_rate": 5e-06,
+      "loss": 0.1223,
+      "step": 15738
+    },
+    {
+      "epoch": 11.847196085811065,
+      "grad_norm": 0.18989330530166626,
+      "learning_rate": 5e-06,
+      "loss": 0.1587,
+      "step": 15739
+    },
+    {
+      "epoch": 11.84794881445239,
+      "grad_norm": 0.17035698890686035,
+      "learning_rate": 5e-06,
+      "loss": 0.0865,
+      "step": 15740
+    },
+    {
+      "epoch": 11.848701543093714,
+      "grad_norm": 0.17176787555217743,
+      "learning_rate": 5e-06,
+      "loss": 0.1035,
+      "step": 15741
+    },
+    {
+      "epoch": 11.84945427173504,
+      "grad_norm": 0.2146095335483551,
+      "learning_rate": 5e-06,
+      "loss": 0.1619,
+      "step": 15742
+    },
+    {
+      "epoch": 11.850207000376365,
+      "grad_norm": 0.18044601380825043,
+      "learning_rate": 5e-06,
+      "loss": 0.096,
+      "step": 15743
+    },
+    {
+      "epoch": 11.850959729017688,
+      "grad_norm": 0.1820582002401352,
+      "learning_rate": 5e-06,
+      "loss": 0.1069,
+      "step": 15744
+    },
+    {
+      "epoch": 11.851712457659014,
+      "grad_norm": 0.17094101011753082,
+      "learning_rate": 5e-06,
+      "loss": 0.0596,
+      "step": 15745
+    },
+    {
+      "epoch": 11.85246518630034,
+      "grad_norm": 0.18634498119354248,
+      "learning_rate": 5e-06,
+      "loss": 0.1194,
+      "step": 15746
+    },
+    {
+      "epoch": 11.853217914941663,
+      "grad_norm": 0.1831444650888443,
+      "learning_rate": 5e-06,
+      "loss": 0.182,
+      "step": 15747
+    },
+    {
+      "epoch": 11.853970643582988,
+      "grad_norm": 0.24476711452007294,
+      "learning_rate": 5e-06,
+      "loss": 0.1714,
+      "step": 15748
+    },
+    {
+      "epoch": 11.854723372224313,
+      "grad_norm": 0.20474599301815033,
+      "learning_rate": 5e-06,
+      "loss": 0.1757,
+      "step": 15749
+    },
+    {
+      "epoch": 11.855476100865637,
+      "grad_norm": 0.18724921345710754,
+      "learning_rate": 5e-06,
+      "loss": 0.1297,
+      "step": 15750
+    },
+    {
+      "epoch": 11.856228829506962,
+      "grad_norm": 0.17093811929225922,
+      "learning_rate": 5e-06,
+      "loss": 0.1712,
+      "step": 15751
+    },
+    {
+      "epoch": 11.856981558148288,
+      "grad_norm": 0.16961728036403656,
+      "learning_rate": 5e-06,
+      "loss": 0.1095,
+      "step": 15752
+    },
+    {
+      "epoch": 11.857734286789613,
+      "grad_norm": 0.18606598675251007,
+      "learning_rate": 5e-06,
+      "loss": 0.1575,
+      "step": 15753
+    },
+    {
+      "epoch": 11.858487015430937,
+      "grad_norm": 0.21540462970733643,
+      "learning_rate": 5e-06,
+      "loss": 0.1678,
+      "step": 15754
+    },
+    {
+      "epoch": 11.859239744072262,
+      "grad_norm": 0.18271593749523163,
+      "learning_rate": 5e-06,
+      "loss": 0.1117,
+      "step": 15755
+    },
+    {
+      "epoch": 11.859992472713587,
+      "grad_norm": 0.19398437440395355,
+      "learning_rate": 5e-06,
+      "loss": 0.1545,
+      "step": 15756
+    },
+    {
+      "epoch": 11.860745201354911,
+      "grad_norm": 0.1844310313463211,
+      "learning_rate": 5e-06,
+      "loss": 0.1475,
+      "step": 15757
+    },
+    {
+      "epoch": 11.861497929996236,
+      "grad_norm": 0.16997162997722626,
+      "learning_rate": 5e-06,
+      "loss": 0.159,
+      "step": 15758
+    },
+    {
+      "epoch": 11.862250658637562,
+      "grad_norm": 0.1847381293773651,
+      "learning_rate": 5e-06,
+      "loss": 0.1231,
+      "step": 15759
+    },
+    {
+      "epoch": 11.863003387278885,
+      "grad_norm": 0.17607392370700836,
+      "learning_rate": 5e-06,
+      "loss": 0.1142,
+      "step": 15760
+    },
+    {
+      "epoch": 11.86375611592021,
+      "grad_norm": 0.21912522614002228,
+      "learning_rate": 5e-06,
+      "loss": 0.1377,
+      "step": 15761
+    },
+    {
+      "epoch": 11.864508844561536,
+      "grad_norm": 0.18031924962997437,
+      "learning_rate": 5e-06,
+      "loss": 0.1244,
+      "step": 15762
+    },
+    {
+      "epoch": 11.86526157320286,
+      "grad_norm": 0.17490030825138092,
+      "learning_rate": 5e-06,
+      "loss": 0.1124,
+      "step": 15763
+    },
+    {
+      "epoch": 11.866014301844185,
+      "grad_norm": 0.19490240514278412,
+      "learning_rate": 5e-06,
+      "loss": 0.1364,
+      "step": 15764
+    },
+    {
+      "epoch": 11.86676703048551,
+      "grad_norm": 0.18615508079528809,
+      "learning_rate": 5e-06,
+      "loss": 0.208,
+      "step": 15765
+    },
+    {
+      "epoch": 11.867519759126834,
+      "grad_norm": 0.18317300081253052,
+      "learning_rate": 5e-06,
+      "loss": 0.1587,
+      "step": 15766
+    },
+    {
+      "epoch": 11.86827248776816,
+      "grad_norm": 0.1636837124824524,
+      "learning_rate": 5e-06,
+      "loss": 0.1262,
+      "step": 15767
+    },
+    {
+      "epoch": 11.869025216409485,
+      "grad_norm": 0.1867399662733078,
+      "learning_rate": 5e-06,
+      "loss": 0.153,
+      "step": 15768
+    },
+    {
+      "epoch": 11.86977794505081,
+      "grad_norm": 0.16204313933849335,
+      "learning_rate": 5e-06,
+      "loss": 0.0939,
+      "step": 15769
+    },
+    {
+      "epoch": 11.870530673692134,
+      "grad_norm": 0.19802862405776978,
+      "learning_rate": 5e-06,
+      "loss": 0.1583,
+      "step": 15770
+    },
+    {
+      "epoch": 11.871283402333459,
+      "grad_norm": 0.18896247446537018,
+      "learning_rate": 5e-06,
+      "loss": 0.1407,
+      "step": 15771
+    },
+    {
+      "epoch": 11.872036130974784,
+      "grad_norm": 0.15503820776939392,
+      "learning_rate": 5e-06,
+      "loss": 0.0814,
+      "step": 15772
+    },
+    {
+      "epoch": 11.872788859616108,
+      "grad_norm": 0.17099490761756897,
+      "learning_rate": 5e-06,
+      "loss": 0.1072,
+      "step": 15773
+    },
+    {
+      "epoch": 11.873541588257433,
+      "grad_norm": 0.17968352138996124,
+      "learning_rate": 5e-06,
+      "loss": 0.1095,
+      "step": 15774
+    },
+    {
+      "epoch": 11.874294316898759,
+      "grad_norm": 0.19749118387699127,
+      "learning_rate": 5e-06,
+      "loss": 0.1563,
+      "step": 15775
+    },
+    {
+      "epoch": 11.875047045540082,
+      "grad_norm": 0.18650910258293152,
+      "learning_rate": 5e-06,
+      "loss": 0.1159,
+      "step": 15776
+    },
+    {
+      "epoch": 11.875799774181408,
+      "grad_norm": 0.17669853568077087,
+      "learning_rate": 5e-06,
+      "loss": 0.0885,
+      "step": 15777
+    },
+    {
+      "epoch": 11.876552502822733,
+      "grad_norm": 0.19192282855510712,
+      "learning_rate": 5e-06,
+      "loss": 0.1647,
+      "step": 15778
+    },
+    {
+      "epoch": 11.877305231464057,
+      "grad_norm": 0.19172199070453644,
+      "learning_rate": 5e-06,
+      "loss": 0.1155,
+      "step": 15779
+    },
+    {
+      "epoch": 11.878057960105382,
+      "grad_norm": 0.17408065497875214,
+      "learning_rate": 5e-06,
+      "loss": 0.1281,
+      "step": 15780
+    },
+    {
+      "epoch": 11.878810688746707,
+      "grad_norm": 0.18837904930114746,
+      "learning_rate": 5e-06,
+      "loss": 0.1051,
+      "step": 15781
+    },
+    {
+      "epoch": 11.879563417388031,
+      "grad_norm": 0.1729263961315155,
+      "learning_rate": 5e-06,
+      "loss": 0.0988,
+      "step": 15782
+    },
+    {
+      "epoch": 11.880316146029356,
+      "grad_norm": 0.19165655970573425,
+      "learning_rate": 5e-06,
+      "loss": 0.1699,
+      "step": 15783
+    },
+    {
+      "epoch": 11.881068874670682,
+      "grad_norm": 0.19474397599697113,
+      "learning_rate": 5e-06,
+      "loss": 0.1203,
+      "step": 15784
+    },
+    {
+      "epoch": 11.881821603312005,
+      "grad_norm": 0.15877071022987366,
+      "learning_rate": 5e-06,
+      "loss": 0.0907,
+      "step": 15785
+    },
+    {
+      "epoch": 11.88257433195333,
+      "grad_norm": 0.163418710231781,
+      "learning_rate": 5e-06,
+      "loss": 0.1255,
+      "step": 15786
+    },
+    {
+      "epoch": 11.883327060594656,
+      "grad_norm": 0.20471054315567017,
+      "learning_rate": 5e-06,
+      "loss": 0.1689,
+      "step": 15787
+    },
+    {
+      "epoch": 11.88407978923598,
+      "grad_norm": 0.1826576143503189,
+      "learning_rate": 5e-06,
+      "loss": 0.1586,
+      "step": 15788
+    },
+    {
+      "epoch": 11.884832517877305,
+      "grad_norm": 0.16014434397220612,
+      "learning_rate": 5e-06,
+      "loss": 0.1191,
+      "step": 15789
+    },
+    {
+      "epoch": 11.88558524651863,
+      "grad_norm": 0.1885817050933838,
+      "learning_rate": 5e-06,
+      "loss": 0.1727,
+      "step": 15790
+    },
+    {
+      "epoch": 11.886337975159956,
+      "grad_norm": 0.19891421496868134,
+      "learning_rate": 5e-06,
+      "loss": 0.1392,
+      "step": 15791
+    },
+    {
+      "epoch": 11.88709070380128,
+      "grad_norm": 0.19133731722831726,
+      "learning_rate": 5e-06,
+      "loss": 0.1716,
+      "step": 15792
+    },
+    {
+      "epoch": 11.887843432442605,
+      "grad_norm": 0.19222739338874817,
+      "learning_rate": 5e-06,
+      "loss": 0.1529,
+      "step": 15793
+    },
+    {
+      "epoch": 11.88859616108393,
+      "grad_norm": 0.18941131234169006,
+      "learning_rate": 5e-06,
+      "loss": 0.1235,
+      "step": 15794
+    },
+    {
+      "epoch": 11.889348889725253,
+      "grad_norm": 0.20708920061588287,
+      "learning_rate": 5e-06,
+      "loss": 0.1463,
+      "step": 15795
+    },
+    {
+      "epoch": 11.890101618366579,
+      "grad_norm": 0.177064910531044,
+      "learning_rate": 5e-06,
+      "loss": 0.102,
+      "step": 15796
+    },
+    {
+      "epoch": 11.890854347007904,
+      "grad_norm": 0.1740979254245758,
+      "learning_rate": 5e-06,
+      "loss": 0.1339,
+      "step": 15797
+    },
+    {
+      "epoch": 11.891607075649228,
+      "grad_norm": 0.16525618731975555,
+      "learning_rate": 5e-06,
+      "loss": 0.085,
+      "step": 15798
+    },
+    {
+      "epoch": 11.892359804290553,
+      "grad_norm": 0.17499953508377075,
+      "learning_rate": 5e-06,
+      "loss": 0.1343,
+      "step": 15799
+    },
+    {
+      "epoch": 11.893112532931879,
+      "grad_norm": 0.19644902646541595,
+      "learning_rate": 5e-06,
+      "loss": 0.1694,
+      "step": 15800
+    },
+    {
+      "epoch": 11.893112532931879,
+      "eval_loss": 0.1841178834438324,
+      "eval_runtime": 452.8708,
+      "eval_samples_per_second": 21.258,
+      "eval_steps_per_second": 0.665,
+      "step": 15800
+    },
+    {
+      "epoch": 11.893865261573202,
+      "grad_norm": 0.19109901785850525,
+      "learning_rate": 5e-06,
+      "loss": 0.1499,
+      "step": 15801
+    },
+    {
+      "epoch": 11.894617990214527,
+      "grad_norm": 0.18780003488063812,
+      "learning_rate": 5e-06,
+      "loss": 0.109,
+      "step": 15802
+    },
+    {
+      "epoch": 11.895370718855853,
+      "grad_norm": 0.17021040618419647,
+      "learning_rate": 5e-06,
+      "loss": 0.0708,
+      "step": 15803
+    },
+    {
+      "epoch": 11.896123447497178,
+      "grad_norm": 0.21026314795017242,
+      "learning_rate": 5e-06,
+      "loss": 0.1371,
+      "step": 15804
+    },
+    {
+      "epoch": 11.896876176138502,
+      "grad_norm": 0.1850854605436325,
+      "learning_rate": 5e-06,
+      "loss": 0.1343,
+      "step": 15805
+    },
+    {
+      "epoch": 11.897628904779827,
+      "grad_norm": 0.1935378462076187,
+      "learning_rate": 5e-06,
+      "loss": 0.0947,
+      "step": 15806
+    },
+    {
+      "epoch": 11.898381633421153,
+      "grad_norm": 0.1658928245306015,
+      "learning_rate": 5e-06,
+      "loss": 0.1063,
+      "step": 15807
+    },
+    {
+      "epoch": 11.899134362062476,
+      "grad_norm": 0.16876693069934845,
+      "learning_rate": 5e-06,
+      "loss": 0.0847,
+      "step": 15808
+    },
+    {
+      "epoch": 11.899887090703801,
+      "grad_norm": 0.20468294620513916,
+      "learning_rate": 5e-06,
+      "loss": 0.1569,
+      "step": 15809
+    },
+    {
+      "epoch": 11.900639819345127,
+      "grad_norm": 0.17733949422836304,
+      "learning_rate": 5e-06,
+      "loss": 0.1266,
+      "step": 15810
+    },
+    {
+      "epoch": 11.90139254798645,
+      "grad_norm": 0.17629463970661163,
+      "learning_rate": 5e-06,
+      "loss": 0.1698,
+      "step": 15811
+    },
+    {
+      "epoch": 11.902145276627776,
+      "grad_norm": 0.1725689321756363,
+      "learning_rate": 5e-06,
+      "loss": 0.0705,
+      "step": 15812
+    },
+    {
+      "epoch": 11.902898005269101,
+      "grad_norm": 0.16115747392177582,
+      "learning_rate": 5e-06,
+      "loss": 0.1526,
+      "step": 15813
+    },
+    {
+      "epoch": 11.903650733910425,
+      "grad_norm": 0.17130880057811737,
+      "learning_rate": 5e-06,
+      "loss": 0.1659,
+      "step": 15814
+    },
+    {
+      "epoch": 11.90440346255175,
+      "grad_norm": 0.17857807874679565,
+      "learning_rate": 5e-06,
+      "loss": 0.1049,
+      "step": 15815
+    },
+    {
+      "epoch": 11.905156191193075,
+      "grad_norm": 0.195856973528862,
+      "learning_rate": 5e-06,
+      "loss": 0.1586,
+      "step": 15816
+    },
+    {
+      "epoch": 11.905908919834399,
+      "grad_norm": 0.19343997538089752,
+      "learning_rate": 5e-06,
+      "loss": 0.1005,
+      "step": 15817
+    },
+    {
+      "epoch": 11.906661648475724,
+      "grad_norm": 0.2399352639913559,
+      "learning_rate": 5e-06,
+      "loss": 0.1709,
+      "step": 15818
+    },
+    {
+      "epoch": 11.90741437711705,
+      "grad_norm": 0.18867146968841553,
+      "learning_rate": 5e-06,
+      "loss": 0.1425,
+      "step": 15819
+    },
+    {
+      "epoch": 11.908167105758373,
+      "grad_norm": 0.17847277224063873,
+      "learning_rate": 5e-06,
+      "loss": 0.1669,
+      "step": 15820
+    },
+    {
+      "epoch": 11.908919834399699,
+      "grad_norm": 0.17172123491764069,
+      "learning_rate": 5e-06,
+      "loss": 0.0751,
+      "step": 15821
+    },
+    {
+      "epoch": 11.909672563041024,
+      "grad_norm": 0.16653773188591003,
+      "learning_rate": 5e-06,
+      "loss": 0.1545,
+      "step": 15822
+    },
+    {
+      "epoch": 11.910425291682348,
+      "grad_norm": 0.1837390512228012,
+      "learning_rate": 5e-06,
+      "loss": 0.1633,
+      "step": 15823
+    },
+    {
+      "epoch": 11.911178020323673,
+      "grad_norm": 0.18020470440387726,
+      "learning_rate": 5e-06,
+      "loss": 0.156,
+      "step": 15824
+    },
+    {
+      "epoch": 11.911930748964998,
+      "grad_norm": 0.17949418723583221,
+      "learning_rate": 5e-06,
+      "loss": 0.1736,
+      "step": 15825
+    },
+    {
+      "epoch": 11.912683477606324,
+      "grad_norm": 0.1546834409236908,
+      "learning_rate": 5e-06,
+      "loss": 0.1145,
+      "step": 15826
+    },
+    {
+      "epoch": 11.913436206247647,
+      "grad_norm": 0.17621029913425446,
+      "learning_rate": 5e-06,
+      "loss": 0.1231,
+      "step": 15827
+    },
+    {
+      "epoch": 11.914188934888973,
+      "grad_norm": 0.1717488020658493,
+      "learning_rate": 5e-06,
+      "loss": 0.1502,
+      "step": 15828
+    },
+    {
+      "epoch": 11.914941663530298,
+      "grad_norm": 0.19009162485599518,
+      "learning_rate": 5e-06,
+      "loss": 0.1011,
+      "step": 15829
+    },
+    {
+      "epoch": 11.915694392171622,
+      "grad_norm": 0.17919807136058807,
+      "learning_rate": 5e-06,
+      "loss": 0.1075,
+      "step": 15830
+    },
+    {
+      "epoch": 11.916447120812947,
+      "grad_norm": 0.18995146453380585,
+      "learning_rate": 5e-06,
+      "loss": 0.1903,
+      "step": 15831
+    },
+    {
+      "epoch": 11.917199849454272,
+      "grad_norm": 0.16764791309833527,
+      "learning_rate": 5e-06,
+      "loss": 0.126,
+      "step": 15832
+    },
+    {
+      "epoch": 11.917952578095596,
+      "grad_norm": 0.18951568007469177,
+      "learning_rate": 5e-06,
+      "loss": 0.1805,
+      "step": 15833
+    },
+    {
+      "epoch": 11.918705306736921,
+      "grad_norm": 0.18188157677650452,
+      "learning_rate": 5e-06,
+      "loss": 0.1233,
+      "step": 15834
+    },
+    {
+      "epoch": 11.919458035378247,
+      "grad_norm": 0.18074744939804077,
+      "learning_rate": 5e-06,
+      "loss": 0.1558,
+      "step": 15835
+    },
+    {
+      "epoch": 11.92021076401957,
+      "grad_norm": 0.22226153314113617,
+      "learning_rate": 5e-06,
+      "loss": 0.1952,
+      "step": 15836
+    },
+    {
+      "epoch": 11.920963492660896,
+      "grad_norm": 0.17485487461090088,
+      "learning_rate": 5e-06,
+      "loss": 0.1542,
+      "step": 15837
+    },
+    {
+      "epoch": 11.921716221302221,
+      "grad_norm": 0.1821981966495514,
+      "learning_rate": 5e-06,
+      "loss": 0.1091,
+      "step": 15838
+    },
+    {
+      "epoch": 11.922468949943545,
+      "grad_norm": 0.17186880111694336,
+      "learning_rate": 5e-06,
+      "loss": 0.1023,
+      "step": 15839
+    },
+    {
+      "epoch": 11.92322167858487,
+      "grad_norm": 0.2041279524564743,
+      "learning_rate": 5e-06,
+      "loss": 0.1908,
+      "step": 15840
+    },
+    {
+      "epoch": 11.923974407226195,
+      "grad_norm": 0.15473997592926025,
+      "learning_rate": 5e-06,
+      "loss": 0.1243,
+      "step": 15841
+    },
+    {
+      "epoch": 11.92472713586752,
+      "grad_norm": 0.19492292404174805,
+      "learning_rate": 5e-06,
+      "loss": 0.1556,
+      "step": 15842
+    },
+    {
+      "epoch": 11.925479864508844,
+      "grad_norm": 0.17906691133975983,
+      "learning_rate": 5e-06,
+      "loss": 0.1542,
+      "step": 15843
+    },
+    {
+      "epoch": 11.92623259315017,
+      "grad_norm": 0.20434528589248657,
+      "learning_rate": 5e-06,
+      "loss": 0.0958,
+      "step": 15844
+    },
+    {
+      "epoch": 11.926985321791495,
+      "grad_norm": 0.17827412486076355,
+      "learning_rate": 5e-06,
+      "loss": 0.1025,
+      "step": 15845
+    },
+    {
