Files
openthoughts3_100k_qwen25_1…/trainer_state.json
ModelHub XC c87c4f234b 初始化项目,由ModelHub XC社区提供模型
Model: mlfoundations-dev/openthoughts3_100k_qwen25_1b_bsz512_lr16e5_epochs5
Source: Original Platform
2026-09-20 09:48:14 +08:00

6868 lines
166 KiB
JSON

{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 4.987212276214834,
"eval_steps": 500,
"global_step": 975,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.005115089514066497,
"grad_norm": 2.9552195401086414,
"learning_rate": 1.6326530612244897e-06,
"loss": 1.4957,
"step": 1
},
{
"epoch": 0.010230179028132993,
"grad_norm": 2.9842410023586825,
"learning_rate": 3.2653061224489794e-06,
"loss": 1.4889,
"step": 2
},
{
"epoch": 0.015345268542199489,
"grad_norm": 2.9723829103846526,
"learning_rate": 4.897959183673469e-06,
"loss": 1.499,
"step": 3
},
{
"epoch": 0.020460358056265986,
"grad_norm": 2.631673524670789,
"learning_rate": 6.530612244897959e-06,
"loss": 1.4779,
"step": 4
},
{
"epoch": 0.02557544757033248,
"grad_norm": 1.98326412510638,
"learning_rate": 8.16326530612245e-06,
"loss": 1.4978,
"step": 5
},
{
"epoch": 0.030690537084398978,
"grad_norm": 1.8204978664324611,
"learning_rate": 9.795918367346939e-06,
"loss": 1.4464,
"step": 6
},
{
"epoch": 0.03580562659846547,
"grad_norm": 1.7948446435007896,
"learning_rate": 1.1428571428571429e-05,
"loss": 1.4445,
"step": 7
},
{
"epoch": 0.04092071611253197,
"grad_norm": 1.8657212421288438,
"learning_rate": 1.3061224489795918e-05,
"loss": 1.4051,
"step": 8
},
{
"epoch": 0.04603580562659847,
"grad_norm": 2.048085010274612,
"learning_rate": 1.469387755102041e-05,
"loss": 1.4095,
"step": 9
},
{
"epoch": 0.05115089514066496,
"grad_norm": 1.6195282538524967,
"learning_rate": 1.63265306122449e-05,
"loss": 1.3885,
"step": 10
},
{
"epoch": 0.056265984654731455,
"grad_norm": 1.5995091492318427,
"learning_rate": 1.795918367346939e-05,
"loss": 1.3675,
"step": 11
},
{
"epoch": 0.061381074168797956,
"grad_norm": 1.1810770323784914,
"learning_rate": 1.9591836734693877e-05,
"loss": 1.3476,
"step": 12
},
{
"epoch": 0.06649616368286446,
"grad_norm": 1.0714645780169445,
"learning_rate": 2.122448979591837e-05,
"loss": 1.3307,
"step": 13
},
{
"epoch": 0.07161125319693094,
"grad_norm": 0.8530777198954964,
"learning_rate": 2.2857142857142858e-05,
"loss": 1.3243,
"step": 14
},
{
"epoch": 0.07672634271099744,
"grad_norm": 1.0117417462565976,
"learning_rate": 2.448979591836735e-05,
"loss": 1.3178,
"step": 15
},
{
"epoch": 0.08184143222506395,
"grad_norm": 0.8038473161125969,
"learning_rate": 2.6122448979591835e-05,
"loss": 1.3075,
"step": 16
},
{
"epoch": 0.08695652173913043,
"grad_norm": 0.6882196853224636,
"learning_rate": 2.775510204081633e-05,
"loss": 1.3308,
"step": 17
},
{
"epoch": 0.09207161125319693,
"grad_norm": 0.6846199240337438,
"learning_rate": 2.938775510204082e-05,
"loss": 1.2891,
"step": 18
},
{
"epoch": 0.09718670076726342,
"grad_norm": 0.6598120744489369,
"learning_rate": 3.102040816326531e-05,
"loss": 1.2973,
"step": 19
},
{
"epoch": 0.10230179028132992,
"grad_norm": 0.7076360903200484,
"learning_rate": 3.26530612244898e-05,
"loss": 1.2698,
"step": 20
},
{
"epoch": 0.10741687979539642,
"grad_norm": 0.5063556390523856,
"learning_rate": 3.4285714285714284e-05,
"loss": 1.2654,
"step": 21
},
{
"epoch": 0.11253196930946291,
"grad_norm": 0.5637324899711151,
"learning_rate": 3.591836734693878e-05,
"loss": 1.2451,
"step": 22
},
{
"epoch": 0.11764705882352941,
"grad_norm": 0.5080004573049448,
"learning_rate": 3.755102040816327e-05,
"loss": 1.261,
"step": 23
},
{
"epoch": 0.12276214833759591,
"grad_norm": 0.5288023327737856,
"learning_rate": 3.9183673469387755e-05,
"loss": 1.2534,
"step": 24
},
{
"epoch": 0.1278772378516624,
"grad_norm": 0.48497307714732846,
"learning_rate": 4.081632653061225e-05,
"loss": 1.2397,
"step": 25
},
{
"epoch": 0.1329923273657289,
"grad_norm": 0.44772110663725717,
"learning_rate": 4.244897959183674e-05,
"loss": 1.2225,
"step": 26
},
{
"epoch": 0.13810741687979539,
"grad_norm": 0.46163915280113776,
"learning_rate": 4.4081632653061226e-05,
"loss": 1.2293,
"step": 27
},
{
"epoch": 0.1432225063938619,
"grad_norm": 0.4382924852998998,
"learning_rate": 4.5714285714285716e-05,
"loss": 1.2546,
"step": 28
},
{
"epoch": 0.1483375959079284,
"grad_norm": 0.520541317113405,
"learning_rate": 4.7346938775510206e-05,
"loss": 1.2408,
"step": 29
},
{
"epoch": 0.1534526854219949,
"grad_norm": 0.5869089470961774,
"learning_rate": 4.89795918367347e-05,
"loss": 1.2252,
"step": 30
},
{
"epoch": 0.1585677749360614,
"grad_norm": 0.6789867896183338,
"learning_rate": 5.0612244897959194e-05,
"loss": 1.2398,
"step": 31
},
{
"epoch": 0.1636828644501279,
"grad_norm": 0.7434952163265829,
"learning_rate": 5.224489795918367e-05,
"loss": 1.2355,
"step": 32
},
{
"epoch": 0.16879795396419436,
"grad_norm": 0.7377379054061958,
"learning_rate": 5.387755102040817e-05,
"loss": 1.2082,
"step": 33
},
{
"epoch": 0.17391304347826086,
"grad_norm": 0.9393308891221348,
"learning_rate": 5.551020408163266e-05,
"loss": 1.199,
"step": 34
},
{
"epoch": 0.17902813299232737,
"grad_norm": 1.0553898126680656,
"learning_rate": 5.714285714285715e-05,
"loss": 1.1774,
"step": 35
},
{
"epoch": 0.18414322250639387,
"grad_norm": 0.9788330468755307,
"learning_rate": 5.877551020408164e-05,
"loss": 1.1986,
"step": 36
},
{
"epoch": 0.18925831202046037,
"grad_norm": 1.069335926561912,
"learning_rate": 6.040816326530612e-05,
"loss": 1.2126,
"step": 37
},
{
"epoch": 0.19437340153452684,
"grad_norm": 1.2716494354568728,
"learning_rate": 6.204081632653062e-05,
"loss": 1.1866,
"step": 38
},
{
"epoch": 0.19948849104859334,
"grad_norm": 0.8844279720716156,
"learning_rate": 6.36734693877551e-05,
"loss": 1.189,
"step": 39
},
{
"epoch": 0.20460358056265984,
"grad_norm": 0.9551479806230622,
"learning_rate": 6.53061224489796e-05,
"loss": 1.1804,
"step": 40
},
{
"epoch": 0.20971867007672634,
"grad_norm": 1.327902580318565,
"learning_rate": 6.693877551020408e-05,
"loss": 1.1794,
"step": 41
},
{
"epoch": 0.21483375959079284,
"grad_norm": 0.6677700681816165,
"learning_rate": 6.857142857142857e-05,
"loss": 1.1811,
"step": 42
},
{
"epoch": 0.21994884910485935,
"grad_norm": 0.7447541171921808,
"learning_rate": 7.020408163265306e-05,
"loss": 1.205,
"step": 43
},
{
"epoch": 0.22506393861892582,
"grad_norm": 1.0810027284299573,
"learning_rate": 7.183673469387756e-05,
"loss": 1.205,
"step": 44
},
{
"epoch": 0.23017902813299232,
"grad_norm": 0.8936813653252775,
"learning_rate": 7.346938775510205e-05,
"loss": 1.1865,
"step": 45
},
{
"epoch": 0.23529411764705882,
"grad_norm": 1.0905660119141343,
"learning_rate": 7.510204081632654e-05,
"loss": 1.1984,
"step": 46
},
{
"epoch": 0.24040920716112532,
"grad_norm": 0.9760034804167205,
"learning_rate": 7.673469387755103e-05,
"loss": 1.1946,
"step": 47
},
{
"epoch": 0.24552429667519182,
"grad_norm": 0.8996610939752795,
"learning_rate": 7.836734693877551e-05,
"loss": 1.1719,
"step": 48
},
{
"epoch": 0.2506393861892583,
"grad_norm": 1.0356496228760523,
"learning_rate": 8e-05,
"loss": 1.1758,
"step": 49
},
{
"epoch": 0.2557544757033248,
"grad_norm": 1.4230939176945363,
"learning_rate": 8.16326530612245e-05,
"loss": 1.1572,
"step": 50
},
{
"epoch": 0.2608695652173913,
"grad_norm": 0.992854215176634,
"learning_rate": 8.326530612244899e-05,
"loss": 1.1754,
"step": 51
},
{
"epoch": 0.2659846547314578,
"grad_norm": 1.4744220816490745,
"learning_rate": 8.489795918367348e-05,
"loss": 1.1505,
"step": 52
},
{
"epoch": 0.2710997442455243,
"grad_norm": 0.8219750016295352,
"learning_rate": 8.653061224489797e-05,
"loss": 1.1431,
"step": 53
},
{
"epoch": 0.27621483375959077,
"grad_norm": 1.7961949310610006,
"learning_rate": 8.816326530612245e-05,
"loss": 1.1621,
"step": 54
},
{
"epoch": 0.2813299232736573,
"grad_norm": 1.1875180294100318,
"learning_rate": 8.979591836734694e-05,
"loss": 1.172,
"step": 55
},
{
"epoch": 0.2864450127877238,
"grad_norm": 1.5278952491212943,
"learning_rate": 9.142857142857143e-05,
"loss": 1.1706,
"step": 56
},
{
"epoch": 0.2915601023017903,
"grad_norm": 1.275696316707438,
"learning_rate": 9.306122448979593e-05,
"loss": 1.1501,
"step": 57
},
{
"epoch": 0.2966751918158568,
"grad_norm": 1.5558971751084085,
"learning_rate": 9.469387755102041e-05,
"loss": 1.1512,
"step": 58
},
{
"epoch": 0.30179028132992325,
"grad_norm": 1.3680230693346969,
"learning_rate": 9.632653061224491e-05,
"loss": 1.1741,
"step": 59
},
{
"epoch": 0.3069053708439898,
"grad_norm": 1.1458227701427397,
"learning_rate": 9.79591836734694e-05,
"loss": 1.1623,
"step": 60
},
{
"epoch": 0.31202046035805625,
"grad_norm": 1.5291418152321128,
"learning_rate": 9.959183673469389e-05,
"loss": 1.1604,
"step": 61
},
{
"epoch": 0.3171355498721228,
"grad_norm": 1.0015727785086583,
"learning_rate": 0.00010122448979591839,
"loss": 1.1521,
"step": 62
},
{
"epoch": 0.32225063938618925,
"grad_norm": 1.7893801666823759,
"learning_rate": 0.00010285714285714287,
"loss": 1.1677,
"step": 63
},
{
"epoch": 0.3273657289002558,
"grad_norm": 1.3172921889393767,
"learning_rate": 0.00010448979591836734,
"loss": 1.1406,
"step": 64
},
{
"epoch": 0.33248081841432225,
"grad_norm": 1.3975261522115878,
"learning_rate": 0.00010612244897959184,
"loss": 1.1575,
"step": 65
},
{
"epoch": 0.3375959079283887,
"grad_norm": 1.6117928405600537,
"learning_rate": 0.00010775510204081634,
"loss": 1.1735,
"step": 66
},
{
"epoch": 0.34271099744245526,
"grad_norm": 1.1265250794696464,
"learning_rate": 0.00010938775510204082,
"loss": 1.1514,
"step": 67
},
{
"epoch": 0.34782608695652173,
"grad_norm": 1.9297479682392855,
"learning_rate": 0.00011102040816326532,
"loss": 1.1772,
"step": 68
},
{
"epoch": 0.35294117647058826,
"grad_norm": 1.5930978557143378,
"learning_rate": 0.0001126530612244898,
"loss": 1.1645,
"step": 69
},
{
"epoch": 0.35805626598465473,
"grad_norm": 0.9806061428167319,
"learning_rate": 0.0001142857142857143,
"loss": 1.1561,
"step": 70
},
{
"epoch": 0.3631713554987212,
"grad_norm": 1.2884019109710794,
"learning_rate": 0.0001159183673469388,
"loss": 1.1349,
"step": 71
},
{
"epoch": 0.36828644501278773,
"grad_norm": 1.5758649002416978,
"learning_rate": 0.00011755102040816328,
"loss": 1.1707,
"step": 72
},
{
"epoch": 0.3734015345268542,
"grad_norm": 0.975871929986924,
"learning_rate": 0.00011918367346938777,
"loss": 1.1524,
"step": 73
},
{
"epoch": 0.37851662404092073,
"grad_norm": 1.892743008362368,
"learning_rate": 0.00012081632653061224,
"loss": 1.1752,
"step": 74
},
{
"epoch": 0.3836317135549872,
"grad_norm": 1.4249669380451713,
"learning_rate": 0.00012244897959183673,
"loss": 1.1404,
"step": 75
},
{
"epoch": 0.3887468030690537,
"grad_norm": 1.3164919802262345,
"learning_rate": 0.00012408163265306124,
"loss": 1.1529,
"step": 76
},
{
"epoch": 0.3938618925831202,
"grad_norm": 0.9860255927097776,
"learning_rate": 0.00012571428571428572,
"loss": 1.1414,
"step": 77
},
{
"epoch": 0.3989769820971867,
"grad_norm": 1.9392772292389309,
"learning_rate": 0.0001273469387755102,
"loss": 1.1743,
"step": 78
},
{
"epoch": 0.4040920716112532,
"grad_norm": 1.0108994479733273,
"learning_rate": 0.00012897959183673472,
"loss": 1.1791,
"step": 79
},
{
"epoch": 0.4092071611253197,
"grad_norm": 1.6054738611871082,
"learning_rate": 0.0001306122448979592,
"loss": 1.1489,
"step": 80
},
{
"epoch": 0.4143222506393862,
"grad_norm": 2.055637102534096,
"learning_rate": 0.00013224489795918368,
"loss": 1.154,
"step": 81
},
{
"epoch": 0.4194373401534527,
"grad_norm": 1.0696040384071863,
"learning_rate": 0.00013387755102040817,
"loss": 1.1548,
"step": 82
},
{
"epoch": 0.42455242966751916,
"grad_norm": 2.067267813751621,
"learning_rate": 0.00013551020408163268,
"loss": 1.1565,
"step": 83
},
{
"epoch": 0.4296675191815857,
"grad_norm": 1.3481917406218638,
"learning_rate": 0.00013714285714285713,
"loss": 1.1689,
"step": 84
},
{
"epoch": 0.43478260869565216,
"grad_norm": 1.9430429763045844,
"learning_rate": 0.00013877551020408165,
"loss": 1.1634,
"step": 85
},
{
"epoch": 0.4398976982097187,
"grad_norm": 1.1269046955736524,
"learning_rate": 0.00014040816326530613,
"loss": 1.1608,
"step": 86
},
{
"epoch": 0.44501278772378516,
"grad_norm": 2.4402015202438188,
"learning_rate": 0.0001420408163265306,
"loss": 1.1912,
"step": 87
},
{
"epoch": 0.45012787723785164,
"grad_norm": 1.7219408702117005,
"learning_rate": 0.00014367346938775512,
"loss": 1.1991,
"step": 88
},
{
"epoch": 0.45524296675191817,
"grad_norm": 1.5354645790141401,
"learning_rate": 0.0001453061224489796,
"loss": 1.1619,
"step": 89
},
{
"epoch": 0.46035805626598464,
"grad_norm": 1.6038175546329443,
"learning_rate": 0.0001469387755102041,
"loss": 1.1569,
"step": 90
},
{
"epoch": 0.46547314578005117,
"grad_norm": 1.6538580895352009,
"learning_rate": 0.0001485714285714286,
"loss": 1.1703,
"step": 91
},
{
"epoch": 0.47058823529411764,
"grad_norm": 1.3451626506992602,
"learning_rate": 0.00015020408163265308,
"loss": 1.1641,
"step": 92
},
{
"epoch": 0.47570332480818417,
"grad_norm": 1.232809474186634,
"learning_rate": 0.00015183673469387757,
"loss": 1.1635,
"step": 93
},
{
"epoch": 0.48081841432225064,
"grad_norm": 0.7505548404279615,
"learning_rate": 0.00015346938775510205,
"loss": 1.1504,
"step": 94
},
{
"epoch": 0.4859335038363171,
"grad_norm": 1.2314120568021114,
"learning_rate": 0.00015510204081632654,
"loss": 1.1495,
"step": 95
},
{
"epoch": 0.49104859335038364,
"grad_norm": 1.0894786985964773,
"learning_rate": 0.00015673469387755102,
"loss": 1.1225,
"step": 96
},
{
"epoch": 0.4961636828644501,
"grad_norm": 1.825853195018056,
"learning_rate": 0.00015836734693877553,
"loss": 1.1309,
"step": 97
},
{
"epoch": 0.5012787723785166,
"grad_norm": 1.1910196454056825,
"learning_rate": 0.00016,
"loss": 1.1688,
"step": 98
},
{
"epoch": 0.5063938618925832,
"grad_norm": 1.5831221112316092,
"learning_rate": 0.0001599994867132121,
"loss": 1.1564,
"step": 99
},
{
"epoch": 0.5115089514066496,
"grad_norm": 1.3473080386538443,
"learning_rate": 0.00015999794685943504,
"loss": 1.1384,
"step": 100
},
{
"epoch": 0.5166240409207161,
"grad_norm": 1.2977978405443058,
"learning_rate": 0.00015999538045842842,
"loss": 1.1233,
"step": 101
},
{
"epoch": 0.5217391304347826,
"grad_norm": 1.077595330517682,
"learning_rate": 0.0001599917875431248,
"loss": 1.1357,
"step": 102
},
{
"epoch": 0.5268542199488491,
"grad_norm": 2.195202839950938,
"learning_rate": 0.00015998716815962901,
"loss": 1.1447,
"step": 103
},
{
"epoch": 0.5319693094629157,
"grad_norm": 1.2189974449985095,
"learning_rate": 0.0001599815223672178,
"loss": 1.1491,
"step": 104
},
{
"epoch": 0.5370843989769821,
"grad_norm": 1.8420232254576474,
"learning_rate": 0.00015997485023833895,
"loss": 1.1306,
"step": 105
},
{
"epoch": 0.5421994884910486,
"grad_norm": 1.580219768099248,
"learning_rate": 0.0001599671518586103,
"loss": 1.1314,
"step": 106
},
{
"epoch": 0.5473145780051151,
"grad_norm": 1.2283391561883323,
"learning_rate": 0.00015995842732681882,
"loss": 1.1245,
"step": 107
},
{
"epoch": 0.5524296675191815,
"grad_norm": 1.6969416482186022,
"learning_rate": 0.00015994867675491915,
"loss": 1.1298,
"step": 108
},
{
"epoch": 0.5575447570332481,
"grad_norm": 0.9258633302440661,
"learning_rate": 0.0001599379002680323,
"loss": 1.1017,
"step": 109
},
{
"epoch": 0.5626598465473146,
"grad_norm": 1.2218775803568054,
"learning_rate": 0.00015992609800444396,
"loss": 1.1393,
"step": 110
},
{
"epoch": 0.5677749360613811,
"grad_norm": 1.7894279716657744,
"learning_rate": 0.00015991327011560282,
"loss": 1.1151,
"step": 111
},
{
"epoch": 0.5728900255754475,
"grad_norm": 1.1093034545807021,
"learning_rate": 0.00015989941676611846,
"loss": 1.128,
"step": 112
},
{
"epoch": 0.578005115089514,
"grad_norm": 1.1968447600020453,
"learning_rate": 0.00015988453813375948,
"loss": 1.1267,
"step": 113
},
{
"epoch": 0.5831202046035806,
"grad_norm": 1.3309032644456822,
"learning_rate": 0.000159868634409451,
"loss": 1.1174,
"step": 114
},
{
"epoch": 0.5882352941176471,
"grad_norm": 1.1303171808005343,
"learning_rate": 0.00015985170579727227,
"loss": 1.1299,
"step": 115
},
{
"epoch": 0.5933503836317136,
"grad_norm": 2.0767735751699545,
"learning_rate": 0.0001598337525144541,
"loss": 1.1258,
"step": 116
},
{
"epoch": 0.59846547314578,
"grad_norm": 0.9546323340961507,
"learning_rate": 0.00015981477479137614,
"loss": 1.1,
"step": 117
},
{
"epoch": 0.6035805626598465,
"grad_norm": 2.5649767556070646,
"learning_rate": 0.00015979477287156373,
"loss": 1.1284,
"step": 118
},
{
"epoch": 0.6086956521739131,
"grad_norm": 1.856177606586923,
"learning_rate": 0.00015977374701168484,
"loss": 1.1253,
"step": 119
},
{
"epoch": 0.6138107416879796,
"grad_norm": 1.8600897833608647,
"learning_rate": 0.00015975169748154693,
"loss": 1.1296,
"step": 120
},
