Files
qwen3-8b-codev-sva-ol-dfs-t…/trainer_state.json
ModelHub XC 6da7369235 初始化项目,由ModelHub XC社区提供模型
Model: Jalik/qwen3-8b-codev-sva-ol-dfs-think
Source: Original Platform
2026-10-05 23:35:23 +08:00

13169 lines
298 KiB
JSON

{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 1875,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0016,
"grad_norm": 34.81804204289023,
"learning_rate": 0.0,
"loss": 1.7558,
"step": 1
},
{
"epoch": 0.0032,
"grad_norm": 29.847368472071775,
"learning_rate": 5.319148936170213e-08,
"loss": 1.5684,
"step": 2
},
{
"epoch": 0.0048,
"grad_norm": 30.71913367274237,
"learning_rate": 1.0638297872340426e-07,
"loss": 1.5991,
"step": 3
},
{
"epoch": 0.0064,
"grad_norm": 24.09766090939983,
"learning_rate": 1.5957446808510638e-07,
"loss": 1.4263,
"step": 4
},
{
"epoch": 0.008,
"grad_norm": 29.105987706228593,
"learning_rate": 2.1276595744680852e-07,
"loss": 1.4938,
"step": 5
},
{
"epoch": 0.0096,
"grad_norm": 19.764415934621866,
"learning_rate": 2.6595744680851066e-07,
"loss": 1.1812,
"step": 6
},
{
"epoch": 0.0112,
"grad_norm": 22.591265036151032,
"learning_rate": 3.1914893617021275e-07,
"loss": 1.3006,
"step": 7
},
{
"epoch": 0.0128,
"grad_norm": 23.598748315811964,
"learning_rate": 3.723404255319149e-07,
"loss": 1.2872,
"step": 8
},
{
"epoch": 0.0144,
"grad_norm": 16.088680684305555,
"learning_rate": 4.2553191489361704e-07,
"loss": 1.0273,
"step": 9
},
{
"epoch": 0.016,
"grad_norm": 26.863393663561652,
"learning_rate": 4.787234042553192e-07,
"loss": 1.5081,
"step": 10
},
{
"epoch": 0.0176,
"grad_norm": 18.999964625925074,
"learning_rate": 5.319148936170213e-07,
"loss": 1.1982,
"step": 11
},
{
"epoch": 0.0192,
"grad_norm": 24.923127976171696,
"learning_rate": 5.851063829787235e-07,
"loss": 1.3732,
"step": 12
},
{
"epoch": 0.0208,
"grad_norm": 23.440205443390177,
"learning_rate": 6.382978723404255e-07,
"loss": 1.3377,
"step": 13
},
{
"epoch": 0.0224,
"grad_norm": 18.25311192687804,
"learning_rate": 6.914893617021278e-07,
"loss": 1.428,
"step": 14
},
{
"epoch": 0.024,
"grad_norm": 12.892517308900176,
"learning_rate": 7.446808510638298e-07,
"loss": 1.1496,
"step": 15
},
{
"epoch": 0.0256,
"grad_norm": 14.100509066254135,
"learning_rate": 7.97872340425532e-07,
"loss": 1.1877,
"step": 16
},
{
"epoch": 0.0272,
"grad_norm": 10.786148534583685,
"learning_rate": 8.510638297872341e-07,
"loss": 1.0393,
"step": 17
},
{
"epoch": 0.0288,
"grad_norm": 14.933814362396397,
"learning_rate": 9.042553191489363e-07,
"loss": 1.2927,
"step": 18
},
{
"epoch": 0.0304,
"grad_norm": 8.660424034825013,
"learning_rate": 9.574468085106384e-07,
"loss": 1.2121,
"step": 19
},
{
"epoch": 0.032,
"grad_norm": 7.461987158895768,
"learning_rate": 1.0106382978723404e-06,
"loss": 1.0941,
"step": 20
},
{
"epoch": 0.0336,
"grad_norm": 8.661728314471343,
"learning_rate": 1.0638297872340427e-06,
"loss": 1.2851,
"step": 21
},
{
"epoch": 0.0352,
"grad_norm": 6.938072005399082,
"learning_rate": 1.1170212765957447e-06,
"loss": 1.0711,
"step": 22
},
{
"epoch": 0.0368,
"grad_norm": 6.478150006679336,
"learning_rate": 1.170212765957447e-06,
"loss": 1.0088,
"step": 23
},
{
"epoch": 0.0384,
"grad_norm": 12.420208557365262,
"learning_rate": 1.223404255319149e-06,
"loss": 0.8255,
"step": 24
},
{
"epoch": 0.04,
"grad_norm": 4.256219375206905,
"learning_rate": 1.276595744680851e-06,
"loss": 0.7457,
"step": 25
},
{
"epoch": 0.0416,
"grad_norm": 5.642485493974883,
"learning_rate": 1.3297872340425533e-06,
"loss": 0.9917,
"step": 26
},
{
"epoch": 0.0432,
"grad_norm": 4.998433770063732,
"learning_rate": 1.3829787234042555e-06,
"loss": 0.9203,
"step": 27
},
{
"epoch": 0.0448,
"grad_norm": 3.910450507681862,
"learning_rate": 1.4361702127659578e-06,
"loss": 0.7581,
"step": 28
},
{
"epoch": 0.0464,
"grad_norm": 3.6802547625847524,
"learning_rate": 1.4893617021276596e-06,
"loss": 0.7426,
"step": 29
},
{
"epoch": 0.048,
"grad_norm": 4.849653088047029,
"learning_rate": 1.5425531914893618e-06,
"loss": 0.9024,
"step": 30
},
{
"epoch": 0.0496,
"grad_norm": 3.9953463767523414,
"learning_rate": 1.595744680851064e-06,
"loss": 0.7524,
"step": 31
},
{
"epoch": 0.0512,
"grad_norm": 3.781836802001676,
"learning_rate": 1.648936170212766e-06,
"loss": 0.7384,
"step": 32
},
{
"epoch": 0.0528,
"grad_norm": 3.917938871237599,
"learning_rate": 1.7021276595744682e-06,
"loss": 0.76,
"step": 33
},
{
"epoch": 0.0544,
"grad_norm": 3.4662569311388762,
"learning_rate": 1.7553191489361704e-06,
"loss": 0.6718,
"step": 34
},
{
"epoch": 0.056,
"grad_norm": 3.2274102891035947,
"learning_rate": 1.8085106382978727e-06,
"loss": 0.6547,
"step": 35
},
{
"epoch": 0.0576,
"grad_norm": 2.5721580950742893,
"learning_rate": 1.8617021276595745e-06,
"loss": 0.489,
"step": 36
},
{
"epoch": 0.0592,
"grad_norm": 3.2354945128207206,
"learning_rate": 1.9148936170212767e-06,
"loss": 0.5163,
"step": 37
},
{
"epoch": 0.0608,
"grad_norm": 2.332315364701189,
"learning_rate": 1.968085106382979e-06,
"loss": 0.4564,
"step": 38
},
{
"epoch": 0.0624,
"grad_norm": 2.691126600320836,
"learning_rate": 2.021276595744681e-06,
"loss": 0.5135,
"step": 39
},
{
"epoch": 0.064,
"grad_norm": 2.463571059719218,
"learning_rate": 2.074468085106383e-06,
"loss": 0.4551,
"step": 40
},
{
"epoch": 0.0656,
"grad_norm": 2.760891972169866,
"learning_rate": 2.1276595744680853e-06,
"loss": 0.4217,
"step": 41
},
{
"epoch": 0.0672,
"grad_norm": 2.6350876508712724,
"learning_rate": 2.1808510638297876e-06,
"loss": 0.4731,
"step": 42
},
{
"epoch": 0.0688,
"grad_norm": 4.289834952715711,
"learning_rate": 2.2340425531914894e-06,
"loss": 0.4462,
"step": 43
},
{
"epoch": 0.0704,
"grad_norm": 3.139932157274447,
"learning_rate": 2.2872340425531916e-06,
"loss": 0.3767,
"step": 44
},
{
"epoch": 0.072,
"grad_norm": 2.858579032192019,
"learning_rate": 2.340425531914894e-06,
"loss": 0.4098,
"step": 45
},
{
"epoch": 0.0736,
"grad_norm": 2.1580095461676962,
"learning_rate": 2.393617021276596e-06,
"loss": 0.2692,
"step": 46
},
{
"epoch": 0.0752,
"grad_norm": 1.9320867319688897,
"learning_rate": 2.446808510638298e-06,
"loss": 0.2546,
"step": 47
},
{
"epoch": 0.0768,
"grad_norm": 2.473999083083898,
"learning_rate": 2.5e-06,
"loss": 0.2815,
"step": 48
},
{
"epoch": 0.0784,
"grad_norm": 2.356216975780871,
"learning_rate": 2.553191489361702e-06,
"loss": 0.2559,
"step": 49
},
{
"epoch": 0.08,
"grad_norm": 2.5350257822407984,
"learning_rate": 2.6063829787234047e-06,
"loss": 0.2376,
"step": 50
},
{
"epoch": 0.0816,
"grad_norm": 2.102101009517552,
"learning_rate": 2.6595744680851065e-06,
"loss": 0.2598,
"step": 51
},
{
"epoch": 0.0832,
"grad_norm": 2.7375926783200133,
"learning_rate": 2.7127659574468084e-06,
"loss": 0.2973,
"step": 52
},
{
"epoch": 0.0848,
"grad_norm": 1.5657418234005696,
"learning_rate": 2.765957446808511e-06,
"loss": 0.2647,
"step": 53
},
{
"epoch": 0.0864,
"grad_norm": 1.550170816967778,
"learning_rate": 2.819148936170213e-06,
"loss": 0.275,
"step": 54
},
{
"epoch": 0.088,
"grad_norm": 2.073432304086172,
"learning_rate": 2.8723404255319155e-06,
"loss": 0.2637,
"step": 55
},
{
"epoch": 0.0896,
"grad_norm": 1.4100443840654642,
"learning_rate": 2.9255319148936174e-06,
"loss": 0.2031,
"step": 56
},
{
"epoch": 0.0912,
"grad_norm": 1.3952044035275684,
"learning_rate": 2.978723404255319e-06,
"loss": 0.2321,
"step": 57
},
{
"epoch": 0.0928,
"grad_norm": 1.4177276174302869,
"learning_rate": 3.031914893617022e-06,
"loss": 0.193,
"step": 58
},
{
"epoch": 0.0944,
"grad_norm": 1.5203039554391684,
"learning_rate": 3.0851063829787237e-06,
"loss": 0.2114,
"step": 59
},
{
"epoch": 0.096,
"grad_norm": 1.698396223303269,
"learning_rate": 3.1382978723404255e-06,
"loss": 0.1947,
"step": 60
},
{
"epoch": 0.0976,
"grad_norm": 1.34959859492364,
"learning_rate": 3.191489361702128e-06,
"loss": 0.2518,
"step": 61
},
{
"epoch": 0.0992,
"grad_norm": 1.4188312579465134,
"learning_rate": 3.24468085106383e-06,
"loss": 0.2451,
"step": 62
},
{
"epoch": 0.1008,
"grad_norm": 1.4195371426245993,
"learning_rate": 3.297872340425532e-06,
"loss": 0.2331,
"step": 63
},
{
"epoch": 0.1024,
"grad_norm": 1.2546229391333108,
"learning_rate": 3.3510638297872345e-06,
"loss": 0.2277,
"step": 64
},
{
"epoch": 0.104,
"grad_norm": 1.2069300037980768,
"learning_rate": 3.4042553191489363e-06,
"loss": 0.2308,
"step": 65
},
{
"epoch": 0.1056,
"grad_norm": 1.189546489104783,
"learning_rate": 3.457446808510639e-06,
"loss": 0.1966,
"step": 66
},
{
"epoch": 0.1072,
"grad_norm": 1.3061420625954332,
"learning_rate": 3.510638297872341e-06,
"loss": 0.1991,
"step": 67
},
{
"epoch": 0.1088,
"grad_norm": 1.8421646589986203,
"learning_rate": 3.5638297872340426e-06,
"loss": 0.2396,
"step": 68
},
{
"epoch": 0.1104,
"grad_norm": 0.9224468400304557,
"learning_rate": 3.6170212765957453e-06,
"loss": 0.2011,
"step": 69
},
{
"epoch": 0.112,
"grad_norm": 1.3114522097104808,
"learning_rate": 3.670212765957447e-06,
"loss": 0.219,
"step": 70
},
{
"epoch": 0.1136,
"grad_norm": 1.4748776486522222,
"learning_rate": 3.723404255319149e-06,
"loss": 0.2534,
"step": 71
},
{
"epoch": 0.1152,
"grad_norm": 1.137834666249036,
"learning_rate": 3.7765957446808516e-06,
"loss": 0.2166,
"step": 72
},
{
"epoch": 0.1168,
"grad_norm": 1.10444394613435,
"learning_rate": 3.8297872340425535e-06,
"loss": 0.2073,
"step": 73
},
{
"epoch": 0.1184,
"grad_norm": 1.1592849866146364,
"learning_rate": 3.882978723404256e-06,
"loss": 0.2513,
"step": 74
},
{
"epoch": 0.12,
"grad_norm": 1.2338439205669296,
"learning_rate": 3.936170212765958e-06,
"loss": 0.2204,
"step": 75
},
{
"epoch": 0.1216,
"grad_norm": 1.5546651999226568,
"learning_rate": 3.98936170212766e-06,
"loss": 0.2407,
"step": 76
},
{
"epoch": 0.1232,
"grad_norm": 1.1031139696989105,
"learning_rate": 4.042553191489362e-06,
"loss": 0.1683,
"step": 77
},
{
"epoch": 0.1248,
"grad_norm": 1.1594158183351086,
"learning_rate": 4.095744680851064e-06,
"loss": 0.2298,
"step": 78
},
{
"epoch": 0.1264,
"grad_norm": 1.2347173901944344,
"learning_rate": 4.148936170212766e-06,
"loss": 0.1921,
"step": 79
},
{
"epoch": 0.128,
"grad_norm": 1.30882550818402,
"learning_rate": 4.202127659574468e-06,
"loss": 0.2406,
"step": 80
},
{
"epoch": 0.1296,
"grad_norm": 0.88988511844779,
"learning_rate": 4.255319148936171e-06,
"loss": 0.1797,
"step": 81
},
{
"epoch": 0.1312,
"grad_norm": 1.1127081693837713,
"learning_rate": 4.308510638297873e-06,
"loss": 0.1994,
"step": 82
},
{
"epoch": 0.1328,
"grad_norm": 1.3917276906514493,
"learning_rate": 4.361702127659575e-06,
"loss": 0.1959,
"step": 83
},
{
"epoch": 0.1344,
"grad_norm": 1.232432337854429,
"learning_rate": 4.414893617021277e-06,
"loss": 0.2391,
"step": 84
},
{
"epoch": 0.136,
"grad_norm": 0.9181391935541895,
"learning_rate": 4.468085106382979e-06,
"loss": 0.2017,
"step": 85
},
{
"epoch": 0.1376,
"grad_norm": 1.4524273144930377,
"learning_rate": 4.521276595744681e-06,
"loss": 0.2232,
"step": 86
},
{
"epoch": 0.1392,
"grad_norm": 1.074195487481202,
"learning_rate": 4.574468085106383e-06,
"loss": 0.1974,
"step": 87
},
{
"epoch": 0.1408,
"grad_norm": 0.8649732503142646,
"learning_rate": 4.6276595744680855e-06,
"loss": 0.1871,
"step": 88
},
{
"epoch": 0.1424,
"grad_norm": 0.9219465940022177,
"learning_rate": 4.680851063829788e-06,
"loss": 0.1812,
"step": 89
},
{
"epoch": 0.144,
"grad_norm": 1.2831146101288808,
"learning_rate": 4.73404255319149e-06,
"loss": 0.2056,
"step": 90
},
{
"epoch": 0.1456,
"grad_norm": 0.9382518652856864,
"learning_rate": 4.787234042553192e-06,
"loss": 0.1915,
"step": 91
},
{
"epoch": 0.1472,
"grad_norm": 0.8726091528264279,
"learning_rate": 4.840425531914894e-06,
"loss": 0.2048,
"step": 92
},
{
"epoch": 0.1488,
"grad_norm": 1.3993412170147372,
"learning_rate": 4.893617021276596e-06,
"loss": 0.2585,
"step": 93
},
{
"epoch": 0.1504,
"grad_norm": 1.098531400595422,
"learning_rate": 4.946808510638298e-06,
"loss": 0.1849,
"step": 94
},
{
"epoch": 0.152,
"grad_norm": 1.005545232975835,
"learning_rate": 5e-06,
"loss": 0.1459,
"step": 95
},
{
"epoch": 0.1536,
"grad_norm": 1.14076201383608,
"learning_rate": 5.053191489361703e-06,
"loss": 0.2284,
"step": 96
},
{
"epoch": 0.1552,
"grad_norm": 0.8900380282086686,
"learning_rate": 5.106382978723404e-06,
"loss": 0.1763,
"step": 97
},
{
"epoch": 0.1568,
"grad_norm": 0.8834382982696954,
"learning_rate": 5.159574468085107e-06,
"loss": 0.2249,
"step": 98
},
{
"epoch": 0.1584,
"grad_norm": 0.9486006436551419,
"learning_rate": 5.212765957446809e-06,
"loss": 0.2234,
"step": 99
},
{
"epoch": 0.16,
"grad_norm": 1.0091476503237242,
"learning_rate": 5.265957446808511e-06,
"loss": 0.2079,
"step": 100
},
{
"epoch": 0.1616,
"grad_norm": 0.998574344088009,
"learning_rate": 5.319148936170213e-06,
"loss": 0.1963,
"step": 101
},
{
"epoch": 0.1632,
"grad_norm": 0.9903941078581274,
"learning_rate": 5.372340425531915e-06,
"loss": 0.2333,
"step": 102
},
{
"epoch": 0.1648,
"grad_norm": 0.9402330236041838,
"learning_rate": 5.425531914893617e-06,
"loss": 0.1805,
"step": 103
},
{
"epoch": 0.1664,
"grad_norm": 0.8931261773188303,
"learning_rate": 5.47872340425532e-06,
"loss": 0.2113,
"step": 104
},
{
"epoch": 0.168,
"grad_norm": 1.3486991572508258,
"learning_rate": 5.531914893617022e-06,
"loss": 0.2161,
"step": 105
},
{
"epoch": 0.1696,
"grad_norm": 1.2139960696646945,
"learning_rate": 5.5851063829787235e-06,
"loss": 0.2309,
"step": 106
},
{
"epoch": 0.1712,
"grad_norm": 0.9467928976819324,
"learning_rate": 5.638297872340426e-06,
"loss": 0.2108,
"step": 107
},
{
"epoch": 0.1728,
"grad_norm": 1.118607529326198,
"learning_rate": 5.691489361702128e-06,
"loss": 0.2215,
"step": 108
},
{
"epoch": 0.1744,
"grad_norm": 1.0730425995868642,
"learning_rate": 5.744680851063831e-06,
"loss": 0.1933,
"step": 109
},
{
"epoch": 0.176,
"grad_norm": 1.0833551828512582,
"learning_rate": 5.7978723404255325e-06,
"loss": 0.1954,
"step": 110
},
{
"epoch": 0.1776,
"grad_norm": 1.0282419711833737,
"learning_rate": 5.851063829787235e-06,
"loss": 0.1497,
"step": 111
},
{
"epoch": 0.1792,
"grad_norm": 1.0337107244189432,
"learning_rate": 5.904255319148937e-06,
"loss": 0.1934,
"step": 112
},
{
"epoch": 0.1808,
"grad_norm": 0.9269854704425737,
"learning_rate": 5.957446808510638e-06,
"loss": 0.1744,
"step": 113
},
{
"epoch": 0.1824,
"grad_norm": 0.9108981439956981,
"learning_rate": 6.010638297872341e-06,
"loss": 0.2075,
"step": 114
},
{
"epoch": 0.184,
"grad_norm": 1.009860498351711,
"learning_rate": 6.063829787234044e-06,
"loss": 0.2381,
"step": 115
},
{
"epoch": 0.1856,
"grad_norm": 1.1908833149233145,
"learning_rate": 6.117021276595745e-06,
"loss": 0.2246,
"step": 116
},
{
"epoch": 0.1872,
"grad_norm": 1.116575749569153,
"learning_rate": 6.170212765957447e-06,
"loss": 0.1884,
"step": 117
},
{
"epoch": 0.1888,
"grad_norm": 0.9646592503991718,
"learning_rate": 6.22340425531915e-06,
"loss": 0.2127,
"step": 118
},
{
"epoch": 0.1904,
"grad_norm": 0.9708025877530861,
"learning_rate": 6.276595744680851e-06,
"loss": 0.2011,
"step": 119
},
{
"epoch": 0.192,
"grad_norm": 1.002147495702246,
"learning_rate": 6.329787234042554e-06,
"loss": 0.1749,
"step": 120
},
{
"epoch": 0.1936,
"grad_norm": 1.10565970599322,
"learning_rate": 6.382978723404256e-06,
"loss": 0.183,
"step": 121
},
{
"epoch": 0.1952,
"grad_norm": 1.0361932640498404,
"learning_rate": 6.436170212765958e-06,
"loss": 0.1852,
"step": 122
},
{
"epoch": 0.1968,
"grad_norm": 0.9678172613312804,
"learning_rate": 6.48936170212766e-06,
"loss": 0.2119,
"step": 123
},
{
"epoch": 0.1984,
"grad_norm": 0.8876958334276951,
"learning_rate": 6.542553191489362e-06,
"loss": 0.1881,
"step": 124
},
{
"epoch": 0.2,
"grad_norm": 0.8265422349576076,
"learning_rate": 6.595744680851064e-06,
"loss": 0.2054,
"step": 125
},
{
"epoch": 0.2016,
"grad_norm": 1.0310525539429798,
"learning_rate": 6.648936170212767e-06,
"loss": 0.175,
"step": 126
},
{
"epoch": 0.2032,
"grad_norm": 0.9481882032169233,
"learning_rate": 6.702127659574469e-06,
"loss": 0.2214,
"step": 127
},
{
"epoch": 0.2048,
"grad_norm": 0.8713333716559208,
"learning_rate": 6.75531914893617e-06,
"loss": 0.1804,
"step": 128
},
{
"epoch": 0.2064,
"grad_norm": 1.0831090182032603,
"learning_rate": 6.808510638297873e-06,
"loss": 0.241,
"step": 129
},
{
"epoch": 0.208,
"grad_norm": 0.936276134068846,
"learning_rate": 6.861702127659575e-06,
"loss": 0.1865,
"step": 130
},
{
"epoch": 0.2096,
"grad_norm": 0.8096308404464423,
"learning_rate": 6.914893617021278e-06,
"loss": 0.1705,
"step": 131
},
{
"epoch": 0.2112,
"grad_norm": 0.8394534605790132,
"learning_rate": 6.968085106382979e-06,
"loss": 0.207,
"step": 132
},
{
"epoch": 0.2128,
"grad_norm": 0.9679523901331609,
"learning_rate": 7.021276595744682e-06,
"loss": 0.1729,
"step": 133
},
{
"epoch": 0.2144,
"grad_norm": 1.0117276742001118,
"learning_rate": 7.074468085106384e-06,
"loss": 0.2235,
"step": 134
},
{
"epoch": 0.216,
"grad_norm": 0.9556918733327034,
"learning_rate": 7.127659574468085e-06,
"loss": 0.1824,
"step": 135
},
{
"epoch": 0.2176,
"grad_norm": 1.0263523377296133,
"learning_rate": 7.1808510638297875e-06,
"loss": 0.1795,
"step": 136
},
{
"epoch": 0.2192,
"grad_norm": 0.7718682019474808,
"learning_rate": 7.234042553191491e-06,
"loss": 0.1698,
"step": 137
},
{
"epoch": 0.2208,
"grad_norm": 0.9931525183507595,
"learning_rate": 7.287234042553192e-06,
"loss": 0.1696,
"step": 138
},
{
"epoch": 0.2224,
"grad_norm": 0.8940617320927301,
"learning_rate": 7.340425531914894e-06,
"loss": 0.1948,
"step": 139
},
{
"epoch": 0.224,
"grad_norm": 0.9608452518771136,
"learning_rate": 7.3936170212765965e-06,
"loss": 0.2156,
"step": 140
},
{
"epoch": 0.2256,
"grad_norm": 0.7050894052585596,
"learning_rate": 7.446808510638298e-06,
"loss": 0.1385,
"step": 141
},
{
"epoch": 0.2272,
"grad_norm": 0.8725390228307975,
"learning_rate": 7.500000000000001e-06,
"loss": 0.1958,
"step": 142
},
{
"epoch": 0.2288,
"grad_norm": 0.7830817074061183,
"learning_rate": 7.553191489361703e-06,
"loss": 0.1659,
"step": 143
},
{
"epoch": 0.2304,
"grad_norm": 0.8636975458408307,
"learning_rate": 7.606382978723405e-06,
"loss": 0.1824,
"step": 144
},
{
"epoch": 0.232,
"grad_norm": 0.9171764954560274,
"learning_rate": 7.659574468085107e-06,
"loss": 0.2313,
"step": 145
},
{
"epoch": 0.2336,
"grad_norm": 1.0573238661438547,
"learning_rate": 7.71276595744681e-06,
"loss": 0.1941,
"step": 146
},
{
"epoch": 0.2352,
"grad_norm": 1.0146849731095173,
"learning_rate": 7.765957446808511e-06,
"loss": 0.21,
"step": 147
},
{
"epoch": 0.2368,
"grad_norm": 0.9059439095236885,
"learning_rate": 7.819148936170213e-06,
"loss": 0.1748,
"step": 148
},
{
"epoch": 0.2384,
"grad_norm": 0.8796921570027063,
"learning_rate": 7.872340425531916e-06,
"loss": 0.181,
"step": 149
},
{
"epoch": 0.24,
"grad_norm": 0.8088354938311507,
"learning_rate": 7.925531914893617e-06,
"loss": 0.1683,
"step": 150
},
{
"epoch": 0.2416,
"grad_norm": 0.7051026540856159,
"learning_rate": 7.97872340425532e-06,
"loss": 0.1748,
"step": 151
},
{
"epoch": 0.2432,
"grad_norm": 0.905965619426617,
"learning_rate": 8.031914893617022e-06,
"loss": 0.176,
"step": 152
},
{
"epoch": 0.2448,
"grad_norm": 0.8117737692873103,
"learning_rate": 8.085106382978723e-06,
"loss": 0.1859,
"step": 153
},
{
"epoch": 0.2464,
"grad_norm": 0.9079152300802589,
"learning_rate": 8.138297872340426e-06,
"loss": 0.1971,
"step": 154
},
{
"epoch": 0.248,
"grad_norm": 0.8851405735530934,
"learning_rate": 8.191489361702128e-06,
"loss": 0.1793,
"step": 155
},
{
"epoch": 0.2496,
"grad_norm": 0.8460401133778855,
"learning_rate": 8.24468085106383e-06,
"loss": 0.1918,
"step": 156
},
{
"epoch": 0.2512,
"grad_norm": 0.885321392566436,
"learning_rate": 8.297872340425532e-06,
"loss": 0.1788,
"step": 157
},
{
"epoch": 0.2528,
"grad_norm": 0.8479159982493871,
"learning_rate": 8.351063829787235e-06,
"loss": 0.1918,
"step": 158
},
{
"epoch": 0.2544,
"grad_norm": 0.8519790007187561,
"learning_rate": 8.404255319148937e-06,
"loss": 0.1769,
"step": 159
},
{
"epoch": 0.256,
"grad_norm": 0.7246956269643212,
"learning_rate": 8.457446808510638e-06,
"loss": 0.1683,
"step": 160
},
{
"epoch": 0.2576,
"grad_norm": 0.901471467782991,
"learning_rate": 8.510638297872341e-06,
"loss": 0.2026,
"step": 161
},
{
"epoch": 0.2592,
"grad_norm": 0.7623628673438987,
"learning_rate": 8.563829787234044e-06,
"loss": 0.1805,
"step": 162
},
{
"epoch": 0.2608,
"grad_norm": 0.979435305786489,
"learning_rate": 8.617021276595746e-06,
"loss": 0.197,
"step": 163
},
{
"epoch": 0.2624,
"grad_norm": 0.7996926495253834,
"learning_rate": 8.670212765957447e-06,
"loss": 0.1695,
"step": 164
},
{
"epoch": 0.264,
"grad_norm": 0.7883660266700971,
"learning_rate": 8.72340425531915e-06,
"loss": 0.2225,
"step": 165
},
{
"epoch": 0.2656,
"grad_norm": 0.6875386736382268,
"learning_rate": 8.776595744680852e-06,
"loss": 0.1554,
"step": 166
},
{
"epoch": 0.2672,
"grad_norm": 0.818740561138253,
"learning_rate": 8.829787234042555e-06,
"loss": 0.1661,
"step": 167
},
{
"epoch": 0.2688,
"grad_norm": 0.8588930589072132,
"learning_rate": 8.882978723404256e-06,
"loss": 0.1854,
"step": 168
},
{
"epoch": 0.2704,
"grad_norm": 0.8348247521804344,
"learning_rate": 8.936170212765958e-06,
"loss": 0.1927,
"step": 169
},
{
"epoch": 0.272,
"grad_norm": 0.8170142968264945,
"learning_rate": 8.98936170212766e-06,
"loss": 0.1995,
"step": 170
},
{
"epoch": 0.2736,
"grad_norm": 0.8390523642741157,
"learning_rate": 9.042553191489362e-06,
"loss": 0.1727,
"step": 171
},
{
"epoch": 0.2752,
"grad_norm": 0.8290260007234339,
"learning_rate": 9.095744680851063e-06,
"loss": 0.1668,
"step": 172
},
{
"epoch": 0.2768,
"grad_norm": 0.8749970870427213,
"learning_rate": 9.148936170212767e-06,
"loss": 0.2063,
"step": 173
},
{
"epoch": 0.2784,
"grad_norm": 0.8327162836404668,
"learning_rate": 9.20212765957447e-06,
"loss": 0.1558,
"step": 174
},
{
"epoch": 0.28,
"grad_norm": 0.8049696620288104,
"learning_rate": 9.255319148936171e-06,
"loss": 0.189,
"step": 175
},
{
"epoch": 0.2816,
"grad_norm": 0.8005808362446974,
"learning_rate": 9.308510638297872e-06,
"loss": 0.1613,
"step": 176
},
{
"epoch": 0.2832,
"grad_norm": 0.9208599859193953,
"learning_rate": 9.361702127659576e-06,
"loss": 0.1994,
"step": 177
},
{
"epoch": 0.2848,
"grad_norm": 0.9725925553911112,
"learning_rate": 9.414893617021279e-06,
"loss": 0.22,
"step": 178
},
{
"epoch": 0.2864,
"grad_norm": 0.9625902359519481,
"learning_rate": 9.46808510638298e-06,
"loss": 0.1778,
"step": 179
},
{
"epoch": 0.288,
"grad_norm": 0.90277120214638,
"learning_rate": 9.521276595744681e-06,
"loss": 0.1869,
"step": 180
},
{
"epoch": 0.2896,
"grad_norm": 0.7845396090783852,
"learning_rate": 9.574468085106385e-06,
"loss": 0.1863,
"step": 181
},
{
"epoch": 0.2912,
"grad_norm": 0.813647612600151,
"learning_rate": 9.627659574468086e-06,
"loss": 0.1901,
"step": 182
},
{
"epoch": 0.2928,
"grad_norm": 0.8887035454042846,
"learning_rate": 9.680851063829787e-06,
"loss": 0.201,
"step": 183
},
{
"epoch": 0.2944,
"grad_norm": 0.9545547384267982,
"learning_rate": 9.73404255319149e-06,
"loss": 0.1777,
"step": 184
},
{
"epoch": 0.296,
"grad_norm": 0.8705843456416451,
"learning_rate": 9.787234042553192e-06,
"loss": 0.1867,
"step": 185
},
{
"epoch": 0.2976,
"grad_norm": 0.9242700800811994,
"learning_rate": 9.840425531914895e-06,
"loss": 0.2043,
"step": 186
},
{
"epoch": 0.2992,
"grad_norm": 0.8154462695304701,
"learning_rate": 9.893617021276596e-06,
"loss": 0.1919,
"step": 187
},
{
"epoch": 0.3008,
"grad_norm": 0.7539977428561881,
"learning_rate": 9.946808510638298e-06,
"loss": 0.1575,
"step": 188
},
{
"epoch": 0.3024,
"grad_norm": 0.8356691934911893,
"learning_rate": 1e-05,
"loss": 0.1869,
"step": 189
},
{
"epoch": 0.304,
"grad_norm": 0.8536194249087785,
"learning_rate": 9.99999133019233e-06,
"loss": 0.1977,
"step": 190
},
{
"epoch": 0.3056,
"grad_norm": 0.768156375426687,
"learning_rate": 9.999965320799377e-06,
"loss": 0.17,
"step": 191
},
{
"epoch": 0.3072,
"grad_norm": 0.7424786234084764,
"learning_rate": 9.999921971911345e-06,
"loss": 0.1876,
"step": 192
},
{
"epoch": 0.3088,
"grad_norm": 0.9852271779497086,
"learning_rate": 9.999861283678563e-06,
"loss": 0.2211,
"step": 193
},
{
"epoch": 0.3104,
"grad_norm": 0.9856877073057979,
"learning_rate": 9.999783256311494e-06,
"loss": 0.2336,
"step": 194
},
{
"epoch": 0.312,
"grad_norm": 0.9780069710109418,
"learning_rate": 9.99968789008073e-06,
"loss": 0.1867,
"step": 195
},
{
"epoch": 0.3136,
"grad_norm": 0.9033871719455824,
"learning_rate": 9.999575185316994e-06,
"loss": 0.1814,
"step": 196
},
{
"epoch": 0.3152,
"grad_norm": 0.540175071720954,
"learning_rate": 9.999445142411139e-06,
"loss": 0.1632,
"step": 197
},
{
"epoch": 0.3168,
"grad_norm": 0.6943058717566751,
"learning_rate": 9.99929776181414e-06,
"loss": 0.1857,
"step": 198
},
{
"epoch": 0.3184,
"grad_norm": 0.8047729285228222,
"learning_rate": 9.999133044037107e-06,
"loss": 0.1882,
"step": 199
},
{
"epoch": 0.32,
"grad_norm": 0.8422459986043794,
"learning_rate": 9.998950989651261e-06,
"loss": 0.1902,
"step": 200
},
{
"epoch": 0.3216,
"grad_norm": 0.9208395244659524,
"learning_rate": 9.99875159928796e-06,
"loss": 0.1752,
"step": 201
},
{
"epoch": 0.3232,
"grad_norm": 0.770941522267793,
"learning_rate": 9.99853487363867e-06,
"loss": 0.164,
"step": 202
},
{
"epoch": 0.3248,
"grad_norm": 0.7315255025361251,
"learning_rate": 9.998300813454981e-06,
"loss": 0.1709,
"step": 203
},
{
"epoch": 0.3264,
"grad_norm": 0.7454580780623087,
"learning_rate": 9.998049419548597e-06,
"loss": 0.1927,
"step": 204
},
{
"epoch": 0.328,
"grad_norm": 0.8608794006880942,
"learning_rate": 9.997780692791329e-06,
"loss": 0.2194,
"step": 205
},
{
"epoch": 0.3296,
"grad_norm": 0.6137485623315554,
"learning_rate": 9.997494634115101e-06,
"loss": 0.1781,
"step": 206
},
{
"epoch": 0.3312,
"grad_norm": 1.0158570417327797,
"learning_rate": 9.997191244511947e-06,
"loss": 0.2266,
"step": 207
},
{
"epoch": 0.3328,
"grad_norm": 0.739236518675017,
"learning_rate": 9.996870525033994e-06,
"loss": 0.2024,
"step": 208
},
{
"epoch": 0.3344,
"grad_norm": 0.6984041111510425,
"learning_rate": 9.996532476793475e-06,
"loss": 0.2049,
"step": 209
},
{
"epoch": 0.336,
"grad_norm": 0.7926701959269863,
"learning_rate": 9.996177100962714e-06,
"loss": 0.1891,
"step": 210
},
{
"epoch": 0.3376,
"grad_norm": 0.7325495334014205,
"learning_rate": 9.995804398774129e-06,
"loss": 0.1939,
"step": 211
},
{
"epoch": 0.3392,
"grad_norm": 0.7857835726430116,
"learning_rate": 9.99541437152022e-06,
"loss": 0.2007,
"step": 212
},
{
"epoch": 0.3408,
"grad_norm": 0.9845202347496962,
"learning_rate": 9.995007020553572e-06,
"loss": 0.209,
"step": 213
},
{
"epoch": 0.3424,
"grad_norm": 0.7875596803896031,
"learning_rate": 9.994582347286849e-06,
"loss": 0.1709,
"step": 214
},
{
"epoch": 0.344,
"grad_norm": 0.8190676375054365,
"learning_rate": 9.994140353192782e-06,
"loss": 0.1587,
"step": 215
},
{
"epoch": 0.3456,
"grad_norm": 0.7225187532590287,
"learning_rate": 9.993681039804176e-06,
"loss": 0.2041,
"step": 216
},
{
"epoch": 0.3472,
"grad_norm": 0.8894647970085704,
"learning_rate": 9.99320440871389e-06,
"loss": 0.1927,
"step": 217
},
{
"epoch": 0.3488,
"grad_norm": 0.7426563853498096,
"learning_rate": 9.99271046157485e-06,
"loss": 0.1963,
"step": 218
},
{
"epoch": 0.3504,
"grad_norm": 0.6656676936275191,
"learning_rate": 9.99219920010002e-06,
"loss": 0.1618,
"step": 219
},
{
"epoch": 0.352,
"grad_norm": 0.8893540686062376,
"learning_rate": 9.991670626062422e-06,
"loss": 0.1994,
"step": 220
},
{
"epoch": 0.3536,
"grad_norm": 0.8377179172819855,
"learning_rate": 9.991124741295106e-06,
"loss": 0.1819,
"step": 221
},
{
"epoch": 0.3552,
"grad_norm": 0.6560502327397431,
"learning_rate": 9.990561547691159e-06,
"loss": 0.1504,
"step": 222
},
{
"epoch": 0.3568,
"grad_norm": 0.6455997710131487,
"learning_rate": 9.989981047203693e-06,
"loss": 0.1628,
"step": 223
},
{
"epoch": 0.3584,
"grad_norm": 0.6973316554031512,
"learning_rate": 9.98938324184584e-06,
"loss": 0.165,
"step": 224
},
{
"epoch": 0.36,
"grad_norm": 0.8465668350347735,
"learning_rate": 9.988768133690741e-06,
"loss": 0.2048,
"step": 225
},
{
"epoch": 0.3616,
"grad_norm": 0.7111891147862526,
"learning_rate": 9.988135724871546e-06,
"loss": 0.1686,
"step": 226
},
{
"epoch": 0.3632,
"grad_norm": 0.723913897728495,
"learning_rate": 9.987486017581401e-06,
"loss": 0.1545,
"step": 227
},
{
"epoch": 0.3648,
"grad_norm": 0.8432617443466722,
"learning_rate": 9.986819014073436e-06,
"loss": 0.1964,
"step": 228
},
{
"epoch": 0.3664,
"grad_norm": 0.8170183822314353,
"learning_rate": 9.986134716660774e-06,
"loss": 0.1509,
"step": 229
},
{
"epoch": 0.368,
"grad_norm": 0.7859910749939657,
"learning_rate": 9.9854331277165e-06,
"loss": 0.1956,
"step": 230
},
{
"epoch": 0.3696,
"grad_norm": 0.7779422825556831,
"learning_rate": 9.984714249673676e-06,
"loss": 0.2122,
"step": 231
},
{
"epoch": 0.3712,
"grad_norm": 0.7688083186033569,
"learning_rate": 9.98397808502531e-06,
"loss": 0.1831,
"step": 232
},
{
"epoch": 0.3728,
"grad_norm": 0.7550273791504949,
"learning_rate": 9.983224636324369e-06,
"loss": 0.2018,
"step": 233
},
{
"epoch": 0.3744,
"grad_norm": 0.7485064720681825,
"learning_rate": 9.982453906183754e-06,
"loss": 0.173,
"step": 234
},
{
"epoch": 0.376,
"grad_norm": 0.8080631676953133,
"learning_rate": 9.981665897276298e-06,
"loss": 0.1766,
"step": 235
},
{
"epoch": 0.3776,
"grad_norm": 0.6751261234025089,
"learning_rate": 9.980860612334753e-06,
"loss": 0.1685,
"step": 236
},
{
"epoch": 0.3792,
"grad_norm": 0.6799215941915618,
"learning_rate": 9.980038054151789e-06,
"loss": 0.1834,
"step": 237
},
{
