llama_3_8b_r_1 / trainer_state.json
sam2ai's picture
Add files using upload-large-folder tool
b3b6f44 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.9938900203665986,
"eval_steps": 500,
"global_step": 735,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.004073319755600814,
"grad_norm": 6.640312047414066,
"learning_rate": 1.3513513513513515e-07,
"loss": 0.2975,
"step": 1
},
{
"epoch": 0.008146639511201629,
"grad_norm": 6.28588556129052,
"learning_rate": 2.702702702702703e-07,
"loss": 0.2922,
"step": 2
},
{
"epoch": 0.012219959266802444,
"grad_norm": 6.53425740231569,
"learning_rate": 4.0540540540540546e-07,
"loss": 0.3306,
"step": 3
},
{
"epoch": 0.016293279022403257,
"grad_norm": 6.4274428823636995,
"learning_rate": 5.405405405405406e-07,
"loss": 0.298,
"step": 4
},
{
"epoch": 0.020366598778004074,
"grad_norm": 6.300509704789429,
"learning_rate": 6.756756756756758e-07,
"loss": 0.2764,
"step": 5
},
{
"epoch": 0.024439918533604887,
"grad_norm": 5.664520140886496,
"learning_rate": 8.108108108108109e-07,
"loss": 0.282,
"step": 6
},
{
"epoch": 0.028513238289205704,
"grad_norm": 5.353461680124434,
"learning_rate": 9.459459459459461e-07,
"loss": 0.3105,
"step": 7
},
{
"epoch": 0.032586558044806514,
"grad_norm": 4.456679664512122,
"learning_rate": 1.0810810810810812e-06,
"loss": 0.2629,
"step": 8
},
{
"epoch": 0.03665987780040733,
"grad_norm": 4.728194529681125,
"learning_rate": 1.2162162162162164e-06,
"loss": 0.2625,
"step": 9
},
{
"epoch": 0.04073319755600815,
"grad_norm": 3.0137557205711736,
"learning_rate": 1.3513513513513515e-06,
"loss": 0.2324,
"step": 10
},
{
"epoch": 0.04480651731160896,
"grad_norm": 3.805111529191594,
"learning_rate": 1.4864864864864868e-06,
"loss": 0.2635,
"step": 11
},
{
"epoch": 0.048879837067209775,
"grad_norm": 3.340077069220912,
"learning_rate": 1.6216216216216219e-06,
"loss": 0.2386,
"step": 12
},
{
"epoch": 0.05295315682281059,
"grad_norm": 3.997290518175738,
"learning_rate": 1.756756756756757e-06,
"loss": 0.2226,
"step": 13
},
{
"epoch": 0.05702647657841141,
"grad_norm": 3.977876990477163,
"learning_rate": 1.8918918918918922e-06,
"loss": 0.2593,
"step": 14
},
{
"epoch": 0.06109979633401222,
"grad_norm": 3.367674913595299,
"learning_rate": 2.0270270270270273e-06,
"loss": 0.2355,
"step": 15
},
{
"epoch": 0.06517311608961303,
"grad_norm": 3.1181805119884842,
"learning_rate": 2.1621621621621623e-06,
"loss": 0.2364,
"step": 16
},
{
"epoch": 0.06924643584521385,
"grad_norm": 2.9154930914971473,
"learning_rate": 2.297297297297298e-06,
"loss": 0.2072,
"step": 17
},
{
"epoch": 0.07331975560081466,
"grad_norm": 2.926858516811969,
"learning_rate": 2.432432432432433e-06,
"loss": 0.193,
"step": 18
},
{
"epoch": 0.07739307535641547,
"grad_norm": 3.2833775927281867,
"learning_rate": 2.5675675675675675e-06,
"loss": 0.2429,
"step": 19
},
{
"epoch": 0.0814663951120163,
"grad_norm": 2.60551584899032,
"learning_rate": 2.702702702702703e-06,
"loss": 0.1904,
"step": 20
},
{
"epoch": 0.0855397148676171,
"grad_norm": 2.4720627934368578,
"learning_rate": 2.837837837837838e-06,
"loss": 0.2241,
"step": 21
},
{
"epoch": 0.08961303462321792,
"grad_norm": 2.7061457567778286,
"learning_rate": 2.9729729729729736e-06,
"loss": 0.1787,
"step": 22
},
{
"epoch": 0.09368635437881874,
"grad_norm": 2.5113937626874554,
"learning_rate": 3.1081081081081082e-06,
"loss": 0.2011,
"step": 23
},
{
"epoch": 0.09775967413441955,
"grad_norm": 2.28679738040873,
"learning_rate": 3.2432432432432437e-06,
"loss": 0.1757,
"step": 24
},
{
"epoch": 0.10183299389002037,
"grad_norm": 1.9954370385777342,
"learning_rate": 3.3783783783783788e-06,
"loss": 0.156,
"step": 25
},
{
"epoch": 0.10590631364562118,
"grad_norm": 1.865602902801458,
"learning_rate": 3.513513513513514e-06,
"loss": 0.1652,
"step": 26
},
{
"epoch": 0.109979633401222,
"grad_norm": 2.2088323944290162,
"learning_rate": 3.648648648648649e-06,
"loss": 0.1481,
"step": 27
},
{
"epoch": 0.11405295315682282,
"grad_norm": 2.3442977739891253,
"learning_rate": 3.7837837837837844e-06,
"loss": 0.1499,
"step": 28
},
{
"epoch": 0.11812627291242363,
"grad_norm": 1.4670114524312288,
"learning_rate": 3.918918918918919e-06,
"loss": 0.1372,
"step": 29
},
{
"epoch": 0.12219959266802444,
"grad_norm": 2.0073103982055134,
"learning_rate": 4.0540540540540545e-06,
"loss": 0.1705,
"step": 30
},
{
"epoch": 0.12627291242362526,
"grad_norm": 1.832956891032527,
"learning_rate": 4.189189189189189e-06,
"loss": 0.1435,
"step": 31
},
{
"epoch": 0.13034623217922606,
"grad_norm": 2.094155576804916,
"learning_rate": 4.324324324324325e-06,
"loss": 0.1775,
"step": 32
},
{
"epoch": 0.13441955193482688,
"grad_norm": 1.8236712859693212,
"learning_rate": 4.45945945945946e-06,
"loss": 0.1754,
"step": 33
},
{
"epoch": 0.1384928716904277,
"grad_norm": 1.593806859793022,
"learning_rate": 4.594594594594596e-06,
"loss": 0.1218,
"step": 34
},
{
"epoch": 0.1425661914460285,
"grad_norm": 1.8242371066883118,
"learning_rate": 4.72972972972973e-06,
"loss": 0.1676,
"step": 35
},
{
"epoch": 0.14663951120162932,
"grad_norm": 1.6067651811563963,
"learning_rate": 4.864864864864866e-06,
"loss": 0.1561,
"step": 36
},
{
"epoch": 0.15071283095723015,
"grad_norm": 1.6523754405520419,
"learning_rate": 5e-06,
"loss": 0.1314,
"step": 37
},
{
"epoch": 0.15478615071283094,
"grad_norm": 1.5370079826272665,
"learning_rate": 5.135135135135135e-06,
"loss": 0.1517,
"step": 38
},
{
"epoch": 0.15885947046843177,
"grad_norm": 1.672563461104479,
"learning_rate": 5.2702702702702705e-06,
"loss": 0.1697,
"step": 39
},
{
"epoch": 0.1629327902240326,
"grad_norm": 1.691060384246397,
"learning_rate": 5.405405405405406e-06,
"loss": 0.1803,
"step": 40
},
{
"epoch": 0.1670061099796334,
"grad_norm": 1.5317972325193578,
"learning_rate": 5.540540540540541e-06,
"loss": 0.131,
"step": 41
},
{
"epoch": 0.1710794297352342,
"grad_norm": 1.4413677201314865,
"learning_rate": 5.675675675675676e-06,
"loss": 0.1195,
"step": 42
},
{
"epoch": 0.17515274949083504,
"grad_norm": 1.7167120970950247,
"learning_rate": 5.810810810810811e-06,
"loss": 0.1872,
"step": 43
},
{
"epoch": 0.17922606924643583,
"grad_norm": 1.415008811809238,
"learning_rate": 5.945945945945947e-06,
"loss": 0.1329,
"step": 44
},
{
"epoch": 0.18329938900203666,
"grad_norm": 1.4698432274933515,
"learning_rate": 6.081081081081082e-06,
"loss": 0.148,
"step": 45
},
{
"epoch": 0.18737270875763748,
"grad_norm": 1.4400084379892886,
"learning_rate": 6.2162162162162164e-06,
"loss": 0.1481,
"step": 46
},
{
"epoch": 0.19144602851323828,
"grad_norm": 2.0030021688014505,
"learning_rate": 6.351351351351351e-06,
"loss": 0.1542,
"step": 47
},
{
"epoch": 0.1955193482688391,
"grad_norm": 1.5204875720098323,
"learning_rate": 6.486486486486487e-06,
"loss": 0.1142,
"step": 48
},
{
"epoch": 0.19959266802443992,
"grad_norm": 1.8023111763307385,
"learning_rate": 6.621621621621622e-06,
"loss": 0.1623,
"step": 49
},
{
"epoch": 0.20366598778004075,
"grad_norm": 1.387841444224276,
"learning_rate": 6.7567567567567575e-06,
"loss": 0.1226,
"step": 50
},
{
"epoch": 0.20773930753564154,
"grad_norm": 1.4770728603330572,
"learning_rate": 6.891891891891892e-06,
"loss": 0.1532,
"step": 51
},
{
"epoch": 0.21181262729124237,
"grad_norm": 1.5526099510243074,
"learning_rate": 7.027027027027028e-06,
"loss": 0.1203,
"step": 52
},
{
"epoch": 0.2158859470468432,
"grad_norm": 1.522290383063382,
"learning_rate": 7.162162162162163e-06,
"loss": 0.1515,
"step": 53
},
{
"epoch": 0.219959266802444,
"grad_norm": 1.5406804163240948,
"learning_rate": 7.297297297297298e-06,
"loss": 0.1469,
"step": 54
},
{
"epoch": 0.2240325865580448,
"grad_norm": 1.527302080895313,
"learning_rate": 7.4324324324324324e-06,
"loss": 0.1352,
"step": 55
},
{
"epoch": 0.22810590631364563,
"grad_norm": 1.558342044097894,
"learning_rate": 7.567567567567569e-06,
"loss": 0.1411,
"step": 56
},
{
"epoch": 0.23217922606924643,
"grad_norm": 1.2770172249503504,
"learning_rate": 7.702702702702704e-06,
"loss": 0.1182,
"step": 57
},
{
"epoch": 0.23625254582484725,
"grad_norm": 1.3969066787405495,
"learning_rate": 7.837837837837838e-06,
"loss": 0.137,
"step": 58
},
{
"epoch": 0.24032586558044808,
"grad_norm": 1.67018108612516,
"learning_rate": 7.972972972972974e-06,
"loss": 0.1798,
"step": 59
},
{
"epoch": 0.24439918533604887,
"grad_norm": 1.5897484049578208,
"learning_rate": 8.108108108108109e-06,
"loss": 0.1659,
"step": 60
},
{
"epoch": 0.2484725050916497,
"grad_norm": 1.158481901678518,
"learning_rate": 8.243243243243245e-06,
"loss": 0.1247,
"step": 61
},
{
"epoch": 0.2525458248472505,
"grad_norm": 1.2947448113190885,
"learning_rate": 8.378378378378378e-06,