{
"epoch": 0.618925831202046,
"grad_norm": 1.675219885511045,
"learning_rate": 0.0001597286245640933,
"loss": 1.1253,
"step": 121
},
{
"epoch": 0.6240409207161125,
"grad_norm": 1.0585377009177594,
"learning_rate": 0.00015970452855539955,
"loss": 1.1438,
"step": 122
},
{
"epoch": 0.629156010230179,
"grad_norm": 2.400147256806906,
"learning_rate": 0.0001596794097646697,
"loss": 1.1355,
"step": 123
},
{
"epoch": 0.6342710997442456,
"grad_norm": 1.7217211127133585,
"learning_rate": 0.0001596532685142324,
"loss": 1.1571,
"step": 124
},
{
"epoch": 0.639386189258312,
"grad_norm": 2.0877878034892046,
"learning_rate": 0.00015962610513953662,
"loss": 1.1381,
"step": 125
},
{
"epoch": 0.6445012787723785,
"grad_norm": 1.2533204273176497,
"learning_rate": 0.00015959791998914732,
"loss": 1.1375,
"step": 126
},
{
"epoch": 0.649616368286445,
"grad_norm": 1.7518209160063163,
"learning_rate": 0.0001595687134247412,
"loss": 1.1418,
"step": 127
},
{
"epoch": 0.6547314578005116,
"grad_norm": 0.8858019521385987,
"learning_rate": 0.0001595384858211018,
"loss": 1.1143,
"step": 128
},
{
"epoch": 0.659846547314578,
"grad_norm": 1.36582303177154,
"learning_rate": 0.0001595072375661149,
"loss": 1.1283,
"step": 129
},
{
"epoch": 0.6649616368286445,
"grad_norm": 0.9251142584655125,
"learning_rate": 0.0001594749690607634,
"loss": 1.1251,
"step": 130
},
{
"epoch": 0.670076726342711,
"grad_norm": 1.107627426852622,
"learning_rate": 0.0001594416807191222,
"loss": 1.0951,
"step": 131
},
{
"epoch": 0.6751918158567775,
"grad_norm": 0.9813952780001111,
"learning_rate": 0.00015940737296835303,
"loss": 1.1351,
"step": 132
},
{
"epoch": 0.680306905370844,
"grad_norm": 1.3700787369710026,
"learning_rate": 0.00015937204624869868,
"loss": 1.1387,
"step": 133
},
{
"epoch": 0.6854219948849105,
"grad_norm": 1.2904194439899275,
"learning_rate": 0.00015933570101347766,
"loss": 1.1005,
"step": 134
},
{
"epoch": 0.690537084398977,
"grad_norm": 1.1820536199055438,
"learning_rate": 0.0001592983377290782,
"loss": 1.114,
"step": 135
},
{
"epoch": 0.6956521739130435,
"grad_norm": 0.9843473721377823,
"learning_rate": 0.0001592599568749523,
"loss": 1.1115,
"step": 136
},
{
"epoch": 0.7007672634271099,
"grad_norm": 1.1033493336912457,
"learning_rate": 0.00015922055894360956,
"loss": 1.1141,
"step": 137
},
{
"epoch": 0.7058823529411765,
"grad_norm": 1.0951574459451423,
"learning_rate": 0.00015918014444061098,
"loss": 1.1099,
"step": 138
},
{
"epoch": 0.710997442455243,
"grad_norm": 1.0641706695758877,
"learning_rate": 0.00015913871388456227,
"loss": 1.1172,
"step": 139
},
{
"epoch": 0.7161125319693095,
"grad_norm": 1.4663299034937627,
"learning_rate": 0.00015909626780710736,
"loss": 1.1155,
"step": 140
},
{
"epoch": 0.7212276214833759,
"grad_norm": 1.0207018070657674,
"learning_rate": 0.00015905280675292154,
"loss": 1.106,
"step": 141
},
{
"epoch": 0.7263427109974424,
"grad_norm": 1.1685679972284322,
"learning_rate": 0.00015900833127970444,
"loss": 1.0792,
"step": 142
},
{
"epoch": 0.731457800511509,
"grad_norm": 1.123687352602435,
"learning_rate": 0.00015896284195817287,
"loss": 1.106,
"step": 143
},
{
"epoch": 0.7365728900255755,
"grad_norm": 1.21454249173366,
"learning_rate": 0.0001589163393720535,
"loss": 1.0982,
"step": 144
},
{
"epoch": 0.7416879795396419,
"grad_norm": 0.9984630775235075,
"learning_rate": 0.00015886882411807543,
"loss": 1.1139,
"step": 145
},
{
"epoch": 0.7468030690537084,
"grad_norm": 1.2001403406405777,
"learning_rate": 0.00015882029680596246,
"loss": 1.0869,
"step": 146
},
{
"epoch": 0.7519181585677749,
"grad_norm": 1.0944141408896968,
"learning_rate": 0.0001587707580584253,
"loss": 1.0939,
"step": 147
},
{
"epoch": 0.7570332480818415,
"grad_norm": 1.0322962653055237,
"learning_rate": 0.00015872020851115355,
"loss": 1.0845,
"step": 148
},
{
"epoch": 0.7621483375959079,
"grad_norm": 1.3470719664767947,
"learning_rate": 0.0001586686488128076,
"loss": 1.121,
"step": 149
},
{
"epoch": 0.7672634271099744,
"grad_norm": 0.9068380122105875,
"learning_rate": 0.0001586160796250102,
"loss": 1.0993,
"step": 150
},
{
"epoch": 0.7723785166240409,
"grad_norm": 1.1687258782812369,
"learning_rate": 0.00015856250162233815,
"loss": 1.115,
"step": 151
},
{
"epoch": 0.7774936061381074,
"grad_norm": 1.340150021571059,
"learning_rate": 0.00015850791549231345,
"loss": 1.0889,
"step": 152
},
{
"epoch": 0.782608695652174,
"grad_norm": 0.9634527414082537,
"learning_rate": 0.00015845232193539453,
"loss": 1.0967,
"step": 153
},
{
"epoch": 0.7877237851662404,
"grad_norm": 1.3305334014339185,
"learning_rate": 0.00015839572166496743,
"loss": 1.0818,
"step": 154
},
{
"epoch": 0.7928388746803069,
"grad_norm": 0.8142121234465892,
"learning_rate": 0.00015833811540733638,
"loss": 1.094,
"step": 155
},
{
"epoch": 0.7979539641943734,
"grad_norm": 1.080425926126085,
"learning_rate": 0.00015827950390171467,
"loss": 1.1184,
"step": 156
},
{
"epoch": 0.80306905370844,
"grad_norm": 1.2956236829957526,
"learning_rate": 0.00015821988790021507,
"loss": 1.0953,
"step": 157
},
{
"epoch": 0.8081841432225064,
"grad_norm": 1.0068070725012628,
"learning_rate": 0.00015815926816784028,
"loss": 1.0822,
"step": 158
},
{
"epoch": 0.8132992327365729,
"grad_norm": 1.443782781796012,
"learning_rate": 0.00015809764548247292,
"loss": 1.1113,
"step": 159
},
{
"epoch": 0.8184143222506394,
"grad_norm": 0.8514747602771416,
"learning_rate": 0.00015803502063486578,
"loss": 1.0763,
"step": 160
},
{
"epoch": 0.8235294117647058,
"grad_norm": 0.9848333369104281,
"learning_rate": 0.0001579713944286315,
"loss": 1.0837,
"step": 161
},
{
"epoch": 0.8286445012787724,
"grad_norm": 1.1641315760933826,
"learning_rate": 0.0001579067676802324,
"loss": 1.1206,
"step": 162
},
{
"epoch": 0.8337595907928389,
"grad_norm": 1.4106475740967381,
"learning_rate": 0.00015784114121896988,
"loss": 1.0873,
"step": 163
},
{
"epoch": 0.8388746803069054,
"grad_norm": 0.7778670199808698,
"learning_rate": 0.0001577745158869738,
"loss": 1.1119,
"step": 164
},
{
"epoch": 0.8439897698209718,
"grad_norm": 0.8492411927300004,
"learning_rate": 0.00015770689253919175,
"loss": 1.1257,
"step": 165
},
{
"epoch": 0.8491048593350383,
"grad_norm": 1.2080069675635858,
"learning_rate": 0.00015763827204337802,
"loss": 1.082,
"step": 166
},
{
"epoch": 0.8542199488491049,
"grad_norm": 1.0708198811222547,
"learning_rate": 0.00015756865528008239,
"loss": 1.0925,
"step": 167
},
{
"epoch": 0.8593350383631714,
"grad_norm": 1.1262913488379183,
"learning_rate": 0.000157498043142639,
"loss": 1.0821,
"step": 168
},
{
"epoch": 0.8644501278772379,
"grad_norm": 1.4945513039406828,
"learning_rate": 0.00015742643653715485,
"loss": 1.1073,
"step": 169
},
{
"epoch": 0.8695652173913043,
"grad_norm": 0.7205364315435271,
"learning_rate": 0.000157353836382498,
"loss": 1.0929,
"step": 170
},
{
"epoch": 0.8746803069053708,
"grad_norm": 1.2951228606069005,
"learning_rate": 0.00015728024361028595,
"loss": 1.089,
"step": 171
},
{
"epoch": 0.8797953964194374,
"grad_norm": 1.012999905443166,
"learning_rate": 0.00015720565916487364,
"loss": 1.0984,
"step": 172
},
{
"epoch": 0.8849104859335039,
"grad_norm": 0.9930104635743521,
"learning_rate": 0.00015713008400334137,
"loss": 1.117,
"step": 173
},
{
"epoch": 0.8900255754475703,
"grad_norm": 1.066494063761974,
"learning_rate": 0.00015705351909548238,
"loss": 1.0696,
"step": 174
},
{
"epoch": 0.8951406649616368,
"grad_norm": 0.9029376592809948,
"learning_rate": 0.00015697596542379061,
"loss": 1.0928,
"step": 175
},
{
"epoch": 0.9002557544757033,
"grad_norm": 0.9912055001368116,
"learning_rate": 0.0001568974239834479,
"loss": 1.1012,
"step": 176
},
{
"epoch": 0.9053708439897699,
"grad_norm": 1.3382093260329788,
"learning_rate": 0.00015681789578231138,
"loss": 1.0989,
"step": 177
},
{
"epoch": 0.9104859335038363,
"grad_norm": 0.9107622512976783,
"learning_rate": 0.00015673738184090034,
"loss": 1.0745,
"step": 178
},
{
"epoch": 0.9156010230179028,
"grad_norm": 1.4524863320621164,
"learning_rate": 0.00015665588319238347,
"loss": 1.1088,
"step": 179
},
{
"epoch": 0.9207161125319693,
"grad_norm": 0.8764168164711392,
"learning_rate": 0.00015657340088256515,
"loss": 1.0827,
"step": 180
},
{
"epoch": 0.9258312020460358,
"grad_norm": 0.9593426709669773,
"learning_rate": 0.0001564899359698724,
"loss": 1.079,
"step": 181
},
{
"epoch": 0.9309462915601023,
"grad_norm": 1.0512399302337887,
"learning_rate": 0.0001564054895253412,
"loss": 1.0938,
"step": 182
},
{
"epoch": 0.9360613810741688,
"grad_norm": 1.5129324295290898,
"learning_rate": 0.00015632006263260258,
"loss": 1.1158,
"step": 183
},
{
"epoch": 0.9411764705882353,
"grad_norm": 0.9429667334268746,
"learning_rate": 0.00015623365638786895,
"loss": 1.0977,
"step": 184
},
{
"epoch": 0.9462915601023018,
"grad_norm": 1.216087618410104,
"learning_rate": 0.00015614627189991992,
"loss": 1.0892,
"step": 185
},
{
"epoch": 0.9514066496163683,
"grad_norm": 0.8077933529981967,
"learning_rate": 0.00015605791029008807,
"loss": 1.1031,
"step": 186
},
{
"epoch": 0.9565217391304348,
"grad_norm": 1.070253998934908,
"learning_rate": 0.00015596857269224455,
"loss": 1.0988,
"step": 187
},
{
"epoch": 0.9616368286445013,
"grad_norm": 0.6240214934142639,
"learning_rate": 0.0001558782602527846,
"loss": 1.0927,
"step": 188
},
{
"epoch": 0.9667519181585678,
"grad_norm": 0.9708959205899346,
"learning_rate": 0.00015578697413061275,
"loss": 1.1022,
"step": 189
},
{
"epoch": 0.9718670076726342,
"grad_norm": 1.150326324835953,
"learning_rate": 0.00015569471549712802,
"loss": 1.0661,
"step": 190
},
{
"epoch": 0.9769820971867008,
"grad_norm": 1.6180812335566688,
"learning_rate": 0.00015560148553620884,
"loss": 1.0898,
"step": 191
},
{
"epoch": 0.9820971867007673,
"grad_norm": 0.8131703242642281,
"learning_rate": 0.00015550728544419791,
"loss": 1.0732,
"step": 192
},
{
"epoch": 0.9872122762148338,
"grad_norm": 0.8943025986368547,
"learning_rate": 0.0001554121164298868,
"loss": 1.1064,
"step": 193
},
{
"epoch": 0.9923273657289002,
"grad_norm": 1.546075493707172,
"learning_rate": 0.0001553159797145004,
"loss": 1.1077,
"step": 194
},
{
"epoch": 0.9974424552429667,
"grad_norm": 1.0147375848592517,
"learning_rate": 0.0001552188765316814,
"loss": 1.0828,
"step": 195
},
{
"epoch": 1.0025575447570332,
"grad_norm": 1.808490849357134,
"learning_rate": 0.00015512080812747433,
"loss": 1.6517,
"step": 196
},
{
"epoch": 1.0076726342710998,
"grad_norm": 0.8186212963172015,
"learning_rate": 0.00015502177576030957,
"loss": 1.0702,
"step": 197
},
{
"epoch": 1.0127877237851663,
"grad_norm": 0.8836359955945491,
"learning_rate": 0.00015492178070098726,
"loss": 1.0641,
"step": 198
},
{
"epoch": 1.0179028132992327,
"grad_norm": 1.2077182953494592,
"learning_rate": 0.00015482082423266098,
"loss": 1.0913,
"step": 199
},
{
"epoch": 1.0230179028132993,
"grad_norm": 1.2857282920795001,
"learning_rate": 0.00015471890765082126,
"loss": 1.079,
"step": 200
},
{
"epoch": 1.0281329923273657,
"grad_norm": 0.7871257715894987,
"learning_rate": 0.000154616032263279,
"loss": 1.0891,
"step": 201
},
{
"epoch": 1.0332480818414322,
"grad_norm": 0.7876971959933657,
"learning_rate": 0.00015451219939014856,
"loss": 1.0594,
"step": 202
},
{
"epoch": 1.0383631713554988,
"grad_norm": 1.1296142868181394,
"learning_rate": 0.00015440741036383107,
"loss": 1.0984,
"step": 203
},
{
"epoch": 1.0434782608695652,
"grad_norm": 1.3346232997413918,
"learning_rate": 0.00015430166652899703,
"loss": 1.0498,
"step": 204
},
{
"epoch": 1.0485933503836318,
"grad_norm": 0.93489371766657,
"learning_rate": 0.00015419496924256934,
"loss": 1.0405,
"step": 205
},
{
"epoch": 1.0537084398976981,
"grad_norm": 0.9075968242325666,
"learning_rate": 0.00015408731987370564,
"loss": 1.0445,
"step": 206
},
{
"epoch": 1.0588235294117647,
"grad_norm": 0.918872998744737,
"learning_rate": 0.00015397871980378088,
"loss": 1.0926,
"step": 207
},
{
"epoch": 1.0639386189258313,
"grad_norm": 1.117540850957588,
"learning_rate": 0.00015386917042636963,
"loss": 1.0647,
"step": 208
},
{
"epoch": 1.0690537084398977,
"grad_norm": 1.4616682738446867,
"learning_rate": 0.00015375867314722808,
"loss": 1.0432,
"step": 209
},
{
"epoch": 1.0741687979539642,
"grad_norm": 0.6968242592336843,
"learning_rate": 0.0001536472293842761,
"loss": 1.0834,
"step": 210
},
{
"epoch": 1.0792838874680306,
"grad_norm": 0.6577127620307978,
"learning_rate": 0.00015353484056757888,
"loss": 1.0566,
"step": 211
},
{
"epoch": 1.0843989769820972,
"grad_norm": 0.8971320106822009,
"learning_rate": 0.00015342150813932884,
"loss": 1.0752,
"step": 212
},
{
"epoch": 1.0895140664961638,
"grad_norm": 1.1517030782858764,
"learning_rate": 0.00015330723355382695,
"loss": 1.0465,
"step": 213
},
{
"epoch": 1.0946291560102301,
"grad_norm": 1.1841809608739537,
"learning_rate": 0.00015319201827746406,
"loss": 1.053,
"step": 214
},
{
"epoch": 1.0997442455242967,
"grad_norm": 0.7140648922452231,
"learning_rate": 0.00015307586378870218,
"loss": 1.0834,
"step": 215
},
{
"epoch": 1.104859335038363,
"grad_norm": 0.9502066657036609,
"learning_rate": 0.00015295877157805535,
"loss": 1.0932,
"step": 216
},
{
"epoch": 1.1099744245524297,
"grad_norm": 1.246817849636434,
"learning_rate": 0.00015284074314807076,
"loss": 1.0693,
"step": 217
},
{
"epoch": 1.1150895140664963,
"grad_norm": 0.6993711715541463,
"learning_rate": 0.00015272178001330918,
"loss": 1.0572,
"step": 218
},
{
"epoch": 1.1202046035805626,
"grad_norm": 1.032505106174948,
"learning_rate": 0.00015260188370032584,
"loss": 1.0402,
"step": 219
},
{
"epoch": 1.1253196930946292,
"grad_norm": 0.955807437091477,
"learning_rate": 0.0001524810557476505,
"loss": 1.0676,
"step": 220
},
{
"epoch": 1.1304347826086956,
"grad_norm": 1.2543689592049212,
"learning_rate": 0.00015235929770576797,
"loss": 1.075,
"step": 221
},
{
"epoch": 1.1355498721227621,
"grad_norm": 0.863162669006661,
"learning_rate": 0.0001522366111370981,
"loss": 1.0457,
"step": 222
},
{
"epoch": 1.1406649616368287,
"grad_norm": 1.0560829950876516,
"learning_rate": 0.0001521129976159758,
"loss": 1.044,
"step": 223
},
{
"epoch": 1.145780051150895,
"grad_norm": 1.155011768272074,
"learning_rate": 0.0001519884587286307,
"loss": 1.0467,
"step": 224
},
{
"epoch": 1.1508951406649617,
"grad_norm": 1.050792646745025,
"learning_rate": 0.00015186299607316693,
"loss": 1.0577,
"step": 225
},
{
"epoch": 1.156010230179028,
"grad_norm": 1.1966316793362106,
"learning_rate": 0.00015173661125954263,
"loss": 1.0727,
"step": 226
},
{
"epoch": 1.1611253196930946,
"grad_norm": 0.8651200013585791,
"learning_rate": 0.0001516093059095491,
"loss": 1.089,
"step": 227
},
{
"epoch": 1.1662404092071612,
"grad_norm": 1.124515137471986,
"learning_rate": 0.00015148108165679027,
"loss": 1.034,
"step": 228
},
{
"epoch": 1.1713554987212276,
"grad_norm": 1.1273200095818452,
"learning_rate": 0.00015135194014666148,
"loss": 1.0455,
"step": 229
},
{
"epoch": 1.1764705882352942,
"grad_norm": 0.899811266105129,
"learning_rate": 0.00015122188303632851,
"loss": 1.0556,
"step": 230
},
{
"epoch": 1.1815856777493605,
"grad_norm": 1.3057611391435584,
"learning_rate": 0.00015109091199470628,
"loss": 1.0862,
"step": 231
},
{
"epoch": 1.186700767263427,
"grad_norm": 1.0223348369617988,
"learning_rate": 0.00015095902870243736,
"loss": 1.0478,
"step": 232
},
{
"epoch": 1.1918158567774937,
"grad_norm": 1.3035543495875737,
"learning_rate": 0.00015082623485187063,
"loss": 1.0646,
"step": 233
},
{
"epoch": 1.19693094629156,
"grad_norm": 0.5147671113021504,
"learning_rate": 0.0001506925321470392,
"loss": 1.0742,
"step": 234
},
{
"epoch": 1.2020460358056266,
"grad_norm": 1.0474593724211334,
"learning_rate": 0.000150557922303639,
"loss": 1.0692,
"step": 235
},
{
"epoch": 1.207161125319693,
"grad_norm": 1.2273985339691533,
"learning_rate": 0.00015042240704900628,
"loss": 1.0323,
"step": 236
},
{
"epoch": 1.2122762148337596,
"grad_norm": 0.9904643189727587,
"learning_rate": 0.0001502859881220958,
"loss": 1.0633,
"step": 237
},
{
"epoch": 1.2173913043478262,
"grad_norm": 1.0126411499371004,
"learning_rate": 0.00015014866727345845,
"loss": 1.036,
"step": 238
},
{
"epoch": 1.2225063938618925,
"grad_norm": 1.478785286579918,
"learning_rate": 0.0001500104462652186,
"loss": 1.0957,
"step": 239
},
{
"epoch": 1.227621483375959,
"grad_norm": 0.7483936584773456,
"learning_rate": 0.0001498713268710517,
"loss": 1.0495,
"step": 240
},
{
"epoch": 1.2327365728900257,
"grad_norm": 0.7541085660661818,
"learning_rate": 0.0001497313108761614,
"loss": 1.0475,
"step": 241
},
{
"epoch": 1.237851662404092,
"grad_norm": 1.5678381398679577,
"learning_rate": 0.00014959040007725676,
"loss": 1.0709,
"step": 242
},
{
"epoch": 1.2429667519181586,
"grad_norm": 0.6371598577267172,