"epoch": 0.3808,
"grad_norm": 0.6635736482383496,
"learning_rate": 9.979198225579968e-06,
"loss": 0.1648,
"step": 238
},
{
"epoch": 0.3824,
"grad_norm": 0.8768185721999313,
"learning_rate": 9.97834112953176e-06,
"loss": 0.1853,
"step": 239
},
{
"epoch": 0.384,
"grad_norm": 0.7511012684386876,
"learning_rate": 9.9774667689795e-06,
"loss": 0.1884,
"step": 240
},
{
"epoch": 0.3856,
"grad_norm": 0.8093211599229085,
"learning_rate": 9.976575146955409e-06,
"loss": 0.1682,
"step": 241
},
{
"epoch": 0.3872,
"grad_norm": 0.6543764192156543,
"learning_rate": 9.97566626655156e-06,
"loss": 0.1659,
"step": 242
},
{
"epoch": 0.3888,
"grad_norm": 0.7759889183889903,
"learning_rate": 9.974740130919883e-06,
"loss": 0.2275,
"step": 243
},
{
"epoch": 0.3904,
"grad_norm": 0.5852882264706131,
"learning_rate": 9.973796743272141e-06,
"loss": 0.1736,
"step": 244
},
{
"epoch": 0.392,
"grad_norm": 0.6537213929728348,
"learning_rate": 9.972836106879936e-06,
"loss": 0.1849,
"step": 245
},
{
"epoch": 0.3936,
"grad_norm": 0.7260390448608465,
"learning_rate": 9.971858225074672e-06,
"loss": 0.1759,
"step": 246
},
{
"epoch": 0.3952,
"grad_norm": 0.6709809367612417,
"learning_rate": 9.970863101247578e-06,
"loss": 0.2074,
"step": 247
},
{
"epoch": 0.3968,
"grad_norm": 0.6515136341965111,
"learning_rate": 9.96985073884966e-06,
"loss": 0.1993,
"step": 248
},
{
"epoch": 0.3984,
"grad_norm": 0.6395755581836635,
"learning_rate": 9.968821141391716e-06,
"loss": 0.1948,
"step": 249
},
{
"epoch": 0.4,
"grad_norm": 0.6523117623124721,
"learning_rate": 9.96777431244431e-06,
"loss": 0.1748,
"step": 250
},
{
"epoch": 0.4016,
"grad_norm": 0.6829817949005532,
"learning_rate": 9.966710255637764e-06,
"loss": 0.1728,
"step": 251
},
{
"epoch": 0.4032,
"grad_norm": 0.729227819156597,
"learning_rate": 9.965628974662145e-06,
"loss": 0.1822,
"step": 252
},
{
"epoch": 0.4048,
"grad_norm": 0.8261375625959386,
"learning_rate": 9.964530473267253e-06,
"loss": 0.199,
"step": 253
},
{
"epoch": 0.4064,
"grad_norm": 0.7181569212285316,
"learning_rate": 9.963414755262606e-06,
"loss": 0.1897,
"step": 254
},
{
"epoch": 0.408,
"grad_norm": 0.7852098756905134,
"learning_rate": 9.962281824517427e-06,
"loss": 0.2056,
"step": 255
},
{
"epoch": 0.4096,
"grad_norm": 0.694703019528023,
"learning_rate": 9.961131684960635e-06,
"loss": 0.1785,
"step": 256
},
{
"epoch": 0.4112,
"grad_norm": 0.6068705767544588,
"learning_rate": 9.959964340580823e-06,
"loss": 0.1851,
"step": 257
},
{
"epoch": 0.4128,
"grad_norm": 0.7057225965922964,
"learning_rate": 9.958779795426253e-06,
"loss": 0.1923,
"step": 258
},
{
"epoch": 0.4144,
"grad_norm": 0.7124877903547212,
"learning_rate": 9.957578053604837e-06,
"loss": 0.1921,
"step": 259
},
{
"epoch": 0.416,
"grad_norm": 0.6489807825871694,
"learning_rate": 9.956359119284123e-06,
"loss": 0.1731,
"step": 260
},
{
"epoch": 0.4176,
"grad_norm": 0.7843288676418799,
"learning_rate": 9.955122996691278e-06,
"loss": 0.182,
"step": 261
},
{
"epoch": 0.4192,
"grad_norm": 0.6487522767383104,
"learning_rate": 9.953869690113085e-06,
"loss": 0.1825,
"step": 262
},
{
"epoch": 0.4208,
"grad_norm": 0.8301584674645015,
"learning_rate": 9.952599203895912e-06,
"loss": 0.1916,
"step": 263
},
{
"epoch": 0.4224,
"grad_norm": 0.7944475583965833,
"learning_rate": 9.95131154244571e-06,
"loss": 0.181,
"step": 264
},
{
"epoch": 0.424,
"grad_norm": 0.7123757126903068,
"learning_rate": 9.950006710227986e-06,
"loss": 0.169,
"step": 265
},
{
"epoch": 0.4256,
"grad_norm": 0.7036540069064404,
"learning_rate": 9.9486847117678e-06,
"loss": 0.1758,
"step": 266
},
{
"epoch": 0.4272,
"grad_norm": 0.7131761102260094,
"learning_rate": 9.947345551649741e-06,
"loss": 0.2225,
"step": 267
},
{
"epoch": 0.4288,
"grad_norm": 0.7001326122449604,
"learning_rate": 9.945989234517913e-06,
"loss": 0.172,
"step": 268
},
{
"epoch": 0.4304,
"grad_norm": 0.7104150287419875,
"learning_rate": 9.94461576507592e-06,
"loss": 0.1555,
"step": 269
},
{
"epoch": 0.432,
"grad_norm": 0.6527994945120155,
"learning_rate": 9.943225148086846e-06,
"loss": 0.1798,
"step": 270
},
{
"epoch": 0.4336,
"grad_norm": 0.6730719264321817,
"learning_rate": 9.941817388373248e-06,
"loss": 0.2018,
"step": 271
},
{
"epoch": 0.4352,
"grad_norm": 0.5467052259203461,
"learning_rate": 9.940392490817124e-06,
"loss": 0.1678,
"step": 272
},
{
"epoch": 0.4368,
"grad_norm": 0.7227643242938706,
"learning_rate": 9.938950460359912e-06,
"loss": 0.1991,
"step": 273
},
{
"epoch": 0.4384,
"grad_norm": 0.699473602607064,
"learning_rate": 9.937491302002462e-06,
"loss": 0.1693,
"step": 274
},
{
"epoch": 0.44,
"grad_norm": 0.6817723927048626,
"learning_rate": 9.936015020805022e-06,
"loss": 0.174,
"step": 275
},
{
"epoch": 0.4416,
"grad_norm": 0.7276462665997606,
"learning_rate": 9.934521621887223e-06,
"loss": 0.1768,
"step": 276
},
{
"epoch": 0.4432,
"grad_norm": 0.6570919513555267,
"learning_rate": 9.933011110428058e-06,
"loss": 0.1811,
"step": 277
},
{
"epoch": 0.4448,
"grad_norm": 0.5752390318675986,
"learning_rate": 9.93148349166586e-06,
"loss": 0.1759,
"step": 278
},
{
"epoch": 0.4464,
"grad_norm": 0.6477771077266749,
"learning_rate": 9.929938770898299e-06,
"loss": 0.1621,
"step": 279
},
{
"epoch": 0.448,
"grad_norm": 0.7920319349855467,
"learning_rate": 9.928376953482343e-06,
"loss": 0.1735,
"step": 280
},
{
"epoch": 0.4496,
"grad_norm": 0.6859439840925039,
"learning_rate": 9.926798044834261e-06,
"loss": 0.1861,
"step": 281
},
{
"epoch": 0.4512,
"grad_norm": 0.6761151428742678,
"learning_rate": 9.92520205042958e-06,
"loss": 0.188,
"step": 282
},
{
"epoch": 0.4528,
"grad_norm": 0.590490022963266,
"learning_rate": 9.92358897580309e-06,
"loss": 0.1612,
"step": 283
},
{
"epoch": 0.4544,
"grad_norm": 0.58318911449323,
"learning_rate": 9.921958826548808e-06,
"loss": 0.1674,
"step": 284
},
{
"epoch": 0.456,
"grad_norm": 0.8092241887905155,
"learning_rate": 9.920311608319968e-06,
"loss": 0.1997,
"step": 285
},
{
"epoch": 0.4576,
"grad_norm": 0.6276358158214642,
"learning_rate": 9.918647326828993e-06,
"loss": 0.1466,
"step": 286
},
{
"epoch": 0.4592,
"grad_norm": 0.7095995627425923,
"learning_rate": 9.916965987847485e-06,
"loss": 0.1933,
"step": 287
},
{
"epoch": 0.4608,
"grad_norm": 0.867387529477996,
"learning_rate": 9.915267597206198e-06,
"loss": 0.222,
"step": 288
},
{
"epoch": 0.4624,
"grad_norm": 0.8376129915670598,
"learning_rate": 9.913552160795022e-06,
"loss": 0.2403,
"step": 289
},
{
"epoch": 0.464,
"grad_norm": 0.5590350389655118,
"learning_rate": 9.911819684562954e-06,
"loss": 0.1545,
"step": 290
},
{
"epoch": 0.4656,
"grad_norm": 0.7632342138710676,
"learning_rate": 9.910070174518093e-06,
"loss": 0.2013,
"step": 291
},
{
"epoch": 0.4672,
"grad_norm": 0.8503071641596915,
"learning_rate": 9.908303636727604e-06,
"loss": 0.1765,
"step": 292
},
{
"epoch": 0.4688,
"grad_norm": 0.7653306675434035,
"learning_rate": 9.9065200773177e-06,
"loss": 0.2066,
"step": 293
},
{
"epoch": 0.4704,
"grad_norm": 0.6057195996308924,
"learning_rate": 9.904719502473635e-06,
"loss": 0.1705,
"step": 294
},
{
"epoch": 0.472,
"grad_norm": 0.7126560934063338,
"learning_rate": 9.902901918439658e-06,
"loss": 0.1894,
"step": 295
},
{
"epoch": 0.4736,
"grad_norm": 0.6328164752234345,
"learning_rate": 9.901067331519013e-06,
"loss": 0.1717,
"step": 296
},
{
"epoch": 0.4752,
"grad_norm": 0.5548560034253891,
"learning_rate": 9.899215748073906e-06,
"loss": 0.1544,
"step": 297
},
{
"epoch": 0.4768,
"grad_norm": 0.7462833122750524,
"learning_rate": 9.897347174525487e-06,
"loss": 0.1969,
"step": 298
},
{
"epoch": 0.4784,
"grad_norm": 0.7339236581196977,
"learning_rate": 9.895461617353823e-06,
"loss": 0.1605,
"step": 299
},
{
"epoch": 0.48,
"grad_norm": 0.6626831103695787,
"learning_rate": 9.893559083097885e-06,
"loss": 0.1905,
"step": 300
},
{
"epoch": 0.4816,
"grad_norm": 0.7474187378820317,
"learning_rate": 9.891639578355511e-06,
"loss": 0.2043,
"step": 301
},
{
"epoch": 0.4832,
"grad_norm": 0.6002938008824829,
"learning_rate": 9.8897031097834e-06,
"loss": 0.1562,
"step": 302
},
{
"epoch": 0.4848,
"grad_norm": 0.6360125618534727,
"learning_rate": 9.887749684097072e-06,
"loss": 0.1523,
"step": 303
},
{
"epoch": 0.4864,
"grad_norm": 0.6692554311788963,
"learning_rate": 9.88577930807086e-06,
"loss": 0.1924,
"step": 304
},
{
"epoch": 0.488,
"grad_norm": 0.5360097067083763,
"learning_rate": 9.883791988537874e-06,
"loss": 0.162,
"step": 305
},
{
"epoch": 0.4896,
"grad_norm": 0.68738204988718,
"learning_rate": 9.881787732389987e-06,
"loss": 0.1993,
"step": 306
},
{
"epoch": 0.4912,
"grad_norm": 0.7445117466234226,
"learning_rate": 9.879766546577805e-06,
"loss": 0.1748,
"step": 307
},
{
"epoch": 0.4928,
"grad_norm": 0.5938100914950023,
"learning_rate": 9.877728438110645e-06,
"loss": 0.1483,
"step": 308
},
{
"epoch": 0.4944,
"grad_norm": 0.7434274659317546,
"learning_rate": 9.87567341405651e-06,
"loss": 0.1603,
"step": 309
},
{
"epoch": 0.496,
"grad_norm": 0.5680275286473279,
"learning_rate": 9.873601481542065e-06,
"loss": 0.1667,
"step": 310
},
{
"epoch": 0.4976,
"grad_norm": 0.6814283646816417,
"learning_rate": 9.871512647752612e-06,
"loss": 0.2075,
"step": 311
},
{
"epoch": 0.4992,
"grad_norm": 0.685459199643075,
"learning_rate": 9.86940691993207e-06,
"loss": 0.1782,
"step": 312
},
{
"epoch": 0.5008,
"grad_norm": 0.7077014460625858,
"learning_rate": 9.867284305382936e-06,
"loss": 0.1815,
"step": 313
},
{
"epoch": 0.5024,
"grad_norm": 0.5637225307080428,
"learning_rate": 9.865144811466275e-06,
"loss": 0.1496,
"step": 314
},
{
"epoch": 0.504,
"grad_norm": 0.8064409601364916,
"learning_rate": 9.86298844560169e-06,
"loss": 0.1944,
"step": 315
},
{
"epoch": 0.5056,
"grad_norm": 0.5785890880117247,
"learning_rate": 9.860815215267288e-06,
"loss": 0.1454,
"step": 316
},
{
"epoch": 0.5072,
"grad_norm": 0.8705402489018914,
"learning_rate": 9.858625127999668e-06,
"loss": 0.2158,
"step": 317
},
{
"epoch": 0.5088,
"grad_norm": 0.6870005018723719,
"learning_rate": 9.856418191393881e-06,
"loss": 0.2095,
"step": 318
},
{
"epoch": 0.5104,
"grad_norm": 0.6153661368789914,
"learning_rate": 9.854194413103418e-06,
"loss": 0.1753,
"step": 319
},
{
"epoch": 0.512,
"grad_norm": 0.7213751974464947,
"learning_rate": 9.851953800840166e-06,
"loss": 0.1885,
"step": 320
},
{
"epoch": 0.5136,
"grad_norm": 0.5995953018784986,
"learning_rate": 9.849696362374399e-06,
"loss": 0.1634,
"step": 321
},
{
"epoch": 0.5152,
"grad_norm": 0.6294118028685393,
"learning_rate": 9.847422105534739e-06,
"loss": 0.1688,
"step": 322
},
{
"epoch": 0.5168,
"grad_norm": 0.6152089659707293,
"learning_rate": 9.845131038208135e-06,
"loss": 0.1688,
"step": 323
},
{
"epoch": 0.5184,
"grad_norm": 0.5450381138637729,
"learning_rate": 9.84282316833983e-06,
"loss": 0.1709,
"step": 324
},
{
"epoch": 0.52,
"grad_norm": 0.659658179238135,
"learning_rate": 9.84049850393334e-06,
"loss": 0.1819,
"step": 325
},
{
"epoch": 0.5216,
"grad_norm": 0.6216923503792764,
"learning_rate": 9.838157053050423e-06,
"loss": 0.1714,
"step": 326
},
{
"epoch": 0.5232,
"grad_norm": 0.630005021776648,
"learning_rate": 9.83579882381105e-06,
"loss": 0.1658,
"step": 327
},
{
"epoch": 0.5248,
"grad_norm": 0.5738821729327643,
"learning_rate": 9.83342382439338e-06,
"loss": 0.1797,
"step": 328
},
{
"epoch": 0.5264,
"grad_norm": 0.6326540599060816,
"learning_rate": 9.831032063033726e-06,
"loss": 0.1804,
"step": 329
},
{
"epoch": 0.528,
"grad_norm": 0.7641984353744667,
"learning_rate": 9.828623548026533e-06,
"loss": 0.1971,
"step": 330
},
{
"epoch": 0.5296,
"grad_norm": 0.7222290428184442,
"learning_rate": 9.826198287724346e-06,
"loss": 0.1982,
"step": 331
},
{
"epoch": 0.5312,
"grad_norm": 0.5176888610353977,
"learning_rate": 9.823756290537783e-06,
"loss": 0.1688,
"step": 332
},
{
"epoch": 0.5328,
"grad_norm": 0.6168478124874316,
"learning_rate": 9.821297564935499e-06,
"loss": 0.1643,
"step": 333
},
{
"epoch": 0.5344,
"grad_norm": 0.6400813769095101,
"learning_rate": 9.81882211944417e-06,
"loss": 0.1722,
"step": 334
},
{
"epoch": 0.536,
"grad_norm": 0.6976773673330412,
"learning_rate": 9.816329962648444e-06,
"loss": 0.1737,
"step": 335
},
{
"epoch": 0.5376,
"grad_norm": 0.6497972501749004,
"learning_rate": 9.813821103190932e-06,
"loss": 0.1656,
"step": 336
},
{
"epoch": 0.5392,
"grad_norm": 0.7410700301130735,
"learning_rate": 9.811295549772169e-06,
"loss": 0.2319,
"step": 337
},
{
"epoch": 0.5408,
"grad_norm": 0.5942080498361239,
"learning_rate": 9.808753311150575e-06,
"loss": 0.1496,
"step": 338
},
{
"epoch": 0.5424,
"grad_norm": 0.6806878422997,
"learning_rate": 9.80619439614244e-06,
"loss": 0.1864,
"step": 339
},
{
"epoch": 0.544,
"grad_norm": 0.5402527241474873,
"learning_rate": 9.803618813621885e-06,
"loss": 0.1521,
"step": 340
},
{
"epoch": 0.5456,
"grad_norm": 0.6502572621978072,
"learning_rate": 9.801026572520832e-06,
"loss": 0.1808,
"step": 341
},
{
"epoch": 0.5472,
"grad_norm": 0.7195862448059469,
"learning_rate": 9.798417681828972e-06,
"loss": 0.1999,
"step": 342
},
{
"epoch": 0.5488,
"grad_norm": 0.6591725180453193,
"learning_rate": 9.795792150593739e-06,
"loss": 0.189,
"step": 343
},
{
"epoch": 0.5504,
"grad_norm": 0.5528283837962463,
"learning_rate": 9.793149987920273e-06,
"loss": 0.1681,
"step": 344
},
{
"epoch": 0.552,
"grad_norm": 0.8284315582025614,
"learning_rate": 9.79049120297139e-06,
"loss": 0.1872,
"step": 345
},
{
"epoch": 0.5536,
"grad_norm": 0.6512745935569575,
"learning_rate": 9.787815804967554e-06,
"loss": 0.179,
"step": 346
},
{
"epoch": 0.5552,
"grad_norm": 0.5977022657869455,
"learning_rate": 9.785123803186834e-06,
"loss": 0.1495,
"step": 347
},
{
"epoch": 0.5568,
"grad_norm": 0.5526892416852104,
"learning_rate": 9.782415206964892e-06,
"loss": 0.1749,
"step": 348
},
{
"epoch": 0.5584,
"grad_norm": 0.607668221802708,
"learning_rate": 9.779690025694926e-06,
"loss": 0.1707,
"step": 349
},
{
"epoch": 0.56,
"grad_norm": 0.7338216851084096,
"learning_rate": 9.776948268827658e-06,
"loss": 0.1834,
"step": 350
},
{
"epoch": 0.5616,
"grad_norm": 0.7414124017125461,
"learning_rate": 9.77418994587129e-06,
"loss": 0.1926,
"step": 351
},
{
"epoch": 0.5632,
"grad_norm": 0.6440217013909703,
"learning_rate": 9.771415066391473e-06,
"loss": 0.1865,
"step": 352
},
{
"epoch": 0.5648,
"grad_norm": 0.5970032306180105,
"learning_rate": 9.768623640011272e-06,
"loss": 0.1652,
"step": 353
},
{
"epoch": 0.5664,
"grad_norm": 0.7114309470243864,
"learning_rate": 9.765815676411145e-06,
"loss": 0.1786,
"step": 354
},
{
"epoch": 0.568,
"grad_norm": 0.6483332745110548,
"learning_rate": 9.762991185328891e-06,
"loss": 0.204,
"step": 355
},
{
"epoch": 0.5696,
"grad_norm": 0.7196681305366737,
"learning_rate": 9.760150176559627e-06,
"loss": 0.1824,
"step": 356
},
{
"epoch": 0.5712,
"grad_norm": 0.6893613255796193,
"learning_rate": 9.757292659955755e-06,
"loss": 0.1878,
"step": 357
},
{
"epoch": 0.5728,
"grad_norm": 0.6585024404535449,
"learning_rate": 9.754418645426919e-06,
"loss": 0.1745,
"step": 358
},
{
"epoch": 0.5744,
"grad_norm": 0.5350822874778702,
"learning_rate": 9.751528142939986e-06,
"loss": 0.1851,
"step": 359
},
{
"epoch": 0.576,
"grad_norm": 0.6178855060828767,
"learning_rate": 9.74862116251899e-06,
"loss": 0.1727,
"step": 360
},
{
"epoch": 0.5776,
"grad_norm": 0.5988641525126333,
"learning_rate": 9.74569771424512e-06,
"loss": 0.1767,
"step": 361
},
{
"epoch": 0.5792,
"grad_norm": 0.6218224309688154,
"learning_rate": 9.742757808256667e-06,
"loss": 0.1919,
"step": 362
},
{
"epoch": 0.5808,
"grad_norm": 0.8126211330417739,
"learning_rate": 9.739801454749e-06,
"loss": 0.1877,
"step": 363
},
{
"epoch": 0.5824,
"grad_norm": 0.7058091825723098,
"learning_rate": 9.736828663974527e-06,
"loss": 0.2,
"step": 364
},
{
"epoch": 0.584,
"grad_norm": 0.5758580039198674,
"learning_rate": 9.733839446242655e-06,
"loss": 0.182,
"step": 365
},
{
"epoch": 0.5856,
"grad_norm": 0.6233156344510485,
"learning_rate": 9.730833811919763e-06,
"loss": 0.1556,
"step": 366
},
{
"epoch": 0.5872,
"grad_norm": 0.6957936152115977,
"learning_rate": 9.727811771429158e-06,
"loss": 0.2041,
"step": 367
},
{
"epoch": 0.5888,
"grad_norm": 0.5095328627893095,
"learning_rate": 9.724773335251046e-06,
"loss": 0.1762,
"step": 368
},
{
"epoch": 0.5904,
"grad_norm": 0.7830290900416893,
"learning_rate": 9.721718513922488e-06,
"loss": 0.1782,
"step": 369
},
{
"epoch": 0.592,
"grad_norm": 0.7340135011511445,
"learning_rate": 9.71864731803737e-06,
"loss": 0.1871,
"step": 370
},
{
"epoch": 0.5936,
"grad_norm": 0.710246175068086,
"learning_rate": 9.715559758246363e-06,
"loss": 0.1918,
"step": 371
},
{
"epoch": 0.5952,
"grad_norm": 0.7143691197093163,
"learning_rate": 9.712455845256888e-06,
"loss": 0.2041,
"step": 372
},
{
"epoch": 0.5968,
"grad_norm": 0.7274655453031303,
"learning_rate": 9.709335589833076e-06,
"loss": 0.1664,
"step": 373
},
{
"epoch": 0.5984,
"grad_norm": 0.6933670057556707,
"learning_rate": 9.70619900279573e-06,
"loss": 0.2134,
"step": 374
},
{
"epoch": 0.6,
"grad_norm": 0.6244287659377605,
"learning_rate": 9.703046095022297e-06,
"loss": 0.1409,
"step": 375
},
{
"epoch": 0.6016,
"grad_norm": 0.6436939607476957,
"learning_rate": 9.699876877446815e-06,
"loss": 0.173,
"step": 376
},
{
"epoch": 0.6032,
"grad_norm": 0.9128550663863582,
"learning_rate": 9.696691361059886e-06,
"loss": 0.1528,
"step": 377
},
{
"epoch": 0.6048,
"grad_norm": 0.6790082900308524,
"learning_rate": 9.693489556908641e-06,
"loss": 0.1896,
"step": 378
},
{
"epoch": 0.6064,
"grad_norm": 0.6250896139768608,
"learning_rate": 9.690271476096686e-06,
"loss": 0.1602,
"step": 379
},
{
"epoch": 0.608,
"grad_norm": 0.7981317385982665,
"learning_rate": 9.68703712978408e-06,
"loss": 0.1743,
"step": 380
},
{
"epoch": 0.6096,
"grad_norm": 0.7876408359294238,
"learning_rate": 9.683786529187287e-06,
"loss": 0.1816,
"step": 381
},
{
"epoch": 0.6112,
"grad_norm": 0.6426931435309722,
"learning_rate": 9.680519685579137e-06,
"loss": 0.1815,
"step": 382
},
{
"epoch": 0.6128,
"grad_norm": 0.6398118115733524,
"learning_rate": 9.677236610288797e-06,
"loss": 0.1751,
"step": 383
},
{
"epoch": 0.6144,
"grad_norm": 0.5832965516597816,
"learning_rate": 9.673937314701714e-06,
"loss": 0.1657,
"step": 384
},
{
"epoch": 0.616,
"grad_norm": 0.7397250944697161,
"learning_rate": 9.670621810259596e-06,
"loss": 0.1837,
"step": 385
},
{
"epoch": 0.6176,
"grad_norm": 0.5919246567645863,
"learning_rate": 9.667290108460354e-06,
"loss": 0.1695,
"step": 386
},
{
"epoch": 0.6192,
"grad_norm": 0.6347478941981485,
"learning_rate": 9.663942220858075e-06,
"loss": 0.1831,
"step": 387
},
{
"epoch": 0.6208,
"grad_norm": 0.5976291586694785,
"learning_rate": 9.660578159062977e-06,
"loss": 0.1709,
"step": 388
},
{
"epoch": 0.6224,
"grad_norm": 0.5759455355417892,
"learning_rate": 9.657197934741366e-06,
"loss": 0.1542,
"step": 389
},
{
"epoch": 0.624,
"grad_norm": 0.65595037788509,
"learning_rate": 9.6538015596156e-06,
"loss": 0.1957,
"step": 390
},
{
"epoch": 0.6256,
"grad_norm": 0.7908438301549425,
"learning_rate": 9.650389045464046e-06,
"loss": 0.1899,
"step": 391
},
{
"epoch": 0.6272,
"grad_norm": 0.7686136073035917,
"learning_rate": 9.646960404121042e-06,
"loss": 0.2329,
"step": 392
},
{
"epoch": 0.6288,
"grad_norm": 0.6924889424616514,
"learning_rate": 9.643515647476851e-06,
"loss": 0.1498,
"step": 393
},
{
"epoch": 0.6304,
"grad_norm": 0.6233870309679708,
"learning_rate": 9.640054787477626e-06,
"loss": 0.1752,
"step": 394
},
{
"epoch": 0.632,
"grad_norm": 0.6542459866032585,
"learning_rate": 9.63657783612536e-06,
"loss": 0.1779,
"step": 395
},
{
"epoch": 0.6336,
"grad_norm": 0.6814798936306584,
"learning_rate": 9.633084805477857e-06,
"loss": 0.192,
"step": 396
},
{
"epoch": 0.6352,
"grad_norm": 0.626379136759705,
"learning_rate": 9.629575707648675e-06,
"loss": 0.1843,
"step": 397
},
{
"epoch": 0.6368,
"grad_norm": 0.594328574633082,
"learning_rate": 9.626050554807096e-06,
"loss": 0.1455,
"step": 398
},
{
"epoch": 0.6384,
"grad_norm": 0.6623352428467473,
"learning_rate": 9.62250935917808e-06,
"loss": 0.1578,
"step": 399
},
{
"epoch": 0.64,
"grad_norm": 0.6391534435541298,
"learning_rate": 9.618952133042223e-06,
"loss": 0.1515,
"step": 400
},
{
"epoch": 0.6416,
"grad_norm": 0.563069765952602,
"learning_rate": 9.615378888735706e-06,
"loss": 0.196,
"step": 401
},
{
"epoch": 0.6432,
"grad_norm": 0.7331340420064142,
"learning_rate": 9.611789638650269e-06,
"loss": 0.2078,
"step": 402
},
{
"epoch": 0.6448,
"grad_norm": 0.6078978103332717,
"learning_rate": 9.608184395233156e-06,
"loss": 0.1982,
"step": 403
},
{
"epoch": 0.6464,
"grad_norm": 0.624260378139288,
"learning_rate": 9.604563170987072e-06,
"loss": 0.1762,
"step": 404
},
{
"epoch": 0.648,
"grad_norm": 0.7232039988465814,
"learning_rate": 9.600925978470143e-06,
"loss": 0.1726,
"step": 405
},
{
"epoch": 0.6496,
"grad_norm": 0.5806952582701647,
"learning_rate": 9.597272830295877e-06,
"loss": 0.1618,
"step": 406
},
{
"epoch": 0.6512,
"grad_norm": 0.6618461091601451,
"learning_rate": 9.593603739133105e-06,
"loss": 0.18,
"step": 407
},
{
"epoch": 0.6528,
"grad_norm": 0.6985536231168606,
"learning_rate": 9.589918717705957e-06,
"loss": 0.1473,
"step": 408
},
{
"epoch": 0.6544,
"grad_norm": 0.5692572841672839,
"learning_rate": 9.586217778793804e-06,
"loss": 0.1483,
"step": 409
},
{
"epoch": 0.656,
"grad_norm": 0.5797177064700132,
"learning_rate": 9.582500935231215e-06,
"loss": 0.1528,
"step": 410
},
{
"epoch": 0.6576,
"grad_norm": 0.6057722592657329,
"learning_rate": 9.57876819990792e-06,
"loss": 0.178,
"step": 411
},
{
"epoch": 0.6592,
"grad_norm": 0.5168924303004003,
"learning_rate": 9.575019585768758e-06,
"loss": 0.1318,
"step": 412
},
{
"epoch": 0.6608,
"grad_norm": 0.6363905653069376,
"learning_rate": 9.571255105813632e-06,
"loss": 0.1957,
"step": 413
},
{
"epoch": 0.6624,
"grad_norm": 0.7023600609250216,
"learning_rate": 9.567474773097469e-06,
"loss": 0.2215,
"step": 414
},
{
"epoch": 0.664,
"grad_norm": 0.9955727364284319,
"learning_rate": 9.563678600730175e-06,
"loss": 0.1622,
"step": 415
},
{
"epoch": 0.6656,
"grad_norm": 0.6554547134502959,
"learning_rate": 9.559866601876581e-06,
"loss": 0.1749,
"step": 416
},
{
"epoch": 0.6672,
"grad_norm": 0.5576684075103858,
"learning_rate": 9.556038789756407e-06,
"loss": 0.1348,
"step": 417
},
{
"epoch": 0.6688,
"grad_norm": 0.6339910625439182,
"learning_rate": 9.55219517764421e-06,
"loss": 0.1866,
"step": 418
},
{
"epoch": 0.6704,
"grad_norm": 0.5826760206698064,
"learning_rate": 9.548335778869342e-06,
"loss": 0.1708,
"step": 419
},
{
"epoch": 0.672,
"grad_norm": 0.5406388806994605,
"learning_rate": 9.544460606815901e-06,
"loss": 0.1768,
"step": 420
},
{
"epoch": 0.6736,
"grad_norm": 0.6095726860692885,
"learning_rate": 9.540569674922685e-06,
"loss": 0.1639,
"step": 421
},
{
"epoch": 0.6752,
"grad_norm": 0.6040137189806664,
"learning_rate": 9.536662996683146e-06,
"loss": 0.1805,
"step": 422
},
{
"epoch": 0.6768,
"grad_norm": 0.6408628255417294,
"learning_rate": 9.532740585645346e-06,
"loss": 0.1696,
"step": 423
},
{
"epoch": 0.6784,
"grad_norm": 0.662234457464977,
"learning_rate": 9.528802455411902e-06,
"loss": 0.1522,
"step": 424
},
{
"epoch": 0.68,
"grad_norm": 0.6538725738243265,
"learning_rate": 9.52484861963995e-06,
"loss": 0.1767,
"step": 425
},
{
"epoch": 0.6816,
"grad_norm": 0.7034571124072382,
"learning_rate": 9.520879092041085e-06,
"loss": 0.1743,
"step": 426
},
{
"epoch": 0.6832,
"grad_norm": 0.6678919334011539,
"learning_rate": 9.516893886381324e-06,
"loss": 0.2079,
"step": 427
},
{
"epoch": 0.6848,
"grad_norm": 0.5640480252090954,
"learning_rate": 9.512893016481053e-06,
"loss": 0.1641,
"step": 428
},
{
"epoch": 0.6864,
"grad_norm": 0.5914103932950491,
"learning_rate": 9.508876496214983e-06,
"loss": 0.1564,
"step": 429
},
{
"epoch": 0.688,
"grad_norm": 2.5351641552422133,
"learning_rate": 9.504844339512096e-06,
"loss": 0.1722,
"step": 430
},
{
"epoch": 0.6896,
"grad_norm": 0.7666833499860748,
"learning_rate": 9.500796560355603e-06,
"loss": 0.1377,
"step": 431
},
{
"epoch": 0.6912,
"grad_norm": 0.6304032618573835,
"learning_rate": 9.496733172782889e-06,
"loss": 0.1899,
"step": 432
},
{
"epoch": 0.6928,
"grad_norm": 0.5888485510557646,
"learning_rate": 9.492654190885469e-06,
"loss": 0.1794,
"step": 433
},
{
"epoch": 0.6944,
"grad_norm": 0.5494152646847476,
"learning_rate": 9.488559628808939e-06,
"loss": 0.1699,
"step": 434
},
{
"epoch": 0.696,
"grad_norm": 0.5705982688366871,
"learning_rate": 9.484449500752927e-06,
"loss": 0.1537,
"step": 435
},
{
"epoch": 0.6976,
"grad_norm": 0.5710787800190167,
"learning_rate": 9.480323820971039e-06,
"loss": 0.1611,
"step": 436
},
{
"epoch": 0.6992,
"grad_norm": 0.634798085327612,
"learning_rate": 9.476182603770814e-06,
"loss": 0.1609,
"step": 437
},
{
"epoch": 0.7008,
"grad_norm": 0.6216634762921098,
"learning_rate": 9.472025863513676e-06,
"loss": 0.2001,
"step": 438
},
{
"epoch": 0.7024,
"grad_norm": 0.6432455592856032,
"learning_rate": 9.467853614614883e-06,
"loss": 0.1821,
"step": 439
},
{
"epoch": 0.704,
"grad_norm": 0.4926302733350555,
"learning_rate": 9.46366587154347e-06,
"loss": 0.1674,
"step": 440
},
{
"epoch": 0.7056,
"grad_norm": 0.5872530489683253,
"learning_rate": 9.459462648822209e-06,
"loss": 0.1491,
"step": 441
},
{
"epoch": 0.7072,
"grad_norm": 0.6799158572655349,
"learning_rate": 9.45524396102755e-06,
"loss": 0.1555,
"step": 442
},
{
"epoch": 0.7088,
"grad_norm": 0.5217004302615952,
"learning_rate": 9.451009822789583e-06,
"loss": 0.1681,
"step": 443
},
{
"epoch": 0.7104,
"grad_norm": 0.5819520653016731,
"learning_rate": 9.44676024879197e-06,
"loss": 0.1725,
"step": 444
},
{
"epoch": 0.712,
"grad_norm": 0.5898354688607719,
"learning_rate": 9.442495253771909e-06,
"loss": 0.1582,
"step": 445
},
{
"epoch": 0.7136,
"grad_norm": 0.5832917468463423,
"learning_rate": 9.438214852520073e-06,
"loss": 0.1988,
"step": 446
},
{
"epoch": 0.7152,
"grad_norm": 0.5847046395515987,
"learning_rate": 9.433919059880564e-06,
"loss": 0.187,
"step": 447
},
{
"epoch": 0.7168,
"grad_norm": 0.6574464356998375,
"learning_rate": 9.429607890750863e-06,
"loss": 0.1809,
"step": 448
},
{
"epoch": 0.7184,
"grad_norm": 0.6144314231164872,
"learning_rate": 9.425281360081769e-06,
"loss": 0.1541,
"step": 449
},
{
"epoch": 0.72,
"grad_norm": 0.686513272576765,
"learning_rate": 9.420939482877359e-06,
"loss": 0.1736,
"step": 450
},
{
"epoch": 0.7216,
"grad_norm": 0.6601593786828654,
"learning_rate": 9.416582274194929e-06,
"loss": 0.2177,
"step": 451
},
{
"epoch": 0.7232,
"grad_norm": 0.5702379301676516,
"learning_rate": 9.412209749144947e-06,
"loss": 0.1569,
"step": 452
},
{
"epoch": 0.7248,
"grad_norm": 0.7023987794127036,
"learning_rate": 9.40782192289099e-06,
"loss": 0.2094,
"step": 453
},
{
"epoch": 0.7264,
"grad_norm": 0.6381639840923098,
"learning_rate": 9.4034188106497e-06,
"loss": 0.1903,
"step": 454
},
{
"epoch": 0.728,
"grad_norm": 0.6534919553717288,
"learning_rate": 9.399000427690736e-06,
"loss": 0.1872,
"step": 455
},
{
"epoch": 0.7296,
"grad_norm": 0.6472495006294313,
"learning_rate": 9.394566789336707e-06,
"loss": 0.1797,
"step": 456
},
{
"epoch": 0.7312,
"grad_norm": 0.685371616316373,
"learning_rate": 9.390117910963132e-06,
"loss": 0.185,
"step": 457
},
{
"epoch": 0.7328,
"grad_norm": 0.6258526512938506,
"learning_rate": 9.385653807998376e-06,
"loss": 0.1686,
"step": 458
},
{
"epoch": 0.7344,
"grad_norm": 0.5510296964462884,
"learning_rate": 9.381174495923608e-06,
"loss": 0.179,
"step": 459
},
{
"epoch": 0.736,
"grad_norm": 0.6678105314771469,
"learning_rate": 9.376679990272736e-06,
"loss": 0.1637,
"step": 460
},
{
"epoch": 0.7376,
"grad_norm": 0.6295338592013524,
"learning_rate": 9.37217030663236e-06,
"loss": 0.2125,
"step": 461
},
{
"epoch": 0.7392,
"grad_norm": 0.7188802602256498,
"learning_rate": 9.367645460641716e-06,
"loss": 0.1918,
"step": 462
},
{
"epoch": 0.7408,
"grad_norm": 0.6063638728540184,
"learning_rate": 9.36310546799262e-06,
"loss": 0.1761,
"step": 463
},
{
"epoch": 0.7424,
"grad_norm": 0.5381982381980303,
"learning_rate": 9.358550344429421e-06,
"loss": 0.1532,
"step": 464
},
{
"epoch": 0.744,
"grad_norm": 0.6527849257416872,
"learning_rate": 9.353980105748934e-06,
"loss": 0.1866,
"step": 465
},
{
"epoch": 0.7456,
"grad_norm": 0.533284139960733,
"learning_rate": 9.349394767800397e-06,
"loss": 0.1588,
"step": 466
},
{
"epoch": 0.7472,
"grad_norm": 0.6335925071705951,
"learning_rate": 9.344794346485408e-06,
"loss": 0.1743,
"step": 467
},
{
"epoch": 0.7488,
"grad_norm": 0.588466297789812,
"learning_rate": 9.340178857757876e-06,
"loss": 0.1721,
"step": 468
},
{
"epoch": 0.7504,
"grad_norm": 0.7309027668707115,
"learning_rate": 9.335548317623957e-06,
"loss": 0.2033,
"step": 469
},
{
"epoch": 0.752,
"grad_norm": 0.6880738949151308,
"learning_rate": 9.330902742142013e-06,
"loss": 0.1699,
"step": 470
},
{
"epoch": 0.7536,
"grad_norm": 0.7528677197123981,
"learning_rate": 9.326242147422538e-06,
"loss": 0.2126,
"step": 471
},
{
"epoch": 0.7552,
"grad_norm": 0.5557708649965533,
"learning_rate": 9.321566549628118e-06,
"loss": 0.1622,
"step": 472
},
{
"epoch": 0.7568,
"grad_norm": 0.678237308627265,
"learning_rate": 9.316875964973366e-06,
"loss": 0.1999,
"step": 473
},
{
"epoch": 0.7584,
"grad_norm": 0.7480113954026755,
"learning_rate": 9.31217040972487e-06,
"loss": 0.1904,
"step": 474
},
{
"epoch": 0.76,
"grad_norm": 0.6729785862304761,
"learning_rate": 9.307449900201132e-06,
"loss": 0.2023,
"step": 475
},
{
"epoch": 0.7616,
"grad_norm": 0.656714305465471,
"learning_rate": 9.302714452772515e-06,