"loss": 0.1219,
"step": 62
},
{
"epoch": 0.25661914460285135,
"grad_norm": 1.489514366083781,
"learning_rate": 8.513513513513514e-06,
"loss": 0.1276,
"step": 63
},
{
"epoch": 0.2606924643584521,
"grad_norm": 1.2576829907966123,
"learning_rate": 8.64864864864865e-06,
"loss": 0.1313,
"step": 64
},
{
"epoch": 0.26476578411405294,
"grad_norm": 1.759594359119281,
"learning_rate": 8.783783783783785e-06,
"loss": 0.207,
"step": 65
},
{
"epoch": 0.26883910386965376,
"grad_norm": 1.2363649869365454,
"learning_rate": 8.91891891891892e-06,
"loss": 0.1334,
"step": 66
},
{
"epoch": 0.2729124236252546,
"grad_norm": 1.466539848741119,
"learning_rate": 9.054054054054054e-06,
"loss": 0.1495,
"step": 67
},
{
"epoch": 0.2769857433808554,
"grad_norm": 1.3201296592763596,
"learning_rate": 9.189189189189191e-06,
"loss": 0.1324,
"step": 68
},
{
"epoch": 0.28105906313645623,
"grad_norm": 1.2111706000441922,
"learning_rate": 9.324324324324325e-06,
"loss": 0.1312,
"step": 69
},
{
"epoch": 0.285132382892057,
"grad_norm": 1.2979543096114095,
"learning_rate": 9.45945945945946e-06,
"loss": 0.1208,
"step": 70
},
{
"epoch": 0.2892057026476578,
"grad_norm": 1.4259159096209377,
"learning_rate": 9.594594594594594e-06,
"loss": 0.124,
"step": 71
},
{
"epoch": 0.29327902240325865,
"grad_norm": 1.216603620132262,
"learning_rate": 9.729729729729732e-06,
"loss": 0.1157,
"step": 72
},
{
"epoch": 0.2973523421588595,
"grad_norm": 1.3357647312150325,
"learning_rate": 9.864864864864865e-06,
"loss": 0.1297,
"step": 73
},
{
"epoch": 0.3014256619144603,
"grad_norm": 1.494099520851434,
"learning_rate": 1e-05,
"loss": 0.1841,
"step": 74
},
{
"epoch": 0.3054989816700611,
"grad_norm": 1.3145157836751054,
"learning_rate": 9.99994352762958e-06,
"loss": 0.1355,
"step": 75
},
{
"epoch": 0.3095723014256619,
"grad_norm": 1.5739902543589488,
"learning_rate": 9.999774111793974e-06,
"loss": 0.1753,
"step": 76
},
{
"epoch": 0.3136456211812627,
"grad_norm": 1.9252674009965682,
"learning_rate": 9.999491756320105e-06,
"loss": 0.1881,
"step": 77
},
{
"epoch": 0.31771894093686354,
"grad_norm": 1.4979094601422043,
"learning_rate": 9.99909646758609e-06,
"loss": 0.1562,
"step": 78
},
{
"epoch": 0.32179226069246436,
"grad_norm": 1.2807564386337615,
"learning_rate": 9.99858825452108e-06,
"loss": 0.1207,
"step": 79
},
{
"epoch": 0.3258655804480652,
"grad_norm": 1.6174801266480536,
"learning_rate": 9.997967128605078e-06,
"loss": 0.2017,
"step": 80
},
{
"epoch": 0.329938900203666,
"grad_norm": 1.4159142195809724,
"learning_rate": 9.997233103868664e-06,
"loss": 0.124,
"step": 81
},
{
"epoch": 0.3340122199592668,
"grad_norm": 1.6942010203200524,
"learning_rate": 9.996386196892683e-06,
"loss": 0.1937,
"step": 82
},
{
"epoch": 0.3380855397148676,
"grad_norm": 1.5919535697796021,
"learning_rate": 9.995426426807875e-06,
"loss": 0.1577,
"step": 83
},
{
"epoch": 0.3421588594704684,
"grad_norm": 1.1350013056899857,
"learning_rate": 9.994353815294438e-06,
"loss": 0.1405,
"step": 84
},
{
"epoch": 0.34623217922606925,
"grad_norm": 1.3406415209215592,
"learning_rate": 9.993168386581533e-06,
"loss": 0.115,
"step": 85
},
{
"epoch": 0.35030549898167007,
"grad_norm": 1.4661144336991712,
"learning_rate": 9.991870167446751e-06,
"loss": 0.1352,
"step": 86
},
{
"epoch": 0.3543788187372709,
"grad_norm": 1.455419254668406,
"learning_rate": 9.990459187215498e-06,
"loss": 0.1292,
"step": 87
},
{
"epoch": 0.35845213849287166,
"grad_norm": 1.5030594271263464,
"learning_rate": 9.98893547776033e-06,
"loss": 0.1561,
"step": 88
},
{
"epoch": 0.3625254582484725,
"grad_norm": 1.6530134787902149,
"learning_rate": 9.987299073500245e-06,
"loss": 0.2027,
"step": 89
},
{
"epoch": 0.3665987780040733,
"grad_norm": 1.2625338529893853,
"learning_rate": 9.985550011399889e-06,
"loss": 0.1272,
"step": 90
},
{
"epoch": 0.37067209775967414,
"grad_norm": 1.2071106713338415,
"learning_rate": 9.98368833096874e-06,
"loss": 0.1666,
"step": 91
},
{
"epoch": 0.37474541751527496,
"grad_norm": 1.6780625112925267,
"learning_rate": 9.981714074260196e-06,
"loss": 0.1846,
"step": 92
},
{
"epoch": 0.3788187372708758,
"grad_norm": 1.3038545888439474,
"learning_rate": 9.979627285870644e-06,
"loss": 0.1259,
"step": 93
},
{
"epoch": 0.38289205702647655,
"grad_norm": 1.8469161081473644,
"learning_rate": 9.977428012938437e-06,
"loss": 0.2386,
"step": 94
},
{
"epoch": 0.3869653767820774,
"grad_norm": 1.4135444068823033,
"learning_rate": 9.975116305142836e-06,
"loss": 0.1359,
"step": 95
},
{
"epoch": 0.3910386965376782,
"grad_norm": 0.9193033843457485,
"learning_rate": 9.97269221470289e-06,
"loss": 0.1211,
"step": 96
},
{
"epoch": 0.395112016293279,
"grad_norm": 1.2085571858788857,
"learning_rate": 9.97015579637625e-06,
"loss": 0.1139,
"step": 97
},
{
"epoch": 0.39918533604887985,
"grad_norm": 1.1994440753116877,
"learning_rate": 9.967507107457942e-06,
"loss": 0.1324,
"step": 98
},
{
"epoch": 0.40325865580448067,
"grad_norm": 1.2284108132440392,
"learning_rate": 9.96474620777906e-06,
"loss": 0.1499,
"step": 99
},
{
"epoch": 0.4073319755600815,
"grad_norm": 1.6212975171489614,
"learning_rate": 9.961873159705426e-06,
"loss": 0.1581,
"step": 100
},
{
"epoch": 0.41140529531568226,
"grad_norm": 1.764584729235939,
"learning_rate": 9.95888802813617e-06,
"loss": 0.2015,
"step": 101
},
{
"epoch": 0.4154786150712831,
"grad_norm": 0.9967692738558999,
"learning_rate": 9.955790880502278e-06,
"loss": 0.1271,
"step": 102
},
{
"epoch": 0.4195519348268839,
"grad_norm": 1.021216540491067,
"learning_rate": 9.952581786765057e-06,
"loss": 0.1194,
"step": 103
},
{
"epoch": 0.42362525458248473,
"grad_norm": 1.803707990576697,
"learning_rate": 9.949260819414557e-06,
"loss": 0.1783,
"step": 104
},
{
"epoch": 0.42769857433808556,
"grad_norm": 1.2598754070444962,
"learning_rate": 9.945828053467939e-06,
"loss": 0.1336,
"step": 105
},
{
"epoch": 0.4317718940936864,
"grad_norm": 1.6420031533088841,
"learning_rate": 9.942283566467773e-06,
"loss": 0.1917,
"step": 106
},
{
"epoch": 0.43584521384928715,
"grad_norm": 1.4491181396568242,
"learning_rate": 9.938627438480295e-06,
"loss": 0.1698,
"step": 107
},
{
"epoch": 0.439918533604888,
"grad_norm": 1.5149064154516412,
"learning_rate": 9.93485975209359e-06,
"loss": 0.1752,
"step": 108
},
{
"epoch": 0.4439918533604888,
"grad_norm": 1.2955834669569637,
"learning_rate": 9.930980592415728e-06,
"loss": 0.1723,
"step": 109
},
{
"epoch": 0.4480651731160896,
"grad_norm": 2.017971717171439,
"learning_rate": 9.926990047072849e-06,
"loss": 0.2711,
"step": 110
},
{
"epoch": 0.45213849287169044,
"grad_norm": 1.3181601641420488,
"learning_rate": 9.922888206207174e-06,
"loss": 0.1275,
"step": 111
},
{
"epoch": 0.45621181262729127,
"grad_norm": 1.1244851230272161,
"learning_rate": 9.918675162474974e-06,
"loss": 0.119,
"step": 112
},
{
"epoch": 0.46028513238289204,
"grad_norm": 1.6894775763034329,
"learning_rate": 9.914351011044472e-06,
"loss": 0.181,
"step": 113
},
{
"epoch": 0.46435845213849286,
"grad_norm": 1.218930853717246,
"learning_rate": 9.909915849593705e-06,
"loss": 0.118,
"step": 114
},
{
"epoch": 0.4684317718940937,
"grad_norm": 1.3249036934529779,
"learning_rate": 9.905369778308304e-06,
"loss": 0.1301,
"step": 115
},
{
"epoch": 0.4725050916496945,
"grad_norm": 1.524266331724149,
"learning_rate": 9.900712899879237e-06,
"loss": 0.1762,
"step": 116
},
{
"epoch": 0.47657841140529533,
"grad_norm": 1.7086702641879,
"learning_rate": 9.895945319500488e-06,
"loss": 0.161,
"step": 117
},
{
"epoch": 0.48065173116089616,
"grad_norm": 1.2161786256312905,
"learning_rate": 9.891067144866687e-06,
"loss": 0.151,
"step": 118
},
{
"epoch": 0.4847250509164969,
"grad_norm": 1.1889624963960783,
"learning_rate": 9.886078486170665e-06,
"loss": 0.11,
"step": 119
},
{
"epoch": 0.48879837067209775,
"grad_norm": 1.365243201547572,
"learning_rate": 9.880979456100974e-06,
"loss": 0.1497,
"step": 120
},
{
"epoch": 0.49287169042769857,
"grad_norm": 1.2585043332112358,
"learning_rate": 9.875770169839343e-06,
"loss": 0.1432,
"step": 121
},
{
"epoch": 0.4969450101832994,
"grad_norm": 1.4002031183579842,
"learning_rate": 9.870450745058066e-06,
"loss": 0.1375,
"step": 122
},
{
"epoch": 0.5010183299389002,
"grad_norm": 1.1440514502080288,
"learning_rate": 9.865021301917358e-06,
"loss": 0.1402,
"step": 123
},
{
"epoch": 0.505091649694501,
"grad_norm": 1.2488108384566508,
"learning_rate": 9.859481963062623e-06,
"loss": 0.1162,
"step": 124
},
{
"epoch": 0.5091649694501018,
"grad_norm": 0.9510210043207896,
"learning_rate": 9.853832853621703e-06,
"loss": 0.1261,
"step": 125
},
{
"epoch": 0.5132382892057027,
"grad_norm": 1.10816343398081,