+      "epoch": 11.927738050432819,
+      "grad_norm": 0.16786909103393555,
+      "learning_rate": 5e-06,
+      "loss": 0.1408,
+      "step": 15846
+    },
+    {
+      "epoch": 11.928490779074144,
+      "grad_norm": 0.18271981179714203,
+      "learning_rate": 5e-06,
+      "loss": 0.1816,
+      "step": 15847
+    },
+    {
+      "epoch": 11.92924350771547,
+      "grad_norm": 0.18823093175888062,
+      "learning_rate": 5e-06,
+      "loss": 0.1159,
+      "step": 15848
+    },
+    {
+      "epoch": 11.929996236356793,
+      "grad_norm": 0.20546692609786987,
+      "learning_rate": 5e-06,
+      "loss": 0.13,
+      "step": 15849
+    },
+    {
+      "epoch": 11.930748964998118,
+      "grad_norm": 0.17504996061325073,
+      "learning_rate": 5e-06,
+      "loss": 0.1145,
+      "step": 15850
+    },
+    {
+      "epoch": 11.931501693639444,
+      "grad_norm": 0.1644221395254135,
+      "learning_rate": 5e-06,
+      "loss": 0.1401,
+      "step": 15851
+    },
+    {
+      "epoch": 11.932254422280767,
+      "grad_norm": 0.16410838067531586,
+      "learning_rate": 5e-06,
+      "loss": 0.0729,
+      "step": 15852
+    },
+    {
+      "epoch": 11.933007150922093,
+      "grad_norm": 0.1641157865524292,
+      "learning_rate": 5e-06,
+      "loss": 0.0814,
+      "step": 15853
+    },
+    {
+      "epoch": 11.933759879563418,
+      "grad_norm": 0.16372352838516235,
+      "learning_rate": 5e-06,
+      "loss": 0.0779,
+      "step": 15854
+    },
+    {
+      "epoch": 11.934512608204741,
+      "grad_norm": 0.15866036713123322,
+      "learning_rate": 5e-06,
+      "loss": 0.0864,
+      "step": 15855
+    },
+    {
+      "epoch": 11.935265336846067,
+      "grad_norm": 0.18627405166625977,
+      "learning_rate": 5e-06,
+      "loss": 0.1132,
+      "step": 15856
+    },
+    {
+      "epoch": 11.936018065487392,
+      "grad_norm": 0.21300727128982544,
+      "learning_rate": 5e-06,
+      "loss": 0.1785,
+      "step": 15857
+    },
+    {
+      "epoch": 11.936770794128716,
+      "grad_norm": 0.19857750833034515,
+      "learning_rate": 5e-06,
+      "loss": 0.1239,
+      "step": 15858
+    },
+    {
+      "epoch": 11.937523522770041,
+      "grad_norm": 0.18405672907829285,
+      "learning_rate": 5e-06,
+      "loss": 0.1361,
+      "step": 15859
+    },
+    {
+      "epoch": 11.938276251411367,
+      "grad_norm": 0.19678451120853424,
+      "learning_rate": 5e-06,
+      "loss": 0.1925,
+      "step": 15860
+    },
+    {
+      "epoch": 11.93902898005269,
+      "grad_norm": 0.16863712668418884,
+      "learning_rate": 5e-06,
+      "loss": 0.1173,
+      "step": 15861
+    },
+    {
+      "epoch": 11.939781708694015,
+      "grad_norm": 0.17059029638767242,
+      "learning_rate": 5e-06,
+      "loss": 0.1029,
+      "step": 15862
+    },
+    {
+      "epoch": 11.94053443733534,
+      "grad_norm": 0.16429825127124786,
+      "learning_rate": 5e-06,
+      "loss": 0.1172,
+      "step": 15863
+    },
+    {
+      "epoch": 11.941287165976666,
+      "grad_norm": 0.1847185343503952,
+      "learning_rate": 5e-06,
+      "loss": 0.1775,
+      "step": 15864
+    },
+    {
+      "epoch": 11.94203989461799,
+      "grad_norm": 0.17778970301151276,
+      "learning_rate": 5e-06,
+      "loss": 0.1624,
+      "step": 15865
+    },
+    {
+      "epoch": 11.942792623259315,
+      "grad_norm": 0.19014601409435272,
+      "learning_rate": 5e-06,
+      "loss": 0.1613,
+      "step": 15866
+    },
+    {
+      "epoch": 11.94354535190064,
+      "grad_norm": 0.1449367254972458,
+      "learning_rate": 5e-06,
+      "loss": 0.0975,
+      "step": 15867
+    },
+    {
+      "epoch": 11.944298080541964,
+      "grad_norm": 0.18240609765052795,
+      "learning_rate": 5e-06,
+      "loss": 0.157,
+      "step": 15868
+    },
+    {
+      "epoch": 11.94505080918329,
+      "grad_norm": 0.1810474693775177,
+      "learning_rate": 5e-06,
+      "loss": 0.1641,
+      "step": 15869
+    },
+    {
+      "epoch": 11.945803537824615,
+      "grad_norm": 0.15508803725242615,
+      "learning_rate": 5e-06,
+      "loss": 0.1615,
+      "step": 15870
+    },
+    {
+      "epoch": 11.946556266465938,
+      "grad_norm": 0.16704583168029785,
+      "learning_rate": 5e-06,
+      "loss": 0.096,
+      "step": 15871
+    },
+    {
+      "epoch": 11.947308995107264,
+      "grad_norm": 0.1738349050283432,
+      "learning_rate": 5e-06,
+      "loss": 0.127,
+      "step": 15872
+    },
+    {
+      "epoch": 11.948061723748589,
+      "grad_norm": 0.1594397872686386,
+      "learning_rate": 5e-06,
+      "loss": 0.1612,
+      "step": 15873
+    },
+    {
+      "epoch": 11.948814452389913,
+      "grad_norm": 0.16370871663093567,
+      "learning_rate": 5e-06,
+      "loss": 0.1129,
+      "step": 15874
+    },
+    {
+      "epoch": 11.949567181031238,
+      "grad_norm": 0.1868942379951477,
+      "learning_rate": 5e-06,
+      "loss": 0.1121,
+      "step": 15875
+    },
+    {
+      "epoch": 11.950319909672563,
+      "grad_norm": 0.1595950573682785,
+      "learning_rate": 5e-06,
+      "loss": 0.1111,
+      "step": 15876
+    },
+    {
+      "epoch": 11.951072638313889,
+      "grad_norm": 0.15983670949935913,
+      "learning_rate": 5e-06,
+      "loss": 0.1047,
+      "step": 15877
+    },
+    {
+      "epoch": 11.951825366955212,
+      "grad_norm": 0.2141847014427185,
+      "learning_rate": 5e-06,
+      "loss": 0.1591,
+      "step": 15878
+    },
+    {
+      "epoch": 11.952578095596538,
+      "grad_norm": 0.17210376262664795,
+      "learning_rate": 5e-06,
+      "loss": 0.1441,
+      "step": 15879
+    },
+    {
+      "epoch": 11.953330824237863,
+      "grad_norm": 0.18140773475170135,
+      "learning_rate": 5e-06,
+      "loss": 0.0914,
+      "step": 15880
+    },
+    {
+      "epoch": 11.954083552879187,
+      "grad_norm": 0.19562189280986786,
+      "learning_rate": 5e-06,
+      "loss": 0.1054,
+      "step": 15881
+    },
+    {
+      "epoch": 11.954836281520512,
+      "grad_norm": 0.20780733227729797,
+      "learning_rate": 5e-06,
+      "loss": 0.1416,
+      "step": 15882
+    },
+    {
+      "epoch": 11.955589010161837,
+      "grad_norm": 0.1702234148979187,
+      "learning_rate": 5e-06,
+      "loss": 0.1935,
+      "step": 15883
+    },
+    {
+      "epoch": 11.956341738803161,
+      "grad_norm": 0.1746790111064911,
+      "learning_rate": 5e-06,
+      "loss": 0.1745,
+      "step": 15884
+    },
+    {
+      "epoch": 11.957094467444486,
+      "grad_norm": 0.18376374244689941,
+      "learning_rate": 5e-06,
+      "loss": 0.1729,
+      "step": 15885
+    },
+    {
+      "epoch": 11.957847196085812,
+      "grad_norm": 0.20088745653629303,
+      "learning_rate": 5e-06,
+      "loss": 0.1073,
+      "step": 15886
+    },
+    {
+      "epoch": 11.958599924727135,
+      "grad_norm": 0.19901973009109497,
+      "learning_rate": 5e-06,
+      "loss": 0.1609,
+      "step": 15887
+    },
+    {
+      "epoch": 11.95935265336846,
+      "grad_norm": 0.1847177892923355,
+      "learning_rate": 5e-06,
+      "loss": 0.1572,
+      "step": 15888
+    },
+    {
+      "epoch": 11.960105382009786,
+      "grad_norm": 0.19335424900054932,
+      "learning_rate": 5e-06,
+      "loss": 0.1388,
+      "step": 15889
+    },
+    {
+      "epoch": 11.96085811065111,
+      "grad_norm": 0.1987866759300232,
+      "learning_rate": 5e-06,
+      "loss": 0.1628,
+      "step": 15890
+    },
+    {
+      "epoch": 11.961610839292435,
+      "grad_norm": 0.17718370258808136,
+      "learning_rate": 5e-06,
+      "loss": 0.1459,
+      "step": 15891
+    },
+    {
+      "epoch": 11.96236356793376,
+      "grad_norm": 0.2166254073381424,
+      "learning_rate": 5e-06,
+      "loss": 0.2011,
+      "step": 15892
+    },
+    {
+      "epoch": 11.963116296575084,
+      "grad_norm": 0.16276316344738007,
+      "learning_rate": 5e-06,
+      "loss": 0.0672,
+      "step": 15893
+    },
+    {
+      "epoch": 11.96386902521641,
+      "grad_norm": 0.16265316307544708,
+      "learning_rate": 5e-06,
+      "loss": 0.1121,
+      "step": 15894
+    },
+    {
+      "epoch": 11.964621753857735,
+      "grad_norm": 0.21414309740066528,
+      "learning_rate": 5e-06,
+      "loss": 0.1522,
+      "step": 15895
+    },
+    {
+      "epoch": 11.965374482499058,
+      "grad_norm": 0.1864224672317505,
+      "learning_rate": 5e-06,
+      "loss": 0.1643,
+      "step": 15896
+    },
+    {
+      "epoch": 11.966127211140384,
+      "grad_norm": 0.1882406324148178,
+      "learning_rate": 5e-06,
+      "loss": 0.1988,
+      "step": 15897
+    },
+    {
+      "epoch": 11.966879939781709,
+      "grad_norm": 0.15943577885627747,
+      "learning_rate": 5e-06,
+      "loss": 0.1162,
+      "step": 15898
+    },
+    {
+      "epoch": 11.967632668423034,
+      "grad_norm": 0.18569818139076233,
+      "learning_rate": 5e-06,
+      "loss": 0.1101,
+      "step": 15899
+    },
+    {
+      "epoch": 11.968385397064358,
+      "grad_norm": 0.1770022213459015,
+      "learning_rate": 5e-06,
+      "loss": 0.1551,
+      "step": 15900
+    },
+    {
+      "epoch": 11.969138125705683,
+      "grad_norm": 0.15835323929786682,
+      "learning_rate": 5e-06,
+      "loss": 0.0916,
+      "step": 15901
+    },
+    {
+      "epoch": 11.969890854347009,
+      "grad_norm": 0.16831035912036896,
+      "learning_rate": 5e-06,
+      "loss": 0.0549,
+      "step": 15902
+    },
+    {
+      "epoch": 11.970643582988332,
+      "grad_norm": 0.18651896715164185,
+      "learning_rate": 5e-06,
+      "loss": 0.1238,
+      "step": 15903
+    },
+    {
+      "epoch": 11.971396311629658,
+      "grad_norm": 0.21088552474975586,
+      "learning_rate": 5e-06,
+      "loss": 0.1815,
+      "step": 15904
+    },
+    {
+      "epoch": 11.972149040270983,
+      "grad_norm": 0.17469839751720428,
+      "learning_rate": 5e-06,
+      "loss": 0.1072,
+      "step": 15905
+    },
+    {
+      "epoch": 11.972901768912307,
+      "grad_norm": 0.18897105753421783,
+      "learning_rate": 5e-06,
+      "loss": 0.1692,
+      "step": 15906
+    },
+    {
+      "epoch": 11.973654497553632,
+      "grad_norm": 0.17712737619876862,
+      "learning_rate": 5e-06,
+      "loss": 0.1282,
+      "step": 15907
+    },
+    {
+      "epoch": 11.974407226194957,
+      "grad_norm": 0.1882035881280899,
+      "learning_rate": 5e-06,
+      "loss": 0.1346,
+      "step": 15908
+    },
+    {
+      "epoch": 11.97515995483628,
+      "grad_norm": 0.18558885157108307,
+      "learning_rate": 5e-06,
+      "loss": 0.1765,
+      "step": 15909
+    },
+    {
+      "epoch": 11.975912683477606,
+      "grad_norm": 0.19620518386363983,
+      "learning_rate": 5e-06,
+      "loss": 0.1711,
+      "step": 15910
+    },
+    {
+      "epoch": 11.976665412118932,
+      "grad_norm": 0.19116321206092834,
+      "learning_rate": 5e-06,
+      "loss": 0.1494,
+      "step": 15911
+    },
+    {
+      "epoch": 11.977418140760255,
+      "grad_norm": 0.22946611046791077,
+      "learning_rate": 5e-06,
+      "loss": 0.2289,
+      "step": 15912
+    },
+    {
+      "epoch": 11.97817086940158,
+      "grad_norm": 0.19617563486099243,
+      "learning_rate": 5e-06,
+      "loss": 0.1842,
+      "step": 15913
+    },
+    {
+      "epoch": 11.978923598042906,
+      "grad_norm": 0.16643890738487244,
+      "learning_rate": 5e-06,
+      "loss": 0.107,
+      "step": 15914
+    },
+    {
+      "epoch": 11.979676326684231,
+      "grad_norm": 0.1832236498594284,
+      "learning_rate": 5e-06,
+      "loss": 0.2363,
+      "step": 15915
+    },
+    {
+      "epoch": 11.980429055325555,
+      "grad_norm": 0.20545150339603424,
+      "learning_rate": 5e-06,
+      "loss": 0.1621,
+      "step": 15916
+    },
+    {
+      "epoch": 11.98118178396688,
+      "grad_norm": 0.17304767668247223,
+      "learning_rate": 5e-06,
+      "loss": 0.148,
+      "step": 15917
+    },
+    {
+      "epoch": 11.981934512608206,
+      "grad_norm": 0.19194236397743225,
+      "learning_rate": 5e-06,
+      "loss": 0.209,
+      "step": 15918
+    },
+    {
+      "epoch": 11.982687241249529,
+      "grad_norm": 0.17870838940143585,
+      "learning_rate": 5e-06,
+      "loss": 0.1211,
+      "step": 15919
+    },
+    {
+      "epoch": 11.983439969890854,
+      "grad_norm": 0.1693127453327179,
+      "learning_rate": 5e-06,
+      "loss": 0.1383,
+      "step": 15920
+    },
+    {
+      "epoch": 11.98419269853218,
+      "grad_norm": 0.18889762461185455,
+      "learning_rate": 5e-06,
+      "loss": 0.1229,
+      "step": 15921
+    },
+    {
+      "epoch": 11.984945427173503,
+      "grad_norm": 0.2154170125722885,
+      "learning_rate": 5e-06,
+      "loss": 0.1716,
+      "step": 15922
+    },
+    {
+      "epoch": 11.985698155814829,
+      "grad_norm": 0.2389429211616516,
+      "learning_rate": 5e-06,
+      "loss": 0.2418,
+      "step": 15923
+    },
+    {
+      "epoch": 11.986450884456154,
+      "grad_norm": 0.18888933956623077,
+      "learning_rate": 5e-06,
+      "loss": 0.0846,
+      "step": 15924
+    },
+    {
+      "epoch": 11.987203613097478,
+      "grad_norm": 0.17474456131458282,
+      "learning_rate": 5e-06,
+      "loss": 0.1413,
+      "step": 15925
+    },
+    {
+      "epoch": 11.987956341738803,
+      "grad_norm": 0.19705012440681458,
+      "learning_rate": 5e-06,
+      "loss": 0.1796,
+      "step": 15926
+    },
+    {
+      "epoch": 11.988709070380128,
+      "grad_norm": 0.1694350391626358,
+      "learning_rate": 5e-06,
+      "loss": 0.0838,
+      "step": 15927
+    },
+    {
+      "epoch": 11.989461799021452,
+      "grad_norm": 0.1681266576051712,
+      "learning_rate": 5e-06,
+      "loss": 0.2166,
+      "step": 15928
+    },
+    {
+      "epoch": 11.990214527662777,
+      "grad_norm": 0.19664040207862854,
+      "learning_rate": 5e-06,
+      "loss": 0.1197,
+      "step": 15929
+    },
+    {
+      "epoch": 11.990967256304103,
+      "grad_norm": 0.1836802363395691,
+      "learning_rate": 5e-06,
+      "loss": 0.1465,
+      "step": 15930
+    },
+    {
+      "epoch": 11.991719984945426,
+      "grad_norm": 0.1863672137260437,
+      "learning_rate": 5e-06,
+      "loss": 0.1534,
+      "step": 15931
+    },
+    {
+      "epoch": 11.992472713586752,
+      "grad_norm": 0.19965387880802155,
+      "learning_rate": 5e-06,
+      "loss": 0.1668,
+      "step": 15932
+    },
+    {
+      "epoch": 11.993225442228077,
+      "grad_norm": 0.19953760504722595,
+      "learning_rate": 5e-06,
+      "loss": 0.1655,
+      "step": 15933
+    },
+    {
+      "epoch": 11.9939781708694,
+      "grad_norm": 0.21508747339248657,
+      "learning_rate": 5e-06,
+      "loss": 0.1509,
+      "step": 15934
+    },
+    {
+      "epoch": 11.994730899510726,
+      "grad_norm": 0.2342425286769867,
+      "learning_rate": 5e-06,
+      "loss": 0.1841,
+      "step": 15935
+    },
+    {
+      "epoch": 11.995483628152051,
+      "grad_norm": 0.2008173018693924,
+      "learning_rate": 5e-06,
+      "loss": 0.1604,
+      "step": 15936
+    },
+    {
+      "epoch": 11.996236356793377,
+      "grad_norm": 0.14616158604621887,
+      "learning_rate": 5e-06,
+      "loss": 0.118,
+      "step": 15937
+    },
+    {
+      "epoch": 11.9969890854347,
+      "grad_norm": 0.188025563955307,
+      "learning_rate": 5e-06,
+      "loss": 0.1438,
+      "step": 15938
+    },
+    {
+      "epoch": 11.997741814076026,
+      "grad_norm": 0.17550110816955566,
+      "learning_rate": 5e-06,
+      "loss": 0.0847,
+      "step": 15939
+    },
+    {
+      "epoch": 11.998494542717351,
+      "grad_norm": 0.1773688644170761,
+      "learning_rate": 5e-06,
+      "loss": 0.1529,
+      "step": 15940
+    },
+    {
+      "epoch": 11.999247271358675,
+      "grad_norm": 0.20628337562084198,
+      "learning_rate": 5e-06,
+      "loss": 0.2201,
+      "step": 15941
+    },
+    {
+      "epoch": 12.0,
+      "grad_norm": 0.18375100195407867,
+      "learning_rate": 5e-06,
+      "loss": 0.1145,
+      "step": 15942
+    },
+    {
+      "epoch": 12.000752728641325,
+      "grad_norm": 0.21224023401737213,
+      "learning_rate": 5e-06,
+      "loss": 0.1687,
+      "step": 15943
+    },
+    {
+      "epoch": 12.001505457282649,
+      "grad_norm": 0.17890746891498566,
+      "learning_rate": 5e-06,
+      "loss": 0.1227,
+      "step": 15944
+    },
+    {
+      "epoch": 12.002258185923974,
+      "grad_norm": 0.17978015542030334,
+      "learning_rate": 5e-06,
+      "loss": 0.088,
+      "step": 15945
+    },
+    {
+      "epoch": 12.0030109145653,
+      "grad_norm": 0.1900380700826645,
+      "learning_rate": 5e-06,
+      "loss": 0.1077,
+      "step": 15946
+    },
+    {
+      "epoch": 12.003763643206623,
+      "grad_norm": 0.18397755920886993,
+      "learning_rate": 5e-06,
+      "loss": 0.1583,
+      "step": 15947
+    },
+    {
+      "epoch": 12.004516371847949,
+      "grad_norm": 0.16087935864925385,
+      "learning_rate": 5e-06,
+      "loss": 0.1071,
+      "step": 15948
+    },
+    {
+      "epoch": 12.005269100489274,
+      "grad_norm": 0.18699905276298523,
+      "learning_rate": 5e-06,
+      "loss": 0.1697,
+      "step": 15949
+    },
+    {
+      "epoch": 12.006021829130598,
+      "grad_norm": 0.22450540959835052,
+      "learning_rate": 5e-06,
+      "loss": 0.1639,
+      "step": 15950
+    },
+    {
+      "epoch": 12.006774557771923,
+      "grad_norm": 0.1905304193496704,
+      "learning_rate": 5e-06,
+      "loss": 0.1701,
+      "step": 15951
+    },
+    {
+      "epoch": 12.007527286413248,
+      "grad_norm": 0.16311493515968323,
+      "learning_rate": 5e-06,
+      "loss": 0.1065,
+      "step": 15952
+    },
+    {
+      "epoch": 12.008280015054574,
+      "grad_norm": 0.18121565878391266,
+      "learning_rate": 5e-06,
+      "loss": 0.1165,
+      "step": 15953
+    },
+    {
+      "epoch": 12.009032743695897,
+      "grad_norm": 0.2011972963809967,
+      "learning_rate": 5e-06,
+      "loss": 0.1716,
+      "step": 15954