"loss": 0.1816,
"step": 476
},
{
"epoch": 0.7632,
"grad_norm": 0.5925014351109619,
"learning_rate": 9.29796408386119e-06,
"loss": 0.1378,
"step": 477
},
{
"epoch": 0.7648,
"grad_norm": 0.5525732379081186,
"learning_rate": 9.293198809941067e-06,
"loss": 0.1908,
"step": 478
},
{
"epoch": 0.7664,
"grad_norm": 0.5638262385914303,
"learning_rate": 9.288418647537752e-06,
"loss": 0.1541,
"step": 479
},
{
"epoch": 0.768,
"grad_norm": 0.601339449329617,
"learning_rate": 9.283623613228479e-06,
"loss": 0.1906,
"step": 480
},
{
"epoch": 0.7696,
"grad_norm": 0.6352818967186328,
"learning_rate": 9.27881372364206e-06,
"loss": 0.1767,
"step": 481
},
{
"epoch": 0.7712,
"grad_norm": 0.6322906198483522,
"learning_rate": 9.27398899545882e-06,
"loss": 0.1655,
"step": 482
},
{
"epoch": 0.7728,
"grad_norm": 0.7204786935021317,
"learning_rate": 9.269149445410545e-06,
"loss": 0.1849,
"step": 483
},
{
"epoch": 0.7744,
"grad_norm": 0.595779370733987,
"learning_rate": 9.264295090280424e-06,
"loss": 0.1888,
"step": 484
},
{
"epoch": 0.776,
"grad_norm": 0.6023200510901364,
"learning_rate": 9.259425946902987e-06,
"loss": 0.1533,
"step": 485
},
{
"epoch": 0.7776,
"grad_norm": 1.0278113910704392,
"learning_rate": 9.254542032164047e-06,
"loss": 0.1727,
"step": 486
},
{
"epoch": 0.7792,
"grad_norm": 0.769793062743826,
"learning_rate": 9.249643363000645e-06,
"loss": 0.203,
"step": 487
},
{
"epoch": 0.7808,
"grad_norm": 0.5931529781151321,
"learning_rate": 9.24472995640099e-06,
"loss": 0.1685,
"step": 488
},
{
"epoch": 0.7824,
"grad_norm": 0.6855430435649229,
"learning_rate": 9.239801829404396e-06,
"loss": 0.1892,
"step": 489
},
{
"epoch": 0.784,
"grad_norm": 0.6547482174999337,
"learning_rate": 9.234858999101232e-06,
"loss": 0.1796,
"step": 490
},
{
"epoch": 0.7856,
"grad_norm": 0.6050031954470846,
"learning_rate": 9.22990148263285e-06,
"loss": 0.1965,
"step": 491
},
{
"epoch": 0.7872,
"grad_norm": 0.7116037902401874,
"learning_rate": 9.224929297191536e-06,
"loss": 0.2087,
"step": 492
},
{
"epoch": 0.7888,
"grad_norm": 0.5565166201964887,
"learning_rate": 9.219942460020447e-06,
"loss": 0.1555,
"step": 493
},
{
"epoch": 0.7904,
"grad_norm": 0.5697772021685539,
"learning_rate": 9.214940988413552e-06,
"loss": 0.2109,
"step": 494
},
{
"epoch": 0.792,
"grad_norm": 0.641495712050076,
"learning_rate": 9.20992489971557e-06,
"loss": 0.1726,
"step": 495
},
{
"epoch": 0.7936,
"grad_norm": 0.9427395842275005,
"learning_rate": 9.204894211321906e-06,
"loss": 0.1426,
"step": 496
},
{
"epoch": 0.7952,
"grad_norm": 0.6564369697644002,
"learning_rate": 9.199848940678607e-06,
"loss": 0.1517,
"step": 497
},
{
"epoch": 0.7968,
"grad_norm": 0.6741401433648226,
"learning_rate": 9.194789105282277e-06,
"loss": 0.1861,
"step": 498
},
{
"epoch": 0.7984,
"grad_norm": 0.6546792940980187,
"learning_rate": 9.189714722680041e-06,
"loss": 0.2103,
"step": 499
},
{
"epoch": 0.8,
"grad_norm": 0.6368229743060805,
"learning_rate": 9.184625810469468e-06,
"loss": 0.1704,
"step": 500
},
{
"epoch": 0.8016,
"grad_norm": 0.5966489859449926,
"learning_rate": 9.179522386298508e-06,
"loss": 0.16,
"step": 501
},
{
"epoch": 0.8032,
"grad_norm": 0.5701757278982061,
"learning_rate": 9.174404467865447e-06,
"loss": 0.1615,
"step": 502
},
{
"epoch": 0.8048,
"grad_norm": 0.6572105119983995,
"learning_rate": 9.169272072918834e-06,
"loss": 0.1536,
"step": 503
},
{
"epoch": 0.8064,
"grad_norm": 0.724287082682101,
"learning_rate": 9.164125219257419e-06,
"loss": 0.1911,
"step": 504
},
{
"epoch": 0.808,
"grad_norm": 0.6019097912167761,
"learning_rate": 9.158963924730092e-06,
"loss": 0.1584,
"step": 505
},
{
"epoch": 0.8096,
"grad_norm": 0.6613360690192642,
"learning_rate": 9.153788207235827e-06,
"loss": 0.1769,
"step": 506
},
{
"epoch": 0.8112,
"grad_norm": 0.5491689672344781,
"learning_rate": 9.148598084723615e-06,
"loss": 0.1815,
"step": 507
},
{
"epoch": 0.8128,
"grad_norm": 0.6897642874622748,
"learning_rate": 9.143393575192402e-06,
"loss": 0.1762,
"step": 508
},
{
"epoch": 0.8144,
"grad_norm": 0.6272256990685852,
"learning_rate": 9.138174696691025e-06,
"loss": 0.1771,
"step": 509
},
{
"epoch": 0.816,
"grad_norm": 0.5635112742337389,
"learning_rate": 9.132941467318152e-06,
"loss": 0.1693,
"step": 510
},
{
"epoch": 0.8176,
"grad_norm": 0.6061721744180999,
"learning_rate": 9.127693905222223e-06,
"loss": 0.1686,
"step": 511
},
{
"epoch": 0.8192,
"grad_norm": 0.5233228290866191,
"learning_rate": 9.122432028601377e-06,
"loss": 0.1924,
"step": 512
},
{
"epoch": 0.8208,
"grad_norm": 0.44540972145091623,
"learning_rate": 9.1171558557034e-06,
"loss": 0.1777,
"step": 513
},
{
"epoch": 0.8224,
"grad_norm": 0.736602336603636,
"learning_rate": 9.111865404825652e-06,
"loss": 0.1837,
"step": 514
},
{
"epoch": 0.824,
"grad_norm": 0.6861888117796532,
"learning_rate": 9.10656069431501e-06,
"loss": 0.1703,
"step": 515
},
{
"epoch": 0.8256,
"grad_norm": 0.5530906032695171,
"learning_rate": 9.101241742567802e-06,
"loss": 0.1589,
"step": 516
},
{
"epoch": 0.8272,
"grad_norm": 0.7051477707023113,
"learning_rate": 9.095908568029741e-06,
"loss": 0.2133,
"step": 517
},
{
"epoch": 0.8288,
"grad_norm": 0.5315136636978075,
"learning_rate": 9.09056118919587e-06,
"loss": 0.1709,
"step": 518
},
{
"epoch": 0.8304,
"grad_norm": 0.5942664473404264,
"learning_rate": 9.085199624610486e-06,
"loss": 0.186,
"step": 519
},
{
"epoch": 0.832,
"grad_norm": 0.6703181769698968,
"learning_rate": 9.079823892867083e-06,
"loss": 0.1749,
"step": 520
},
{
"epoch": 0.8336,
"grad_norm": 0.5644359640048555,
"learning_rate": 9.074434012608282e-06,
"loss": 0.1492,
"step": 521
},
{
"epoch": 0.8352,
"grad_norm": 0.5851826516678555,
"learning_rate": 9.069030002525777e-06,
"loss": 0.1535,
"step": 522
},
{
"epoch": 0.8368,
"grad_norm": 0.844038649379113,
"learning_rate": 9.063611881360258e-06,
"loss": 0.168,
"step": 523
},
{
"epoch": 0.8384,
"grad_norm": 0.8697091130024973,
"learning_rate": 9.05817966790135e-06,
"loss": 0.2165,
"step": 524
},
{
"epoch": 0.84,
"grad_norm": 0.5188460853273824,
"learning_rate": 9.052733380987555e-06,
"loss": 0.1501,
"step": 525
},
{
"epoch": 0.8416,
"grad_norm": 0.48801229169041654,
"learning_rate": 9.047273039506174e-06,
"loss": 0.1532,
"step": 526
},
{
"epoch": 0.8432,
"grad_norm": 0.5417158043030413,
"learning_rate": 9.041798662393255e-06,
"loss": 0.1467,
"step": 527
},
{
"epoch": 0.8448,
"grad_norm": 0.4400068556748351,
"learning_rate": 9.036310268633515e-06,
"loss": 0.1503,
"step": 528
},
{
"epoch": 0.8464,
"grad_norm": 0.6122832880799541,
"learning_rate": 9.030807877260278e-06,
"loss": 0.1936,
"step": 529
},
{
"epoch": 0.848,
"grad_norm": 0.5052101992731955,
"learning_rate": 9.025291507355419e-06,
"loss": 0.1537,
"step": 530
},
{
"epoch": 0.8496,
"grad_norm": 0.5192551747179266,
"learning_rate": 9.01976117804928e-06,
"loss": 0.1547,
"step": 531
},
{
"epoch": 0.8512,
"grad_norm": 0.49213393820731605,
"learning_rate": 9.014216908520619e-06,
"loss": 0.1553,
"step": 532
},
{
"epoch": 0.8528,
"grad_norm": 0.5071140102167218,
"learning_rate": 9.008658717996538e-06,
"loss": 0.1554,
"step": 533
},
{
"epoch": 0.8544,
"grad_norm": 0.6827933439312666,
"learning_rate": 9.003086625752414e-06,
"loss": 0.1932,
"step": 534
},
{
"epoch": 0.856,
"grad_norm": 0.5864360512922673,
"learning_rate": 8.997500651111833e-06,
"loss": 0.1716,
"step": 535
},
{
"epoch": 0.8576,
"grad_norm": 0.5944034789319267,
"learning_rate": 8.991900813446523e-06,
"loss": 0.1634,
"step": 536
},
{
"epoch": 0.8592,
"grad_norm": 0.5091240513478156,
"learning_rate": 8.986287132176295e-06,
"loss": 0.1685,
"step": 537
},
{
"epoch": 0.8608,
"grad_norm": 0.7494199102144861,
"learning_rate": 8.980659626768961e-06,
"loss": 0.1934,
"step": 538
},
{
"epoch": 0.8624,
"grad_norm": 0.7974878146648244,
"learning_rate": 8.975018316740278e-06,
"loss": 0.1911,
"step": 539
},
{
"epoch": 0.864,
"grad_norm": 0.5998400965825982,
"learning_rate": 8.969363221653875e-06,
"loss": 0.1823,
"step": 540
},
{
"epoch": 0.8656,
"grad_norm": 0.6585459265108777,
"learning_rate": 8.963694361121186e-06,
"loss": 0.17,
"step": 541
},
{
"epoch": 0.8672,
"grad_norm": 0.6261077154019918,
"learning_rate": 8.958011754801383e-06,
"loss": 0.1653,
"step": 542
},
{
"epoch": 0.8688,
"grad_norm": 0.6006776918263226,
"learning_rate": 8.952315422401307e-06,
"loss": 0.154,
"step": 543
},
{
"epoch": 0.8704,
"grad_norm": 0.5703728114864435,
"learning_rate": 8.946605383675403e-06,
"loss": 0.176,
"step": 544
},
{
"epoch": 0.872,
"grad_norm": 0.5322731571703057,
"learning_rate": 8.940881658425645e-06,
"loss": 0.1652,
"step": 545
},
{
"epoch": 0.8736,
"grad_norm": 0.5122205501487338,
"learning_rate": 8.93514426650147e-06,
"loss": 0.163,
"step": 546
},
{
"epoch": 0.8752,
"grad_norm": 0.7586848792086728,
"learning_rate": 8.929393227799715e-06,
"loss": 0.1972,
"step": 547
},
{
"epoch": 0.8768,
"grad_norm": 0.8346895892281893,
"learning_rate": 8.923628562264536e-06,
"loss": 0.1682,
"step": 548
},
{
"epoch": 0.8784,
"grad_norm": 0.6068018291774427,
"learning_rate": 8.917850289887353e-06,
"loss": 0.1785,
"step": 549
},
{
"epoch": 0.88,
"grad_norm": 0.5904092423937719,
"learning_rate": 8.91205843070677e-06,
"loss": 0.1937,
"step": 550
},
{
"epoch": 0.8816,
"grad_norm": 0.4562130293447332,
"learning_rate": 8.906253004808506e-06,
"loss": 0.1539,
"step": 551
},
{
"epoch": 0.8832,
"grad_norm": 0.4668657484984759,
"learning_rate": 8.900434032325332e-06,
"loss": 0.1609,
"step": 552
},
{
"epoch": 0.8848,
"grad_norm": 0.5608698910579404,
"learning_rate": 8.894601533437e-06,
"loss": 0.1599,
"step": 553
},
{
"epoch": 0.8864,
"grad_norm": 0.7505426795256622,
"learning_rate": 8.888755528370163e-06,
"loss": 0.2107,
"step": 554
},
{
"epoch": 0.888,
"grad_norm": 0.5486422743371283,
"learning_rate": 8.882896037398322e-06,
"loss": 0.1732,
"step": 555
},
{
"epoch": 0.8896,
"grad_norm": 0.7009410607796757,
"learning_rate": 8.877023080841739e-06,
"loss": 0.1779,
"step": 556
},
{
"epoch": 0.8912,
"grad_norm": 0.6454158742141469,
"learning_rate": 8.871136679067372e-06,
"loss": 0.1772,
"step": 557
},
{
"epoch": 0.8928,
"grad_norm": 0.5551002041392086,
"learning_rate": 8.865236852488813e-06,
"loss": 0.1653,
"step": 558
},
{
"epoch": 0.8944,
"grad_norm": 0.5403970849306158,
"learning_rate": 8.859323621566207e-06,
"loss": 0.1579,
"step": 559
},
{
"epoch": 0.896,
"grad_norm": 0.6128649217554931,
"learning_rate": 8.853397006806183e-06,
"loss": 0.1481,
"step": 560
},
{
"epoch": 0.8976,
"grad_norm": 0.665475459969274,
"learning_rate": 8.847457028761783e-06,
"loss": 0.1971,
"step": 561
},
{
"epoch": 0.8992,
"grad_norm": 0.6004953239776809,
"learning_rate": 8.841503708032398e-06,
"loss": 0.1491,
"step": 562
},
{
"epoch": 0.9008,
"grad_norm": 0.5914230592401661,
"learning_rate": 8.835537065263684e-06,
"loss": 0.1821,
"step": 563
},
{
"epoch": 0.9024,
"grad_norm": 0.5749955666836588,
"learning_rate": 8.829557121147499e-06,
"loss": 0.175,
"step": 564
},
{
"epoch": 0.904,
"grad_norm": 0.6239721634846342,
"learning_rate": 8.82356389642183e-06,
"loss": 0.1783,
"step": 565
},
{
"epoch": 0.9056,
"grad_norm": 0.5421733789957934,
"learning_rate": 8.817557411870717e-06,
"loss": 0.1672,
"step": 566
},
{
"epoch": 0.9072,
"grad_norm": 0.6062737124758995,
"learning_rate": 8.811537688324187e-06,
"loss": 0.1745,
"step": 567
},
{
"epoch": 0.9088,
"grad_norm": 0.6257024584050931,
"learning_rate": 8.805504746658183e-06,
"loss": 0.1525,
"step": 568
},
{
"epoch": 0.9104,
"grad_norm": 0.5075795451502729,
"learning_rate": 8.799458607794476e-06,
"loss": 0.1425,
"step": 569
},
{
"epoch": 0.912,
"grad_norm": 0.44624708990402145,
"learning_rate": 8.793399292700616e-06,
"loss": 0.142,
"step": 570
},
{
"epoch": 0.9136,
"grad_norm": 0.6388320641777617,
"learning_rate": 8.787326822389836e-06,
"loss": 0.1928,
"step": 571
},
{
"epoch": 0.9152,
"grad_norm": 0.6599001616999065,
"learning_rate": 8.781241217921e-06,
"loss": 0.175,
"step": 572
},
{
"epoch": 0.9168,
"grad_norm": 0.6623953017673031,
"learning_rate": 8.775142500398513e-06,
"loss": 0.2091,
"step": 573
},
{
"epoch": 0.9184,
"grad_norm": 0.6158010376098083,
"learning_rate": 8.769030690972262e-06,
"loss": 0.1569,
"step": 574
},
{
"epoch": 0.92,
"grad_norm": 0.6827048775451922,
"learning_rate": 8.76290581083753e-06,
"loss": 0.1867,
"step": 575
},
{
"epoch": 0.9216,
"grad_norm": 0.6019517066540375,
"learning_rate": 8.756767881234928e-06,
"loss": 0.1825,
"step": 576
},
{
"epoch": 0.9232,
"grad_norm": 0.659537013603221,
"learning_rate": 8.750616923450328e-06,
"loss": 0.1909,
"step": 577
},
{
"epoch": 0.9248,
"grad_norm": 0.6344392219230661,
"learning_rate": 8.744452958814775e-06,
"loss": 0.1776,
"step": 578
},
{
"epoch": 0.9264,
"grad_norm": 0.7571656346750517,
"learning_rate": 8.738276008704426e-06,
"loss": 0.1615,
"step": 579
},
{
"epoch": 0.928,
"grad_norm": 0.4547296468636667,
"learning_rate": 8.732086094540467e-06,
"loss": 0.1578,
"step": 580
},
{
"epoch": 0.9296,
"grad_norm": 0.6922972346413194,
"learning_rate": 8.725883237789046e-06,
"loss": 0.1889,
"step": 581
},
{
"epoch": 0.9312,
"grad_norm": 0.5060844731139387,
"learning_rate": 8.719667459961191e-06,
"loss": 0.1651,
"step": 582
},
{
"epoch": 0.9328,
"grad_norm": 0.5288442745980257,
"learning_rate": 8.713438782612743e-06,
"loss": 0.1826,
"step": 583
},
{
"epoch": 0.9344,
"grad_norm": 0.5825524252725381,
"learning_rate": 8.707197227344275e-06,
"loss": 0.1727,
"step": 584
},
{
"epoch": 0.936,
"grad_norm": 0.7344163773647593,
"learning_rate": 8.700942815801023e-06,
"loss": 0.1613,
"step": 585
},
{
"epoch": 0.9376,
"grad_norm": 0.6644691472505426,
"learning_rate": 8.6946755696728e-06,
"loss": 0.1799,
"step": 586
},
{
"epoch": 0.9392,
"grad_norm": 0.5871410218360695,
"learning_rate": 8.688395510693939e-06,
"loss": 0.1994,
"step": 587
},
{
"epoch": 0.9408,
"grad_norm": 0.7195765913474886,
"learning_rate": 8.682102660643196e-06,
"loss": 0.1902,
"step": 588
},
{
"epoch": 0.9424,
"grad_norm": 0.5180235746334956,
"learning_rate": 8.675797041343696e-06,
"loss": 0.1658,
"step": 589
},
{
"epoch": 0.944,
"grad_norm": 0.5867823086515722,
"learning_rate": 8.669478674662839e-06,
"loss": 0.1769,
"step": 590
},
{
"epoch": 0.9456,
"grad_norm": 0.565979625704949,
"learning_rate": 8.663147582512232e-06,
"loss": 0.1658,
"step": 591
},
{
"epoch": 0.9472,
"grad_norm": 0.5188310172570304,
"learning_rate": 8.65680378684762e-06,
"loss": 0.1669,
"step": 592
},
{
"epoch": 0.9488,
"grad_norm": 0.659095809585939,
"learning_rate": 8.6504473096688e-06,
"loss": 0.19,
"step": 593
},
{
"epoch": 0.9504,
"grad_norm": 0.5845162435392918,
"learning_rate": 8.64407817301954e-06,
"loss": 0.1894,
"step": 594
},
{
"epoch": 0.952,
"grad_norm": 0.6088727199715791,
"learning_rate": 8.637696398987517e-06,
"loss": 0.1808,
"step": 595
},
{
"epoch": 0.9536,
"grad_norm": 0.5873903314833873,
"learning_rate": 8.631302009704235e-06,
"loss": 0.1709,
"step": 596
},
{
"epoch": 0.9552,
"grad_norm": 0.5767520947977285,
"learning_rate": 8.624895027344943e-06,
"loss": 0.1699,
"step": 597
},
{
"epoch": 0.9568,
"grad_norm": 0.521565787320087,
"learning_rate": 8.618475474128563e-06,
"loss": 0.1651,
"step": 598
},
{
"epoch": 0.9584,
"grad_norm": 0.6767699363448209,
"learning_rate": 8.61204337231761e-06,
"loss": 0.1843,
"step": 599
},
{
"epoch": 0.96,
"grad_norm": 0.7269456774715906,
"learning_rate": 8.605598744218122e-06,
"loss": 0.1855,
"step": 600
},
{
"epoch": 0.9616,
"grad_norm": 0.5522334276002403,
"learning_rate": 8.599141612179572e-06,
"loss": 0.1611,
"step": 601
},
{
"epoch": 0.9632,
"grad_norm": 0.5551303422625761,
"learning_rate": 8.592671998594794e-06,
"loss": 0.1628,
"step": 602
},
{
"epoch": 0.9648,
"grad_norm": 0.5915089065372864,
"learning_rate": 8.586189925899913e-06,
"loss": 0.194,
"step": 603
},
{
"epoch": 0.9664,
"grad_norm": 0.6691761361907202,
"learning_rate": 8.57969541657426e-06,
"loss": 0.2066,
"step": 604
},
{
"epoch": 0.968,
"grad_norm": 0.5513075924107017,
"learning_rate": 8.57318849314029e-06,
"loss": 0.1763,
"step": 605
},
{
"epoch": 0.9696,
"grad_norm": 0.47749823840789024,
"learning_rate": 8.566669178163513e-06,
"loss": 0.172,
"step": 606
},
{
"epoch": 0.9712,
"grad_norm": 0.6083515006679234,
"learning_rate": 8.560137494252416e-06,
"loss": 0.2038,
"step": 607
},
{
"epoch": 0.9728,
"grad_norm": 0.5524391811017402,
"learning_rate": 8.553593464058374e-06,
"loss": 0.162,
"step": 608
},
{
"epoch": 0.9744,
"grad_norm": 0.6273363800912689,
"learning_rate": 8.54703711027558e-06,
"loss": 0.1848,
"step": 609
},
{
"epoch": 0.976,
"grad_norm": 0.563056499001572,
"learning_rate": 8.540468455640964e-06,
"loss": 0.1587,
"step": 610
},
{
"epoch": 0.9776,
"grad_norm": 0.6847750981353311,
"learning_rate": 8.533887522934114e-06,
"loss": 0.1675,
"step": 611
},
{
"epoch": 0.9792,
"grad_norm": 0.5742173459852481,
"learning_rate": 8.527294334977201e-06,
"loss": 0.1624,
"step": 612
},
{
"epoch": 0.9808,
"grad_norm": 0.6637993249088606,
"learning_rate": 8.520688914634894e-06,
"loss": 0.1803,
"step": 613
},
{
"epoch": 0.9824,
"grad_norm": 0.4369590177773895,
"learning_rate": 8.51407128481428e-06,
"loss": 0.1525,
"step": 614
},
{
"epoch": 0.984,
"grad_norm": 0.7532630872446077,
"learning_rate": 8.507441468464792e-06,
"loss": 0.2135,
"step": 615
},
{
"epoch": 0.9856,
"grad_norm": 0.564427516392782,
"learning_rate": 8.50079948857812e-06,
"loss": 0.1843,
"step": 616
},
{
"epoch": 0.9872,
"grad_norm": 0.5638136132536073,
"learning_rate": 8.494145368188143e-06,
"loss": 0.1858,
"step": 617
},
{
"epoch": 0.9888,
"grad_norm": 0.5458252617750412,
"learning_rate": 8.487479130370838e-06,
"loss": 0.1702,
"step": 618
},
{
"epoch": 0.9904,
"grad_norm": 0.5478253457855483,
"learning_rate": 8.480800798244202e-06,
"loss": 0.1617,
"step": 619
},
{
"epoch": 0.992,
"grad_norm": 0.6549597329864674,
"learning_rate": 8.47411039496818e-06,
"loss": 0.2115,
"step": 620
},
{
"epoch": 0.9936,
"grad_norm": 0.6231575958677863,
"learning_rate": 8.467407943744574e-06,
"loss": 0.1666,
"step": 621
},
{
"epoch": 0.9952,
"grad_norm": 0.6445917115081824,
"learning_rate": 8.460693467816972e-06,
"loss": 0.2234,
"step": 622
},
{
"epoch": 0.9968,
"grad_norm": 0.6814348004031786,
"learning_rate": 8.453966990470656e-06,
"loss": 0.174,
"step": 623
},
{
"epoch": 0.9984,
"grad_norm": 0.6503814372726033,
"learning_rate": 8.447228535032536e-06,
"loss": 0.2096,
"step": 624
},
{
"epoch": 1.0,
"grad_norm": 0.6177140626876225,
"learning_rate": 8.440478124871054e-06,
"loss": 0.1699,
"step": 625
},
{
"epoch": 1.0016,
"grad_norm": 0.6279296362469027,
"learning_rate": 8.433715783396115e-06,
"loss": 0.1593,
"step": 626
},
{
"epoch": 1.0032,
"grad_norm": 0.6169920243765645,
"learning_rate": 8.426941534058999e-06,
"loss": 0.1758,
"step": 627
},
{
"epoch": 1.0048,
"grad_norm": 0.5504261554913911,
"learning_rate": 8.420155400352279e-06,
"loss": 0.1435,
"step": 628
},
{
"epoch": 1.0064,
"grad_norm": 0.5328785878205684,
"learning_rate": 8.413357405809748e-06,
"loss": 0.1409,
"step": 629
},
{
"epoch": 1.008,
"grad_norm": 0.5906202874951293,
"learning_rate": 8.406547574006326e-06,
"loss": 0.1616,
"step": 630
},
{
"epoch": 1.0096,
"grad_norm": 0.4976338334935746,
"learning_rate": 8.399725928557985e-06,
"loss": 0.1679,
"step": 631
},
{
"epoch": 1.0112,
"grad_norm": 0.5073160477324073,
"learning_rate": 8.39289249312167e-06,
"loss": 0.1829,
"step": 632
},
{
"epoch": 1.0128,
"grad_norm": 0.433808154094448,
"learning_rate": 8.386047291395208e-06,
"loss": 0.1489,
"step": 633
},
{
"epoch": 1.0144,
"grad_norm": 0.6031025254613697,
"learning_rate": 8.37919034711723e-06,
"loss": 0.1505,
"step": 634
},
{
"epoch": 1.016,
"grad_norm": 0.5780776063205046,
"learning_rate": 8.372321684067092e-06,
"loss": 0.1441,
"step": 635
},
{
"epoch": 1.0176,
"grad_norm": 0.5742093607633564,
"learning_rate": 8.36544132606479e-06,
"loss": 0.1705,
"step": 636
},
{
"epoch": 1.0192,
"grad_norm": 0.4364812464665491,
"learning_rate": 8.358549296970877e-06,
"loss": 0.1485,
"step": 637
},
{
"epoch": 1.0208,
"grad_norm": 0.5816408322219182,
"learning_rate": 8.351645620686377e-06,
"loss": 0.1248,
"step": 638
},
{
"epoch": 1.0224,
"grad_norm": 0.5213840195202285,
"learning_rate": 8.34473032115271e-06,
"loss": 0.1274,
"step": 639
},
{
"epoch": 1.024,
"grad_norm": 0.5331917494486439,
"learning_rate": 8.337803422351602e-06,
"loss": 0.1403,
"step": 640
},
{
"epoch": 1.0256,
"grad_norm": 0.6014549854455337,
"learning_rate": 8.33086494830501e-06,
"loss": 0.1669,
"step": 641
},
{
"epoch": 1.0272,
"grad_norm": 0.5713419607455482,
"learning_rate": 8.323914923075018e-06,
"loss": 0.167,
"step": 642
},
{
"epoch": 1.0288,
"grad_norm": 0.6256277902771666,
"learning_rate": 8.316953370763788e-06,
"loss": 0.1573,
"step": 643
},
{
"epoch": 1.0304,
"grad_norm": 0.4790706972867777,
"learning_rate": 8.309980315513444e-06,
"loss": 0.1482,
"step": 644
},
{
"epoch": 1.032,
"grad_norm": 0.4907914743900495,
"learning_rate": 8.302995781506007e-06,
"loss": 0.1499,
"step": 645
},
{
"epoch": 1.0336,
"grad_norm": 0.5497397467829843,
"learning_rate": 8.295999792963301e-06,
"loss": 0.145,
"step": 646
},
{
"epoch": 1.0352,
"grad_norm": 0.4916588139180812,
"learning_rate": 8.288992374146878e-06,
"loss": 0.1341,
"step": 647
},
{
"epoch": 1.0368,
"grad_norm": 0.5691768666597691,
"learning_rate": 8.281973549357927e-06,
"loss": 0.1423,
"step": 648
},
{
"epoch": 1.0384,
"grad_norm": 0.5659702014394414,
"learning_rate": 8.274943342937191e-06,
"loss": 0.1721,
"step": 649
},
{
"epoch": 1.04,
"grad_norm": 0.4533206244016209,
"learning_rate": 8.267901779264889e-06,
"loss": 0.117,
"step": 650
},
{
"epoch": 1.0416,
"grad_norm": 0.5278376568158641,
"learning_rate": 8.260848882760616e-06,
"loss": 0.1515,
"step": 651
},
{
"epoch": 1.0432,
"grad_norm": 0.5719475420028524,
"learning_rate": 8.25378467788328e-06,
"loss": 0.1453,
"step": 652
},
{
"epoch": 1.0448,
"grad_norm": 0.6822870435613364,
"learning_rate": 8.246709189130997e-06,
"loss": 0.1705,
"step": 653
},
{
"epoch": 1.0464,
"grad_norm": 0.6190177276506774,
"learning_rate": 8.23962244104102e-06,
"loss": 0.1583,
"step": 654
},
{
"epoch": 1.048,
"grad_norm": 0.6738420565245757,
"learning_rate": 8.232524458189644e-06,
"loss": 0.1564,
"step": 655
},
{
"epoch": 1.0496,
"grad_norm": 0.5036375360542049,
"learning_rate": 8.225415265192126e-06,
"loss": 0.1452,
"step": 656
},
{
"epoch": 1.0512,
"grad_norm": 0.6382256818876165,
"learning_rate": 8.218294886702606e-06,
"loss": 0.1522,
"step": 657
},
{
"epoch": 1.0528,
"grad_norm": 0.6400283751172734,
"learning_rate": 8.211163347414005e-06,
"loss": 0.1747,
"step": 658
},
{
"epoch": 1.0544,
"grad_norm": 0.5185706085217897,
"learning_rate": 8.20402067205795e-06,
"loss": 0.1347,
"step": 659
},
{
"epoch": 1.056,
"grad_norm": 0.5067954848845849,
"learning_rate": 8.196866885404697e-06,
"loss": 0.1305,
"step": 660
},
{
"epoch": 1.0576,
"grad_norm": 0.5293454043755162,
"learning_rate": 8.18970201226302e-06,
"loss": 0.1472,
"step": 661
},
{
"epoch": 1.0592,
"grad_norm": 0.5889077640609076,
"learning_rate": 8.182526077480153e-06,
"loss": 0.1508,
"step": 662
},
{
"epoch": 1.0608,
"grad_norm": 0.551049920014103,
"learning_rate": 8.175339105941685e-06,
"loss": 0.1724,
"step": 663
},
{
"epoch": 1.0624,
"grad_norm": 0.5319387589984294,
"learning_rate": 8.168141122571478e-06,
"loss": 0.1382,
"step": 664
},
{
"epoch": 1.064,
"grad_norm": 0.6362794700471304,
"learning_rate": 8.160932152331587e-06,
"loss": 0.1757,
"step": 665
},
{
"epoch": 1.0656,
"grad_norm": 0.5372073650506101,
"learning_rate": 8.153712220222163e-06,
"loss": 0.1456,
"step": 666
},
{
"epoch": 1.0672,
"grad_norm": 0.4562474453373547,
"learning_rate": 8.14648135128138e-06,
"loss": 0.1537,
"step": 667
},
{
"epoch": 1.0688,
"grad_norm": 0.46286161303394696,
"learning_rate": 8.139239570585334e-06,
"loss": 0.1332,
"step": 668
},
{
"epoch": 1.0704,
"grad_norm": 0.7522795611502965,
"learning_rate": 8.131986903247959e-06,
"loss": 0.1882,
"step": 669
},
{
"epoch": 1.072,
"grad_norm": 0.5461474780631845,
"learning_rate": 8.124723374420951e-06,
"loss": 0.1583,
"step": 670
},
{
"epoch": 1.0735999999999999,
"grad_norm": 0.5729738194109378,
"learning_rate": 8.117449009293668e-06,
"loss": 0.1679,
"step": 671
},
{
"epoch": 1.0752,
"grad_norm": 0.5919046963114191,
"learning_rate": 8.11016383309305e-06,
"loss": 0.1667,
"step": 672
},
{
"epoch": 1.0768,
"grad_norm": 0.569697464956142,
"learning_rate": 8.102867871083528e-06,
"loss": 0.1731,
"step": 673
},
{
"epoch": 1.0784,
"grad_norm": 0.5591989676561727,
"learning_rate": 8.095561148566932e-06,
"loss": 0.1614,
"step": 674
},
{
"epoch": 1.08,
"grad_norm": 0.7113701376019653,
"learning_rate": 8.088243690882421e-06,
"loss": 0.1616,
"step": 675
},
{
"epoch": 1.0816,
"grad_norm": 0.6738955651124401,
"learning_rate": 8.080915523406371e-06,
"loss": 0.1543,
"step": 676
},
{
"epoch": 1.0832,
"grad_norm": 0.6512101878698159,
"learning_rate": 8.073576671552303e-06,
"loss": 0.1703,
"step": 677
},
{
"epoch": 1.0848,
"grad_norm": 0.5387775822430224,
"learning_rate": 8.06622716077079e-06,
"loss": 0.1401,
"step": 678
},
{
"epoch": 1.0864,
"grad_norm": 0.5436367967786133,
"learning_rate": 8.058867016549372e-06,
"loss": 0.1583,
"step": 679
},
{
"epoch": 1.088,
"grad_norm": 0.5500941360414712,
"learning_rate": 8.051496264412464e-06,
"loss": 0.16,
"step": 680
},
{
"epoch": 1.0896,
"grad_norm": 0.5346718159425781,
"learning_rate": 8.044114929921264e-06,
"loss": 0.1629,
"step": 681
},
{
"epoch": 1.0912,
"grad_norm": 0.6198957900170741,
"learning_rate": 8.036723038673675e-06,
"loss": 0.1647,
"step": 682
},
{
"epoch": 1.0928,
"grad_norm": 0.583906768989803,
"learning_rate": 8.029320616304204e-06,
"loss": 0.1306,
"step": 683
},
{
"epoch": 1.0944,
"grad_norm": 0.53211966582671,
"learning_rate": 8.021907688483885e-06,
"loss": 0.1412,
"step": 684
},
{
"epoch": 1.096,
"grad_norm": 0.7074151711973038,
"learning_rate": 8.01448428092018e-06,
"loss": 0.1493,
"step": 685
},
{
"epoch": 1.0976,
"grad_norm": 0.6404806694672004,
"learning_rate": 8.007050419356898e-06,
"loss": 0.1874,
"step": 686
},
{
"epoch": 1.0992,
"grad_norm": 0.5929215250068229,
"learning_rate": 7.999606129574096e-06,
"loss": 0.177,
"step": 687
},
{
"epoch": 1.1008,
"grad_norm": 0.6145735157226404,
"learning_rate": 7.992151437387999e-06,
"loss": 0.1736,
"step": 688
},
{
"epoch": 1.1024,
"grad_norm": 0.5870984741370409,
"learning_rate": 7.984686368650907e-06,
"loss": 0.156,
"step": 689
},
{
"epoch": 1.104,
"grad_norm": 0.6212957365777209,
"learning_rate": 7.977210949251102e-06,
"loss": 0.1679,
"step": 690
},
{
"epoch": 1.1056,
"grad_norm": 0.5836207483589525,
"learning_rate": 7.969725205112766e-06,
"loss": 0.1657,
"step": 691
},
{
"epoch": 1.1072,
"grad_norm": 0.604056146161596,
"learning_rate": 7.962229162195882e-06,
"loss": 0.1774,
"step": 692
},
{
"epoch": 1.1088,
"grad_norm": 0.586972544170573,
"learning_rate": 7.95472284649615e-06,
"loss": 0.1541,
"step": 693
},
{
"epoch": 1.1104,
"grad_norm": 0.5573257958042075,
"learning_rate": 7.947206284044896e-06,
"loss": 0.1746,
"step": 694
},
{
"epoch": 1.112,
"grad_norm": 0.5613769117424418,
"learning_rate": 7.939679500908982e-06,
"loss": 0.1708,
"step": 695
},
{
"epoch": 1.1136,
"grad_norm": 0.5052214450888993,
"learning_rate": 7.932142523190711e-06,
"loss": 0.1494,
"step": 696
},
{
"epoch": 1.1152,
"grad_norm": 0.5119675218605533,
"learning_rate": 7.924595377027741e-06,
"loss": 0.169,
"step": 697
},
{
"epoch": 1.1168,
"grad_norm": 0.7543158919111078,
"learning_rate": 7.917038088592997e-06,
"loss": 0.1424,
"step": 698
},
{
"epoch": 1.1184,
"grad_norm": 0.5603695885292748,
"learning_rate": 7.90947068409457e-06,
"loss": 0.149,
"step": 699
},
{
"epoch": 1.12,
"grad_norm": 0.7499772960621266,
"learning_rate": 7.90189318977564e-06,
"loss": 0.1486,
"step": 700
},
{
"epoch": 1.1216,
"grad_norm": 0.5079373049915685,
"learning_rate": 7.894305631914373e-06,
"loss": 0.1339,
"step": 701
},
{
"epoch": 1.1232,
"grad_norm": 0.6065323032378716,
"learning_rate": 7.886708036823838e-06,
"loss": 0.1657,
"step": 702
},
{
"epoch": 1.1248,
"grad_norm": 0.561729287241411,
"learning_rate": 7.879100430851907e-06,
"loss": 0.1505,
"step": 703
},
{
"epoch": 1.1264,
"grad_norm": 0.7225921944339428,
"learning_rate": 7.871482840381174e-06,
"loss": 0.1633,
"step": 704
},
{
"epoch": 1.1280000000000001,
"grad_norm": 0.46357686320666275,
"learning_rate": 7.863855291828857e-06,
"loss": 0.1421,
"step": 705
},
{
"epoch": 1.1296,
"grad_norm": 0.510849628162838,
"learning_rate": 7.856217811646707e-06,
"loss": 0.1434,
"step": 706
},
{
"epoch": 1.1312,
"grad_norm": 0.5485276231464398,
"learning_rate": 7.848570426320918e-06,
"loss": 0.1482,
"step": 707
},
{
"epoch": 1.1328,
"grad_norm": 0.5803395448690748,
"learning_rate": 7.840913162372032e-06,
"loss": 0.1606,
"step": 708
},
{
"epoch": 1.1344,
"grad_norm": 0.6277608491894551,
"learning_rate": 7.833246046354856e-06,
"loss": 0.1661,
"step": 709
},
{
"epoch": 1.1360000000000001,
"grad_norm": 0.5374340387383901,
"learning_rate": 7.825569104858353e-06,
"loss": 0.1272,
"step": 710
},
{
"epoch": 1.1376,
"grad_norm": 0.6787355141433786,
"learning_rate": 7.81788236450557e-06,
"loss": 0.1713,
"step": 711
},
{
"epoch": 1.1392,
"grad_norm": 0.5987285377899805,
"learning_rate": 7.810185851953529e-06,
"loss": 0.1458,
"step": 712
},
{
"epoch": 1.1408,
"grad_norm": 0.7467984641938552,
"learning_rate": 7.802479593893142e-06,
"loss": 0.1605,
"step": 713
},
{
"epoch": 1.1424,
"grad_norm": 0.5043746986362968,
"learning_rate": 7.794763617049124e-06,