"learning_rate": 0.00014944859628252902,
"loss": 1.0567,
"step": 243
},
{
"epoch": 1.248081841432225,
"grad_norm": 1.319879889605403,
"learning_rate": 0.00014930590131162852,
"loss": 1.047,
"step": 244
},
{
"epoch": 1.2531969309462916,
"grad_norm": 0.8776665619364019,
"learning_rate": 0.0001491623169956414,
"loss": 1.0872,
"step": 245
},
{
"epoch": 1.258312020460358,
"grad_norm": 0.9790185512929787,
"learning_rate": 0.00014901784517706593,
"loss": 1.0652,
"step": 246
},
{
"epoch": 1.2634271099744245,
"grad_norm": 1.3594792853072082,
"learning_rate": 0.000148872487709789,
"loss": 1.0502,
"step": 247
},
{
"epoch": 1.2685421994884911,
"grad_norm": 0.7969835384302694,
"learning_rate": 0.00014872624645906232,
"loss": 1.0616,
"step": 248
},
{
"epoch": 1.2736572890025575,
"grad_norm": 0.727091253801579,
"learning_rate": 0.00014857912330147846,
"loss": 1.0387,
"step": 249
},
{
"epoch": 1.278772378516624,
"grad_norm": 1.0041139209180052,
"learning_rate": 0.00014843112012494667,
"loss": 1.0501,
"step": 250
},
{
"epoch": 1.2838874680306906,
"grad_norm": 1.6348417676022182,
"learning_rate": 0.00014828223882866888,
"loss": 1.0598,
"step": 251
},
{
"epoch": 1.289002557544757,
"grad_norm": 0.6602961275800052,
"learning_rate": 0.0001481324813231151,
"loss": 1.0662,
"step": 252
},
{
"epoch": 1.2941176470588236,
"grad_norm": 1.432213174915101,
"learning_rate": 0.00014798184952999915,
"loss": 1.0487,
"step": 253
},
{
"epoch": 1.29923273657289,
"grad_norm": 0.9222257083643778,
"learning_rate": 0.00014783034538225367,
"loss": 1.0669,
"step": 254
},
{
"epoch": 1.3043478260869565,
"grad_norm": 1.3404135459336588,
"learning_rate": 0.00014767797082400563,
"loss": 1.0496,
"step": 255
},
{
"epoch": 1.309462915601023,
"grad_norm": 0.9659792167335154,
"learning_rate": 0.0001475247278105512,
"loss": 1.0723,
"step": 256
},
{
"epoch": 1.3145780051150895,
"grad_norm": 1.0170255045927854,
"learning_rate": 0.00014737061830833074,
"loss": 1.0446,
"step": 257
},
{
"epoch": 1.319693094629156,
"grad_norm": 1.1193991179751086,
"learning_rate": 0.00014721564429490358,
"loss": 1.041,
"step": 258
},
{
"epoch": 1.3248081841432224,
"grad_norm": 1.1403672639024403,
"learning_rate": 0.0001470598077589225,
"loss": 1.0531,
"step": 259
},
{
"epoch": 1.329923273657289,
"grad_norm": 1.1390597701387402,
"learning_rate": 0.00014690311070010838,
"loss": 1.0718,
"step": 260
},
{
"epoch": 1.3350383631713556,
"grad_norm": 0.7892617568723064,
"learning_rate": 0.0001467455551292245,
"loss": 1.0629,
"step": 261
},
{
"epoch": 1.340153452685422,
"grad_norm": 0.7600813249002821,
"learning_rate": 0.00014658714306805068,
"loss": 1.0698,
"step": 262
},
{
"epoch": 1.3452685421994885,
"grad_norm": 0.6627024387905829,
"learning_rate": 0.00014642787654935735,
"loss": 1.0607,
"step": 263
},
{
"epoch": 1.350383631713555,
"grad_norm": 0.6219187835291411,
"learning_rate": 0.0001462677576168795,
"loss": 1.0697,
"step": 264
},
{
"epoch": 1.3554987212276215,
"grad_norm": 0.7860491135288276,
"learning_rate": 0.00014610678832529045,
"loss": 1.0252,
"step": 265
},
{
"epoch": 1.3606138107416879,
"grad_norm": 1.1586096133091477,
"learning_rate": 0.00014594497074017548,
"loss": 1.0488,
"step": 266
},
{
"epoch": 1.3657289002557544,
"grad_norm": 1.0597220126722886,
"learning_rate": 0.00014578230693800532,
"loss": 1.0832,
"step": 267
},
{
"epoch": 1.370843989769821,
"grad_norm": 0.9932062967120353,
"learning_rate": 0.00014561879900610946,
"loss": 1.0695,
"step": 268
},
{
"epoch": 1.3759590792838874,
"grad_norm": 1.22903542245493,
"learning_rate": 0.00014545444904264938,
"loss": 1.0699,
"step": 269
},
{
"epoch": 1.381074168797954,
"grad_norm": 0.8220411539596157,
"learning_rate": 0.00014528925915659177,
"loss": 1.0403,
"step": 270
},
{
"epoch": 1.3861892583120206,
"grad_norm": 0.6890142797754029,
"learning_rate": 0.00014512323146768127,
"loss": 1.0876,
"step": 271
},
{
"epoch": 1.391304347826087,
"grad_norm": 0.6649187014482618,
"learning_rate": 0.0001449563681064133,
"loss": 1.0341,
"step": 272
},
{
"epoch": 1.3964194373401535,
"grad_norm": 0.7764352724568059,
"learning_rate": 0.00014478867121400695,
"loss": 1.032,
"step": 273
},
{
"epoch": 1.40153452685422,
"grad_norm": 1.1369915666314934,
"learning_rate": 0.00014462014294237705,
"loss": 1.061,
"step": 274
},
{
"epoch": 1.4066496163682864,
"grad_norm": 1.1768286724701913,
"learning_rate": 0.0001444507854541071,
"loss": 1.0544,
"step": 275
},
{
"epoch": 1.4117647058823528,
"grad_norm": 0.570432672332846,
"learning_rate": 0.00014428060092242106,
"loss": 1.0563,
"step": 276
},
{
"epoch": 1.4168797953964194,
"grad_norm": 0.8815392140240954,
"learning_rate": 0.00014410959153115578,
"loss": 1.0556,
"step": 277
},
{
"epoch": 1.421994884910486,
"grad_norm": 1.5313592848267774,
"learning_rate": 0.00014393775947473274,
"loss": 1.0709,
"step": 278
},
{
"epoch": 1.4271099744245523,
"grad_norm": 0.521957544153483,
"learning_rate": 0.00014376510695813003,
"loss": 1.0587,
"step": 279
},
{
"epoch": 1.432225063938619,
"grad_norm": 1.346248541102802,
"learning_rate": 0.0001435916361968541,
"loss": 1.0428,
"step": 280
},
{
"epoch": 1.4373401534526855,
"grad_norm": 0.7504790380490416,
"learning_rate": 0.00014341734941691116,
"loss": 1.0378,
"step": 281
},
{
"epoch": 1.4424552429667519,
"grad_norm": 0.8545960490019313,
"learning_rate": 0.00014324224885477877,
"loss": 1.0872,
"step": 282
},
{
"epoch": 1.4475703324808185,
"grad_norm": 0.8147945107029777,
"learning_rate": 0.00014306633675737702,
"loss": 1.0674,
"step": 283
},
{
"epoch": 1.452685421994885,
"grad_norm": 1.1893545689604992,
"learning_rate": 0.00014288961538203983,
"loss": 1.0489,
"step": 284
},
{
"epoch": 1.4578005115089514,
"grad_norm": 1.2942117095522307,
"learning_rate": 0.00014271208699648586,
"loss": 1.0477,
"step": 285
},
{
"epoch": 1.4629156010230178,
"grad_norm": 0.7378283326633418,
"learning_rate": 0.00014253375387878953,
"loss": 1.0555,
"step": 286
},
{
"epoch": 1.4680306905370843,
"grad_norm": 0.7888946066967314,
"learning_rate": 0.00014235461831735156,
"loss": 1.0553,
"step": 287
},
{
"epoch": 1.473145780051151,
"grad_norm": 0.9641958232815284,
"learning_rate": 0.00014217468261086994,
"loss": 1.0715,
"step": 288
},
{
"epoch": 1.4782608695652173,
"grad_norm": 1.0802957905096358,
"learning_rate": 0.00014199394906831022,
"loss": 1.0573,
"step": 289
},
{
"epoch": 1.4833759590792839,
"grad_norm": 0.889710644387899,
"learning_rate": 0.00014181242000887583,
"loss": 1.0416,
"step": 290
},
{
"epoch": 1.4884910485933505,
"grad_norm": 0.9839182472936752,
"learning_rate": 0.0001416300977619785,
"loss": 1.0683,
"step": 291
},
{
"epoch": 1.4936061381074168,
"grad_norm": 1.4570363489326466,
"learning_rate": 0.00014144698466720824,
"loss": 1.0682,
"step": 292
},
{
"epoch": 1.4987212276214834,
"grad_norm": 0.6421264437106724,
"learning_rate": 0.00014126308307430336,
"loss": 1.036,
"step": 293
},
{
"epoch": 1.50383631713555,
"grad_norm": 0.7698446433339242,
"learning_rate": 0.0001410783953431203,
"loss": 1.0543,
"step": 294
},
{
"epoch": 1.5089514066496164,
"grad_norm": 1.135470942833899,
"learning_rate": 0.0001408929238436033,
"loss": 1.0494,
"step": 295
},
{
"epoch": 1.5140664961636827,
"grad_norm": 1.2574882854332994,
"learning_rate": 0.00014070667095575424,
"loss": 1.0329,
"step": 296
},
{
"epoch": 1.5191815856777495,
"grad_norm": 0.831583885097438,
"learning_rate": 0.0001405196390696017,
"loss": 1.0354,
"step": 297
},
{
"epoch": 1.5242966751918159,
"grad_norm": 0.8122054069061297,
"learning_rate": 0.00014033183058517062,
"loss": 1.0373,
"step": 298
},
{
"epoch": 1.5294117647058822,
"grad_norm": 0.789271302529768,
"learning_rate": 0.00014014324791245133,
"loss": 1.0661,
"step": 299
},
{
"epoch": 1.5345268542199488,
"grad_norm": 0.8876298298886537,
"learning_rate": 0.00013995389347136867,
"loss": 1.0772,
"step": 300
},
{
"epoch": 1.5396419437340154,
"grad_norm": 1.222594187103921,
"learning_rate": 0.00013976376969175098,
"loss": 1.0295,
"step": 301
},
{
"epoch": 1.5447570332480818,
"grad_norm": 0.8015203752241604,
"learning_rate": 0.00013957287901329886,
"loss": 1.0439,
"step": 302
},
{
"epoch": 1.5498721227621484,
"grad_norm": 0.7636174781609883,
"learning_rate": 0.0001393812238855539,
"loss": 1.0256,
"step": 303
},
{
"epoch": 1.554987212276215,
"grad_norm": 0.8361332402971631,
"learning_rate": 0.00013918880676786722,
"loss": 1.0638,
"step": 304
},
{
"epoch": 1.5601023017902813,
"grad_norm": 0.7824565656315564,
"learning_rate": 0.0001389956301293679,
"loss": 1.0348,
"step": 305
},
{
"epoch": 1.5652173913043477,
"grad_norm": 0.7422549311900589,
"learning_rate": 0.00013880169644893135,
"loss": 1.0527,
"step": 306
},
{
"epoch": 1.5703324808184145,
"grad_norm": 0.7467065337377603,
"learning_rate": 0.0001386070082151475,
"loss": 1.0486,
"step": 307
},
{
"epoch": 1.5754475703324808,
"grad_norm": 0.8423161393459444,
"learning_rate": 0.0001384115679262888,
"loss": 1.0466,
"step": 308
},
{
"epoch": 1.5805626598465472,
"grad_norm": 1.090652519689791,
"learning_rate": 0.00013821537809027818,
"loss": 1.0468,
"step": 309
},
{
"epoch": 1.5856777493606138,
"grad_norm": 1.12758760950078,
"learning_rate": 0.00013801844122465694,
"loss": 1.0219,
"step": 310
},
{
"epoch": 1.5907928388746804,
"grad_norm": 0.9857228161657801,
"learning_rate": 0.0001378207598565523,
"loss": 1.0627,
"step": 311
},
{
"epoch": 1.5959079283887467,
"grad_norm": 1.0964519178575107,
"learning_rate": 0.00013762233652264518,
"loss": 1.0654,
"step": 312
},
{
"epoch": 1.6010230179028133,
"grad_norm": 0.9862402374018855,
"learning_rate": 0.00013742317376913744,
"loss": 1.0593,
"step": 313
},
{
"epoch": 1.60613810741688,
"grad_norm": 1.0057422548976498,
"learning_rate": 0.00013722327415171932,
"loss": 1.0503,
"step": 314
},
{
"epoch": 1.6112531969309463,
"grad_norm": 1.0581487003208225,
"learning_rate": 0.00013702264023553672,
"loss": 1.0354,
"step": 315
},
{
"epoch": 1.6163682864450126,
"grad_norm": 1.0063487888499572,
"learning_rate": 0.000136821274595158,
"loss": 1.0325,
"step": 316
},
{
"epoch": 1.6214833759590794,
"grad_norm": 0.8988296059520002,
"learning_rate": 0.00013661917981454126,
"loss": 1.0509,
"step": 317
},
{
"epoch": 1.6265984654731458,
"grad_norm": 0.8858205141372513,
"learning_rate": 0.00013641635848700103,
"loss": 1.0396,
"step": 318
},
{
"epoch": 1.6317135549872122,
"grad_norm": 0.9577786403021353,
"learning_rate": 0.00013621281321517497,
"loss": 1.0605,
"step": 319
},
{
"epoch": 1.6368286445012787,
"grad_norm": 1.177906683540007,
"learning_rate": 0.00013600854661099062,
"loss": 1.032,
"step": 320
},
{
"epoch": 1.6419437340153453,
"grad_norm": 0.7028559175743067,
"learning_rate": 0.00013580356129563162,
"loss": 1.023,
"step": 321
},
{
"epoch": 1.6470588235294117,
"grad_norm": 0.6254370463820547,
"learning_rate": 0.00013559785989950438,
"loss": 1.0894,
"step": 322
},
{
"epoch": 1.6521739130434783,
"grad_norm": 0.8358491408593351,
"learning_rate": 0.00013539144506220414,
"loss": 1.0277,
"step": 323
},
{
"epoch": 1.6572890025575449,
"grad_norm": 0.9314769370329481,
"learning_rate": 0.00013518431943248103,
"loss": 1.0307,
"step": 324
},
{
"epoch": 1.6624040920716112,
"grad_norm": 0.9766494499709318,
"learning_rate": 0.0001349764856682064,
"loss": 1.0354,
"step": 325
},
{
"epoch": 1.6675191815856778,
"grad_norm": 1.0430759085655377,
"learning_rate": 0.00013476794643633828,
"loss": 1.0634,
"step": 326
},
{
"epoch": 1.6726342710997444,
"grad_norm": 1.1584154642373328,
"learning_rate": 0.00013455870441288754,
"loss": 1.0477,
"step": 327
},
{
"epoch": 1.6777493606138107,
"grad_norm": 0.8209211656616354,
"learning_rate": 0.00013434876228288328,
"loss": 1.0518,
"step": 328
},
{
"epoch": 1.682864450127877,
"grad_norm": 0.5950142309084981,
"learning_rate": 0.0001341381227403386,
"loss": 1.0413,
"step": 329
},
{
"epoch": 1.6879795396419437,
"grad_norm": 0.4248896581193454,
"learning_rate": 0.00013392678848821583,
"loss": 1.0595,
"step": 330
},
{
"epoch": 1.6930946291560103,
"grad_norm": 0.40787219995022117,
"learning_rate": 0.00013371476223839195,
"loss": 1.0542,
"step": 331
},
{
"epoch": 1.6982097186700766,
"grad_norm": 0.5982166226472746,
"learning_rate": 0.00013350204671162377,
"loss": 1.0473,
"step": 332
},
{
"epoch": 1.7033248081841432,
"grad_norm": 0.8234498529525044,
"learning_rate": 0.00013328864463751306,
"loss": 1.0422,
"step": 333
},
{
"epoch": 1.7084398976982098,
"grad_norm": 0.9120498916678802,
"learning_rate": 0.0001330745587544714,
"loss": 1.0445,
"step": 334
},
{
"epoch": 1.7135549872122762,
"grad_norm": 0.8536664285634687,
"learning_rate": 0.0001328597918096852,
"loss": 1.0591,
"step": 335
},
{
"epoch": 1.7186700767263428,
"grad_norm": 0.7798989542046489,
"learning_rate": 0.00013264434655908038,
"loss": 1.042,
"step": 336
},
{
"epoch": 1.7237851662404093,
"grad_norm": 0.7012097027827464,
"learning_rate": 0.0001324282257672869,
"loss": 1.0699,
"step": 337
},
{
"epoch": 1.7289002557544757,
"grad_norm": 0.5965066691156631,
"learning_rate": 0.00013221143220760342,
"loss": 1.0337,
"step": 338
},
{
"epoch": 1.734015345268542,
"grad_norm": 0.5689064722949222,
"learning_rate": 0.0001319939686619617,
"loss": 1.0305,
"step": 339
},
{
"epoch": 1.7391304347826086,
"grad_norm": 0.6528231761426132,
"learning_rate": 0.00013177583792089095,
"loss": 1.037,
"step": 340
},
{
"epoch": 1.7442455242966752,
"grad_norm": 0.6357980288661895,
"learning_rate": 0.00013155704278348173,
"loss": 1.05,
"step": 341
},
{
"epoch": 1.7493606138107416,
"grad_norm": 0.6725173799471241,
"learning_rate": 0.00013133758605735044,
"loss": 1.0486,
"step": 342
},
{
"epoch": 1.7544757033248082,
"grad_norm": 0.8324964493472449,
"learning_rate": 0.00013111747055860303,
"loss": 1.0654,
"step": 343
},
{
"epoch": 1.7595907928388748,
"grad_norm": 1.01836225761936,
"learning_rate": 0.00013089669911179892,
"loss": 1.0426,
"step": 344
},
{
"epoch": 1.7647058823529411,
"grad_norm": 1.2706031653553156,
"learning_rate": 0.00013067527454991475,
"loss": 1.041,
"step": 345
},
{
"epoch": 1.7698209718670077,
"grad_norm": 0.6881061023678446,
"learning_rate": 0.00013045319971430813,
"loss": 1.0352,
"step": 346
},
{
"epoch": 1.7749360613810743,
"grad_norm": 0.41782627403046635,
"learning_rate": 0.000130230477454681,
"loss": 1.0365,
"step": 347
},
{
"epoch": 1.7800511508951407,
"grad_norm": 0.6511442040505736,
"learning_rate": 0.0001300071106290432,
"loss": 1.0424,
"step": 348
},
{
"epoch": 1.785166240409207,
"grad_norm": 1.00122755561264,
"learning_rate": 0.00012978310210367573,
"loss": 1.0496,
"step": 349
},
{
"epoch": 1.7902813299232738,
"grad_norm": 1.2999707124683983,
"learning_rate": 0.00012955845475309404,
"loss": 1.0485,
"step": 350
},
{
"epoch": 1.7953964194373402,
"grad_norm": 0.7604920157178511,
"learning_rate": 0.00012933317146001101,
"loss": 1.0449,
"step": 351
},
{
"epoch": 1.8005115089514065,
"grad_norm": 0.7313280145451542,
"learning_rate": 0.0001291072551153001,
"loss": 1.0388,
"step": 352
},
{
"epoch": 1.8056265984654731,
"grad_norm": 0.7209698592252862,
"learning_rate": 0.00012888070861795817,
"loss": 1.0314,
"step": 353
},
{
"epoch": 1.8107416879795397,
"grad_norm": 0.4704918814827189,
"learning_rate": 0.00012865353487506837,
"loss": 1.0166,
"step": 354
},
{
"epoch": 1.815856777493606,
"grad_norm": 0.5594052291382118,
"learning_rate": 0.00012842573680176266,
"loss": 1.0162,
"step": 355
},
{
"epoch": 1.8209718670076727,
"grad_norm": 0.8592492119844757,
"learning_rate": 0.0001281973173211846,
"loss": 1.0713,
"step": 356
},
{
"epoch": 1.8260869565217392,
"grad_norm": 0.9047214294696422,
"learning_rate": 0.00012796827936445174,
"loss": 1.0344,
"step": 357
},
{
"epoch": 1.8312020460358056,
"grad_norm": 0.7971596491179705,
"learning_rate": 0.000127738625870618,
"loss": 1.0302,
"step": 358
},
{
"epoch": 1.836317135549872,
"grad_norm": 1.001864978195367,
"learning_rate": 0.00012750835978663595,
"loss": 1.0274,
"step": 359
},
{
"epoch": 1.8414322250639388,
"grad_norm": 1.519342002494586,
"learning_rate": 0.00012727748406731907,
"loss": 1.0617,
"step": 360
},
{
"epoch": 1.8465473145780051,
"grad_norm": 0.5804834547212457,
"learning_rate": 0.00012704600167530382,
"loss": 1.0474,
"step": 361
},
{
"epoch": 1.8516624040920715,
"grad_norm": 1.244178515922692,
"learning_rate": 0.00012681391558101149,
"loss": 1.0337,
"step": 362
},
{
"epoch": 1.856777493606138,
"grad_norm": 1.523561449211884,
"learning_rate": 0.00012658122876261021,
"loss": 1.0351,
"step": 363
},
{
"epoch": 1.8618925831202047,
"grad_norm": 0.6774947177487926,
"learning_rate": 0.00012634794420597674,
"loss": 1.0175,
"step": 364
},
{
"epoch": 1.867007672634271,
"grad_norm": 1.5925817066330552,
"learning_rate": 0.00012611406490465812,
"loss": 1.0642,
"step": 365
},
{