+    },
+    {
+      "epoch": 12.009785472337223,
+      "grad_norm": 0.16064511239528656,
+      "learning_rate": 5e-06,
+      "loss": 0.1248,
+      "step": 15955
+    },
+    {
+      "epoch": 12.010538200978548,
+      "grad_norm": 0.17212779819965363,
+      "learning_rate": 5e-06,
+      "loss": 0.1741,
+      "step": 15956
+    },
+    {
+      "epoch": 12.011290929619872,
+      "grad_norm": 0.19771724939346313,
+      "learning_rate": 5e-06,
+      "loss": 0.1524,
+      "step": 15957
+    },
+    {
+      "epoch": 12.012043658261197,
+      "grad_norm": 0.17265476286411285,
+      "learning_rate": 5e-06,
+      "loss": 0.0998,
+      "step": 15958
+    },
+    {
+      "epoch": 12.012796386902522,
+      "grad_norm": 0.19088590145111084,
+      "learning_rate": 5e-06,
+      "loss": 0.1352,
+      "step": 15959
+    },
+    {
+      "epoch": 12.013549115543846,
+      "grad_norm": 0.16837705671787262,
+      "learning_rate": 5e-06,
+      "loss": 0.1535,
+      "step": 15960
+    },
+    {
+      "epoch": 12.014301844185171,
+      "grad_norm": 0.1653982698917389,
+      "learning_rate": 5e-06,
+      "loss": 0.1543,
+      "step": 15961
+    },
+    {
+      "epoch": 12.015054572826497,
+      "grad_norm": 0.1908964067697525,
+      "learning_rate": 5e-06,
+      "loss": 0.1188,
+      "step": 15962
+    },
+    {
+      "epoch": 12.01580730146782,
+      "grad_norm": 0.1698518842458725,
+      "learning_rate": 5e-06,
+      "loss": 0.0846,
+      "step": 15963
+    },
+    {
+      "epoch": 12.016560030109146,
+      "grad_norm": 0.17975497245788574,
+      "learning_rate": 5e-06,
+      "loss": 0.0932,
+      "step": 15964
+    },
+    {
+      "epoch": 12.017312758750471,
+      "grad_norm": 0.16435496509075165,
+      "learning_rate": 5e-06,
+      "loss": 0.0948,
+      "step": 15965
+    },
+    {
+      "epoch": 12.018065487391794,
+      "grad_norm": 0.17618808150291443,
+      "learning_rate": 5e-06,
+      "loss": 0.1294,
+      "step": 15966
+    },
+    {
+      "epoch": 12.01881821603312,
+      "grad_norm": 0.18261471390724182,
+      "learning_rate": 5e-06,
+      "loss": 0.1316,
+      "step": 15967
+    },
+    {
+      "epoch": 12.019570944674445,
+      "grad_norm": 0.1683439165353775,
+      "learning_rate": 5e-06,
+      "loss": 0.1179,
+      "step": 15968
+    },
+    {
+      "epoch": 12.020323673315769,
+      "grad_norm": 0.19664332270622253,
+      "learning_rate": 5e-06,
+      "loss": 0.1227,
+      "step": 15969
+    },
+    {
+      "epoch": 12.021076401957094,
+      "grad_norm": 0.16778962314128876,
+      "learning_rate": 5e-06,
+      "loss": 0.0869,
+      "step": 15970
+    },
+    {
+      "epoch": 12.02182913059842,
+      "grad_norm": 0.15306462347507477,
+      "learning_rate": 5e-06,
+      "loss": 0.0998,
+      "step": 15971
+    },
+    {
+      "epoch": 12.022581859239745,
+      "grad_norm": 0.17583759129047394,
+      "learning_rate": 5e-06,
+      "loss": 0.082,
+      "step": 15972
+    },
+    {
+      "epoch": 12.023334587881068,
+      "grad_norm": 0.17836283147335052,
+      "learning_rate": 5e-06,
+      "loss": 0.1583,
+      "step": 15973
+    },
+    {
+      "epoch": 12.024087316522394,
+      "grad_norm": 0.20084436237812042,
+      "learning_rate": 5e-06,
+      "loss": 0.1852,
+      "step": 15974
+    },
+    {
+      "epoch": 12.02484004516372,
+      "grad_norm": 0.18943966925144196,
+      "learning_rate": 5e-06,
+      "loss": 0.1225,
+      "step": 15975
+    },
+    {
+      "epoch": 12.025592773805043,
+      "grad_norm": 0.21092389523983002,
+      "learning_rate": 5e-06,
+      "loss": 0.1564,
+      "step": 15976
+    },
+    {
+      "epoch": 12.026345502446368,
+      "grad_norm": 0.17468616366386414,
+      "learning_rate": 5e-06,
+      "loss": 0.1475,
+      "step": 15977
+    },
+    {
+      "epoch": 12.027098231087693,
+      "grad_norm": 0.18198414146900177,
+      "learning_rate": 5e-06,
+      "loss": 0.1531,
+      "step": 15978
+    },
+    {
+      "epoch": 12.027850959729017,
+      "grad_norm": 0.19232577085494995,
+      "learning_rate": 5e-06,
+      "loss": 0.1761,
+      "step": 15979
+    },
+    {
+      "epoch": 12.028603688370342,
+      "grad_norm": 0.16224084794521332,
+      "learning_rate": 5e-06,
+      "loss": 0.1447,
+      "step": 15980
+    },
+    {
+      "epoch": 12.029356417011668,
+      "grad_norm": 0.17120203375816345,
+      "learning_rate": 5e-06,
+      "loss": 0.1075,
+      "step": 15981
+    },
+    {
+      "epoch": 12.030109145652991,
+      "grad_norm": 0.16560524702072144,
+      "learning_rate": 5e-06,
+      "loss": 0.1693,
+      "step": 15982
+    },
+    {
+      "epoch": 12.030861874294317,
+      "grad_norm": 0.201870858669281,
+      "learning_rate": 5e-06,
+      "loss": 0.2041,
+      "step": 15983
+    },
+    {
+      "epoch": 12.031614602935642,
+      "grad_norm": 0.16625316441059113,
+      "learning_rate": 5e-06,
+      "loss": 0.0973,
+      "step": 15984
+    },
+    {
+      "epoch": 12.032367331576966,
+      "grad_norm": 0.19709765911102295,
+      "learning_rate": 5e-06,
+      "loss": 0.1629,
+      "step": 15985
+    },
+    {
+      "epoch": 12.033120060218291,
+      "grad_norm": 0.20327383279800415,
+      "learning_rate": 5e-06,
+      "loss": 0.1845,
+      "step": 15986
+    },
+    {
+      "epoch": 12.033872788859616,
+      "grad_norm": 0.19748549163341522,
+      "learning_rate": 5e-06,
+      "loss": 0.1291,
+      "step": 15987
+    },
+    {
+      "epoch": 12.034625517500942,
+      "grad_norm": 0.17357246577739716,
+      "learning_rate": 5e-06,
+      "loss": 0.0849,
+      "step": 15988
+    },
+    {
+      "epoch": 12.035378246142265,
+      "grad_norm": 0.15106219053268433,
+      "learning_rate": 5e-06,
+      "loss": 0.062,
+      "step": 15989
+    },
+    {
+      "epoch": 12.03613097478359,
+      "grad_norm": 0.2049635648727417,
+      "learning_rate": 5e-06,
+      "loss": 0.1578,
+      "step": 15990
+    },
+    {
+      "epoch": 12.036883703424916,
+      "grad_norm": 0.19930531084537506,
+      "learning_rate": 5e-06,
+      "loss": 0.1668,
+      "step": 15991
+    },
+    {
+      "epoch": 12.03763643206624,
+      "grad_norm": 0.1633257418870926,
+      "learning_rate": 5e-06,
+      "loss": 0.1292,
+      "step": 15992
+    },
+    {
+      "epoch": 12.038389160707565,
+      "grad_norm": 0.1781502366065979,
+      "learning_rate": 5e-06,
+      "loss": 0.1085,
+      "step": 15993
+    },
+    {
+      "epoch": 12.03914188934889,
+      "grad_norm": 0.1778683364391327,
+      "learning_rate": 5e-06,
+      "loss": 0.1002,
+      "step": 15994
+    },
+    {
+      "epoch": 12.039894617990214,
+      "grad_norm": 0.22183936834335327,
+      "learning_rate": 5e-06,
+      "loss": 0.2196,
+      "step": 15995
+    },
+    {
+      "epoch": 12.04064734663154,
+      "grad_norm": 0.19910109043121338,
+      "learning_rate": 5e-06,
+      "loss": 0.2025,
+      "step": 15996
+    },
+    {
+      "epoch": 12.041400075272865,
+      "grad_norm": 0.1582460105419159,
+      "learning_rate": 5e-06,
+      "loss": 0.0913,
+      "step": 15997
+    },
+    {
+      "epoch": 12.042152803914188,
+      "grad_norm": 0.20248724520206451,
+      "learning_rate": 5e-06,
+      "loss": 0.1729,
+      "step": 15998
+    },
+    {
+      "epoch": 12.042905532555514,
+      "grad_norm": 0.1675233542919159,
+      "learning_rate": 5e-06,
+      "loss": 0.1481,
+      "step": 15999
+    },
+    {
+      "epoch": 12.043658261196839,
+      "grad_norm": 0.17274805903434753,
+      "learning_rate": 5e-06,
+      "loss": 0.0957,
+      "step": 16000
+    },
+    {
+      "epoch": 12.043658261196839,
+      "eval_loss": 0.18335658311843872,
+      "eval_runtime": 452.9604,
+      "eval_samples_per_second": 21.254,
+      "eval_steps_per_second": 0.665,
+      "step": 16000
+    },
+    {
+      "epoch": 12.044410989838163,
+      "grad_norm": 0.16334906220436096,
+      "learning_rate": 5e-06,
+      "loss": 0.1239,
+      "step": 16001
+    },
+    {
+      "epoch": 12.045163718479488,
+      "grad_norm": 0.19723041355609894,
+      "learning_rate": 5e-06,
+      "loss": 0.1056,
+      "step": 16002
+    },
+    {
+      "epoch": 12.045916447120813,
+      "grad_norm": 0.18961508572101593,
+      "learning_rate": 5e-06,
+      "loss": 0.1956,
+      "step": 16003
+    },
+    {
+      "epoch": 12.046669175762137,
+      "grad_norm": 0.15420182049274445,
+      "learning_rate": 5e-06,
+      "loss": 0.0482,
+      "step": 16004
+    },
+    {
+      "epoch": 12.047421904403462,
+      "grad_norm": 0.17865145206451416,
+      "learning_rate": 5e-06,
+      "loss": 0.1072,
+      "step": 16005
+    },
+    {
+      "epoch": 12.048174633044788,
+      "grad_norm": 0.18314824998378754,
+      "learning_rate": 5e-06,
+      "loss": 0.1306,
+      "step": 16006
+    },
+    {
+      "epoch": 12.048927361686113,
+      "grad_norm": 0.15307648479938507,
+      "learning_rate": 5e-06,
+      "loss": 0.0821,
+      "step": 16007
+    },
+    {
+      "epoch": 12.049680090327437,
+      "grad_norm": 0.18395981192588806,
+      "learning_rate": 5e-06,
+      "loss": 0.1095,
+      "step": 16008
+    },
+    {
+      "epoch": 12.050432818968762,
+      "grad_norm": 0.21078704297542572,
+      "learning_rate": 5e-06,
+      "loss": 0.1203,
+      "step": 16009
+    },
+    {
+      "epoch": 12.051185547610087,
+      "grad_norm": 0.1723247915506363,
+      "learning_rate": 5e-06,
+      "loss": 0.1793,
+      "step": 16010
+    },
+    {
+      "epoch": 12.051938276251411,
+      "grad_norm": 0.17418988049030304,
+      "learning_rate": 5e-06,
+      "loss": 0.1342,
+      "step": 16011
+    },
+    {
+      "epoch": 12.052691004892736,
+      "grad_norm": 0.19715817272663116,
+      "learning_rate": 5e-06,
+      "loss": 0.1235,
+      "step": 16012
+    },
+    {
+      "epoch": 12.053443733534062,
+      "grad_norm": 0.18909041583538055,
+      "learning_rate": 5e-06,
+      "loss": 0.1242,
+      "step": 16013
+    },
+    {
+      "epoch": 12.054196462175385,
+      "grad_norm": 0.1516982614994049,
+      "learning_rate": 5e-06,
+      "loss": 0.1008,
+      "step": 16014
+    },
+    {
+      "epoch": 12.05494919081671,
+      "grad_norm": 0.19468465447425842,
+      "learning_rate": 5e-06,
+      "loss": 0.1624,
+      "step": 16015
+    },
+    {
+      "epoch": 12.055701919458036,
+      "grad_norm": 0.18058833479881287,
+      "learning_rate": 5e-06,
+      "loss": 0.148,
+      "step": 16016
+    },
+    {
+      "epoch": 12.05645464809936,
+      "grad_norm": 0.1615714132785797,
+      "learning_rate": 5e-06,
+      "loss": 0.1307,
+      "step": 16017
+    },
+    {
+      "epoch": 12.057207376740685,
+      "grad_norm": 0.17503008246421814,
+      "learning_rate": 5e-06,
+      "loss": 0.1555,
+      "step": 16018
+    },
+    {
+      "epoch": 12.05796010538201,
+      "grad_norm": 0.1853693425655365,
+      "learning_rate": 5e-06,
+      "loss": 0.1232,
+      "step": 16019
+    },
+    {
+      "epoch": 12.058712834023334,
+      "grad_norm": 0.16883400082588196,
+      "learning_rate": 5e-06,
+      "loss": 0.1084,
+      "step": 16020
+    },
+    {
+      "epoch": 12.05946556266466,
+      "grad_norm": 0.18723750114440918,
+      "learning_rate": 5e-06,
+      "loss": 0.096,
+      "step": 16021
+    },
+    {
+      "epoch": 12.060218291305985,
+      "grad_norm": 0.20472940802574158,
+      "learning_rate": 5e-06,
+      "loss": 0.177,
+      "step": 16022
+    },
+    {
+      "epoch": 12.060971019947308,
+      "grad_norm": 0.17868462204933167,
+      "learning_rate": 5e-06,
+      "loss": 0.1794,
+      "step": 16023
+    },
+    {
+      "epoch": 12.061723748588633,
+      "grad_norm": 0.19650214910507202,
+      "learning_rate": 5e-06,
+      "loss": 0.1589,
+      "step": 16024
+    },
+    {
+      "epoch": 12.062476477229959,
+      "grad_norm": 0.22567296028137207,
+      "learning_rate": 5e-06,
+      "loss": 0.1099,
+      "step": 16025
+    },
+    {
+      "epoch": 12.063229205871284,
+      "grad_norm": 0.18402211368083954,
+      "learning_rate": 5e-06,
+      "loss": 0.1962,
+      "step": 16026
+    },
+    {
+      "epoch": 12.063981934512608,
+      "grad_norm": 0.2051773965358734,
+      "learning_rate": 5e-06,
+      "loss": 0.1285,
+      "step": 16027
+    },
+    {
+      "epoch": 12.064734663153933,
+      "grad_norm": 0.183771014213562,
+      "learning_rate": 5e-06,
+      "loss": 0.1291,
+      "step": 16028
+    },
+    {
+      "epoch": 12.065487391795259,
+      "grad_norm": 0.17572662234306335,
+      "learning_rate": 5e-06,
+      "loss": 0.0861,
+      "step": 16029
+    },
+    {
+      "epoch": 12.066240120436582,
+      "grad_norm": 0.17893145978450775,
+      "learning_rate": 5e-06,
+      "loss": 0.0897,
+      "step": 16030
+    },
+    {
+      "epoch": 12.066992849077907,
+      "grad_norm": 0.19315388798713684,
+      "learning_rate": 5e-06,
+      "loss": 0.1355,
+      "step": 16031
+    },
+    {
+      "epoch": 12.067745577719233,
+      "grad_norm": 0.18231625854969025,
+      "learning_rate": 5e-06,
+      "loss": 0.1008,
+      "step": 16032
+    },
+    {
+      "epoch": 12.068498306360556,
+      "grad_norm": 0.17690417170524597,
+      "learning_rate": 5e-06,
+      "loss": 0.086,
+      "step": 16033
+    },
+    {
+      "epoch": 12.069251035001882,
+      "grad_norm": 0.17275677621364594,
+      "learning_rate": 5e-06,
+      "loss": 0.1397,
+      "step": 16034
+    },
+    {
+      "epoch": 12.070003763643207,
+      "grad_norm": 0.17728370428085327,
+      "learning_rate": 5e-06,
+      "loss": 0.1812,
+      "step": 16035
+    },
+    {
+      "epoch": 12.07075649228453,
+      "grad_norm": 0.19481337070465088,
+      "learning_rate": 5e-06,
+      "loss": 0.0846,
+      "step": 16036
+    },
+    {
+      "epoch": 12.071509220925856,
+      "grad_norm": 0.21173498034477234,
+      "learning_rate": 5e-06,
+      "loss": 0.1645,
+      "step": 16037
+    },
+    {
+      "epoch": 12.072261949567181,
+      "grad_norm": 0.19076354801654816,
+      "learning_rate": 5e-06,
+      "loss": 0.0685,
+      "step": 16038
+    },
+    {
+      "epoch": 12.073014678208505,
+      "grad_norm": 0.19880187511444092,
+      "learning_rate": 5e-06,
+      "loss": 0.0999,
+      "step": 16039
+    },
+    {
+      "epoch": 12.07376740684983,
+      "grad_norm": 0.15709610283374786,
+      "learning_rate": 5e-06,
+      "loss": 0.065,
+      "step": 16040
+    },
+    {
+      "epoch": 12.074520135491156,
+      "grad_norm": 0.1753903031349182,
+      "learning_rate": 5e-06,
+      "loss": 0.1622,
+      "step": 16041
+    },
+    {
+      "epoch": 12.07527286413248,
+      "grad_norm": 0.156229630112648,
+      "learning_rate": 5e-06,
+      "loss": 0.1105,
+      "step": 16042
+    },
+    {
+      "epoch": 12.076025592773805,
+      "grad_norm": 0.19306917488574982,
+      "learning_rate": 5e-06,
+      "loss": 0.1406,
+      "step": 16043
+    },
+    {
+      "epoch": 12.07677832141513,
+      "grad_norm": 0.1958913803100586,
+      "learning_rate": 5e-06,
+      "loss": 0.0956,
+      "step": 16044
+    },
+    {
+      "epoch": 12.077531050056455,
+      "grad_norm": 0.1810644418001175,
+      "learning_rate": 5e-06,
+      "loss": 0.1117,
+      "step": 16045
+    },
+    {
+      "epoch": 12.078283778697779,
+      "grad_norm": 0.16932061314582825,
+      "learning_rate": 5e-06,
+      "loss": 0.135,
+      "step": 16046
+    },
+    {
+      "epoch": 12.079036507339104,
+      "grad_norm": 0.2043260633945465,
+      "learning_rate": 5e-06,
+      "loss": 0.1546,
+      "step": 16047
+    },
+    {
+      "epoch": 12.07978923598043,
+      "grad_norm": 0.20933598279953003,
+      "learning_rate": 5e-06,
+      "loss": 0.1806,
+      "step": 16048
+    },
+    {
+      "epoch": 12.080541964621753,
+      "grad_norm": 0.16441138088703156,
+      "learning_rate": 5e-06,
+      "loss": 0.195,
+      "step": 16049
+    },
+    {
+      "epoch": 12.081294693263079,
+      "grad_norm": 0.16551783680915833,
+      "learning_rate": 5e-06,
+      "loss": 0.0734,
+      "step": 16050
+    },
+    {
+      "epoch": 12.082047421904404,
+      "grad_norm": 0.1751224845647812,
+      "learning_rate": 5e-06,
+      "loss": 0.1217,
+      "step": 16051
+    },
+    {
+      "epoch": 12.082800150545728,
+      "grad_norm": 0.18251767754554749,
+      "learning_rate": 5e-06,
+      "loss": 0.1194,
+      "step": 16052
+    },
+    {
+      "epoch": 12.083552879187053,
+      "grad_norm": 0.1660604476928711,
+      "learning_rate": 5e-06,
+      "loss": 0.1619,
+      "step": 16053
+    },
+    {
+      "epoch": 12.084305607828378,
+      "grad_norm": 0.17619571089744568,
+      "learning_rate": 5e-06,
+      "loss": 0.1645,
+      "step": 16054
+    },
+    {
+      "epoch": 12.085058336469702,
+      "grad_norm": 0.18589283525943756,
+      "learning_rate": 5e-06,
+      "loss": 0.1102,
+      "step": 16055
+    },
+    {
+      "epoch": 12.085811065111027,
+      "grad_norm": 0.1998787224292755,
+      "learning_rate": 5e-06,
+      "loss": 0.1622,
+      "step": 16056
+    },
+    {
+      "epoch": 12.086563793752353,
+      "grad_norm": 0.18509165942668915,
+      "learning_rate": 5e-06,
+      "loss": 0.1231,
+      "step": 16057
+    },
+    {
+      "epoch": 12.087316522393676,
+      "grad_norm": 0.15664739906787872,
+      "learning_rate": 5e-06,
+      "loss": 0.1169,
+      "step": 16058
+    },
+    {
+      "epoch": 12.088069251035002,
+      "grad_norm": 0.16993239521980286,
+      "learning_rate": 5e-06,
+      "loss": 0.1557,
+      "step": 16059
+    },
+    {
+      "epoch": 12.088821979676327,
+      "grad_norm": 0.171518936753273,
+      "learning_rate": 5e-06,
+      "loss": 0.214,
+      "step": 16060
+    },
+    {
+      "epoch": 12.089574708317652,
+      "grad_norm": 0.19068464636802673,
+      "learning_rate": 5e-06,
+      "loss": 0.213,
+      "step": 16061
+    },
+    {
+      "epoch": 12.090327436958976,
+      "grad_norm": 0.17354817688465118,