"learning_rate": 9.848074101202037e-06,
"loss": 0.13,
"step": 126
},
{
"epoch": 0.5173116089613035,
"grad_norm": 1.1626046136273134,
"learning_rate": 9.842205835887785e-06,
"loss": 0.1272,
"step": 127
},
{
"epoch": 0.5213849287169042,
"grad_norm": 1.1143169256996153,
"learning_rate": 9.836228190236892e-06,
"loss": 0.1511,
"step": 128
},
{
"epoch": 0.5254582484725051,
"grad_norm": 1.4130055894133877,
"learning_rate": 9.83014129927808e-06,
"loss": 0.1429,
"step": 129
},
{
"epoch": 0.5295315682281059,
"grad_norm": 1.1740739283614359,
"learning_rate": 9.823945300507815e-06,
"loss": 0.1482,
"step": 130
},
{
"epoch": 0.5336048879837068,
"grad_norm": 1.2043935670338368,
"learning_rate": 9.817640333887194e-06,
"loss": 0.1469,
"step": 131
},
{
"epoch": 0.5376782077393075,
"grad_norm": 1.0821808237384798,
"learning_rate": 9.81122654183878e-06,
"loss": 0.1166,
"step": 132
},
{
"epoch": 0.5417515274949084,
"grad_norm": 0.8388713275721755,
"learning_rate": 9.804704069243389e-06,
"loss": 0.1184,
"step": 133
},
{
"epoch": 0.5458248472505092,
"grad_norm": 0.8619134174032774,
"learning_rate": 9.798073063436815e-06,
"loss": 0.1129,
"step": 134
},
{
"epoch": 0.5498981670061099,
"grad_norm": 1.5195117909833376,
"learning_rate": 9.791333674206507e-06,
"loss": 0.1981,
"step": 135
},
{
"epoch": 0.5539714867617108,
"grad_norm": 1.0850667490173038,
"learning_rate": 9.784486053788179e-06,
"loss": 0.1107,
"step": 136
},
{
"epoch": 0.5580448065173116,
"grad_norm": 5.378203932427923,
"learning_rate": 9.77753035686237e-06,
"loss": 0.1653,
"step": 137
},
{
"epoch": 0.5621181262729125,
"grad_norm": 1.4728169236524777,
"learning_rate": 9.770466740550963e-06,
"loss": 0.1684,
"step": 138
},
{
"epoch": 0.5661914460285132,
"grad_norm": 1.0273885824594953,
"learning_rate": 9.763295364413616e-06,
"loss": 0.1255,
"step": 139
},
{
"epoch": 0.570264765784114,
"grad_norm": 1.1688264950391403,
"learning_rate": 9.756016390444174e-06,
"loss": 0.1485,
"step": 140
},
{
"epoch": 0.5743380855397149,
"grad_norm": 1.0399349052714333,
"learning_rate": 9.748629983067004e-06,
"loss": 0.1393,
"step": 141
},
{
"epoch": 0.5784114052953157,
"grad_norm": 1.7232917988923804,
"learning_rate": 9.741136309133279e-06,
"loss": 0.1983,
"step": 142
},
{
"epoch": 0.5824847250509165,
"grad_norm": 1.0879663731187623,
"learning_rate": 9.733535537917211e-06,
"loss": 0.1274,
"step": 143
},
{
"epoch": 0.5865580448065173,
"grad_norm": 0.9368564524590517,
"learning_rate": 9.725827841112226e-06,
"loss": 0.1207,
"step": 144
},
{
"epoch": 0.5906313645621182,
"grad_norm": 1.0798454525817731,
"learning_rate": 9.718013392827087e-06,
"loss": 0.1163,
"step": 145
},
{
"epoch": 0.594704684317719,
"grad_norm": 1.5318145919287482,
"learning_rate": 9.710092369581966e-06,
"loss": 0.1623,
"step": 146
},
{
"epoch": 0.5987780040733197,
"grad_norm": 1.203723637498308,
"learning_rate": 9.702064950304442e-06,
"loss": 0.1287,
"step": 147
},
{
"epoch": 0.6028513238289206,
"grad_norm": 1.0956856988556738,
"learning_rate": 9.693931316325473e-06,
"loss": 0.1027,
"step": 148
},
{
"epoch": 0.6069246435845214,
"grad_norm": 1.1062672875600013,
"learning_rate": 9.685691651375297e-06,
"loss": 0.1069,
"step": 149
},
{
"epoch": 0.6109979633401222,
"grad_norm": 0.9590820963065577,
"learning_rate": 9.677346141579277e-06,
"loss": 0.1073,
"step": 150
},
{
"epoch": 0.615071283095723,
"grad_norm": 1.3618508422829732,
"learning_rate": 9.668894975453705e-06,
"loss": 0.1664,
"step": 151
},
{
"epoch": 0.6191446028513238,
"grad_norm": 1.4665857948785328,
"learning_rate": 9.66033834390153e-06,
"loss": 0.1471,
"step": 152
},
{
"epoch": 0.6232179226069247,
"grad_norm": 1.1291116955509388,
"learning_rate": 9.65167644020806e-06,
"loss": 0.1352,
"step": 153
},
{
"epoch": 0.6272912423625254,
"grad_norm": 0.9222800631795732,
"learning_rate": 9.64290946003659e-06,
"loss": 0.104,
"step": 154
},
{
"epoch": 0.6313645621181263,
"grad_norm": 1.022018748144035,
"learning_rate": 9.63403760142398e-06,
"loss": 0.1065,
"step": 155
},
{
"epoch": 0.6354378818737271,
"grad_norm": 1.001598283035824,
"learning_rate": 9.625061064776183e-06,
"loss": 0.1166,
"step": 156
},
{
"epoch": 0.639511201629328,
"grad_norm": 0.8769241639145234,
"learning_rate": 9.61598005286372e-06,
"loss": 0.0995,
"step": 157
},
{
"epoch": 0.6435845213849287,
"grad_norm": 1.8569649444104026,
"learning_rate": 9.606794770817102e-06,
"loss": 0.203,
"step": 158
},
{
"epoch": 0.6476578411405295,
"grad_norm": 1.2047388338203417,
"learning_rate": 9.597505426122184e-06,
"loss": 0.1697,
"step": 159
},
{
"epoch": 0.6517311608961304,
"grad_norm": 1.2752618352547809,
"learning_rate": 9.588112228615495e-06,
"loss": 0.1917,
"step": 160
},
{
"epoch": 0.6558044806517311,
"grad_norm": 1.2319452847872203,
"learning_rate": 9.57861539047949e-06,
"loss": 0.144,
"step": 161
},
{
"epoch": 0.659877800407332,
"grad_norm": 1.5388901828747517,
"learning_rate": 9.569015126237744e-06,
"loss": 0.1687,
"step": 162
},
{
"epoch": 0.6639511201629328,
"grad_norm": 0.8967339263601748,
"learning_rate": 9.559311652750135e-06,
"loss": 0.1214,
"step": 163
},
{
"epoch": 0.6680244399185336,
"grad_norm": 0.8769857565124224,
"learning_rate": 9.549505189207924e-06,
"loss": 0.1026,
"step": 164
},
{
"epoch": 0.6720977596741344,
"grad_norm": 1.3154047046175807,
"learning_rate": 9.539595957128803e-06,
"loss": 0.1849,
"step": 165
},
{
"epoch": 0.6761710794297352,
"grad_norm": 0.9498449430689078,
"learning_rate": 9.529584180351902e-06,
"loss": 0.126,
"step": 166
},
{
"epoch": 0.6802443991853361,
"grad_norm": 1.249038222628632,
"learning_rate": 9.519470085032733e-06,
"loss": 0.1389,
"step": 167
},
{
"epoch": 0.6843177189409368,
"grad_norm": 1.0021120176773701,
"learning_rate": 9.509253899638066e-06,
"loss": 0.1097,
"step": 168
},
{
"epoch": 0.6883910386965377,
"grad_norm": 1.8720586325440431,
"learning_rate": 9.498935854940785e-06,
"loss": 0.191,
"step": 169
},
{
"epoch": 0.6924643584521385,
"grad_norm": 0.8779925461874577,
"learning_rate": 9.488516184014667e-06,
"loss": 0.1145,
"step": 170
},
{
"epoch": 0.6965376782077393,
"grad_norm": 1.1255708822925372,
"learning_rate": 9.477995122229117e-06,
"loss": 0.1648,
"step": 171
},
{
"epoch": 0.7006109979633401,
"grad_norm": 0.816369690322416,
"learning_rate": 9.467372907243858e-06,
"loss": 0.1067,
"step": 172
},
{
"epoch": 0.7046843177189409,
"grad_norm": 1.108258499952871,
"learning_rate": 9.456649779003548e-06,
"loss": 0.1333,
"step": 173
},
{
"epoch": 0.7087576374745418,
"grad_norm": 0.9004600691246489,
"learning_rate": 9.44582597973238e-06,
"loss": 0.1326,
"step": 174
},
{
"epoch": 0.7128309572301426,
"grad_norm": 1.1680632813950595,
"learning_rate": 9.434901753928593e-06,
"loss": 0.156,
"step": 175
},
{
"epoch": 0.7169042769857433,
"grad_norm": 0.9349361326216877,
"learning_rate": 9.423877348358956e-06,
"loss": 0.1068,
"step": 176
},
{
"epoch": 0.7209775967413442,
"grad_norm": 0.7647245750773263,
"learning_rate": 9.4127530120532e-06,
"loss": 0.1071,
"step": 177
},
{
"epoch": 0.725050916496945,
"grad_norm": 1.5837338501339817,
"learning_rate": 9.401528996298375e-06,
"loss": 0.1804,
"step": 178
},
{
"epoch": 0.7291242362525459,
"grad_norm": 0.8862736019513471,
"learning_rate": 9.390205554633193e-06,
"loss": 0.1141,
"step": 179
},
{
"epoch": 0.7331975560081466,
"grad_norm": 1.6364898998524855,
"learning_rate": 9.378782942842292e-06,
"loss": 0.154,
"step": 180
},
{
"epoch": 0.7372708757637475,
"grad_norm": 1.4722723655088397,
"learning_rate": 9.367261418950459e-06,
"loss": 0.2054,
"step": 181
},
{
"epoch": 0.7413441955193483,
"grad_norm": 1.2922211250976714,
"learning_rate": 9.355641243216798e-06,
"loss": 0.1869,
"step": 182
},
{
"epoch": 0.745417515274949,
"grad_norm": 1.4988694289422715,
"learning_rate": 9.343922678128854e-06,
"loss": 0.1205,
"step": 183
},
{
"epoch": 0.7494908350305499,
"grad_norm": 1.1189091721473459,
"learning_rate": 9.332105988396692e-06,
"loss": 0.1346,
"step": 184
},
{
"epoch": 0.7535641547861507,
"grad_norm": 0.9495523540335931,
"learning_rate": 9.3201914409469e-06,
"loss": 0.1414,
"step": 185
},
{
"epoch": 0.7576374745417516,
"grad_norm": 1.1449056285447028,
"learning_rate": 9.308179304916573e-06,
"loss": 0.1296,
"step": 186
},
{
"epoch": 0.7617107942973523,
"grad_norm": 0.9587320256117761,
"learning_rate": 9.29606985164723e-06,
"loss": 0.1092,
"step": 187
},
{
"epoch": 0.7657841140529531,
"grad_norm": 1.0410633462297467,
"learning_rate": 9.283863354678683e-06,
"loss": 0.1433,
"step": 188
},
{
"epoch": 0.769857433808554,
"grad_norm": 1.1206598030991681,
"learning_rate": 9.27156008974286e-06,
"loss": 0.143,