"epoch": 1.8721227621483376,
"grad_norm": 0.9024164489305665,
"learning_rate": 0.00012587959385983324,
"loss": 1.0234,
"step": 366
},
{
"epoch": 1.8772378516624042,
"grad_norm": 1.115764120453565,
"learning_rate": 0.00012564453408027426,
"loss": 1.0397,
"step": 367
},
{
"epoch": 1.8823529411764706,
"grad_norm": 1.1284878583531806,
"learning_rate": 0.00012540888858230824,
"loss": 1.0617,
"step": 368
},
{
"epoch": 1.887468030690537,
"grad_norm": 1.369054346796234,
"learning_rate": 0.00012517266038977818,
"loss": 1.0484,
"step": 369
},
{
"epoch": 1.8925831202046037,
"grad_norm": 0.5471933614937924,
"learning_rate": 0.0001249358525340043,
"loss": 1.0218,
"step": 370
},
{
"epoch": 1.89769820971867,
"grad_norm": 1.6768791794846525,
"learning_rate": 0.0001246984680537452,
"loss": 1.0132,
"step": 371
},
{
"epoch": 1.9028132992327365,
"grad_norm": 0.613549742278663,
"learning_rate": 0.00012446050999515888,
"loss": 1.0453,
"step": 372
},
{
"epoch": 1.907928388746803,
"grad_norm": 1.5773005910681697,
"learning_rate": 0.00012422198141176345,
"loss": 1.0459,
"step": 373
},
{
"epoch": 1.9130434782608696,
"grad_norm": 0.8312979883835351,
"learning_rate": 0.0001239828853643982,
"loss": 1.0369,
"step": 374
},
{
"epoch": 1.918158567774936,
"grad_norm": 1.3845792221589313,
"learning_rate": 0.0001237432249211842,
"loss": 1.0463,
"step": 375
},
{
"epoch": 1.9232736572890026,
"grad_norm": 1.0736341423406188,
"learning_rate": 0.00012350300315748484,
"loss": 1.0561,
"step": 376
},
{
"epoch": 1.9283887468030692,
"grad_norm": 0.8621125590588978,
"learning_rate": 0.00012326222315586668,
"loss": 1.0426,
"step": 377
},
{
"epoch": 1.9335038363171355,
"grad_norm": 0.904722020788263,
"learning_rate": 0.00012302088800605952,
"loss": 1.0259,
"step": 378
},
{
"epoch": 1.938618925831202,
"grad_norm": 0.6732585033574245,
"learning_rate": 0.00012277900080491692,
"loss": 1.0313,
"step": 379
},
{
"epoch": 1.9437340153452687,
"grad_norm": 0.6416698124531928,
"learning_rate": 0.00012253656465637653,
"loss": 1.0711,
"step": 380
},
{
"epoch": 1.948849104859335,
"grad_norm": 0.7122434752056842,
"learning_rate": 0.00012229358267142008,
"loss": 1.0713,
"step": 381
},
{
"epoch": 1.9539641943734014,
"grad_norm": 0.54611607531493,
"learning_rate": 0.00012205005796803375,
"loss": 1.0335,
"step": 382
},
{
"epoch": 1.959079283887468,
"grad_norm": 0.5359712603831651,
"learning_rate": 0.00012180599367116776,
"loss": 1.0232,
"step": 383
},
{
"epoch": 1.9641943734015346,
"grad_norm": 0.5280063176632849,
"learning_rate": 0.00012156139291269664,
"loss": 1.027,
"step": 384
},
{
"epoch": 1.969309462915601,
"grad_norm": 0.535083976279752,
"learning_rate": 0.0001213162588313788,
"loss": 1.0486,
"step": 385
},
{
"epoch": 1.9744245524296675,
"grad_norm": 0.4782403691033259,
"learning_rate": 0.00012107059457281637,
"loss": 1.0256,
"step": 386
},
{
"epoch": 1.979539641943734,
"grad_norm": 0.565136073142885,
"learning_rate": 0.00012082440328941484,
"loss": 1.0443,
"step": 387
},
{
"epoch": 1.9846547314578005,
"grad_norm": 0.5877927821168869,
"learning_rate": 0.0001205776881403425,
"loss": 1.0313,
"step": 388
},
{
"epoch": 1.989769820971867,
"grad_norm": 0.6255407764215224,
"learning_rate": 0.00012033045229149005,
"loss": 1.0342,
"step": 389
},
{
"epoch": 1.9948849104859336,
"grad_norm": 0.6759056742280855,
"learning_rate": 0.00012008269891542978,
"loss": 1.0514,
"step": 390
},
{
"epoch": 2.0,
"grad_norm": 0.8561590379204816,
"learning_rate": 0.00011983443119137513,
"loss": 1.5296,
"step": 391
},
{
"epoch": 2.0051150895140664,
"grad_norm": 1.381016128078137,
"learning_rate": 0.00011958565230513966,
"loss": 1.0089,
"step": 392
},
{
"epoch": 2.010230179028133,
"grad_norm": 0.8032002356442994,
"learning_rate": 0.00011933636544909622,
"loss": 1.0454,
"step": 393
},
{
"epoch": 2.0153452685421995,
"grad_norm": 0.43347228006186905,
"learning_rate": 0.00011908657382213605,
"loss": 1.0244,
"step": 394
},
{
"epoch": 2.020460358056266,
"grad_norm": 0.9300842333247236,
"learning_rate": 0.00011883628062962773,
"loss": 1.0108,
"step": 395
},
{
"epoch": 2.0255754475703327,
"grad_norm": 1.2033668729998865,
"learning_rate": 0.00011858548908337595,
"loss": 1.0058,
"step": 396
},
{
"epoch": 2.030690537084399,
"grad_norm": 0.8070429133656032,
"learning_rate": 0.00011833420240158042,
"loss": 1.0024,
"step": 397
},
{
"epoch": 2.0358056265984654,
"grad_norm": 0.5467994463899325,
"learning_rate": 0.00011808242380879444,
"loss": 1.0095,
"step": 398
},
{
"epoch": 2.040920716112532,
"grad_norm": 0.38547874756738976,
"learning_rate": 0.00011783015653588369,
"loss": 1.0221,
"step": 399
},
{
"epoch": 2.0460358056265986,
"grad_norm": 0.45167488291589547,
"learning_rate": 0.00011757740381998459,
"loss": 1.0307,
"step": 400
},
{
"epoch": 2.051150895140665,
"grad_norm": 0.5728713415264914,
"learning_rate": 0.00011732416890446289,
"loss": 1.0102,
"step": 401
},
{
"epoch": 2.0562659846547313,
"grad_norm": 0.5622015885826027,
"learning_rate": 0.00011707045503887201,
"loss": 1.0321,
"step": 402
},
{
"epoch": 2.061381074168798,
"grad_norm": 0.47465694453487833,
"learning_rate": 0.0001168162654789113,
"loss": 1.0022,
"step": 403
},
{
"epoch": 2.0664961636828645,
"grad_norm": 0.38423696046132405,
"learning_rate": 0.00011656160348638436,
"loss": 1.0368,
"step": 404
},
{
"epoch": 2.071611253196931,
"grad_norm": 0.35468814431123347,
"learning_rate": 0.00011630647232915709,
"loss": 1.0022,
"step": 405
},
{
"epoch": 2.0767263427109977,
"grad_norm": 0.3621831137612044,
"learning_rate": 0.00011605087528111576,
"loss": 1.0124,
"step": 406
},
{
"epoch": 2.081841432225064,
"grad_norm": 0.4590860171853655,
"learning_rate": 0.00011579481562212512,
"loss": 1.0342,
"step": 407
},
{
"epoch": 2.0869565217391304,
"grad_norm": 0.6112384224299995,
"learning_rate": 0.0001155382966379861,
"loss": 0.9972,
"step": 408
},
{
"epoch": 2.0920716112531967,
"grad_norm": 0.695206097014345,
"learning_rate": 0.00011528132162039391,
"loss": 1.0037,
"step": 409
},
{
"epoch": 2.0971867007672635,
"grad_norm": 0.788245237616969,
"learning_rate": 0.00011502389386689551,
"loss": 0.9974,
"step": 410
},
{
"epoch": 2.10230179028133,
"grad_norm": 0.9202020610857758,
"learning_rate": 0.00011476601668084759,
"loss": 1.0061,
"step": 411
},
{
"epoch": 2.1074168797953963,
"grad_norm": 1.0046129321237383,
"learning_rate": 0.00011450769337137389,
"loss": 1.025,
"step": 412
},
{
"epoch": 2.112531969309463,
"grad_norm": 0.83434322052857,
"learning_rate": 0.000114248927253323,
"loss": 1.0215,
"step": 413
},
{
"epoch": 2.1176470588235294,
"grad_norm": 0.5524509049937103,
"learning_rate": 0.00011398972164722564,
"loss": 1.0125,
"step": 414
},
{
"epoch": 2.122762148337596,
"grad_norm": 0.3945345722394861,
"learning_rate": 0.00011373007987925218,
"loss": 1.0342,
"step": 415
},
{
"epoch": 2.1278772378516626,
"grad_norm": 0.5147364920457403,
"learning_rate": 0.00011347000528116976,
"loss": 1.0367,
"step": 416
},
{
"epoch": 2.132992327365729,
"grad_norm": 0.6536344129940352,
"learning_rate": 0.00011320950119029985,
"loss": 1.0158,
"step": 417
},
{
"epoch": 2.1381074168797953,
"grad_norm": 0.6676422009570077,
"learning_rate": 0.00011294857094947507,
"loss": 1.0329,
"step": 418
},
{
"epoch": 2.1432225063938617,
"grad_norm": 0.5735086201100545,
"learning_rate": 0.0001126872179069966,
"loss": 1.0116,
"step": 419
},
{
"epoch": 2.1483375959079285,
"grad_norm": 0.47601350406210197,
"learning_rate": 0.00011242544541659102,
"loss": 1.0285,
"step": 420
},
{
"epoch": 2.153452685421995,
"grad_norm": 0.4143559734940057,
"learning_rate": 0.00011216325683736734,
"loss": 1.0246,
"step": 421
},
{
"epoch": 2.1585677749360612,
"grad_norm": 0.34742299383983577,
"learning_rate": 0.00011190065553377392,
"loss": 1.0288,
"step": 422
},
{
"epoch": 2.163682864450128,
"grad_norm": 0.4446200862863137,
"learning_rate": 0.00011163764487555522,
"loss": 1.0089,
"step": 423
},
{
"epoch": 2.1687979539641944,
"grad_norm": 0.3904742024949866,
"learning_rate": 0.00011137422823770866,
"loss": 1.0162,
"step": 424
},
{
"epoch": 2.1739130434782608,
"grad_norm": 0.3468757223042228,
"learning_rate": 0.00011111040900044125,
"loss": 1.0203,
"step": 425
},
{
"epoch": 2.1790281329923276,
"grad_norm": 0.39737691764617955,
"learning_rate": 0.00011084619054912619,
"loss": 1.0335,
"step": 426
},
{
"epoch": 2.184143222506394,
"grad_norm": 0.3383994927388332,
"learning_rate": 0.00011058157627425948,
"loss": 0.9695,
"step": 427
},
{
"epoch": 2.1892583120204603,
"grad_norm": 0.33040617935487243,
"learning_rate": 0.00011031656957141643,
"loss": 1.0244,
"step": 428
},
{
"epoch": 2.1943734015345266,
"grad_norm": 0.4212857692090894,
"learning_rate": 0.00011005117384120798,
"loss": 1.0466,
"step": 429
},
{
"epoch": 2.1994884910485935,
"grad_norm": 0.6213113924496735,
"learning_rate": 0.00010978539248923722,
"loss": 1.0372,
"step": 430
},
{
"epoch": 2.20460358056266,
"grad_norm": 0.7462378811042474,
"learning_rate": 0.00010951922892605553,
"loss": 1.0139,
"step": 431
},
{
"epoch": 2.209718670076726,
"grad_norm": 0.6984427005704308,
"learning_rate": 0.00010925268656711894,
"loss": 1.023,
"step": 432
},
{
"epoch": 2.214833759590793,
"grad_norm": 0.6346802419956531,
"learning_rate": 0.0001089857688327442,
"loss": 0.9984,
"step": 433
},
{
"epoch": 2.2199488491048593,
"grad_norm": 0.612789907689284,
"learning_rate": 0.00010871847914806503,
"loss": 1.0259,
"step": 434
},
{
"epoch": 2.2250639386189257,
"grad_norm": 0.5284138979478009,
"learning_rate": 0.00010845082094298794,
"loss": 1.0258,
"step": 435
},
{
"epoch": 2.2301790281329925,
"grad_norm": 0.36597869072338884,
"learning_rate": 0.00010818279765214853,
"loss": 1.0084,
"step": 436
},
{
"epoch": 2.235294117647059,
"grad_norm": 0.29982136342979027,
"learning_rate": 0.00010791441271486708,
"loss": 1.0287,
"step": 437
},
{
"epoch": 2.2404092071611252,
"grad_norm": 0.30479749358568886,
"learning_rate": 0.00010764566957510469,
"loss": 1.0012,
"step": 438
},
{
"epoch": 2.2455242966751916,
"grad_norm": 0.3305269824424927,
"learning_rate": 0.00010737657168141889,
"loss": 1.0037,
"step": 439
},
{
"epoch": 2.2506393861892584,
"grad_norm": 0.35155297727172763,
"learning_rate": 0.00010710712248691956,
"loss": 0.995,
"step": 440
},
{
"epoch": 2.2557544757033248,
"grad_norm": 0.4137242643914776,
"learning_rate": 0.00010683732544922449,
"loss": 0.987,
"step": 441
},
{
"epoch": 2.260869565217391,
"grad_norm": 0.3905105509743982,
"learning_rate": 0.00010656718403041502,
"loss": 1.007,
"step": 442
},
{
"epoch": 2.265984654731458,
"grad_norm": 0.2770119849444069,
"learning_rate": 0.0001062967016969917,
"loss": 1.0284,
"step": 443
},
{
"epoch": 2.2710997442455243,
"grad_norm": 0.26449698605636457,
"learning_rate": 0.00010602588191982973,
"loss": 0.9899,
"step": 444
},
{
"epoch": 2.2762148337595907,
"grad_norm": 0.31210485637648877,
"learning_rate": 0.00010575472817413442,
"loss": 1.0327,
"step": 445
},
{
"epoch": 2.2813299232736575,
"grad_norm": 0.2858229366718099,
"learning_rate": 0.00010548324393939672,
"loss": 1.0193,
"step": 446
},
{
"epoch": 2.286445012787724,
"grad_norm": 0.22011619436189744,
"learning_rate": 0.00010521143269934834,
"loss": 1.0272,
"step": 447
},
{
"epoch": 2.29156010230179,
"grad_norm": 0.27090570626400357,
"learning_rate": 0.00010493929794191724,
"loss": 1.0181,
"step": 448
},
{
"epoch": 2.296675191815857,
"grad_norm": 0.31785604120748834,
"learning_rate": 0.00010466684315918282,
"loss": 1.0006,
"step": 449
},
{
"epoch": 2.3017902813299234,
"grad_norm": 0.3032636927731663,
"learning_rate": 0.0001043940718473311,
"loss": 1.0147,
"step": 450
},
{
"epoch": 2.3069053708439897,
"grad_norm": 0.2375690504332674,
"learning_rate": 0.00010412098750660978,
"loss": 1.0018,
"step": 451
},
{
"epoch": 2.312020460358056,
"grad_norm": 0.3456416612358553,
"learning_rate": 0.00010384759364128357,
"loss": 0.9881,
"step": 452
},
{
"epoch": 2.317135549872123,
"grad_norm": 0.4194030878444893,
"learning_rate": 0.00010357389375958886,
"loss": 0.9927,
"step": 453
},
{
"epoch": 2.3222506393861893,
"grad_norm": 0.46414687932122983,
"learning_rate": 0.00010329989137368903,
"loss": 1.0227,
"step": 454
},
{
"epoch": 2.3273657289002556,
"grad_norm": 0.6333868297695803,
"learning_rate": 0.00010302558999962913,
"loss": 1.0179,
"step": 455
},
{
"epoch": 2.3324808184143224,
"grad_norm": 0.8246141961755575,
"learning_rate": 0.00010275099315729098,
"loss": 1.0122,
"step": 456
},
{
"epoch": 2.337595907928389,
"grad_norm": 0.9329035035972839,
"learning_rate": 0.00010247610437034788,
"loss": 1.0256,
"step": 457
},
{
"epoch": 2.342710997442455,
"grad_norm": 0.8920959003753987,
"learning_rate": 0.00010220092716621933,
"loss": 1.0336,
"step": 458
},
{
"epoch": 2.3478260869565215,
"grad_norm": 0.7649026682124002,
"learning_rate": 0.00010192546507602599,
"loss": 0.9979,
"step": 459
},
{
"epoch": 2.3529411764705883,
"grad_norm": 0.5318293860445822,
"learning_rate": 0.00010164972163454407,
"loss": 0.9965,
"step": 460
},
{
"epoch": 2.3580562659846547,
"grad_norm": 0.3706902670999193,
"learning_rate": 0.00010137370038016026,
"loss": 1.0004,
"step": 461
},
{
"epoch": 2.363171355498721,
"grad_norm": 0.361701979402021,
"learning_rate": 0.00010109740485482612,
"loss": 1.0072,
"step": 462
},
{
"epoch": 2.368286445012788,
"grad_norm": 0.4019014044381587,
"learning_rate": 0.0001008208386040127,
"loss": 0.986,
"step": 463
},
{
"epoch": 2.373401534526854,
"grad_norm": 0.4620567850345528,
"learning_rate": 0.00010054400517666505,
"loss": 1.0342,
"step": 464
},
{
"epoch": 2.3785166240409206,
"grad_norm": 0.4681673124155279,
"learning_rate": 0.00010026690812515676,
"loss": 1.0193,
"step": 465
},
{
"epoch": 2.3836317135549874,
"grad_norm": 0.42130381882854007,
"learning_rate": 9.998955100524414e-05,
"loss": 1.0097,
"step": 466
},
{
"epoch": 2.3887468030690537,
"grad_norm": 0.33474077626940735,
"learning_rate": 9.971193737602085e-05,
"loss": 1.0008,
"step": 467
},
{
"epoch": 2.39386189258312,
"grad_norm": 0.29874381080717,
"learning_rate": 9.94340707998721e-05,
"loss": 1.0015,
"step": 468
},
{
"epoch": 2.398976982097187,
"grad_norm": 0.2804936921637105,
"learning_rate": 9.915595484242893e-05,
"loss": 1.0164,
"step": 469
},
{
"epoch": 2.4040920716112533,
"grad_norm": 0.28970206873273624,
"learning_rate": 9.887759307252252e-05,
"loss": 1.0069,
"step": 470
},
{
"epoch": 2.4092071611253196,
"grad_norm": 0.2663880382326739,
"learning_rate": 9.859898906213834e-05,
"loss": 1.0084,
"step": 471
},
{
"epoch": 2.414322250639386,
"grad_norm": 0.3201947655467473,
"learning_rate": 9.832014638637028e-05,
"loss": 1.0157,
"step": 472
},
{
"epoch": 2.419437340153453,
"grad_norm": 0.2791216226508958,
"learning_rate": 9.804106862337496e-05,
"loss": 0.9785,
"step": 473
},
{
"epoch": 2.424552429667519,
"grad_norm": 0.31822985832697637,
"learning_rate": 9.77617593543255e-05,
"loss": 1.0181,
"step": 474
},
{
"epoch": 2.4296675191815855,
"grad_norm": 0.2927332305658229,
"learning_rate": 9.748222216336591e-05,
"loss": 1.0018,
"step": 475
},
{
"epoch": 2.4347826086956523,
"grad_norm": 0.30006184933469426,
"learning_rate": 9.720246063756482e-05,
"loss": 0.9921,
"step": 476
},
{
"epoch": 2.4398976982097187,
"grad_norm": 0.3416495170384536,
"learning_rate": 9.692247836686964e-05,
"loss": 1.0278,
"step": 477
},
{
"epoch": 2.445012787723785,
"grad_norm": 0.3687081034325788,
"learning_rate": 9.664227894406033e-05,
"loss": 1.0093,
"step": 478
},
{
"epoch": 2.4501278772378514,
"grad_norm": 0.370813800196682,
"learning_rate": 9.63618659647035e-05,
"loss": 0.9817,
"step": 479
},
{
"epoch": 2.455242966751918,
"grad_norm": 0.3369139030783905,
"learning_rate": 9.608124302710603e-05,
"loss": 0.9864,
"step": 480
},
{
"epoch": 2.4603580562659846,
"grad_norm": 0.3663218481972242,
"learning_rate": 9.580041373226909e-05,
"loss": 1.0331,
"step": 481
},
{
"epoch": 2.4654731457800514,
"grad_norm": 0.30714759554897403,
"learning_rate": 9.551938168384183e-05,
"loss": 1.0296,
"step": 482
},
{
"epoch": 2.4705882352941178,
"grad_norm": 0.31231457503575155,
"learning_rate": 9.523815048807525e-05,
"loss": 0.9932,
"step": 483
},
{
"epoch": 2.475703324808184,
"grad_norm": 0.28265453071623914,
"learning_rate": 9.495672375377572e-05,
"loss": 1.0106,
"step": 484
},
{
"epoch": 2.4808184143222505,
"grad_norm": 0.6253742051342691,
"learning_rate": 9.467510509225887e-05,
"loss": 1.0009,
"step": 485
},
{
"epoch": 2.4859335038363173,
"grad_norm": 0.3301521421025129,
"learning_rate": 9.439329811730313e-05,
"loss": 1.0149,
"step": 486
},
{
"epoch": 2.4910485933503836,
"grad_norm": 0.3339607112416421,
"learning_rate": 9.411130644510346e-05,
"loss": 1.0063,
"step": 487
},
{
"epoch": 2.49616368286445,
"grad_norm": 0.43097055987261607,
"learning_rate": 9.382913369422483e-05,