+      "learning_rate": 5e-06,
+      "loss": 0.1331,
+      "step": 16062
+    },
+    {
+      "epoch": 12.091080165600301,
+      "grad_norm": 0.1750434786081314,
+      "learning_rate": 5e-06,
+      "loss": 0.1323,
+      "step": 16063
+    },
+    {
+      "epoch": 12.091832894241627,
+      "grad_norm": 0.18610763549804688,
+      "learning_rate": 5e-06,
+      "loss": 0.2262,
+      "step": 16064
+    },
+    {
+      "epoch": 12.09258562288295,
+      "grad_norm": 0.18086741864681244,
+      "learning_rate": 5e-06,
+      "loss": 0.1807,
+      "step": 16065
+    },
+    {
+      "epoch": 12.093338351524276,
+      "grad_norm": 0.1736088991165161,
+      "learning_rate": 5e-06,
+      "loss": 0.149,
+      "step": 16066
+    },
+    {
+      "epoch": 12.094091080165601,
+      "grad_norm": 0.16473053395748138,
+      "learning_rate": 5e-06,
+      "loss": 0.0591,
+      "step": 16067
+    },
+    {
+      "epoch": 12.094843808806925,
+      "grad_norm": 0.20104554295539856,
+      "learning_rate": 5e-06,
+      "loss": 0.1329,
+      "step": 16068
+    },
+    {
+      "epoch": 12.09559653744825,
+      "grad_norm": 0.1603412926197052,
+      "learning_rate": 5e-06,
+      "loss": 0.1391,
+      "step": 16069
+    },
+    {
+      "epoch": 12.096349266089575,
+      "grad_norm": 0.1581060141324997,
+      "learning_rate": 5e-06,
+      "loss": 0.1407,
+      "step": 16070
+    },
+    {
+      "epoch": 12.097101994730899,
+      "grad_norm": 0.2023864984512329,
+      "learning_rate": 5e-06,
+      "loss": 0.1267,
+      "step": 16071
+    },
+    {
+      "epoch": 12.097854723372224,
+      "grad_norm": 0.16505871713161469,
+      "learning_rate": 5e-06,
+      "loss": 0.1031,
+      "step": 16072
+    },
+    {
+      "epoch": 12.09860745201355,
+      "grad_norm": 0.18126307427883148,
+      "learning_rate": 5e-06,
+      "loss": 0.149,
+      "step": 16073
+    },
+    {
+      "epoch": 12.099360180654873,
+      "grad_norm": 0.19744041562080383,
+      "learning_rate": 5e-06,
+      "loss": 0.0911,
+      "step": 16074
+    },
+    {
+      "epoch": 12.100112909296199,
+      "grad_norm": 0.19508522748947144,
+      "learning_rate": 5e-06,
+      "loss": 0.2289,
+      "step": 16075
+    },
+    {
+      "epoch": 12.100865637937524,
+      "grad_norm": 0.19465571641921997,
+      "learning_rate": 5e-06,
+      "loss": 0.184,
+      "step": 16076
+    },
+    {
+      "epoch": 12.101618366578847,
+      "grad_norm": 0.1947617083787918,
+      "learning_rate": 5e-06,
+      "loss": 0.0897,
+      "step": 16077
+    },
+    {
+      "epoch": 12.102371095220173,
+      "grad_norm": 0.19853661954402924,
+      "learning_rate": 5e-06,
+      "loss": 0.2166,
+      "step": 16078
+    },
+    {
+      "epoch": 12.103123823861498,
+      "grad_norm": 0.25993767380714417,
+      "learning_rate": 5e-06,
+      "loss": 0.1989,
+      "step": 16079
+    },
+    {
+      "epoch": 12.103876552502824,
+      "grad_norm": 0.18951073288917542,
+      "learning_rate": 5e-06,
+      "loss": 0.1606,
+      "step": 16080
+    },
+    {
+      "epoch": 12.104629281144147,
+      "grad_norm": 0.194418802857399,
+      "learning_rate": 5e-06,
+      "loss": 0.1626,
+      "step": 16081
+    },
+    {
+      "epoch": 12.105382009785473,
+      "grad_norm": 0.1934630125761032,
+      "learning_rate": 5e-06,
+      "loss": 0.1557,
+      "step": 16082
+    },
+    {
+      "epoch": 12.106134738426798,
+      "grad_norm": 0.16652651131153107,
+      "learning_rate": 5e-06,
+      "loss": 0.1174,
+      "step": 16083
+    },
+    {
+      "epoch": 12.106887467068121,
+      "grad_norm": 0.1686403751373291,
+      "learning_rate": 5e-06,
+      "loss": 0.1543,
+      "step": 16084
+    },
+    {
+      "epoch": 12.107640195709447,
+      "grad_norm": 0.2207777500152588,
+      "learning_rate": 5e-06,
+      "loss": 0.1924,
+      "step": 16085
+    },
+    {
+      "epoch": 12.108392924350772,
+      "grad_norm": 0.17770113050937653,
+      "learning_rate": 5e-06,
+      "loss": 0.0668,
+      "step": 16086
+    },
+    {
+      "epoch": 12.109145652992096,
+      "grad_norm": 0.22244858741760254,
+      "learning_rate": 5e-06,
+      "loss": 0.1192,
+      "step": 16087
+    },
+    {
+      "epoch": 12.109898381633421,
+      "grad_norm": 0.19627724587917328,
+      "learning_rate": 5e-06,
+      "loss": 0.1514,
+      "step": 16088
+    },
+    {
+      "epoch": 12.110651110274747,
+      "grad_norm": 0.16487397253513336,
+      "learning_rate": 5e-06,
+      "loss": 0.0965,
+      "step": 16089
+    },
+    {
+      "epoch": 12.11140383891607,
+      "grad_norm": 0.1994566172361374,
+      "learning_rate": 5e-06,
+      "loss": 0.205,
+      "step": 16090
+    },
+    {
+      "epoch": 12.112156567557395,
+      "grad_norm": 0.19886493682861328,
+      "learning_rate": 5e-06,
+      "loss": 0.1837,
+      "step": 16091
+    },
+    {
+      "epoch": 12.11290929619872,
+      "grad_norm": 0.17417240142822266,
+      "learning_rate": 5e-06,
+      "loss": 0.1676,
+      "step": 16092
+    },
+    {
+      "epoch": 12.113662024840044,
+      "grad_norm": 0.19069723784923553,
+      "learning_rate": 5e-06,
+      "loss": 0.1471,
+      "step": 16093
+    },
+    {
+      "epoch": 12.11441475348137,
+      "grad_norm": 0.17288683354854584,
+      "learning_rate": 5e-06,
+      "loss": 0.1025,
+      "step": 16094
+    },
+    {
+      "epoch": 12.115167482122695,
+      "grad_norm": 0.16807246208190918,
+      "learning_rate": 5e-06,
+      "loss": 0.1611,
+      "step": 16095
+    },
+    {
+      "epoch": 12.115920210764019,
+      "grad_norm": 0.16197003424167633,
+      "learning_rate": 5e-06,
+      "loss": 0.1494,
+      "step": 16096
+    },
+    {
+      "epoch": 12.116672939405344,
+      "grad_norm": 0.19195327162742615,
+      "learning_rate": 5e-06,
+      "loss": 0.1715,
+      "step": 16097
+    },
+    {
+      "epoch": 12.11742566804667,
+      "grad_norm": 0.20664916932582855,
+      "learning_rate": 5e-06,
+      "loss": 0.1664,
+      "step": 16098
+    },
+    {
+      "epoch": 12.118178396687995,
+      "grad_norm": 0.22246523201465607,
+      "learning_rate": 5e-06,
+      "loss": 0.1621,
+      "step": 16099
+    },
+    {
+      "epoch": 12.118931125329318,
+      "grad_norm": 0.17550352215766907,
+      "learning_rate": 5e-06,
+      "loss": 0.1215,
+      "step": 16100
+    },
+    {
+      "epoch": 12.119683853970644,
+      "grad_norm": 0.15292437374591827,
+      "learning_rate": 5e-06,
+      "loss": 0.0741,
+      "step": 16101
+    },
+    {
+      "epoch": 12.120436582611969,
+      "grad_norm": 0.1965799629688263,
+      "learning_rate": 5e-06,
+      "loss": 0.2067,
+      "step": 16102
+    },
+    {
+      "epoch": 12.121189311253293,
+      "grad_norm": 0.23601096868515015,
+      "learning_rate": 5e-06,
+      "loss": 0.1218,
+      "step": 16103
+    },
+    {
+      "epoch": 12.121942039894618,
+      "grad_norm": 0.18303100764751434,
+      "learning_rate": 5e-06,
+      "loss": 0.1634,
+      "step": 16104
+    },
+    {
+      "epoch": 12.122694768535943,
+      "grad_norm": 0.19264088571071625,
+      "learning_rate": 5e-06,
+      "loss": 0.1556,
+      "step": 16105
+    },
+    {
+      "epoch": 12.123447497177267,
+      "grad_norm": 0.16889338195323944,
+      "learning_rate": 5e-06,
+      "loss": 0.2034,
+      "step": 16106
+    },
+    {
+      "epoch": 12.124200225818592,
+      "grad_norm": 0.1609502136707306,
+      "learning_rate": 5e-06,
+      "loss": 0.1227,
+      "step": 16107
+    },
+    {
+      "epoch": 12.124952954459918,
+      "grad_norm": 0.163804292678833,
+      "learning_rate": 5e-06,
+      "loss": 0.1117,
+      "step": 16108
+    },
+    {
+      "epoch": 12.125705683101241,
+      "grad_norm": 0.18322400748729706,
+      "learning_rate": 5e-06,
+      "loss": 0.0988,
+      "step": 16109
+    },
+    {
+      "epoch": 12.126458411742567,
+      "grad_norm": 0.21675826609134674,
+      "learning_rate": 5e-06,
+      "loss": 0.1925,
+      "step": 16110
+    },
+    {
+      "epoch": 12.127211140383892,
+      "grad_norm": 0.17542840540409088,
+      "learning_rate": 5e-06,
+      "loss": 0.1939,
+      "step": 16111
+    },
+    {
+      "epoch": 12.127963869025216,
+      "grad_norm": 0.1727764755487442,
+      "learning_rate": 5e-06,
+      "loss": 0.1074,
+      "step": 16112
+    },
+    {
+      "epoch": 12.128716597666541,
+      "grad_norm": 0.16818509995937347,
+      "learning_rate": 5e-06,
+      "loss": 0.1135,
+      "step": 16113
+    },
+    {
+      "epoch": 12.129469326307866,
+      "grad_norm": 0.17817793786525726,
+      "learning_rate": 5e-06,
+      "loss": 0.1649,
+      "step": 16114
+    },
+    {
+      "epoch": 12.13022205494919,
+      "grad_norm": 0.19929780066013336,
+      "learning_rate": 5e-06,
+      "loss": 0.1825,
+      "step": 16115
+    },
+    {
+      "epoch": 12.130974783590515,
+      "grad_norm": 0.1662866473197937,
+      "learning_rate": 5e-06,
+      "loss": 0.0817,
+      "step": 16116
+    },
+    {
+      "epoch": 12.13172751223184,
+      "grad_norm": 0.17082682251930237,
+      "learning_rate": 5e-06,
+      "loss": 0.1044,
+      "step": 16117
+    },
+    {
+      "epoch": 12.132480240873166,
+      "grad_norm": 0.19436535239219666,
+      "learning_rate": 5e-06,
+      "loss": 0.1519,
+      "step": 16118
+    },
+    {
+      "epoch": 12.13323296951449,
+      "grad_norm": 0.1813293844461441,
+      "learning_rate": 5e-06,
+      "loss": 0.1466,
+      "step": 16119
+    },
+    {
+      "epoch": 12.133985698155815,
+      "grad_norm": 0.21104174852371216,
+      "learning_rate": 5e-06,
+      "loss": 0.1777,
+      "step": 16120
+    },
+    {
+      "epoch": 12.13473842679714,
+      "grad_norm": 0.1607496440410614,
+      "learning_rate": 5e-06,
+      "loss": 0.0763,
+      "step": 16121
+    },
+    {
+      "epoch": 12.135491155438464,
+      "grad_norm": 0.19094112515449524,
+      "learning_rate": 5e-06,
+      "loss": 0.1612,
+      "step": 16122
+    },
+    {
+      "epoch": 12.13624388407979,
+      "grad_norm": 0.17743568122386932,
+      "learning_rate": 5e-06,
+      "loss": 0.0932,
+      "step": 16123
+    },
+    {
+      "epoch": 12.136996612721115,
+      "grad_norm": 0.1973499208688736,
+      "learning_rate": 5e-06,
+      "loss": 0.1597,
+      "step": 16124
+    },
+    {
+      "epoch": 12.137749341362438,
+      "grad_norm": 0.19571737945079803,
+      "learning_rate": 5e-06,
+      "loss": 0.1585,
+      "step": 16125
+    },
+    {
+      "epoch": 12.138502070003764,
+      "grad_norm": 0.16797980666160583,
+      "learning_rate": 5e-06,
+      "loss": 0.1294,
+      "step": 16126
+    },
+    {
+      "epoch": 12.139254798645089,
+      "grad_norm": 0.19307048618793488,
+      "learning_rate": 5e-06,
+      "loss": 0.1293,
+      "step": 16127
+    },
+    {
+      "epoch": 12.140007527286413,
+      "grad_norm": 0.17609140276908875,
+      "learning_rate": 5e-06,
+      "loss": 0.1861,
+      "step": 16128
+    },
+    {
+      "epoch": 12.140760255927738,
+      "grad_norm": 0.2217116504907608,
+      "learning_rate": 5e-06,
+      "loss": 0.2003,
+      "step": 16129
+    },
+    {
+      "epoch": 12.141512984569063,
+      "grad_norm": 0.16645167768001556,
+      "learning_rate": 5e-06,
+      "loss": 0.2002,
+      "step": 16130
+    },
+    {
+      "epoch": 12.142265713210387,
+      "grad_norm": 0.19638386368751526,
+      "learning_rate": 5e-06,
+      "loss": 0.1301,
+      "step": 16131
+    },
+    {
+      "epoch": 12.143018441851712,
+      "grad_norm": 0.215654656291008,
+      "learning_rate": 5e-06,
+      "loss": 0.1852,
+      "step": 16132
+    },
+    {
+      "epoch": 12.143771170493038,
+      "grad_norm": 0.16966475546360016,
+      "learning_rate": 5e-06,
+      "loss": 0.0909,
+      "step": 16133
+    },
+    {
+      "epoch": 12.144523899134363,
+      "grad_norm": 0.17494668066501617,
+      "learning_rate": 5e-06,
+      "loss": 0.107,
+      "step": 16134
+    },
+    {
+      "epoch": 12.145276627775687,
+      "grad_norm": 0.18685102462768555,
+      "learning_rate": 5e-06,
+      "loss": 0.1295,
+      "step": 16135
+    },
+    {
+      "epoch": 12.146029356417012,
+      "grad_norm": 0.16115252673625946,
+      "learning_rate": 5e-06,
+      "loss": 0.0834,
+      "step": 16136
+    },
+    {
+      "epoch": 12.146782085058337,
+      "grad_norm": 0.1802385002374649,
+      "learning_rate": 5e-06,
+      "loss": 0.1445,
+      "step": 16137
+    },
+    {
+      "epoch": 12.14753481369966,
+      "grad_norm": 0.21174106001853943,
+      "learning_rate": 5e-06,
+      "loss": 0.1538,
+      "step": 16138
+    },
+    {
+      "epoch": 12.148287542340986,
+      "grad_norm": 0.15693525969982147,
+      "learning_rate": 5e-06,
+      "loss": 0.1039,
+      "step": 16139
+    },
+    {
+      "epoch": 12.149040270982312,
+      "grad_norm": 0.17187146842479706,
+      "learning_rate": 5e-06,
+      "loss": 0.1039,
+      "step": 16140
+    },
+    {
+      "epoch": 12.149792999623635,
+      "grad_norm": 0.21176807582378387,
+      "learning_rate": 5e-06,
+      "loss": 0.2091,
+      "step": 16141
+    },
+    {
+      "epoch": 12.15054572826496,
+      "grad_norm": 0.18280695378780365,
+      "learning_rate": 5e-06,
+      "loss": 0.1143,
+      "step": 16142
+    },
+    {
+      "epoch": 12.151298456906286,
+      "grad_norm": 0.1574554294347763,
+      "learning_rate": 5e-06,
+      "loss": 0.117,
+      "step": 16143
+    },
+    {
+      "epoch": 12.15205118554761,
+      "grad_norm": 0.18422996997833252,
+      "learning_rate": 5e-06,
+      "loss": 0.1253,
+      "step": 16144
+    },
+    {
+      "epoch": 12.152803914188935,
+      "grad_norm": 0.1721314787864685,
+      "learning_rate": 5e-06,
+      "loss": 0.1601,
+      "step": 16145
+    },
+    {
+      "epoch": 12.15355664283026,
+      "grad_norm": 0.21187685430049896,
+      "learning_rate": 5e-06,
+      "loss": 0.1877,
+      "step": 16146
+    },
+    {
+      "epoch": 12.154309371471584,
+      "grad_norm": 0.18177469074726105,
+      "learning_rate": 5e-06,
+      "loss": 0.1505,
+      "step": 16147
+    },
+    {
+      "epoch": 12.155062100112909,
+      "grad_norm": 0.1697554886341095,
+      "learning_rate": 5e-06,
+      "loss": 0.1175,
+      "step": 16148
+    },
+    {
+      "epoch": 12.155814828754234,
+      "grad_norm": 0.19052977859973907,
+      "learning_rate": 5e-06,
+      "loss": 0.1123,
+      "step": 16149
+    },
+    {
+      "epoch": 12.156567557395558,
+      "grad_norm": 0.19467014074325562,
+      "learning_rate": 5e-06,
+      "loss": 0.1758,
+      "step": 16150
+    },
+    {
+      "epoch": 12.157320286036883,
+      "grad_norm": 0.19930431246757507,
+      "learning_rate": 5e-06,
+      "loss": 0.1527,
+      "step": 16151
+    },
+    {
+      "epoch": 12.158073014678209,
+      "grad_norm": 0.19130364060401917,
+      "learning_rate": 5e-06,
+      "loss": 0.1301,
+      "step": 16152
+    },
+    {
+      "epoch": 12.158825743319534,
+      "grad_norm": 0.18191158771514893,
+      "learning_rate": 5e-06,
+      "loss": 0.1108,
+      "step": 16153
+    },
+    {
+      "epoch": 12.159578471960858,
+      "grad_norm": 0.17718029022216797,
+      "learning_rate": 5e-06,
+      "loss": 0.1084,
+      "step": 16154
+    },
+    {
+      "epoch": 12.160331200602183,
+      "grad_norm": 0.1916397362947464,
+      "learning_rate": 5e-06,
+      "loss": 0.1008,
+      "step": 16155
+    },
+    {
+      "epoch": 12.161083929243508,
+      "grad_norm": 0.19094111025333405,
+      "learning_rate": 5e-06,
+      "loss": 0.1053,
+      "step": 16156
+    },
+    {
+      "epoch": 12.161836657884832,
+      "grad_norm": 0.1919124722480774,
+      "learning_rate": 5e-06,
+      "loss": 0.1455,
+      "step": 16157
+    },
+    {
+      "epoch": 12.162589386526157,
+      "grad_norm": 0.1803586483001709,
+      "learning_rate": 5e-06,
+      "loss": 0.1175,
+      "step": 16158
+    },
+    {
+      "epoch": 12.163342115167483,
+      "grad_norm": 0.19459311664104462,
+      "learning_rate": 5e-06,
+      "loss": 0.1171,
+      "step": 16159
+    },
+    {
+      "epoch": 12.164094843808806,
+      "grad_norm": 0.1905137598514557,
+      "learning_rate": 5e-06,
+      "loss": 0.0843,
+      "step": 16160
+    },
+    {
+      "epoch": 12.164847572450132,
+      "grad_norm": 0.22281785309314728,
+      "learning_rate": 5e-06,
+      "loss": 0.1816,
+      "step": 16161
+    },
+    {
+      "epoch": 12.165600301091457,
+      "grad_norm": 0.17944201827049255,
+      "learning_rate": 5e-06,
+      "loss": 0.1562,
+      "step": 16162
+    },
+    {
+      "epoch": 12.16635302973278,
+      "grad_norm": 0.17254768311977386,
+      "learning_rate": 5e-06,
+      "loss": 0.0775,
+      "step": 16163
+    },
+    {
+      "epoch": 12.167105758374106,
+      "grad_norm": 0.18713954091072083,
+      "learning_rate": 5e-06,
+      "loss": 0.1461,
+      "step": 16164
+    },
+    {
+      "epoch": 12.167858487015431,
+      "grad_norm": 0.1808730661869049,
+      "learning_rate": 5e-06,
+      "loss": 0.2048,
+      "step": 16165
+    },
+    {
+      "epoch": 12.168611215656755,
+      "grad_norm": 0.1913261115550995,
+      "learning_rate": 5e-06,
+      "loss": 0.1293,
+      "step": 16166
+    },
+    {
+      "epoch": 12.16936394429808,
+      "grad_norm": 0.16405357420444489,
+      "learning_rate": 5e-06,
+      "loss": 0.1078,
+      "step": 16167
+    },
+    {
+      "epoch": 12.170116672939406,
+      "grad_norm": 0.21036168932914734,
+      "learning_rate": 5e-06,
+      "loss": 0.1934,
+      "step": 16168
+    },
+    {
+      "epoch": 12.17086940158073,
+      "grad_norm": 0.17477872967720032,
+      "learning_rate": 5e-06,
+      "loss": 0.15,
+      "step": 16169
+    },
+    {
+      "epoch": 12.171622130222055,
+      "grad_norm": 0.19170540571212769,
+      "learning_rate": 5e-06,
+      "loss": 0.1933,
+      "step": 16170
+    },
+    {
+      "epoch": 12.17237485886338,