"step": 189
},
{
"epoch": 0.7739307535641547,
"grad_norm": 0.9317753678463256,
"learning_rate": 9.259160334757575e-06,
"loss": 0.1112,
"step": 190
},
{
"epoch": 0.7780040733197556,
"grad_norm": 0.9788621413861975,
"learning_rate": 9.246664369820249e-06,
"loss": 0.1406,
"step": 191
},
{
"epoch": 0.7820773930753564,
"grad_norm": 2.0096874215766145,
"learning_rate": 9.234072477201588e-06,
"loss": 0.2634,
"step": 192
},
{
"epoch": 0.7861507128309573,
"grad_norm": 1.4500526896303214,
"learning_rate": 9.2213849413392e-06,
"loss": 0.1386,
"step": 193
},
{
"epoch": 0.790224032586558,
"grad_norm": 0.7432983646218648,
"learning_rate": 9.208602048831176e-06,
"loss": 0.109,
"step": 194
},
{
"epoch": 0.7942973523421588,
"grad_norm": 0.8971636740777943,
"learning_rate": 9.195724088429611e-06,
"loss": 0.1153,
"step": 195
},
{
"epoch": 0.7983706720977597,
"grad_norm": 0.8965234237275304,
"learning_rate": 9.18275135103409e-06,
"loss": 0.1197,
"step": 196
},
{
"epoch": 0.8024439918533605,
"grad_norm": 1.0574527032014713,
"learning_rate": 9.169684129685099e-06,
"loss": 0.1417,
"step": 197
},
{
"epoch": 0.8065173116089613,
"grad_norm": 1.6715970219344671,
"learning_rate": 9.156522719557428e-06,
"loss": 0.2119,
"step": 198
},
{
"epoch": 0.8105906313645621,
"grad_norm": 1.4942152004701117,
"learning_rate": 9.143267417953486e-06,
"loss": 0.1713,
"step": 199
},
{
"epoch": 0.814663951120163,
"grad_norm": 1.1953125576245964,
"learning_rate": 9.129918524296596e-06,
"loss": 0.2013,
"step": 200
},
{
"epoch": 0.8187372708757638,
"grad_norm": 0.8487901135540641,
"learning_rate": 9.11647634012422e-06,
"loss": 0.1071,
"step": 201
},
{
"epoch": 0.8228105906313645,
"grad_norm": 0.8755395856306362,
"learning_rate": 9.102941169081167e-06,
"loss": 0.1225,
"step": 202
},
{
"epoch": 0.8268839103869654,
"grad_norm": 1.0216007592933207,
"learning_rate": 9.089313316912708e-06,
"loss": 0.1519,
"step": 203
},
{
"epoch": 0.8309572301425662,
"grad_norm": 0.9504018753400427,
"learning_rate": 9.075593091457692e-06,
"loss": 0.1232,
"step": 204
},
{
"epoch": 0.835030549898167,
"grad_norm": 0.836150880582385,
"learning_rate": 9.061780802641582e-06,
"loss": 0.1239,
"step": 205
},
{
"epoch": 0.8391038696537678,
"grad_norm": 0.7970664385278868,
"learning_rate": 9.047876762469451e-06,
"loss": 0.1089,
"step": 206
},
{
"epoch": 0.8431771894093686,
"grad_norm": 0.8842436763981146,
"learning_rate": 9.033881285018945e-06,
"loss": 0.1125,
"step": 207
},
{
"epoch": 0.8472505091649695,
"grad_norm": 1.2825326027052009,
"learning_rate": 9.019794686433174e-06,
"loss": 0.177,
"step": 208
},
{
"epoch": 0.8513238289205702,
"grad_norm": 0.9060622603360827,
"learning_rate": 9.005617284913586e-06,
"loss": 0.1051,
"step": 209
},
{
"epoch": 0.8553971486761711,
"grad_norm": 1.6864332980917753,
"learning_rate": 8.991349400712772e-06,
"loss": 0.127,
"step": 210
},
{
"epoch": 0.8594704684317719,
"grad_norm": 1.0001473814777286,
"learning_rate": 8.976991356127225e-06,
"loss": 0.1325,
"step": 211
},
{
"epoch": 0.8635437881873728,
"grad_norm": 0.7291694120330555,
"learning_rate": 8.962543475490068e-06,
"loss": 0.1121,
"step": 212
},
{
"epoch": 0.8676171079429735,
"grad_norm": 0.9272720834880637,
"learning_rate": 8.948006085163735e-06,
"loss": 0.1119,
"step": 213
},
{
"epoch": 0.8716904276985743,
"grad_norm": 0.8312647759455691,
"learning_rate": 8.933379513532575e-06,
"loss": 0.1112,
"step": 214
},
{
"epoch": 0.8757637474541752,
"grad_norm": 0.8149067559673082,
"learning_rate": 8.91866409099546e-06,
"loss": 0.1093,
"step": 215
},
{
"epoch": 0.879837067209776,
"grad_norm": 0.8297195108998483,
"learning_rate": 8.903860149958308e-06,
"loss": 0.1069,
"step": 216
},
{
"epoch": 0.8839103869653768,
"grad_norm": 1.0028275547991912,
"learning_rate": 8.888968024826575e-06,
"loss": 0.1382,
"step": 217
},
{
"epoch": 0.8879837067209776,
"grad_norm": 1.0402053627346783,
"learning_rate": 8.873988051997702e-06,
"loss": 0.1077,
"step": 218
},
{
"epoch": 0.8920570264765784,
"grad_norm": 1.016561804402662,
"learning_rate": 8.85892056985352e-06,
"loss": 0.1432,
"step": 219
},
{
"epoch": 0.8961303462321792,
"grad_norm": 1.6116327142859865,
"learning_rate": 8.8437659187526e-06,
"loss": 0.2553,
"step": 220
},
{
"epoch": 0.90020366598778,
"grad_norm": 1.9529266780373014,
"learning_rate": 8.828524441022575e-06,
"loss": 0.194,
"step": 221
},
{
"epoch": 0.9042769857433809,
"grad_norm": 0.7615347250958462,
"learning_rate": 8.813196480952393e-06,
"loss": 0.1043,
"step": 222
},
{
"epoch": 0.9083503054989817,
"grad_norm": 0.8628795475365344,
"learning_rate": 8.797782384784549e-06,
"loss": 0.1258,
"step": 223
},
{
"epoch": 0.9124236252545825,
"grad_norm": 0.8101851952339439,
"learning_rate": 8.782282500707262e-06,
"loss": 0.1067,
"step": 224
},
{
"epoch": 0.9164969450101833,
"grad_norm": 0.8946136137938691,
"learning_rate": 8.766697178846611e-06,
"loss": 0.1312,
"step": 225
},
{
"epoch": 0.9205702647657841,
"grad_norm": 1.1880991953150588,
"learning_rate": 8.751026771258622e-06,
"loss": 0.1457,
"step": 226
},
{
"epoch": 0.924643584521385,
"grad_norm": 0.7683685928101024,
"learning_rate": 8.735271631921322e-06,
"loss": 0.1094,
"step": 227
},
{
"epoch": 0.9287169042769857,
"grad_norm": 1.0839731638909997,
"learning_rate": 8.719432116726738e-06,
"loss": 0.1446,
"step": 228
},
{
"epoch": 0.9327902240325866,
"grad_norm": 1.0311134631629166,
"learning_rate": 8.703508583472855e-06,
"loss": 0.1551,
"step": 229
},
{
"epoch": 0.9368635437881874,
"grad_norm": 0.9245375301250945,
"learning_rate": 8.68750139185554e-06,
"loss": 0.1316,
"step": 230
},
{
"epoch": 0.9409368635437881,
"grad_norm": 0.7855768495227055,
"learning_rate": 8.671410903460416e-06,
"loss": 0.1256,
"step": 231
},
{
"epoch": 0.945010183299389,
"grad_norm": 1.4592779355487666,
"learning_rate": 8.65523748175469e-06,
"loss": 0.1715,
"step": 232
},
{
"epoch": 0.9490835030549898,
"grad_norm": 1.444988409307209,
"learning_rate": 8.63898149207895e-06,
"loss": 0.1826,
"step": 233
},
{
"epoch": 0.9531568228105907,
"grad_norm": 1.064651622477415,
"learning_rate": 8.622643301638902e-06,
"loss": 0.1451,
"step": 234
},
{
"epoch": 0.9572301425661914,
"grad_norm": 0.7774230497311596,
"learning_rate": 8.606223279497081e-06,
"loss": 0.0999,
"step": 235
},
{
"epoch": 0.9613034623217923,
"grad_norm": 0.9665945798764777,
"learning_rate": 8.589721796564521e-06,
"loss": 0.1011,
"step": 236
},
{
"epoch": 0.9653767820773931,
"grad_norm": 0.9332915183840536,
"learning_rate": 8.57313922559236e-06,
"loss": 0.131,
"step": 237
},
{
"epoch": 0.9694501018329938,
"grad_norm": 1.2006420660113526,
"learning_rate": 8.556475941163436e-06,
"loss": 0.1206,
"step": 238
},
{
"epoch": 0.9735234215885947,
"grad_norm": 1.3281125416185249,
"learning_rate": 8.539732319683817e-06,
"loss": 0.1687,
"step": 239
},
{
"epoch": 0.9775967413441955,
"grad_norm": 0.9158134622590804,
"learning_rate": 8.5229087393743e-06,
"loss": 0.1204,
"step": 240
},
{
"epoch": 0.9816700610997964,
"grad_norm": 1.2823822319441516,
"learning_rate": 8.506005580261872e-06,
"loss": 0.1613,
"step": 241
},
{
"epoch": 0.9857433808553971,
"grad_norm": 0.9217849677879558,
"learning_rate": 8.489023224171114e-06,
"loss": 0.1151,
"step": 242
},
{
"epoch": 0.9898167006109979,
"grad_norm": 0.8536566198123717,
"learning_rate": 8.47196205471559e-06,
"loss": 0.0979,
"step": 243
},
{
"epoch": 0.9938900203665988,
"grad_norm": 1.2576661581152648,
"learning_rate": 8.45482245728917e-06,
"loss": 0.1824,
"step": 244
},
{
"epoch": 0.9979633401221996,
"grad_norm": 1.601767617645757,
"learning_rate": 8.437604819057336e-06,
"loss": 0.1619,
"step": 245
},
{
"epoch": 1.0020366598778003,
"grad_norm": 0.9672208490405952,
"learning_rate": 8.420309528948422e-06,
"loss": 0.1146,
"step": 246
},
{
"epoch": 1.0061099796334012,
"grad_norm": 0.832508397003989,
"learning_rate": 8.40293697764484e-06,
"loss": 0.0915,
"step": 247
},
{
"epoch": 1.010183299389002,
"grad_norm": 0.904130963567214,
"learning_rate": 8.385487557574253e-06,
"loss": 0.0944,
"step": 248
},
{
"epoch": 1.0142566191446027,
"grad_norm": 0.7450068739811258,
"learning_rate": 8.367961662900704e-06,
"loss": 0.0827,
"step": 249
},
{
"epoch": 1.0183299389002036,
"grad_norm": 0.9476412140789163,
"learning_rate": 8.35035968951572e-06,
"loss": 0.1092,
"step": 250
},
{
"epoch": 1.0224032586558045,
"grad_norm": 0.9386152878297552,
"learning_rate": 8.33268203502937e-06,
"loss": 0.1095,
"step": 251
},
{
"epoch": 1.0264765784114054,
"grad_norm": 0.9603179696847995,
"learning_rate": 8.314929098761268e-06,
"loss": 0.0915,
"step": 252
},
{
"epoch": 1.030549898167006,
"grad_norm": 0.8304219961514915,