"loss": 0.1487,
"step": 714
},
{
"epoch": 1.144,
"grad_norm": 0.604071307126098,
"learning_rate": 7.787037948179884e-06,
"loss": 0.1453,
"step": 715
},
{
"epoch": 1.1456,
"grad_norm": 0.4556742185325212,
"learning_rate": 7.779302614077449e-06,
"loss": 0.1608,
"step": 716
},
{
"epoch": 1.1472,
"grad_norm": 0.47744612146512305,
"learning_rate": 7.771557641567363e-06,
"loss": 0.1496,
"step": 717
},
{
"epoch": 1.1488,
"grad_norm": 0.6323409937695268,
"learning_rate": 7.763803057508594e-06,
"loss": 0.1657,
"step": 718
},
{
"epoch": 1.1504,
"grad_norm": 0.5856309914852824,
"learning_rate": 7.756038888793446e-06,
"loss": 0.1382,
"step": 719
},
{
"epoch": 1.152,
"grad_norm": 0.5179981888857819,
"learning_rate": 7.748265162347455e-06,
"loss": 0.1569,
"step": 720
},
{
"epoch": 1.1536,
"grad_norm": 0.5490129622372413,
"learning_rate": 7.740481905129307e-06,
"loss": 0.1447,
"step": 721
},
{
"epoch": 1.1552,
"grad_norm": 0.5498722673930562,
"learning_rate": 7.732689144130741e-06,
"loss": 0.161,
"step": 722
},
{
"epoch": 1.1568,
"grad_norm": 0.5597760588731585,
"learning_rate": 7.724886906376451e-06,
"loss": 0.1438,
"step": 723
},
{
"epoch": 1.1584,
"grad_norm": 0.6481079187988128,
"learning_rate": 7.717075218923998e-06,
"loss": 0.183,
"step": 724
},
{
"epoch": 1.16,
"grad_norm": 0.6139842046656744,
"learning_rate": 7.709254108863714e-06,
"loss": 0.171,
"step": 725
},
{
"epoch": 1.1616,
"grad_norm": 0.6212254029001303,
"learning_rate": 7.701423603318605e-06,
"loss": 0.1647,
"step": 726
},
{
"epoch": 1.1632,
"grad_norm": 0.608744973839724,
"learning_rate": 7.693583729444263e-06,
"loss": 0.1723,
"step": 727
},
{
"epoch": 1.1648,
"grad_norm": 0.5487659031122327,
"learning_rate": 7.685734514428767e-06,
"loss": 0.1467,
"step": 728
},
{
"epoch": 1.1663999999999999,
"grad_norm": 0.5400176548099397,
"learning_rate": 7.677875985492591e-06,
"loss": 0.1484,
"step": 729
},
{
"epoch": 1.168,
"grad_norm": 0.5318741683739623,
"learning_rate": 7.67000816988851e-06,
"loss": 0.1453,
"step": 730
},
{
"epoch": 1.1696,
"grad_norm": 0.5515436460713169,
"learning_rate": 7.662131094901499e-06,
"loss": 0.1554,
"step": 731
},
{
"epoch": 1.1712,
"grad_norm": 0.6595695762130074,
"learning_rate": 7.654244787848655e-06,
"loss": 0.1569,
"step": 732
},
{
"epoch": 1.1728,
"grad_norm": 0.5787190224694766,
"learning_rate": 7.646349276079079e-06,
"loss": 0.1659,
"step": 733
},
{
"epoch": 1.1743999999999999,
"grad_norm": 0.729230198001716,
"learning_rate": 7.6384445869738e-06,
"loss": 0.1943,
"step": 734
},
{
"epoch": 1.176,
"grad_norm": 0.5836025705757528,
"learning_rate": 7.630530747945672e-06,
"loss": 0.1329,
"step": 735
},
{
"epoch": 1.1776,
"grad_norm": 0.656440732853705,
"learning_rate": 7.622607786439279e-06,
"loss": 0.1451,
"step": 736
},
{
"epoch": 1.1792,
"grad_norm": 0.5215030466523584,
"learning_rate": 7.6146757299308406e-06,
"loss": 0.1541,
"step": 737
},
{
"epoch": 1.1808,
"grad_norm": 0.5830246808269974,
"learning_rate": 7.606734605928123e-06,
"loss": 0.1369,
"step": 738
},
{
"epoch": 1.1824,
"grad_norm": 0.49170586991389953,
"learning_rate": 7.598784441970329e-06,
"loss": 0.1373,
"step": 739
},
{
"epoch": 1.184,
"grad_norm": 0.48377737237174273,
"learning_rate": 7.590825265628019e-06,
"loss": 0.162,
"step": 740
},
{
"epoch": 1.1856,
"grad_norm": 0.5612309538994934,
"learning_rate": 7.5828571045030005e-06,
"loss": 0.1559,
"step": 741
},
{
"epoch": 1.1872,
"grad_norm": 0.7738721031229652,
"learning_rate": 7.574879986228245e-06,
"loss": 0.1874,
"step": 742
},
{
"epoch": 1.1888,
"grad_norm": 0.5181952631455986,
"learning_rate": 7.566893938467788e-06,
"loss": 0.1243,
"step": 743
},
{
"epoch": 1.1904,
"grad_norm": 0.5322199470099435,
"learning_rate": 7.558898988916624e-06,
"loss": 0.1562,
"step": 744
},
{
"epoch": 1.192,
"grad_norm": 0.5711196347653461,
"learning_rate": 7.550895165300626e-06,
"loss": 0.1457,
"step": 745
},
{
"epoch": 1.1936,
"grad_norm": 0.5644921475966505,
"learning_rate": 7.542882495376437e-06,
"loss": 0.1544,
"step": 746
},
{
"epoch": 1.1952,
"grad_norm": 0.44082119467399256,
"learning_rate": 7.5348610069313795e-06,
"loss": 0.152,
"step": 747
},
{
"epoch": 1.1968,
"grad_norm": 0.4637786621995717,
"learning_rate": 7.5268307277833605e-06,
"loss": 0.1474,
"step": 748
},
{
"epoch": 1.1984,
"grad_norm": 0.6429446450374807,
"learning_rate": 7.518791685780769e-06,
"loss": 0.1542,
"step": 749
},
{
"epoch": 1.2,
"grad_norm": 0.570674502228561,
"learning_rate": 7.5107439088023845e-06,
"loss": 0.169,
"step": 750
},
{
"epoch": 1.2016,
"grad_norm": 0.5220699747319325,
"learning_rate": 7.502687424757278e-06,
"loss": 0.122,
"step": 751
},
{
"epoch": 1.2032,
"grad_norm": 0.6561686013477482,
"learning_rate": 7.4946222615847165e-06,
"loss": 0.1591,
"step": 752
},
{
"epoch": 1.2048,
"grad_norm": 0.5624109704097062,
"learning_rate": 7.486548447254065e-06,
"loss": 0.1227,
"step": 753
},
{
"epoch": 1.2064,
"grad_norm": 0.4853983491296869,
"learning_rate": 7.478466009764692e-06,
"loss": 0.1549,
"step": 754
},
{
"epoch": 1.208,
"grad_norm": 0.524585540494509,
"learning_rate": 7.470374977145867e-06,
"loss": 0.1375,
"step": 755
},
{
"epoch": 1.2096,
"grad_norm": 0.517879320449341,
"learning_rate": 7.462275377456671e-06,
"loss": 0.1521,
"step": 756
},
{
"epoch": 1.2112,
"grad_norm": 0.5106256553752204,
"learning_rate": 7.4541672387858895e-06,
"loss": 0.1434,
"step": 757
},
{
"epoch": 1.2128,
"grad_norm": 0.5309484212130765,
"learning_rate": 7.446050589251928e-06,
"loss": 0.1355,
"step": 758
},
{
"epoch": 1.2144,
"grad_norm": 0.6337563744932865,
"learning_rate": 7.437925457002697e-06,
"loss": 0.1684,
"step": 759
},
{
"epoch": 1.216,
"grad_norm": 0.5037724076452551,
"learning_rate": 7.429791870215535e-06,
"loss": 0.1525,
"step": 760
},
{
"epoch": 1.2176,
"grad_norm": 0.5521881595794954,
"learning_rate": 7.421649857097092e-06,
"loss": 0.1617,
"step": 761
},
{
"epoch": 1.2192,
"grad_norm": 0.5531885468537365,
"learning_rate": 7.413499445883245e-06,
"loss": 0.1429,
"step": 762
},
{
"epoch": 1.2208,
"grad_norm": 0.5498011201020139,
"learning_rate": 7.405340664838994e-06,
"loss": 0.1607,
"step": 763
},
{
"epoch": 1.2224,
"grad_norm": 0.544062105621307,
"learning_rate": 7.39717354225836e-06,
"loss": 0.1328,
"step": 764
},
{
"epoch": 1.224,
"grad_norm": 0.5639317179411057,
"learning_rate": 7.3889981064643e-06,
"loss": 0.1656,
"step": 765
},
{
"epoch": 1.2256,
"grad_norm": 0.5981495862361838,
"learning_rate": 7.380814385808594e-06,
"loss": 0.1451,
"step": 766
},
{
"epoch": 1.2272,
"grad_norm": 0.5997605000524621,
"learning_rate": 7.372622408671757e-06,
"loss": 0.1506,
"step": 767
},
{
"epoch": 1.2288000000000001,
"grad_norm": 0.6848779464476087,
"learning_rate": 7.364422203462935e-06,
"loss": 0.1627,
"step": 768
},
{
"epoch": 1.2304,
"grad_norm": 0.5069993151541472,
"learning_rate": 7.3562137986198065e-06,
"loss": 0.1513,
"step": 769
},
{
"epoch": 1.232,
"grad_norm": 0.5020139809651705,
"learning_rate": 7.3479972226084925e-06,
"loss": 0.1482,
"step": 770
},
{
"epoch": 1.2336,
"grad_norm": 0.5441488663971923,
"learning_rate": 7.339772503923445e-06,
"loss": 0.1532,
"step": 771
},
{
"epoch": 1.2352,
"grad_norm": 0.6069308433419601,
"learning_rate": 7.331539671087353e-06,
"loss": 0.149,
"step": 772
},
{
"epoch": 1.2368000000000001,
"grad_norm": 0.6408050457608785,
"learning_rate": 7.32329875265105e-06,
"loss": 0.1464,
"step": 773
},
{
"epoch": 1.2384,
"grad_norm": 0.5627210618459557,
"learning_rate": 7.315049777193407e-06,
"loss": 0.1367,
"step": 774
},
{
"epoch": 1.24,
"grad_norm": 0.6136501470019226,
"learning_rate": 7.306792773321234e-06,
"loss": 0.1476,
"step": 775
},
{
"epoch": 1.2416,
"grad_norm": 0.6226943009429379,
"learning_rate": 7.298527769669188e-06,
"loss": 0.1603,
"step": 776
},
{
"epoch": 1.2432,
"grad_norm": 0.4813284291522472,
"learning_rate": 7.290254794899665e-06,
"loss": 0.1619,
"step": 777
},
{
"epoch": 1.2448,
"grad_norm": 0.40950753577111415,
"learning_rate": 7.281973877702705e-06,
"loss": 0.1279,
"step": 778
},
{
"epoch": 1.2464,
"grad_norm": 0.45648472868558637,
"learning_rate": 7.2736850467958905e-06,
"loss": 0.139,
"step": 779
},
{
"epoch": 1.248,
"grad_norm": 0.42575428780079727,
"learning_rate": 7.26538833092425e-06,
"loss": 0.1357,
"step": 780
},
{
"epoch": 1.2496,
"grad_norm": 0.5100880549828651,
"learning_rate": 7.257083758860159e-06,
"loss": 0.1465,
"step": 781
},
{
"epoch": 1.2511999999999999,
"grad_norm": 0.5246108217810566,
"learning_rate": 7.248771359403231e-06,
"loss": 0.1415,
"step": 782
},
{
"epoch": 1.2528000000000001,
"grad_norm": 0.5624386704603905,
"learning_rate": 7.240451161380226e-06,
"loss": 0.1685,
"step": 783
},
{
"epoch": 1.2544,
"grad_norm": 0.5243691082628384,
"learning_rate": 7.232123193644957e-06,
"loss": 0.1362,
"step": 784
},
{
"epoch": 1.256,
"grad_norm": 0.5743633306738232,
"learning_rate": 7.22378748507817e-06,
"loss": 0.1507,
"step": 785
},
{
"epoch": 1.2576,
"grad_norm": 0.6343901102305041,
"learning_rate": 7.215444064587462e-06,
"loss": 0.1615,
"step": 786
},
{
"epoch": 1.2591999999999999,
"grad_norm": 0.5564320455817214,
"learning_rate": 7.207092961107176e-06,
"loss": 0.1548,
"step": 787
},
{
"epoch": 1.2608,
"grad_norm": 0.5253941035405657,
"learning_rate": 7.198734203598294e-06,
"loss": 0.1692,
"step": 788
},
{
"epoch": 1.2624,
"grad_norm": 0.825255107618771,
"learning_rate": 7.190367821048346e-06,
"loss": 0.145,
"step": 789
},
{
"epoch": 1.264,
"grad_norm": 0.5111747021665888,
"learning_rate": 7.181993842471301e-06,
"loss": 0.1273,
"step": 790
},
{
"epoch": 1.2656,
"grad_norm": 0.4766009489404193,
"learning_rate": 7.173612296907473e-06,
"loss": 0.1577,
"step": 791
},
{
"epoch": 1.2671999999999999,
"grad_norm": 0.6230147518519226,
"learning_rate": 7.165223213423416e-06,
"loss": 0.1413,
"step": 792
},
{
"epoch": 1.2688,
"grad_norm": 0.6012707541973911,
"learning_rate": 7.15682662111183e-06,
"loss": 0.163,
"step": 793
},
{
"epoch": 1.2704,
"grad_norm": 0.5519377349178405,
"learning_rate": 7.148422549091447e-06,
"loss": 0.1632,
"step": 794
},
{
"epoch": 1.272,
"grad_norm": 0.4801986718973068,
"learning_rate": 7.140011026506945e-06,
"loss": 0.1331,
"step": 795
},
{
"epoch": 1.2736,
"grad_norm": 0.6142087167483267,
"learning_rate": 7.131592082528837e-06,
"loss": 0.1693,
"step": 796
},
{
"epoch": 1.2752,
"grad_norm": 0.5391896700166154,
"learning_rate": 7.12316574635337e-06,
"loss": 0.1432,
"step": 797
},
{
"epoch": 1.2768,
"grad_norm": 0.5877642944054979,
"learning_rate": 7.114732047202433e-06,
"loss": 0.1558,
"step": 798
},
{
"epoch": 1.2784,
"grad_norm": 0.6208516283697577,
"learning_rate": 7.106291014323445e-06,
"loss": 0.1606,
"step": 799
},
{
"epoch": 1.28,
"grad_norm": 0.5914362379543822,
"learning_rate": 7.0978426769892585e-06,
"loss": 0.1478,
"step": 800
},
{
"epoch": 1.2816,
"grad_norm": 0.5493827670833178,
"learning_rate": 7.089387064498057e-06,
"loss": 0.1644,
"step": 801
},
{
"epoch": 1.2832,
"grad_norm": 0.4298951821491084,
"learning_rate": 7.080924206173253e-06,
"loss": 0.1459,
"step": 802
},
{
"epoch": 1.2848,
"grad_norm": 0.5564387294953059,
"learning_rate": 7.072454131363391e-06,
"loss": 0.1476,
"step": 803
},
{
"epoch": 1.2864,
"grad_norm": 0.5754097526572193,
"learning_rate": 7.063976869442037e-06,
"loss": 0.1496,
"step": 804
},
{
"epoch": 1.288,
"grad_norm": 0.5550888572756114,
"learning_rate": 7.055492449807684e-06,
"loss": 0.1504,
"step": 805
},
{
"epoch": 1.2896,
"grad_norm": 0.567171453948784,
"learning_rate": 7.047000901883646e-06,
"loss": 0.1432,
"step": 806
},
{
"epoch": 1.2912,
"grad_norm": 0.5082166399579625,
"learning_rate": 7.038502255117957e-06,
"loss": 0.1391,
"step": 807
},
{
"epoch": 1.2928,
"grad_norm": 0.6337338127086184,
"learning_rate": 7.029996538983273e-06,
"loss": 0.1531,
"step": 808
},
{
"epoch": 1.2944,
"grad_norm": 0.5456478057031151,
"learning_rate": 7.021483782976759e-06,
"loss": 0.1323,
"step": 809
},
{
"epoch": 1.296,
"grad_norm": 0.5597415355737146,
"learning_rate": 7.012964016620002e-06,
"loss": 0.1617,
"step": 810
},
{
"epoch": 1.2976,
"grad_norm": 0.5330856236520068,
"learning_rate": 7.004437269458894e-06,
"loss": 0.1364,
"step": 811
},
{
"epoch": 1.2992,
"grad_norm": 0.6012969782223988,
"learning_rate": 6.995903571063541e-06,
"loss": 0.17,
"step": 812
},
{
"epoch": 1.3008,
"grad_norm": 0.6084230749942078,
"learning_rate": 6.987362951028147e-06,
"loss": 0.1311,
"step": 813
},
{
"epoch": 1.3024,
"grad_norm": 0.4686087225933573,
"learning_rate": 6.97881543897093e-06,
"loss": 0.1398,
"step": 814
},
{
"epoch": 1.304,
"grad_norm": 0.5911287790476738,
"learning_rate": 6.970261064534003e-06,
"loss": 0.1528,
"step": 815
},
{
"epoch": 1.3056,
"grad_norm": 0.6388720637973209,
"learning_rate": 6.961699857383279e-06,
"loss": 0.1826,
"step": 816
},
{
"epoch": 1.3072,
"grad_norm": 0.5516837183234556,
"learning_rate": 6.953131847208365e-06,
"loss": 0.155,
"step": 817
},
{
"epoch": 1.3088,
"grad_norm": 0.575298300574076,
"learning_rate": 6.944557063722459e-06,
"loss": 0.1615,
"step": 818
},
{
"epoch": 1.3104,
"grad_norm": 0.44762110958001544,
"learning_rate": 6.935975536662254e-06,
"loss": 0.1359,
"step": 819
},
{
"epoch": 1.312,
"grad_norm": 0.5578738184942572,
"learning_rate": 6.9273872957878255e-06,
"loss": 0.1494,
"step": 820
},
{
"epoch": 1.3136,
"grad_norm": 0.6172346964417539,
"learning_rate": 6.91879237088253e-06,
"loss": 0.157,
"step": 821
},
{
"epoch": 1.3152,
"grad_norm": 0.5361703723110776,
"learning_rate": 6.910190791752907e-06,
"loss": 0.152,
"step": 822
},
{
"epoch": 1.3168,
"grad_norm": 0.5620161571748259,
"learning_rate": 6.90158258822857e-06,
"loss": 0.1413,
"step": 823
},
{
"epoch": 1.3184,
"grad_norm": 0.5435786383442265,
"learning_rate": 6.892967790162109e-06,
"loss": 0.1546,
"step": 824
},
{
"epoch": 1.32,
"grad_norm": 0.5406728173733499,
"learning_rate": 6.884346427428978e-06,
"loss": 0.1373,
"step": 825
},
{
"epoch": 1.3216,
"grad_norm": 0.5725273169790757,
"learning_rate": 6.875718529927404e-06,
"loss": 0.1527,
"step": 826
},
{
"epoch": 1.3232,
"grad_norm": 0.6239622747013701,
"learning_rate": 6.867084127578267e-06,
"loss": 0.1286,
"step": 827
},
{
"epoch": 1.3248,
"grad_norm": 0.5396781697020108,
"learning_rate": 6.858443250325013e-06,
"loss": 0.148,
"step": 828
},
{
"epoch": 1.3264,
"grad_norm": 0.556069357553151,
"learning_rate": 6.849795928133538e-06,
"loss": 0.1588,
"step": 829
},
{
"epoch": 1.328,
"grad_norm": 0.6774883749142859,
"learning_rate": 6.841142190992092e-06,
"loss": 0.1663,
"step": 830
},
{
"epoch": 1.3296000000000001,
"grad_norm": 0.6323757392485342,
"learning_rate": 6.832482068911167e-06,
"loss": 0.1719,
"step": 831
},
{
"epoch": 1.3312,
"grad_norm": 0.6613058841340895,
"learning_rate": 6.823815591923402e-06,
"loss": 0.1591,
"step": 832
},
{
"epoch": 1.3328,
"grad_norm": 0.5989112297405516,
"learning_rate": 6.815142790083473e-06,
"loss": 0.1761,
"step": 833
},
{
"epoch": 1.3344,
"grad_norm": 0.5563628334117668,
"learning_rate": 6.8064636934679885e-06,
"loss": 0.1549,
"step": 834
},
{
"epoch": 1.336,
"grad_norm": 0.556391503346919,
"learning_rate": 6.797778332175387e-06,
"loss": 0.1775,
"step": 835
},
{
"epoch": 1.3376000000000001,
"grad_norm": 0.6099848597293082,
"learning_rate": 6.789086736325834e-06,
"loss": 0.1551,
"step": 836
},
{
"epoch": 1.3392,
"grad_norm": 0.520527739132676,
"learning_rate": 6.780388936061118e-06,
"loss": 0.1601,
"step": 837
},
{
"epoch": 1.3408,
"grad_norm": 0.5236449600727803,
"learning_rate": 6.771684961544537e-06,
"loss": 0.1356,
"step": 838
},
{
"epoch": 1.3424,
"grad_norm": 0.5269077608757189,
"learning_rate": 6.7629748429608076e-06,
"loss": 0.1497,
"step": 839
},
{
"epoch": 1.3439999999999999,
"grad_norm": 0.592694905381524,
"learning_rate": 6.754258610515949e-06,
"loss": 0.1744,
"step": 840
},
{
"epoch": 1.3456000000000001,
"grad_norm": 0.5440171139434928,
"learning_rate": 6.745536294437187e-06,
"loss": 0.1389,
"step": 841
},
{
"epoch": 1.3472,
"grad_norm": 0.5025512744260899,
"learning_rate": 6.736807924972841e-06,
"loss": 0.1483,
"step": 842
},
{
"epoch": 1.3488,
"grad_norm": 0.5764421624282045,
"learning_rate": 6.728073532392226e-06,
"loss": 0.1721,
"step": 843
},
{
"epoch": 1.3504,
"grad_norm": 0.6255857137496629,
"learning_rate": 6.719333146985544e-06,
"loss": 0.1574,
"step": 844
},
{
"epoch": 1.3519999999999999,
"grad_norm": 0.6611496792564259,
"learning_rate": 6.710586799063777e-06,
"loss": 0.1531,
"step": 845
},
{
"epoch": 1.3536000000000001,
"grad_norm": 0.5802836119065187,
"learning_rate": 6.701834518958587e-06,
"loss": 0.1269,
"step": 846
},
{
"epoch": 1.3552,
"grad_norm": 0.6582808926096432,
"learning_rate": 6.6930763370222104e-06,
"loss": 0.1521,
"step": 847
},
{
"epoch": 1.3568,
"grad_norm": 0.6474062125297559,
"learning_rate": 6.684312283627348e-06,
"loss": 0.1482,
"step": 848
},
{
"epoch": 1.3584,
"grad_norm": 0.5869443972798116,
"learning_rate": 6.6755423891670605e-06,
"loss": 0.1575,
"step": 849
},
{
"epoch": 1.3599999999999999,
"grad_norm": 0.4816289960840906,
"learning_rate": 6.6667666840546685e-06,
"loss": 0.1465,
"step": 850
},
{
"epoch": 1.3616,
"grad_norm": 0.5773341751638496,
"learning_rate": 6.6579851987236435e-06,
"loss": 0.1573,
"step": 851
},
{
"epoch": 1.3632,
"grad_norm": 0.5062507234891164,
"learning_rate": 6.649197963627497e-06,
"loss": 0.1637,
"step": 852
},
{
"epoch": 1.3648,
"grad_norm": 0.5009639734955896,
"learning_rate": 6.640405009239689e-06,
"loss": 0.1393,
"step": 853
},
{
"epoch": 1.3664,
"grad_norm": 0.4539215560606831,
"learning_rate": 6.631606366053507e-06,
"loss": 0.1424,
"step": 854
},
{
"epoch": 1.3679999999999999,
"grad_norm": 0.6247871913131606,
"learning_rate": 6.622802064581968e-06,
"loss": 0.1826,
"step": 855
},
{
"epoch": 1.3696,
"grad_norm": 0.5265634773339909,
"learning_rate": 6.613992135357713e-06,
"loss": 0.1809,
"step": 856
},
{
"epoch": 1.3712,
"grad_norm": 0.5117738515896976,
"learning_rate": 6.605176608932897e-06,
"loss": 0.1465,
"step": 857
},
{
"epoch": 1.3728,
"grad_norm": 0.5245651195289621,
"learning_rate": 6.596355515879091e-06,
"loss": 0.1365,
"step": 858
},
{
"epoch": 1.3744,
"grad_norm": 0.6989735364854152,
"learning_rate": 6.587528886787165e-06,
"loss": 0.1659,
"step": 859
},
{
"epoch": 1.376,
"grad_norm": 0.5781222840043198,
"learning_rate": 6.578696752267189e-06,
"loss": 0.1491,
"step": 860
},
{
"epoch": 1.3776,
"grad_norm": 0.4823935197967017,
"learning_rate": 6.5698591429483286e-06,
"loss": 0.1465,
"step": 861
},
{
"epoch": 1.3792,
"grad_norm": 0.5240539510431244,
"learning_rate": 6.5610160894787275e-06,
"loss": 0.137,
"step": 862
},
{
"epoch": 1.3808,
"grad_norm": 0.6261921153544949,
"learning_rate": 6.552167622525421e-06,
"loss": 0.1521,
"step": 863
},
{
"epoch": 1.3824,
"grad_norm": 0.5715464635783474,
"learning_rate": 6.543313772774209e-06,
"loss": 0.1714,
"step": 864
},
{
"epoch": 1.384,
"grad_norm": 0.5725005230153112,
"learning_rate": 6.534454570929563e-06,
"loss": 0.1671,
"step": 865
},
{
"epoch": 1.3856,
"grad_norm": 0.6437703140813323,
"learning_rate": 6.52559004771451e-06,
"loss": 0.1471,
"step": 866
},
{
"epoch": 1.3872,
"grad_norm": 0.5790323320386902,
"learning_rate": 6.516720233870538e-06,
"loss": 0.1704,
"step": 867
},
{
"epoch": 1.3888,
"grad_norm": 0.5310317929848538,
"learning_rate": 6.507845160157476e-06,
"loss": 0.1522,
"step": 868
},
{
"epoch": 1.3904,
"grad_norm": 0.5236019767695259,
"learning_rate": 6.498964857353401e-06,
"loss": 0.1249,
"step": 869
},
{
"epoch": 1.392,
"grad_norm": 0.7212928222907986,
"learning_rate": 6.4900793562545165e-06,
"loss": 0.1503,
"step": 870
},
{
"epoch": 1.3936,
"grad_norm": 0.4665171635057242,
"learning_rate": 6.481188687675057e-06,
"loss": 0.1318,
"step": 871
},
{
"epoch": 1.3952,
"grad_norm": 0.5267917227851336,
"learning_rate": 6.47229288244718e-06,
"loss": 0.1355,
"step": 872
},
{
"epoch": 1.3968,
"grad_norm": 0.5038532558598493,
"learning_rate": 6.46339197142085e-06,
"loss": 0.1461,
"step": 873
},
{
"epoch": 1.3984,
"grad_norm": 0.5814020305901507,
"learning_rate": 6.454485985463742e-06,
"loss": 0.1738,
"step": 874
},
{
"epoch": 1.4,
"grad_norm": 0.5479101878525573,
"learning_rate": 6.445574955461134e-06,
"loss": 0.1541,
"step": 875
},
{
"epoch": 1.4016,
"grad_norm": 0.6678311234826609,
"learning_rate": 6.436658912315789e-06,
"loss": 0.1563,
"step": 876
},
{
"epoch": 1.4032,
"grad_norm": 0.6314397815071712,
"learning_rate": 6.427737886947859e-06,
"loss": 0.1466,
"step": 877
},
{
"epoch": 1.4048,
"grad_norm": 0.5717954182771735,
"learning_rate": 6.418811910294776e-06,
"loss": 0.1312,
"step": 878
},
{
"epoch": 1.4064,
"grad_norm": 0.577582181286399,
"learning_rate": 6.409881013311136e-06,
"loss": 0.1545,
"step": 879
},
{
"epoch": 1.408,
"grad_norm": 0.5788454875463872,
"learning_rate": 6.400945226968607e-06,
"loss": 0.1529,
"step": 880
},
{
"epoch": 1.4096,
"grad_norm": 0.5108931012176959,
"learning_rate": 6.392004582255807e-06,
"loss": 0.1614,
"step": 881
},
{
"epoch": 1.4112,
"grad_norm": 0.6091362567652464,
"learning_rate": 6.383059110178205e-06,
"loss": 0.1729,
"step": 882
},
{
"epoch": 1.4128,
"grad_norm": 0.5416356573906651,
"learning_rate": 6.374108841758006e-06,
"loss": 0.1505,
"step": 883
},
{
"epoch": 1.4144,
"grad_norm": 0.5179051501414724,
"learning_rate": 6.365153808034057e-06,
"loss": 0.1333,
"step": 884
},
{
"epoch": 1.416,
"grad_norm": 0.7281382714892222,
"learning_rate": 6.356194040061725e-06,
"loss": 0.1831,
"step": 885
},
{
"epoch": 1.4176,
"grad_norm": 0.4678339849048488,
"learning_rate": 6.3472295689127946e-06,
"loss": 0.1506,
"step": 886
},
{
"epoch": 1.4192,
"grad_norm": 0.556580811093487,
"learning_rate": 6.338260425675365e-06,
"loss": 0.1584,
"step": 887
},
{
"epoch": 1.4208,
"grad_norm": 0.5655944800548436,
"learning_rate": 6.329286641453729e-06,
"loss": 0.1654,
"step": 888
},
{
"epoch": 1.4224,
"grad_norm": 0.6008523764232223,
"learning_rate": 6.320308247368285e-06,
"loss": 0.1701,
"step": 889
},
{
"epoch": 1.424,
"grad_norm": 0.5477593602599127,
"learning_rate": 6.311325274555413e-06,
"loss": 0.1677,
"step": 890
},
{
"epoch": 1.4256,
"grad_norm": 1.2025396767647096,
"learning_rate": 6.302337754167369e-06,
"loss": 0.1635,
"step": 891
},
{
"epoch": 1.4272,
"grad_norm": 0.5374645713839911,
"learning_rate": 6.2933457173721855e-06,
"loss": 0.1789,
"step": 892
},
{
"epoch": 1.4288,
"grad_norm": 0.6306865105152074,
"learning_rate": 6.2843491953535515e-06,
"loss": 0.1498,
"step": 893
},
{
"epoch": 1.4304000000000001,
"grad_norm": 0.5579061672686437,
"learning_rate": 6.275348219310715e-06,
"loss": 0.147,
"step": 894
},
{
"epoch": 1.432,
"grad_norm": 0.5367505545237147,
"learning_rate": 6.266342820458366e-06,
"loss": 0.1536,
"step": 895
},
{
"epoch": 1.4336,
"grad_norm": 0.5080713718054941,
"learning_rate": 6.2573330300265375e-06,
"loss": 0.1455,
"step": 896
},
{
"epoch": 1.4352,
"grad_norm": 0.5209714319157174,
"learning_rate": 6.248318879260488e-06,
"loss": 0.1369,
"step": 897
},
{
"epoch": 1.4368,
"grad_norm": 0.5895412437474985,
"learning_rate": 6.239300399420601e-06,
"loss": 0.1691,
"step": 898
},
{
"epoch": 1.4384000000000001,
"grad_norm": 0.4974895425799255,
"learning_rate": 6.230277621782269e-06,
"loss": 0.1646,
"step": 899
},
{
"epoch": 1.44,
"grad_norm": 0.4902887835647808,
"learning_rate": 6.221250577635791e-06,
"loss": 0.128,
"step": 900
},
{
"epoch": 1.4416,
"grad_norm": 0.5533602703229357,
"learning_rate": 6.2122192982862615e-06,
"loss": 0.1621,
"step": 901
},
{
"epoch": 1.4432,
"grad_norm": 0.5360878159223267,
"learning_rate": 6.203183815053463e-06,
"loss": 0.1528,
"step": 902
},
{
"epoch": 1.4447999999999999,
"grad_norm": 0.5309701308876584,
"learning_rate": 6.1941441592717564e-06,
"loss": 0.1473,
"step": 903
},
{
"epoch": 1.4464000000000001,
"grad_norm": 0.5844769399580534,
"learning_rate": 6.185100362289972e-06,
"loss": 0.1414,
"step": 904
},
{
"epoch": 1.448,
"grad_norm": 0.6653453224405401,
"learning_rate": 6.176052455471302e-06,
"loss": 0.1554,
"step": 905
},
{
"epoch": 1.4496,
"grad_norm": 0.5051864022225347,
"learning_rate": 6.167000470193189e-06,
"loss": 0.1416,
"step": 906
},
{
"epoch": 1.4512,
"grad_norm": 0.6063180096496212,
"learning_rate": 6.157944437847226e-06,
"loss": 0.1589,
"step": 907
},
{
"epoch": 1.4527999999999999,
"grad_norm": 0.5212788754712189,
"learning_rate": 6.148884389839035e-06,
"loss": 0.1325,
"step": 908
},
{
"epoch": 1.4544000000000001,
"grad_norm": 0.6890835864660986,
"learning_rate": 6.1398203575881645e-06,
"loss": 0.1541,
"step": 909
},
{
"epoch": 1.456,
"grad_norm": 0.5372841712620403,
"learning_rate": 6.130752372527981e-06,
"loss": 0.1414,
"step": 910
},
{
"epoch": 1.4576,
"grad_norm": 0.6425977194067422,
"learning_rate": 6.121680466105559e-06,
"loss": 0.1579,
"step": 911
},
{
"epoch": 1.4592,
"grad_norm": 0.5911894517349429,
"learning_rate": 6.112604669781572e-06,
"loss": 0.1724,
"step": 912
},
{
"epoch": 1.4607999999999999,
"grad_norm": 0.5100025876387476,
"learning_rate": 6.1035250150301864e-06,
"loss": 0.1454,
"step": 913
},
{
"epoch": 1.4624,
"grad_norm": 0.5488845386134333,
"learning_rate": 6.0944415333389405e-06,
"loss": 0.1576,
"step": 914
},
{
"epoch": 1.464,
"grad_norm": 0.6126346878765391,
"learning_rate": 6.085354256208655e-06,
"loss": 0.1463,
"step": 915
},
{
"epoch": 1.4656,
"grad_norm": 0.492372467642877,
"learning_rate": 6.076263215153308e-06,
"loss": 0.1308,
"step": 916
},
{
"epoch": 1.4672,
"grad_norm": 0.5540249076039442,
"learning_rate": 6.067168441699927e-06,
"loss": 0.1716,
"step": 917
},
{
"epoch": 1.4687999999999999,
"grad_norm": 0.5537376875409453,
"learning_rate": 6.058069967388489e-06,
"loss": 0.1511,
"step": 918
},
{
"epoch": 1.4704,
"grad_norm": 0.6016252934442746,
"learning_rate": 6.048967823771802e-06,
"loss": 0.1505,
"step": 919
},
{
"epoch": 1.472,
"grad_norm": 0.5319120725182609,
"learning_rate": 6.039862042415401e-06,
"loss": 0.1494,
"step": 920
},
{
"epoch": 1.4736,
"grad_norm": 0.5594347322227284,
"learning_rate": 6.030752654897435e-06,
"loss": 0.1338,
"step": 921
},
{
"epoch": 1.4752,
"grad_norm": 0.6414965362642412,
"learning_rate": 6.021639692808558e-06,
"loss": 0.1764,
"step": 922
},
{
"epoch": 1.4768,
"grad_norm": 0.5605639890258036,
"learning_rate": 6.0125231877518205e-06,
"loss": 0.1745,
"step": 923
},
{
"epoch": 1.4784,
"grad_norm": 0.6111094109016982,
"learning_rate": 6.0034031713425636e-06,
"loss": 0.1746,
"step": 924
},
{
"epoch": 1.48,
"grad_norm": 0.6050555341264038,
"learning_rate": 5.994279675208302e-06,
"loss": 0.1475,
"step": 925
},
{
"epoch": 1.4816,
"grad_norm": 0.4818427880196344,
"learning_rate": 5.985152730988617e-06,
"loss": 0.1534,
"step": 926
},
{
"epoch": 1.4832,
"grad_norm": 1.0009042597155746,
"learning_rate": 5.9760223703350495e-06,
"loss": 0.1577,
"step": 927
},
{
"epoch": 1.4848,
"grad_norm": 0.5999536913098751,
"learning_rate": 5.966888624910989e-06,
"loss": 0.145,
"step": 928
},
{
"epoch": 1.4864,
"grad_norm": 0.5748903356491881,
"learning_rate": 5.957751526391558e-06,
"loss": 0.1394,
"step": 929
},
{
"epoch": 1.488,
"grad_norm": 0.5592583688906009,
"learning_rate": 5.948611106463518e-06,
"loss": 0.1653,
"step": 930
},
{
"epoch": 1.4896,
"grad_norm": 0.652109041733027,
"learning_rate": 5.939467396825137e-06,
"loss": 0.1676,
"step": 931
},
{
"epoch": 1.4912,
"grad_norm": 0.5275222704865555,
"learning_rate": 5.9303204291860975e-06,
"loss": 0.1537,
"step": 932
},
{
"epoch": 1.4928,
"grad_norm": 0.49523295226963077,
"learning_rate": 5.92117023526738e-06,
"loss": 0.1508,
"step": 933
},
{
"epoch": 1.4944,
"grad_norm": 0.5548087289012306,
"learning_rate": 5.912016846801153e-06,
"loss": 0.1616,
"step": 934
},
{
"epoch": 1.496,
"grad_norm": 0.5245425345634828,
"learning_rate": 5.902860295530665e-06,
"loss": 0.1489,
"step": 935
},
{
"epoch": 1.4976,
"grad_norm": 0.5008181114027274,
"learning_rate": 5.893700613210128e-06,
"loss": 0.1644,
"step": 936
},
{
"epoch": 1.4992,
"grad_norm": 0.6987493301547839,
"learning_rate": 5.88453783160462e-06,
"loss": 0.1561,
"step": 937
},
{
"epoch": 1.5008,
"grad_norm": 0.6553146935497204,
"learning_rate": 5.875371982489959e-06,
"loss": 0.1955,
"step": 938
},
{
"epoch": 1.5024,
"grad_norm": 0.5923450230892061,
"learning_rate": 5.866203097652605e-06,
"loss": 0.1486,
"step": 939
},
{
"epoch": 1.504,
"grad_norm": 0.6165262791634114,
"learning_rate": 5.857031208889548e-06,
"loss": 0.1614,
"step": 940
},
{
"epoch": 1.5056,
"grad_norm": 0.5433426322787046,
"learning_rate": 5.847856348008188e-06,
"loss": 0.1383,
"step": 941
},
{
"epoch": 1.5072,
"grad_norm": 0.47737207741102855,
"learning_rate": 5.838678546826242e-06,
"loss": 0.1563,
"step": 942
},
{
"epoch": 1.5088,
"grad_norm": 0.512689405450477,
"learning_rate": 5.829497837171616e-06,
"loss": 0.1795,
"step": 943
},
{
"epoch": 1.5104,
"grad_norm": 0.5674460676548242,
"learning_rate": 5.820314250882304e-06,
"loss": 0.1685,
"step": 944
},
{
"epoch": 1.512,
"grad_norm": 0.5160793953715768,
"learning_rate": 5.811127819806277e-06,
"loss": 0.1445,
"step": 945
},
{
"epoch": 1.5135999999999998,
"grad_norm": 0.5345491942221383,
"learning_rate": 5.801938575801372e-06,
"loss": 0.1551,
"step": 946
},
{
"epoch": 1.5152,
"grad_norm": 0.5757619952946275,
"learning_rate": 5.792746550735182e-06,
"loss": 0.1582,
"step": 947
},
{
"epoch": 1.5168,
"grad_norm": 0.6174589791928229,
"learning_rate": 5.7835517764849395e-06,
"loss": 0.165,
"step": 948
},
{
"epoch": 1.5184,
"grad_norm": 0.6016003086199918,
"learning_rate": 5.7743542849374155e-06,
"loss": 0.1415,
"step": 949
},
{