+      "grad_norm": 0.16740447282791138,
+      "learning_rate": 5e-06,
+      "loss": 0.0886,
+      "step": 16171
+    },
+    {
+      "epoch": 12.173127587504705,
+      "grad_norm": 0.19341962039470673,
+      "learning_rate": 5e-06,
+      "loss": 0.1552,
+      "step": 16172
+    },
+    {
+      "epoch": 12.173880316146029,
+      "grad_norm": 0.1649918109178543,
+      "learning_rate": 5e-06,
+      "loss": 0.172,
+      "step": 16173
+    },
+    {
+      "epoch": 12.174633044787354,
+      "grad_norm": 0.15387386083602905,
+      "learning_rate": 5e-06,
+      "loss": 0.089,
+      "step": 16174
+    },
+    {
+      "epoch": 12.17538577342868,
+      "grad_norm": 0.18667180836200714,
+      "learning_rate": 5e-06,
+      "loss": 0.1405,
+      "step": 16175
+    },
+    {
+      "epoch": 12.176138502070003,
+      "grad_norm": 0.17994536459445953,
+      "learning_rate": 5e-06,
+      "loss": 0.0934,
+      "step": 16176
+    },
+    {
+      "epoch": 12.176891230711329,
+      "grad_norm": 0.20036792755126953,
+      "learning_rate": 5e-06,
+      "loss": 0.2428,
+      "step": 16177
+    },
+    {
+      "epoch": 12.177643959352654,
+      "grad_norm": 0.20076705515384674,
+      "learning_rate": 5e-06,
+      "loss": 0.253,
+      "step": 16178
+    },
+    {
+      "epoch": 12.178396687993978,
+      "grad_norm": 0.17971593141555786,
+      "learning_rate": 5e-06,
+      "loss": 0.1298,
+      "step": 16179
+    },
+    {
+      "epoch": 12.179149416635303,
+      "grad_norm": 0.19609184563159943,
+      "learning_rate": 5e-06,
+      "loss": 0.1296,
+      "step": 16180
+    },
+    {
+      "epoch": 12.179902145276628,
+      "grad_norm": 0.18165840208530426,
+      "learning_rate": 5e-06,
+      "loss": 0.1873,
+      "step": 16181
+    },
+    {
+      "epoch": 12.180654873917952,
+      "grad_norm": 0.21887865662574768,
+      "learning_rate": 5e-06,
+      "loss": 0.1371,
+      "step": 16182
+    },
+    {
+      "epoch": 12.181407602559277,
+      "grad_norm": 0.17211468517780304,
+      "learning_rate": 5e-06,
+      "loss": 0.0741,
+      "step": 16183
+    },
+    {
+      "epoch": 12.182160331200603,
+      "grad_norm": 0.17055760324001312,
+      "learning_rate": 5e-06,
+      "loss": 0.1268,
+      "step": 16184
+    },
+    {
+      "epoch": 12.182913059841926,
+      "grad_norm": 0.17653048038482666,
+      "learning_rate": 5e-06,
+      "loss": 0.1584,
+      "step": 16185
+    },
+    {
+      "epoch": 12.183665788483252,
+      "grad_norm": 0.16333116590976715,
+      "learning_rate": 5e-06,
+      "loss": 0.1717,
+      "step": 16186
+    },
+    {
+      "epoch": 12.184418517124577,
+      "grad_norm": 0.16513296961784363,
+      "learning_rate": 5e-06,
+      "loss": 0.1463,
+      "step": 16187
+    },
+    {
+      "epoch": 12.1851712457659,
+      "grad_norm": 0.17069099843502045,
+      "learning_rate": 5e-06,
+      "loss": 0.1491,
+      "step": 16188
+    },
+    {
+      "epoch": 12.185923974407226,
+      "grad_norm": 0.16497863829135895,
+      "learning_rate": 5e-06,
+      "loss": 0.1341,
+      "step": 16189
+    },
+    {
+      "epoch": 12.186676703048551,
+      "grad_norm": 0.16887323558330536,
+      "learning_rate": 5e-06,
+      "loss": 0.1523,
+      "step": 16190
+    },
+    {
+      "epoch": 12.187429431689877,
+      "grad_norm": 0.17341890931129456,
+      "learning_rate": 5e-06,
+      "loss": 0.1176,
+      "step": 16191
+    },
+    {
+      "epoch": 12.1881821603312,
+      "grad_norm": 0.19735218584537506,
+      "learning_rate": 5e-06,
+      "loss": 0.162,
+      "step": 16192
+    },
+    {
+      "epoch": 12.188934888972526,
+      "grad_norm": 0.16732025146484375,
+      "learning_rate": 5e-06,
+      "loss": 0.1253,
+      "step": 16193
+    },
+    {
+      "epoch": 12.189687617613851,
+      "grad_norm": 0.17652837932109833,
+      "learning_rate": 5e-06,
+      "loss": 0.1303,
+      "step": 16194
+    },
+    {
+      "epoch": 12.190440346255174,
+      "grad_norm": 0.20673251152038574,
+      "learning_rate": 5e-06,
+      "loss": 0.103,
+      "step": 16195
+    },
+    {
+      "epoch": 12.1911930748965,
+      "grad_norm": 0.18477796018123627,
+      "learning_rate": 5e-06,
+      "loss": 0.1847,
+      "step": 16196
+    },
+    {
+      "epoch": 12.191945803537825,
+      "grad_norm": 0.19669607281684875,
+      "learning_rate": 5e-06,
+      "loss": 0.1457,
+      "step": 16197
+    },
+    {
+      "epoch": 12.192698532179149,
+      "grad_norm": 0.17614971101284027,
+      "learning_rate": 5e-06,
+      "loss": 0.1285,
+      "step": 16198
+    },
+    {
+      "epoch": 12.193451260820474,
+      "grad_norm": 0.17309173941612244,
+      "learning_rate": 5e-06,
+      "loss": 0.1452,
+      "step": 16199
+    },
+    {
+      "epoch": 12.1942039894618,
+      "grad_norm": 0.19592811167240143,
+      "learning_rate": 5e-06,
+      "loss": 0.138,
+      "step": 16200
+    },
+    {
+      "epoch": 12.1942039894618,
+      "eval_loss": 0.18366855382919312,
+      "eval_runtime": 452.8857,
+      "eval_samples_per_second": 21.257,
+      "eval_steps_per_second": 0.665,
+      "step": 16200
+    },
+    {
+      "epoch": 12.194956718103123,
+      "grad_norm": 0.18742206692695618,
+      "learning_rate": 5e-06,
+      "loss": 0.1166,
+      "step": 16201
+    },
+    {
+      "epoch": 12.195709446744448,
+      "grad_norm": 0.18842940032482147,
+      "learning_rate": 5e-06,
+      "loss": 0.175,
+      "step": 16202
+    },
+    {
+      "epoch": 12.196462175385774,
+      "grad_norm": 0.163364440202713,
+      "learning_rate": 5e-06,
+      "loss": 0.101,
+      "step": 16203
+    },
+    {
+      "epoch": 12.197214904027097,
+      "grad_norm": 0.23110376298427582,
+      "learning_rate": 5e-06,
+      "loss": 0.1671,
+      "step": 16204
+    },
+    {
+      "epoch": 12.197967632668423,
+      "grad_norm": 0.1814289540052414,
+      "learning_rate": 5e-06,
+      "loss": 0.1534,
+      "step": 16205
+    },
+    {
+      "epoch": 12.198720361309748,
+      "grad_norm": 0.17492879927158356,
+      "learning_rate": 5e-06,
+      "loss": 0.1337,
+      "step": 16206
+    },
+    {
+      "epoch": 12.199473089951073,
+      "grad_norm": 0.20875076949596405,
+      "learning_rate": 5e-06,
+      "loss": 0.1895,
+      "step": 16207
+    },
+    {
+      "epoch": 12.200225818592397,
+      "grad_norm": 0.19151626527309418,
+      "learning_rate": 5e-06,
+      "loss": 0.1272,
+      "step": 16208
+    },
+    {
+      "epoch": 12.200978547233722,
+      "grad_norm": 0.17591863870620728,
+      "learning_rate": 5e-06,
+      "loss": 0.126,
+      "step": 16209
+    },
+    {
+      "epoch": 12.201731275875048,
+      "grad_norm": 0.18472479283809662,
+      "learning_rate": 5e-06,
+      "loss": 0.1963,
+      "step": 16210
+    },
+    {
+      "epoch": 12.202484004516371,
+      "grad_norm": 0.1598520576953888,
+      "learning_rate": 5e-06,
+      "loss": 0.0823,
+      "step": 16211
+    },
+    {
+      "epoch": 12.203236733157697,
+      "grad_norm": 0.1800847053527832,
+      "learning_rate": 5e-06,
+      "loss": 0.1177,
+      "step": 16212
+    },
+    {
+      "epoch": 12.203989461799022,
+      "grad_norm": 0.21487219631671906,
+      "learning_rate": 5e-06,
+      "loss": 0.1124,
+      "step": 16213
+    },
+    {
+      "epoch": 12.204742190440346,
+      "grad_norm": 0.19024331867694855,
+      "learning_rate": 5e-06,
+      "loss": 0.0953,
+      "step": 16214
+    },
+    {
+      "epoch": 12.205494919081671,
+      "grad_norm": 0.21360337734222412,
+      "learning_rate": 5e-06,
+      "loss": 0.1777,
+      "step": 16215
+    },
+    {
+      "epoch": 12.206247647722996,
+      "grad_norm": 0.193354070186615,
+      "learning_rate": 5e-06,
+      "loss": 0.1141,
+      "step": 16216
+    },
+    {
+      "epoch": 12.20700037636432,
+      "grad_norm": 0.16598330438137054,
+      "learning_rate": 5e-06,
+      "loss": 0.0811,
+      "step": 16217
+    },
+    {
+      "epoch": 12.207753105005645,
+      "grad_norm": 0.18747512996196747,
+      "learning_rate": 5e-06,
+      "loss": 0.1568,
+      "step": 16218
+    },
+    {
+      "epoch": 12.20850583364697,
+      "grad_norm": 0.210530623793602,
+      "learning_rate": 5e-06,
+      "loss": 0.2379,
+      "step": 16219
+    },
+    {
+      "epoch": 12.209258562288294,
+      "grad_norm": 0.1713327318429947,
+      "learning_rate": 5e-06,
+      "loss": 0.1127,
+      "step": 16220
+    },
+    {
+      "epoch": 12.21001129092962,
+      "grad_norm": 0.1711658239364624,
+      "learning_rate": 5e-06,
+      "loss": 0.1621,
+      "step": 16221
+    },
+    {
+      "epoch": 12.210764019570945,
+      "grad_norm": 0.17899005115032196,
+      "learning_rate": 5e-06,
+      "loss": 0.1525,
+      "step": 16222
+    },
+    {
+      "epoch": 12.211516748212269,
+      "grad_norm": 0.2156423181295395,
+      "learning_rate": 5e-06,
+      "loss": 0.1332,
+      "step": 16223
+    },
+    {
+      "epoch": 12.212269476853594,
+      "grad_norm": 0.1662418693304062,
+      "learning_rate": 5e-06,
+      "loss": 0.1458,
+      "step": 16224
+    },
+    {
+      "epoch": 12.21302220549492,
+      "grad_norm": 0.18587641417980194,
+      "learning_rate": 5e-06,
+      "loss": 0.1462,
+      "step": 16225
+    },
+    {
+      "epoch": 12.213774934136245,
+      "grad_norm": 0.1863671839237213,
+      "learning_rate": 5e-06,
+      "loss": 0.0973,
+      "step": 16226
+    },
+    {
+      "epoch": 12.214527662777568,
+      "grad_norm": 0.1918182373046875,
+      "learning_rate": 5e-06,
+      "loss": 0.0958,
+      "step": 16227
+    },
+    {
+      "epoch": 12.215280391418894,
+      "grad_norm": 0.16248483955860138,
+      "learning_rate": 5e-06,
+      "loss": 0.1351,
+      "step": 16228
+    },
+    {
+      "epoch": 12.216033120060219,
+      "grad_norm": 0.19127260148525238,
+      "learning_rate": 5e-06,
+      "loss": 0.1175,
+      "step": 16229
+    },
+    {
+      "epoch": 12.216785848701543,
+      "grad_norm": 0.21964311599731445,
+      "learning_rate": 5e-06,
+      "loss": 0.1725,
+      "step": 16230
+    },
+    {
+      "epoch": 12.217538577342868,
+      "grad_norm": 0.19423313438892365,
+      "learning_rate": 5e-06,
+      "loss": 0.183,
+      "step": 16231
+    },
+    {
+      "epoch": 12.218291305984193,
+      "grad_norm": 0.21459472179412842,
+      "learning_rate": 5e-06,
+      "loss": 0.1454,
+      "step": 16232
+    },
+    {
+      "epoch": 12.219044034625517,
+      "grad_norm": 0.17635872960090637,
+      "learning_rate": 5e-06,
+      "loss": 0.1749,
+      "step": 16233
+    },
+    {
+      "epoch": 12.219796763266842,
+      "grad_norm": 0.21784847974777222,
+      "learning_rate": 5e-06,
+      "loss": 0.1762,
+      "step": 16234
+    },
+    {
+      "epoch": 12.220549491908168,
+      "grad_norm": 0.1973576545715332,
+      "learning_rate": 5e-06,
+      "loss": 0.1688,
+      "step": 16235
+    },
+    {
+      "epoch": 12.221302220549491,
+      "grad_norm": 0.19418860971927643,
+      "learning_rate": 5e-06,
+      "loss": 0.1108,
+      "step": 16236
+    },
+    {
+      "epoch": 12.222054949190817,
+      "grad_norm": 0.19950012862682343,
+      "learning_rate": 5e-06,
+      "loss": 0.1072,
+      "step": 16237
+    },
+    {
+      "epoch": 12.222807677832142,
+      "grad_norm": 0.1842374950647354,
+      "learning_rate": 5e-06,
+      "loss": 0.235,
+      "step": 16238
+    },
+    {
+      "epoch": 12.223560406473466,
+      "grad_norm": 0.18180342018604279,
+      "learning_rate": 5e-06,
+      "loss": 0.1808,
+      "step": 16239
+    },
+    {
+      "epoch": 12.224313135114791,
+      "grad_norm": 0.16858592629432678,
+      "learning_rate": 5e-06,
+      "loss": 0.0789,
+      "step": 16240
+    },
+    {
+      "epoch": 12.225065863756116,
+      "grad_norm": 0.1985209584236145,
+      "learning_rate": 5e-06,
+      "loss": 0.1632,
+      "step": 16241
+    },
+    {
+      "epoch": 12.225818592397442,
+      "grad_norm": 0.21338121592998505,
+      "learning_rate": 5e-06,
+      "loss": 0.1618,
+      "step": 16242
+    },
+    {
+      "epoch": 12.226571321038765,
+      "grad_norm": 0.18417668342590332,
+      "learning_rate": 5e-06,
+      "loss": 0.1195,
+      "step": 16243
+    },
+    {
+      "epoch": 12.22732404968009,
+      "grad_norm": 0.18052928149700165,
+      "learning_rate": 5e-06,
+      "loss": 0.1234,
+      "step": 16244
+    },
+    {
+      "epoch": 12.228076778321416,
+      "grad_norm": 0.18571561574935913,
+      "learning_rate": 5e-06,
+      "loss": 0.1883,
+      "step": 16245
+    },
+    {
+      "epoch": 12.22882950696274,
+      "grad_norm": 0.16694006323814392,
+      "learning_rate": 5e-06,
+      "loss": 0.135,
+      "step": 16246
+    },
+    {
+      "epoch": 12.229582235604065,
+      "grad_norm": 0.17139677703380585,
+      "learning_rate": 5e-06,
+      "loss": 0.1148,
+      "step": 16247
+    },
+    {
+      "epoch": 12.23033496424539,
+      "grad_norm": 0.18424804508686066,
+      "learning_rate": 5e-06,
+      "loss": 0.1459,
+      "step": 16248
+    },
+    {
+      "epoch": 12.231087692886714,
+      "grad_norm": 0.1863631159067154,
+      "learning_rate": 5e-06,
+      "loss": 0.0716,
+      "step": 16249
+    },
+    {
+      "epoch": 12.23184042152804,
+      "grad_norm": 0.15912102162837982,
+      "learning_rate": 5e-06,
+      "loss": 0.0901,
+      "step": 16250
+    },
+    {
+      "epoch": 12.232593150169365,
+      "grad_norm": 0.17864903807640076,
+      "learning_rate": 5e-06,
+      "loss": 0.1588,
+      "step": 16251
+    },
+    {
+      "epoch": 12.233345878810688,
+      "grad_norm": 0.18323136866092682,
+      "learning_rate": 5e-06,
+      "loss": 0.1224,
+      "step": 16252
+    },
+    {
+      "epoch": 12.234098607452013,
+      "grad_norm": 0.21343457698822021,
+      "learning_rate": 5e-06,
+      "loss": 0.1632,
+      "step": 16253
+    },
+    {
+      "epoch": 12.234851336093339,
+      "grad_norm": 0.1789226084947586,
+      "learning_rate": 5e-06,
+      "loss": 0.0885,
+      "step": 16254
+    },
+    {
+      "epoch": 12.235604064734662,
+      "grad_norm": 0.1778271198272705,
+      "learning_rate": 5e-06,
+      "loss": 0.1699,
+      "step": 16255
+    },
+    {
+      "epoch": 12.236356793375988,
+      "grad_norm": 0.1420617699623108,
+      "learning_rate": 5e-06,
+      "loss": 0.0561,
+      "step": 16256
+    },
+    {
+      "epoch": 12.237109522017313,
+      "grad_norm": 0.1794150471687317,
+      "learning_rate": 5e-06,
+      "loss": 0.0718,
+      "step": 16257
+    },
+    {
+      "epoch": 12.237862250658637,
+      "grad_norm": 0.19360190629959106,
+      "learning_rate": 5e-06,
+      "loss": 0.1591,
+      "step": 16258
+    },
+    {
+      "epoch": 12.238614979299962,
+      "grad_norm": 0.1829816699028015,
+      "learning_rate": 5e-06,
+      "loss": 0.1356,
+      "step": 16259
+    },
+    {
+      "epoch": 12.239367707941287,
+      "grad_norm": 0.19692383706569672,
+      "learning_rate": 5e-06,
+      "loss": 0.2083,
+      "step": 16260
+    },
+    {
+      "epoch": 12.240120436582613,
+      "grad_norm": 0.16894695162773132,
+      "learning_rate": 5e-06,
+      "loss": 0.0841,
+      "step": 16261
+    },
+    {
+      "epoch": 12.240873165223936,
+      "grad_norm": 0.17717179656028748,
+      "learning_rate": 5e-06,
+      "loss": 0.1045,
+      "step": 16262
+    },
+    {
+      "epoch": 12.241625893865262,
+      "grad_norm": 0.17512331902980804,
+      "learning_rate": 5e-06,
+      "loss": 0.1733,
+      "step": 16263
+    },
+    {
+      "epoch": 12.242378622506587,
+      "grad_norm": 0.18775801360607147,
+      "learning_rate": 5e-06,
+      "loss": 0.1348,
+      "step": 16264
+    },
+    {
+      "epoch": 12.24313135114791,
+      "grad_norm": 0.18880629539489746,
+      "learning_rate": 5e-06,
+      "loss": 0.148,
+      "step": 16265
+    },
+    {
+      "epoch": 12.243884079789236,
+      "grad_norm": 0.18422041833400726,
+      "learning_rate": 5e-06,
+      "loss": 0.1752,
+      "step": 16266
+    },
+    {
+      "epoch": 12.244636808430561,
+      "grad_norm": 0.14323066174983978,
+      "learning_rate": 5e-06,
+      "loss": 0.1085,
+      "step": 16267
+    },
+    {
+      "epoch": 12.245389537071885,
+      "grad_norm": 0.21295511722564697,
+      "learning_rate": 5e-06,
+      "loss": 0.1883,
+      "step": 16268
+    },
+    {
+      "epoch": 12.24614226571321,
+      "grad_norm": 0.16899248957633972,
+      "learning_rate": 5e-06,
+      "loss": 0.155,
+      "step": 16269
+    },
+    {
+      "epoch": 12.246894994354536,
+      "grad_norm": 0.17464947700500488,
+      "learning_rate": 5e-06,
+      "loss": 0.1093,
+      "step": 16270
+    },
+    {
+      "epoch": 12.24764772299586,
+      "grad_norm": 0.1619589626789093,
+      "learning_rate": 5e-06,
+      "loss": 0.1607,
+      "step": 16271
+    },
+    {
+      "epoch": 12.248400451637185,
+      "grad_norm": 0.17853209376335144,
+      "learning_rate": 5e-06,
+      "loss": 0.0871,
+      "step": 16272
+    },
+    {
+      "epoch": 12.24915318027851,
+      "grad_norm": 0.18439055979251862,
+      "learning_rate": 5e-06,
+      "loss": 0.0902,
+      "step": 16273
+    },
+    {
+      "epoch": 12.249905908919834,
+      "grad_norm": 0.1675071269273758,
+      "learning_rate": 5e-06,
+      "loss": 0.1699,
+      "step": 16274
+    },
+    {
+      "epoch": 12.250658637561159,
+      "grad_norm": 0.19085784256458282,
+      "learning_rate": 5e-06,
+      "loss": 0.0981,
+      "step": 16275
+    },
+    {
+      "epoch": 12.251411366202484,
+      "grad_norm": 0.18758268654346466,
+      "learning_rate": 5e-06,
+      "loss": 0.163,
+      "step": 16276
+    },
+    {
+      "epoch": 12.252164094843808,
+      "grad_norm": 0.20926252007484436,
+      "learning_rate": 5e-06,
+      "loss": 0.1852,
+      "step": 16277
+    },
+    {
+      "epoch": 12.252916823485133,
+      "grad_norm": 0.19155727326869965,
+      "learning_rate": 5e-06,