"loss": 1.0006,
"step": 488
},
{
"epoch": 2.501278772378517,
"grad_norm": 0.4628233259981002,
"learning_rate": 9.354678348555588e-05,
"loss": 1.01,
"step": 489
},
{
"epoch": 2.506393861892583,
"grad_norm": 0.42109201238862093,
"learning_rate": 9.32642594422624e-05,
"loss": 1.0005,
"step": 490
},
{
"epoch": 2.5115089514066495,
"grad_norm": 0.3000190268265243,
"learning_rate": 9.298156518974083e-05,
"loss": 0.9828,
"step": 491
},
{
"epoch": 2.516624040920716,
"grad_norm": 0.34447764215788584,
"learning_rate": 9.269870435557182e-05,
"loss": 0.9983,
"step": 492
},
{
"epoch": 2.5217391304347827,
"grad_norm": 0.34719463151150515,
"learning_rate": 9.241568056947357e-05,
"loss": 0.9957,
"step": 493
},
{
"epoch": 2.526854219948849,
"grad_norm": 0.3562431622604642,
"learning_rate": 9.213249746325538e-05,
"loss": 1.009,
"step": 494
},
{
"epoch": 2.531969309462916,
"grad_norm": 0.4561876850442379,
"learning_rate": 9.184915867077087e-05,
"loss": 1.0251,
"step": 495
},
{
"epoch": 2.5370843989769822,
"grad_norm": 0.4511984999410279,
"learning_rate": 9.156566782787153e-05,
"loss": 1.0187,
"step": 496
},
{
"epoch": 2.5421994884910486,
"grad_norm": 0.44832399361859254,
"learning_rate": 9.128202857235994e-05,
"loss": 1.001,
"step": 497
},
{
"epoch": 2.547314578005115,
"grad_norm": 0.33173037324641885,
"learning_rate": 9.09982445439432e-05,
"loss": 0.994,
"step": 498
},
{
"epoch": 2.5524296675191813,
"grad_norm": 0.29896063500798026,
"learning_rate": 9.07143193841861e-05,
"loss": 1.0045,
"step": 499
},
{
"epoch": 2.557544757033248,
"grad_norm": 0.3423239422564292,
"learning_rate": 9.04302567364645e-05,
"loss": 1.0003,
"step": 500
},
{
"epoch": 2.5626598465473145,
"grad_norm": 0.36432147271465287,
"learning_rate": 9.014606024591843e-05,
"loss": 0.9986,
"step": 501
},
{
"epoch": 2.5677749360613813,
"grad_norm": 0.3596896213509184,
"learning_rate": 8.986173355940555e-05,
"loss": 1.0247,
"step": 502
},
{
"epoch": 2.5728900255754477,
"grad_norm": 0.4405726852473915,
"learning_rate": 8.95772803254541e-05,
"loss": 0.9924,
"step": 503
},
{
"epoch": 2.578005115089514,
"grad_norm": 0.3170631360066317,
"learning_rate": 8.92927041942163e-05,
"loss": 1.0303,
"step": 504
},
{
"epoch": 2.5831202046035804,
"grad_norm": 0.3387391303505355,
"learning_rate": 8.900800881742133e-05,
"loss": 1.0151,
"step": 505
},
{
"epoch": 2.588235294117647,
"grad_norm": 0.28833695365918766,
"learning_rate": 8.872319784832858e-05,
"loss": 1.013,
"step": 506
},
{
"epoch": 2.5933503836317136,
"grad_norm": 0.20607008360250567,
"learning_rate": 8.843827494168073e-05,
"loss": 1.0073,
"step": 507
},
{
"epoch": 2.59846547314578,
"grad_norm": 0.2626303665983081,
"learning_rate": 8.815324375365687e-05,
"loss": 0.9952,
"step": 508
},
{
"epoch": 2.6035805626598467,
"grad_norm": 0.3287929618310504,
"learning_rate": 8.78681079418256e-05,
"loss": 1.01,
"step": 509
},
{
"epoch": 2.608695652173913,
"grad_norm": 0.3133004696328574,
"learning_rate": 8.758287116509804e-05,
"loss": 0.9878,
"step": 510
},
{
"epoch": 2.6138107416879794,
"grad_norm": 0.26137737343431855,
"learning_rate": 8.729753708368088e-05,
"loss": 1.0001,
"step": 511
},
{
"epoch": 2.618925831202046,
"grad_norm": 0.2417595400284646,
"learning_rate": 8.701210935902947e-05,
"loss": 1.0085,
"step": 512
},
{
"epoch": 2.6240409207161126,
"grad_norm": 0.20556326952478393,
"learning_rate": 8.672659165380083e-05,
"loss": 1.0005,
"step": 513
},
{
"epoch": 2.629156010230179,
"grad_norm": 0.21749270028177237,
"learning_rate": 8.644098763180664e-05,
"loss": 1.0139,
"step": 514
},
{
"epoch": 2.634271099744246,
"grad_norm": 0.22529865204782643,
"learning_rate": 8.61553009579661e-05,
"loss": 1.0147,
"step": 515
},
{
"epoch": 2.639386189258312,
"grad_norm": 0.29286565968536116,
"learning_rate": 8.586953529825917e-05,
"loss": 0.9954,
"step": 516
},
{
"epoch": 2.6445012787723785,
"grad_norm": 0.3699915833902878,
"learning_rate": 8.558369431967922e-05,
"loss": 1.0262,
"step": 517
},
{
"epoch": 2.649616368286445,
"grad_norm": 0.3086840195749082,
"learning_rate": 8.529778169018625e-05,
"loss": 1.0062,
"step": 518
},
{
"epoch": 2.6547314578005117,
"grad_norm": 0.2876573818915736,
"learning_rate": 8.501180107865959e-05,
"loss": 1.0012,
"step": 519
},
{
"epoch": 2.659846547314578,
"grad_norm": 0.23893237612799384,
"learning_rate": 8.472575615485099e-05,
"loss": 1.0174,
"step": 520
},
{
"epoch": 2.6649616368286444,
"grad_norm": 0.23873637581404566,
"learning_rate": 8.44396505893375e-05,
"loss": 1.0141,
"step": 521
},
{
"epoch": 2.670076726342711,
"grad_norm": 0.19494201056390806,
"learning_rate": 8.415348805347421e-05,
"loss": 0.9877,
"step": 522
},
{
"epoch": 2.6751918158567776,
"grad_norm": 0.21348437759871342,
"learning_rate": 8.386727221934741e-05,
"loss": 0.9866,
"step": 523
},
{
"epoch": 2.680306905370844,
"grad_norm": 0.2610271703225153,
"learning_rate": 8.35810067597272e-05,
"loss": 1.0117,
"step": 524
},
{
"epoch": 2.6854219948849103,
"grad_norm": 0.20408308006728992,
"learning_rate": 8.329469534802058e-05,
"loss": 0.9993,
"step": 525
},
{
"epoch": 2.690537084398977,
"grad_norm": 0.17654912107193485,
"learning_rate": 8.300834165822416e-05,
"loss": 1.0052,
"step": 526
},
{
"epoch": 2.6956521739130435,
"grad_norm": 0.23701778310743846,
"learning_rate": 8.272194936487709e-05,
"loss": 0.9946,
"step": 527
},
{
"epoch": 2.70076726342711,
"grad_norm": 0.20975206836600477,
"learning_rate": 8.243552214301382e-05,
"loss": 1.0019,
"step": 528
},
{
"epoch": 2.7058823529411766,
"grad_norm": 0.2175874800056513,
"learning_rate": 8.214906366811713e-05,
"loss": 1.0069,
"step": 529
},
{
"epoch": 2.710997442455243,
"grad_norm": 0.1921986755255322,
"learning_rate": 8.186257761607076e-05,
"loss": 1.0055,
"step": 530
},
{
"epoch": 2.7161125319693094,
"grad_norm": 0.24795865579153492,
"learning_rate": 8.157606766311236e-05,
"loss": 1.024,
"step": 531
},
{
"epoch": 2.7212276214833757,
"grad_norm": 0.19549655493129472,
"learning_rate": 8.128953748578625e-05,
"loss": 1.0271,
"step": 532
},
{
"epoch": 2.7263427109974425,
"grad_norm": 0.23282883600049625,
"learning_rate": 8.100299076089628e-05,
"loss": 0.9891,
"step": 533
},
{
"epoch": 2.731457800511509,
"grad_norm": 0.26082933290129945,
"learning_rate": 8.071643116545865e-05,
"loss": 1.0149,
"step": 534
},
{
"epoch": 2.7365728900255757,
"grad_norm": 0.26152603055694823,
"learning_rate": 8.042986237665475e-05,
"loss": 1.0227,
"step": 535
},
{
"epoch": 2.741687979539642,
"grad_norm": 0.2426042651939562,
"learning_rate": 8.014328807178386e-05,
"loss": 1.0098,
"step": 536
},
{
"epoch": 2.7468030690537084,
"grad_norm": 0.22784410485991166,
"learning_rate": 7.985671192821616e-05,
"loss": 1.0003,
"step": 537
},
{
"epoch": 2.7519181585677748,
"grad_norm": 0.22699737171816037,
"learning_rate": 7.957013762334525e-05,
"loss": 1.0246,
"step": 538
},
{
"epoch": 2.7570332480818416,
"grad_norm": 0.23210188945309745,
"learning_rate": 7.928356883454137e-05,
"loss": 1.0036,
"step": 539
},
{
"epoch": 2.762148337595908,
"grad_norm": 0.24048763890708685,
"learning_rate": 7.899700923910376e-05,
"loss": 0.9906,
"step": 540
},
{
"epoch": 2.7672634271099743,
"grad_norm": 0.257128135830221,
"learning_rate": 7.871046251421379e-05,
"loss": 0.9925,
"step": 541
},
{
"epoch": 2.772378516624041,
"grad_norm": 0.23202894799634624,
"learning_rate": 7.842393233688767e-05,
"loss": 1.0127,
"step": 542
},
{
"epoch": 2.7774936061381075,
"grad_norm": 0.21076498907777885,
"learning_rate": 7.813742238392924e-05,
"loss": 1.0138,
"step": 543
},
{
"epoch": 2.782608695652174,
"grad_norm": 0.22887397533756398,
"learning_rate": 7.785093633188289e-05,
"loss": 0.9787,
"step": 544
},
{
"epoch": 2.78772378516624,
"grad_norm": 0.2404840944612172,
"learning_rate": 7.75644778569862e-05,
"loss": 0.9963,
"step": 545
},
{
"epoch": 2.792838874680307,
"grad_norm": 0.18629613157074487,
"learning_rate": 7.727805063512294e-05,
"loss": 0.9864,
"step": 546
},
{
"epoch": 2.7979539641943734,
"grad_norm": 0.22879442074029022,
"learning_rate": 7.699165834177585e-05,
"loss": 1.0102,
"step": 547
},
{
"epoch": 2.80306905370844,
"grad_norm": 0.23593054489382964,
"learning_rate": 7.670530465197942e-05,
"loss": 1.0227,
"step": 548
},
{
"epoch": 2.8081841432225065,
"grad_norm": 0.20189988478036155,
"learning_rate": 7.641899324027281e-05,
"loss": 1.0219,
"step": 549
},
{
"epoch": 2.813299232736573,
"grad_norm": 0.24828137261438246,
"learning_rate": 7.613272778065263e-05,
"loss": 1.0045,
"step": 550
},
{
"epoch": 2.8184143222506393,
"grad_norm": 0.2157701472534619,
"learning_rate": 7.584651194652581e-05,
"loss": 1.0125,
"step": 551
},
{
"epoch": 2.8235294117647056,
"grad_norm": 0.17015517958877327,
"learning_rate": 7.556034941066256e-05,
"loss": 1.0053,
"step": 552
},
{
"epoch": 2.8286445012787724,
"grad_norm": 0.22130484910335416,
"learning_rate": 7.527424384514903e-05,
"loss": 0.999,
"step": 553
},
{
"epoch": 2.833759590792839,
"grad_norm": 0.20391913170666773,
"learning_rate": 7.498819892134045e-05,
"loss": 1.0163,
"step": 554
},
{
"epoch": 2.8388746803069056,
"grad_norm": 0.20720182960275763,
"learning_rate": 7.470221830981378e-05,
"loss": 1.0151,
"step": 555
},
{
"epoch": 2.843989769820972,
"grad_norm": 0.21753937615745583,
"learning_rate": 7.441630568032079e-05,
"loss": 1.0092,
"step": 556
},
{
"epoch": 2.8491048593350383,
"grad_norm": 0.22387737130429428,
"learning_rate": 7.413046470174087e-05,
"loss": 1.0021,
"step": 557
},
{
"epoch": 2.8542199488491047,
"grad_norm": 0.22178475826624403,
"learning_rate": 7.38446990420339e-05,
"loss": 1.0033,
"step": 558
},
{
"epoch": 2.8593350383631715,
"grad_norm": 0.2714637281511857,
"learning_rate": 7.355901236819339e-05,
"loss": 1.0008,
"step": 559
},
{
"epoch": 2.864450127877238,
"grad_norm": 0.24531134435835011,
"learning_rate": 7.327340834619917e-05,
"loss": 1.0173,
"step": 560
},
{
"epoch": 2.869565217391304,
"grad_norm": 0.23304836763846518,
"learning_rate": 7.298789064097055e-05,
"loss": 1.0041,
"step": 561
},
{
"epoch": 2.874680306905371,
"grad_norm": 0.23283268931818618,
"learning_rate": 7.270246291631918e-05,
"loss": 1.0065,
"step": 562
},
{
"epoch": 2.8797953964194374,
"grad_norm": 0.28906409841345,
"learning_rate": 7.241712883490198e-05,
"loss": 1.0183,
"step": 563
},
{
"epoch": 2.8849104859335037,
"grad_norm": 0.26482896625331576,
"learning_rate": 7.213189205817443e-05,
"loss": 0.9837,
"step": 564
},
{
"epoch": 2.89002557544757,
"grad_norm": 0.23112874854066667,
"learning_rate": 7.184675624634311e-05,
"loss": 1.0054,
"step": 565
},
{
"epoch": 2.895140664961637,
"grad_norm": 0.24317449030985097,
"learning_rate": 7.15617250583193e-05,
"loss": 0.999,
"step": 566
},
{
"epoch": 2.9002557544757033,
"grad_norm": 0.32732824347638817,
"learning_rate": 7.127680215167146e-05,
"loss": 0.9891,
"step": 567
},
{
"epoch": 2.90537084398977,
"grad_norm": 0.30595491785320433,
"learning_rate": 7.099199118257869e-05,
"loss": 0.9984,
"step": 568
},
{
"epoch": 2.9104859335038364,
"grad_norm": 0.270848644431639,
"learning_rate": 7.070729580578373e-05,
"loss": 1.0151,
"step": 569
},
{
"epoch": 2.915601023017903,
"grad_norm": 0.2644114393044732,
"learning_rate": 7.042271967454591e-05,
"loss": 0.9987,
"step": 570
},
{
"epoch": 2.920716112531969,
"grad_norm": 0.21941955834716276,
"learning_rate": 7.013826644059448e-05,
"loss": 1.0138,
"step": 571
},
{
"epoch": 2.9258312020460355,
"grad_norm": 0.2825462760834641,
"learning_rate": 6.985393975408161e-05,
"loss": 1.0058,
"step": 572
},
{
"epoch": 2.9309462915601023,
"grad_norm": 0.29876161391962847,
"learning_rate": 6.956974326353553e-05,
"loss": 0.9933,
"step": 573
},
{
"epoch": 2.9360613810741687,
"grad_norm": 0.21038863146466047,
"learning_rate": 6.928568061581392e-05,
"loss": 1.0045,
"step": 574
},
{
"epoch": 2.9411764705882355,
"grad_norm": 0.24871666978117873,
"learning_rate": 6.90017554560568e-05,
"loss": 1.0137,
"step": 575
},
{
"epoch": 2.946291560102302,
"grad_norm": 0.2509893555667477,
"learning_rate": 6.871797142764008e-05,
"loss": 1.0075,
"step": 576
},
{
"epoch": 2.9514066496163682,
"grad_norm": 0.18743939396884352,
"learning_rate": 6.843433217212851e-05,
"loss": 1.0183,
"step": 577
},
{
"epoch": 2.9565217391304346,
"grad_norm": 0.20396334066646526,
"learning_rate": 6.815084132922916e-05,
"loss": 1.0243,
"step": 578
},
{
"epoch": 2.9616368286445014,
"grad_norm": 0.2143958956025943,
"learning_rate": 6.786750253674466e-05,
"loss": 1.0003,
"step": 579
},
{
"epoch": 2.9667519181585678,
"grad_norm": 0.18697396090837168,
"learning_rate": 6.758431943052643e-05,
"loss": 1.0061,
"step": 580
},
{
"epoch": 2.971867007672634,
"grad_norm": 0.23047142469720475,
"learning_rate": 6.73012956444282e-05,
"loss": 0.9919,
"step": 581
},
{
"epoch": 2.976982097186701,
"grad_norm": 0.22718676594117237,
"learning_rate": 6.70184348102592e-05,
"loss": 1.011,
"step": 582
},
{
"epoch": 2.9820971867007673,
"grad_norm": 0.17951187476574207,
"learning_rate": 6.673574055773763e-05,
"loss": 1.0128,
"step": 583
},
{
"epoch": 2.9872122762148337,
"grad_norm": 0.21411989833562928,
"learning_rate": 6.645321651444413e-05,
"loss": 1.0125,
"step": 584
},
{
"epoch": 2.9923273657289,
"grad_norm": 0.20680740641794057,
"learning_rate": 6.617086630577517e-05,
"loss": 0.9937,
"step": 585
},
{
"epoch": 2.997442455242967,
"grad_norm": 0.21453037325893434,
"learning_rate": 6.588869355489655e-05,
"loss": 1.01,
"step": 586
},
{
"epoch": 3.002557544757033,
"grad_norm": 0.3030812536548564,
"learning_rate": 6.560670188269691e-05,
"loss": 1.5313,
"step": 587
},
{
"epoch": 3.0076726342710995,
"grad_norm": 0.26671231746704943,
"learning_rate": 6.532489490774116e-05,
"loss": 1.0068,
"step": 588
},
{
"epoch": 3.0127877237851663,
"grad_norm": 0.226380753216681,
"learning_rate": 6.504327624622432e-05,
"loss": 1.0075,
"step": 589
},
{
"epoch": 3.0179028132992327,
"grad_norm": 0.22198165995246655,
"learning_rate": 6.476184951192478e-05,
"loss": 0.9587,
"step": 590
},
{
"epoch": 3.023017902813299,
"grad_norm": 0.25142010177475915,
"learning_rate": 6.448061831615818e-05,
"loss": 0.9979,
"step": 591
},
{
"epoch": 3.028132992327366,
"grad_norm": 0.23039021907280735,
"learning_rate": 6.419958626773094e-05,
"loss": 0.991,
"step": 592
},
{
"epoch": 3.0332480818414322,
"grad_norm": 0.19358315586895924,
"learning_rate": 6.3918756972894e-05,
"loss": 0.9699,
"step": 593
},
{
"epoch": 3.0383631713554986,
"grad_norm": 0.20409950024884876,
"learning_rate": 6.363813403529654e-05,
"loss": 0.9685,
"step": 594
},
{
"epoch": 3.0434782608695654,
"grad_norm": 0.20796747128888024,
"learning_rate": 6.335772105593967e-05,
"loss": 0.9803,
"step": 595
},
{
"epoch": 3.0485933503836318,
"grad_norm": 0.19636346924665282,
"learning_rate": 6.307752163313039e-05,
"loss": 0.9709,
"step": 596
},
{
"epoch": 3.053708439897698,
"grad_norm": 0.20953419178425575,
"learning_rate": 6.279753936243518e-05,
"loss": 0.9833,
"step": 597
},
{
"epoch": 3.0588235294117645,
"grad_norm": 0.21995568214122566,
"learning_rate": 6.25177778366341e-05,
"loss": 0.9853,
"step": 598
},
{
"epoch": 3.0639386189258313,
"grad_norm": 0.20012803133949741,
"learning_rate": 6.223824064567454e-05,
"loss": 0.9842,
"step": 599
},
{
"epoch": 3.0690537084398977,
"grad_norm": 0.20316280739878417,
"learning_rate": 6.195893137662507e-05,
"loss": 0.9971,
"step": 600
},
{
"epoch": 3.074168797953964,
"grad_norm": 0.22980549010912854,
"learning_rate": 6.167985361362975e-05,
"loss": 0.9855,
"step": 601
},
{
"epoch": 3.079283887468031,
"grad_norm": 0.19483410893806968,
"learning_rate": 6.140101093786168e-05,
"loss": 0.9957,
"step": 602
},
{
"epoch": 3.084398976982097,
"grad_norm": 0.2171122750771049,
"learning_rate": 6.11224069274775e-05,
"loss": 0.9943,
"step": 603
},
{
"epoch": 3.0895140664961636,
"grad_norm": 0.2151667311212257,
"learning_rate": 6.08440451575711e-05,
"loss": 1.0177,
"step": 604
},
{
"epoch": 3.0946291560102304,
"grad_norm": 0.2063702673694027,
"learning_rate": 6.056592920012792e-05,
"loss": 0.992,
"step": 605
},
{
"epoch": 3.0997442455242967,
"grad_norm": 0.20418368098246234,
"learning_rate": 6.028806262397917e-05,
"loss": 0.967,
"step": 606
},
{
"epoch": 3.104859335038363,
"grad_norm": 0.21838960722876233,
"learning_rate": 6.001044899475588e-05,
"loss": 1.0032,
"step": 607
},
{
"epoch": 3.10997442455243,
"grad_norm": 0.16997842731362967,
"learning_rate": 5.9733091874843276e-05,
"loss": 0.9931,
"step": 608
},
{
"epoch": 3.1150895140664963,
"grad_norm": 0.17472409767100683,
"learning_rate": 5.945599482333496e-05,
"loss": 0.9905,
"step": 609
},
{
"epoch": 3.1202046035805626,
"grad_norm": 0.19017138271851242,
"learning_rate": 5.917916139598733e-05,
"loss": 0.997,
"step": 610
},
{
"epoch": 3.125319693094629,
"grad_norm": 0.20991449377873234,