"epoch": 1.52,
"grad_norm": 0.5913220584757023,
"learning_rate": 5.765154107988803e-06,
"loss": 0.1451,
"step": 950
},
{
"epoch": 1.5215999999999998,
"grad_norm": 0.5680764781796656,
"learning_rate": 5.755951277544607e-06,
"loss": 0.1609,
"step": 951
},
{
"epoch": 1.5232,
"grad_norm": 0.549575985658153,
"learning_rate": 5.746745825519539e-06,
"loss": 0.1404,
"step": 952
},
{
"epoch": 1.5248,
"grad_norm": 0.49838090089444537,
"learning_rate": 5.737537783837395e-06,
"loss": 0.1713,
"step": 953
},
{
"epoch": 1.5264,
"grad_norm": 0.6034000185836031,
"learning_rate": 5.728327184430955e-06,
"loss": 0.1472,
"step": 954
},
{
"epoch": 1.528,
"grad_norm": 0.5473487424443946,
"learning_rate": 5.719114059241871e-06,
"loss": 0.1515,
"step": 955
},
{
"epoch": 1.5295999999999998,
"grad_norm": 0.5890724815269632,
"learning_rate": 5.709898440220552e-06,
"loss": 0.1549,
"step": 956
},
{
"epoch": 1.5312000000000001,
"grad_norm": 0.5689660309977727,
"learning_rate": 5.700680359326055e-06,
"loss": 0.1618,
"step": 957
},
{
"epoch": 1.5328,
"grad_norm": 0.6497643809638114,
"learning_rate": 5.691459848525977e-06,
"loss": 0.1321,
"step": 958
},
{
"epoch": 1.5344,
"grad_norm": 0.5864206877805881,
"learning_rate": 5.682236939796337e-06,
"loss": 0.1627,
"step": 959
},
{
"epoch": 1.536,
"grad_norm": 0.5220854083855688,
"learning_rate": 5.673011665121477e-06,
"loss": 0.1634,
"step": 960
},
{
"epoch": 1.5375999999999999,
"grad_norm": 0.4770724015769505,
"learning_rate": 5.663784056493936e-06,
"loss": 0.1632,
"step": 961
},
{
"epoch": 1.5392000000000001,
"grad_norm": 0.5448424890220768,
"learning_rate": 5.6545541459143535e-06,
"loss": 0.1427,
"step": 962
},
{
"epoch": 1.5408,
"grad_norm": 0.5634888021059619,
"learning_rate": 5.6453219653913495e-06,
"loss": 0.1538,
"step": 963
},
{
"epoch": 1.5424,
"grad_norm": 0.550479460086095,
"learning_rate": 5.636087546941413e-06,
"loss": 0.1384,
"step": 964
},
{
"epoch": 1.544,
"grad_norm": 0.49015354505862657,
"learning_rate": 5.6268509225888005e-06,
"loss": 0.1543,
"step": 965
},
{
"epoch": 1.5455999999999999,
"grad_norm": 0.5797817967979403,
"learning_rate": 5.617612124365411e-06,
"loss": 0.153,
"step": 966
},
{
"epoch": 1.5472000000000001,
"grad_norm": 0.5431613707156535,
"learning_rate": 5.608371184310688e-06,
"loss": 0.1487,
"step": 967
},
{
"epoch": 1.5488,
"grad_norm": 0.5954289503327552,
"learning_rate": 5.5991281344714984e-06,
"loss": 0.1621,
"step": 968
},
{
"epoch": 1.5504,
"grad_norm": 0.6268875595546433,
"learning_rate": 5.5898830069020325e-06,
"loss": 0.1535,
"step": 969
},
{
"epoch": 1.552,
"grad_norm": 0.5764322941969324,
"learning_rate": 5.580635833663679e-06,
"loss": 0.1421,
"step": 970
},
{
"epoch": 1.5535999999999999,
"grad_norm": 0.5281465161842288,
"learning_rate": 5.5713866468249235e-06,
"loss": 0.1654,
"step": 971
},
{
"epoch": 1.5552000000000001,
"grad_norm": 0.5420048214696946,
"learning_rate": 5.562135478461234e-06,
"loss": 0.1442,
"step": 972
},
{
"epoch": 1.5568,
"grad_norm": 0.7430654031549809,
"learning_rate": 5.55288236065495e-06,
"loss": 0.159,
"step": 973
},
{
"epoch": 1.5584,
"grad_norm": 0.5100462850336434,
"learning_rate": 5.5436273254951734e-06,
"loss": 0.1619,
"step": 974
},
{
"epoch": 1.56,
"grad_norm": 0.4761662737416338,
"learning_rate": 5.5343704050776535e-06,
"loss": 0.148,
"step": 975
},
{
"epoch": 1.5615999999999999,
"grad_norm": 0.5031242076822473,
"learning_rate": 5.5251116315046785e-06,
"loss": 0.1579,
"step": 976
},
{
"epoch": 1.5632000000000001,
"grad_norm": 0.46710008957176047,
"learning_rate": 5.515851036884964e-06,
"loss": 0.1283,
"step": 977
},
{
"epoch": 1.5648,
"grad_norm": 0.5814683968028678,
"learning_rate": 5.5065886533335355e-06,
"loss": 0.1722,
"step": 978
},
{
"epoch": 1.5664,
"grad_norm": 0.5529000930611109,
"learning_rate": 5.497324512971632e-06,
"loss": 0.1368,
"step": 979
},
{
"epoch": 1.568,
"grad_norm": 0.5777074047566567,
"learning_rate": 5.4880586479265774e-06,
"loss": 0.1488,
"step": 980
},
{
"epoch": 1.5695999999999999,
"grad_norm": 0.5095014247385022,
"learning_rate": 5.478791090331677e-06,
"loss": 0.1322,
"step": 981
},
{
"epoch": 1.5712000000000002,
"grad_norm": 0.5176107334221077,
"learning_rate": 5.4695218723261115e-06,
"loss": 0.1438,
"step": 982
},
{
"epoch": 1.5728,
"grad_norm": 0.6041370840372803,
"learning_rate": 5.46025102605481e-06,
"loss": 0.1567,
"step": 983
},
{
"epoch": 1.5744,
"grad_norm": 0.5276503605330135,
"learning_rate": 5.4509785836683606e-06,
"loss": 0.1492,
"step": 984
},
{
"epoch": 1.576,
"grad_norm": 0.5456581063904966,
"learning_rate": 5.441704577322877e-06,
"loss": 0.181,
"step": 985
},
{
"epoch": 1.5776,
"grad_norm": 0.4867577906303481,
"learning_rate": 5.4324290391798995e-06,
"loss": 0.1554,
"step": 986
},
{
"epoch": 1.5792000000000002,
"grad_norm": 0.6879700895288914,
"learning_rate": 5.423152001406282e-06,
"loss": 0.1445,
"step": 987
},
{
"epoch": 1.5808,
"grad_norm": 0.5771100859920616,
"learning_rate": 5.413873496174077e-06,
"loss": 0.1442,
"step": 988
},
{
"epoch": 1.5824,
"grad_norm": 0.5269332660185532,
"learning_rate": 5.404593555660424e-06,
"loss": 0.1366,
"step": 989
},
{
"epoch": 1.584,
"grad_norm": 0.5468050646240755,
"learning_rate": 5.39531221204745e-06,
"loss": 0.1631,
"step": 990
},
{
"epoch": 1.5856,
"grad_norm": 0.5950422010039307,
"learning_rate": 5.3860294975221335e-06,
"loss": 0.1783,
"step": 991
},
{
"epoch": 1.5872000000000002,
"grad_norm": 0.5280271811650628,
"learning_rate": 5.376745444276219e-06,
"loss": 0.1434,
"step": 992
},
{
"epoch": 1.5888,
"grad_norm": 0.5802973893217669,
"learning_rate": 5.3674600845060856e-06,
"loss": 0.1552,
"step": 993
},
{
"epoch": 1.5904,
"grad_norm": 0.57923826108895,
"learning_rate": 5.358173450412649e-06,
"loss": 0.1607,
"step": 994
},
{
"epoch": 1.592,
"grad_norm": 0.5437428597987385,
"learning_rate": 5.34888557420124e-06,
"loss": 0.1393,
"step": 995
},
{
"epoch": 1.5936,
"grad_norm": 0.6200014248689855,
"learning_rate": 5.339596488081501e-06,
"loss": 0.1569,
"step": 996
},
{
"epoch": 1.5952,
"grad_norm": 0.6935759520364619,
"learning_rate": 5.330306224267268e-06,
"loss": 0.1773,
"step": 997
},
{
"epoch": 1.5968,
"grad_norm": 2.119606937884837,
"learning_rate": 5.321014814976459e-06,
"loss": 0.1344,
"step": 998
},
{
"epoch": 1.5984,
"grad_norm": 0.598018632010377,
"learning_rate": 5.311722292430966e-06,
"loss": 0.1372,
"step": 999
},
{
"epoch": 1.6,
"grad_norm": 0.434843941317161,
"learning_rate": 5.302428688856544e-06,
"loss": 0.1301,
"step": 1000
},
{
"epoch": 1.6016,
"grad_norm": 0.6210468649065004,
"learning_rate": 5.293134036482697e-06,
"loss": 0.1713,
"step": 1001
},
{
"epoch": 1.6032,
"grad_norm": 0.7565529338559617,
"learning_rate": 5.283838367542562e-06,
"loss": 0.1279,
"step": 1002
},
{
"epoch": 1.6048,
"grad_norm": 0.5030840530629843,
"learning_rate": 5.274541714272805e-06,
"loss": 0.1328,
"step": 1003
},
{
"epoch": 1.6064,
"grad_norm": 0.5462206182913272,
"learning_rate": 5.265244108913503e-06,
"loss": 0.1408,
"step": 1004
},
{
"epoch": 1.608,
"grad_norm": 0.6816423534686401,
"learning_rate": 5.255945583708037e-06,
"loss": 0.1552,
"step": 1005
},
{
"epoch": 1.6096,
"grad_norm": 0.4576340574079896,
"learning_rate": 5.2466461709029755e-06,
"loss": 0.1267,
"step": 1006
},
{
"epoch": 1.6112,
"grad_norm": 0.6062951784508979,
"learning_rate": 5.237345902747969e-06,
"loss": 0.1519,
"step": 1007
},
{
"epoch": 1.6128,
"grad_norm": 0.5663362096909292,
"learning_rate": 5.228044811495632e-06,
"loss": 0.1284,
"step": 1008
},
{
"epoch": 1.6143999999999998,
"grad_norm": 0.5776896863378876,
"learning_rate": 5.218742929401432e-06,
"loss": 0.138,
"step": 1009
},
{
"epoch": 1.616,
"grad_norm": 0.6044987945150759,
"learning_rate": 5.2094402887235805e-06,
"loss": 0.1592,
"step": 1010
},
{
"epoch": 1.6176,
"grad_norm": 0.5816038633429572,
"learning_rate": 5.200136921722919e-06,
"loss": 0.1554,
"step": 1011
},
{
"epoch": 1.6192,
"grad_norm": 0.5368415472576993,
"learning_rate": 5.1908328606628114e-06,
"loss": 0.1551,
"step": 1012
},
{
"epoch": 1.6208,
"grad_norm": 0.4503704825593342,
"learning_rate": 5.181528137809023e-06,
"loss": 0.1206,
"step": 1013
},
{
"epoch": 1.6223999999999998,
"grad_norm": 0.5168666910381308,
"learning_rate": 5.1722227854296195e-06,
"loss": 0.1351,
"step": 1014
},
{
"epoch": 1.624,
"grad_norm": 0.5452190550410263,
"learning_rate": 5.162916835794843e-06,
"loss": 0.175,
"step": 1015
},
{
"epoch": 1.6256,
"grad_norm": 0.6070758563689632,
"learning_rate": 5.1536103211770135e-06,
"loss": 0.1432,
"step": 1016
},
{
"epoch": 1.6272,
"grad_norm": 0.5639999642393143,
"learning_rate": 5.14430327385041e-06,
"loss": 0.1738,
"step": 1017
},
{
"epoch": 1.6288,
"grad_norm": 0.6175381005420875,
"learning_rate": 5.134995726091152e-06,
"loss": 0.1455,
"step": 1018
},
{
"epoch": 1.6303999999999998,
"grad_norm": 0.6846837134675778,
"learning_rate": 5.1256877101771015e-06,
"loss": 0.1586,
"step": 1019
},
{
"epoch": 1.6320000000000001,
"grad_norm": 0.6738134485856369,
"learning_rate": 5.116379258387742e-06,
"loss": 0.1828,
"step": 1020
},
{
"epoch": 1.6336,
"grad_norm": 0.5595185721578609,
"learning_rate": 5.1070704030040675e-06,
"loss": 0.1428,
"step": 1021
},
{
"epoch": 1.6352,
"grad_norm": 0.4825499919823439,
"learning_rate": 5.097761176308471e-06,
"loss": 0.1359,
"step": 1022
},
{
"epoch": 1.6368,
"grad_norm": 0.46648507687026325,
"learning_rate": 5.088451610584638e-06,
"loss": 0.1295,
"step": 1023
},
{
"epoch": 1.6383999999999999,
"grad_norm": 0.6321115397285245,
"learning_rate": 5.079141738117423e-06,
"loss": 0.1589,
"step": 1024
},
{
"epoch": 1.6400000000000001,
"grad_norm": 0.6341827547232484,
"learning_rate": 5.06983159119275e-06,
"loss": 0.1362,
"step": 1025
},
{
"epoch": 1.6416,
"grad_norm": 0.6309257516615531,
"learning_rate": 5.060521202097491e-06,
"loss": 0.1691,
"step": 1026
},
{
"epoch": 1.6432,
"grad_norm": 0.6388624293554819,
"learning_rate": 5.051210603119358e-06,
"loss": 0.1522,
"step": 1027
},
{
"epoch": 1.6448,
"grad_norm": 0.5544418906642719,
"learning_rate": 5.041899826546791e-06,
"loss": 0.1435,
"step": 1028
},
{
"epoch": 1.6463999999999999,
"grad_norm": 0.5543950396767072,
"learning_rate": 5.032588904668851e-06,
"loss": 0.1402,
"step": 1029
},
{
"epoch": 1.6480000000000001,
"grad_norm": 0.4670790538253044,
"learning_rate": 5.023277869775097e-06,
"loss": 0.1274,
"step": 1030
},
{
"epoch": 1.6496,
"grad_norm": 0.5072929164199254,
"learning_rate": 5.013966754155482e-06,
"loss": 0.1487,
"step": 1031
},
{
"epoch": 1.6512,
"grad_norm": 0.5102664597123158,
"learning_rate": 5.004655590100238e-06,
"loss": 0.1436,
"step": 1032
},
{
"epoch": 1.6528,
"grad_norm": 0.534891531237293,
"learning_rate": 4.995344409899764e-06,
"loss": 0.1327,
"step": 1033
},
{
"epoch": 1.6543999999999999,
"grad_norm": 0.5750404739726822,
"learning_rate": 4.986033245844519e-06,
"loss": 0.1557,
"step": 1034
},
{
"epoch": 1.6560000000000001,
"grad_norm": 0.49728487115678194,
"learning_rate": 4.976722130224904e-06,
"loss": 0.1512,
"step": 1035
},
{
"epoch": 1.6576,
"grad_norm": 0.5615634112729294,
"learning_rate": 4.967411095331149e-06,
"loss": 0.1505,
"step": 1036
},
{
"epoch": 1.6592,
"grad_norm": 0.5184057455854084,
"learning_rate": 4.95810017345321e-06,
"loss": 0.1247,
"step": 1037
},
{
"epoch": 1.6608,
"grad_norm": 0.6000127741728973,
"learning_rate": 4.948789396880644e-06,
"loss": 0.1305,
"step": 1038
},
{
"epoch": 1.6623999999999999,
"grad_norm": 0.5566332993656999,
"learning_rate": 4.939478797902512e-06,
"loss": 0.1608,
"step": 1039
},
{
"epoch": 1.6640000000000001,
"grad_norm": 0.6332750992380535,
"learning_rate": 4.930168408807252e-06,
"loss": 0.1618,
"step": 1040
},
{
"epoch": 1.6656,
"grad_norm": 0.5707147075839376,
"learning_rate": 4.920858261882578e-06,
"loss": 0.1515,
"step": 1041
},
{
"epoch": 1.6672,
"grad_norm": 0.49964553963105474,
"learning_rate": 4.911548389415363e-06,
"loss": 0.1455,
"step": 1042
},
{
"epoch": 1.6688,
"grad_norm": 0.55013018082513,
"learning_rate": 4.9022388236915306e-06,
"loss": 0.15,
"step": 1043
},
{
"epoch": 1.6703999999999999,
"grad_norm": 0.4989862513730279,
"learning_rate": 4.892929596995934e-06,
"loss": 0.1361,
"step": 1044
},
{
"epoch": 1.6720000000000002,
"grad_norm": 0.5938870189814183,
"learning_rate": 4.883620741612259e-06,
"loss": 0.153,
"step": 1045
},
{
"epoch": 1.6736,
"grad_norm": 0.7945971306105271,
"learning_rate": 4.8743122898229e-06,
"loss": 0.1424,
"step": 1046
},
{
"epoch": 1.6752,
"grad_norm": 0.5351186257354851,
"learning_rate": 4.865004273908851e-06,
"loss": 0.1358,
"step": 1047
},
{
"epoch": 1.6768,
"grad_norm": 0.4525182330642793,
"learning_rate": 4.855696726149593e-06,
"loss": 0.1423,
"step": 1048
},
{
"epoch": 1.6784,
"grad_norm": 0.5726462642548055,
"learning_rate": 4.846389678822987e-06,
"loss": 0.1612,
"step": 1049
},
{
"epoch": 1.6800000000000002,
"grad_norm": 0.64359822665452,
"learning_rate": 4.837083164205159e-06,
"loss": 0.1526,
"step": 1050
},
{
"epoch": 1.6816,
"grad_norm": 0.41924847937452997,
"learning_rate": 4.827777214570384e-06,
"loss": 0.1373,
"step": 1051
},
{
"epoch": 1.6832,
"grad_norm": 0.5703091103281263,
"learning_rate": 4.818471862190979e-06,
"loss": 0.1593,
"step": 1052
},
{
"epoch": 1.6848,
"grad_norm": 0.5132692358499653,
"learning_rate": 4.809167139337191e-06,
"loss": 0.1456,
"step": 1053
},
{
"epoch": 1.6864,
"grad_norm": 0.6555935155675325,
"learning_rate": 4.799863078277082e-06,
"loss": 0.1588,
"step": 1054
},
{
"epoch": 1.688,
"grad_norm": 0.5162971067239667,
"learning_rate": 4.790559711276422e-06,
"loss": 0.1261,
"step": 1055
},
{
"epoch": 1.6896,
"grad_norm": 0.5500167098905193,
"learning_rate": 4.781257070598571e-06,
"loss": 0.1765,
"step": 1056
},
{
"epoch": 1.6912,
"grad_norm": 0.5867029208444309,
"learning_rate": 4.771955188504371e-06,
"loss": 0.17,
"step": 1057
},
{
"epoch": 1.6928,
"grad_norm": 0.5837859900116088,
"learning_rate": 4.762654097252033e-06,
"loss": 0.1646,
"step": 1058
},
{
"epoch": 1.6944,
"grad_norm": 0.6012610066783571,
"learning_rate": 4.753353829097025e-06,
"loss": 0.1708,
"step": 1059
},
{
"epoch": 1.696,
"grad_norm": 0.5591457963060675,
"learning_rate": 4.7440544162919645e-06,
"loss": 0.1484,
"step": 1060
},
{
"epoch": 1.6976,
"grad_norm": 0.5699413788965874,
"learning_rate": 4.734755891086498e-06,
"loss": 0.15,
"step": 1061
},
{
"epoch": 1.6992,
"grad_norm": 0.5673409756283209,
"learning_rate": 4.725458285727195e-06,
"loss": 0.159,
"step": 1062
},
{
"epoch": 1.7008,
"grad_norm": 0.5828123340888676,
"learning_rate": 4.716161632457438e-06,
"loss": 0.1693,
"step": 1063
},
{
"epoch": 1.7024,
"grad_norm": 0.6373884450885641,
"learning_rate": 4.7068659635173034e-06,
"loss": 0.1785,
"step": 1064
},
{
"epoch": 1.704,
"grad_norm": 0.5582686576560517,
"learning_rate": 4.6975713111434556e-06,
"loss": 0.1627,
"step": 1065
},
{
"epoch": 1.7056,
"grad_norm": 0.504669660734267,
"learning_rate": 4.688277707569035e-06,
"loss": 0.1454,
"step": 1066
},
{
"epoch": 1.7072,
"grad_norm": 0.5421465640724115,
"learning_rate": 4.678985185023542e-06,
"loss": 0.1579,
"step": 1067
},
{
"epoch": 1.7088,
"grad_norm": 0.5906370282739937,
"learning_rate": 4.669693775732733e-06,
"loss": 0.1362,
"step": 1068
},
{
"epoch": 1.7104,
"grad_norm": 0.45391188693563655,
"learning_rate": 4.660403511918499e-06,
"loss": 0.1553,
"step": 1069
},
{
"epoch": 1.712,
"grad_norm": 0.5274830084578019,
"learning_rate": 4.65111442579876e-06,
"loss": 0.1433,
"step": 1070
},
{
"epoch": 1.7136,
"grad_norm": 0.5615290456145285,
"learning_rate": 4.641826549587352e-06,
"loss": 0.1519,
"step": 1071
},
{
"epoch": 1.7151999999999998,
"grad_norm": 0.6398198211082029,
"learning_rate": 4.632539915493915e-06,
"loss": 0.1419,
"step": 1072
},
{
"epoch": 1.7168,
"grad_norm": 0.6106262391808214,
"learning_rate": 4.623254555723783e-06,
"loss": 0.1784,
"step": 1073
},
{
"epoch": 1.7184,
"grad_norm": 0.7744977724497079,
"learning_rate": 4.613970502477867e-06,
"loss": 0.2017,
"step": 1074
},
{
"epoch": 1.72,
"grad_norm": 0.7755010921629375,
"learning_rate": 4.604687787952552e-06,
"loss": 0.1622,
"step": 1075
},
{
"epoch": 1.7216,
"grad_norm": 0.5621036718255097,
"learning_rate": 4.5954064443395765e-06,
"loss": 0.1419,
"step": 1076
},
{
"epoch": 1.7231999999999998,
"grad_norm": 0.6379666360795156,
"learning_rate": 4.586126503825925e-06,
"loss": 0.1765,
"step": 1077
},
{
"epoch": 1.7248,
"grad_norm": 0.5645087223518565,
"learning_rate": 4.57684799859372e-06,
"loss": 0.1353,
"step": 1078
},
{
"epoch": 1.7264,
"grad_norm": 0.5975176537754664,
"learning_rate": 4.567570960820101e-06,
"loss": 0.1804,
"step": 1079
},
{
"epoch": 1.728,
"grad_norm": 0.5658181117102297,
"learning_rate": 4.558295422677124e-06,
"loss": 0.1535,
"step": 1080
},
{
"epoch": 1.7296,
"grad_norm": 0.5923205116373064,
"learning_rate": 4.54902141633164e-06,
"loss": 0.1719,
"step": 1081
},
{
"epoch": 1.7311999999999999,
"grad_norm": 0.5060424429304817,
"learning_rate": 4.539748973945191e-06,
"loss": 0.1417,
"step": 1082
},
{
"epoch": 1.7328000000000001,
"grad_norm": 0.5481319567236494,
"learning_rate": 4.53047812767389e-06,
"loss": 0.1239,
"step": 1083
},
{
"epoch": 1.7344,
"grad_norm": 0.5446158867414732,
"learning_rate": 4.5212089096683234e-06,
"loss": 0.1439,
"step": 1084
},
{
"epoch": 1.736,
"grad_norm": 0.575261403362818,
"learning_rate": 4.511941352073424e-06,
"loss": 0.1622,
"step": 1085
},
{
"epoch": 1.7376,
"grad_norm": 0.5401889081739879,
"learning_rate": 4.5026754870283695e-06,
"loss": 0.1463,
"step": 1086
},
{
"epoch": 1.7391999999999999,
"grad_norm": 0.551054503389811,
"learning_rate": 4.493411346666465e-06,
"loss": 0.1492,
"step": 1087
},
{
"epoch": 1.7408000000000001,
"grad_norm": 0.574159784159435,
"learning_rate": 4.484148963115038e-06,
"loss": 0.1746,
"step": 1088
},
{
"epoch": 1.7424,
"grad_norm": 0.5181416723954063,
"learning_rate": 4.474888368495322e-06,
"loss": 0.123,
"step": 1089
},
{
"epoch": 1.744,
"grad_norm": 0.4889927670637744,
"learning_rate": 4.465629594922348e-06,
"loss": 0.149,
"step": 1090
},
{
"epoch": 1.7456,
"grad_norm": 0.6253591582357048,
"learning_rate": 4.456372674504828e-06,
"loss": 0.1561,
"step": 1091
},
{
"epoch": 1.7471999999999999,
"grad_norm": 0.49811080971784044,
"learning_rate": 4.447117639345052e-06,
"loss": 0.1409,
"step": 1092
},
{
"epoch": 1.7488000000000001,
"grad_norm": 0.5875722630354159,
"learning_rate": 4.437864521538768e-06,
"loss": 0.1694,
"step": 1093
},
{
"epoch": 1.7504,
"grad_norm": 0.47545254750852645,
"learning_rate": 4.428613353175078e-06,
"loss": 0.1527,
"step": 1094
},
{
"epoch": 1.752,
"grad_norm": 0.6132312422718952,
"learning_rate": 4.4193641663363214e-06,
"loss": 0.1604,
"step": 1095
},
{
"epoch": 1.7536,
"grad_norm": 0.5838116668483884,
"learning_rate": 4.410116993097968e-06,
"loss": 0.1742,
"step": 1096
},
{
"epoch": 1.7551999999999999,
"grad_norm": 0.5739353888617148,
"learning_rate": 4.400871865528502e-06,
"loss": 0.1268,
"step": 1097
},
{
"epoch": 1.7568000000000001,
"grad_norm": 0.5547254380078593,
"learning_rate": 4.391628815689314e-06,
"loss": 0.132,
"step": 1098
},
{
"epoch": 1.7584,
"grad_norm": 0.6450753500376499,
"learning_rate": 4.382387875634592e-06,
"loss": 0.19,
"step": 1099
},
{
"epoch": 1.76,
"grad_norm": 0.6297316489989471,
"learning_rate": 4.373149077411203e-06,
"loss": 0.1589,
"step": 1100
},
{
"epoch": 1.7616,
"grad_norm": 0.4956570946445073,
"learning_rate": 4.363912453058589e-06,
"loss": 0.129,
"step": 1101
},
{
"epoch": 1.7631999999999999,
"grad_norm": 0.47066284232886313,
"learning_rate": 4.354678034608654e-06,
"loss": 0.1432,
"step": 1102
},
{
"epoch": 1.7648000000000001,
"grad_norm": 0.5762049143646805,
"learning_rate": 4.345445854085649e-06,
"loss": 0.1391,
"step": 1103
},
{
"epoch": 1.7664,
"grad_norm": 0.6618154136373311,
"learning_rate": 4.336215943506066e-06,
"loss": 0.1531,
"step": 1104
},
{
"epoch": 1.768,
"grad_norm": 0.5632165205243514,
"learning_rate": 4.326988334878526e-06,
"loss": 0.1667,
"step": 1105
},
{
"epoch": 1.7696,
"grad_norm": 0.5551559730122675,
"learning_rate": 4.317763060203665e-06,
"loss": 0.1374,
"step": 1106
},
{
"epoch": 1.7711999999999999,
"grad_norm": 0.5842111113157518,
"learning_rate": 4.308540151474027e-06,
"loss": 0.1709,
"step": 1107
},
{
"epoch": 1.7728000000000002,
"grad_norm": 0.5362820641528826,
"learning_rate": 4.299319640673948e-06,
"loss": 0.1584,
"step": 1108
},
{
"epoch": 1.7744,
"grad_norm": 0.46994692475077166,
"learning_rate": 4.290101559779451e-06,
"loss": 0.1608,
"step": 1109
},
{
"epoch": 1.776,
"grad_norm": 0.5311789145433026,
"learning_rate": 4.280885940758131e-06,
"loss": 0.1491,
"step": 1110
},
{
"epoch": 1.7776,
"grad_norm": 0.5467368739694822,
"learning_rate": 4.271672815569047e-06,
"loss": 0.1456,
"step": 1111
},
{
"epoch": 1.7792,
"grad_norm": 0.6387088875694875,
"learning_rate": 4.262462216162606e-06,
"loss": 0.1453,
"step": 1112
},
{
"epoch": 1.7808000000000002,
"grad_norm": 0.46474324561408825,
"learning_rate": 4.253254174480462e-06,
"loss": 0.1405,
"step": 1113
},
{
"epoch": 1.7824,
"grad_norm": 0.5467344791657153,
"learning_rate": 4.244048722455393e-06,
"loss": 0.1462,
"step": 1114
},
{
"epoch": 1.784,
"grad_norm": 0.4978488172283637,
"learning_rate": 4.234845892011198e-06,
"loss": 0.1645,
"step": 1115
},
{
"epoch": 1.7856,
"grad_norm": 0.4878277443646257,
"learning_rate": 4.225645715062585e-06,
"loss": 0.1358,
"step": 1116
},
{
"epoch": 1.7872,
"grad_norm": 0.5472507121867703,
"learning_rate": 4.216448223515061e-06,
"loss": 0.1639,
"step": 1117
},
{
"epoch": 1.7888,
"grad_norm": 0.5904234289030031,
"learning_rate": 4.2072534492648184e-06,
"loss": 0.1393,
"step": 1118
},
{
"epoch": 1.7904,
"grad_norm": 0.5590571229144006,
"learning_rate": 4.198061424198627e-06,
"loss": 0.1546,
"step": 1119
},
{
"epoch": 1.792,
"grad_norm": 0.5282918459750269,
"learning_rate": 4.188872180193723e-06,
"loss": 0.1566,
"step": 1120
},
{
"epoch": 1.7936,
"grad_norm": 0.5971210953457523,
"learning_rate": 4.179685749117698e-06,
"loss": 0.159,
"step": 1121
},
{
"epoch": 1.7952,
"grad_norm": 0.5645737272399435,
"learning_rate": 4.170502162828385e-06,
"loss": 0.1594,
"step": 1122
},
{
"epoch": 1.7968,
"grad_norm": 0.5916158666862895,
"learning_rate": 4.161321453173759e-06,
"loss": 0.1594,
"step": 1123
},
{
"epoch": 1.7984,
"grad_norm": 0.45117265340652973,
"learning_rate": 4.152143651991812e-06,
"loss": 0.1528,
"step": 1124
},
{
"epoch": 1.8,
"grad_norm": 0.6840157221545611,
"learning_rate": 4.142968791110455e-06,
"loss": 0.167,
"step": 1125
},
{
"epoch": 1.8016,
"grad_norm": 0.6144945633178218,
"learning_rate": 4.133796902347397e-06,
"loss": 0.1365,
"step": 1126
},
{
"epoch": 1.8032,
"grad_norm": 0.5391602325633188,
"learning_rate": 4.124628017510043e-06,
"loss": 0.1518,
"step": 1127
},
{
"epoch": 1.8048,
"grad_norm": 0.5648800256244909,
"learning_rate": 4.115462168395382e-06,
"loss": 0.142,
"step": 1128
},
{
"epoch": 1.8064,
"grad_norm": 0.5582799889567649,
"learning_rate": 4.106299386789873e-06,
"loss": 0.1428,
"step": 1129
},
{
"epoch": 1.808,
"grad_norm": 0.45546158844511153,
"learning_rate": 4.097139704469337e-06,
"loss": 0.1506,
"step": 1130
},
{
"epoch": 1.8096,
"grad_norm": 0.5158500311316843,
"learning_rate": 4.0879831531988485e-06,
"loss": 0.1447,
"step": 1131
},
{
"epoch": 1.8112,
"grad_norm": 0.6121030702714039,
"learning_rate": 4.078829764732621e-06,
"loss": 0.1716,
"step": 1132
},
{
"epoch": 1.8128,
"grad_norm": 0.48790923688154414,
"learning_rate": 4.069679570813903e-06,
"loss": 0.1419,
"step": 1133
},
{
"epoch": 1.8144,
"grad_norm": 0.49833768912955706,
"learning_rate": 4.060532603174865e-06,
"loss": 0.1298,
"step": 1134
},
{
"epoch": 1.8159999999999998,
"grad_norm": 0.5940353144465385,
"learning_rate": 4.051388893536484e-06,
"loss": 0.1276,
"step": 1135
},
{
"epoch": 1.8176,
"grad_norm": 0.5749055916597056,
"learning_rate": 4.042248473608442e-06,
"loss": 0.1655,
"step": 1136
},
{
"epoch": 1.8192,
"grad_norm": 0.601304986223328,
"learning_rate": 4.033111375089013e-06,
"loss": 0.1656,
"step": 1137
},
{
"epoch": 1.8208,
"grad_norm": 0.5436322058476164,
"learning_rate": 4.023977629664951e-06,
"loss": 0.1609,
"step": 1138
},
{
"epoch": 1.8224,
"grad_norm": 0.5369126585279858,
"learning_rate": 4.0148472690113845e-06,
"loss": 0.1527,
"step": 1139
},
{
"epoch": 1.8239999999999998,
"grad_norm": 0.5371126437672011,
"learning_rate": 4.0057203247917e-06,
"loss": 0.1395,
"step": 1140
},
{
"epoch": 1.8256000000000001,
"grad_norm": 0.47824337984541443,
"learning_rate": 3.996596828657437e-06,
"loss": 0.1496,
"step": 1141
},
{
"epoch": 1.8272,
"grad_norm": 0.5048332154968177,
"learning_rate": 3.987476812248181e-06,
"loss": 0.1486,
"step": 1142
},
{
"epoch": 1.8288,
"grad_norm": 0.4993966001290576,
"learning_rate": 3.978360307191444e-06,
"loss": 0.145,
"step": 1143
},
{
"epoch": 1.8304,
"grad_norm": 0.5449026362320392,
"learning_rate": 3.969247345102567e-06,
"loss": 0.1474,
"step": 1144
},
{
"epoch": 1.8319999999999999,
"grad_norm": 0.5379892097748908,
"learning_rate": 3.960137957584601e-06,
"loss": 0.1563,
"step": 1145
},
{
"epoch": 1.8336000000000001,
"grad_norm": 0.5074611873476278,
"learning_rate": 3.9510321762282e-06,
"loss": 0.1529,
"step": 1146
},
{
"epoch": 1.8352,
"grad_norm": 0.41739174447444305,
"learning_rate": 3.941930032611513e-06,
"loss": 0.1231,
"step": 1147
},
{
"epoch": 1.8368,
"grad_norm": 0.4390596333219584,
"learning_rate": 3.932831558300074e-06,
"loss": 0.136,
"step": 1148
},
{
"epoch": 1.8384,
"grad_norm": 0.5056064931187001,
"learning_rate": 3.923736784846693e-06,
"loss": 0.1427,
"step": 1149
},
{
"epoch": 1.8399999999999999,
"grad_norm": 0.5360752230971821,
"learning_rate": 3.914645743791346e-06,
"loss": 0.1441,
"step": 1150
},
{
"epoch": 1.8416000000000001,
"grad_norm": 0.5681398659807825,
"learning_rate": 3.90555846666106e-06,
"loss": 0.1669,
"step": 1151
},
{
"epoch": 1.8432,
"grad_norm": 0.5648832309701712,
"learning_rate": 3.896474984969817e-06,
"loss": 0.1678,
"step": 1152
},
{
"epoch": 1.8448,
"grad_norm": 0.5728057323886478,
"learning_rate": 3.887395330218429e-06,
"loss": 0.1572,
"step": 1153
},
{
"epoch": 1.8464,
"grad_norm": 0.48622959165707924,
"learning_rate": 3.878319533894443e-06,
"loss": 0.1563,
"step": 1154
},
{
"epoch": 1.8479999999999999,
"grad_norm": 0.541891863292097,
"learning_rate": 3.869247627472021e-06,
"loss": 0.1625,
"step": 1155
},
{
"epoch": 1.8496000000000001,
"grad_norm": 0.6313115633021836,
"learning_rate": 3.860179642411838e-06,
"loss": 0.1778,
"step": 1156
},
{
"epoch": 1.8512,
"grad_norm": 0.5573084867562419,
"learning_rate": 3.851115610160967e-06,
"loss": 0.1473,
"step": 1157
},
{
"epoch": 1.8528,
"grad_norm": 0.5886813218514988,
"learning_rate": 3.842055562152775e-06,
"loss": 0.1549,
"step": 1158
},
{
"epoch": 1.8544,
"grad_norm": 0.4837158791412468,
"learning_rate": 3.8329995298068114e-06,
"loss": 0.1383,
"step": 1159
},
{
"epoch": 1.8559999999999999,
"grad_norm": 0.5208123862437744,
"learning_rate": 3.8239475445287015e-06,
"loss": 0.165,
"step": 1160
},
{
"epoch": 1.8576000000000001,
"grad_norm": 0.5328737371790516,
"learning_rate": 3.814899637710031e-06,
"loss": 0.1655,
"step": 1161
},
{
"epoch": 1.8592,
"grad_norm": 0.5191910560103351,
"learning_rate": 3.8058558407282465e-06,
"loss": 0.1538,
"step": 1162
},
{
"epoch": 1.8608,
"grad_norm": 0.5262222884973183,
"learning_rate": 3.7968161849465395e-06,
"loss": 0.166,
"step": 1163
},
{
"epoch": 1.8624,
"grad_norm": 0.44935212276312425,
"learning_rate": 3.78778070171374e-06,
"loss": 0.144,
"step": 1164
},
{
"epoch": 1.8639999999999999,
"grad_norm": 0.5804161058913064,
"learning_rate": 3.7787494223642096e-06,
"loss": 0.1724,
"step": 1165
},
{
"epoch": 1.8656000000000001,
"grad_norm": 0.5680580352400117,
"learning_rate": 3.7697223782177304e-06,
"loss": 0.1521,
"step": 1166
},
{
"epoch": 1.8672,
"grad_norm": 0.6090438937904407,
"learning_rate": 3.760699600579399e-06,
"loss": 0.1564,
"step": 1167
},
{
"epoch": 1.8688,
"grad_norm": 0.6203057152306043,
"learning_rate": 3.7516811207395116e-06,
"loss": 0.1569,
"step": 1168
},
{
"epoch": 1.8704,
"grad_norm": 0.5669580338963477,
"learning_rate": 3.742666969973463e-06,
"loss": 0.1458,
"step": 1169
},
{
"epoch": 1.8719999999999999,
"grad_norm": 0.5929690326988141,
"learning_rate": 3.733657179541635e-06,
"loss": 0.1635,
"step": 1170
},
{
"epoch": 1.8736000000000002,
"grad_norm": 0.6111980945025998,
"learning_rate": 3.724651780689286e-06,
"loss": 0.1709,
"step": 1171
},
{
"epoch": 1.8752,
"grad_norm": 0.5758206055570084,
"learning_rate": 3.715650804646449e-06,
"loss": 0.1425,
"step": 1172
},
{
"epoch": 1.8768,
"grad_norm": 0.5574715711895325,
"learning_rate": 3.7066542826278153e-06,
"loss": 0.1617,
"step": 1173
},
{
"epoch": 1.8784,
"grad_norm": 0.6088474601944908,
"learning_rate": 3.6976622458326308e-06,
"loss": 0.1525,
"step": 1174
},
{
"epoch": 1.88,
"grad_norm": 0.629125289484949,
"learning_rate": 3.6886747254445877e-06,
"loss": 0.1571,
"step": 1175
},
{
"epoch": 1.8816000000000002,
"grad_norm": 0.5360350349530443,
"learning_rate": 3.6796917526317153e-06,
"loss": 0.1494,
"step": 1176
},
{
"epoch": 1.8832,
"grad_norm": 0.4813628582407898,
"learning_rate": 3.6707133585462713e-06,
"loss": 0.1343,
"step": 1177
},
{
"epoch": 1.8848,
"grad_norm": 0.5619353254121278,
"learning_rate": 3.6617395743246375e-06,
"loss": 0.1418,
"step": 1178
},
{
"epoch": 1.8864,
"grad_norm": 0.46956104366697576,
"learning_rate": 3.652770431087206e-06,
"loss": 0.1624,
"step": 1179
},
{
"epoch": 1.888,