+      "loss": 0.1713,
+      "step": 16278
+    },
+    {
+      "epoch": 12.253669552126459,
+      "grad_norm": 0.18136075139045715,
+      "learning_rate": 5e-06,
+      "loss": 0.1363,
+      "step": 16279
+    },
+    {
+      "epoch": 12.254422280767784,
+      "grad_norm": 0.17898112535476685,
+      "learning_rate": 5e-06,
+      "loss": 0.1368,
+      "step": 16280
+    },
+    {
+      "epoch": 12.255175009409108,
+      "grad_norm": 0.19875845313072205,
+      "learning_rate": 5e-06,
+      "loss": 0.1261,
+      "step": 16281
+    },
+    {
+      "epoch": 12.255927738050433,
+      "grad_norm": 0.17846640944480896,
+      "learning_rate": 5e-06,
+      "loss": 0.1227,
+      "step": 16282
+    },
+    {
+      "epoch": 12.256680466691758,
+      "grad_norm": 0.20386742055416107,
+      "learning_rate": 5e-06,
+      "loss": 0.1469,
+      "step": 16283
+    },
+    {
+      "epoch": 12.257433195333082,
+      "grad_norm": 0.18238097429275513,
+      "learning_rate": 5e-06,
+      "loss": 0.0778,
+      "step": 16284
+    },
+    {
+      "epoch": 12.258185923974407,
+      "grad_norm": 0.1798887550830841,
+      "learning_rate": 5e-06,
+      "loss": 0.1393,
+      "step": 16285
+    },
+    {
+      "epoch": 12.258938652615733,
+      "grad_norm": 0.16933754086494446,
+      "learning_rate": 5e-06,
+      "loss": 0.1173,
+      "step": 16286
+    },
+    {
+      "epoch": 12.259691381257056,
+      "grad_norm": 0.1553378701210022,
+      "learning_rate": 5e-06,
+      "loss": 0.0865,
+      "step": 16287
+    },
+    {
+      "epoch": 12.260444109898382,
+      "grad_norm": 0.1538388580083847,
+      "learning_rate": 5e-06,
+      "loss": 0.0606,
+      "step": 16288
+    },
+    {
+      "epoch": 12.261196838539707,
+      "grad_norm": 0.16727672517299652,
+      "learning_rate": 5e-06,
+      "loss": 0.1041,
+      "step": 16289
+    },
+    {
+      "epoch": 12.26194956718103,
+      "grad_norm": 0.20590193569660187,
+      "learning_rate": 5e-06,
+      "loss": 0.1107,
+      "step": 16290
+    },
+    {
+      "epoch": 12.262702295822356,
+      "grad_norm": 0.16499803960323334,
+      "learning_rate": 5e-06,
+      "loss": 0.1265,
+      "step": 16291
+    },
+    {
+      "epoch": 12.263455024463681,
+      "grad_norm": 0.18282733857631683,
+      "learning_rate": 5e-06,
+      "loss": 0.1427,
+      "step": 16292
+    },
+    {
+      "epoch": 12.264207753105005,
+      "grad_norm": 0.1756349802017212,
+      "learning_rate": 5e-06,
+      "loss": 0.0819,
+      "step": 16293
+    },
+    {
+      "epoch": 12.26496048174633,
+      "grad_norm": 0.1618480235338211,
+      "learning_rate": 5e-06,
+      "loss": 0.0844,
+      "step": 16294
+    },
+    {
+      "epoch": 12.265713210387656,
+      "grad_norm": 0.21079474687576294,
+      "learning_rate": 5e-06,
+      "loss": 0.1241,
+      "step": 16295
+    },
+    {
+      "epoch": 12.26646593902898,
+      "grad_norm": 0.18288303911685944,
+      "learning_rate": 5e-06,
+      "loss": 0.1004,
+      "step": 16296
+    },
+    {
+      "epoch": 12.267218667670305,
+      "grad_norm": 0.17122583091259003,
+      "learning_rate": 5e-06,
+      "loss": 0.1172,
+      "step": 16297
+    },
+    {
+      "epoch": 12.26797139631163,
+      "grad_norm": 0.17926675081253052,
+      "learning_rate": 5e-06,
+      "loss": 0.1062,
+      "step": 16298
+    },
+    {
+      "epoch": 12.268724124952955,
+      "grad_norm": 0.1839887797832489,
+      "learning_rate": 5e-06,
+      "loss": 0.1655,
+      "step": 16299
+    },
+    {
+      "epoch": 12.269476853594279,
+      "grad_norm": 0.16466102004051208,
+      "learning_rate": 5e-06,
+      "loss": 0.1065,
+      "step": 16300
+    },
+    {
+      "epoch": 12.270229582235604,
+      "grad_norm": 0.1706903725862503,
+      "learning_rate": 5e-06,
+      "loss": 0.225,
+      "step": 16301
+    },
+    {
+      "epoch": 12.27098231087693,
+      "grad_norm": 0.18569231033325195,
+      "learning_rate": 5e-06,
+      "loss": 0.1809,
+      "step": 16302
+    },
+    {
+      "epoch": 12.271735039518253,
+      "grad_norm": 0.21409937739372253,
+      "learning_rate": 5e-06,
+      "loss": 0.2496,
+      "step": 16303
+    },
+    {
+      "epoch": 12.272487768159579,
+      "grad_norm": 0.15354380011558533,
+      "learning_rate": 5e-06,
+      "loss": 0.0945,
+      "step": 16304
+    },
+    {
+      "epoch": 12.273240496800904,
+      "grad_norm": 0.1629246473312378,
+      "learning_rate": 5e-06,
+      "loss": 0.1388,
+      "step": 16305
+    },
+    {
+      "epoch": 12.273993225442227,
+      "grad_norm": 0.2034575194120407,
+      "learning_rate": 5e-06,
+      "loss": 0.1172,
+      "step": 16306
+    },
+    {
+      "epoch": 12.274745954083553,
+      "grad_norm": 0.18936596810817719,
+      "learning_rate": 5e-06,
+      "loss": 0.0652,
+      "step": 16307
+    },
+    {
+      "epoch": 12.275498682724878,
+      "grad_norm": 0.18367531895637512,
+      "learning_rate": 5e-06,
+      "loss": 0.1455,
+      "step": 16308
+    },
+    {
+      "epoch": 12.276251411366202,
+      "grad_norm": 0.1882247030735016,
+      "learning_rate": 5e-06,
+      "loss": 0.1608,
+      "step": 16309
+    },
+    {
+      "epoch": 12.277004140007527,
+      "grad_norm": 0.18682609498500824,
+      "learning_rate": 5e-06,
+      "loss": 0.1434,
+      "step": 16310
+    },
+    {
+      "epoch": 12.277756868648853,
+      "grad_norm": 0.1979360580444336,
+      "learning_rate": 5e-06,
+      "loss": 0.174,
+      "step": 16311
+    },
+    {
+      "epoch": 12.278509597290176,
+      "grad_norm": 0.17412324249744415,
+      "learning_rate": 5e-06,
+      "loss": 0.1528,
+      "step": 16312
+    },
+    {
+      "epoch": 12.279262325931501,
+      "grad_norm": 0.16508854925632477,
+      "learning_rate": 5e-06,
+      "loss": 0.0534,
+      "step": 16313
+    },
+    {
+      "epoch": 12.280015054572827,
+      "grad_norm": 0.16616332530975342,
+      "learning_rate": 5e-06,
+      "loss": 0.1232,
+      "step": 16314
+    },
+    {
+      "epoch": 12.280767783214152,
+      "grad_norm": 0.1765039563179016,
+      "learning_rate": 5e-06,
+      "loss": 0.1392,
+      "step": 16315
+    },
+    {
+      "epoch": 12.281520511855476,
+      "grad_norm": 0.17779353260993958,
+      "learning_rate": 5e-06,
+      "loss": 0.1108,
+      "step": 16316
+    },
+    {
+      "epoch": 12.282273240496801,
+      "grad_norm": 0.14655640721321106,
+      "learning_rate": 5e-06,
+      "loss": 0.0412,
+      "step": 16317
+    },
+    {
+      "epoch": 12.283025969138126,
+      "grad_norm": 0.19376908242702484,
+      "learning_rate": 5e-06,
+      "loss": 0.1419,
+      "step": 16318
+    },
+    {
+      "epoch": 12.28377869777945,
+      "grad_norm": 0.19310884177684784,
+      "learning_rate": 5e-06,
+      "loss": 0.1395,
+      "step": 16319
+    },
+    {
+      "epoch": 12.284531426420775,
+      "grad_norm": 0.17334799468517303,
+      "learning_rate": 5e-06,
+      "loss": 0.1462,
+      "step": 16320
+    },
+    {
+      "epoch": 12.2852841550621,
+      "grad_norm": 0.19028131663799286,
+      "learning_rate": 5e-06,
+      "loss": 0.1535,
+      "step": 16321
+    },
+    {
+      "epoch": 12.286036883703424,
+      "grad_norm": 0.18745626509189606,
+      "learning_rate": 5e-06,
+      "loss": 0.1206,
+      "step": 16322
+    },
+    {
+      "epoch": 12.28678961234475,
+      "grad_norm": 0.1929057389497757,
+      "learning_rate": 5e-06,
+      "loss": 0.1548,
+      "step": 16323
+    },
+    {
+      "epoch": 12.287542340986075,
+      "grad_norm": 0.15301361680030823,
+      "learning_rate": 5e-06,
+      "loss": 0.0844,
+      "step": 16324
+    },
+    {
+      "epoch": 12.288295069627399,
+      "grad_norm": 0.1775735467672348,
+      "learning_rate": 5e-06,
+      "loss": 0.17,
+      "step": 16325
+    },
+    {
+      "epoch": 12.289047798268724,
+      "grad_norm": 0.17024016380310059,
+      "learning_rate": 5e-06,
+      "loss": 0.1867,
+      "step": 16326
+    },
+    {
+      "epoch": 12.28980052691005,
+      "grad_norm": 0.19784900546073914,
+      "learning_rate": 5e-06,
+      "loss": 0.1607,
+      "step": 16327
+    },
+    {
+      "epoch": 12.290553255551373,
+      "grad_norm": 0.1663035750389099,
+      "learning_rate": 5e-06,
+      "loss": 0.1719,
+      "step": 16328
+    },
+    {
+      "epoch": 12.291305984192698,
+      "grad_norm": 0.20605067908763885,
+      "learning_rate": 5e-06,
+      "loss": 0.195,
+      "step": 16329
+    },
+    {
+      "epoch": 12.292058712834024,
+      "grad_norm": 0.18910950422286987,
+      "learning_rate": 5e-06,
+      "loss": 0.132,
+      "step": 16330
+    },
+    {
+      "epoch": 12.292811441475347,
+      "grad_norm": 0.18079042434692383,
+      "learning_rate": 5e-06,
+      "loss": 0.1428,
+      "step": 16331
+    },
+    {
+      "epoch": 12.293564170116673,
+      "grad_norm": 0.19360822439193726,
+      "learning_rate": 5e-06,
+      "loss": 0.166,
+      "step": 16332
+    },
+    {
+      "epoch": 12.294316898757998,
+      "grad_norm": 0.1814558207988739,
+      "learning_rate": 5e-06,
+      "loss": 0.1156,
+      "step": 16333
+    },
+    {
+      "epoch": 12.295069627399322,
+      "grad_norm": 0.17681929469108582,
+      "learning_rate": 5e-06,
+      "loss": 0.1343,
+      "step": 16334
+    },
+    {
+      "epoch": 12.295822356040647,
+      "grad_norm": 0.17688651382923126,
+      "learning_rate": 5e-06,
+      "loss": 0.1521,
+      "step": 16335
+    },
+    {
+      "epoch": 12.296575084681972,
+      "grad_norm": 0.191006138920784,
+      "learning_rate": 5e-06,
+      "loss": 0.1365,
+      "step": 16336
+    },
+    {
+      "epoch": 12.297327813323298,
+      "grad_norm": 0.17894406616687775,
+      "learning_rate": 5e-06,
+      "loss": 0.155,
+      "step": 16337
+    },
+    {
+      "epoch": 12.298080541964621,
+      "grad_norm": 0.16717121005058289,
+      "learning_rate": 5e-06,
+      "loss": 0.0917,
+      "step": 16338
+    },
+    {
+      "epoch": 12.298833270605947,
+      "grad_norm": 0.1806029975414276,
+      "learning_rate": 5e-06,
+      "loss": 0.158,
+      "step": 16339
+    },
+    {
+      "epoch": 12.299585999247272,
+      "grad_norm": 0.19165070354938507,
+      "learning_rate": 5e-06,
+      "loss": 0.1284,
+      "step": 16340
+    },
+    {
+      "epoch": 12.300338727888596,
+      "grad_norm": 0.18953044712543488,
+      "learning_rate": 5e-06,
+      "loss": 0.1195,
+      "step": 16341
+    },
+    {
+      "epoch": 12.301091456529921,
+      "grad_norm": 0.15255144238471985,
+      "learning_rate": 5e-06,
+      "loss": 0.1431,
+      "step": 16342
+    },
+    {
+      "epoch": 12.301844185171246,
+      "grad_norm": 0.2183181494474411,
+      "learning_rate": 5e-06,
+      "loss": 0.1587,
+      "step": 16343
+    },
+    {
+      "epoch": 12.30259691381257,
+      "grad_norm": 0.18191951513290405,
+      "learning_rate": 5e-06,
+      "loss": 0.2315,
+      "step": 16344
+    },
+    {
+      "epoch": 12.303349642453895,
+      "grad_norm": 0.1917026937007904,
+      "learning_rate": 5e-06,
+      "loss": 0.1676,
+      "step": 16345
+    },
+    {
+      "epoch": 12.30410237109522,
+      "grad_norm": 0.1994270384311676,
+      "learning_rate": 5e-06,
+      "loss": 0.1111,
+      "step": 16346
+    },
+    {
+      "epoch": 12.304855099736544,
+      "grad_norm": 0.2263205498456955,
+      "learning_rate": 5e-06,
+      "loss": 0.1586,
+      "step": 16347
+    },
+    {
+      "epoch": 12.30560782837787,
+      "grad_norm": 0.19199687242507935,
+      "learning_rate": 5e-06,
+      "loss": 0.1228,
+      "step": 16348
+    },
+    {
+      "epoch": 12.306360557019195,
+      "grad_norm": 0.15998195111751556,
+      "learning_rate": 5e-06,
+      "loss": 0.1106,
+      "step": 16349
+    },
+    {
+      "epoch": 12.307113285660519,
+      "grad_norm": 0.2166939526796341,
+      "learning_rate": 5e-06,
+      "loss": 0.1539,
+      "step": 16350
+    },
+    {
+      "epoch": 12.307866014301844,
+      "grad_norm": 0.165801003575325,
+      "learning_rate": 5e-06,
+      "loss": 0.1386,
+      "step": 16351
+    },
+    {
+      "epoch": 12.30861874294317,
+      "grad_norm": 0.1852656453847885,
+      "learning_rate": 5e-06,
+      "loss": 0.1572,
+      "step": 16352
+    },
+    {
+      "epoch": 12.309371471584495,
+      "grad_norm": 0.19597208499908447,
+      "learning_rate": 5e-06,
+      "loss": 0.0969,
+      "step": 16353
+    },
+    {
+      "epoch": 12.310124200225818,
+      "grad_norm": 0.20292524993419647,
+      "learning_rate": 5e-06,
+      "loss": 0.158,
+      "step": 16354
+    },
+    {
+      "epoch": 12.310876928867144,
+      "grad_norm": 0.17120352387428284,
+      "learning_rate": 5e-06,
+      "loss": 0.1544,
+      "step": 16355
+    },
+    {
+      "epoch": 12.311629657508469,
+      "grad_norm": 0.17709960043430328,
+      "learning_rate": 5e-06,
+      "loss": 0.1076,
+      "step": 16356
+    },
+    {
+      "epoch": 12.312382386149793,
+      "grad_norm": 0.2061699777841568,
+      "learning_rate": 5e-06,
+      "loss": 0.208,
+      "step": 16357
+    },
+    {
+      "epoch": 12.313135114791118,
+      "grad_norm": 0.15608902275562286,
+      "learning_rate": 5e-06,
+      "loss": 0.1797,
+      "step": 16358
+    },
+    {
+      "epoch": 12.313887843432443,
+      "grad_norm": 0.19166719913482666,
+      "learning_rate": 5e-06,
+      "loss": 0.1369,
+      "step": 16359
+    },
+    {
+      "epoch": 12.314640572073767,
+      "grad_norm": 0.1949223428964615,
+      "learning_rate": 5e-06,
+      "loss": 0.1002,
+      "step": 16360
+    },
+    {
+      "epoch": 12.315393300715092,
+      "grad_norm": 0.19755792617797852,
+      "learning_rate": 5e-06,
+      "loss": 0.1841,
+      "step": 16361
+    },
+    {
+      "epoch": 12.316146029356418,
+      "grad_norm": 0.20485863089561462,
+      "learning_rate": 5e-06,
+      "loss": 0.1632,
+      "step": 16362
+    },
+    {
+      "epoch": 12.316898757997741,
+      "grad_norm": 0.19817258417606354,
+      "learning_rate": 5e-06,
+      "loss": 0.1565,
+      "step": 16363
+    },
+    {
+      "epoch": 12.317651486639067,
+      "grad_norm": 0.19730812311172485,
+      "learning_rate": 5e-06,
+      "loss": 0.1359,
+      "step": 16364
+    },
+    {
+      "epoch": 12.318404215280392,
+      "grad_norm": 0.17654277384281158,
+      "learning_rate": 5e-06,
+      "loss": 0.0898,
+      "step": 16365
+    },
+    {
+      "epoch": 12.319156943921715,
+      "grad_norm": 0.19507934153079987,
+      "learning_rate": 5e-06,
+      "loss": 0.1559,
+      "step": 16366
+    },
+    {
+      "epoch": 12.31990967256304,
+      "grad_norm": 0.21533741056919098,
+      "learning_rate": 5e-06,
+      "loss": 0.1467,
+      "step": 16367
+    },
+    {
+      "epoch": 12.320662401204366,
+      "grad_norm": 0.1881830096244812,
+      "learning_rate": 5e-06,
+      "loss": 0.1524,
+      "step": 16368
+    },
+    {
+      "epoch": 12.32141512984569,
+      "grad_norm": 0.18695271015167236,
+      "learning_rate": 5e-06,
+      "loss": 0.1475,
+      "step": 16369
+    },
+    {
+      "epoch": 12.322167858487015,
+      "grad_norm": 0.18109944462776184,
+      "learning_rate": 5e-06,
+      "loss": 0.138,
+      "step": 16370
+    },
+    {
+      "epoch": 12.32292058712834,
+      "grad_norm": 0.20543424785137177,
+      "learning_rate": 5e-06,
+      "loss": 0.1804,
+      "step": 16371
+    },
+    {
+      "epoch": 12.323673315769666,
+      "grad_norm": 0.18001331388950348,
+      "learning_rate": 5e-06,
+      "loss": 0.1091,
+      "step": 16372
+    },
+    {
+      "epoch": 12.32442604441099,
+      "grad_norm": 0.18388020992279053,
+      "learning_rate": 5e-06,
+      "loss": 0.1345,
+      "step": 16373
+    },
+    {
+      "epoch": 12.325178773052315,
+      "grad_norm": 0.1620633453130722,
+      "learning_rate": 5e-06,
+      "loss": 0.0825,
+      "step": 16374
+    },
+    {
+      "epoch": 12.32593150169364,
+      "grad_norm": 0.19907796382904053,
+      "learning_rate": 5e-06,
+      "loss": 0.1311,
+      "step": 16375
+    },
+    {
+      "epoch": 12.326684230334964,
+      "grad_norm": 0.18831048905849457,
+      "learning_rate": 5e-06,
+      "loss": 0.1324,
+      "step": 16376
+    },
+    {
+      "epoch": 12.327436958976289,
+      "grad_norm": 0.15917664766311646,
+      "learning_rate": 5e-06,
+      "loss": 0.0576,
+      "step": 16377
+    },
+    {
+      "epoch": 12.328189687617614,
+      "grad_norm": 0.18531253933906555,
+      "learning_rate": 5e-06,
+      "loss": 0.1992,
+      "step": 16378
+    },
+    {
+      "epoch": 12.328942416258938,
+      "grad_norm": 0.14673441648483276,
+      "learning_rate": 5e-06,
+      "loss": 0.1179,
+      "step": 16379
+    },
+    {
+      "epoch": 12.329695144900263,
+      "grad_norm": 0.1802399456501007,
+      "learning_rate": 5e-06,
+      "loss": 0.1102,
+      "step": 16380
+    },
+    {
+      "epoch": 12.330447873541589,
+      "grad_norm": 0.17527376115322113,
+      "learning_rate": 5e-06,
+      "loss": 0.165,
+      "step": 16381
+    },
+    {
+      "epoch": 12.331200602182912,
+      "grad_norm": 0.18322093784809113,
+      "learning_rate": 5e-06,
+      "loss": 0.1028,
+      "step": 16382
+    },
+    {
+      "epoch": 12.331953330824238,
+      "grad_norm": 0.1856260448694229,
+      "learning_rate": 5e-06,
+      "loss": 0.1302,
+      "step": 16383
+    },
+    {
+      "epoch": 12.332706059465563,
+      "grad_norm": 0.17769794166088104,
+      "learning_rate": 5e-06,
+      "loss": 0.1813,
+      "step": 16384
+    },
+    {
+      "epoch": 12.333458788106887,
+      "grad_norm": 0.19093315303325653,
+      "learning_rate": 5e-06,
+      "loss": 0.162,
+      "step": 16385
+    },
+    {
+      "epoch": 12.334211516748212,
+      "grad_norm": 0.14671075344085693,
+      "learning_rate": 5e-06,
+      "loss": 0.1425,
+      "step": 16386
+    },
+    {
+      "epoch": 12.334964245389537,
+      "grad_norm": 0.1581646203994751,