"learning_rate": 5.890259514517391e-05,
"loss": 0.9424,
"step": 611
},
{
"epoch": 3.130434782608696,
"grad_norm": 0.18292524869621102,
"learning_rate": 5.8626299619839746e-05,
"loss": 0.9861,
"step": 612
},
{
"epoch": 3.135549872122762,
"grad_norm": 0.2017383191767403,
"learning_rate": 5.835027836545595e-05,
"loss": 0.9789,
"step": 613
},
{
"epoch": 3.1406649616368285,
"grad_norm": 0.19413435157237083,
"learning_rate": 5.807453492397406e-05,
"loss": 0.9979,
"step": 614
},
{
"epoch": 3.1457800511508953,
"grad_norm": 0.2084729542731396,
"learning_rate": 5.779907283378068e-05,
"loss": 0.9986,
"step": 615
},
{
"epoch": 3.1508951406649617,
"grad_norm": 0.18070671193483825,
"learning_rate": 5.752389562965216e-05,
"loss": 0.9904,
"step": 616
},
{
"epoch": 3.156010230179028,
"grad_norm": 0.2048424498193481,
"learning_rate": 5.724900684270904e-05,
"loss": 1.0189,
"step": 617
},
{
"epoch": 3.1611253196930944,
"grad_norm": 0.1791173036862784,
"learning_rate": 5.69744100003709e-05,
"loss": 0.993,
"step": 618
},
{
"epoch": 3.166240409207161,
"grad_norm": 0.175304426693732,
"learning_rate": 5.6700108626311026e-05,
"loss": 0.98,
"step": 619
},
{
"epoch": 3.1713554987212276,
"grad_norm": 0.20407301683806356,
"learning_rate": 5.6426106240411154e-05,
"loss": 0.9753,
"step": 620
},
{
"epoch": 3.176470588235294,
"grad_norm": 0.1722496184742862,
"learning_rate": 5.615240635871646e-05,
"loss": 0.9667,
"step": 621
},
{
"epoch": 3.1815856777493607,
"grad_norm": 0.17495115657326407,
"learning_rate": 5.587901249339021e-05,
"loss": 0.9717,
"step": 622
},
{
"epoch": 3.186700767263427,
"grad_norm": 0.15159682079825595,
"learning_rate": 5.560592815266894e-05,
"loss": 0.9698,
"step": 623
},
{
"epoch": 3.1918158567774935,
"grad_norm": 0.17422493946730463,
"learning_rate": 5.5333156840817196e-05,
"loss": 0.9917,
"step": 624
},
{
"epoch": 3.1969309462915603,
"grad_norm": 0.17647551731047653,
"learning_rate": 5.506070205808279e-05,
"loss": 0.9753,
"step": 625
},
{
"epoch": 3.2020460358056266,
"grad_norm": 0.16465455332731427,
"learning_rate": 5.478856730065171e-05,
"loss": 0.9877,
"step": 626
},
{
"epoch": 3.207161125319693,
"grad_norm": 0.17093246811891893,
"learning_rate": 5.4516756060603296e-05,
"loss": 0.9623,
"step": 627
},
{
"epoch": 3.21227621483376,
"grad_norm": 0.16871112747821665,
"learning_rate": 5.424527182586559e-05,
"loss": 0.9731,
"step": 628
},
{
"epoch": 3.217391304347826,
"grad_norm": 0.15991295449327614,
"learning_rate": 5.39741180801703e-05,
"loss": 0.9797,
"step": 629
},
{
"epoch": 3.2225063938618925,
"grad_norm": 0.1676223599207677,
"learning_rate": 5.370329830300833e-05,
"loss": 0.9965,
"step": 630
},
{
"epoch": 3.227621483375959,
"grad_norm": 0.1705782466914915,
"learning_rate": 5.343281596958501e-05,
"loss": 0.9719,
"step": 631
},
{
"epoch": 3.2327365728900257,
"grad_norm": 0.15413417287199227,
"learning_rate": 5.316267455077553e-05,
"loss": 0.9773,
"step": 632
},
{
"epoch": 3.237851662404092,
"grad_norm": 0.17023531257677904,
"learning_rate": 5.289287751308046e-05,
"loss": 0.9783,
"step": 633
},
{
"epoch": 3.2429667519181584,
"grad_norm": 0.15013670492365472,
"learning_rate": 5.262342831858112e-05,
"loss": 0.9604,
"step": 634
},
{
"epoch": 3.2480818414322252,
"grad_norm": 0.15777685772691763,
"learning_rate": 5.2354330424895345e-05,
"loss": 0.9745,
"step": 635
},
{
"epoch": 3.2531969309462916,
"grad_norm": 0.15975460515508463,
"learning_rate": 5.2085587285132967e-05,
"loss": 0.9852,
"step": 636
},
{
"epoch": 3.258312020460358,
"grad_norm": 0.14670011867682328,
"learning_rate": 5.181720234785148e-05,
"loss": 0.9707,
"step": 637
},
{
"epoch": 3.2634271099744243,
"grad_norm": 0.1895012451013448,
"learning_rate": 5.154917905701208e-05,
"loss": 0.9822,
"step": 638
},
{
"epoch": 3.268542199488491,
"grad_norm": 0.16100593761541052,
"learning_rate": 5.128152085193499e-05,
"loss": 0.9886,
"step": 639
},
{
"epoch": 3.2736572890025575,
"grad_norm": 0.17025980362682186,
"learning_rate": 5.101423116725581e-05,
"loss": 1.0043,
"step": 640
},
{
"epoch": 3.2787723785166243,
"grad_norm": 0.1787908815981198,
"learning_rate": 5.074731343288109e-05,
"loss": 0.9969,
"step": 641
},
{
"epoch": 3.2838874680306906,
"grad_norm": 0.1758036828408357,
"learning_rate": 5.0480771073944486e-05,
"loss": 0.9798,
"step": 642
},
{
"epoch": 3.289002557544757,
"grad_norm": 0.1846074907509873,
"learning_rate": 5.02146075107628e-05,
"loss": 0.9787,
"step": 643
},
{
"epoch": 3.2941176470588234,
"grad_norm": 0.15520603706600486,
"learning_rate": 4.994882615879203e-05,
"loss": 0.98,
"step": 644
},
{
"epoch": 3.29923273657289,
"grad_norm": 0.15960013221511546,
"learning_rate": 4.968343042858359e-05,
"loss": 0.9537,
"step": 645
},
{
"epoch": 3.3043478260869565,
"grad_norm": 0.14562090268864258,
"learning_rate": 4.941842372574053e-05,
"loss": 0.9868,
"step": 646
},
{
"epoch": 3.309462915601023,
"grad_norm": 0.14991384950017142,
"learning_rate": 4.9153809450873817e-05,
"loss": 0.9587,
"step": 647
},
{
"epoch": 3.3145780051150897,
"grad_norm": 0.15129530354741377,
"learning_rate": 4.8889590999558766e-05,
"loss": 0.9806,
"step": 648
},
{
"epoch": 3.319693094629156,
"grad_norm": 0.15107676444695042,
"learning_rate": 4.862577176229134e-05,
"loss": 0.9646,
"step": 649
},
{
"epoch": 3.3248081841432224,
"grad_norm": 0.16800107731911698,
"learning_rate": 4.8362355124444786e-05,
"loss": 1.0041,
"step": 650
},
{
"epoch": 3.329923273657289,
"grad_norm": 0.15636911731976455,
"learning_rate": 4.809934446622612e-05,
"loss": 0.9814,
"step": 651
},
{
"epoch": 3.3350383631713556,
"grad_norm": 0.15615192809689907,
"learning_rate": 4.783674316263268e-05,
"loss": 0.9742,
"step": 652
},
{
"epoch": 3.340153452685422,
"grad_norm": 0.16353049930603708,
"learning_rate": 4.757455458340902e-05,
"loss": 0.9935,
"step": 653
},
{
"epoch": 3.3452685421994883,
"grad_norm": 0.1460818311738437,
"learning_rate": 4.731278209300343e-05,
"loss": 0.9567,
"step": 654
},
{
"epoch": 3.350383631713555,
"grad_norm": 0.17401947735960155,
"learning_rate": 4.705142905052496e-05,
"loss": 0.993,
"step": 655
},
{
"epoch": 3.3554987212276215,
"grad_norm": 0.17900895951737225,
"learning_rate": 4.679049880970018e-05,
"loss": 0.9758,
"step": 656
},
{
"epoch": 3.360613810741688,
"grad_norm": 0.15979951703347306,
"learning_rate": 4.652999471883025e-05,
"loss": 0.9822,
"step": 657
},
{
"epoch": 3.3657289002557547,
"grad_norm": 0.18194476580955435,
"learning_rate": 4.626992012074787e-05,
"loss": 0.9647,
"step": 658
},
{
"epoch": 3.370843989769821,
"grad_norm": 0.15634352291536172,
"learning_rate": 4.6010278352774344e-05,
"loss": 0.9987,
"step": 659
},
{
"epoch": 3.3759590792838874,
"grad_norm": 0.17941844911904992,
"learning_rate": 4.575107274667701e-05,
"loss": 0.974,
"step": 660
},
{
"epoch": 3.381074168797954,
"grad_norm": 0.16314330503637559,
"learning_rate": 4.5492306628626126e-05,
"loss": 0.9575,
"step": 661
},
{
"epoch": 3.3861892583120206,
"grad_norm": 0.1629084188455543,
"learning_rate": 4.523398331915246e-05,
"loss": 0.9731,
"step": 662
},
{
"epoch": 3.391304347826087,
"grad_norm": 0.1771258163728386,
"learning_rate": 4.4976106133104507e-05,
"loss": 0.995,
"step": 663
},
{
"epoch": 3.3964194373401533,
"grad_norm": 0.13731181190472927,
"learning_rate": 4.471867837960611e-05,
"loss": 0.9573,
"step": 664
},
{
"epoch": 3.40153452685422,
"grad_norm": 0.19967565876780527,
"learning_rate": 4.446170336201393e-05,
"loss": 1.0048,
"step": 665
},
{
"epoch": 3.4066496163682864,
"grad_norm": 0.1500644782724652,
"learning_rate": 4.4205184377874914e-05,
"loss": 0.9872,
"step": 666
},
{
"epoch": 3.411764705882353,
"grad_norm": 0.18298053656228014,
"learning_rate": 4.394912471888424e-05,
"loss": 0.9768,
"step": 667
},
{
"epoch": 3.4168797953964196,
"grad_norm": 0.18233641990424151,
"learning_rate": 4.3693527670842943e-05,
"loss": 0.9943,
"step": 668
},
{
"epoch": 3.421994884910486,
"grad_norm": 0.17142258997448992,
"learning_rate": 4.343839651361565e-05,
"loss": 0.9631,
"step": 669
},
{
"epoch": 3.4271099744245523,
"grad_norm": 0.2068694812660131,
"learning_rate": 4.3183734521088724e-05,
"loss": 0.9994,
"step": 670
},
{
"epoch": 3.4322250639386187,
"grad_norm": 0.14564368721220933,
"learning_rate": 4.292954496112799e-05,
"loss": 0.9987,
"step": 671
},
{
"epoch": 3.4373401534526855,
"grad_norm": 0.17021085709995834,
"learning_rate": 4.2675831095537124e-05,
"loss": 0.9712,
"step": 672
},
{
"epoch": 3.442455242966752,
"grad_norm": 0.1672240638166061,
"learning_rate": 4.2422596180015445e-05,
"loss": 0.992,
"step": 673
},
{
"epoch": 3.4475703324808182,
"grad_norm": 0.14924296026603445,
"learning_rate": 4.216984346411634e-05,
"loss": 0.9803,
"step": 674
},
{
"epoch": 3.452685421994885,
"grad_norm": 0.18423800174722993,
"learning_rate": 4.1917576191205566e-05,
"loss": 1.0249,
"step": 675
},
{
"epoch": 3.4578005115089514,
"grad_norm": 0.15161100552887438,
"learning_rate": 4.166579759841959e-05,
"loss": 0.9859,
"step": 676
},
{
"epoch": 3.4629156010230178,
"grad_norm": 0.17383191362114958,
"learning_rate": 4.141451091662408e-05,
"loss": 0.9678,
"step": 677
},
{
"epoch": 3.4680306905370846,
"grad_norm": 0.17148536065945263,
"learning_rate": 4.11637193703723e-05,
"loss": 0.9785,
"step": 678
},
{
"epoch": 3.473145780051151,
"grad_norm": 0.17211711658902415,
"learning_rate": 4.091342617786396e-05,
"loss": 0.9802,
"step": 679
},
{
"epoch": 3.4782608695652173,
"grad_norm": 0.1668260684828794,
"learning_rate": 4.0663634550903826e-05,
"loss": 1.009,
"step": 680
},
{
"epoch": 3.483375959079284,
"grad_norm": 0.1372640845646432,
"learning_rate": 4.041434769486037e-05,
"loss": 0.9742,
"step": 681
},
{
"epoch": 3.4884910485933505,
"grad_norm": 0.18730966256521414,
"learning_rate": 4.016556880862487e-05,
"loss": 0.9839,
"step": 682
},
{
"epoch": 3.493606138107417,
"grad_norm": 0.1619578928694765,
"learning_rate": 3.9917301084570244e-05,
"loss": 1.0194,
"step": 683
},
{
"epoch": 3.498721227621483,
"grad_norm": 0.16048357434682578,
"learning_rate": 3.966954770850999e-05,
"loss": 0.9598,
"step": 684
},
{
"epoch": 3.50383631713555,
"grad_norm": 0.19097500487439395,
"learning_rate": 3.942231185965754e-05,
"loss": 1.001,
"step": 685
},
{
"epoch": 3.5089514066496164,
"grad_norm": 0.15734764403319154,
"learning_rate": 3.917559671058519e-05,
"loss": 1.0114,
"step": 686
},
{
"epoch": 3.5140664961636827,
"grad_norm": 0.2035274811662844,
"learning_rate": 3.892940542718364e-05,
"loss": 0.9836,
"step": 687
},
{
"epoch": 3.5191815856777495,
"grad_norm": 0.17015176173828528,
"learning_rate": 3.868374116862125e-05,
"loss": 0.9577,
"step": 688
},
{
"epoch": 3.524296675191816,
"grad_norm": 0.17994075514564073,
"learning_rate": 3.84386070873034e-05,
"loss": 0.9854,
"step": 689
},
{
"epoch": 3.5294117647058822,
"grad_norm": 0.19672729889233553,
"learning_rate": 3.819400632883226e-05,
"loss": 0.9929,
"step": 690
},
{
"epoch": 3.5345268542199486,
"grad_norm": 0.15431718642181724,
"learning_rate": 3.794994203196628e-05,
"loss": 1.0188,
"step": 691
},
{
"epoch": 3.5396419437340154,
"grad_norm": 0.168114456736758,
"learning_rate": 3.770641732857994e-05,
"loss": 0.9841,
"step": 692
},
{
"epoch": 3.544757033248082,
"grad_norm": 0.14748373779839796,
"learning_rate": 3.746343534362351e-05,
"loss": 0.9866,
"step": 693
},
{
"epoch": 3.5498721227621486,
"grad_norm": 0.16571097284750733,
"learning_rate": 3.72209991950831e-05,
"loss": 0.9777,
"step": 694
},
{
"epoch": 3.554987212276215,
"grad_norm": 0.14053074538150237,
"learning_rate": 3.6979111993940515e-05,
"loss": 1.0066,
"step": 695
},
{
"epoch": 3.5601023017902813,
"grad_norm": 0.16166313541175412,
"learning_rate": 3.673777684413333e-05,
"loss": 0.9734,
"step": 696
},
{
"epoch": 3.5652173913043477,
"grad_norm": 0.14913778377381534,
"learning_rate": 3.649699684251518e-05,
"loss": 0.9985,
"step": 697
},
{
"epoch": 3.5703324808184145,
"grad_norm": 0.1375979709766286,
"learning_rate": 3.6256775078815826e-05,
"loss": 0.9801,
"step": 698
},
{
"epoch": 3.575447570332481,
"grad_norm": 0.14848717762308264,
"learning_rate": 3.6017114635601824e-05,
"loss": 0.9813,
"step": 699
},
{
"epoch": 3.580562659846547,
"grad_norm": 0.14915296878982703,
"learning_rate": 3.57780185882366e-05,
"loss": 0.9868,
"step": 700
},
{
"epoch": 3.585677749360614,
"grad_norm": 0.16767338171047977,
"learning_rate": 3.553949000484115e-05,
"loss": 0.9778,
"step": 701
},
{
"epoch": 3.5907928388746804,
"grad_norm": 0.14034215936654634,
"learning_rate": 3.53015319462548e-05,
"loss": 0.9797,
"step": 702
},
{
"epoch": 3.5959079283887467,
"grad_norm": 0.18513480001929697,
"learning_rate": 3.5064147465995715e-05,
"loss": 0.9845,
"step": 703
},
{
"epoch": 3.601023017902813,
"grad_norm": 0.15060863329159177,
"learning_rate": 3.482733961022186e-05,
"loss": 0.9969,
"step": 704
},
{
"epoch": 3.60613810741688,
"grad_norm": 0.16008029711196534,
"learning_rate": 3.459111141769178e-05,
"loss": 0.9557,
"step": 705
},
{
"epoch": 3.6112531969309463,
"grad_norm": 0.14139469110818648,
"learning_rate": 3.4355465919725746e-05,
"loss": 0.9964,
"step": 706
},
{
"epoch": 3.6163682864450126,
"grad_norm": 0.18008083689325374,
"learning_rate": 3.4120406140166805e-05,
"loss": 0.9766,
"step": 707
},
{
"epoch": 3.6214833759590794,
"grad_norm": 0.14065787507859065,
"learning_rate": 3.388593509534189e-05,
"loss": 0.9751,
"step": 708
},
{
"epoch": 3.626598465473146,
"grad_norm": 0.1959838264971315,
"learning_rate": 3.365205579402329e-05,
"loss": 0.9655,
"step": 709
},
{
"epoch": 3.631713554987212,
"grad_norm": 0.12818286962222172,
"learning_rate": 3.341877123738982e-05,
"loss": 1.0021,
"step": 710
},
{
"epoch": 3.6368286445012785,
"grad_norm": 0.5122578200262962,
"learning_rate": 3.318608441898854e-05,
"loss": 0.9839,
"step": 711
},
{
"epoch": 3.6419437340153453,
"grad_norm": 0.15511326652130292,
"learning_rate": 3.295399832469622e-05,
"loss": 0.9753,
"step": 712
},
{
"epoch": 3.6470588235294117,
"grad_norm": 0.15230817767413712,
"learning_rate": 3.272251593268094e-05,
"loss": 0.9958,
"step": 713
},
{
"epoch": 3.6521739130434785,
"grad_norm": 0.14863844136180315,
"learning_rate": 3.249164021336407e-05,
"loss": 0.9667,
"step": 714
},
{
"epoch": 3.657289002557545,
"grad_norm": 0.15837188038589173,
"learning_rate": 3.226137412938205e-05,
"loss": 0.9762,
"step": 715
},
{
"epoch": 3.662404092071611,
"grad_norm": 0.137972231146731,
"learning_rate": 3.203172063554828e-05,
"loss": 1.0006,
"step": 716
},
{
"epoch": 3.6675191815856776,
"grad_norm": 0.14570282964804776,
"learning_rate": 3.18026826788154e-05,
"loss": 0.9582,
"step": 717
},
{
"epoch": 3.6726342710997444,
"grad_norm": 0.14367337787422657,
"learning_rate": 3.157426319823734e-05,
"loss": 0.9958,
"step": 718
},
{
"epoch": 3.6777493606138107,
"grad_norm": 0.13664360556373642,
"learning_rate": 3.1346465124931654e-05,
"loss": 1.0013,
"step": 719
},
{
"epoch": 3.682864450127877,
"grad_norm": 0.132973918091974,
"learning_rate": 3.111929138204185e-05,
"loss": 0.9652,
"step": 720
},
{
"epoch": 3.687979539641944,
"grad_norm": 0.36095258416718945,
"learning_rate": 3.08927448846999e-05,
"loss": 0.9689,
"step": 721
},
{
"epoch": 3.6930946291560103,
"grad_norm": 0.11932193774380136,
"learning_rate": 3.066682853998902e-05,
"loss": 0.964,
"step": 722
},
{
"epoch": 3.6982097186700766,
"grad_norm": 0.13955395924036143,
"learning_rate": 3.0441545246905977e-05,
"loss": 0.9803,
"step": 723
},
{
"epoch": 3.703324808184143,
"grad_norm": 0.13898112971972146,
"learning_rate": 3.0216897896324292e-05,
"loss": 0.9962,
"step": 724
},
{
"epoch": 3.70843989769821,
"grad_norm": 0.13054535182435784,
"learning_rate": 2.999288937095682e-05,
"loss": 0.9722,
"step": 725
},
{
"epoch": 3.713554987212276,
"grad_norm": 0.14497101318689892,
"learning_rate": 2.9769522545319016e-05,
"loss": 0.9684,
"step": 726
},
{
"epoch": 3.718670076726343,
"grad_norm": 0.1515086696396122,
"learning_rate": 2.9546800285691907e-05,
"loss": 0.9848,
"step": 727
},
{
"epoch": 3.7237851662404093,
"grad_norm": 0.12165415305310626,
"learning_rate": 2.9324725450085264e-05,
"loss": 0.9637,
"step": 728
},
{
"epoch": 3.7289002557544757,
"grad_norm": 0.15221342193107146,
"learning_rate": 2.9103300888201106e-05,
"loss": 0.995,
"step": 729
},
{
"epoch": 3.734015345268542,
"grad_norm": 0.12949528095065918,
"learning_rate": 2.888252944139698e-05,
"loss": 0.9936,
"step": 730
},
{
"epoch": 3.7391304347826084,
"grad_norm": 0.13806390326837292,
"learning_rate": 2.8662413942649575e-05,
"loss": 0.9579,
"step": 731
},
{
"epoch": 3.7442455242966752,
"grad_norm": 0.14578589598937078,
"learning_rate": 2.84429572165183e-05,
"loss": 0.9911,
"step": 732
},
{
"epoch": 3.7493606138107416,
"grad_norm": 0.13332385850023673,
"learning_rate": 2.822416207910906e-05,