"grad_norm": 0.5205570918620398,
"learning_rate": 3.6438059599382765e-06,
"loss": 0.1317,
"step": 1180
},
{
"epoch": 1.8896,
"grad_norm": 0.5340970796663593,
"learning_rate": 3.634846191965944e-06,
"loss": 0.1607,
"step": 1181
},
{
"epoch": 1.8912,
"grad_norm": 0.6394861221997985,
"learning_rate": 3.625891158241994e-06,
"loss": 0.1506,
"step": 1182
},
{
"epoch": 1.8928,
"grad_norm": 0.6162441458777921,
"learning_rate": 3.6169408898217973e-06,
"loss": 0.1552,
"step": 1183
},
{
"epoch": 1.8944,
"grad_norm": 0.5424684347535607,
"learning_rate": 3.6079954177441945e-06,
"loss": 0.1161,
"step": 1184
},
{
"epoch": 1.896,
"grad_norm": 0.6033445508651539,
"learning_rate": 3.599054773031394e-06,
"loss": 0.161,
"step": 1185
},
{
"epoch": 1.8976,
"grad_norm": 0.48506147577639774,
"learning_rate": 3.5901189866888654e-06,
"loss": 0.1441,
"step": 1186
},
{
"epoch": 1.8992,
"grad_norm": 0.6220741323856972,
"learning_rate": 3.581188089705226e-06,
"loss": 0.1461,
"step": 1187
},
{
"epoch": 1.9008,
"grad_norm": 0.5415714865899466,
"learning_rate": 3.572262113052142e-06,
"loss": 0.1253,
"step": 1188
},
{
"epoch": 1.9024,
"grad_norm": 0.5589115267643062,
"learning_rate": 3.563341087684213e-06,
"loss": 0.131,
"step": 1189
},
{
"epoch": 1.904,
"grad_norm": 0.4719049950452641,
"learning_rate": 3.554425044538868e-06,
"loss": 0.1301,
"step": 1190
},
{
"epoch": 1.9056,
"grad_norm": 0.5341342657868913,
"learning_rate": 3.5455140145362587e-06,
"loss": 0.1567,
"step": 1191
},
{
"epoch": 1.9072,
"grad_norm": 0.5393219953660934,
"learning_rate": 3.5366080285791516e-06,
"loss": 0.1459,
"step": 1192
},
{
"epoch": 1.9088,
"grad_norm": 0.6161673024827632,
"learning_rate": 3.527707117552822e-06,
"loss": 0.1393,
"step": 1193
},
{
"epoch": 1.9104,
"grad_norm": 0.5528858066287445,
"learning_rate": 3.5188113123249435e-06,
"loss": 0.1736,
"step": 1194
},
{
"epoch": 1.912,
"grad_norm": 0.64792187316954,
"learning_rate": 3.5099206437454852e-06,
"loss": 0.164,
"step": 1195
},
{
"epoch": 1.9136,
"grad_norm": 0.5118602571349121,
"learning_rate": 3.5010351426466006e-06,
"loss": 0.1497,
"step": 1196
},
{
"epoch": 1.9152,
"grad_norm": 0.6668269787294003,
"learning_rate": 3.4921548398425246e-06,
"loss": 0.1555,
"step": 1197
},
{
"epoch": 1.9167999999999998,
"grad_norm": 0.5066295835870492,
"learning_rate": 3.4832797661294633e-06,
"loss": 0.1344,
"step": 1198
},
{
"epoch": 1.9184,
"grad_norm": 0.5919476639653518,
"learning_rate": 3.4744099522854914e-06,
"loss": 0.152,
"step": 1199
},
{
"epoch": 1.92,
"grad_norm": 0.4623571458903057,
"learning_rate": 3.4655454290704393e-06,
"loss": 0.1285,
"step": 1200
},
{
"epoch": 1.9216,
"grad_norm": 0.45500270007117355,
"learning_rate": 3.4566862272257923e-06,
"loss": 0.1254,
"step": 1201
},
{
"epoch": 1.9232,
"grad_norm": 0.6468194510804908,
"learning_rate": 3.44783237747458e-06,
"loss": 0.1485,
"step": 1202
},
{
"epoch": 1.9247999999999998,
"grad_norm": 0.5768858767514213,
"learning_rate": 3.438983910521273e-06,
"loss": 0.1132,
"step": 1203
},
{
"epoch": 1.9264000000000001,
"grad_norm": 0.5293773127766422,
"learning_rate": 3.430140857051675e-06,
"loss": 0.1647,
"step": 1204
},
{
"epoch": 1.928,
"grad_norm": 0.5398552616516444,
"learning_rate": 3.421303247732813e-06,
"loss": 0.1359,
"step": 1205
},
{
"epoch": 1.9296,
"grad_norm": 0.6117195843783928,
"learning_rate": 3.4124711132128374e-06,
"loss": 0.159,
"step": 1206
},
{
"epoch": 1.9312,
"grad_norm": 0.5634281177652379,
"learning_rate": 3.4036444841209113e-06,
"loss": 0.1603,
"step": 1207
},
{
"epoch": 1.9327999999999999,
"grad_norm": 0.5950789674093822,
"learning_rate": 3.3948233910671036e-06,
"loss": 0.1227,
"step": 1208
},
{
"epoch": 1.9344000000000001,
"grad_norm": 0.514753411115372,
"learning_rate": 3.3860078646422894e-06,
"loss": 0.13,
"step": 1209
},
{
"epoch": 1.936,
"grad_norm": 0.5058728544474614,
"learning_rate": 3.3771979354180343e-06,
"loss": 0.1379,
"step": 1210
},
{
"epoch": 1.9376,
"grad_norm": 0.5621859944254937,
"learning_rate": 3.3683936339464957e-06,
"loss": 0.1489,
"step": 1211
},
{
"epoch": 1.9392,
"grad_norm": 0.6030285925469078,
"learning_rate": 3.359594990760313e-06,
"loss": 0.1531,
"step": 1212
},
{
"epoch": 1.9407999999999999,
"grad_norm": 0.7263070275606974,
"learning_rate": 3.3508020363725043e-06,
"loss": 0.1624,
"step": 1213
},
{
"epoch": 1.9424000000000001,
"grad_norm": 0.7069528500845071,
"learning_rate": 3.34201480127636e-06,
"loss": 0.1456,
"step": 1214
},
{
"epoch": 1.944,
"grad_norm": 0.6634405058615528,
"learning_rate": 3.333233315945333e-06,
"loss": 0.1564,
"step": 1215
},
{
"epoch": 1.9456,
"grad_norm": 0.5505517296772173,
"learning_rate": 3.324457610832942e-06,
"loss": 0.133,
"step": 1216
},
{
"epoch": 1.9472,
"grad_norm": 0.5055000429287526,
"learning_rate": 3.315687716372655e-06,
"loss": 0.1618,
"step": 1217
},
{
"epoch": 1.9487999999999999,
"grad_norm": 0.4199928096467762,
"learning_rate": 3.306923662977789e-06,
"loss": 0.1392,
"step": 1218
},
{
"epoch": 1.9504000000000001,
"grad_norm": 0.5141327706511539,
"learning_rate": 3.2981654810414128e-06,
"loss": 0.1612,
"step": 1219
},
{
"epoch": 1.952,
"grad_norm": 0.5835451067896915,
"learning_rate": 3.2894132009362245e-06,
"loss": 0.1474,
"step": 1220
},
{
"epoch": 1.9536,
"grad_norm": 0.5757731649735447,
"learning_rate": 3.280666853014457e-06,
"loss": 0.1449,
"step": 1221
},
{
"epoch": 1.9552,
"grad_norm": 0.4916938579821752,
"learning_rate": 3.271926467607774e-06,
"loss": 0.1309,
"step": 1222
},
{
"epoch": 1.9567999999999999,
"grad_norm": 0.5267469536859699,
"learning_rate": 3.2631920750271594e-06,
"loss": 0.1379,
"step": 1223
},
{
"epoch": 1.9584000000000001,
"grad_norm": 0.5056655908272999,
"learning_rate": 3.2544637055628135e-06,
"loss": 0.1367,
"step": 1224
},
{
"epoch": 1.96,
"grad_norm": 0.5329618646916798,
"learning_rate": 3.2457413894840516e-06,
"loss": 0.1491,
"step": 1225
},
{
"epoch": 1.9616,
"grad_norm": 0.5510240384949934,
"learning_rate": 3.2370251570391933e-06,
"loss": 0.1495,
"step": 1226
},
{
"epoch": 1.9632,
"grad_norm": 0.4943356938896233,
"learning_rate": 3.2283150384554642e-06,
"loss": 0.1553,
"step": 1227
},
{
"epoch": 1.9647999999999999,
"grad_norm": 0.5512771427186363,
"learning_rate": 3.219611063938883e-06,
"loss": 0.1421,
"step": 1228
},
{
"epoch": 1.9664000000000001,
"grad_norm": 0.523127577596483,
"learning_rate": 3.210913263674166e-06,
"loss": 0.1379,
"step": 1229
},
{
"epoch": 1.968,
"grad_norm": 0.4777885177258767,
"learning_rate": 3.2022216678246145e-06,
"loss": 0.1382,
"step": 1230
},
{
"epoch": 1.9696,
"grad_norm": 0.6054036973505424,
"learning_rate": 3.193536306532013e-06,
"loss": 0.1677,
"step": 1231
},
{
"epoch": 1.9712,
"grad_norm": 0.6293170540326982,
"learning_rate": 3.184857209916528e-06,
"loss": 0.1284,
"step": 1232
},
{
"epoch": 1.9727999999999999,
"grad_norm": 0.5320629738784454,
"learning_rate": 3.1761844080765993e-06,
"loss": 0.1546,
"step": 1233
},
{
"epoch": 1.9744000000000002,
"grad_norm": 0.4771451543962523,
"learning_rate": 3.1675179310888344e-06,
"loss": 0.1524,
"step": 1234
},
{
"epoch": 1.976,
"grad_norm": 0.5949360004308281,
"learning_rate": 3.15885780900791e-06,
"loss": 0.1512,
"step": 1235
},
{
"epoch": 1.9776,
"grad_norm": 0.50300876963979,
"learning_rate": 3.150204071866464e-06,
"loss": 0.1516,
"step": 1236
},
{
"epoch": 1.9792,
"grad_norm": 0.4803725474827863,
"learning_rate": 3.141556749674988e-06,
"loss": 0.1422,
"step": 1237
},
{
"epoch": 1.9808,
"grad_norm": 0.631057829773093,
"learning_rate": 3.132915872421734e-06,
"loss": 0.1629,
"step": 1238
},
{
"epoch": 1.9824000000000002,
"grad_norm": 0.5567533090045178,
"learning_rate": 3.1242814700725977e-06,
"loss": 0.1732,
"step": 1239
},
{
"epoch": 1.984,
"grad_norm": 0.5446883880440788,
"learning_rate": 3.1156535725710224e-06,
"loss": 0.148,
"step": 1240
},
{
"epoch": 1.9856,
"grad_norm": 0.5021577667238679,
"learning_rate": 3.1070322098378925e-06,
"loss": 0.1372,
"step": 1241
},
{
"epoch": 1.9872,
"grad_norm": 0.5687572837238439,
"learning_rate": 3.0984174117714306e-06,
"loss": 0.1615,
"step": 1242
},
{
"epoch": 1.9888,
"grad_norm": 0.48344458689299913,
"learning_rate": 3.0898092082470943e-06,
"loss": 0.167,
"step": 1243
},
{
"epoch": 1.9904,
"grad_norm": 0.4926315846538951,
"learning_rate": 3.081207629117472e-06,
"loss": 0.1505,
"step": 1244
},
{
"epoch": 1.992,
"grad_norm": 0.6442645787108856,
"learning_rate": 3.0726127042121766e-06,
"loss": 0.1482,
"step": 1245
},
{
"epoch": 1.9936,
"grad_norm": 0.5487674769769518,
"learning_rate": 3.064024463337747e-06,
"loss": 0.1764,
"step": 1246
},
{
"epoch": 1.9952,
"grad_norm": 0.526079169160587,
"learning_rate": 3.0554429362775417e-06,
"loss": 0.1656,
"step": 1247
},
{
"epoch": 1.9968,
"grad_norm": 0.5118848435027314,
"learning_rate": 3.046868152791638e-06,
"loss": 0.1444,
"step": 1248
},
{
"epoch": 1.9984,
"grad_norm": 0.5978413661102846,
"learning_rate": 3.038300142616723e-06,
"loss": 0.1523,
"step": 1249
},
{
"epoch": 2.0,
"grad_norm": 0.5609329120357919,
"learning_rate": 3.0297389354659984e-06,
"loss": 0.1529,
"step": 1250
},
{
"epoch": 2.0016,
"grad_norm": 0.43639588134558943,
"learning_rate": 3.021184561029071e-06,
"loss": 0.1223,
"step": 1251
},
{
"epoch": 2.0032,
"grad_norm": 0.5175198520916784,
"learning_rate": 3.0126370489718537e-06,
"loss": 0.1257,
"step": 1252
},
{
"epoch": 2.0048,
"grad_norm": 0.534898910324941,
"learning_rate": 3.0040964289364618e-06,
"loss": 0.1328,
"step": 1253
},
{
"epoch": 2.0064,
"grad_norm": 0.5100720765383592,
"learning_rate": 2.9955627305411074e-06,
"loss": 0.1101,
"step": 1254
},
{
"epoch": 2.008,
"grad_norm": 0.4569461251288956,
"learning_rate": 2.9870359833799994e-06,
"loss": 0.1514,
"step": 1255
},
{
"epoch": 2.0096,
"grad_norm": 0.4361589900488507,
"learning_rate": 2.978516217023243e-06,
"loss": 0.1265,
"step": 1256
},
{
"epoch": 2.0112,
"grad_norm": 0.5527136660504377,
"learning_rate": 2.97000346101673e-06,
"loss": 0.1367,
"step": 1257
},
{
"epoch": 2.0128,
"grad_norm": 0.501009867764785,
"learning_rate": 2.9614977448820444e-06,
"loss": 0.1306,
"step": 1258
},
{
"epoch": 2.0144,
"grad_norm": 0.5901847474003876,
"learning_rate": 2.952999098116356e-06,
"loss": 0.1237,
"step": 1259
},
{
"epoch": 2.016,
"grad_norm": 0.46784980449800667,
"learning_rate": 2.944507550192318e-06,
"loss": 0.1462,
"step": 1260
},
{
"epoch": 2.0176,
"grad_norm": 0.5277652695547297,
"learning_rate": 2.9360231305579645e-06,
"loss": 0.1411,
"step": 1261
},
{
"epoch": 2.0192,
"grad_norm": 0.4799819268810399,
"learning_rate": 2.9275458686366108e-06,
"loss": 0.1248,
"step": 1262
},
{
"epoch": 2.0208,
"grad_norm": 0.6314687057557924,
"learning_rate": 2.9190757938267477e-06,
"loss": 0.1328,
"step": 1263
},
{
"epoch": 2.0224,
"grad_norm": 0.6071318965588209,
"learning_rate": 2.9106129355019464e-06,
"loss": 0.1165,
"step": 1264
},
{
"epoch": 2.024,
"grad_norm": 0.4632893787471104,
"learning_rate": 2.9021573230107436e-06,
"loss": 0.1123,
"step": 1265
},
{
"epoch": 2.0256,
"grad_norm": 0.6461220356788484,
"learning_rate": 2.8937089856765564e-06,
"loss": 0.1307,
"step": 1266
},
{
"epoch": 2.0272,
"grad_norm": 0.5454503184815178,
"learning_rate": 2.885267952797569e-06,
"loss": 0.1312,
"step": 1267
},
{
"epoch": 2.0288,
"grad_norm": 0.5046516808647794,
"learning_rate": 2.876834253646631e-06,
"loss": 0.1172,
"step": 1268
},
{
"epoch": 2.0304,
"grad_norm": 0.705304087696635,
"learning_rate": 2.8684079174711665e-06,
"loss": 0.1156,
"step": 1269
},
{
"epoch": 2.032,
"grad_norm": 0.43826286571009326,
"learning_rate": 2.8599889734930548e-06,
"loss": 0.1083,
"step": 1270
},
{
"epoch": 2.0336,
"grad_norm": 0.6704604684457001,
"learning_rate": 2.8515774509085535e-06,
"loss": 0.1235,
"step": 1271
},
{
"epoch": 2.0352,
"grad_norm": 0.47611253055234465,
"learning_rate": 2.8431733788881703e-06,
"loss": 0.1227,
"step": 1272
},
{
"epoch": 2.0368,
"grad_norm": 0.6194318157437627,
"learning_rate": 2.8347767865765828e-06,
"loss": 0.1508,
"step": 1273
},
{
"epoch": 2.0384,
"grad_norm": 0.48695778405378304,
"learning_rate": 2.826387703092528e-06,
"loss": 0.1257,
"step": 1274
},
{
"epoch": 2.04,
"grad_norm": 0.6512313492718865,
"learning_rate": 2.8180061575286995e-06,
"loss": 0.1334,
"step": 1275
},
{
"epoch": 2.0416,
"grad_norm": 0.602842467084012,
"learning_rate": 2.8096321789516557e-06,
"loss": 0.1141,
"step": 1276
},
{
"epoch": 2.0432,
"grad_norm": 0.4536179609439741,
"learning_rate": 2.8012657964017056e-06,
"loss": 0.1219,
"step": 1277
},
{
"epoch": 2.0448,
"grad_norm": 0.6842457999016375,
"learning_rate": 2.792907038892823e-06,
"loss": 0.1214,
"step": 1278
},
{
"epoch": 2.0464,
"grad_norm": 0.5532564457913459,
"learning_rate": 2.784555935412538e-06,
"loss": 0.1322,
"step": 1279
},
{
"epoch": 2.048,
"grad_norm": 0.5227144208433524,
"learning_rate": 2.77621251492183e-06,
"loss": 0.1139,
"step": 1280
},
{
"epoch": 2.0496,
"grad_norm": 0.5715212971351321,
"learning_rate": 2.7678768063550454e-06,
"loss": 0.1236,
"step": 1281
},
{
"epoch": 2.0512,
"grad_norm": 0.6213097024340799,
"learning_rate": 2.759548838619774e-06,
"loss": 0.153,
"step": 1282
},
{
"epoch": 2.0528,
"grad_norm": 0.7209988778579537,
"learning_rate": 2.7512286405967726e-06,
"loss": 0.1128,
"step": 1283
},
{
"epoch": 2.0544,
"grad_norm": 0.46240652361057705,
"learning_rate": 2.742916241139843e-06,
"loss": 0.1171,
"step": 1284
},
{
"epoch": 2.056,
"grad_norm": 0.6619700569115736,
"learning_rate": 2.7346116690757496e-06,
"loss": 0.1155,
"step": 1285
},
{
"epoch": 2.0576,
"grad_norm": 0.6128092643082029,
"learning_rate": 2.726314953204111e-06,
"loss": 0.1136,
"step": 1286
},
{
"epoch": 2.0592,
"grad_norm": 0.5066417226272464,
"learning_rate": 2.718026122297297e-06,
"loss": 0.1233,
"step": 1287
},
{
"epoch": 2.0608,
"grad_norm": 0.5014164475875741,
"learning_rate": 2.7097452051003375e-06,
"loss": 0.1161,
"step": 1288
},
{
"epoch": 2.0624,
"grad_norm": 0.4475784602769177,
"learning_rate": 2.701472230330813e-06,
"loss": 0.1104,
"step": 1289
},
{
"epoch": 2.064,
"grad_norm": 0.6893935323029474,
"learning_rate": 2.693207226678767e-06,
"loss": 0.1248,
"step": 1290
},
{
"epoch": 2.0656,
"grad_norm": 0.6144002858193115,
"learning_rate": 2.684950222806596e-06,
"loss": 0.1056,
"step": 1291
},
{
"epoch": 2.0672,
"grad_norm": 0.6190088465077782,
"learning_rate": 2.676701247348951e-06,
"loss": 0.1297,
"step": 1292
},
{
"epoch": 2.0688,
"grad_norm": 0.4799662504423945,
"learning_rate": 2.6684603289126492e-06,
"loss": 0.1265,
"step": 1293
},
{
"epoch": 2.0704,
"grad_norm": 0.5826386978140625,
"learning_rate": 2.660227496076557e-06,
"loss": 0.1379,
"step": 1294
},
{
"epoch": 2.072,
"grad_norm": 0.7376030031301402,
"learning_rate": 2.6520027773915075e-06,
"loss": 0.157,
"step": 1295
},
{
"epoch": 2.0736,
"grad_norm": 0.6715007615581003,
"learning_rate": 2.643786201380194e-06,
"loss": 0.1246,
"step": 1296
},
{
"epoch": 2.0752,
"grad_norm": 0.7553025581846244,
"learning_rate": 2.6355777965370665e-06,
"loss": 0.1319,
"step": 1297
},
{
"epoch": 2.0768,
"grad_norm": 0.5252532640292111,
"learning_rate": 2.627377591328245e-06,
"loss": 0.1273,
"step": 1298
},
{
"epoch": 2.0784,
"grad_norm": 0.6295780787020748,
"learning_rate": 2.6191856141914074e-06,
"loss": 0.1241,
"step": 1299
},
{
"epoch": 2.08,
"grad_norm": 0.5432854539698122,
"learning_rate": 2.6110018935357005e-06,
"loss": 0.133,
"step": 1300
},
{
"epoch": 2.0816,
"grad_norm": 0.7262536036001677,
"learning_rate": 2.6028264577416418e-06,
"loss": 0.1204,
"step": 1301
},
{
"epoch": 2.0832,
"grad_norm": 0.5337284193274364,
"learning_rate": 2.594659335161008e-06,
"loss": 0.114,
"step": 1302
},
{
"epoch": 2.0848,
"grad_norm": 0.5744479689434387,
"learning_rate": 2.586500554116757e-06,
"loss": 0.1218,
"step": 1303
},
{
"epoch": 2.0864,
"grad_norm": 0.8231572317963317,
"learning_rate": 2.578350142902909e-06,
"loss": 0.1137,
"step": 1304
},
{
"epoch": 2.088,
"grad_norm": 0.660038206744017,
"learning_rate": 2.570208129784466e-06,
"loss": 0.1207,
"step": 1305
},
{
"epoch": 2.0896,
"grad_norm": 0.4610609360026432,
"learning_rate": 2.562074542997305e-06,
"loss": 0.11,
"step": 1306
},
{
"epoch": 2.0912,
"grad_norm": 0.6202859735931181,
"learning_rate": 2.5539494107480746e-06,
"loss": 0.1238,
"step": 1307
},
{
"epoch": 2.0928,
"grad_norm": 0.6508774034942395,
"learning_rate": 2.545832761214112e-06,
"loss": 0.1338,
"step": 1308
},
{
"epoch": 2.0944,
"grad_norm": 0.5397566454341794,
"learning_rate": 2.5377246225433306e-06,
"loss": 0.1077,
"step": 1309
},
{
"epoch": 2.096,
"grad_norm": 0.5468537671283503,
"learning_rate": 2.529625022854133e-06,
"loss": 0.1215,
"step": 1310
},
{
"epoch": 2.0976,
"grad_norm": 0.47822082349206285,
"learning_rate": 2.5215339902353097e-06,
"loss": 0.1191,
"step": 1311
},
{
"epoch": 2.0992,
"grad_norm": 0.48894021884821226,
"learning_rate": 2.513451552745936e-06,
"loss": 0.1217,
"step": 1312
},
{
"epoch": 2.1008,
"grad_norm": 0.4841894937328627,
"learning_rate": 2.505377738415286e-06,
"loss": 0.1247,
"step": 1313
},
{
"epoch": 2.1024,
"grad_norm": 0.5556297674391704,
"learning_rate": 2.4973125752427243e-06,
"loss": 0.127,
"step": 1314
},
{
"epoch": 2.104,
"grad_norm": 0.5297785155741399,
"learning_rate": 2.4892560911976167e-06,
"loss": 0.1095,
"step": 1315
},
{
"epoch": 2.1056,
"grad_norm": 0.533313580258,
"learning_rate": 2.481208314219233e-06,
"loss": 0.1288,
"step": 1316
},
{
"epoch": 2.1072,
"grad_norm": 0.8790691699100301,
"learning_rate": 2.4731692722166408e-06,
"loss": 0.1364,
"step": 1317
},
{
"epoch": 2.1088,
"grad_norm": 0.6879278864141557,
"learning_rate": 2.4651389930686226e-06,
"loss": 0.1353,
"step": 1318
},
{
"epoch": 2.1104,
"grad_norm": 0.6411748656773347,
"learning_rate": 2.457117504623565e-06,
"loss": 0.1312,
"step": 1319
},
{
"epoch": 2.112,
"grad_norm": 0.5660312840636772,
"learning_rate": 2.4491048346993756e-06,
"loss": 0.1285,
"step": 1320
},
{
"epoch": 2.1136,
"grad_norm": 0.4790990711350736,
"learning_rate": 2.4411010110833783e-06,
"loss": 0.1036,
"step": 1321
},
{
"epoch": 2.1152,
"grad_norm": 0.5846605455212341,
"learning_rate": 2.433106061532214e-06,
"loss": 0.1272,
"step": 1322
},
{
"epoch": 2.1168,
"grad_norm": 0.6041836948091586,
"learning_rate": 2.4251200137717545e-06,
"loss": 0.1332,
"step": 1323
},
{
"epoch": 2.1184,
"grad_norm": 0.5238581024964356,
"learning_rate": 2.4171428954969995e-06,
"loss": 0.1226,
"step": 1324
},
{
"epoch": 2.12,
"grad_norm": 0.6115731456658904,
"learning_rate": 2.4091747343719828e-06,
"loss": 0.1128,
"step": 1325
},
{
"epoch": 2.1216,
"grad_norm": 0.5251496878486472,
"learning_rate": 2.401215558029671e-06,
"loss": 0.1199,
"step": 1326
},
{
"epoch": 2.1232,
"grad_norm": 0.5774652766778592,
"learning_rate": 2.3932653940718784e-06,
"loss": 0.1501,
"step": 1327
},
{
"epoch": 2.1248,
"grad_norm": 0.5869279907069584,
"learning_rate": 2.3853242700691594e-06,
"loss": 0.1235,
"step": 1328
},
{
"epoch": 2.1264,
"grad_norm": 0.593468161051104,
"learning_rate": 2.3773922135607217e-06,
"loss": 0.12,
"step": 1329
},
{
"epoch": 2.128,
"grad_norm": 0.6270289088054396,
"learning_rate": 2.3694692520543293e-06,
"loss": 0.1093,
"step": 1330
},
{
"epoch": 2.1296,
"grad_norm": 0.7504874233573358,
"learning_rate": 2.3615554130262003e-06,
"loss": 0.1358,
"step": 1331
},
{
"epoch": 2.1312,
"grad_norm": 0.654969028893796,
"learning_rate": 2.3536507239209223e-06,
"loss": 0.1293,
"step": 1332
},
{
"epoch": 2.1328,
"grad_norm": 0.6653037174243946,
"learning_rate": 2.3457552121513455e-06,
"loss": 0.1453,
"step": 1333
},
{
"epoch": 2.1344,
"grad_norm": 0.6041085811024695,
"learning_rate": 2.337868905098499e-06,
"loss": 0.1314,
"step": 1334
},
{
"epoch": 2.136,
"grad_norm": 0.7657057690645885,
"learning_rate": 2.329991830111492e-06,
"loss": 0.1358,
"step": 1335
},
{
"epoch": 2.1376,
"grad_norm": 0.5266987915303966,
"learning_rate": 2.32212401450741e-06,
"loss": 0.1216,
"step": 1336
},
{
"epoch": 2.1391999999999998,
"grad_norm": 0.576653406122159,
"learning_rate": 2.3142654855712353e-06,
"loss": 0.1395,
"step": 1337
},
{
"epoch": 2.1408,
"grad_norm": 0.5280958434601437,
"learning_rate": 2.3064162705557387e-06,
"loss": 0.1351,
"step": 1338
},
{
"epoch": 2.1424,
"grad_norm": 0.6482056432990362,
"learning_rate": 2.2985763966813963e-06,
"loss": 0.1223,
"step": 1339
},
{
"epoch": 2.144,
"grad_norm": 0.5549529190884563,
"learning_rate": 2.2907458911362885e-06,
"loss": 0.1248,
"step": 1340
},
{
"epoch": 2.1456,
"grad_norm": 0.47630320911013324,
"learning_rate": 2.2829247810760023e-06,
"loss": 0.1169,
"step": 1341
},
{
"epoch": 2.1471999999999998,
"grad_norm": 0.4929139863696648,
"learning_rate": 2.275113093623551e-06,
"loss": 0.1033,
"step": 1342
},
{
"epoch": 2.1488,
"grad_norm": 0.4884199203955449,
"learning_rate": 2.2673108558692603e-06,
"loss": 0.1191,
"step": 1343
},
{
"epoch": 2.1504,
"grad_norm": 0.6204030551858207,
"learning_rate": 2.259518094870693e-06,
"loss": 0.137,
"step": 1344
},
{
"epoch": 2.152,
"grad_norm": 0.5736806374801147,
"learning_rate": 2.251734837652547e-06,
"loss": 0.1314,
"step": 1345
},
{
"epoch": 2.1536,
"grad_norm": 0.709304948389432,
"learning_rate": 2.243961111206555e-06,
"loss": 0.1369,
"step": 1346
},
{
"epoch": 2.1552,
"grad_norm": 0.6232527286728418,
"learning_rate": 2.236196942491407e-06,
"loss": 0.1278,
"step": 1347
},
{
"epoch": 2.1568,
"grad_norm": 0.47351357431440344,
"learning_rate": 2.228442358432638e-06,
"loss": 0.1177,
"step": 1348
},
{
"epoch": 2.1584,
"grad_norm": 0.7769623670201542,
"learning_rate": 2.2206973859225518e-06,
"loss": 0.1232,
"step": 1349
},
{
"epoch": 2.16,
"grad_norm": 0.5486752460405687,
"learning_rate": 2.2129620518201184e-06,
"loss": 0.1283,
"step": 1350
},
{
"epoch": 2.1616,
"grad_norm": 0.546364497000948,
"learning_rate": 2.2052363829508776e-06,
"loss": 0.1413,
"step": 1351
},
{
"epoch": 2.1632,
"grad_norm": 0.6686865910510724,
"learning_rate": 2.1975204061068594e-06,
"loss": 0.1297,
"step": 1352
},
{
"epoch": 2.1648,
"grad_norm": 0.6760829982130596,
"learning_rate": 2.189814148046473e-06,
"loss": 0.117,
"step": 1353
},
{
"epoch": 2.1664,
"grad_norm": 0.5073327520918374,
"learning_rate": 2.182117635494431e-06,
"loss": 0.1273,
"step": 1354
},
{
"epoch": 2.168,
"grad_norm": 0.8517224164074038,
"learning_rate": 2.1744308951416483e-06,
"loss": 0.1583,
"step": 1355
},
{
"epoch": 2.1696,
"grad_norm": 0.5903890411657852,
"learning_rate": 2.1667539536451455e-06,
"loss": 0.1356,
"step": 1356
},
{
"epoch": 2.1712,
"grad_norm": 0.589092062161417,
"learning_rate": 2.1590868376279693e-06,
"loss": 0.1265,
"step": 1357
},
{
"epoch": 2.1728,
"grad_norm": 0.5197151319812391,
"learning_rate": 2.151429573679084e-06,
"loss": 0.1212,
"step": 1358
},
{
"epoch": 2.1744,
"grad_norm": 0.6892255686405191,
"learning_rate": 2.1437821883532956e-06,
"loss": 0.14,
"step": 1359
},
{
"epoch": 2.176,
"grad_norm": 0.6375422110297063,
"learning_rate": 2.136144708171145e-06,
"loss": 0.1106,
"step": 1360
},
{
"epoch": 2.1776,
"grad_norm": 0.5501240987770825,
"learning_rate": 2.128517159618827e-06,
"loss": 0.1216,
"step": 1361
},
{
"epoch": 2.1792,
"grad_norm": 0.6697214780356907,
"learning_rate": 2.1208995691480947e-06,
"loss": 0.1395,
"step": 1362
},
{
"epoch": 2.1808,
"grad_norm": 0.5757990202316203,
"learning_rate": 2.1132919631761637e-06,
"loss": 0.1198,
"step": 1363
},
{
"epoch": 2.1824,
"grad_norm": 0.6792181090809793,
"learning_rate": 2.1056943680856286e-06,
"loss": 0.1352,
"step": 1364
},
{
"epoch": 2.184,
"grad_norm": 0.5842439346676801,
"learning_rate": 2.098106810224362e-06,
"loss": 0.1318,
"step": 1365
},
{
"epoch": 2.1856,
"grad_norm": 0.45888614528391547,
"learning_rate": 2.0905293159054315e-06,
"loss": 0.1172,
"step": 1366
},
{
"epoch": 2.1872,
"grad_norm": 0.6654110111317498,
"learning_rate": 2.0829619114070068e-06,
"loss": 0.1132,
"step": 1367
},
{
"epoch": 2.1888,
"grad_norm": 0.666483655951972,
"learning_rate": 2.075404622972261e-06,
"loss": 0.1238,
"step": 1368
},
{
"epoch": 2.1904,
"grad_norm": 0.5425147457691434,
"learning_rate": 2.0678574768092926e-06,
"loss": 0.1242,
"step": 1369
},
{
"epoch": 2.192,
"grad_norm": 0.5943112395215109,
"learning_rate": 2.0603204990910195e-06,
"loss": 0.1238,
"step": 1370
},
{
"epoch": 2.1936,
"grad_norm": 0.4729152914155383,
"learning_rate": 2.0527937159551044e-06,
"loss": 0.1223,
"step": 1371
},
{
"epoch": 2.1952,
"grad_norm": 0.5245337186483764,
"learning_rate": 2.0452771535038518e-06,
"loss": 0.1206,
"step": 1372
},
{
"epoch": 2.1968,
"grad_norm": 0.6533505444040253,
"learning_rate": 2.03777083780412e-06,
"loss": 0.1074,
"step": 1373
},
{
"epoch": 2.1984,
"grad_norm": 0.7135059351954349,
"learning_rate": 2.030274794887237e-06,
"loss": 0.1191,
"step": 1374
},
{
"epoch": 2.2,
"grad_norm": 0.5718000361544356,
"learning_rate": 2.0227890507488993e-06,
"loss": 0.1263,
"step": 1375
},
{
"epoch": 2.2016,
"grad_norm": 0.4815044344331988,
"learning_rate": 2.0153136313490945e-06,
"loss": 0.1268,
"step": 1376
},
{
"epoch": 2.2032,
"grad_norm": 0.4903026662424045,
"learning_rate": 2.0078485626120015e-06,
"loss": 0.1192,
"step": 1377
},
{
"epoch": 2.2048,
"grad_norm": 0.5904261829538481,
"learning_rate": 2.000393870425904e-06,
"loss": 0.1099,
"step": 1378
},
{
"epoch": 2.2064,
"grad_norm": 0.560542059078168,
"learning_rate": 1.9929495806431024e-06,
"loss": 0.1309,
"step": 1379
},
{
"epoch": 2.208,
"grad_norm": 0.5424165967971996,
"learning_rate": 1.985515719079819e-06,
"loss": 0.1244,
"step": 1380
},
{
"epoch": 2.2096,
"grad_norm": 0.4194165193326586,
"learning_rate": 1.978092311516116e-06,
"loss": 0.1262,
"step": 1381
},
{
"epoch": 2.2112,
"grad_norm": 0.6298386300438156,
"learning_rate": 1.9706793836957964e-06,
"loss": 0.1067,
"step": 1382
},
{
"epoch": 2.2128,
"grad_norm": 0.497357860168961,
"learning_rate": 1.963276961326326e-06,
"loss": 0.1407,
"step": 1383
},
{
"epoch": 2.2144,
"grad_norm": 0.4935974860455067,
"learning_rate": 1.955885070078737e-06,
"loss": 0.1268,
"step": 1384
},
{
"epoch": 2.216,
"grad_norm": 0.578752533892835,
"learning_rate": 1.948503735587537e-06,
"loss": 0.1107,
"step": 1385
},
{
"epoch": 2.2176,
"grad_norm": 0.8849218762418022,
"learning_rate": 1.9411329834506286e-06,
"loss": 0.1364,
"step": 1386
},
{
"epoch": 2.2192,
"grad_norm": 0.563079128412417,
"learning_rate": 1.9337728392292104e-06,
"loss": 0.1438,
"step": 1387
},
{
"epoch": 2.2208,
"grad_norm": 0.7508804367335684,
"learning_rate": 1.926423328447698e-06,
"loss": 0.1376,
"step": 1388
},
{
"epoch": 2.2224,
"grad_norm": 0.701829486393775,
"learning_rate": 1.919084476593631e-06,
"loss": 0.1096,
"step": 1389
},
{
"epoch": 2.224,
"grad_norm": 0.6429039739799077,
"learning_rate": 1.9117563091175795e-06,
"loss": 0.1,
"step": 1390
},
{
"epoch": 2.2256,
"grad_norm": 0.5388899595819046,
"learning_rate": 1.904438851433068e-06,
"loss": 0.1129,
"step": 1391
},
{
"epoch": 2.2272,
"grad_norm": 0.5811185674895967,
"learning_rate": 1.897132128916474e-06,
"loss": 0.1205,
"step": 1392
},
{
"epoch": 2.2288,
"grad_norm": 0.6116313951715844,
"learning_rate": 1.8898361669069497e-06,
"loss": 0.1153,
"step": 1393
},
{
"epoch": 2.2304,
"grad_norm": 0.5593909657546566,
"learning_rate": 1.8825509907063328e-06,
"loss": 0.1273,
"step": 1394
},
{
"epoch": 2.232,
"grad_norm": 0.5402185545920631,
"learning_rate": 1.87527662557905e-06,
"loss": 0.1227,
"step": 1395
},
{
"epoch": 2.2336,
"grad_norm": 0.5964749325602555,
"learning_rate": 1.8680130967520433e-06,
"loss": 0.1188,
"step": 1396
},
{
"epoch": 2.2352,
"grad_norm": 0.5498041486161213,
"learning_rate": 1.8607604294146685e-06,
"loss": 0.1311,
"step": 1397
},
{
"epoch": 2.2368,
"grad_norm": 0.6257356883083792,
"learning_rate": 1.8535186487186213e-06,
"loss": 0.1227,
"step": 1398
},
{
"epoch": 2.2384,
"grad_norm": 1.1310991279873195,
"learning_rate": 1.8462877797778367e-06,
"loss": 0.1123,
"step": 1399
},
{
"epoch": 2.24,
"grad_norm": 0.503422219008381,
"learning_rate": 1.8390678476684143e-06,
"loss": 0.1211,
"step": 1400
},
{
"epoch": 2.2416,
"grad_norm": 0.5590382127260044,
"learning_rate": 1.831858877428524e-06,
"loss": 0.1252,
"step": 1401
},
{
"epoch": 2.2432,
"grad_norm": 0.5219268517097256,
"learning_rate": 1.8246608940583166e-06,
"loss": 0.1364,
"step": 1402
},
{
"epoch": 2.2448,
"grad_norm": 0.6042083786813122,
"learning_rate": 1.8174739225198485e-06,
"loss": 0.1156,
"step": 1403
},
{
"epoch": 2.2464,
"grad_norm": 0.5027386240423649,
"learning_rate": 1.8102979877369808e-06,
"loss": 0.1204,
"step": 1404
},
{
"epoch": 2.248,
"grad_norm": 0.7544688991257215,
"learning_rate": 1.8031331145953047e-06,
"loss": 0.1326,
"step": 1405
},
{
"epoch": 2.2496,
"grad_norm": 0.6508846605553914,
"learning_rate": 1.7959793279420507e-06,
"loss": 0.1262,
"step": 1406
},
{
"epoch": 2.2512,
"grad_norm": 0.5860670301599696,
"learning_rate": 1.7888366525859968e-06,
"loss": 0.1295,
"step": 1407
},
{
"epoch": 2.2528,
"grad_norm": 0.9354478730993057,
"learning_rate": 1.781705113297396e-06,
"loss": 0.1433,
"step": 1408
},
{
"epoch": 2.2544,
"grad_norm": 0.5818299899834042,
"learning_rate": 1.7745847348078742e-06,
"loss": 0.1389,
"step": 1409
},
{
"epoch": 2.2560000000000002,
"grad_norm": 0.5286095685587858,
"learning_rate": 1.7674755418103578e-06,
"loss": 0.1323,
"step": 1410
},
{
"epoch": 2.2576,
"grad_norm": 0.5637985240201602,
"learning_rate": 1.7603775589589821e-06,
"loss": 0.1241,
"step": 1411
},
{
"epoch": 2.2592,
"grad_norm": 0.4414412502779432,
"learning_rate": 1.7532908108690038e-06,
"loss": 0.1227,
"step": 1412
},
{
"epoch": 2.2608,
"grad_norm": 0.5652329993436173,