+      "learning_rate": 5e-06,
+      "loss": 0.0805,
+      "step": 16387
+    },
+    {
+      "epoch": 12.335716974030863,
+      "grad_norm": 0.21477222442626953,
+      "learning_rate": 5e-06,
+      "loss": 0.1816,
+      "step": 16388
+    },
+    {
+      "epoch": 12.336469702672186,
+      "grad_norm": 0.16503342986106873,
+      "learning_rate": 5e-06,
+      "loss": 0.0597,
+      "step": 16389
+    },
+    {
+      "epoch": 12.337222431313512,
+      "grad_norm": 0.18897731602191925,
+      "learning_rate": 5e-06,
+      "loss": 0.144,
+      "step": 16390
+    },
+    {
+      "epoch": 12.337975159954837,
+      "grad_norm": 0.16565033793449402,
+      "learning_rate": 5e-06,
+      "loss": 0.134,
+      "step": 16391
+    },
+    {
+      "epoch": 12.33872788859616,
+      "grad_norm": 0.15643735229969025,
+      "learning_rate": 5e-06,
+      "loss": 0.1183,
+      "step": 16392
+    },
+    {
+      "epoch": 12.339480617237486,
+      "grad_norm": 0.20042620599269867,
+      "learning_rate": 5e-06,
+      "loss": 0.2111,
+      "step": 16393
+    },
+    {
+      "epoch": 12.340233345878811,
+      "grad_norm": 0.17685523629188538,
+      "learning_rate": 5e-06,
+      "loss": 0.1229,
+      "step": 16394
+    },
+    {
+      "epoch": 12.340986074520135,
+      "grad_norm": 0.18911702930927277,
+      "learning_rate": 5e-06,
+      "loss": 0.1332,
+      "step": 16395
+    },
+    {
+      "epoch": 12.34173880316146,
+      "grad_norm": 0.1650569587945938,
+      "learning_rate": 5e-06,
+      "loss": 0.169,
+      "step": 16396
+    },
+    {
+      "epoch": 12.342491531802786,
+      "grad_norm": 0.16311071813106537,
+      "learning_rate": 5e-06,
+      "loss": 0.0916,
+      "step": 16397
+    },
+    {
+      "epoch": 12.34324426044411,
+      "grad_norm": 0.21958495676517487,
+      "learning_rate": 5e-06,
+      "loss": 0.1562,
+      "step": 16398
+    },
+    {
+      "epoch": 12.343996989085435,
+      "grad_norm": 0.21069474518299103,
+      "learning_rate": 5e-06,
+      "loss": 0.1196,
+      "step": 16399
+    },
+    {
+      "epoch": 12.34474971772676,
+      "grad_norm": 0.19292771816253662,
+      "learning_rate": 5e-06,
+      "loss": 0.1128,
+      "step": 16400
+    },
+    {
+      "epoch": 12.34474971772676,
+      "eval_loss": 0.1828315258026123,
+      "eval_runtime": 452.6735,
+      "eval_samples_per_second": 21.267,
+      "eval_steps_per_second": 0.665,
+      "step": 16400
+    },
+    {
+      "epoch": 12.345502446368084,
+      "grad_norm": 0.18746373057365417,
+      "learning_rate": 5e-06,
+      "loss": 0.2037,
+      "step": 16401
+    },
+    {
+      "epoch": 12.346255175009409,
+      "grad_norm": 0.16054680943489075,
+      "learning_rate": 5e-06,
+      "loss": 0.1304,
+      "step": 16402
+    },
+    {
+      "epoch": 12.347007903650734,
+      "grad_norm": 0.1906135231256485,
+      "learning_rate": 5e-06,
+      "loss": 0.1786,
+      "step": 16403
+    },
+    {
+      "epoch": 12.347760632292058,
+      "grad_norm": 0.19602078199386597,
+      "learning_rate": 5e-06,
+      "loss": 0.1513,
+      "step": 16404
+    },
+    {
+      "epoch": 12.348513360933383,
+      "grad_norm": 0.15783073008060455,
+      "learning_rate": 5e-06,
+      "loss": 0.094,
+      "step": 16405
+    },
+    {
+      "epoch": 12.349266089574709,
+      "grad_norm": 0.17483225464820862,
+      "learning_rate": 5e-06,
+      "loss": 0.112,
+      "step": 16406
+    },
+    {
+      "epoch": 12.350018818216034,
+      "grad_norm": 0.15555787086486816,
+      "learning_rate": 5e-06,
+      "loss": 0.1336,
+      "step": 16407
+    },
+    {
+      "epoch": 12.350771546857358,
+      "grad_norm": 0.20411111414432526,
+      "learning_rate": 5e-06,
+      "loss": 0.2243,
+      "step": 16408
+    },
+    {
+      "epoch": 12.351524275498683,
+      "grad_norm": 0.18997564911842346,
+      "learning_rate": 5e-06,
+      "loss": 0.1384,
+      "step": 16409
+    },
+    {
+      "epoch": 12.352277004140008,
+      "grad_norm": 0.19980065524578094,
+      "learning_rate": 5e-06,
+      "loss": 0.1822,
+      "step": 16410
+    },
+    {
+      "epoch": 12.353029732781332,
+      "grad_norm": 0.19732649624347687,
+      "learning_rate": 5e-06,
+      "loss": 0.1705,
+      "step": 16411
+    },
+    {
+      "epoch": 12.353782461422657,
+      "grad_norm": 0.15924568474292755,
+      "learning_rate": 5e-06,
+      "loss": 0.0889,
+      "step": 16412
+    },
+    {
+      "epoch": 12.354535190063983,
+      "grad_norm": 0.1554485261440277,
+      "learning_rate": 5e-06,
+      "loss": 0.104,
+      "step": 16413
+    },
+    {
+      "epoch": 12.355287918705306,
+      "grad_norm": 0.17834526300430298,
+      "learning_rate": 5e-06,
+      "loss": 0.0872,
+      "step": 16414
+    },
+    {
+      "epoch": 12.356040647346632,
+      "grad_norm": 0.1805206686258316,
+      "learning_rate": 5e-06,
+      "loss": 0.1444,
+      "step": 16415
+    },
+    {
+      "epoch": 12.356793375987957,
+      "grad_norm": 0.17971915006637573,
+      "learning_rate": 5e-06,
+      "loss": 0.1235,
+      "step": 16416
+    },
+    {
+      "epoch": 12.35754610462928,
+      "grad_norm": 0.18291467428207397,
+      "learning_rate": 5e-06,
+      "loss": 0.1026,
+      "step": 16417
+    },
+    {
+      "epoch": 12.358298833270606,
+      "grad_norm": 0.17308515310287476,
+      "learning_rate": 5e-06,
+      "loss": 0.08,
+      "step": 16418
+    },
+    {
+      "epoch": 12.359051561911931,
+      "grad_norm": 0.20954082906246185,
+      "learning_rate": 5e-06,
+      "loss": 0.1305,
+      "step": 16419
+    },
+    {
+      "epoch": 12.359804290553255,
+      "grad_norm": 0.15352171659469604,
+      "learning_rate": 5e-06,
+      "loss": 0.1199,
+      "step": 16420
+    },
+    {
+      "epoch": 12.36055701919458,
+      "grad_norm": 0.20545895397663116,
+      "learning_rate": 5e-06,
+      "loss": 0.182,
+      "step": 16421
+    },
+    {
+      "epoch": 12.361309747835906,
+      "grad_norm": 0.15674103796482086,
+      "learning_rate": 5e-06,
+      "loss": 0.1662,
+      "step": 16422
+    },
+    {
+      "epoch": 12.362062476477229,
+      "grad_norm": 0.19403839111328125,
+      "learning_rate": 5e-06,
+      "loss": 0.1062,
+      "step": 16423
+    },
+    {
+      "epoch": 12.362815205118554,
+      "grad_norm": 0.19188626110553741,
+      "learning_rate": 5e-06,
+      "loss": 0.1071,
+      "step": 16424
+    },
+    {
+      "epoch": 12.36356793375988,
+      "grad_norm": 0.19028037786483765,
+      "learning_rate": 5e-06,
+      "loss": 0.19,
+      "step": 16425
+    },
+    {
+      "epoch": 12.364320662401205,
+      "grad_norm": 0.16812138259410858,
+      "learning_rate": 5e-06,
+      "loss": 0.0931,
+      "step": 16426
+    },
+    {
+      "epoch": 12.365073391042529,
+      "grad_norm": 0.1885511875152588,
+      "learning_rate": 5e-06,
+      "loss": 0.1044,
+      "step": 16427
+    },
+    {
+      "epoch": 12.365826119683854,
+      "grad_norm": 0.177470862865448,
+      "learning_rate": 5e-06,
+      "loss": 0.1248,
+      "step": 16428
+    },
+    {
+      "epoch": 12.36657884832518,
+      "grad_norm": 0.1666792333126068,
+      "learning_rate": 5e-06,
+      "loss": 0.0963,
+      "step": 16429
+    },
+    {
+      "epoch": 12.367331576966503,
+      "grad_norm": 0.1686277836561203,
+      "learning_rate": 5e-06,
+      "loss": 0.1793,
+      "step": 16430
+    },
+    {
+      "epoch": 12.368084305607828,
+      "grad_norm": 0.17627058923244476,
+      "learning_rate": 5e-06,
+      "loss": 0.1285,
+      "step": 16431
+    },
+    {
+      "epoch": 12.368837034249154,
+      "grad_norm": 0.22985665500164032,
+      "learning_rate": 5e-06,
+      "loss": 0.1645,
+      "step": 16432
+    },
+    {
+      "epoch": 12.369589762890477,
+      "grad_norm": 0.18862120807170868,
+      "learning_rate": 5e-06,
+      "loss": 0.141,
+      "step": 16433
+    },
+    {
+      "epoch": 12.370342491531803,
+      "grad_norm": 0.15603171288967133,
+      "learning_rate": 5e-06,
+      "loss": 0.069,
+      "step": 16434
+    },
+    {
+      "epoch": 12.371095220173128,
+      "grad_norm": 0.18599912524223328,
+      "learning_rate": 5e-06,
+      "loss": 0.1461,
+      "step": 16435
+    },
+    {
+      "epoch": 12.371847948814452,
+      "grad_norm": 0.19589346647262573,
+      "learning_rate": 5e-06,
+      "loss": 0.1222,
+      "step": 16436
+    },
+    {
+      "epoch": 12.372600677455777,
+      "grad_norm": 0.16017144918441772,
+      "learning_rate": 5e-06,
+      "loss": 0.1223,
+      "step": 16437
+    },
+    {
+      "epoch": 12.373353406097102,
+      "grad_norm": 0.17290884256362915,
+      "learning_rate": 5e-06,
+      "loss": 0.0851,
+      "step": 16438
+    },
+    {
+      "epoch": 12.374106134738426,
+      "grad_norm": 0.20001864433288574,
+      "learning_rate": 5e-06,
+      "loss": 0.1527,
+      "step": 16439
+    },
+    {
+      "epoch": 12.374858863379751,
+      "grad_norm": 0.20896016061306,
+      "learning_rate": 5e-06,
+      "loss": 0.1976,
+      "step": 16440
+    },
+    {
+      "epoch": 12.375611592021077,
+      "grad_norm": 0.18491367995738983,
+      "learning_rate": 5e-06,
+      "loss": 0.1782,
+      "step": 16441
+    },
+    {
+      "epoch": 12.3763643206624,
+      "grad_norm": 0.20589250326156616,
+      "learning_rate": 5e-06,
+      "loss": 0.1485,
+      "step": 16442
+    },
+    {
+      "epoch": 12.377117049303726,
+      "grad_norm": 0.19331789016723633,
+      "learning_rate": 5e-06,
+      "loss": 0.1911,
+      "step": 16443
+    },
+    {
+      "epoch": 12.377869777945051,
+      "grad_norm": 0.1891690194606781,
+      "learning_rate": 5e-06,
+      "loss": 0.154,
+      "step": 16444
+    },
+    {
+      "epoch": 12.378622506586376,
+      "grad_norm": 0.16675126552581787,
+      "learning_rate": 5e-06,
+      "loss": 0.0866,
+      "step": 16445
+    },
+    {
+      "epoch": 12.3793752352277,
+      "grad_norm": 0.1952163130044937,
+      "learning_rate": 5e-06,
+      "loss": 0.124,
+      "step": 16446
+    },
+    {
+      "epoch": 12.380127963869025,
+      "grad_norm": 0.16965952515602112,
+      "learning_rate": 5e-06,
+      "loss": 0.1171,
+      "step": 16447
+    },
+    {
+      "epoch": 12.38088069251035,
+      "grad_norm": 0.19350378215312958,
+      "learning_rate": 5e-06,
+      "loss": 0.1081,
+      "step": 16448
+    },
+    {
+      "epoch": 12.381633421151674,
+      "grad_norm": 0.2099548876285553,
+      "learning_rate": 5e-06,
+      "loss": 0.2434,
+      "step": 16449
+    },
+    {
+      "epoch": 12.382386149793,
+      "grad_norm": 0.16325588524341583,
+      "learning_rate": 5e-06,
+      "loss": 0.1509,
+      "step": 16450
+    },
+    {
+      "epoch": 12.383138878434325,
+      "grad_norm": 0.1926242709159851,
+      "learning_rate": 5e-06,
+      "loss": 0.1664,
+      "step": 16451
+    },
+    {
+      "epoch": 12.383891607075649,
+      "grad_norm": 0.15656450390815735,
+      "learning_rate": 5e-06,
+      "loss": 0.1724,
+      "step": 16452
+    },
+    {
+      "epoch": 12.384644335716974,
+      "grad_norm": 0.18480846285820007,
+      "learning_rate": 5e-06,
+      "loss": 0.0804,
+      "step": 16453
+    },
+    {
+      "epoch": 12.3853970643583,
+      "grad_norm": 0.17932158708572388,
+      "learning_rate": 5e-06,
+      "loss": 0.0775,
+      "step": 16454
+    },
+    {
+      "epoch": 12.386149792999623,
+      "grad_norm": 0.21090492606163025,
+      "learning_rate": 5e-06,
+      "loss": 0.1854,
+      "step": 16455
+    },
+    {
+      "epoch": 12.386902521640948,
+      "grad_norm": 0.2126559317111969,
+      "learning_rate": 5e-06,
+      "loss": 0.1179,
+      "step": 16456
+    },
+    {
+      "epoch": 12.387655250282274,
+      "grad_norm": 0.1947508156299591,
+      "learning_rate": 5e-06,
+      "loss": 0.1346,
+      "step": 16457
+    },
+    {
+      "epoch": 12.388407978923597,
+      "grad_norm": 0.19020992517471313,
+      "learning_rate": 5e-06,
+      "loss": 0.1627,
+      "step": 16458
+    },
+    {
+      "epoch": 12.389160707564923,
+      "grad_norm": 0.205296590924263,
+      "learning_rate": 5e-06,
+      "loss": 0.1671,
+      "step": 16459
+    },
+    {
+      "epoch": 12.389913436206248,
+      "grad_norm": 0.17308054864406586,
+      "learning_rate": 5e-06,
+      "loss": 0.0802,
+      "step": 16460
+    },
+    {
+      "epoch": 12.390666164847573,
+      "grad_norm": 0.1823367476463318,
+      "learning_rate": 5e-06,
+      "loss": 0.1687,
+      "step": 16461
+    },
+    {
+      "epoch": 12.391418893488897,
+      "grad_norm": 0.18554289638996124,
+      "learning_rate": 5e-06,
+      "loss": 0.1216,
+      "step": 16462
+    },
+    {
+      "epoch": 12.392171622130222,
+      "grad_norm": 0.19485986232757568,
+      "learning_rate": 5e-06,
+      "loss": 0.2077,
+      "step": 16463
+    },
+    {
+      "epoch": 12.392924350771548,
+      "grad_norm": 0.1690843105316162,
+      "learning_rate": 5e-06,
+      "loss": 0.1229,
+      "step": 16464
+    },
+    {
+      "epoch": 12.393677079412871,
+      "grad_norm": 0.22166655957698822,
+      "learning_rate": 5e-06,
+      "loss": 0.1595,
+      "step": 16465
+    },
+    {
+      "epoch": 12.394429808054197,
+      "grad_norm": 0.19262465834617615,
+      "learning_rate": 5e-06,
+      "loss": 0.1345,
+      "step": 16466
+    },
+    {
+      "epoch": 12.395182536695522,
+      "grad_norm": 0.16600672900676727,
+      "learning_rate": 5e-06,
+      "loss": 0.1193,
+      "step": 16467
+    },
+    {
+      "epoch": 12.395935265336846,
+      "grad_norm": 0.17783516645431519,
+      "learning_rate": 5e-06,
+      "loss": 0.1227,
+      "step": 16468
+    },
+    {
+      "epoch": 12.396687993978171,
+      "grad_norm": 0.21201340854167938,
+      "learning_rate": 5e-06,
+      "loss": 0.1951,
+      "step": 16469
+    },
+    {
+      "epoch": 12.397440722619496,
+      "grad_norm": 0.16407395899295807,
+      "learning_rate": 5e-06,
+      "loss": 0.0819,
+      "step": 16470
+    },
+    {
+      "epoch": 12.39819345126082,
+      "grad_norm": 0.18667346239089966,
+      "learning_rate": 5e-06,
+      "loss": 0.176,
+      "step": 16471
+    },
+    {
+      "epoch": 12.398946179902145,
+      "grad_norm": 0.16106466948986053,
+      "learning_rate": 5e-06,
+      "loss": 0.1622,
+      "step": 16472
+    },
+    {
+      "epoch": 12.39969890854347,
+      "grad_norm": 0.17063061892986298,
+      "learning_rate": 5e-06,
+      "loss": 0.1511,
+      "step": 16473
+    },
+    {
+      "epoch": 12.400451637184794,
+      "grad_norm": 0.21400243043899536,
+      "learning_rate": 5e-06,
+      "loss": 0.1212,
+      "step": 16474
+    },
+    {
+      "epoch": 12.40120436582612,
+      "grad_norm": 0.17671099305152893,
+      "learning_rate": 5e-06,
+      "loss": 0.1654,
+      "step": 16475
+    },
+    {
+      "epoch": 12.401957094467445,
+      "grad_norm": 0.2095586210489273,
+      "learning_rate": 5e-06,
+      "loss": 0.1575,
+      "step": 16476
+    },
+    {
+      "epoch": 12.402709823108768,
+      "grad_norm": 0.19170859456062317,
+      "learning_rate": 5e-06,
+      "loss": 0.1601,
+      "step": 16477
+    },
+    {
+      "epoch": 12.403462551750094,
+      "grad_norm": 0.1679413765668869,
+      "learning_rate": 5e-06,
+      "loss": 0.1293,
+      "step": 16478
+    },
+    {
+      "epoch": 12.40421528039142,
+      "grad_norm": 0.18701474368572235,
+      "learning_rate": 5e-06,
+      "loss": 0.1559,
+      "step": 16479
+    },
+    {
+      "epoch": 12.404968009032745,
+      "grad_norm": 0.1921035498380661,
+      "learning_rate": 5e-06,
+      "loss": 0.1149,
+      "step": 16480
+    },
+    {
+      "epoch": 12.405720737674068,
+      "grad_norm": 0.18923313915729523,
+      "learning_rate": 5e-06,
+      "loss": 0.146,
+      "step": 16481
+    },
+    {
+      "epoch": 12.406473466315393,
+      "grad_norm": 0.2055039256811142,
+      "learning_rate": 5e-06,
+      "loss": 0.1613,
+      "step": 16482
+    },
+    {
+      "epoch": 12.407226194956719,
+      "grad_norm": 0.17611029744148254,
+      "learning_rate": 5e-06,
+      "loss": 0.0952,
+      "step": 16483
+    },
+    {
+      "epoch": 12.407978923598042,
+      "grad_norm": 0.1679159253835678,
+      "learning_rate": 5e-06,
+      "loss": 0.14,
+      "step": 16484
+    },
+    {
+      "epoch": 12.408731652239368,
+      "grad_norm": 0.1489751935005188,
+      "learning_rate": 5e-06,
+      "loss": 0.1071,
+      "step": 16485
+    },
+    {
+      "epoch": 12.409484380880693,
+      "grad_norm": 0.16391603648662567,
+      "learning_rate": 5e-06,
+      "loss": 0.0645,
+      "step": 16486
+    },
+    {
+      "epoch": 12.410237109522017,
+      "grad_norm": 0.1709279716014862,
+      "learning_rate": 5e-06,
+      "loss": 0.1632,
+      "step": 16487
+    },
+    {
+      "epoch": 12.410989838163342,
+      "grad_norm": 0.16751523315906525,
+      "learning_rate": 5e-06,
+      "loss": 0.1441,
+      "step": 16488
+    },
+    {
+      "epoch": 12.411742566804667,
+      "grad_norm": 0.1709282547235489,
+      "learning_rate": 5e-06,
+      "loss": 0.0772,
+      "step": 16489
+    },
+    {
+      "epoch": 12.412495295445991,
+      "grad_norm": 0.18285027146339417,
+      "learning_rate": 5e-06,
+      "loss": 0.1129,
+      "step": 16490
+    },
+    {
+      "epoch": 12.413248024087316,
+      "grad_norm": 0.16641297936439514,
+      "learning_rate": 5e-06,
+      "loss": 0.0928,
+      "step": 16491
+    },
+    {
+      "epoch": 12.414000752728642,
+      "grad_norm": 0.16867852210998535,
+      "learning_rate": 5e-06,
+      "loss": 0.061,
+      "step": 16492
+    },
+    {
+      "epoch": 12.414753481369965,
+      "grad_norm": 0.20078156888484955,
+      "learning_rate": 5e-06,
+      "loss": 0.1112,
+      "step": 16493
+    },
+    {
+      "epoch": 12.41550621001129,
+      "grad_norm": 0.16208767890930176,
+      "learning_rate": 5e-06,
+      "loss": 0.1023,
+      "step": 16494
+    },
+    {