"learning_rate": 8.297101281731576e-06,
"loss": 0.0954,
"step": 253
},
{
"epoch": 1.034623217922607,
"grad_norm": 0.9195780795898366,
"learning_rate": 8.279198986651925e-06,
"loss": 0.1024,
"step": 254
},
{
"epoch": 1.0386965376782078,
"grad_norm": 0.8515089836255187,
"learning_rate": 8.261222617916335e-06,
"loss": 0.0841,
"step": 255
},
{
"epoch": 1.0427698574338085,
"grad_norm": 0.9554349159166925,
"learning_rate": 8.243172581592066e-06,
"loss": 0.1058,
"step": 256
},
{
"epoch": 1.0468431771894093,
"grad_norm": 0.6631561369471801,
"learning_rate": 8.22504928541045e-06,
"loss": 0.0886,
"step": 257
},
{
"epoch": 1.0509164969450102,
"grad_norm": 0.9048066298357775,
"learning_rate": 8.206853138757687e-06,
"loss": 0.0844,
"step": 258
},
{
"epoch": 1.054989816700611,
"grad_norm": 0.9562971602799583,
"learning_rate": 8.188584552665592e-06,
"loss": 0.0915,
"step": 259
},
{
"epoch": 1.0590631364562118,
"grad_norm": 1.002361514741711,
"learning_rate": 8.17024393980231e-06,
"loss": 0.1016,
"step": 260
},
{
"epoch": 1.0631364562118126,
"grad_norm": 0.7452508340661453,
"learning_rate": 8.15183171446299e-06,
"loss": 0.0674,
"step": 261
},
{
"epoch": 1.0672097759674135,
"grad_norm": 0.817626319509027,
"learning_rate": 8.133348292560442e-06,
"loss": 0.0785,
"step": 262
},
{
"epoch": 1.0712830957230142,
"grad_norm": 1.1811024484035837,
"learning_rate": 8.114794091615718e-06,
"loss": 0.1265,
"step": 263
},
{
"epoch": 1.075356415478615,
"grad_norm": 0.603310266983341,
"learning_rate": 8.096169530748708e-06,
"loss": 0.0638,
"step": 264
},
{
"epoch": 1.079429735234216,
"grad_norm": 0.7947710103588282,
"learning_rate": 8.077475030668647e-06,
"loss": 0.096,
"step": 265
},
{
"epoch": 1.0835030549898166,
"grad_norm": 0.9274759793959164,
"learning_rate": 8.058711013664633e-06,
"loss": 0.0879,
"step": 266
},
{
"epoch": 1.0875763747454175,
"grad_norm": 0.7085855447631716,
"learning_rate": 8.039877903596069e-06,
"loss": 0.0847,
"step": 267
},
{
"epoch": 1.0916496945010183,
"grad_norm": 0.7746156648229091,
"learning_rate": 8.020976125883105e-06,
"loss": 0.0814,
"step": 268
},
{
"epoch": 1.0957230142566192,
"grad_norm": 0.9487432139661005,
"learning_rate": 8.002006107497018e-06,
"loss": 0.099,
"step": 269
},
{
"epoch": 1.0997963340122199,
"grad_norm": 0.7861823082141532,
"learning_rate": 7.982968276950568e-06,
"loss": 0.0719,
"step": 270
},
{
"epoch": 1.1038696537678208,
"grad_norm": 1.2995214283719732,
"learning_rate": 7.963863064288326e-06,
"loss": 0.1165,
"step": 271
},
{
"epoch": 1.1079429735234216,
"grad_norm": 0.6796566871802787,
"learning_rate": 7.944690901076949e-06,
"loss": 0.078,
"step": 272
},
{
"epoch": 1.1120162932790225,
"grad_norm": 0.853762040034195,
"learning_rate": 7.925452220395436e-06,
"loss": 0.0994,
"step": 273
},
{
"epoch": 1.1160896130346232,
"grad_norm": 0.8516487725429964,
"learning_rate": 7.906147456825349e-06,
"loss": 0.0988,
"step": 274
},
{
"epoch": 1.120162932790224,
"grad_norm": 0.7864403540144408,
"learning_rate": 7.886777046440993e-06,
"loss": 0.0814,
"step": 275
},
{
"epoch": 1.124236252545825,
"grad_norm": 1.1716856642298268,
"learning_rate": 7.867341426799562e-06,
"loss": 0.1396,
"step": 276
},
{
"epoch": 1.1283095723014256,
"grad_norm": 0.7736692798031964,
"learning_rate": 7.847841036931263e-06,
"loss": 0.0963,
"step": 277
},
{
"epoch": 1.1323828920570265,
"grad_norm": 0.627853542856467,
"learning_rate": 7.828276317329388e-06,
"loss": 0.0796,
"step": 278
},
{
"epoch": 1.1364562118126273,
"grad_norm": 0.6987894756370886,
"learning_rate": 7.80864770994038e-06,
"loss": 0.0673,
"step": 279
},
{
"epoch": 1.140529531568228,
"grad_norm": 0.7372216853011646,
"learning_rate": 7.788955658153829e-06,
"loss": 0.0861,
"step": 280
},
{
"epoch": 1.1446028513238289,
"grad_norm": 0.6309328369589573,
"learning_rate": 7.769200606792476e-06,
"loss": 0.0732,
"step": 281
},
{
"epoch": 1.1486761710794298,
"grad_norm": 1.5872377960488204,
"learning_rate": 7.749383002102147e-06,
"loss": 0.098,
"step": 282
},
{
"epoch": 1.1527494908350306,
"grad_norm": 0.8441328057902189,
"learning_rate": 7.72950329174169e-06,
"loss": 0.0906,
"step": 283
},
{
"epoch": 1.1568228105906313,
"grad_norm": 0.843705996727692,
"learning_rate": 7.709561924772855e-06,
"loss": 0.0979,
"step": 284
},
{
"epoch": 1.1608961303462322,
"grad_norm": 0.977169845042871,
"learning_rate": 7.689559351650142e-06,
"loss": 0.1124,
"step": 285
},
{
"epoch": 1.164969450101833,
"grad_norm": 0.9463259076204176,
"learning_rate": 7.66949602421064e-06,
"loss": 0.1068,
"step": 286
},
{
"epoch": 1.1690427698574337,
"grad_norm": 0.811424080065267,
"learning_rate": 7.649372395663816e-06,
"loss": 0.072,
"step": 287
},
{
"epoch": 1.1731160896130346,
"grad_norm": 0.8132304851241069,
"learning_rate": 7.629188920581267e-06,
"loss": 0.0884,
"step": 288
},
{
"epoch": 1.1771894093686355,
"grad_norm": 1.4632519153044663,
"learning_rate": 7.608946054886468e-06,
"loss": 0.0991,
"step": 289
},
{
"epoch": 1.1812627291242364,
"grad_norm": 0.9085288701945666,
"learning_rate": 7.588644255844464e-06,
"loss": 0.1015,
"step": 290
},
{
"epoch": 1.185336048879837,
"grad_norm": 0.7812384333509852,
"learning_rate": 7.568283982051538e-06,
"loss": 0.0773,
"step": 291
},
{
"epoch": 1.189409368635438,
"grad_norm": 0.6164352412170983,
"learning_rate": 7.5478656934248626e-06,
"loss": 0.0742,
"step": 292
},
{
"epoch": 1.1934826883910388,
"grad_norm": 0.7441220005875893,
"learning_rate": 7.527389851192099e-06,
"loss": 0.0805,
"step": 293
},
{
"epoch": 1.1975560081466394,
"grad_norm": 0.8878229183151843,
"learning_rate": 7.506856917880989e-06,
"loss": 0.0831,
"step": 294
},
{
"epoch": 1.2016293279022403,
"grad_norm": 0.6283198281323106,
"learning_rate": 7.486267357308896e-06,
"loss": 0.0807,
"step": 295
},
{
"epoch": 1.2057026476578412,
"grad_norm": 0.8264549827812461,
"learning_rate": 7.465621634572336e-06,
"loss": 0.0857,
"step": 296
},
{
"epoch": 1.2097759674134418,
"grad_norm": 1.0321645288915193,
"learning_rate": 7.444920216036473e-06,
"loss": 0.0969,
"step": 297
},
{
"epoch": 1.2138492871690427,
"grad_norm": 0.8253428462864323,
"learning_rate": 7.4241635693245766e-06,
"loss": 0.0745,
"step": 298
},
{
"epoch": 1.2179226069246436,
"grad_norm": 0.612362708583028,
"learning_rate": 7.40335216330746e-06,
"loss": 0.0724,
"step": 299
},
{
"epoch": 1.2219959266802445,
"grad_norm": 0.7460056721898694,
"learning_rate": 7.382486468092899e-06,
"loss": 0.0826,
"step": 300
},
{
"epoch": 1.2260692464358451,
"grad_norm": 2.407874545569998,
"learning_rate": 7.361566955014999e-06,
"loss": 0.0938,
"step": 301
},
{
"epoch": 1.230142566191446,
"grad_norm": 1.531711526524773,
"learning_rate": 7.340594096623559e-06,
"loss": 0.0895,
"step": 302
},
{
"epoch": 1.234215885947047,
"grad_norm": 0.6587659825194164,
"learning_rate": 7.319568366673389e-06,
"loss": 0.0791,
"step": 303
},
{
"epoch": 1.2382892057026478,
"grad_norm": 0.8381388182575081,
"learning_rate": 7.2984902401136115e-06,
"loss": 0.0908,
"step": 304
},
{
"epoch": 1.2423625254582484,
"grad_norm": 0.8130694806248154,
"learning_rate": 7.277360193076936e-06,
"loss": 0.0837,
"step": 305
},
{
"epoch": 1.2464358452138493,
"grad_norm": 0.6905135676271011,
"learning_rate": 7.256178702868899e-06,
"loss": 0.0787,
"step": 306
},
{
"epoch": 1.2505091649694502,
"grad_norm": 0.7451067168863142,
"learning_rate": 7.234946247957087e-06,
"loss": 0.0859,
"step": 307
},
{
"epoch": 1.2545824847250509,
"grad_norm": 0.9065093126808257,
"learning_rate": 7.213663307960321e-06,
"loss": 0.093,
"step": 308
},
{
"epoch": 1.2586558044806517,
"grad_norm": 1.1386079619114733,
"learning_rate": 7.192330363637832e-06,
"loss": 0.1148,
"step": 309
},
{
"epoch": 1.2627291242362526,
"grad_norm": 0.8040303061288451,
"learning_rate": 7.170947896878392e-06,
"loss": 0.0794,
"step": 310
},
{
"epoch": 1.2668024439918533,
"grad_norm": 0.7068240451041161,
"learning_rate": 7.149516390689433e-06,
"loss": 0.0772,
"step": 311
},
{
"epoch": 1.2708757637474541,
"grad_norm": 0.713991995225983,
"learning_rate": 7.12803632918614e-06,
"loss": 0.0773,
"step": 312
},
{
"epoch": 1.274949083503055,
"grad_norm": 0.8238519181704003,
"learning_rate": 7.1065081975805086e-06,
"loss": 0.0772,
"step": 313
},
{
"epoch": 1.2790224032586557,
"grad_norm": 1.100568471941002,
"learning_rate": 7.084932482170385e-06,
"loss": 0.1006,
"step": 314
},
{
"epoch": 1.2830957230142566,
"grad_norm": 0.7759792325387684,
"learning_rate": 7.063309670328491e-06,
"loss": 0.0884,
"step": 315
},
{
"epoch": 1.2871690427698574,
"grad_norm": 0.7869101421309218,
"learning_rate": 7.041640250491398e-06,
"loss": 0.076,