"learning_rate": 1.7462153221167222e-06,
"loss": 0.1174,
"step": 1413
},
{
"epoch": 2.2624,
"grad_norm": 0.4799799828203231,
"learning_rate": 1.7391511172393849e-06,
"loss": 0.1206,
"step": 1414
},
{
"epoch": 2.2640000000000002,
"grad_norm": 0.7023832841247929,
"learning_rate": 1.7320982207351128e-06,
"loss": 0.115,
"step": 1415
},
{
"epoch": 2.2656,
"grad_norm": 0.6422637303644735,
"learning_rate": 1.7250566570628103e-06,
"loss": 0.1177,
"step": 1416
},
{
"epoch": 2.2672,
"grad_norm": 0.5562590591637463,
"learning_rate": 1.7180264506420746e-06,
"loss": 0.1256,
"step": 1417
},
{
"epoch": 2.2688,
"grad_norm": 0.5060906362167427,
"learning_rate": 1.7110076258531244e-06,
"loss": 0.1265,
"step": 1418
},
{
"epoch": 2.2704,
"grad_norm": 0.6166253199957842,
"learning_rate": 1.7040002070367006e-06,
"loss": 0.1274,
"step": 1419
},
{
"epoch": 2.2720000000000002,
"grad_norm": 0.5799993528309908,
"learning_rate": 1.6970042184939943e-06,
"loss": 0.122,
"step": 1420
},
{
"epoch": 2.2736,
"grad_norm": 0.65001665776756,
"learning_rate": 1.6900196844865575e-06,
"loss": 0.1389,
"step": 1421
},
{
"epoch": 2.2752,
"grad_norm": 0.46599256500503206,
"learning_rate": 1.683046629236213e-06,
"loss": 0.1061,
"step": 1422
},
{
"epoch": 2.2768,
"grad_norm": 0.5532665378973851,
"learning_rate": 1.6760850769249837e-06,
"loss": 0.1299,
"step": 1423
},
{
"epoch": 2.2784,
"grad_norm": 0.6069406862269126,
"learning_rate": 1.669135051694994e-06,
"loss": 0.129,
"step": 1424
},
{
"epoch": 2.2800000000000002,
"grad_norm": 0.4578237396071794,
"learning_rate": 1.662196577648398e-06,
"loss": 0.1268,
"step": 1425
},
{
"epoch": 2.2816,
"grad_norm": 0.51246561516751,
"learning_rate": 1.6552696788472921e-06,
"loss": 0.1219,
"step": 1426
},
{
"epoch": 2.2832,
"grad_norm": 0.7058831632624211,
"learning_rate": 1.6483543793136247e-06,
"loss": 0.1319,
"step": 1427
},
{
"epoch": 2.2848,
"grad_norm": 0.55878918286923,
"learning_rate": 1.6414507030291249e-06,
"loss": 0.1144,
"step": 1428
},
{
"epoch": 2.2864,
"grad_norm": 0.5710678966071876,
"learning_rate": 1.6345586739352105e-06,
"loss": 0.1388,
"step": 1429
},
{
"epoch": 2.288,
"grad_norm": 0.4814051013116807,
"learning_rate": 1.6276783159329095e-06,
"loss": 0.1417,
"step": 1430
},
{
"epoch": 2.2896,
"grad_norm": 0.7276094982836895,
"learning_rate": 1.6208096528827717e-06,
"loss": 0.1307,
"step": 1431
},
{
"epoch": 2.2912,
"grad_norm": 0.646262466918538,
"learning_rate": 1.6139527086047929e-06,
"loss": 0.1216,
"step": 1432
},
{
"epoch": 2.2928,
"grad_norm": 0.471672606701089,
"learning_rate": 1.6071075068783303e-06,
"loss": 0.1241,
"step": 1433
},
{
"epoch": 2.2944,
"grad_norm": 0.6289384584879771,
"learning_rate": 1.600274071442014e-06,
"loss": 0.1188,
"step": 1434
},
{
"epoch": 2.296,
"grad_norm": 0.5856187418444901,
"learning_rate": 1.5934524259936757e-06,
"loss": 0.1209,
"step": 1435
},
{
"epoch": 2.2976,
"grad_norm": 0.8339572005265362,
"learning_rate": 1.5866425941902524e-06,
"loss": 0.1267,
"step": 1436
},
{
"epoch": 2.2992,
"grad_norm": 2.616906401327472,
"learning_rate": 1.5798445996477219e-06,
"loss": 0.1185,
"step": 1437
},
{
"epoch": 2.3008,
"grad_norm": 0.5789903587995018,
"learning_rate": 1.573058465941002e-06,
"loss": 0.1164,
"step": 1438
},
{
"epoch": 2.3024,
"grad_norm": 0.656052075546998,
"learning_rate": 1.5662842166038844e-06,
"loss": 0.1311,
"step": 1439
},
{
"epoch": 2.304,
"grad_norm": 0.5712611356018166,
"learning_rate": 1.5595218751289465e-06,
"loss": 0.1344,
"step": 1440
},
{
"epoch": 2.3056,
"grad_norm": 0.791499871534804,
"learning_rate": 1.5527714649674641e-06,
"loss": 0.1238,
"step": 1441
},
{
"epoch": 2.3072,
"grad_norm": 0.8521155473455099,
"learning_rate": 1.5460330095293447e-06,
"loss": 0.118,
"step": 1442
},
{
"epoch": 2.3088,
"grad_norm": 0.6632883546513534,
"learning_rate": 1.5393065321830292e-06,
"loss": 0.1135,
"step": 1443
},
{
"epoch": 2.3104,
"grad_norm": 0.7201094033551724,
"learning_rate": 1.5325920562554259e-06,
"loss": 0.1279,
"step": 1444
},
{
"epoch": 2.312,
"grad_norm": 0.5873467221476779,
"learning_rate": 1.5258896050318217e-06,
"loss": 0.1305,
"step": 1445
},
{
"epoch": 2.3136,
"grad_norm": 0.7386109461856626,
"learning_rate": 1.5191992017557994e-06,
"loss": 0.129,
"step": 1446
},
{
"epoch": 2.3152,
"grad_norm": 0.49838787171862453,
"learning_rate": 1.512520869629165e-06,
"loss": 0.1217,
"step": 1447
},
{
"epoch": 2.3168,
"grad_norm": 0.7533734997168323,
"learning_rate": 1.5058546318118583e-06,
"loss": 0.1363,
"step": 1448
},
{
"epoch": 2.3184,
"grad_norm": 0.45388316231064185,
"learning_rate": 1.4992005114218805e-06,
"loss": 0.1199,
"step": 1449
},
{
"epoch": 2.32,
"grad_norm": 0.5161701391887932,
"learning_rate": 1.4925585315352108e-06,
"loss": 0.128,
"step": 1450
},
{
"epoch": 2.3216,
"grad_norm": 0.5391684121848352,
"learning_rate": 1.485928715185721e-06,
"loss": 0.1223,
"step": 1451
},
{
"epoch": 2.3232,
"grad_norm": 0.5799422120118404,
"learning_rate": 1.4793110853651077e-06,
"loss": 0.1252,
"step": 1452
},
{
"epoch": 2.3247999999999998,
"grad_norm": 0.6247166059758802,
"learning_rate": 1.472705665022799e-06,
"loss": 0.1152,
"step": 1453
},
{
"epoch": 2.3264,
"grad_norm": 0.6356278067513945,
"learning_rate": 1.4661124770658857e-06,
"loss": 0.1255,
"step": 1454
},
{
"epoch": 2.328,
"grad_norm": 0.8549426086401078,
"learning_rate": 1.459531544359038e-06,
"loss": 0.1309,
"step": 1455
},
{
"epoch": 2.3296,
"grad_norm": 0.6729068029293481,
"learning_rate": 1.4529628897244214e-06,
"loss": 0.1208,
"step": 1456
},
{
"epoch": 2.3312,
"grad_norm": 0.5790339849442727,
"learning_rate": 1.4464065359416274e-06,
"loss": 0.1203,
"step": 1457
},
{
"epoch": 2.3327999999999998,
"grad_norm": 0.7737785425521434,
"learning_rate": 1.4398625057475845e-06,
"loss": 0.1178,
"step": 1458
},
{
"epoch": 2.3344,
"grad_norm": 0.7250086149611722,
"learning_rate": 1.4333308218364861e-06,
"loss": 0.1351,
"step": 1459
},
{
"epoch": 2.336,
"grad_norm": 0.6632917124385066,
"learning_rate": 1.4268115068597122e-06,
"loss": 0.122,
"step": 1460
},
{
"epoch": 2.3376,
"grad_norm": 0.605989583750546,
"learning_rate": 1.4203045834257418e-06,
"loss": 0.1228,
"step": 1461
},
{
"epoch": 2.3392,
"grad_norm": 0.6858879227342829,
"learning_rate": 1.4138100741000888e-06,
"loss": 0.133,
"step": 1462
},
{
"epoch": 2.3407999999999998,
"grad_norm": 0.5265758818259707,
"learning_rate": 1.4073280014052077e-06,
"loss": 0.1103,
"step": 1463
},
{
"epoch": 2.3424,
"grad_norm": 0.7531926617297378,
"learning_rate": 1.4008583878204297e-06,
"loss": 0.1366,
"step": 1464
},
{
"epoch": 2.344,
"grad_norm": 0.663435435277719,
"learning_rate": 1.3944012557818793e-06,
"loss": 0.1315,
"step": 1465
},
{
"epoch": 2.3456,
"grad_norm": 0.600321849796885,
"learning_rate": 1.3879566276823896e-06,
"loss": 0.1361,
"step": 1466
},
{
"epoch": 2.3472,
"grad_norm": 0.5448754342766675,
"learning_rate": 1.3815245258714393e-06,
"loss": 0.1407,
"step": 1467
},
{
"epoch": 2.3487999999999998,
"grad_norm": 0.6442440506766789,
"learning_rate": 1.3751049726550587e-06,
"loss": 0.1476,
"step": 1468
},
{
"epoch": 2.3504,
"grad_norm": 0.6730024053183656,
"learning_rate": 1.368697990295766e-06,
"loss": 0.1173,
"step": 1469
},
{
"epoch": 2.352,
"grad_norm": 0.5489852664785198,
"learning_rate": 1.3623036010124845e-06,
"loss": 0.1257,
"step": 1470
},
{
"epoch": 2.3536,
"grad_norm": 0.4942744073084671,
"learning_rate": 1.3559218269804624e-06,
"loss": 0.1198,
"step": 1471
},
{
"epoch": 2.3552,
"grad_norm": 0.6406123482068938,
"learning_rate": 1.3495526903312029e-06,
"loss": 0.1276,
"step": 1472
},
{
"epoch": 2.3568,
"grad_norm": 0.5982150492787023,
"learning_rate": 1.3431962131523796e-06,
"loss": 0.1254,
"step": 1473
},
{
"epoch": 2.3584,
"grad_norm": 0.6118680514944181,
"learning_rate": 1.3368524174877679e-06,
"loss": 0.1223,
"step": 1474
},
{
"epoch": 2.36,
"grad_norm": 0.605198939118349,
"learning_rate": 1.330521325337164e-06,
"loss": 0.1336,
"step": 1475
},
{
"epoch": 2.3616,
"grad_norm": 0.5874385833375948,
"learning_rate": 1.3242029586563054e-06,
"loss": 0.1142,
"step": 1476
},
{
"epoch": 2.3632,
"grad_norm": 0.631308843494122,
"learning_rate": 1.3178973393568055e-06,
"loss": 0.117,
"step": 1477
},
{
"epoch": 2.3648,
"grad_norm": 0.5943460352325605,
"learning_rate": 1.3116044893060637e-06,
"loss": 0.1218,
"step": 1478
},
{
"epoch": 2.3664,
"grad_norm": 0.5735833802932462,
"learning_rate": 1.3053244303272022e-06,
"loss": 0.1316,
"step": 1479
},
{
"epoch": 2.368,
"grad_norm": 0.62082794575548,
"learning_rate": 1.2990571841989796e-06,
"loss": 0.1246,
"step": 1480
},
{
"epoch": 2.3696,
"grad_norm": 0.6109112736249646,
"learning_rate": 1.2928027726557257e-06,
"loss": 0.1158,
"step": 1481
},
{
"epoch": 2.3712,
"grad_norm": 1.7507320031040519,
"learning_rate": 1.2865612173872577e-06,
"loss": 0.1165,
"step": 1482
},
{
"epoch": 2.3728,
"grad_norm": 0.7323551878225706,
"learning_rate": 1.2803325400388095e-06,
"loss": 0.1285,
"step": 1483
},
{
"epoch": 2.3744,
"grad_norm": 0.5360667727565115,
"learning_rate": 1.2741167622109557e-06,
"loss": 0.1075,
"step": 1484
},
{
"epoch": 2.376,
"grad_norm": 0.7701076675889028,
"learning_rate": 1.2679139054595335e-06,
"loss": 0.1174,
"step": 1485
},
{
"epoch": 2.3776,
"grad_norm": 0.5494228769510036,
"learning_rate": 1.261723991295576e-06,
"loss": 0.1133,
"step": 1486
},
{
"epoch": 2.3792,
"grad_norm": 0.5457917152335158,
"learning_rate": 1.2555470411852262e-06,
"loss": 0.0991,
"step": 1487
},
{
"epoch": 2.3808,
"grad_norm": 0.6174184251358676,
"learning_rate": 1.2493830765496724e-06,
"loss": 0.1134,
"step": 1488
},
{
"epoch": 2.3824,
"grad_norm": 0.596239722660635,
"learning_rate": 1.2432321187650726e-06,
"loss": 0.137,
"step": 1489
},
{
"epoch": 2.384,
"grad_norm": 0.5172840120449287,
"learning_rate": 1.237094189162471e-06,
"loss": 0.1245,
"step": 1490
},
{
"epoch": 2.3856,
"grad_norm": 0.7540145661346448,
"learning_rate": 1.2309693090277392e-06,
"loss": 0.1329,
"step": 1491
},
{
"epoch": 2.3872,
"grad_norm": 0.6520506766936994,
"learning_rate": 1.2248574996014872e-06,
"loss": 0.1155,
"step": 1492
},
{
"epoch": 2.3888,
"grad_norm": 0.617360521827684,
"learning_rate": 1.218758782079001e-06,
"loss": 0.1367,
"step": 1493
},
{
"epoch": 2.3904,
"grad_norm": 0.7009226755415091,
"learning_rate": 1.2126731776101657e-06,
"loss": 0.1242,
"step": 1494
},
{
"epoch": 2.392,
"grad_norm": 0.6825235611401647,
"learning_rate": 1.2066007072993856e-06,
"loss": 0.1334,
"step": 1495
},
{
"epoch": 2.3936,
"grad_norm": 0.806338244642499,
"learning_rate": 1.2005413922055247e-06,
"loss": 0.119,
"step": 1496
},
{
"epoch": 2.3952,
"grad_norm": 0.5225484590985218,
"learning_rate": 1.194495253341818e-06,
"loss": 0.1262,
"step": 1497
},
{
"epoch": 2.3968,
"grad_norm": 0.7304079949333458,
"learning_rate": 1.1884623116758121e-06,
"loss": 0.0981,
"step": 1498
},
{
"epoch": 2.3984,
"grad_norm": 0.5794262765287543,
"learning_rate": 1.1824425881292846e-06,
"loss": 0.1392,
"step": 1499
},
{
"epoch": 2.4,
"grad_norm": 0.6568036931324175,
"learning_rate": 1.1764361035781718e-06,
"loss": 0.1257,
"step": 1500
},
{
"epoch": 2.4016,
"grad_norm": 0.5871838767067157,
"learning_rate": 1.170442878852503e-06,
"loss": 0.1177,
"step": 1501
},
{
"epoch": 2.4032,
"grad_norm": 0.7908664995850833,
"learning_rate": 1.1644629347363173e-06,
"loss": 0.1312,
"step": 1502
},
{
"epoch": 2.4048,
"grad_norm": 0.5595386868535858,
"learning_rate": 1.1584962919676024e-06,
"loss": 0.1342,
"step": 1503
},
{
"epoch": 2.4064,
"grad_norm": 0.4828380773998647,
"learning_rate": 1.1525429712382175e-06,
"loss": 0.1199,
"step": 1504
},
{
"epoch": 2.408,
"grad_norm": 0.5529623223296712,
"learning_rate": 1.1466029931938182e-06,
"loss": 0.134,
"step": 1505
},
{
"epoch": 2.4096,
"grad_norm": 0.5792271217679253,
"learning_rate": 1.1406763784337948e-06,
"loss": 0.124,
"step": 1506
},
{
"epoch": 2.4112,
"grad_norm": 0.6352011574353768,
"learning_rate": 1.1347631475111882e-06,
"loss": 0.1228,
"step": 1507
},
{
"epoch": 2.4128,
"grad_norm": 0.6608171700559409,
"learning_rate": 1.1288633209326288e-06,
"loss": 0.118,
"step": 1508
},
{
"epoch": 2.4144,
"grad_norm": 0.9143305938814551,
"learning_rate": 1.122976919158264e-06,
"loss": 0.1126,
"step": 1509
},
{
"epoch": 2.416,
"grad_norm": 0.6559094909349412,
"learning_rate": 1.1171039626016789e-06,
"loss": 0.1094,
"step": 1510
},
{
"epoch": 2.4176,
"grad_norm": 0.6690568973178005,
"learning_rate": 1.1112444716298381e-06,
"loss": 0.118,
"step": 1511
},
{
"epoch": 2.4192,
"grad_norm": 0.4966328703274408,
"learning_rate": 1.1053984665630025e-06,
"loss": 0.1198,
"step": 1512
},
{
"epoch": 2.4208,
"grad_norm": 0.5093968505131875,
"learning_rate": 1.0995659676746706e-06,
"loss": 0.1257,
"step": 1513
},
{
"epoch": 2.4224,
"grad_norm": 0.49080984960512064,
"learning_rate": 1.093746995191497e-06,
"loss": 0.1169,
"step": 1514
},
{
"epoch": 2.424,
"grad_norm": 0.5628609548837707,
"learning_rate": 1.0879415692932328e-06,
"loss": 0.1235,
"step": 1515
},
{
"epoch": 2.4256,
"grad_norm": 0.8328126451427691,
"learning_rate": 1.0821497101126487e-06,
"loss": 0.1363,
"step": 1516
},
{
"epoch": 2.4272,
"grad_norm": 0.6851303264670734,
"learning_rate": 1.076371437735465e-06,
"loss": 0.1334,
"step": 1517
},
{
"epoch": 2.4288,
"grad_norm": 0.6795388327610195,
"learning_rate": 1.0706067722002877e-06,
"loss": 0.1293,
"step": 1518
},
{
"epoch": 2.4304,
"grad_norm": 0.5226867651951489,
"learning_rate": 1.064855733498531e-06,
"loss": 0.1272,
"step": 1519
},
{
"epoch": 2.432,
"grad_norm": 0.6356614005983517,
"learning_rate": 1.0591183415743562e-06,
"loss": 0.1275,
"step": 1520
},
{
"epoch": 2.4336,
"grad_norm": 0.7067696700585631,
"learning_rate": 1.0533946163245984e-06,
"loss": 0.1308,
"step": 1521
},
{
"epoch": 2.4352,
"grad_norm": 0.7872408474635217,
"learning_rate": 1.047684577598694e-06,
"loss": 0.1206,
"step": 1522
},
{
"epoch": 2.4368,
"grad_norm": 0.7525161383258674,
"learning_rate": 1.0419882451986197e-06,
"loss": 0.1316,
"step": 1523
},
{
"epoch": 2.4384,
"grad_norm": 0.4943562576767839,
"learning_rate": 1.0363056388788162e-06,
"loss": 0.1244,
"step": 1524
},
{
"epoch": 2.44,
"grad_norm": 0.6922742203136746,
"learning_rate": 1.0306367783461258e-06,
"loss": 0.1383,
"step": 1525
},
{
"epoch": 2.4416,
"grad_norm": 0.7215974126746628,
"learning_rate": 1.024981683259723e-06,
"loss": 0.1223,
"step": 1526
},
{
"epoch": 2.4432,
"grad_norm": 0.6240248416505015,
"learning_rate": 1.0193403732310392e-06,
"loss": 0.1169,
"step": 1527
},
{
"epoch": 2.4448,
"grad_norm": 0.6347216475844211,
"learning_rate": 1.0137128678237062e-06,
"loss": 0.1523,
"step": 1528
},
{
"epoch": 2.4464,
"grad_norm": 0.7489677944188922,
"learning_rate": 1.0080991865534773e-06,
"loss": 0.1214,
"step": 1529
},
{
"epoch": 2.448,
"grad_norm": 0.5083655643124518,
"learning_rate": 1.002499348888169e-06,
"loss": 0.1003,
"step": 1530
},
{
"epoch": 2.4496,
"grad_norm": 0.5918749098198338,
"learning_rate": 9.969133742475883e-07,
"loss": 0.1276,
"step": 1531
},
{
"epoch": 2.4512,
"grad_norm": 0.8305375352370075,
"learning_rate": 9.913412820034629e-07,
"loss": 0.1151,
"step": 1532
},
{
"epoch": 2.4528,
"grad_norm": 0.6901398319677366,
"learning_rate": 9.857830914793827e-07,
"loss": 0.1265,
"step": 1533
},
{
"epoch": 2.4544,
"grad_norm": 0.5753891916363378,
"learning_rate": 9.802388219507215e-07,
"loss": 0.1267,
"step": 1534
},
{
"epoch": 2.456,
"grad_norm": 0.5173343676247624,
"learning_rate": 9.747084926445839e-07,
"loss": 0.1306,
"step": 1535
},
{
"epoch": 2.4576000000000002,
"grad_norm": 0.4935950906614236,
"learning_rate": 9.691921227397227e-07,
"loss": 0.1423,
"step": 1536
},
{
"epoch": 2.4592,
"grad_norm": 0.8696320690998582,
"learning_rate": 9.63689731366486e-07,
"loss": 0.1261,
"step": 1537
},
{
"epoch": 2.4608,
"grad_norm": 0.5539057994405642,
"learning_rate": 9.58201337606745e-07,
"loss": 0.1249,
"step": 1538
},
{
"epoch": 2.4624,
"grad_norm": 0.69830425994133,
"learning_rate": 9.527269604938249e-07,
"loss": 0.1321,
"step": 1539
},
{
"epoch": 2.464,
"grad_norm": 1.2209044420260544,
"learning_rate": 9.472666190124457e-07,
"loss": 0.1239,
"step": 1540
},
{
"epoch": 2.4656000000000002,
"grad_norm": 0.5888181883391111,
"learning_rate": 9.418203320986502e-07,
"loss": 0.1281,
"step": 1541
},
{
"epoch": 2.4672,
"grad_norm": 0.5742386164467689,
"learning_rate": 9.363881186397434e-07,
"loss": 0.1231,
"step": 1542
},
{
"epoch": 2.4688,
"grad_norm": 0.46717929897935223,
"learning_rate": 9.309699974742243e-07,
"loss": 0.1105,
"step": 1543
},
{
"epoch": 2.4704,
"grad_norm": 0.6074722443170127,
"learning_rate": 9.255659873917183e-07,
"loss": 0.1277,
"step": 1544
},
{
"epoch": 2.472,
"grad_norm": 0.5016482574881039,
"learning_rate": 9.201761071329196e-07,
"loss": 0.1061,
"step": 1545
},
{
"epoch": 2.4736000000000002,
"grad_norm": 0.6520352958821154,
"learning_rate": 9.148003753895146e-07,
"loss": 0.1079,
"step": 1546
},
{
"epoch": 2.4752,
"grad_norm": 0.5285473698965738,
"learning_rate": 9.094388108041302e-07,
"loss": 0.1262,
"step": 1547
},
{
"epoch": 2.4768,
"grad_norm": 0.5969921097713612,
"learning_rate": 9.040914319702598e-07,
"loss": 0.1265,
"step": 1548
},
{
"epoch": 2.4784,
"grad_norm": 0.5542949307708024,
"learning_rate": 8.987582574321996e-07,
"loss": 0.1292,
"step": 1549
},
{
"epoch": 2.48,
"grad_norm": 0.5635602853422518,
"learning_rate": 8.934393056849921e-07,
"loss": 0.1222,
"step": 1550
},
{
"epoch": 2.4816,
"grad_norm": 0.5677332222065206,
"learning_rate": 8.881345951743486e-07,
"loss": 0.1113,
"step": 1551
},
{
"epoch": 2.4832,
"grad_norm": 0.45869085306886925,
"learning_rate": 8.828441442966013e-07,
"loss": 0.1275,
"step": 1552
},
{
"epoch": 2.4848,
"grad_norm": 0.6820178172243544,
"learning_rate": 8.775679713986235e-07,
"loss": 0.1448,
"step": 1553
},
{
"epoch": 2.4864,
"grad_norm": 0.5774054354763961,
"learning_rate": 8.723060947777778e-07,
"loss": 0.1341,
"step": 1554
},
{
"epoch": 2.488,
"grad_norm": 0.5425072566689275,
"learning_rate": 8.670585326818493e-07,
"loss": 0.125,
"step": 1555
},
{
"epoch": 2.4896,
"grad_norm": 0.655280760130869,
"learning_rate": 8.618253033089768e-07,
"loss": 0.1333,
"step": 1556
},
{
"epoch": 2.4912,
"grad_norm": 0.6650048674401643,
"learning_rate": 8.566064248076001e-07,
"loss": 0.1253,
"step": 1557
},
{
"epoch": 2.4928,
"grad_norm": 0.584837340674422,
"learning_rate": 8.514019152763852e-07,
"loss": 0.1125,
"step": 1558
},
{
"epoch": 2.4944,
"grad_norm": 0.6574849395373898,
"learning_rate": 8.462117927641733e-07,
"loss": 0.1165,
"step": 1559
},
{
"epoch": 2.496,
"grad_norm": 0.5419076837772993,
"learning_rate": 8.410360752699099e-07,
"loss": 0.1215,
"step": 1560
},
{
"epoch": 2.4976,
"grad_norm": 0.5801088254281928,
"learning_rate": 8.358747807425827e-07,
"loss": 0.1084,
"step": 1561
},
{
"epoch": 2.4992,
"grad_norm": 0.615814151680793,
"learning_rate": 8.307279270811675e-07,
"loss": 0.1264,
"step": 1562
},
{
"epoch": 2.5008,
"grad_norm": 0.5515339409568563,
"learning_rate": 8.255955321345533e-07,
"loss": 0.1156,
"step": 1563
},
{
"epoch": 2.5023999999999997,
"grad_norm": 0.5394140285328092,
"learning_rate": 8.20477613701493e-07,
"loss": 0.1207,
"step": 1564
},
{
"epoch": 2.504,
"grad_norm": 0.6737057399468097,
"learning_rate": 8.153741895305351e-07,
"loss": 0.1147,
"step": 1565
},
{
"epoch": 2.5056000000000003,
"grad_norm": 0.5342346841612867,
"learning_rate": 8.102852773199588e-07,
"loss": 0.1286,
"step": 1566
},
{
"epoch": 2.5072,
"grad_norm": 0.7586471501221933,
"learning_rate": 8.052108947177234e-07,
"loss": 0.1271,
"step": 1567
},
{
"epoch": 2.5088,
"grad_norm": 0.573243651462557,
"learning_rate": 8.001510593213946e-07,
"loss": 0.1332,
"step": 1568
},
{
"epoch": 2.5103999999999997,
"grad_norm": 0.5187448503933086,
"learning_rate": 7.951057886780939e-07,
"loss": 0.111,
"step": 1569
},
{
"epoch": 2.512,
"grad_norm": 0.7462093149974235,
"learning_rate": 7.900751002844326e-07,
"loss": 0.1192,
"step": 1570
},
{
"epoch": 2.5136,
"grad_norm": 0.5995907898413535,
"learning_rate": 7.850590115864481e-07,
"loss": 0.1239,
"step": 1571
},
{
"epoch": 2.5152,
"grad_norm": 0.7352417664888046,
"learning_rate": 7.80057539979554e-07,
"loss": 0.135,
"step": 1572
},
{
"epoch": 2.5168,
"grad_norm": 0.6865885554544564,
"learning_rate": 7.750707028084653e-07,
"loss": 0.098,
"step": 1573
},
{
"epoch": 2.5183999999999997,
"grad_norm": 1.3662580241464843,
"learning_rate": 7.70098517367151e-07,
"loss": 0.1366,
"step": 1574
},
{
"epoch": 2.52,
"grad_norm": 0.4796750348580511,
"learning_rate": 7.651410008987698e-07,
"loss": 0.1172,
"step": 1575
},
{
"epoch": 2.5216,
"grad_norm": 0.6604694028269259,
"learning_rate": 7.601981705956041e-07,
"loss": 0.12,
"step": 1576
},
{
"epoch": 2.5232,
"grad_norm": 0.5303557578066007,
"learning_rate": 7.552700435990123e-07,
"loss": 0.141,
"step": 1577
},
{
"epoch": 2.5248,
"grad_norm": 0.7417151282793515,
"learning_rate": 7.503566369993564e-07,
"loss": 0.1271,
"step": 1578
},
{
"epoch": 2.5263999999999998,
"grad_norm": 0.8307394493008035,
"learning_rate": 7.454579678359547e-07,
"loss": 0.1241,
"step": 1579
},
{
"epoch": 2.528,
"grad_norm": 0.8100521212942532,
"learning_rate": 7.405740530970157e-07,
"loss": 0.1367,
"step": 1580
},
{
"epoch": 2.5296,
"grad_norm": 0.6897753202111327,
"learning_rate": 7.357049097195773e-07,
"loss": 0.1224,
"step": 1581
},
{
"epoch": 2.5312,
"grad_norm": 0.6160125438618,
"learning_rate": 7.308505545894567e-07,
"loss": 0.1206,
"step": 1582
},
{
"epoch": 2.5328,
"grad_norm": 0.49697405958864155,
"learning_rate": 7.260110045411816e-07,
"loss": 0.1272,
"step": 1583
},
{
"epoch": 2.5343999999999998,
"grad_norm": 0.5822102904628198,
"learning_rate": 7.211862763579414e-07,
"loss": 0.1272,
"step": 1584
},
{
"epoch": 2.536,
"grad_norm": 0.5511970779445157,
"learning_rate": 7.163763867715218e-07,
"loss": 0.1346,
"step": 1585
},
{
"epoch": 2.5376,
"grad_norm": 0.6128640820273042,
"learning_rate": 7.115813524622489e-07,
"loss": 0.1284,
"step": 1586
},
{
"epoch": 2.5392,
"grad_norm": 0.6972410602619332,
"learning_rate": 7.068011900589333e-07,
"loss": 0.1333,
"step": 1587
},
{
"epoch": 2.5408,
"grad_norm": 0.68427107766501,
"learning_rate": 7.020359161388108e-07,
"loss": 0.1325,
"step": 1588
},
{
"epoch": 2.5423999999999998,
"grad_norm": 0.48781326938420877,
"learning_rate": 6.972855472274853e-07,
"loss": 0.1217,
"step": 1589
},
{
"epoch": 2.544,
"grad_norm": 0.5568292075788861,
"learning_rate": 6.925500997988694e-07,
"loss": 0.1183,
"step": 1590
},
{
"epoch": 2.5456,
"grad_norm": 0.5459236401808886,
"learning_rate": 6.87829590275132e-07,
"loss": 0.1179,
"step": 1591
},
{
"epoch": 2.5472,
"grad_norm": 0.5982958435506572,
"learning_rate": 6.83124035026635e-07,
"loss": 0.1492,
"step": 1592
},
{
"epoch": 2.5488,
"grad_norm": 0.5720780498315257,
"learning_rate": 6.784334503718826e-07,
"loss": 0.1222,
"step": 1593
},
{
"epoch": 2.5504,
"grad_norm": 0.5903124291966404,
"learning_rate": 6.737578525774636e-07,
"loss": 0.1175,
"step": 1594
},
{
"epoch": 2.552,
"grad_norm": 0.7475099258106006,
"learning_rate": 6.690972578579886e-07,
"loss": 0.1186,
"step": 1595
},
{
"epoch": 2.5536,
"grad_norm": 0.5601898793999107,
"learning_rate": 6.644516823760439e-07,
"loss": 0.1258,
"step": 1596
},
{
"epoch": 2.5552,
"grad_norm": 0.6395335283310418,
"learning_rate": 6.598211422421258e-07,
"loss": 0.1229,
"step": 1597
},
{
"epoch": 2.5568,
"grad_norm": 0.534435951570097,
"learning_rate": 6.552056535145917e-07,
"loss": 0.1232,
"step": 1598
},
{
"epoch": 2.5584,
"grad_norm": 0.6176685570004835,
"learning_rate": 6.506052321996037e-07,
"loss": 0.1185,
"step": 1599
},
{
"epoch": 2.56,
"grad_norm": 0.604216811894863,
"learning_rate": 6.46019894251066e-07,
"loss": 0.1223,
"step": 1600
},
{
"epoch": 2.5616,
"grad_norm": 0.7710939984727455,
"learning_rate": 6.414496555705802e-07,
"loss": 0.1352,
"step": 1601
},
{
"epoch": 2.5632,
"grad_norm": 0.5913133497639669,
"learning_rate": 6.368945320073799e-07,
"loss": 0.1555,
"step": 1602
},
{
"epoch": 2.5648,
"grad_norm": 0.5355844502669492,
"learning_rate": 6.323545393582847e-07,
"loss": 0.1234,
"step": 1603
},
{
"epoch": 2.5664,
"grad_norm": 0.6024474573408983,
"learning_rate": 6.278296933676414e-07,
"loss": 0.1232,
"step": 1604
},
{
"epoch": 2.568,
"grad_norm": 0.608328646075386,
"learning_rate": 6.233200097272646e-07,
"loss": 0.1308,
"step": 1605
},
{
"epoch": 2.5696,
"grad_norm": 0.5913002597898629,
"learning_rate": 6.188255040763929e-07,
"loss": 0.1182,
"step": 1606
},
{
"epoch": 2.5712,
"grad_norm": 0.7717824154192828,
"learning_rate": 6.143461920016247e-07,
"loss": 0.1349,
"step": 1607
},
{
"epoch": 2.5728,
"grad_norm": 0.4769238274509385,
"learning_rate": 6.098820890368696e-07,
"loss": 0.1114,
"step": 1608
},
{
"epoch": 2.5744,
"grad_norm": 0.50010406684415,
"learning_rate": 6.054332106632943e-07,
"loss": 0.1464,
"step": 1609
},
{
"epoch": 2.576,
"grad_norm": 0.5847655987586389,
"learning_rate": 6.009995723092655e-07,
"loss": 0.1233,
"step": 1610
},
{
"epoch": 2.5776,
"grad_norm": 0.7082604708165138,
"learning_rate": 5.965811893503015e-07,
"loss": 0.1224,
"step": 1611
},
{
"epoch": 2.5792,
"grad_norm": 0.5000246734339294,
"learning_rate": 5.921780771090124e-07,
"loss": 0.1322,
"step": 1612
},
{
"epoch": 2.5808,
"grad_norm": 0.5705063368264416,
"learning_rate": 5.877902508550542e-07,
"loss": 0.1331,
"step": 1613
},
{
"epoch": 2.5824,
"grad_norm": 0.6744806111731407,
"learning_rate": 5.834177258050711e-07,
"loss": 0.1311,
"step": 1614
},
{
"epoch": 2.584,
"grad_norm": 0.6120102769245663,
"learning_rate": 5.790605171226421e-07,
"loss": 0.1144,
"step": 1615
},
{
"epoch": 2.5856,
"grad_norm": 0.5917366812127909,
"learning_rate": 5.747186399182336e-07,
"loss": 0.1431,
"step": 1616
},
{
"epoch": 2.5872,
"grad_norm": 0.573017071596183,
"learning_rate": 5.703921092491393e-07,
"loss": 0.116,
"step": 1617
},
{
"epoch": 2.5888,
"grad_norm": 0.6112343589310972,
"learning_rate": 5.660809401194362e-07,
"loss": 0.1152,
"step": 1618
},
{
"epoch": 2.5904,
"grad_norm": 0.8609513603817291,
"learning_rate": 5.617851474799285e-07,
"loss": 0.1292,
"step": 1619
},
{
"epoch": 2.592,
"grad_norm": 0.7615381802655957,
"learning_rate": 5.575047462280919e-07,
"loss": 0.1153,
"step": 1620
},
{
"epoch": 2.5936,
"grad_norm": 0.5110866329630191,
"learning_rate": 5.532397512080306e-07,
"loss": 0.1117,
"step": 1621
},
{
"epoch": 2.5952,
"grad_norm": 0.7494070785939076,
"learning_rate": 5.489901772104178e-07,
"loss": 0.1293,
"step": 1622
},
{
"epoch": 2.5968,
"grad_norm": 0.5327524980171712,
"learning_rate": 5.447560389724499e-07,
"loss": 0.1246,
"step": 1623
},
{
"epoch": 2.5984,
"grad_norm": 0.6090772082837987,
"learning_rate": 5.405373511777939e-07,
"loss": 0.1417,
"step": 1624
},
{
"epoch": 2.6,
"grad_norm": 0.584004742479714,
"learning_rate": 5.363341284565316e-07,
"loss": 0.11,
"step": 1625
},
{
"epoch": 2.6016,
"grad_norm": 0.5890499568060071,
"learning_rate": 5.321463853851189e-07,
"loss": 0.1067,
"step": 1626
},
{
"epoch": 2.6032,
"grad_norm": 0.6154747318909511,
"learning_rate": 5.279741364863244e-07,
"loss": 0.1051,
"step": 1627
},
{
"epoch": 2.6048,
"grad_norm": 0.5226642336746629,
"learning_rate": 5.238173962291881e-07,
"loss": 0.1165,
"step": 1628
},
{
"epoch": 2.6064,
"grad_norm": 0.5385886083756561,
"learning_rate": 5.196761790289639e-07,
"loss": 0.1218,
"step": 1629
},
{
"epoch": 2.608,
"grad_norm": 0.5674281733477105,
"learning_rate": 5.155504992470751e-07,
"loss": 0.1316,
"step": 1630
},
{
"epoch": 2.6096,
"grad_norm": 0.8158086288613888,
"learning_rate": 5.114403711910631e-07,
"loss": 0.1324,
"step": 1631
},
{
"epoch": 2.6112,
"grad_norm": 0.5106111231666498,
"learning_rate": 5.073458091145328e-07,
"loss": 0.1131,
"step": 1632
},
{
"epoch": 2.6128,
"grad_norm": 0.6062824817101468,
"learning_rate": 5.032668272171138e-07,
"loss": 0.1319,
"step": 1633
},
{
"epoch": 2.6144,
"grad_norm": 0.45526403974850765,
"learning_rate": 4.99203439644399e-07,
"loss": 0.1336,
"step": 1634
},
{
"epoch": 2.616,
"grad_norm": 0.524329352627648,
"learning_rate": 4.951556604879049e-07,
"loss": 0.1164,
"step": 1635
},
{
"epoch": 2.6176,
"grad_norm": 0.6217504270293189,
"learning_rate": 4.911235037850187e-07,
"loss": 0.1212,
"step": 1636
},
{
"epoch": 2.6192,
"grad_norm": 0.5765914967979414,
"learning_rate": 4.871069835189485e-07,
"loss": 0.1132,
"step": 1637
},
{
"epoch": 2.6208,
"grad_norm": 0.6492090303922368,
"learning_rate": 4.831061136186787e-07,
"loss": 0.1021,
"step": 1638
},
{
"epoch": 2.6224,
"grad_norm": 0.5694711852095302,
"learning_rate": 4.791209079589165e-07,
"loss": 0.1106,
"step": 1639
},
{
"epoch": 2.624,
"grad_norm": 0.5728656307017238,
"learning_rate": 4.7515138036005157e-07,
"loss": 0.1116,
"step": 1640
},
{
"epoch": 2.6256,
"grad_norm": 0.6106883224379283,
"learning_rate": 4.7119754458809727e-07,
"loss": 0.1335,
"step": 1641
},
{
"epoch": 2.6272,
"grad_norm": 0.5650807694637036,
"learning_rate": 4.672594143546538e-07,
"loss": 0.1151,
"step": 1642
},
{
"epoch": 2.6288,
"grad_norm": 0.5020457823361345,
"learning_rate": 4.6333700331685385e-07,
"loss": 0.144,
"step": 1643
},
{
"epoch": 2.6304,
"grad_norm": 0.5292755323303459,
"learning_rate": 4.594303250773152e-07,
"loss": 0.1344,
"step": 1644
},
{
"epoch": 2.632,
"grad_norm": 0.48109810268355935,
"learning_rate": 4.555393931841001e-07,
"loss": 0.1077,
"step": 1645
},
{
"epoch": 2.6336,
"grad_norm": 0.674566270483141,
"learning_rate": 4.5166422113065877e-07,
"loss": 0.1107,
"step": 1646
},
{
"epoch": 2.6352,