+      "epoch": 12.416258938652616,
+      "grad_norm": 0.17744843661785126,
+      "learning_rate": 5e-06,
+      "loss": 0.1138,
+      "step": 16495
+    },
+    {
+      "epoch": 12.41701166729394,
+      "grad_norm": 0.20024070143699646,
+      "learning_rate": 5e-06,
+      "loss": 0.1519,
+      "step": 16496
+    },
+    {
+      "epoch": 12.417764395935265,
+      "grad_norm": 0.21755200624465942,
+      "learning_rate": 5e-06,
+      "loss": 0.1426,
+      "step": 16497
+    },
+    {
+      "epoch": 12.41851712457659,
+      "grad_norm": 0.17720159888267517,
+      "learning_rate": 5e-06,
+      "loss": 0.1438,
+      "step": 16498
+    },
+    {
+      "epoch": 12.419269853217916,
+      "grad_norm": 0.19144147634506226,
+      "learning_rate": 5e-06,
+      "loss": 0.1862,
+      "step": 16499
+    },
+    {
+      "epoch": 12.42002258185924,
+      "grad_norm": 0.20211797952651978,
+      "learning_rate": 5e-06,
+      "loss": 0.1701,
+      "step": 16500
+    },
+    {
+      "epoch": 12.420775310500565,
+      "grad_norm": 0.1795448511838913,
+      "learning_rate": 5e-06,
+      "loss": 0.1212,
+      "step": 16501
+    },
+    {
+      "epoch": 12.42152803914189,
+      "grad_norm": 0.1782751828432083,
+      "learning_rate": 5e-06,
+      "loss": 0.0784,
+      "step": 16502
+    },
+    {
+      "epoch": 12.422280767783214,
+      "grad_norm": 0.17541635036468506,
+      "learning_rate": 5e-06,
+      "loss": 0.1425,
+      "step": 16503
+    },
+    {
+      "epoch": 12.423033496424539,
+      "grad_norm": 0.18949659168720245,
+      "learning_rate": 5e-06,
+      "loss": 0.1815,
+      "step": 16504
+    },
+    {
+      "epoch": 12.423786225065864,
+      "grad_norm": 0.17389480769634247,
+      "learning_rate": 5e-06,
+      "loss": 0.1253,
+      "step": 16505
+    },
+    {
+      "epoch": 12.424538953707188,
+      "grad_norm": 0.20093709230422974,
+      "learning_rate": 5e-06,
+      "loss": 0.1562,
+      "step": 16506
+    },
+    {
+      "epoch": 12.425291682348513,
+      "grad_norm": 0.20569293200969696,
+      "learning_rate": 5e-06,
+      "loss": 0.1689,
+      "step": 16507
+    },
+    {
+      "epoch": 12.426044410989839,
+      "grad_norm": 0.17160524427890778,
+      "learning_rate": 5e-06,
+      "loss": 0.1087,
+      "step": 16508
+    },
+    {
+      "epoch": 12.426797139631162,
+      "grad_norm": 0.22213931381702423,
+      "learning_rate": 5e-06,
+      "loss": 0.2153,
+      "step": 16509
+    },
+    {
+      "epoch": 12.427549868272488,
+      "grad_norm": 0.2200566977262497,
+      "learning_rate": 5e-06,
+      "loss": 0.1718,
+      "step": 16510
+    },
+    {
+      "epoch": 12.428302596913813,
+      "grad_norm": 0.16709363460540771,
+      "learning_rate": 5e-06,
+      "loss": 0.0664,
+      "step": 16511
+    },
+    {
+      "epoch": 12.429055325555137,
+      "grad_norm": 0.2065098136663437,
+      "learning_rate": 5e-06,
+      "loss": 0.1314,
+      "step": 16512
+    },
+    {
+      "epoch": 12.429808054196462,
+      "grad_norm": 0.1823834478855133,
+      "learning_rate": 5e-06,
+      "loss": 0.1315,
+      "step": 16513
+    },
+    {
+      "epoch": 12.430560782837787,
+      "grad_norm": 0.16944894194602966,
+      "learning_rate": 5e-06,
+      "loss": 0.1029,
+      "step": 16514
+    },
+    {
+      "epoch": 12.431313511479111,
+      "grad_norm": 0.17131972312927246,
+      "learning_rate": 5e-06,
+      "loss": 0.0962,
+      "step": 16515
+    },
+    {
+      "epoch": 12.432066240120436,
+      "grad_norm": 0.1902579814195633,
+      "learning_rate": 5e-06,
+      "loss": 0.1373,
+      "step": 16516
+    },
+    {
+      "epoch": 12.432818968761762,
+      "grad_norm": 0.2117292582988739,
+      "learning_rate": 5e-06,
+      "loss": 0.1842,
+      "step": 16517
+    },
+    {
+      "epoch": 12.433571697403087,
+      "grad_norm": 0.18800170719623566,
+      "learning_rate": 5e-06,
+      "loss": 0.1527,
+      "step": 16518
+    },
+    {
+      "epoch": 12.43432442604441,
+      "grad_norm": 0.21264468133449554,
+      "learning_rate": 5e-06,
+      "loss": 0.1578,
+      "step": 16519
+    },
+    {
+      "epoch": 12.435077154685736,
+      "grad_norm": 0.1830260306596756,
+      "learning_rate": 5e-06,
+      "loss": 0.1052,
+      "step": 16520
+    },
+    {
+      "epoch": 12.435829883327061,
+      "grad_norm": 0.1711675524711609,
+      "learning_rate": 5e-06,
+      "loss": 0.0649,
+      "step": 16521
+    },
+    {
+      "epoch": 12.436582611968385,
+      "grad_norm": 0.16086405515670776,
+      "learning_rate": 5e-06,
+      "loss": 0.0766,
+      "step": 16522
+    },
+    {
+      "epoch": 12.43733534060971,
+      "grad_norm": 0.18603526055812836,
+      "learning_rate": 5e-06,
+      "loss": 0.1952,
+      "step": 16523
+    },
+    {
+      "epoch": 12.438088069251036,
+      "grad_norm": 0.17737479507923126,
+      "learning_rate": 5e-06,
+      "loss": 0.1615,
+      "step": 16524
+    },
+    {
+      "epoch": 12.43884079789236,
+      "grad_norm": 0.2014082670211792,
+      "learning_rate": 5e-06,
+      "loss": 0.1791,
+      "step": 16525
+    },
+    {
+      "epoch": 12.439593526533685,
+      "grad_norm": 0.21587951481342316,
+      "learning_rate": 5e-06,
+      "loss": 0.1686,
+      "step": 16526
+    },
+    {
+      "epoch": 12.44034625517501,
+      "grad_norm": 0.18347042798995972,
+      "learning_rate": 5e-06,
+      "loss": 0.147,
+      "step": 16527
+    },
+    {
+      "epoch": 12.441098983816334,
+      "grad_norm": 0.21411964297294617,
+      "learning_rate": 5e-06,
+      "loss": 0.1412,
+      "step": 16528
+    },
+    {
+      "epoch": 12.441851712457659,
+      "grad_norm": 0.18644966185092926,
+      "learning_rate": 5e-06,
+      "loss": 0.1414,
+      "step": 16529
+    },
+    {
+      "epoch": 12.442604441098984,
+      "grad_norm": 0.18910911679267883,
+      "learning_rate": 5e-06,
+      "loss": 0.1263,
+      "step": 16530
+    },
+    {
+      "epoch": 12.443357169740308,
+      "grad_norm": 0.18671244382858276,
+      "learning_rate": 5e-06,
+      "loss": 0.1462,
+      "step": 16531
+    },
+    {
+      "epoch": 12.444109898381633,
+      "grad_norm": 0.2015087604522705,
+      "learning_rate": 5e-06,
+      "loss": 0.1459,
+      "step": 16532
+    },
+    {
+      "epoch": 12.444862627022959,
+      "grad_norm": 0.16716895997524261,
+      "learning_rate": 5e-06,
+      "loss": 0.1145,
+      "step": 16533
+    },
+    {
+      "epoch": 12.445615355664284,
+      "grad_norm": 0.18355511128902435,
+      "learning_rate": 5e-06,
+      "loss": 0.1286,
+      "step": 16534
+    },
+    {
+      "epoch": 12.446368084305607,
+      "grad_norm": 0.18215180933475494,
+      "learning_rate": 5e-06,
+      "loss": 0.1003,
+      "step": 16535
+    },
+    {
+      "epoch": 12.447120812946933,
+      "grad_norm": 0.1819421499967575,
+      "learning_rate": 5e-06,
+      "loss": 0.1959,
+      "step": 16536
+    },
+    {
+      "epoch": 12.447873541588258,
+      "grad_norm": 0.18577468395233154,
+      "learning_rate": 5e-06,
+      "loss": 0.163,
+      "step": 16537
+    },
+    {
+      "epoch": 12.448626270229582,
+      "grad_norm": 0.18156149983406067,
+      "learning_rate": 5e-06,
+      "loss": 0.0866,
+      "step": 16538
+    },
+    {
+      "epoch": 12.449378998870907,
+      "grad_norm": 0.16745905578136444,
+      "learning_rate": 5e-06,
+      "loss": 0.1293,
+      "step": 16539
+    },
+    {
+      "epoch": 12.450131727512233,
+      "grad_norm": 0.1871601939201355,
+      "learning_rate": 5e-06,
+      "loss": 0.2077,
+      "step": 16540
+    },
+    {
+      "epoch": 12.450884456153556,
+      "grad_norm": 0.1895197480916977,
+      "learning_rate": 5e-06,
+      "loss": 0.1782,
+      "step": 16541
+    },
+    {
+      "epoch": 12.451637184794881,
+      "grad_norm": 0.13989010453224182,
+      "learning_rate": 5e-06,
+      "loss": 0.0705,
+      "step": 16542
+    },
+    {
+      "epoch": 12.452389913436207,
+      "grad_norm": 0.17687197029590607,
+      "learning_rate": 5e-06,
+      "loss": 0.1391,
+      "step": 16543
+    },
+    {
+      "epoch": 12.45314264207753,
+      "grad_norm": 0.17326565086841583,
+      "learning_rate": 5e-06,
+      "loss": 0.1704,
+      "step": 16544
+    },
+    {
+      "epoch": 12.453895370718856,
+      "grad_norm": 0.18479380011558533,
+      "learning_rate": 5e-06,
+      "loss": 0.1814,
+      "step": 16545
+    },
+    {
+      "epoch": 12.454648099360181,
+      "grad_norm": 0.1958581954240799,
+      "learning_rate": 5e-06,
+      "loss": 0.1025,
+      "step": 16546
+    },
+    {
+      "epoch": 12.455400828001505,
+      "grad_norm": 0.16894973814487457,
+      "learning_rate": 5e-06,
+      "loss": 0.1273,
+      "step": 16547
+    },
+    {
+      "epoch": 12.45615355664283,
+      "grad_norm": 0.2136932909488678,
+      "learning_rate": 5e-06,
+      "loss": 0.1491,
+      "step": 16548
+    },
+    {
+      "epoch": 12.456906285284155,
+      "grad_norm": 0.1643250435590744,
+      "learning_rate": 5e-06,
+      "loss": 0.1722,
+      "step": 16549
+    },
+    {
+      "epoch": 12.457659013925479,
+      "grad_norm": 0.20017732679843903,
+      "learning_rate": 5e-06,
+      "loss": 0.2121,
+      "step": 16550
+    },
+    {
+      "epoch": 12.458411742566804,
+      "grad_norm": 0.17801311612129211,
+      "learning_rate": 5e-06,
+      "loss": 0.2353,
+      "step": 16551
+    },
+    {
+      "epoch": 12.45916447120813,
+      "grad_norm": 0.20430725812911987,
+      "learning_rate": 5e-06,
+      "loss": 0.1169,
+      "step": 16552
+    },
+    {
+      "epoch": 12.459917199849455,
+      "grad_norm": 0.19065473973751068,
+      "learning_rate": 5e-06,
+      "loss": 0.2385,
+      "step": 16553
+    },
+    {
+      "epoch": 12.460669928490779,
+      "grad_norm": 0.17720924317836761,
+      "learning_rate": 5e-06,
+      "loss": 0.0808,
+      "step": 16554
+    },
+    {
+      "epoch": 12.461422657132104,
+      "grad_norm": 0.2039174884557724,
+      "learning_rate": 5e-06,
+      "loss": 0.1679,
+      "step": 16555
+    },
+    {
+      "epoch": 12.46217538577343,
+      "grad_norm": 0.19382302463054657,
+      "learning_rate": 5e-06,
+      "loss": 0.2088,
+      "step": 16556
+    },
+    {
+      "epoch": 12.462928114414753,
+      "grad_norm": 0.18189771473407745,
+      "learning_rate": 5e-06,
+      "loss": 0.1749,
+      "step": 16557
+    },
+    {
+      "epoch": 12.463680843056078,
+      "grad_norm": 0.19151462614536285,
+      "learning_rate": 5e-06,
+      "loss": 0.2463,
+      "step": 16558
+    },
+    {
+      "epoch": 12.464433571697404,
+      "grad_norm": 0.17013759911060333,
+      "learning_rate": 5e-06,
+      "loss": 0.1707,
+      "step": 16559
+    },
+    {
+      "epoch": 12.465186300338727,
+      "grad_norm": 0.18152375519275665,
+      "learning_rate": 5e-06,
+      "loss": 0.1445,
+      "step": 16560
+    },
+    {
+      "epoch": 12.465939028980053,
+      "grad_norm": 0.19773972034454346,
+      "learning_rate": 5e-06,
+      "loss": 0.1683,
+      "step": 16561
+    },
+    {
+      "epoch": 12.466691757621378,
+      "grad_norm": 0.18033060431480408,
+      "learning_rate": 5e-06,
+      "loss": 0.1356,
+      "step": 16562
+    },
+    {
+      "epoch": 12.467444486262702,
+      "grad_norm": 0.19281268119812012,
+      "learning_rate": 5e-06,
+      "loss": 0.1411,
+      "step": 16563
+    },
+    {
+      "epoch": 12.468197214904027,
+      "grad_norm": 0.18899838626384735,
+      "learning_rate": 5e-06,
+      "loss": 0.1149,
+      "step": 16564
+    },
+    {
+      "epoch": 12.468949943545352,
+      "grad_norm": 0.24755772948265076,
+      "learning_rate": 5e-06,
+      "loss": 0.1771,
+      "step": 16565
+    },
+    {
+      "epoch": 12.469702672186676,
+      "grad_norm": 0.1505463868379593,
+      "learning_rate": 5e-06,
+      "loss": 0.0705,
+      "step": 16566
+    },
+    {
+      "epoch": 12.470455400828001,
+      "grad_norm": 0.19650940597057343,
+      "learning_rate": 5e-06,
+      "loss": 0.1171,
+      "step": 16567
+    },
+    {
+      "epoch": 12.471208129469327,
+      "grad_norm": 0.17221875488758087,
+      "learning_rate": 5e-06,
+      "loss": 0.1501,
+      "step": 16568
+    },
+    {
+      "epoch": 12.471960858110652,
+      "grad_norm": 0.19599780440330505,
+      "learning_rate": 5e-06,
+      "loss": 0.1405,
+      "step": 16569
+    },
+    {
+      "epoch": 12.472713586751976,
+      "grad_norm": 0.17176491022109985,
+      "learning_rate": 5e-06,
+      "loss": 0.0727,
+      "step": 16570
+    },
+    {
+      "epoch": 12.473466315393301,
+      "grad_norm": 0.19671185314655304,
+      "learning_rate": 5e-06,
+      "loss": 0.139,
+      "step": 16571
+    },
+    {
+      "epoch": 12.474219044034626,
+      "grad_norm": 0.16313610970973969,
+      "learning_rate": 5e-06,
+      "loss": 0.1588,
+      "step": 16572
+    },
+    {
+      "epoch": 12.47497177267595,
+      "grad_norm": 0.1778470128774643,
+      "learning_rate": 5e-06,
+      "loss": 0.1217,
+      "step": 16573
+    },
+    {
+      "epoch": 12.475724501317275,
+      "grad_norm": 0.1929701864719391,
+      "learning_rate": 5e-06,
+      "loss": 0.1122,
+      "step": 16574
+    },
+    {
+      "epoch": 12.4764772299586,
+      "grad_norm": 0.2007606029510498,
+      "learning_rate": 5e-06,
+      "loss": 0.1551,
+      "step": 16575
+    },
+    {
+      "epoch": 12.477229958599924,
+      "grad_norm": 0.17681623995304108,
+      "learning_rate": 5e-06,
+      "loss": 0.106,
+      "step": 16576
+    },
+    {
+      "epoch": 12.47798268724125,
+      "grad_norm": 0.19030672311782837,
+      "learning_rate": 5e-06,
+      "loss": 0.1203,
+      "step": 16577
+    },
+    {
+      "epoch": 12.478735415882575,
+      "grad_norm": 0.1815904676914215,
+      "learning_rate": 5e-06,
+      "loss": 0.0979,
+      "step": 16578
+    },
+    {
+      "epoch": 12.479488144523899,
+      "grad_norm": 0.19462968409061432,
+      "learning_rate": 5e-06,
+      "loss": 0.163,
+      "step": 16579
+    },
+    {
+      "epoch": 12.480240873165224,
+      "grad_norm": 0.1759449988603592,
+      "learning_rate": 5e-06,
+      "loss": 0.1492,
+      "step": 16580
+    },
+    {
+      "epoch": 12.48099360180655,
+      "grad_norm": 0.17896424233913422,
+      "learning_rate": 5e-06,
+      "loss": 0.1609,
+      "step": 16581
+    },
+    {
+      "epoch": 12.481746330447873,
+      "grad_norm": 0.14895297586917877,
+      "learning_rate": 5e-06,
+      "loss": 0.101,
+      "step": 16582
+    },
+    {
+      "epoch": 12.482499059089198,
+      "grad_norm": 0.19768796861171722,
+      "learning_rate": 5e-06,
+      "loss": 0.1396,
+      "step": 16583
+    },
+    {
+      "epoch": 12.483251787730524,
+      "grad_norm": 0.1815672218799591,
+      "learning_rate": 5e-06,
+      "loss": 0.1811,
+      "step": 16584
+    },
+    {
+      "epoch": 12.484004516371847,
+      "grad_norm": 0.1520002782344818,
+      "learning_rate": 5e-06,
+      "loss": 0.1084,
+      "step": 16585
+    },
+    {
+      "epoch": 12.484757245013173,
+      "grad_norm": 0.18618255853652954,
+      "learning_rate": 5e-06,
+      "loss": 0.1327,
+      "step": 16586
+    },
+    {
+      "epoch": 12.485509973654498,
+      "grad_norm": 0.1776125729084015,
+      "learning_rate": 5e-06,
+      "loss": 0.0746,
+      "step": 16587
+    },
+    {
+      "epoch": 12.486262702295821,
+      "grad_norm": 0.17942270636558533,
+      "learning_rate": 5e-06,
+      "loss": 0.0833,
+      "step": 16588
+    },
+    {
+      "epoch": 12.487015430937147,
+      "grad_norm": 0.1927299052476883,
+      "learning_rate": 5e-06,
+      "loss": 0.1083,
+      "step": 16589
+    },
+    {
+      "epoch": 12.487768159578472,
+      "grad_norm": 0.1904297024011612,
+      "learning_rate": 5e-06,
+      "loss": 0.133,
+      "step": 16590
+    },
+    {
+      "epoch": 12.488520888219798,
+      "grad_norm": 0.17745979130268097,
+      "learning_rate": 5e-06,
+      "loss": 0.0949,
+      "step": 16591
+    },
+    {
+      "epoch": 12.489273616861121,
+      "grad_norm": 0.21787822246551514,
+      "learning_rate": 5e-06,
+      "loss": 0.1665,
+      "step": 16592
+    },
+    {
+      "epoch": 12.490026345502447,
+      "grad_norm": 0.16623730957508087,
+      "learning_rate": 5e-06,
+      "loss": 0.1405,
+      "step": 16593
+    },
+    {
+      "epoch": 12.490779074143772,
+      "grad_norm": 0.19282789528369904,
+      "learning_rate": 5e-06,
+      "loss": 0.183,
+      "step": 16594
+    },
+    {
+      "epoch": 12.491531802785095,
+      "grad_norm": 0.17969831824302673,
+      "learning_rate": 5e-06,
+      "loss": 0.1562,
+      "step": 16595
+    },
+    {
+      "epoch": 12.49228453142642,
+      "grad_norm": 0.21430274844169617,
+      "learning_rate": 5e-06,
+      "loss": 0.213,
+      "step": 16596
+    },
+    {
+      "epoch": 12.493037260067746,
+      "grad_norm": 0.21763411164283752,
+      "learning_rate": 5e-06,
+      "loss": 0.1795,
+      "step": 16597
+    },
+    {
+      "epoch": 12.49378998870907,
+      "grad_norm": 0.1666228026151657,
+      "learning_rate": 5e-06,
+      "loss": 0.195,
+      "step": 16598
+    },
+    {
+      "epoch": 12.494542717350395,
+      "grad_norm": 0.19373956322669983,
+      "learning_rate": 5e-06,
+      "loss": 0.1654,
+      "step": 16599
+    },
+    {
+      "epoch": 12.49529544599172,
+      "grad_norm": 0.1482105702161789,
+      "learning_rate": 5e-06,
+      "loss": 0.0638,
+      "step": 16600
+    },
+    {
+      "epoch": 12.49529544599172,
+      "eval_loss": 0.1833503395318985,
+      "eval_runtime": 453.0031,
+      "eval_samples_per_second": 21.252,
+      "eval_steps_per_second": 0.664,
+      "step": 16600
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1000000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 754,
+  "save_steps": 200,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 8.995170222958287e+19,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}