"loss": 0.98,
"step": 733
},
{
"epoch": 3.7544757033248084,
"grad_norm": 0.13795834892848188,
"learning_rate": 2.8006031338038294e-05,
"loss": 0.9688,
"step": 734
},
{
"epoch": 3.7595907928388748,
"grad_norm": 0.1379970205043612,
"learning_rate": 2.7788567792396598e-05,
"loss": 0.9824,
"step": 735
},
{
"epoch": 3.764705882352941,
"grad_norm": 0.13245965794958156,
"learning_rate": 2.7571774232713156e-05,
"loss": 0.9898,
"step": 736
},
{
"epoch": 3.7698209718670075,
"grad_norm": 3.216146863532581,
"learning_rate": 2.735565344091965e-05,
"loss": 1.0116,
"step": 737
},
{
"epoch": 3.7749360613810743,
"grad_norm": 0.16955205361202183,
"learning_rate": 2.7140208190314788e-05,
"loss": 0.9771,
"step": 738
},
{
"epoch": 3.7800511508951407,
"grad_norm": 0.1169665628971288,
"learning_rate": 2.6925441245528624e-05,
"loss": 0.9703,
"step": 739
},
{
"epoch": 3.785166240409207,
"grad_norm": 0.14477921623852658,
"learning_rate": 2.6711355362486967e-05,
"loss": 0.9641,
"step": 740
},
{
"epoch": 3.790281329923274,
"grad_norm": 0.1320973718883581,
"learning_rate": 2.6497953288376234e-05,
"loss": 0.9765,
"step": 741
},
{
"epoch": 3.79539641943734,
"grad_norm": 0.1312668538179555,
"learning_rate": 2.6285237761608084e-05,
"loss": 1.0241,
"step": 742
},
{
"epoch": 3.8005115089514065,
"grad_norm": 0.1250506576902336,
"learning_rate": 2.607321151178419e-05,
"loss": 1.0064,
"step": 743
},
{
"epoch": 3.805626598465473,
"grad_norm": 0.12488907245170656,
"learning_rate": 2.5861877259661407e-05,
"loss": 0.9863,
"step": 744
},
{
"epoch": 3.8107416879795397,
"grad_norm": 0.11830419244600696,
"learning_rate": 2.5651237717116715e-05,
"loss": 0.9917,
"step": 745
},
{
"epoch": 3.815856777493606,
"grad_norm": 0.13708129163485394,
"learning_rate": 2.5441295587112498e-05,
"loss": 0.9752,
"step": 746
},
{
"epoch": 3.820971867007673,
"grad_norm": 0.12160381151388162,
"learning_rate": 2.5232053563661758e-05,
"loss": 0.9697,
"step": 747
},
{
"epoch": 3.8260869565217392,
"grad_norm": 0.12470170773969481,
"learning_rate": 2.5023514331793637e-05,
"loss": 1.0127,
"step": 748
},
{
"epoch": 3.8312020460358056,
"grad_norm": 0.12216238637358023,
"learning_rate": 2.481568056751897e-05,
"loss": 0.9721,
"step": 749
},
{
"epoch": 3.836317135549872,
"grad_norm": 0.12167084392403932,
"learning_rate": 2.4608554937795886e-05,
"loss": 0.9859,
"step": 750
},
{
"epoch": 3.8414322250639388,
"grad_norm": 0.15146248060779338,
"learning_rate": 2.4402140100495632e-05,
"loss": 0.9774,
"step": 751
},
{
"epoch": 3.846547314578005,
"grad_norm": 0.1321258229124532,
"learning_rate": 2.419643870436839e-05,
"loss": 0.9798,
"step": 752
},
{
"epoch": 3.8516624040920715,
"grad_norm": 0.1251068879283419,
"learning_rate": 2.3991453389009402e-05,
"loss": 0.9899,
"step": 753
},
{
"epoch": 3.8567774936061383,
"grad_norm": 0.1268521777949983,
"learning_rate": 2.378718678482504e-05,
"loss": 0.9989,
"step": 754
},
{
"epoch": 3.8618925831202047,
"grad_norm": 0.12726325379028405,
"learning_rate": 2.3583641512998997e-05,
"loss": 0.9811,
"step": 755
},
{
"epoch": 3.867007672634271,
"grad_norm": 0.12193614533361499,
"learning_rate": 2.338082018545875e-05,
"loss": 0.9633,
"step": 756
},
{
"epoch": 3.8721227621483374,
"grad_norm": 0.12067252983591115,
"learning_rate": 2.317872540484201e-05,
"loss": 0.9656,
"step": 757
},
{
"epoch": 3.877237851662404,
"grad_norm": 0.12513901653710313,
"learning_rate": 2.2977359764463304e-05,
"loss": 0.9907,
"step": 758
},
{
"epoch": 3.8823529411764706,
"grad_norm": 0.1256893578341812,
"learning_rate": 2.2776725848280688e-05,
"loss": 1.0123,
"step": 759
},
{
"epoch": 3.887468030690537,
"grad_norm": 0.11430778894464487,
"learning_rate": 2.2576826230862565e-05,
"loss": 1.0062,
"step": 760
},
{
"epoch": 3.8925831202046037,
"grad_norm": 0.12097149256274506,
"learning_rate": 2.2377663477354846e-05,
"loss": 0.9403,
"step": 761
},
{
"epoch": 3.89769820971867,
"grad_norm": 0.124502270810866,
"learning_rate": 2.217924014344772e-05,
"loss": 0.9793,
"step": 762
},
{
"epoch": 3.9028132992327365,
"grad_norm": 0.11763618100617718,
"learning_rate": 2.198155877534311e-05,
"loss": 1.0038,
"step": 763
},
{
"epoch": 3.907928388746803,
"grad_norm": 0.11641223364110288,
"learning_rate": 2.1784621909721838e-05,
"loss": 0.9839,
"step": 764
},
{
"epoch": 3.9130434782608696,
"grad_norm": 0.11065271657834529,
"learning_rate": 2.1588432073711215e-05,
"loss": 0.9613,
"step": 765
},
{
"epoch": 3.918158567774936,
"grad_norm": 0.11720974573827699,
"learning_rate": 2.1392991784852518e-05,
"loss": 0.9934,
"step": 766
},
{
"epoch": 3.923273657289003,
"grad_norm": 0.10378602749677512,
"learning_rate": 2.1198303551068663e-05,
"loss": 0.9849,
"step": 767
},
{
"epoch": 3.928388746803069,
"grad_norm": 0.11073948338502751,
"learning_rate": 2.100436987063213e-05,
"loss": 0.9836,
"step": 768
},
{
"epoch": 3.9335038363171355,
"grad_norm": 0.1080108854502703,
"learning_rate": 2.0811193232132824e-05,
"loss": 0.977,
"step": 769
},
{
"epoch": 3.938618925831202,
"grad_norm": 0.09801026976054619,
"learning_rate": 2.0618776114446126e-05,
"loss": 0.9683,
"step": 770
},
{
"epoch": 3.9437340153452687,
"grad_norm": 0.12038757535371046,
"learning_rate": 2.0427120986701173e-05,
"loss": 0.9593,
"step": 771
},
{
"epoch": 3.948849104859335,
"grad_norm": 0.11456427512297371,
"learning_rate": 2.0236230308249025e-05,
"loss": 0.9795,
"step": 772
},
{
"epoch": 3.9539641943734014,
"grad_norm": 0.10618965347944075,
"learning_rate": 2.004610652863135e-05,
"loss": 0.9788,
"step": 773
},
{
"epoch": 3.959079283887468,
"grad_norm": 0.12283547115037487,
"learning_rate": 1.985675208754871e-05,
"loss": 0.9667,
"step": 774
},
{
"epoch": 3.9641943734015346,
"grad_norm": 0.10722031562677722,
"learning_rate": 1.9668169414829403e-05,
"loss": 1.0291,
"step": 775
},
{
"epoch": 3.969309462915601,
"grad_norm": 0.11255749532177638,
"learning_rate": 1.9480360930398304e-05,
"loss": 0.987,
"step": 776
},
{
"epoch": 3.9744245524296673,
"grad_norm": 0.11147499924790798,
"learning_rate": 1.929332904424577e-05,
"loss": 0.985,
"step": 777
},
{
"epoch": 3.979539641943734,
"grad_norm": 0.10389011868423567,
"learning_rate": 1.9107076156396705e-05,
"loss": 1.001,
"step": 778
},
{
"epoch": 3.9846547314578005,
"grad_norm": 0.1171327392852703,
"learning_rate": 1.8921604656879742e-05,
"loss": 0.9858,
"step": 779
},
{
"epoch": 3.9897698209718673,
"grad_norm": 0.11314714000913269,
"learning_rate": 1.873691692569665e-05,
"loss": 0.9826,
"step": 780
},
{
"epoch": 3.9948849104859336,
"grad_norm": 0.10973505326358149,
"learning_rate": 1.8553015332791772e-05,
"loss": 1.0062,
"step": 781
},
{
"epoch": 4.0,
"grad_norm": 0.19830487981983497,
"learning_rate": 1.8369902238021497e-05,
"loss": 1.4739,
"step": 782
},
{
"epoch": 4.005115089514066,
"grad_norm": 0.11616653497917728,
"learning_rate": 1.8187579991124166e-05,
"loss": 0.963,
"step": 783
},
{
"epoch": 4.010230179028133,
"grad_norm": 0.15277895417241139,
"learning_rate": 1.80060509316898e-05,
"loss": 0.9534,
"step": 784
},
{
"epoch": 4.015345268542199,
"grad_norm": 0.13631336060650912,
"learning_rate": 1.782531738913006e-05,
"loss": 0.9765,
"step": 785
},
{
"epoch": 4.020460358056266,
"grad_norm": 0.13037547872846397,
"learning_rate": 1.7645381682648474e-05,
"loss": 0.9657,
"step": 786
},
{
"epoch": 4.025575447570333,
"grad_norm": 0.13126122893844305,
"learning_rate": 1.7466246121210525e-05,
"loss": 0.9585,
"step": 787
},
{
"epoch": 4.030690537084399,
"grad_norm": 0.12590197949144366,
"learning_rate": 1.7287913003514143e-05,
"loss": 0.9698,
"step": 788
},
{
"epoch": 4.035805626598465,
"grad_norm": 0.13056581907239462,
"learning_rate": 1.711038461796019e-05,
"loss": 0.9591,
"step": 789
},
{
"epoch": 4.040920716112532,
"grad_norm": 0.12269534388195832,
"learning_rate": 1.693366324262299e-05,
"loss": 0.9837,
"step": 790
},
{
"epoch": 4.046035805626598,
"grad_norm": 0.1310163348216626,
"learning_rate": 1.675775114522125e-05,
"loss": 0.96,
"step": 791
},
{
"epoch": 4.051150895140665,
"grad_norm": 0.12447472625743548,
"learning_rate": 1.658265058308884e-05,
"loss": 0.9537,
"step": 792
},
{
"epoch": 4.056265984654732,
"grad_norm": 0.11548954475540298,
"learning_rate": 1.6408363803145916e-05,
"loss": 0.9817,
"step": 793
},
{
"epoch": 4.061381074168798,
"grad_norm": 0.10914046862924401,
"learning_rate": 1.6234893041869977e-05,
"loss": 0.942,
"step": 794
},
{
"epoch": 4.0664961636828645,
"grad_norm": 0.12534956939361333,
"learning_rate": 1.6062240525267282e-05,
"loss": 0.9764,
"step": 795
},
{
"epoch": 4.071611253196931,
"grad_norm": 0.11626368570662485,
"learning_rate": 1.5890408468844247e-05,
"loss": 0.9922,
"step": 796
},
{
"epoch": 4.076726342710997,
"grad_norm": 0.10699708302716189,
"learning_rate": 1.5719399077578932e-05,
"loss": 0.9981,
"step": 797
},
{
"epoch": 4.081841432225064,
"grad_norm": 0.11112817596289239,
"learning_rate": 1.554921454589293e-05,
"loss": 0.9731,
"step": 798
},
{
"epoch": 4.086956521739131,
"grad_norm": 0.11328621420104991,
"learning_rate": 1.537985705762294e-05,
"loss": 0.9559,
"step": 799
},
{
"epoch": 4.092071611253197,
"grad_norm": 0.11746598830065647,
"learning_rate": 1.5211328785993088e-05,
"loss": 0.9615,
"step": 800
},
{
"epoch": 4.0971867007672635,
"grad_norm": 0.09651748717766548,
"learning_rate": 1.5043631893586707e-05,
"loss": 0.9638,
"step": 801
},
{
"epoch": 4.10230179028133,
"grad_norm": 0.1214177512550134,
"learning_rate": 1.4876768532318768e-05,
"loss": 0.9656,
"step": 802
},
{
"epoch": 4.107416879795396,
"grad_norm": 0.1012144891018282,
"learning_rate": 1.4710740843408248e-05,
"loss": 0.9735,
"step": 803
},
{
"epoch": 4.112531969309463,
"grad_norm": 0.09990576165665672,
"learning_rate": 1.4545550957350635e-05,
"loss": 0.956,
"step": 804
},
{
"epoch": 4.117647058823529,
"grad_norm": 0.1195577503786533,
"learning_rate": 1.4381200993890593e-05,
"loss": 0.9751,
"step": 805
},
{
"epoch": 4.122762148337596,
"grad_norm": 0.10584701347111745,
"learning_rate": 1.4217693061994697e-05,
"loss": 0.9607,
"step": 806
},
{
"epoch": 4.127877237851663,
"grad_norm": 0.11811713034338012,
"learning_rate": 1.4055029259824512e-05,
"loss": 0.9756,
"step": 807
},
{
"epoch": 4.132992327365729,
"grad_norm": 0.11133570038229246,
"learning_rate": 1.3893211674709562e-05,
"loss": 0.9717,
"step": 808
},
{
"epoch": 4.138107416879795,
"grad_norm": 0.10750621531215204,
"learning_rate": 1.3732242383120524e-05,
"loss": 0.965,
"step": 809
},
{
"epoch": 4.143222506393862,
"grad_norm": 0.09963142743312713,
"learning_rate": 1.3572123450642693e-05,
"loss": 0.9658,
"step": 810
},
{
"epoch": 4.148337595907928,
"grad_norm": 0.13474038037677824,
"learning_rate": 1.3412856931949345e-05,
"loss": 0.9749,
"step": 811
},
{
"epoch": 4.153452685421995,
"grad_norm": 0.10191264255745801,
"learning_rate": 1.325444487077551e-05,
"loss": 0.9715,
"step": 812
},
{
"epoch": 4.158567774936062,
"grad_norm": 0.09831207587627867,
"learning_rate": 1.3096889299891635e-05,
"loss": 0.9714,
"step": 813
},
{
"epoch": 4.163682864450128,
"grad_norm": 0.11453013377243791,
"learning_rate": 1.2940192241077525e-05,
"loss": 0.9763,
"step": 814
},
{
"epoch": 4.168797953964194,
"grad_norm": 0.10046839458939905,
"learning_rate": 1.2784355705096432e-05,
"loss": 0.981,
"step": 815
},
{
"epoch": 4.173913043478261,
"grad_norm": 0.09714449670617001,
"learning_rate": 1.2629381691669269e-05,
"loss": 0.9827,
"step": 816
},
{
"epoch": 4.179028132992327,
"grad_norm": 0.10070672573661457,
"learning_rate": 1.2475272189448818e-05,
"loss": 0.9659,
"step": 817
},
{
"epoch": 4.1841432225063935,
"grad_norm": 0.09591739352603393,
"learning_rate": 1.2322029175994393e-05,
"loss": 0.9955,
"step": 818
},
{
"epoch": 4.189258312020461,
"grad_norm": 0.1049048046610583,
"learning_rate": 1.2169654617746338e-05,
"loss": 0.9738,
"step": 819
},
{
"epoch": 4.194373401534527,
"grad_norm": 0.09283193710402415,
"learning_rate": 1.2018150470000874e-05,
"loss": 0.9617,
"step": 820
},
{
"epoch": 4.1994884910485935,
"grad_norm": 0.09291829854983867,
"learning_rate": 1.1867518676884901e-05,
"loss": 0.9608,
"step": 821
},
{
"epoch": 4.20460358056266,
"grad_norm": 0.097503759394023,
"learning_rate": 1.1717761171331126e-05,
"loss": 0.9623,
"step": 822
},
{
"epoch": 4.209718670076726,
"grad_norm": 0.10131361216152728,
"learning_rate": 1.1568879875053347e-05,
"loss": 0.9618,
"step": 823
},
{
"epoch": 4.2148337595907925,
"grad_norm": 0.09685571836703365,
"learning_rate": 1.1420876698521562e-05,
"loss": 0.9592,
"step": 824
},
{
"epoch": 4.21994884910486,
"grad_norm": 0.11054448273687466,
"learning_rate": 1.1273753540937685e-05,
"loss": 0.9515,
"step": 825
},
{
"epoch": 4.225063938618926,
"grad_norm": 0.10163114230776238,
"learning_rate": 1.1127512290211002e-05,
"loss": 0.9698,
"step": 826
},
{
"epoch": 4.2301790281329925,
"grad_norm": 0.09358052612607201,
"learning_rate": 1.0982154822934076e-05,
"loss": 0.971,
"step": 827
},
{
"epoch": 4.235294117647059,
"grad_norm": 0.11554962294172144,
"learning_rate": 1.083768300435863e-05,
"loss": 0.9832,
"step": 828
},
{
"epoch": 4.240409207161125,
"grad_norm": 0.11068234980519318,
"learning_rate": 1.0694098688371492e-05,
"loss": 0.9734,
"step": 829
},
{
"epoch": 4.245524296675192,
"grad_norm": 0.10022383801325921,
"learning_rate": 1.0551403717471013e-05,
"loss": 0.9743,
"step": 830
},
{
"epoch": 4.250639386189258,
"grad_norm": 0.10227963784373037,
"learning_rate": 1.040959992274326e-05,
"loss": 0.9812,
"step": 831
},
{
"epoch": 4.255754475703325,
"grad_norm": 0.10250069997045962,
"learning_rate": 1.0268689123838613e-05,
"loss": 0.9589,
"step": 832
},
{
"epoch": 4.260869565217392,
"grad_norm": 0.09870324404565264,
"learning_rate": 1.0128673128948336e-05,
"loss": 0.9634,
"step": 833
},
{
"epoch": 4.265984654731458,
"grad_norm": 0.09924319017501951,
"learning_rate": 9.989553734781405e-06,
"loss": 0.9498,
"step": 834
},
{
"epoch": 4.271099744245524,
"grad_norm": 0.10540784765184293,
"learning_rate": 9.851332726541564e-06,
"loss": 0.9733,
"step": 835
},
{
"epoch": 4.276214833759591,
"grad_norm": 0.11352756005239198,
"learning_rate": 9.71401187790419e-06,
"loss": 0.9608,
"step": 836
},
{
"epoch": 4.281329923273657,
"grad_norm": 0.10509362314164922,
"learning_rate": 9.577592950993745e-06,
"loss": 0.9408,
"step": 837
},
{
"epoch": 4.286445012787723,
"grad_norm": 0.09023255808041263,
"learning_rate": 9.44207769636103e-06,
"loss": 0.9677,
"step": 838
},
{
"epoch": 4.291560102301791,
"grad_norm": 0.10065024730730968,
"learning_rate": 9.307467852960799e-06,
"loss": 0.9784,
"step": 839
},
{
"epoch": 4.296675191815857,
"grad_norm": 0.10290767513184404,
"learning_rate": 9.173765148129416e-06,
"loss": 0.9737,
"step": 840
},
{
"epoch": 4.301790281329923,
"grad_norm": 0.09617538927014685,
"learning_rate": 9.040971297562655e-06,
"loss": 0.9785,
"step": 841
},
{
"epoch": 4.30690537084399,
"grad_norm": 0.0947242418806391,
"learning_rate": 8.909088005293749e-06,
"loss": 0.9918,
"step": 842
},
{
"epoch": 4.312020460358056,
"grad_norm": 0.09234986153646492,
"learning_rate": 8.778116963671502e-06,
"loss": 0.9685,
"step": 843
},
{
"epoch": 4.3171355498721224,
"grad_norm": 0.09275374402110674,
"learning_rate": 8.648059853338524e-06,
"loss": 0.9462,
"step": 844
},
{
"epoch": 4.322250639386189,
"grad_norm": 0.09708880926006762,
"learning_rate": 8.518918343209725e-06,
"loss": 0.9825,
"step": 845
},
{
"epoch": 4.327365728900256,
"grad_norm": 0.09058592523348291,
"learning_rate": 8.390694090450889e-06,
"loss": 0.9614,
"step": 846
},
{
"epoch": 4.332480818414322,
"grad_norm": 0.18799766380452354,
"learning_rate": 8.263388740457405e-06,
"loss": 0.9739,
"step": 847
},
{
"epoch": 4.337595907928389,
"grad_norm": 0.09510070455955731,
"learning_rate": 8.137003926833099e-06,
"loss": 0.9738,
"step": 848
},
{
"epoch": 4.342710997442455,
"grad_norm": 0.10287662253560259,
"learning_rate": 8.011541271369343e-06,
"loss": 0.9523,
"step": 849
},
{
"epoch": 4.3478260869565215,
"grad_norm": 0.09595568952373368,
"learning_rate": 7.887002384024236e-06,
"loss": 0.973,
"step": 850
},
{
"epoch": 4.352941176470588,
"grad_norm": 0.0859996090122486,
"learning_rate": 7.763388862901906e-06,
"loss": 0.9595,
"step": 851
},
{
"epoch": 4.358056265984655,
"grad_norm": 0.08656712958397209,
"learning_rate": 7.640702294232051e-06,
"loss": 0.977,
"step": 852
},
{
"epoch": 4.3631713554987215,
"grad_norm": 0.09563739575689091,
"learning_rate": 7.518944252349522e-06,
"loss": 0.9773,
"step": 853
},
{
"epoch": 4.368286445012788,
"grad_norm": 0.08811063326460167,
"learning_rate": 7.398116299674174e-06,
"loss": 0.9659,
"step": 854
},
{
"epoch": 4.373401534526854,
"grad_norm": 0.08936621498489415,
"learning_rate": 7.278219986690822e-06,
"loss": 0.9808,
"step": 855
},
{