"step": 316
},
{
"epoch": 1.2912423625254583,
"grad_norm": 0.6354916164352619,
"learning_rate": 7.019924712148511e-06,
"loss": 0.0719,
"step": 317
},
{
"epoch": 1.2953156822810592,
"grad_norm": 0.9646300732606192,
"learning_rate": 6.998163545830998e-06,
"loss": 0.0896,
"step": 318
},
{
"epoch": 1.2993890020366599,
"grad_norm": 0.7887777312539676,
"learning_rate": 6.976357243100718e-06,
"loss": 0.0894,
"step": 319
},
{
"epoch": 1.3034623217922607,
"grad_norm": 0.8543058333089151,
"learning_rate": 6.954506296539112e-06,
"loss": 0.0924,
"step": 320
},
{
"epoch": 1.3075356415478616,
"grad_norm": 0.8888556729496842,
"learning_rate": 6.9326111997360775e-06,
"loss": 0.089,
"step": 321
},
{
"epoch": 1.3116089613034623,
"grad_norm": 1.0511089795556172,
"learning_rate": 6.910672447278827e-06,
"loss": 0.0996,
"step": 322
},
{
"epoch": 1.3156822810590632,
"grad_norm": 0.6640960066694046,
"learning_rate": 6.8886905347406985e-06,
"loss": 0.0889,
"step": 323
},
{
"epoch": 1.319755600814664,
"grad_norm": 0.7474191392306333,
"learning_rate": 6.866665958669976e-06,
"loss": 0.1038,
"step": 324
},
{
"epoch": 1.3238289205702647,
"grad_norm": 0.6625825963174811,
"learning_rate": 6.844599216578667e-06,
"loss": 0.0785,
"step": 325
},
{
"epoch": 1.3279022403258656,
"grad_norm": 1.0051676188526157,
"learning_rate": 6.822490806931262e-06,
"loss": 0.1002,
"step": 326
},
{
"epoch": 1.3319755600814664,
"grad_norm": 0.6599233049529285,
"learning_rate": 6.800341229133486e-06,
"loss": 0.0869,
"step": 327
},
{
"epoch": 1.336048879837067,
"grad_norm": 0.9213204884412184,
"learning_rate": 6.778150983520999e-06,
"loss": 0.1235,
"step": 328
},
{
"epoch": 1.340122199592668,
"grad_norm": 0.7390992342429102,
"learning_rate": 6.755920571348111e-06,
"loss": 0.0839,
"step": 329
},
{
"epoch": 1.3441955193482689,
"grad_norm": 0.8491729200395206,
"learning_rate": 6.73365049477645e-06,
"loss": 0.0851,
"step": 330
},
{
"epoch": 1.3482688391038695,
"grad_norm": 1.9733113662315056,
"learning_rate": 6.711341256863623e-06,
"loss": 0.121,
"step": 331
},
{
"epoch": 1.3523421588594704,
"grad_norm": 0.7374060548134543,
"learning_rate": 6.688993361551847e-06,
"loss": 0.0829,
"step": 332
},
{
"epoch": 1.3564154786150713,
"grad_norm": 0.9445537578309379,
"learning_rate": 6.66660731365657e-06,
"loss": 0.1029,
"step": 333
},
{
"epoch": 1.3604887983706722,
"grad_norm": 0.6167725635362412,
"learning_rate": 6.64418361885507e-06,
"loss": 0.0846,
"step": 334
},
{
"epoch": 1.364562118126273,
"grad_norm": 0.9582564113338518,
"learning_rate": 6.621722783675024e-06,
"loss": 0.0884,
"step": 335
},
{
"epoch": 1.3686354378818737,
"grad_norm": 0.7341939098462013,
"learning_rate": 6.599225315483076e-06,
"loss": 0.0822,
"step": 336
},
{
"epoch": 1.3727087576374746,
"grad_norm": 0.6266005187010786,
"learning_rate": 6.576691722473368e-06,
"loss": 0.0741,
"step": 337
},
{
"epoch": 1.3767820773930755,
"grad_norm": 1.0420040214165822,
"learning_rate": 6.554122513656065e-06,
"loss": 0.1013,
"step": 338
},
{
"epoch": 1.3808553971486761,
"grad_norm": 0.7579135385444917,
"learning_rate": 6.531518198845854e-06,
"loss": 0.0869,
"step": 339
},
{
"epoch": 1.384928716904277,
"grad_norm": 0.9115989624081831,
"learning_rate": 6.508879288650431e-06,
"loss": 0.0942,
"step": 340
},
{
"epoch": 1.3890020366598779,
"grad_norm": 1.0153542283519323,
"learning_rate": 6.486206294458966e-06,
"loss": 0.1146,
"step": 341
},
{
"epoch": 1.3930753564154785,
"grad_norm": 0.732006592180605,
"learning_rate": 6.463499728430549e-06,
"loss": 0.0766,
"step": 342
},
{
"epoch": 1.3971486761710794,
"grad_norm": 1.0103652295314396,
"learning_rate": 6.4407601034826225e-06,
"loss": 0.1056,
"step": 343
},
{
"epoch": 1.4012219959266803,
"grad_norm": 0.645303692033008,
"learning_rate": 6.417987933279397e-06,
"loss": 0.082,
"step": 344
},
{
"epoch": 1.405295315682281,
"grad_norm": 0.6637967514075462,
"learning_rate": 6.395183732220242e-06,
"loss": 0.0715,
"step": 345
},
{
"epoch": 1.4093686354378818,
"grad_norm": 0.8370194911957503,
"learning_rate": 6.372348015428077e-06,
"loss": 0.0809,
"step": 346
},
{
"epoch": 1.4134419551934827,
"grad_norm": 0.7044123550494559,
"learning_rate": 6.349481298737723e-06,
"loss": 0.0861,
"step": 347
},
{
"epoch": 1.4175152749490836,
"grad_norm": 0.6093242409125569,
"learning_rate": 6.32658409868426e-06,
"loss": 0.0684,
"step": 348
},
{
"epoch": 1.4215885947046842,
"grad_norm": 0.632063149807712,
"learning_rate": 6.303656932491349e-06,
"loss": 0.0802,
"step": 349
},
{
"epoch": 1.4256619144602851,
"grad_norm": 1.0427713522649937,
"learning_rate": 6.280700318059563e-06,
"loss": 0.1154,
"step": 350
},
{
"epoch": 1.429735234215886,
"grad_norm": 1.0764304967406033,
"learning_rate": 6.257714773954674e-06,
"loss": 0.1164,
"step": 351
},
{
"epoch": 1.4338085539714869,
"grad_norm": 0.7365645776407571,
"learning_rate": 6.234700819395946e-06,
"loss": 0.0901,
"step": 352
},
{
"epoch": 1.4378818737270875,
"grad_norm": 0.7945282541810098,
"learning_rate": 6.211658974244407e-06,
"loss": 0.0908,
"step": 353
},
{
"epoch": 1.4419551934826884,
"grad_norm": 0.6918143511185231,
"learning_rate": 6.1885897589911e-06,
"loss": 0.0845,
"step": 354
},
{
"epoch": 1.4460285132382893,
"grad_norm": 1.1124806520090096,
"learning_rate": 6.1654936947453355e-06,
"loss": 0.1019,
"step": 355
},
{
"epoch": 1.45010183299389,
"grad_norm": 0.6025591013529961,
"learning_rate": 6.142371303222909e-06,
"loss": 0.0755,
"step": 356
},
{
"epoch": 1.4541751527494908,
"grad_norm": 0.8630513814588915,
"learning_rate": 6.119223106734328e-06,
"loss": 0.0895,
"step": 357
},
{
"epoch": 1.4582484725050917,
"grad_norm": 0.7965271271909935,
"learning_rate": 6.0960496281729995e-06,
"loss": 0.0783,
"step": 358
},
{
"epoch": 1.4623217922606924,
"grad_norm": 0.7038891889592316,
"learning_rate": 6.072851391003432e-06,
"loss": 0.0796,
"step": 359
},
{
"epoch": 1.4663951120162932,
"grad_norm": 0.6384512393476599,
"learning_rate": 6.0496289192494e-06,
"loss": 0.0881,
"step": 360
},
{
"epoch": 1.4704684317718941,
"grad_norm": 0.940655034735333,
"learning_rate": 6.026382737482116e-06,
"loss": 0.102,
"step": 361
},
{
"epoch": 1.4745417515274948,
"grad_norm": 0.9682934531343439,
"learning_rate": 6.003113370808375e-06,
"loss": 0.1029,
"step": 362
},
{
"epoch": 1.4786150712830957,
"grad_norm": 1.0247819631277137,
"learning_rate": 5.979821344858695e-06,
"loss": 0.1197,
"step": 363
},
{
"epoch": 1.4826883910386965,
"grad_norm": 0.7672663757155214,
"learning_rate": 5.956507185775441e-06,
"loss": 0.0917,
"step": 364
},
{
"epoch": 1.4867617107942974,
"grad_norm": 0.6578195533765551,
"learning_rate": 5.933171420200946e-06,
"loss": 0.0811,
"step": 365
},
{
"epoch": 1.4908350305498983,
"grad_norm": 0.6603035021161331,
"learning_rate": 5.909814575265609e-06,
"loss": 0.0819,
"step": 366
},
{
"epoch": 1.494908350305499,
"grad_norm": 0.583612907656864,
"learning_rate": 5.88643717857599e-06,
"loss": 0.0664,
"step": 367
},
{
"epoch": 1.4989816700610998,
"grad_norm": 0.6176322927829936,
"learning_rate": 5.863039758202889e-06,
"loss": 0.0771,
"step": 368
},
{
"epoch": 1.5030549898167007,
"grad_norm": 0.7024687823342335,
"learning_rate": 5.839622842669423e-06,
"loss": 0.0815,
"step": 369
},
{
"epoch": 1.5071283095723014,
"grad_norm": 0.8718566327697364,
"learning_rate": 5.816186960939084e-06,
"loss": 0.0974,
"step": 370
},
{
"epoch": 1.5112016293279023,
"grad_norm": 0.7361502684673189,
"learning_rate": 5.7927326424037875e-06,
"loss": 0.0848,
"step": 371
},
{
"epoch": 1.5152749490835031,
"grad_norm": 0.6902398893653613,
"learning_rate": 5.7692604168719225e-06,
"loss": 0.0783,
"step": 372
},
{
"epoch": 1.5193482688391038,
"grad_norm": 0.572845083615776,
"learning_rate": 5.745770814556373e-06,
"loss": 0.0763,
"step": 373
},
{
"epoch": 1.5234215885947047,
"grad_norm": 1.0372051056861917,
"learning_rate": 5.722264366062549e-06,
"loss": 0.1234,
"step": 374
},
{
"epoch": 1.5274949083503055,
"grad_norm": 0.9263356887643026,
"learning_rate": 5.698741602376395e-06,
"loss": 0.0958,
"step": 375
},
{
"epoch": 1.5315682281059062,
"grad_norm": 0.6171130691440203,
"learning_rate": 5.675203054852403e-06,
"loss": 0.0772,
"step": 376
},
{
"epoch": 1.535641547861507,
"grad_norm": 0.686848420732314,
"learning_rate": 5.651649255201603e-06,
"loss": 0.091,
"step": 377
},
{
"epoch": 1.539714867617108,
"grad_norm": 0.6841747229884891,
"learning_rate": 5.628080735479553e-06,
"loss": 0.0861,
"step": 378
},
{
"epoch": 1.5437881873727086,
"grad_norm": 0.5919736957226754,
"learning_rate": 5.604498028074323e-06,
"loss": 0.0702,
"step": 379
},
{
"epoch": 1.5478615071283097,