"grad_norm": 0.6677219383231198,
"learning_rate": 4.478048223557907e-07,
"loss": 0.1241,
"step": 1647
},
{
"epoch": 2.6368,
"grad_norm": 3.207723756777255,
"learning_rate": 4.439612102435942e-07,
"loss": 0.139,
"step": 1648
},
{
"epoch": 2.6384,
"grad_norm": 0.6670367130969531,
"learning_rate": 4.401333981234196e-07,
"loss": 0.1122,
"step": 1649
},
{
"epoch": 2.64,
"grad_norm": 0.668947400770753,
"learning_rate": 4.3632139926982676e-07,
"loss": 0.1374,
"step": 1650
},
{
"epoch": 2.6416,
"grad_norm": 0.5745367357537344,
"learning_rate": 4.325252269025315e-07,
"loss": 0.1142,
"step": 1651
},
{
"epoch": 2.6432,
"grad_norm": 0.5527128303827578,
"learning_rate": 4.287448941863692e-07,
"loss": 0.1337,
"step": 1652
},
{
"epoch": 2.6448,
"grad_norm": 0.6171076750159349,
"learning_rate": 4.249804142312436e-07,
"loss": 0.15,
"step": 1653
},
{
"epoch": 2.6464,
"grad_norm": 0.8209830429366161,
"learning_rate": 4.2123180009207956e-07,
"loss": 0.1135,
"step": 1654
},
{
"epoch": 2.648,
"grad_norm": 0.6183799106977054,
"learning_rate": 4.1749906476878486e-07,
"loss": 0.1227,
"step": 1655
},
{
"epoch": 2.6496,
"grad_norm": 0.6328711531358652,
"learning_rate": 4.137822212061965e-07,
"loss": 0.1213,
"step": 1656
},
{
"epoch": 2.6512000000000002,
"grad_norm": 0.6081588965617674,
"learning_rate": 4.100812822940431e-07,
"loss": 0.1349,
"step": 1657
},
{
"epoch": 2.6528,
"grad_norm": 0.6199953626306821,
"learning_rate": 4.063962608668959e-07,
"loss": 0.1194,
"step": 1658
},
{
"epoch": 2.6544,
"grad_norm": 0.6198443912523248,
"learning_rate": 4.0272716970412516e-07,
"loss": 0.1286,
"step": 1659
},
{
"epoch": 2.656,
"grad_norm": 0.5798820466341014,
"learning_rate": 3.990740215298583e-07,
"loss": 0.1038,
"step": 1660
},
{
"epoch": 2.6576,
"grad_norm": 0.6125597387369447,
"learning_rate": 3.954368290129301e-07,
"loss": 0.129,
"step": 1661
},
{
"epoch": 2.6592000000000002,
"grad_norm": 0.8104693569639763,
"learning_rate": 3.918156047668453e-07,
"loss": 0.1133,
"step": 1662
},
{
"epoch": 2.6608,
"grad_norm": 0.39990094126010445,
"learning_rate": 3.882103613497318e-07,
"loss": 0.1311,
"step": 1663
},
{
"epoch": 2.6624,
"grad_norm": 0.5493945954546839,
"learning_rate": 3.84621111264295e-07,
"loss": 0.1292,
"step": 1664
},
{
"epoch": 2.664,
"grad_norm": 0.5511140595838893,
"learning_rate": 3.810478669577794e-07,
"loss": 0.1322,
"step": 1665
},
{
"epoch": 2.6656,
"grad_norm": 0.6038073993666218,
"learning_rate": 3.7749064082191976e-07,
"loss": 0.1094,
"step": 1666
},
{
"epoch": 2.6672000000000002,
"grad_norm": 0.5322196820437206,
"learning_rate": 3.739494451929049e-07,
"loss": 0.1186,
"step": 1667
},
{
"epoch": 2.6688,
"grad_norm": 0.7566012475602327,
"learning_rate": 3.7042429235132625e-07,
"loss": 0.1183,
"step": 1668
},
{
"epoch": 2.6704,
"grad_norm": 0.5853005135211712,
"learning_rate": 3.6691519452214387e-07,
"loss": 0.123,
"step": 1669
},
{
"epoch": 2.672,
"grad_norm": 0.6264802316083258,
"learning_rate": 3.6342216387464047e-07,
"loss": 0.1383,
"step": 1670
},
{
"epoch": 2.6736,
"grad_norm": 0.5682086267972235,
"learning_rate": 3.5994521252237516e-07,
"loss": 0.1166,
"step": 1671
},
{
"epoch": 2.6752000000000002,
"grad_norm": 0.5850994958673718,
"learning_rate": 3.564843525231498e-07,
"loss": 0.1245,
"step": 1672
},
{
"epoch": 2.6768,
"grad_norm": 0.7119987545145817,
"learning_rate": 3.53039595878959e-07,
"loss": 0.1159,
"step": 1673
},
{
"epoch": 2.6784,
"grad_norm": 0.5011451347803199,
"learning_rate": 3.496109545359544e-07,
"loss": 0.1066,
"step": 1674
},
{
"epoch": 2.68,
"grad_norm": 0.5910071224168866,
"learning_rate": 3.461984403844015e-07,
"loss": 0.1466,
"step": 1675
},
{
"epoch": 2.6816,
"grad_norm": 0.582163374242144,
"learning_rate": 3.42802065258635e-07,
"loss": 0.1345,
"step": 1676
},
{
"epoch": 2.6832000000000003,
"grad_norm": 0.6116553034928778,
"learning_rate": 3.394218409370242e-07,
"loss": 0.1217,
"step": 1677
},
{
"epoch": 2.6848,
"grad_norm": 0.6420769008924371,
"learning_rate": 3.360577791419256e-07,
"loss": 0.1059,
"step": 1678
},
{
"epoch": 2.6864,
"grad_norm": 0.5749332703664223,
"learning_rate": 3.3270989153964707e-07,
"loss": 0.1086,
"step": 1679
},
{
"epoch": 2.6879999999999997,
"grad_norm": 0.4708129830539554,
"learning_rate": 3.2937818974040637e-07,
"loss": 0.1116,
"step": 1680
},
{
"epoch": 2.6896,
"grad_norm": 0.586336296860425,
"learning_rate": 3.260626852982873e-07,
"loss": 0.1258,
"step": 1681
},
{
"epoch": 2.6912000000000003,
"grad_norm": 0.5162340202373774,
"learning_rate": 3.227633897112059e-07,
"loss": 0.1292,
"step": 1682
},
{
"epoch": 2.6928,
"grad_norm": 0.5342294671618151,
"learning_rate": 3.194803144208636e-07,
"loss": 0.12,
"step": 1683
},
{
"epoch": 2.6944,
"grad_norm": 0.5090430977103828,
"learning_rate": 3.16213470812714e-07,
"loss": 0.1353,
"step": 1684
},
{
"epoch": 2.6959999999999997,
"grad_norm": 0.6251269471736536,
"learning_rate": 3.129628702159204e-07,
"loss": 0.1302,
"step": 1685
},
{
"epoch": 2.6976,
"grad_norm": 0.47163049692183523,
"learning_rate": 3.097285239033138e-07,
"loss": 0.111,
"step": 1686
},
{
"epoch": 2.6992000000000003,
"grad_norm": 0.6074180592926421,
"learning_rate": 3.0651044309136016e-07,
"loss": 0.0994,
"step": 1687
},
{
"epoch": 2.7008,
"grad_norm": 0.8525974992494059,
"learning_rate": 3.033086389401141e-07,
"loss": 0.122,
"step": 1688
},
{
"epoch": 2.7024,
"grad_norm": 0.5359667640254151,
"learning_rate": 3.0012312255318696e-07,
"loss": 0.1386,
"step": 1689
},
{
"epoch": 2.7039999999999997,
"grad_norm": 0.8221075128341616,
"learning_rate": 2.9695390497770535e-07,
"loss": 0.1391,
"step": 1690
},
{
"epoch": 2.7056,
"grad_norm": 0.5516455630845664,
"learning_rate": 2.93800997204271e-07,
"loss": 0.1155,
"step": 1691
},
{
"epoch": 2.7072000000000003,
"grad_norm": 0.7569992216863624,
"learning_rate": 2.9066441016692594e-07,
"loss": 0.1127,
"step": 1692
},
{
"epoch": 2.7088,
"grad_norm": 0.49490463017347885,
"learning_rate": 2.8754415474311235e-07,
"loss": 0.114,
"step": 1693
},
{
"epoch": 2.7104,
"grad_norm": 0.6162250336936629,
"learning_rate": 2.844402417536374e-07,
"loss": 0.1335,
"step": 1694
},
{
"epoch": 2.7119999999999997,
"grad_norm": 0.5722797436738771,
"learning_rate": 2.8135268196263055e-07,
"loss": 0.1152,
"step": 1695
},
{
"epoch": 2.7136,
"grad_norm": 0.5709609888528145,
"learning_rate": 2.782814860775124e-07,
"loss": 0.1209,
"step": 1696
},
{
"epoch": 2.7152,
"grad_norm": 0.6254651292823231,
"learning_rate": 2.752266647489549e-07,
"loss": 0.1139,
"step": 1697
},
{
"epoch": 2.7168,
"grad_norm": 0.6592631981929915,
"learning_rate": 2.7218822857084217e-07,
"loss": 0.1209,
"step": 1698
},
{
"epoch": 2.7184,
"grad_norm": 0.5400486321421776,
"learning_rate": 2.691661880802382e-07,
"loss": 0.1172,
"step": 1699
},
{
"epoch": 2.7199999999999998,
"grad_norm": 0.5606531570274645,
"learning_rate": 2.661605537573453e-07,
"loss": 0.1271,
"step": 1700
},
{
"epoch": 2.7216,
"grad_norm": 0.8131174792629867,
"learning_rate": 2.631713360254734e-07,
"loss": 0.125,
"step": 1701
},
{
"epoch": 2.7232,
"grad_norm": 0.4885001706796863,
"learning_rate": 2.6019854525099977e-07,
"loss": 0.1216,
"step": 1702
},
{
"epoch": 2.7248,
"grad_norm": 0.627595798159668,
"learning_rate": 2.572421917433332e-07,
"loss": 0.1156,
"step": 1703
},
{
"epoch": 2.7264,
"grad_norm": 0.8880016201864134,
"learning_rate": 2.5430228575488156e-07,
"loss": 0.1058,
"step": 1704
},
{
"epoch": 2.7279999999999998,
"grad_norm": 0.5518965232423588,
"learning_rate": 2.513788374810111e-07,
"loss": 0.1153,
"step": 1705
},
{
"epoch": 2.7296,
"grad_norm": 0.7450244465110475,
"learning_rate": 2.4847185706001643e-07,
"loss": 0.1321,
"step": 1706
},
{
"epoch": 2.7312,
"grad_norm": 0.4687152932488462,
"learning_rate": 2.455813545730812e-07,
"loss": 0.1345,
"step": 1707
},
{
"epoch": 2.7328,
"grad_norm": 0.5393196574050693,
"learning_rate": 2.4270734004424643e-07,
"loss": 0.1349,
"step": 1708
},
{
"epoch": 2.7344,
"grad_norm": 0.6071860478776072,
"learning_rate": 2.39849823440374e-07,
"loss": 0.1342,
"step": 1709
},
{
"epoch": 2.7359999999999998,
"grad_norm": 0.7378054124140381,
"learning_rate": 2.3700881467111025e-07,
"loss": 0.1292,
"step": 1710
},
{
"epoch": 2.7376,
"grad_norm": 0.6010891884147711,
"learning_rate": 2.3418432358885633e-07,
"loss": 0.1279,
"step": 1711
},
{
"epoch": 2.7392,
"grad_norm": 0.5288414801025237,
"learning_rate": 2.3137635998872808e-07,
"loss": 0.1205,
"step": 1712
},
{
"epoch": 2.7408,
"grad_norm": 0.7112989919409005,
"learning_rate": 2.285849336085294e-07,
"loss": 0.1381,
"step": 1713
},
{
"epoch": 2.7424,
"grad_norm": 0.6821053930178859,
"learning_rate": 2.258100541287117e-07,
"loss": 0.1265,
"step": 1714
},
{
"epoch": 2.7439999999999998,
"grad_norm": 0.7119019089361959,
"learning_rate": 2.2305173117234236e-07,
"loss": 0.1138,
"step": 1715
},
{
"epoch": 2.7456,
"grad_norm": 0.5345854428057913,
"learning_rate": 2.2030997430507462e-07,
"loss": 0.1102,
"step": 1716
},
{
"epoch": 2.7472,
"grad_norm": 0.5022015625396683,
"learning_rate": 2.1758479303510937e-07,
"loss": 0.1352,
"step": 1717
},
{
"epoch": 2.7488,
"grad_norm": 0.688063806232108,
"learning_rate": 2.148761968131663e-07,
"loss": 0.1476,
"step": 1718
},
{
"epoch": 2.7504,
"grad_norm": 0.6117011903074215,
"learning_rate": 2.121841950324488e-07,
"loss": 0.1187,
"step": 1719
},
{
"epoch": 2.752,
"grad_norm": 0.6608374040176052,
"learning_rate": 2.0950879702861082e-07,
"loss": 0.1249,
"step": 1720
},
{
"epoch": 2.7536,
"grad_norm": 0.6037962372447667,
"learning_rate": 2.0685001207972843e-07,
"loss": 0.1291,
"step": 1721
},
{
"epoch": 2.7552,
"grad_norm": 0.5116346912127422,
"learning_rate": 2.042078494062616e-07,
"loss": 0.1151,
"step": 1722
},
{
"epoch": 2.7568,
"grad_norm": 0.6154298295733341,
"learning_rate": 2.0158231817102858e-07,
"loss": 0.124,
"step": 1723
},
{
"epoch": 2.7584,
"grad_norm": 0.6686585881999877,
"learning_rate": 1.9897342747916938e-07,
"loss": 0.1133,
"step": 1724
},
{
"epoch": 2.76,
"grad_norm": 0.6537413008185268,
"learning_rate": 1.9638118637811564e-07,
"loss": 0.1453,
"step": 1725
},
{
"epoch": 2.7616,
"grad_norm": 0.6651876711050978,
"learning_rate": 1.9380560385756088e-07,
"loss": 0.1279,
"step": 1726
},
{
"epoch": 2.7632,
"grad_norm": 0.5505418831099438,
"learning_rate": 1.9124668884942632e-07,
"loss": 0.1312,
"step": 1727
},
{
"epoch": 2.7648,
"grad_norm": 0.6299015851527305,
"learning_rate": 1.8870445022783234e-07,
"loss": 0.1426,
"step": 1728
},
{
"epoch": 2.7664,
"grad_norm": 0.5879161271033048,
"learning_rate": 1.861788968090683e-07,
"loss": 0.1248,
"step": 1729
},
{
"epoch": 2.768,
"grad_norm": 0.509027301695031,
"learning_rate": 1.8367003735155764e-07,
"loss": 0.1178,
"step": 1730
},
{
"epoch": 2.7696,
"grad_norm": 0.5791747697524069,
"learning_rate": 1.8117788055583286e-07,
"loss": 0.1292,
"step": 1731
},
{
"epoch": 2.7712,
"grad_norm": 0.5496391017826657,
"learning_rate": 1.7870243506450113e-07,
"loss": 0.1113,
"step": 1732
},
{
"epoch": 2.7728,
"grad_norm": 0.564290811947087,
"learning_rate": 1.762437094622177e-07,
"loss": 0.1258,
"step": 1733
},
{
"epoch": 2.7744,
"grad_norm": 0.5072066025828119,
"learning_rate": 1.738017122756541e-07,
"loss": 0.1206,
"step": 1734
},
{
"epoch": 2.776,
"grad_norm": 0.6918835154095383,
"learning_rate": 1.713764519734673e-07,
"loss": 0.1298,
"step": 1735
},
{
"epoch": 2.7776,
"grad_norm": 0.6702026642610149,
"learning_rate": 1.68967936966275e-07,
"loss": 0.1267,
"step": 1736
},
{
"epoch": 2.7792,
"grad_norm": 0.5528682773813236,
"learning_rate": 1.6657617560662088e-07,
"loss": 0.1265,
"step": 1737
},
{
"epoch": 2.7808,
"grad_norm": 0.5560555476095583,
"learning_rate": 1.6420117618895003e-07,
"loss": 0.1026,
"step": 1738
},
{
"epoch": 2.7824,
"grad_norm": 0.5915861152185402,
"learning_rate": 1.6184294694957747e-07,
"loss": 0.1193,
"step": 1739
},
{
"epoch": 2.784,
"grad_norm": 0.5404956883495103,
"learning_rate": 1.5950149606666077e-07,
"loss": 0.1121,
"step": 1740
},
{
"epoch": 2.7856,
"grad_norm": 0.4801285410003388,
"learning_rate": 1.5717683166017184e-07,
"loss": 0.1096,
"step": 1741
},
{
"epoch": 2.7872,
"grad_norm": 0.5904169080677608,
"learning_rate": 1.5486896179186693e-07,
"loss": 0.1158,
"step": 1742
},
{
"epoch": 2.7888,
"grad_norm": 0.605840518157387,
"learning_rate": 1.5257789446526172e-07,
"loss": 0.1223,
"step": 1743
},
{
"epoch": 2.7904,
"grad_norm": 0.6092744859859395,
"learning_rate": 1.5030363762560228e-07,
"loss": 0.1216,
"step": 1744
},
{
"epoch": 2.792,
"grad_norm": 0.6679141709717694,
"learning_rate": 1.480461991598353e-07,
"loss": 0.1179,
"step": 1745
},
{
"epoch": 2.7936,
"grad_norm": 0.5292627252565983,
"learning_rate": 1.458055868965841e-07,
"loss": 0.1165,
"step": 1746
},
{
"epoch": 2.7952,
"grad_norm": 0.6164888336048616,
"learning_rate": 1.4358180860611913e-07,
"loss": 0.1228,
"step": 1747
},
{
"epoch": 2.7968,
"grad_norm": 0.5182774606277455,
"learning_rate": 1.4137487200033383e-07,
"loss": 0.1178,
"step": 1748
},
{
"epoch": 2.7984,
"grad_norm": 0.4999840953251745,
"learning_rate": 1.3918478473271325e-07,
"loss": 0.1365,
"step": 1749
},
{
"epoch": 2.8,
"grad_norm": 0.5572717645841915,
"learning_rate": 1.3701155439831249e-07,
"loss": 0.1334,
"step": 1750
},
{
"epoch": 2.8016,
"grad_norm": 0.6966418649624069,
"learning_rate": 1.3485518853372625e-07,
"loss": 0.1283,
"step": 1751
},
{
"epoch": 2.8032,
"grad_norm": 0.553241091520799,
"learning_rate": 1.3271569461706547e-07,
"loss": 0.1279,
"step": 1752
},
{
"epoch": 2.8048,
"grad_norm": 0.5837355106522873,
"learning_rate": 1.305930800679317e-07,
"loss": 0.1226,
"step": 1753
},
{
"epoch": 2.8064,
"grad_norm": 0.6042772297563292,
"learning_rate": 1.2848735224738729e-07,
"loss": 0.1111,
"step": 1754
},
{
"epoch": 2.808,
"grad_norm": 0.663059306343012,
"learning_rate": 1.2639851845793583e-07,
"loss": 0.1201,
"step": 1755
},
{
"epoch": 2.8096,
"grad_norm": 0.5614631003954963,
"learning_rate": 1.2432658594349113e-07,
"loss": 0.1453,
"step": 1756
},
{
"epoch": 2.8112,
"grad_norm": 0.605420680626281,
"learning_rate": 1.2227156188935552e-07,
"loss": 0.1097,
"step": 1757
},
{
"epoch": 2.8128,
"grad_norm": 0.6106841783064658,
"learning_rate": 1.202334534221955e-07,
"loss": 0.1179,
"step": 1758
},
{
"epoch": 2.8144,
"grad_norm": 0.7587168903110397,
"learning_rate": 1.1821226761001391e-07,
"loss": 0.1265,
"step": 1759
},
{
"epoch": 2.816,
"grad_norm": 0.8084315026479338,
"learning_rate": 1.1620801146212723e-07,
"loss": 0.1317,
"step": 1760
},
{
"epoch": 2.8176,
"grad_norm": 0.4590565272624402,
"learning_rate": 1.1422069192914221e-07,
"loss": 0.115,
"step": 1761
},
{
"epoch": 2.8192,
"grad_norm": 0.6776379078873239,
"learning_rate": 1.1225031590292923e-07,
"loss": 0.1325,
"step": 1762
},
{
"epoch": 2.8208,
"grad_norm": 0.5479995794917868,
"learning_rate": 1.1029689021660183e-07,
"loss": 0.1289,
"step": 1763
},
{
"epoch": 2.8224,
"grad_norm": 0.7132907246069775,
"learning_rate": 1.0836042164448945e-07,
"loss": 0.1509,
"step": 1764
},
{
"epoch": 2.824,
"grad_norm": 0.7417827576754505,
"learning_rate": 1.0644091690211633e-07,
"loss": 0.1227,
"step": 1765
},
{
"epoch": 2.8256,
"grad_norm": 0.6509952242894306,
"learning_rate": 1.0453838264617711e-07,
"loss": 0.1199,
"step": 1766
},
{
"epoch": 2.8272,
"grad_norm": 0.5711565520328492,
"learning_rate": 1.0265282547451405e-07,
"loss": 0.1235,
"step": 1767
},
{
"epoch": 2.8288,
"grad_norm": 0.5547174133495429,
"learning_rate": 1.0078425192609487e-07,
"loss": 0.1151,
"step": 1768
},
{
"epoch": 2.8304,
"grad_norm": 0.5741995518197428,
"learning_rate": 9.893266848098826e-08,
"loss": 0.1219,
"step": 1769
},
{
"epoch": 2.832,
"grad_norm": 0.4925073617875599,
"learning_rate": 9.709808156034394e-08,
"loss": 0.1193,
"step": 1770
},
{
"epoch": 2.8336,
"grad_norm": 0.678871259076856,
"learning_rate": 9.528049752636714e-08,
"loss": 0.1122,
"step": 1771
},
{
"epoch": 2.8352,
"grad_norm": 0.9126382093423114,
"learning_rate": 9.347992268230022e-08,
"loss": 0.117,
"step": 1772
},
{
"epoch": 2.8368,
"grad_norm": 0.9651993250827703,
"learning_rate": 9.169636327239883e-08,
"loss": 0.1201,
"step": 1773
},
{
"epoch": 2.8384,
"grad_norm": 0.591775893701563,
"learning_rate": 8.992982548190809e-08,
"loss": 0.1289,
"step": 1774
},
{
"epoch": 2.84,
"grad_norm": 0.5782820796840922,
"learning_rate": 8.818031543704641e-08,
"loss": 0.1226,
"step": 1775
},
{
"epoch": 2.8416,
"grad_norm": 0.4817839203347196,
"learning_rate": 8.644783920498001e-08,
"loss": 0.1153,
"step": 1776
},
{
"epoch": 2.8432,
"grad_norm": 0.6201647534672659,
"learning_rate": 8.473240279380235e-08,
"loss": 0.1222,
"step": 1777
},
{
"epoch": 2.8448,
"grad_norm": 0.5639150945444918,
"learning_rate": 8.303401215251583e-08,
"loss": 0.1253,
"step": 1778
},
{
"epoch": 2.8464,
"grad_norm": 0.5447717087410271,
"learning_rate": 8.135267317100792e-08,
"loss": 0.1133,
"step": 1779
},
{
"epoch": 2.848,
"grad_norm": 0.7942411082263157,
"learning_rate": 7.968839168003395e-08,
"loss": 0.1303,
"step": 1780
},
{
"epoch": 2.8496,
"grad_norm": 0.5504263382483138,
"learning_rate": 7.804117345119266e-08,
"loss": 0.1141,
"step": 1781
},
{
"epoch": 2.8512,
"grad_norm": 0.6111951780064648,
"learning_rate": 7.64110241969107e-08,
"loss": 0.1203,
"step": 1782
},
{
"epoch": 2.8528000000000002,
"grad_norm": 0.6382521212284029,
"learning_rate": 7.479794957042041e-08,
"loss": 0.1323,
"step": 1783
},
{
"epoch": 2.8544,
"grad_norm": 0.5737860571266434,
"learning_rate": 7.320195516574036e-08,
"loss": 0.1204,
"step": 1784
},
{
"epoch": 2.856,
"grad_norm": 0.6215264864278954,
"learning_rate": 7.16230465176565e-08,
"loss": 0.0822,
"step": 1785
},
{
"epoch": 2.8576,
"grad_norm": 0.5454862381493043,
"learning_rate": 7.00612291017022e-08,
"loss": 0.1162,
"step": 1786
},
{
"epoch": 2.8592,
"grad_norm": 0.7638928319845638,
"learning_rate": 6.851650833414103e-08,
"loss": 0.1294,
"step": 1787
},
{
"epoch": 2.8608000000000002,
"grad_norm": 0.5545884695767085,
"learning_rate": 6.698888957194505e-08,
"loss": 0.1237,
"step": 1788
},
{
"epoch": 2.8624,
"grad_norm": 0.5184627270134425,
"learning_rate": 6.547837811277824e-08,
"loss": 0.1135,
"step": 1789
},
{
"epoch": 2.864,
"grad_norm": 0.622176656828793,
"learning_rate": 6.39849791949787e-08,
"loss": 0.1192,
"step": 1790
},
{
"epoch": 2.8656,
"grad_norm": 0.6338405174138664,
"learning_rate": 6.250869799753866e-08,
"loss": 0.1298,
"step": 1791
},
{
"epoch": 2.8672,
"grad_norm": 0.47502034942621923,
"learning_rate": 6.104953964008897e-08,
"loss": 0.1071,
"step": 1792
},
{
"epoch": 2.8688000000000002,
"grad_norm": 0.5810524718166514,
"learning_rate": 5.960750918287627e-08,
"loss": 0.1231,
"step": 1793
},
{
"epoch": 2.8704,
"grad_norm": 0.5648855468150723,
"learning_rate": 5.818261162675309e-08,
"loss": 0.1166,
"step": 1794
},
{
"epoch": 2.872,
"grad_norm": 0.5347635999690873,
"learning_rate": 5.677485191315391e-08,
"loss": 0.1134,
"step": 1795
},
{
"epoch": 2.8736,
"grad_norm": 0.6473683179170193,
"learning_rate": 5.538423492408129e-08,
"loss": 0.1196,
"step": 1796
},
{
"epoch": 2.8752,
"grad_norm": 0.7074484466102422,
"learning_rate": 5.401076548208761e-08,
"loss": 0.1315,
"step": 1797
},
{
"epoch": 2.8768000000000002,
"grad_norm": 0.5277075532355038,
"learning_rate": 5.265444835025946e-08,
"loss": 0.1234,
"step": 1798
},
{
"epoch": 2.8784,
"grad_norm": 0.6736162431978554,
"learning_rate": 5.1315288232201e-08,
"loss": 0.1156,
"step": 1799
},
{
"epoch": 2.88,
"grad_norm": 0.5430961317505087,
"learning_rate": 4.9993289772015116e-08,
"loss": 0.1349,
"step": 1800
},
{
"epoch": 2.8816,
"grad_norm": 0.5228554751240404,
"learning_rate": 4.8688457554291746e-08,
"loss": 0.1179,
"step": 1801
},
{
"epoch": 2.8832,
"grad_norm": 0.5857376858950251,
"learning_rate": 4.7400796104088434e-08,
"loss": 0.1254,
"step": 1802
},
{
"epoch": 2.8848000000000003,
"grad_norm": 0.522763015041997,
"learning_rate": 4.613030988691536e-08,
"loss": 0.1113,
"step": 1803
},
{
"epoch": 2.8864,
"grad_norm": 0.6355966376024335,
"learning_rate": 4.4877003308722575e-08,
"loss": 0.1216,
"step": 1804
},
{
"epoch": 2.888,
"grad_norm": 0.5055099230235414,
"learning_rate": 4.364088071587891e-08,
"loss": 0.1113,
"step": 1805
},
{
"epoch": 2.8895999999999997,
"grad_norm": 0.6315262071616932,
"learning_rate": 4.2421946395164174e-08,
"loss": 0.1281,
"step": 1806
},
{
"epoch": 2.8912,
"grad_norm": 0.5152266606162141,
"learning_rate": 4.1220204573747534e-08,
"loss": 0.1201,
"step": 1807
},
{
"epoch": 2.8928000000000003,
"grad_norm": 0.6057143805022445,
"learning_rate": 4.0035659419178086e-08,
"loss": 0.1351,
"step": 1808
},
{
"epoch": 2.8944,
"grad_norm": 0.5678226365092268,
"learning_rate": 3.88683150393665e-08,
"loss": 0.1175,
"step": 1809
},
{
"epoch": 2.896,
"grad_norm": 0.5848121006003573,
"learning_rate": 3.771817548257395e-08,
"loss": 0.1264,
"step": 1810
},
{
"epoch": 2.8975999999999997,
"grad_norm": 1.0577628024040964,
"learning_rate": 3.658524473739544e-08,
"loss": 0.1228,
"step": 1811
},
{
"epoch": 2.8992,
"grad_norm": 0.5974050615353388,
"learning_rate": 3.546952673274817e-08,
"loss": 0.1316,
"step": 1812
},
{
"epoch": 2.9008000000000003,
"grad_norm": 0.5702494602443352,
"learning_rate": 3.437102533785541e-08,
"loss": 0.1245,
"step": 1813
},
{
"epoch": 2.9024,
"grad_norm": 0.6012515395235518,
"learning_rate": 3.328974436223709e-08,
"loss": 0.1276,
"step": 1814
},
{
"epoch": 2.904,
"grad_norm": 0.5992390071092003,
"learning_rate": 3.2225687555690886e-08,
"loss": 0.1255,
"step": 1815
},
{
"epoch": 2.9055999999999997,
"grad_norm": 0.5933694959224938,
"learning_rate": 3.117885860828396e-08,
"loss": 0.1118,
"step": 1816
},
{
"epoch": 2.9072,
"grad_norm": 0.6427090696823885,
"learning_rate": 3.014926115034012e-08,
"loss": 0.1217,
"step": 1817
},
{
"epoch": 2.9088000000000003,
"grad_norm": 0.6071274179775048,
"learning_rate": 2.9136898752422648e-08,
"loss": 0.117,
"step": 1818
},
{
"epoch": 2.9104,
"grad_norm": 0.5757517558130233,
"learning_rate": 2.8141774925327103e-08,
"loss": 0.1318,
"step": 1819
},
{
"epoch": 2.912,
"grad_norm": 0.6194274620287339,
"learning_rate": 2.7163893120066288e-08,
"loss": 0.1138,
"step": 1820
},
{
"epoch": 2.9135999999999997,
"grad_norm": 0.6659093498270136,
"learning_rate": 2.6203256727859172e-08,
"loss": 0.1198,
"step": 1821
},
{
"epoch": 2.9152,
"grad_norm": 0.6500624637169251,
"learning_rate": 2.5259869080118127e-08,
"loss": 0.1353,
"step": 1822
},
{
"epoch": 2.9168,
"grad_norm": 0.6334904487218471,
"learning_rate": 2.4333733448440033e-08,
"loss": 0.1093,
"step": 1823
},
{
"epoch": 2.9184,
"grad_norm": 0.6397958740870024,
"learning_rate": 2.34248530445913e-08,
"loss": 0.1336,
"step": 1824
},
{
"epoch": 2.92,
"grad_norm": 0.821775012600362,
"learning_rate": 2.2533231020499536e-08,
"loss": 0.1175,
"step": 1825
},
{
"epoch": 2.9215999999999998,
"grad_norm": 0.853402346059548,
"learning_rate": 2.1658870468241332e-08,
"loss": 0.1321,
"step": 1826
},
{
"epoch": 2.9232,
"grad_norm": 0.5700131123645366,
"learning_rate": 2.0801774420031172e-08,
"loss": 0.1291,
"step": 1827
},
{
"epoch": 2.9248,
"grad_norm": 0.6314515623817292,
"learning_rate": 1.9961945848213092e-08,
"loss": 0.1228,
"step": 1828
},
{
"epoch": 2.9264,
"grad_norm": 0.5033981358378709,
"learning_rate": 1.9139387665247922e-08,
"loss": 0.122,
"step": 1829
},
{
"epoch": 2.928,
"grad_norm": 0.6061791960718723,
"learning_rate": 1.8334102723703286e-08,
"loss": 0.1334,
"step": 1830
},
{
"epoch": 2.9295999999999998,
"grad_norm": 0.5929330456180494,
"learning_rate": 1.754609381624639e-08,
"loss": 0.1291,
"step": 1831
},
{
"epoch": 2.9312,
"grad_norm": 0.5620984127132138,
"learning_rate": 1.677536367563126e-08,
"loss": 0.1121,
"step": 1832
},
{
"epoch": 2.9328,
"grad_norm": 0.5643386329264197,
"learning_rate": 1.6021914974690413e-08,
"loss": 0.1194,
"step": 1833
},
{
"epoch": 2.9344,
"grad_norm": 0.6436418946698571,
"learning_rate": 1.5285750326325953e-08,
"loss": 0.1449,
"step": 1834
},
{
"epoch": 2.936,
"grad_norm": 0.5955174063162612,
"learning_rate": 1.4566872283500733e-08,
"loss": 0.1013,
"step": 1835
},
{
"epoch": 2.9375999999999998,
"grad_norm": 0.7573844523944555,
"learning_rate": 1.3865283339228319e-08,
"loss": 0.113,
"step": 1836
},
{
"epoch": 2.9392,
"grad_norm": 0.6353115609045219,
"learning_rate": 1.3180985926564693e-08,
"loss": 0.1284,
"step": 1837
},
{
"epoch": 2.9408,
"grad_norm": 0.7802770309850776,
"learning_rate": 1.2513982418601024e-08,
"loss": 0.1317,
"step": 1838
},
{
"epoch": 2.9424,
"grad_norm": 0.5853040123883677,
"learning_rate": 1.1864275128454783e-08,
"loss": 0.1193,
"step": 1839
},
{
"epoch": 2.944,
"grad_norm": 0.6302848731136266,
"learning_rate": 1.1231866309259764e-08,
"loss": 0.1066,
"step": 1840
},
{
"epoch": 2.9455999999999998,
"grad_norm": 0.7317355768502972,
"learning_rate": 1.0616758154161633e-08,
"loss": 0.1202,
"step": 1841
},
{
"epoch": 2.9472,
"grad_norm": 0.569323884781362,
"learning_rate": 1.0018952796307934e-08,
"loss": 0.13,
"step": 1842
},
{
"epoch": 2.9488,
"grad_norm": 0.6578467266711805,
"learning_rate": 9.438452308841995e-09,
"loss": 0.1306,
"step": 1843
},
{
"epoch": 2.9504,
"grad_norm": 0.8308252015902933,
"learning_rate": 8.87525870489514e-09,
"loss": 0.1455,
"step": 1844
},
{
"epoch": 2.952,
"grad_norm": 0.559135582948155,
"learning_rate": 8.329373937578378e-09,
"loss": 0.1065,
"step": 1845
},
{
"epoch": 2.9536,
"grad_norm": 0.5307478335199818,
"learning_rate": 7.800799899979061e-09,
"loss": 0.1176,
"step": 1846
},
{
"epoch": 2.9552,
"grad_norm": 0.63506245719894,
"learning_rate": 7.289538425150899e-09,
"loss": 0.1253,
"step": 1847
},
{
"epoch": 2.9568,
"grad_norm": 0.549216287695388,
"learning_rate": 6.7955912861095155e-09,
"loss": 0.1388,
"step": 1848
},
{
"epoch": 2.9584,
"grad_norm": 1.1618041275136664,
"learning_rate": 6.31896019582523e-09,
"loss": 0.1268,
"step": 1849
},
{
"epoch": 2.96,
"grad_norm": 1.0005096250016008,
"learning_rate": 5.8596468072180665e-09,
"loss": 0.1187,
"step": 1850
},
{
"epoch": 2.9616,
"grad_norm": 0.658748161661329,
"learning_rate": 5.417652713152199e-09,
"loss": 0.1412,
"step": 1851
},
{
"epoch": 2.9632,
"grad_norm": 0.6061273591339458,
"learning_rate": 4.992979446428736e-09,
"loss": 0.1204,
"step": 1852
},
{
"epoch": 2.9648,
"grad_norm": 0.5677575670277463,
"learning_rate": 4.585628479781279e-09,
"loss": 0.1064,
"step": 1853
},
{
"epoch": 2.9664,
"grad_norm": 0.49744277512190194,
"learning_rate": 4.195601225872592e-09,
"loss": 0.1188,
"step": 1854
},
{
"epoch": 2.968,
"grad_norm": 0.6696521950901151,
"learning_rate": 3.822899037286276e-09,
"loss": 0.1031,
"step": 1855
},
{
"epoch": 2.9696,
"grad_norm": 2.2047954286301206,
"learning_rate": 3.4675232065256583e-09,
"loss": 0.1052,
"step": 1856
},
{
"epoch": 2.9712,
"grad_norm": 0.5297447166582149,
"learning_rate": 3.129474966006574e-09,
"loss": 0.1158,
"step": 1857
},
{
"epoch": 2.9728,
"grad_norm": 0.48166574817142765,
"learning_rate": 2.808755488054038e-09,
"loss": 0.1215,
"step": 1858
},
{
"epoch": 2.9744,
"grad_norm": 0.5286377648531796,
"learning_rate": 2.5053658848989137e-09,
"loss": 0.1079,
"step": 1859
},
{
"epoch": 2.976,
"grad_norm": 0.5144940454190761,
"learning_rate": 2.219307208672361e-09,
"loss": 0.1292,
"step": 1860
},
{
"epoch": 2.9776,
"grad_norm": 0.5938876308373047,
"learning_rate": 1.9505804514047266e-09,
"loss": 0.1188,
"step": 1861
},
{
"epoch": 2.9792,
"grad_norm": 0.7195759843588121,
"learning_rate": 1.6991865450188827e-09,
"loss": 0.1227,
"step": 1862
},
{
"epoch": 2.9808,
"grad_norm": 1.053302453809404,
"learning_rate": 1.465126361330227e-09,
"loss": 0.1202,
"step": 1863
},
{
"epoch": 2.9824,
"grad_norm": 0.5611986756309829,
"learning_rate": 1.2484007120411312e-09,
"loss": 0.1293,
"step": 1864
},
{
"epoch": 2.984,
"grad_norm": 0.6273623273337964,
"learning_rate": 1.0490103487392766e-09,
"loss": 0.1352,
"step": 1865
},
{
"epoch": 2.9856,
"grad_norm": 0.7957994004997752,
"learning_rate": 8.669559628954327e-10,
"loss": 0.1082,
"step": 1866
},
{
"epoch": 2.9872,
"grad_norm": 0.5300756566545812,
"learning_rate": 7.02238185860682e-10,
"loss": 0.1289,
"step": 1867
},
{
"epoch": 2.9888,
"grad_norm": 0.4958503737015836,
"learning_rate": 5.54857588862534e-10,
"loss": 0.1131,
"step": 1868
},
{
"epoch": 2.9904,
"grad_norm": 0.5367274406949382,
"learning_rate": 4.2481468300603625e-10,
"loss": 0.1166,
"step": 1869
},
{
"epoch": 2.992,
"grad_norm": 0.569752118377782,
"learning_rate": 3.1210991927044244e-10,
"loss": 0.1422,
"step": 1870
},
{
"epoch": 2.9936,
"grad_norm": 0.5824280004872976,
"learning_rate": 2.167436885064378e-10,
"loss": 0.1258,
"step": 1871
},
{
"epoch": 2.9952,
"grad_norm": 0.6232476405595663,
"learning_rate": 1.387163214372489e-10,
"loss": 0.1314,
"step": 1872
},
{
"epoch": 2.9968,
"grad_norm": 0.6923552082565289,
"learning_rate": 7.80280886558682e-11,
"loss": 0.1281,
"step": 1873
},
{
"epoch": 2.9984,
"grad_norm": 0.5622680510465953,
"learning_rate": 3.467920062394381e-11,
"loss": 0.1402,
"step": 1874
},
{
"epoch": 3.0,
"grad_norm": 0.5672700396607967,
"learning_rate": 8.669807672334606e-12,
"loss": 0.1312,
"step": 1875
},
{
"epoch": 3.0,
"step": 1875,
"total_flos": 43566831157248.0,
"train_loss": 0.17250009640455247,
"train_runtime": 4072.8546,
"train_samples_per_second": 3.683,
"train_steps_per_second": 0.46
}
],
"logging_steps": 1.0,
"max_steps": 1875,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 200,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 43566831157248.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}