"epoch": 4.378516624040921,
"grad_norm": 0.08480759324528003,
"learning_rate": 7.1592568519292685e-06,
"loss": 0.995,
"step": 856
},
{
"epoch": 4.383631713554987,
"grad_norm": 0.09425882010067797,
"learning_rate": 7.041228421944661e-06,
"loss": 0.97,
"step": 857
},
{
"epoch": 4.388746803069053,
"grad_norm": 0.09049164262625532,
"learning_rate": 6.924136211297843e-06,
"loss": 0.9735,
"step": 858
},
{
"epoch": 4.3938618925831205,
"grad_norm": 0.09432852817748705,
"learning_rate": 6.807981722535948e-06,
"loss": 0.9747,
"step": 859
},
{
"epoch": 4.398976982097187,
"grad_norm": 0.08257744048556204,
"learning_rate": 6.692766446173071e-06,
"loss": 0.9694,
"step": 860
},
{
"epoch": 4.404092071611253,
"grad_norm": 0.08806501756544037,
"learning_rate": 6.578491860671161e-06,
"loss": 0.9915,
"step": 861
},
{
"epoch": 4.40920716112532,
"grad_norm": 0.08791530570750992,
"learning_rate": 6.465159432421146e-06,
"loss": 0.9728,
"step": 862
},
{
"epoch": 4.414322250639386,
"grad_norm": 0.08625144926305722,
"learning_rate": 6.352770615723938e-06,
"loss": 0.9695,
"step": 863
},
{
"epoch": 4.419437340153452,
"grad_norm": 0.08918874201887288,
"learning_rate": 6.241326852771936e-06,
"loss": 0.9602,
"step": 864
},
{
"epoch": 4.42455242966752,
"grad_norm": 0.08881643339862394,
"learning_rate": 6.130829573630382e-06,
"loss": 0.9856,
"step": 865
},
{
"epoch": 4.429667519181586,
"grad_norm": 0.08855426391586113,
"learning_rate": 6.021280196219135e-06,
"loss": 0.9483,
"step": 866
},
{
"epoch": 4.434782608695652,
"grad_norm": 0.0839058511965793,
"learning_rate": 5.912680126294401e-06,
"loss": 0.968,
"step": 867
},
{
"epoch": 4.439897698209719,
"grad_norm": 0.0820171015241334,
"learning_rate": 5.80503075743069e-06,
"loss": 0.991,
"step": 868
},
{
"epoch": 4.445012787723785,
"grad_norm": 0.08439945432430783,
"learning_rate": 5.698333471002979e-06,
"loss": 0.9625,
"step": 869
},
{
"epoch": 4.450127877237851,
"grad_norm": 0.08365696475232975,
"learning_rate": 5.592589636168963e-06,
"loss": 0.9611,
"step": 870
},
{
"epoch": 4.455242966751918,
"grad_norm": 0.08184968767415474,
"learning_rate": 5.487800609851457e-06,
"loss": 0.9796,
"step": 871
},
{
"epoch": 4.460358056265985,
"grad_norm": 0.08210234669126502,
"learning_rate": 5.383967736721047e-06,
"loss": 0.9623,
"step": 872
},
{
"epoch": 4.465473145780051,
"grad_norm": 0.0802871627505241,
"learning_rate": 5.281092349178751e-06,
"loss": 0.9748,
"step": 873
},
{
"epoch": 4.470588235294118,
"grad_norm": 0.08204680683957348,
"learning_rate": 5.1791757673390395e-06,
"loss": 0.9668,
"step": 874
},
{
"epoch": 4.475703324808184,
"grad_norm": 0.08431340131792082,
"learning_rate": 5.078219299012768e-06,
"loss": 0.9855,
"step": 875
},
{
"epoch": 4.4808184143222505,
"grad_norm": 0.0797907578774786,
"learning_rate": 4.978224239690459e-06,
"loss": 0.9679,
"step": 876
},
{
"epoch": 4.485933503836317,
"grad_norm": 0.08306039121373762,
"learning_rate": 4.879191872525688e-06,
"loss": 0.9893,
"step": 877
},
{
"epoch": 4.491048593350383,
"grad_norm": 0.07917647536758707,
"learning_rate": 4.7811234683186045e-06,
"loss": 0.9704,
"step": 878
},
{
"epoch": 4.4961636828644505,
"grad_norm": 0.0807065804302448,
"learning_rate": 4.684020285499627e-06,
"loss": 0.9661,
"step": 879
},
{
"epoch": 4.501278772378517,
"grad_norm": 0.09035035909083011,
"learning_rate": 4.587883570113238e-06,
"loss": 0.9658,
"step": 880
},
{
"epoch": 4.506393861892583,
"grad_norm": 0.08298798682382963,
"learning_rate": 4.492714555802104e-06,
"loss": 0.9889,
"step": 881
},
{
"epoch": 4.5115089514066495,
"grad_norm": 0.08471598730967986,
"learning_rate": 4.398514463791177e-06,
"loss": 0.9679,
"step": 882
},
{
"epoch": 4.516624040920716,
"grad_norm": 0.07914286941445323,
"learning_rate": 4.305284502872003e-06,
"loss": 0.9654,
"step": 883
},
{
"epoch": 4.521739130434782,
"grad_norm": 0.083989141142803,
"learning_rate": 4.213025869387268e-06,
"loss": 0.9895,
"step": 884
},
{
"epoch": 4.526854219948849,
"grad_norm": 0.08827292350699108,
"learning_rate": 4.121739747215428e-06,
"loss": 0.9594,
"step": 885
},
{
"epoch": 4.531969309462916,
"grad_norm": 0.08779622646413142,
"learning_rate": 4.031427307755466e-06,
"loss": 0.9676,
"step": 886
},
{
"epoch": 4.537084398976982,
"grad_norm": 0.08480663829610874,
"learning_rate": 3.942089709911958e-06,
"loss": 0.9572,
"step": 887
},
{
"epoch": 4.542199488491049,
"grad_norm": 0.08079594054352263,
"learning_rate": 3.853728100080085e-06,
"loss": 0.9553,
"step": 888
},
{
"epoch": 4.547314578005115,
"grad_norm": 0.08303712831781242,
"learning_rate": 3.766343612131067e-06,
"loss": 0.9807,
"step": 889
},
{
"epoch": 4.552429667519181,
"grad_norm": 0.07950916145013891,
"learning_rate": 3.67993736739745e-06,
"loss": 0.9659,
"step": 890
},
{
"epoch": 4.557544757033249,
"grad_norm": 0.07990116254069944,
"learning_rate": 3.5945104746588368e-06,
"loss": 0.9801,
"step": 891
},
{
"epoch": 4.562659846547315,
"grad_norm": 0.08494691861889211,
"learning_rate": 3.5100640301276003e-06,
"loss": 0.9823,
"step": 892
},
{
"epoch": 4.567774936061381,
"grad_norm": 0.08687436970325295,
"learning_rate": 3.4265991174348633e-06,
"loss": 0.9502,
"step": 893
},
{
"epoch": 4.572890025575448,
"grad_norm": 0.08783549738630972,
"learning_rate": 3.344116807616562e-06,
"loss": 0.9583,
"step": 894
},
{
"epoch": 4.578005115089514,
"grad_norm": 0.08639468624179081,
"learning_rate": 3.2626181590996507e-06,
"loss": 0.9895,
"step": 895
},
{
"epoch": 4.58312020460358,
"grad_norm": 0.08133146999758012,
"learning_rate": 3.1821042176886483e-06,
"loss": 0.9692,
"step": 896
},
{
"epoch": 4.588235294117647,
"grad_norm": 0.07799889788482668,
"learning_rate": 3.102576016552119e-06,
"loss": 0.9669,
"step": 897
},
{
"epoch": 4.593350383631714,
"grad_norm": 0.07963888148355315,
"learning_rate": 3.0240345762094114e-06,
"loss": 0.96,
"step": 898
},
{
"epoch": 4.59846547314578,
"grad_norm": 0.0873175266502377,
"learning_rate": 2.946480904517639e-06,
"loss": 0.9618,
"step": 899
},
{
"epoch": 4.603580562659847,
"grad_norm": 0.07985201563600253,
"learning_rate": 2.869915996658641e-06,
"loss": 0.9732,
"step": 900
},
{
"epoch": 4.608695652173913,
"grad_norm": 0.0807698076031068,
"learning_rate": 2.7943408351263613e-06,
"loss": 0.9451,
"step": 901
},
{
"epoch": 4.6138107416879794,
"grad_norm": 0.07892811163095617,
"learning_rate": 2.7197563897140677e-06,
"loss": 0.9485,
"step": 902
},
{
"epoch": 4.618925831202046,
"grad_norm": 0.08593035915604227,
"learning_rate": 2.6461636175020155e-06,
"loss": 0.9573,
"step": 903
},
{
"epoch": 4.624040920716112,
"grad_norm": 0.07772596675848412,
"learning_rate": 2.5735634628451543e-06,
"loss": 0.9902,
"step": 904
},
{
"epoch": 4.629156010230179,
"grad_norm": 0.08924244768639207,
"learning_rate": 2.5019568573609877e-06,
"loss": 0.9761,
"step": 905
},
{
"epoch": 4.634271099744246,
"grad_norm": 0.0780396889441693,
"learning_rate": 2.4313447199176344e-06,
"loss": 0.981,
"step": 906
},
{
"epoch": 4.639386189258312,
"grad_norm": 0.07560972029315807,
"learning_rate": 2.3617279566220084e-06,
"loss": 0.9689,
"step": 907
},
{
"epoch": 4.6445012787723785,
"grad_norm": 0.08169326714820896,
"learning_rate": 2.293107460808246e-06,
"loss": 0.9938,
"step": 908
},
{
"epoch": 4.649616368286445,
"grad_norm": 0.08112758121761164,
"learning_rate": 2.2254841130262017e-06,
"loss": 0.985,
"step": 909
},
{
"epoch": 4.654731457800511,
"grad_norm": 0.0774355250755235,
"learning_rate": 2.1588587810301265e-06,
"loss": 0.9508,
"step": 910
},
{
"epoch": 4.659846547314578,
"grad_norm": 0.0783097671672101,
"learning_rate": 2.0932323197675996e-06,
"loss": 0.9725,
"step": 911
},
{
"epoch": 4.664961636828645,
"grad_norm": 0.07486265565311617,
"learning_rate": 2.028605571368507e-06,
"loss": 0.9524,
"step": 912
},
{
"epoch": 4.670076726342711,
"grad_norm": 0.07660204929906744,
"learning_rate": 1.96497936513425e-06,
"loss": 0.9595,
"step": 913
},
{
"epoch": 4.675191815856778,
"grad_norm": 0.07487200606734074,
"learning_rate": 1.9023545175271118e-06,
"loss": 0.9557,
"step": 914
},
{
"epoch": 4.680306905370844,
"grad_norm": 0.08373911344578278,
"learning_rate": 1.8407318321597456e-06,
"loss": 0.9719,
"step": 915
},
{
"epoch": 4.68542199488491,
"grad_norm": 0.08229171455758384,
"learning_rate": 1.780112099784921e-06,
"loss": 0.9786,
"step": 916
},
{
"epoch": 4.690537084398977,
"grad_norm": 0.07552750889444408,
"learning_rate": 1.720496098285338e-06,
"loss": 0.9881,
"step": 917
},
{
"epoch": 4.695652173913043,
"grad_norm": 0.08111232611724761,
"learning_rate": 1.6618845926636274e-06,
"loss": 0.9858,
"step": 918
},
{
"epoch": 4.70076726342711,
"grad_norm": 0.0751305874093556,
"learning_rate": 1.6042783350325785e-06,
"loss": 0.9571,
"step": 919
},
{
"epoch": 4.705882352941177,
"grad_norm": 0.07727030303219494,
"learning_rate": 1.5476780646054779e-06,
"loss": 0.9682,
"step": 920
},
{
"epoch": 4.710997442455243,
"grad_norm": 0.07757885991942609,
"learning_rate": 1.4920845076865953e-06,
"loss": 0.9569,
"step": 921
},
{
"epoch": 4.716112531969309,
"grad_norm": 0.07457519345270389,
"learning_rate": 1.4374983776618768e-06,
"loss": 0.9692,
"step": 922
},
{
"epoch": 4.721227621483376,
"grad_norm": 0.07806209448258153,
"learning_rate": 1.3839203749898133e-06,
"loss": 0.9699,
"step": 923
},
{
"epoch": 4.726342710997442,
"grad_norm": 0.07651285936775651,
"learning_rate": 1.3313511871924357e-06,
"loss": 0.9836,
"step": 924
},
{
"epoch": 4.731457800511509,
"grad_norm": 0.07547635355971415,
"learning_rate": 1.279791488846467e-06,
"loss": 0.9808,
"step": 925
},
{
"epoch": 4.736572890025576,
"grad_norm": 0.0767435027207511,
"learning_rate": 1.229241941574717e-06,
"loss": 0.9645,
"step": 926
},
{
"epoch": 4.741687979539642,
"grad_norm": 0.07605938895578472,
"learning_rate": 1.179703194037538e-06,
"loss": 0.9963,
"step": 927
},
{
"epoch": 4.746803069053708,
"grad_norm": 0.0754173818204729,
"learning_rate": 1.1311758819245732e-06,
"loss": 0.9552,
"step": 928
},
{
"epoch": 4.751918158567775,
"grad_norm": 0.07458945556127608,
"learning_rate": 1.0836606279465145e-06,
"loss": 0.973,
"step": 929
},
{
"epoch": 4.757033248081841,
"grad_norm": 0.0753169058732559,
"learning_rate": 1.0371580418271443e-06,
"loss": 0.9712,
"step": 930
},
{
"epoch": 4.762148337595908,
"grad_norm": 0.07811363964701583,
"learning_rate": 9.916687202955643e-07,
"loss": 0.9756,
"step": 931
},
{
"epoch": 4.767263427109975,
"grad_norm": 0.07417154658551314,
"learning_rate": 9.471932470784595e-07,
"loss": 0.9896,
"step": 932
},
{
"epoch": 4.772378516624041,
"grad_norm": 0.07612039964436539,
"learning_rate": 9.037321928926546e-07,
"loss": 0.9718,
"step": 933
},
{
"epoch": 4.7774936061381075,
"grad_norm": 0.07615126747924418,
"learning_rate": 8.612861154377517e-07,
"loss": 0.96,
"step": 934
},
{
"epoch": 4.782608695652174,
"grad_norm": 0.07488030342345021,
"learning_rate": 8.198555593890423e-07,
"loss": 0.9701,
"step": 935
},
{
"epoch": 4.78772378516624,
"grad_norm": 0.07450161516562323,
"learning_rate": 7.794410563904465e-07,
"loss": 0.9832,
"step": 936
},
{
"epoch": 4.792838874680307,
"grad_norm": 0.07694219467907164,
"learning_rate": 7.400431250477269e-07,
"loss": 0.9482,
"step": 937
},
{
"epoch": 4.797953964194374,
"grad_norm": 0.07728359019431556,
"learning_rate": 7.016622709218102e-07,
"loss": 0.958,
"step": 938
},
{
"epoch": 4.80306905370844,
"grad_norm": 0.0772171077931656,
"learning_rate": 6.642989865223471e-07,
"loss": 0.9755,
"step": 939
},
{
"epoch": 4.8081841432225065,
"grad_norm": 0.07371810004441909,
"learning_rate": 6.279537513013268e-07,
"loss": 0.9518,
"step": 940
},
{
"epoch": 4.813299232736573,
"grad_norm": 0.079983562262525,
"learning_rate": 5.926270316470017e-07,
"loss": 0.9644,
"step": 941
},
{
"epoch": 4.818414322250639,
"grad_norm": 0.07392991977596462,
"learning_rate": 5.583192808778037e-07,
"loss": 0.9522,
"step": 942
},
{
"epoch": 4.823529411764706,
"grad_norm": 0.07380800604090382,
"learning_rate": 5.250309392366237e-07,
"loss": 0.9689,
"step": 943
},
{
"epoch": 4.828644501278772,
"grad_norm": 0.07306370523077808,
"learning_rate": 4.927624338851189e-07,
"loss": 0.9767,
"step": 944
},
{
"epoch": 4.833759590792839,
"grad_norm": 0.07486655104269185,
"learning_rate": 4.615141788982147e-07,
"loss": 0.9524,
"step": 945
},
{
"epoch": 4.838874680306906,
"grad_norm": 0.07533207321369925,
"learning_rate": 4.3128657525882024e-07,
"loss": 0.9599,
"step": 946
},
{
"epoch": 4.843989769820972,
"grad_norm": 0.07414133822569055,
"learning_rate": 4.0208001085268567e-07,
"loss": 0.9468,
"step": 947
},
{
"epoch": 4.849104859335038,
"grad_norm": 0.08044922294041608,
"learning_rate": 3.7389486046340185e-07,
"loss": 0.9768,
"step": 948
},
{
"epoch": 4.854219948849105,
"grad_norm": 0.07410782395709221,
"learning_rate": 3.467314857675952e-07,
"loss": 0.9689,
"step": 949
},
{
"epoch": 4.859335038363171,
"grad_norm": 0.07573674261063944,
"learning_rate": 3.2059023533030033e-07,
"loss": 0.9803,
"step": 950
},
{
"epoch": 4.864450127877237,
"grad_norm": 0.07388156420140238,
"learning_rate": 2.954714446004747e-07,
"loss": 0.9617,
"step": 951
},
{
"epoch": 4.869565217391305,
"grad_norm": 0.07512578680798108,
"learning_rate": 2.713754359067089e-07,
"loss": 0.9808,
"step": 952
},
{
"epoch": 4.874680306905371,
"grad_norm": 0.07479347896842965,
"learning_rate": 2.483025184530785e-07,
"loss": 0.9745,
"step": 953
},
{
"epoch": 4.879795396419437,
"grad_norm": 0.07390137699328536,
"learning_rate": 2.2625298831517428e-07,
"loss": 0.9858,
"step": 954
},
{
"epoch": 4.884910485933504,
"grad_norm": 0.07642394694193964,
"learning_rate": 2.0522712843630056e-07,
"loss": 0.9829,
"step": 955
},
{
"epoch": 4.89002557544757,
"grad_norm": 0.07703256064419331,
"learning_rate": 1.8522520862386929e-07,
"loss": 0.9658,
"step": 956
},
{
"epoch": 4.8951406649616365,
"grad_norm": 0.07233300328021432,
"learning_rate": 1.6624748554590064e-07,
"loss": 0.9758,
"step": 957
},
{
"epoch": 4.900255754475703,
"grad_norm": 0.07197673456343069,
"learning_rate": 1.482942027277634e-07,
"loss": 0.9818,
"step": 958
},
{
"epoch": 4.90537084398977,
"grad_norm": 0.0753125286919161,
"learning_rate": 1.3136559054903075e-07,
"loss": 0.9384,
"step": 959
},
{
"epoch": 4.910485933503836,
"grad_norm": 0.07437079440706637,
"learning_rate": 1.1546186624052269e-07,
"loss": 0.9805,
"step": 960
},
{
"epoch": 4.915601023017903,
"grad_norm": 0.07524998197309374,
"learning_rate": 1.0058323388153491e-07,
"loss": 0.9765,
"step": 961
},
{
"epoch": 4.920716112531969,
"grad_norm": 0.07268873818022609,
"learning_rate": 8.672988439719199e-08,
"loss": 0.9568,
"step": 962
},
{
"epoch": 4.9258312020460355,
"grad_norm": 0.07629672988928426,
"learning_rate": 7.39019955560405e-08,
"loss": 0.9934,
"step": 963
},
{
"epoch": 4.930946291560103,
"grad_norm": 0.07396116828111508,
"learning_rate": 6.209973196771301e-08,
"loss": 0.9743,
"step": 964
},
{
"epoch": 4.936061381074169,
"grad_norm": 0.07416699993453406,
"learning_rate": 5.132324508085873e-08,
"loss": 0.9879,
"step": 965
},
{
"epoch": 4.9411764705882355,
"grad_norm": 0.07436538410708188,
"learning_rate": 4.1572673181198283e-08,
"loss": 0.9872,
"step": 966
},
{
"epoch": 4.946291560102302,
"grad_norm": 0.07136305198796307,
"learning_rate": 3.284814138971193e-08,
"loss": 0.9611,
"step": 967
},
{
"epoch": 4.951406649616368,
"grad_norm": 0.07282223951429675,
"learning_rate": 2.514976166108518e-08,
"loss": 0.9606,
"step": 968
},
{
"epoch": 4.956521739130435,
"grad_norm": 0.07408153822019757,
"learning_rate": 1.8477632782216703e-08,
"loss": 0.9564,
"step": 969
},
{
"epoch": 4.961636828644501,
"grad_norm": 1.6648677122846838,
"learning_rate": 1.2831840371010373e-08,
"loss": 0.9987,
"step": 970
},
{
"epoch": 4.966751918158568,
"grad_norm": 0.07577180986663225,
"learning_rate": 8.212456875229535e-09,
"loss": 0.9713,
"step": 971
},
{
"epoch": 4.971867007672635,
"grad_norm": 0.07315087204820622,
"learning_rate": 4.619541571582176e-09,
"loss": 0.9586,
"step": 972
},
{
"epoch": 4.976982097186701,
"grad_norm": 0.07628720300728932,
"learning_rate": 2.053140564974854e-09,
"loss": 0.9833,
"step": 973
},
{
"epoch": 4.982097186700767,
"grad_norm": 0.07411652218403934,
"learning_rate": 5.132867878909764e-10,
"loss": 0.9588,
"step": 974
},
{
"epoch": 4.987212276214834,
"grad_norm": 0.07429401114883186,
"learning_rate": 0.0,
"loss": 0.9949,
"step": 975
},
{
"epoch": 4.987212276214834,
"step": 975,
"total_flos": 7100358600687616.0,
"train_loss": 1.0383980158047799,
"train_runtime": 35038.9958,
"train_samples_per_second": 14.27,
"train_steps_per_second": 0.028
}
],
"logging_steps": 1.0,
"max_steps": 975,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 7100358600687616.0,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}