"grad_norm": 0.5832264712402621,
"learning_rate": 5.580901665694471e-06,
"loss": 0.0744,
"step": 380
},
{
"epoch": 1.5519348268839104,
"grad_norm": 0.8947933752883832,
"learning_rate": 5.557292181357003e-06,
"loss": 0.099,
"step": 381
},
{
"epoch": 1.556008146639511,
"grad_norm": 0.7901652985659288,
"learning_rate": 5.533670108375334e-06,
"loss": 0.0808,
"step": 382
},
{
"epoch": 1.5600814663951121,
"grad_norm": 0.7423631802302912,
"learning_rate": 5.510035980347249e-06,
"loss": 0.0691,
"step": 383
},
{
"epoch": 1.5641547861507128,
"grad_norm": 1.1196975716000397,
"learning_rate": 5.486390331142841e-06,
"loss": 0.14,
"step": 384
},
{
"epoch": 1.5682281059063137,
"grad_norm": 1.1562803291517847,
"learning_rate": 5.462733694892452e-06,
"loss": 0.1132,
"step": 385
},
{
"epoch": 1.5723014256619146,
"grad_norm": 0.5971769426579886,
"learning_rate": 5.439066605974615e-06,
"loss": 0.0918,
"step": 386
},
{
"epoch": 1.5763747454175152,
"grad_norm": 0.7622249533746938,
"learning_rate": 5.415389599003972e-06,
"loss": 0.0834,
"step": 387
},
{
"epoch": 1.580448065173116,
"grad_norm": 0.8225191930854497,
"learning_rate": 5.391703208819209e-06,
"loss": 0.0901,
"step": 388
},
{
"epoch": 1.584521384928717,
"grad_norm": 0.903243980934119,
"learning_rate": 5.368007970470964e-06,
"loss": 0.0877,
"step": 389
},
{
"epoch": 1.5885947046843176,
"grad_norm": 0.8517368255821154,
"learning_rate": 5.344304419209748e-06,
"loss": 0.0982,
"step": 390
},
{
"epoch": 1.5926680244399185,
"grad_norm": 0.5949993054618524,
"learning_rate": 5.3205930904738544e-06,
"loss": 0.069,
"step": 391
},
{
"epoch": 1.5967413441955194,
"grad_norm": 0.9588124561719404,
"learning_rate": 5.296874519877256e-06,
"loss": 0.1151,
"step": 392
},
{
"epoch": 1.60081466395112,
"grad_norm": 0.7682037010435501,
"learning_rate": 5.273149243197517e-06,
"loss": 0.1008,
"step": 393
},
{
"epoch": 1.6048879837067211,
"grad_norm": 0.6305755320656267,
"learning_rate": 5.2494177963636785e-06,
"loss": 0.0809,
"step": 394
},
{
"epoch": 1.6089613034623218,
"grad_norm": 0.8293942768800023,
"learning_rate": 5.225680715444168e-06,
"loss": 0.0933,
"step": 395
},
{
"epoch": 1.6130346232179225,
"grad_norm": 0.6400765956762063,
"learning_rate": 5.201938536634674e-06,
"loss": 0.083,
"step": 396
},
{
"epoch": 1.6171079429735236,
"grad_norm": 0.6237036767747972,
"learning_rate": 5.178191796246043e-06,
"loss": 0.0737,
"step": 397
},
{
"epoch": 1.6211812627291242,
"grad_norm": 0.76728636527015,
"learning_rate": 5.154441030692162e-06,
"loss": 0.0925,
"step": 398
},
{
"epoch": 1.625254582484725,
"grad_norm": 0.6271220608058555,
"learning_rate": 5.1306867764778445e-06,
"loss": 0.0667,
"step": 399
},
{
"epoch": 1.629327902240326,
"grad_norm": 0.563944778713813,
"learning_rate": 5.106929570186706e-06,
"loss": 0.0637,
"step": 400
},
{
"epoch": 1.6334012219959266,
"grad_norm": 0.8300464332472979,
"learning_rate": 5.083169948469049e-06,
"loss": 0.0979,
"step": 401
},
{
"epoch": 1.6374745417515275,
"grad_norm": 0.8778673983216725,
"learning_rate": 5.059408448029737e-06,
"loss": 0.1003,
"step": 402
},
{
"epoch": 1.6415478615071284,
"grad_norm": 0.9971185861157015,
"learning_rate": 5.0356456056160715e-06,
"loss": 0.1058,
"step": 403
},
{
"epoch": 1.645621181262729,
"grad_norm": 0.6245905119244692,
"learning_rate": 5.0118819580056686e-06,
"loss": 0.0732,
"step": 404
},
{
"epoch": 1.64969450101833,
"grad_norm": 0.6093763548801165,
"learning_rate": 4.988118041994332e-06,
"loss": 0.073,
"step": 405
},
{
"epoch": 1.6537678207739308,
"grad_norm": 0.732574864023735,
"learning_rate": 4.964354394383929e-06,
"loss": 0.0784,
"step": 406
},
{
"epoch": 1.6578411405295315,
"grad_norm": 0.5623878101115458,
"learning_rate": 4.940591551970264e-06,
"loss": 0.0705,
"step": 407
},
{
"epoch": 1.6619144602851323,
"grad_norm": 0.6836109996389221,
"learning_rate": 4.9168300515309515e-06,
"loss": 0.0759,
"step": 408
},
{
"epoch": 1.6659877800407332,
"grad_norm": 1.0751642454268762,
"learning_rate": 4.8930704298132965e-06,
"loss": 0.1147,
"step": 409
},
{
"epoch": 1.6700610997963339,
"grad_norm": 0.6925121226298128,
"learning_rate": 4.869313223522159e-06,
"loss": 0.0773,
"step": 410
},
{
"epoch": 1.674134419551935,
"grad_norm": 0.6533487029013805,
"learning_rate": 4.845558969307839e-06,
"loss": 0.0769,
"step": 411
},
{
"epoch": 1.6782077393075356,
"grad_norm": 0.6834593440184977,
"learning_rate": 4.821808203753959e-06,
"loss": 0.0848,
"step": 412
},
{
"epoch": 1.6822810590631363,
"grad_norm": 0.5761353995488768,
"learning_rate": 4.798061463365327e-06,
"loss": 0.0784,
"step": 413
},
{
"epoch": 1.6863543788187374,
"grad_norm": 0.5422511517223145,
"learning_rate": 4.774319284555833e-06,
"loss": 0.0744,
"step": 414
},
{
"epoch": 1.690427698574338,
"grad_norm": 0.617398792027579,
"learning_rate": 4.7505822036363214e-06,
"loss": 0.0732,
"step": 415
},
{
"epoch": 1.694501018329939,
"grad_norm": 0.9403552399900484,
"learning_rate": 4.726850756802486e-06,
"loss": 0.085,
"step": 416
},
{
"epoch": 1.6985743380855398,
"grad_norm": 0.5589577658327266,
"learning_rate": 4.703125480122747e-06,
"loss": 0.07,
"step": 417
},
{
"epoch": 1.7026476578411405,
"grad_norm": 0.8384307665718973,
"learning_rate": 4.679406909526147e-06,
"loss": 0.1021,
"step": 418
},
{
"epoch": 1.7067209775967414,
"grad_norm": 0.6378594747202708,
"learning_rate": 4.655695580790254e-06,
"loss": 0.0826,
"step": 419
},
{
"epoch": 1.7107942973523422,
"grad_norm": 0.9335835620964789,
"learning_rate": 4.631992029529037e-06,
"loss": 0.1112,
"step": 420
},
{
"epoch": 1.7148676171079429,
"grad_norm": 0.6170098686203401,
"learning_rate": 4.608296791180793e-06,
"loss": 0.0848,
"step": 421
},
{
"epoch": 1.7189409368635438,
"grad_norm": 0.6186288012037354,
"learning_rate": 4.584610400996028e-06,
"loss": 0.0795,
"step": 422
},
{
"epoch": 1.7230142566191446,
"grad_norm": 0.7591412034838235,
"learning_rate": 4.560933394025386e-06,
"loss": 0.0865,
"step": 423
},
{
"epoch": 1.7270875763747453,
"grad_norm": 0.6645225808347813,
"learning_rate": 4.537266305107549e-06,
"loss": 0.087,
"step": 424
},
{
"epoch": 1.7311608961303462,
"grad_norm": 0.5774310704171809,
"learning_rate": 4.513609668857162e-06,
"loss": 0.0683,
"step": 425
},
{
"epoch": 1.735234215885947,
"grad_norm": 0.5396004853223214,
"learning_rate": 4.489964019652752e-06,
"loss": 0.0645,
"step": 426
},
{
"epoch": 1.7393075356415477,
"grad_norm": 0.8193272698607866,
"learning_rate": 4.4663298916246665e-06,
"loss": 0.0966,
"step": 427
},
{
"epoch": 1.7433808553971488,
"grad_norm": 0.8438620683447817,
"learning_rate": 4.442707818642999e-06,
"loss": 0.0862,
"step": 428
},
{
"epoch": 1.7474541751527495,
"grad_norm": 0.8986877868414216,
"learning_rate": 4.419098334305529e-06,
"loss": 0.0909,
"step": 429
},
{
"epoch": 1.7515274949083504,
"grad_norm": 0.7300173025939761,
"learning_rate": 4.395501971925677e-06,
"loss": 0.0807,
"step": 430
},
{
"epoch": 1.7556008146639512,
"grad_norm": 0.8080156287729479,
"learning_rate": 4.371919264520449e-06,
"loss": 0.0865,
"step": 431
},
{
"epoch": 1.759674134419552,
"grad_norm": 0.8071100091965066,
"learning_rate": 4.348350744798399e-06,
"loss": 0.0845,
"step": 432
},
{
"epoch": 1.7637474541751528,
"grad_norm": 0.6345944880929723,
"learning_rate": 4.324796945147598e-06,
"loss": 0.0746,
"step": 433
},
{
"epoch": 1.7678207739307537,
"grad_norm": 0.7862831637321498,
"learning_rate": 4.301258397623606e-06,
"loss": 0.0876,
"step": 434
},
{
"epoch": 1.7718940936863543,
"grad_norm": 0.7699681856032118,
"learning_rate": 4.2777356339374526e-06,
"loss": 0.0919,
"step": 435
},
{
"epoch": 1.7759674134419552,
"grad_norm": 1.260183839778813,
"learning_rate": 4.254229185443628e-06,
"loss": 0.0864,
"step": 436
},
{
"epoch": 1.780040733197556,
"grad_norm": 0.6829744950654246,
"learning_rate": 4.230739583128078e-06,
"loss": 0.0746,
"step": 437
},
{
"epoch": 1.7841140529531567,
"grad_norm": 1.2716252017981826,
"learning_rate": 4.2072673575962125e-06,
"loss": 0.1001,
"step": 438
},
{
"epoch": 1.7881873727087576,
"grad_norm": 0.997750452895151,
"learning_rate": 4.183813039060919e-06,
"loss": 0.1182,
"step": 439
},
{
"epoch": 1.7922606924643585,
"grad_norm": 0.6022378654035607,
"learning_rate": 4.160377157330579e-06,
"loss": 0.0851,
"step": 440
},
{
"epoch": 1.7963340122199591,
"grad_norm": 0.7627379348442356,
"learning_rate": 4.136960241797113e-06,
"loss": 0.0675,
"step": 441
},
{
"epoch": 1.8004073319755602,
"grad_norm": 0.7651666287124916,
"learning_rate": 4.113562821424012e-06,
"loss": 0.0916,
"step": 442
},
{
"epoch": 1.804480651731161,
"grad_norm": 0.7330363703477193,
"learning_rate": 4.090185424734392e-06,
"loss": 0.0838,
"step": 443
},
{
"epoch": 1.8085539714867616,
"grad_norm": 0.7101706692426837,
"learning_rate": 4.066828579799054e-06,
"loss": 0.0791,
"step": 444
},
{
"epoch": 1.8126272912423627,
"grad_norm": 0.5688223211779385,
"learning_rate": 4.043492814224559e-06,
"loss": 0.0709,
"step": 445
},
{
"epoch": 1.8167006109979633,
"grad_norm": 0.8376097479341676,
"learning_rate": 4.020178655141307e-06,
"loss": 0.0826,
"step": 446
},
{
"epoch": 1.8207739307535642,
"grad_norm": 0.7808635217722791,
"learning_rate": 3.9968866291916254e-06,
"loss": 0.0895,
"step": 447
},
{
"epoch": 1.824847250509165,
"grad_norm": 0.6170736835701749,
"learning_rate": 3.973617262517886e-06,
"loss": 0.0682,
"step": 448
},
{
"epoch": 1.8289205702647657,
"grad_norm": 0.7757035878498671,
"learning_rate": 3.950371080750602e-06,
"loss": 0.0856,
"step": 449
},
{
"epoch": 1.8329938900203666,
"grad_norm": 1.0217188455954884,
"learning_rate": 3.927148608996569e-06,
"loss": 0.1131,
"step": 450
},
{
"epoch": 1.8370672097759675,
"grad_norm": 0.9683593480288422,
"learning_rate": 3.903950371827001e-06,
"loss": 0.0921,
"step": 451
},
{
"epoch": 1.8411405295315681,
"grad_norm": 0.6337703466634409,
"learning_rate": 3.880776893265673e-06,
"loss": 0.0827,
"step": 452
},
{
"epoch": 1.845213849287169,
"grad_norm": 0.7361682132709152,
"learning_rate": 3.85762869677709e-06,
"loss": 0.0636,
"step": 453
},
{
"epoch": 1.84928716904277,
"grad_norm": 0.7754176214282682,
"learning_rate": 3.834506305254667e-06,
"loss": 0.0961,
"step": 454
},
{
"epoch": 1.8533604887983706,
"grad_norm": 0.828382619398843,
"learning_rate": 3.811410241008902e-06,
"loss": 0.0951,
"step": 455
},
{
"epoch": 1.8574338085539714,
"grad_norm": 0.5945605581238207,
"learning_rate": 3.788341025755595e-06,
"loss": 0.0763,
"step": 456
},
{
"epoch": 1.8615071283095723,
"grad_norm": 0.7934942467804943,
"learning_rate": 3.765299180604055e-06,
"loss": 0.0934,
"step": 457
},
{
"epoch": 1.865580448065173,
"grad_norm": 0.6628615018421619,
"learning_rate": 3.7422852260453274e-06,
"loss": 0.0753,
"step": 458
},
{
"epoch": 1.869653767820774,
"grad_norm": 0.8426888513729046,
"learning_rate": 3.719299681940437e-06,
"loss": 0.0983,
"step": 459
},
{
"epoch": 1.8737270875763747,
"grad_norm": 0.8833092682930523,
"learning_rate": 3.696343067508651e-06,
"loss": 0.0841,
"step": 460
},
{
"epoch": 1.8778004073319754,
"grad_norm": 0.5875978979383154,
"learning_rate": 3.673415901315743e-06,
"loss": 0.0693,
"step": 461
},
{
"epoch": 1.8818737270875765,
"grad_norm": 0.6322423557535214,
"learning_rate": 3.650518701262278e-06,
"loss": 0.0781,
"step": 462
},
{
"epoch": 1.8859470468431772,
"grad_norm": 0.7821903785073223,
"learning_rate": 3.6276519845719237e-06,
"loss": 0.077,
"step": 463
},
{
"epoch": 1.890020366598778,
"grad_norm": 0.8674854155144539,
"learning_rate": 3.6048162677797595e-06,
"loss": 0.0927,
"step": 464
},
{
"epoch": 1.894093686354379,
"grad_norm": 0.59647645188795,
"learning_rate": 3.582012066720605e-06,
"loss": 0.0718,
"step": 465
},
{
"epoch": 1.8981670061099796,
"grad_norm": 0.6297079342542728,
"learning_rate": 3.559239896517379e-06,
"loss": 0.0854,
"step": 466
},
{
"epoch": 1.9022403258655805,
"grad_norm": 1.041008043566478,
"learning_rate": 3.536500271569452e-06,
"loss": 0.0873,
"step": 467
},
{
"epoch": 1.9063136456211813,
"grad_norm": 0.6843579823593634,
"learning_rate": 3.5137937055410343e-06,
"loss": 0.0716,
"step": 468
},
{
"epoch": 1.910386965376782,
"grad_norm": 0.8825368664636357,
"learning_rate": 3.4911207113495703e-06,
"loss": 0.0955,
"step": 469
},
{
"epoch": 1.9144602851323829,
"grad_norm": 0.7381839178415016,
"learning_rate": 3.4684818011541484e-06,
"loss": 0.0815,
"step": 470
},
{
"epoch": 1.9185336048879837,
"grad_norm": 0.7678147459965866,
"learning_rate": 3.4458774863439366e-06,
"loss": 0.0895,
"step": 471
},
{
"epoch": 1.9226069246435844,
"grad_norm": 0.8220888342463925,
"learning_rate": 3.423308277526633e-06,
"loss": 0.0914,
"step": 472
},
{
"epoch": 1.9266802443991853,
"grad_norm": 0.7021908890281842,
"learning_rate": 3.4007746845169253e-06,
"loss": 0.0906,
"step": 473
},
{
"epoch": 1.9307535641547862,
"grad_norm": 0.7284906476786358,
"learning_rate": 3.3782772163249767e-06,
"loss": 0.0725,
"step": 474
},
{
"epoch": 1.9348268839103868,
"grad_norm": 0.5533573194210741,
"learning_rate": 3.3558163811449317e-06,
"loss": 0.0731,
"step": 475
},
{
"epoch": 1.938900203665988,
"grad_norm": 0.9405832752114753,
"learning_rate": 3.3333926863434317e-06,
"loss": 0.0962,
"step": 476
},
{
"epoch": 1.9429735234215886,
"grad_norm": 0.6819165425758139,
"learning_rate": 3.311006638448155e-06,
"loss": 0.0767,
"step": 477
},
{
"epoch": 1.9470468431771895,
"grad_norm": 0.6845642751352256,
"learning_rate": 3.288658743136378e-06,
"loss": 0.0695,
"step": 478
},
{
"epoch": 1.9511201629327903,
"grad_norm": 1.012736448173238,
"learning_rate": 3.2663495052235505e-06,
"loss": 0.1222,
"step": 479
},
{
"epoch": 1.955193482688391,
"grad_norm": 0.6928154487833174,
"learning_rate": 3.2440794286518896e-06,
"loss": 0.0903,
"step": 480
},
{
"epoch": 1.9592668024439919,
"grad_norm": 0.6484293467201061,
"learning_rate": 3.2218490164790015e-06,
"loss": 0.0696,
"step": 481
},
{
"epoch": 1.9633401221995928,
"grad_norm": 0.6538882594004228,
"learning_rate": 3.199658770866515e-06,
"loss": 0.0787,
"step": 482
},
{
"epoch": 1.9674134419551934,
"grad_norm": 0.5752828743925392,
"learning_rate": 3.1775091930687374e-06,
"loss": 0.0675,
"step": 483
},
{
"epoch": 1.9714867617107943,
"grad_norm": 0.5743945825655228,
"learning_rate": 3.1554007834213357e-06,
"loss": 0.0715,
"step": 484
},
{
"epoch": 1.9755600814663952,
"grad_norm": 0.8159794403070233,
"learning_rate": 3.1333340413300263e-06,
"loss": 0.0921,
"step": 485
},
{
"epoch": 1.9796334012219958,
"grad_norm": 0.6636547264511075,
"learning_rate": 3.1113094652593023e-06,
"loss": 0.0759,
"step": 486
},
{
"epoch": 1.9837067209775967,
"grad_norm": 1.0216030221101335,
"learning_rate": 3.0893275527211742e-06,
"loss": 0.1142,
"step": 487
},
{
"epoch": 1.9877800407331976,
"grad_norm": 0.6661497473358808,
"learning_rate": 3.067388800263923e-06,
"loss": 0.0861,
"step": 488
},
{
"epoch": 1.9918533604887982,
"grad_norm": 0.6220758643441348,
"learning_rate": 3.04549370346089e-06,
"loss": 0.0724,
"step": 489
},
{
"epoch": 1.9959266802443993,
"grad_norm": 0.6380698066205512,
"learning_rate": 3.0236427568992845e-06,
"loss": 0.081,
"step": 490
},
{
"epoch": 2.0,
"grad_norm": 1.0732156070255234,
"learning_rate": 3.0018364541690048e-06,
"loss": 0.0967,
"step": 491
},
{
"epoch": 2.0040733197556007,
"grad_norm": 0.46382587645965473,
"learning_rate": 2.9800752878514903e-06,
"loss": 0.0486,
"step": 492
},
{
"epoch": 2.0081466395112018,
"grad_norm": 0.6061183961253724,
"learning_rate": 2.958359749508603e-06,
"loss": 0.0551,
"step": 493
},
{
"epoch": 2.0122199592668024,
"grad_norm": 0.4921733736983431,
"learning_rate": 2.936690329671511e-06,
"loss": 0.0454,
"step": 494
},
{
"epoch": 2.016293279022403,
"grad_norm": 0.43183509951355314,
"learning_rate": 2.915067517829615e-06,
"loss": 0.0466,
"step": 495
},
{
"epoch": 2.020366598778004,
"grad_norm": 0.49575364629777624,
"learning_rate": 2.893491802419492e-06,
"loss": 0.049,
"step": 496
},
{
"epoch": 2.024439918533605,
"grad_norm": 0.4271031180271253,
"learning_rate": 2.871963670813861e-06,
"loss": 0.0393,
"step": 497
},
{
"epoch": 2.0285132382892055,
"grad_norm": 0.5825597063224182,
"learning_rate": 2.850483609310567e-06,
"loss": 0.0439,
"step": 498
},
{
"epoch": 2.0325865580448066,
"grad_norm": 0.7134060092750841,
"learning_rate": 2.829052103121611e-06,
"loss": 0.0507,
"step": 499
},
{
"epoch": 2.0366598778004072,
"grad_norm": 0.5359846386854068,
"learning_rate": 2.807669636362169e-06,
"loss": 0.0465,
"step": 500
},
{
"epoch": 2.0407331975560083,
"grad_norm": 1.2307395730228317,
"learning_rate": 2.7863366920396805e-06,
"loss": 0.0609,
"step": 501
},
{
"epoch": 2.044806517311609,
"grad_norm": 0.7016648303052435,
"learning_rate": 2.765053752042915e-06,
"loss": 0.0542,
"step": 502
},
{
"epoch": 2.0488798370672097,
"grad_norm": 0.6321598695043944,
"learning_rate": 2.7438212971311016e-06,
"loss": 0.0462,
"step": 503
},
{
"epoch": 2.0529531568228108,
"grad_norm": 0.524422558245556,
"learning_rate": 2.722639806923066e-06,
"loss": 0.0387,
"step": 504
},
{
"epoch": 2.0570264765784114,
"grad_norm": 0.8384795501307071,
"learning_rate": 2.7015097598863906e-06,
"loss": 0.0534,
"step": 505
},
{
"epoch": 2.061099796334012,
"grad_norm": 0.6240843386469663,
"learning_rate": 2.680431633326614e-06,
"loss": 0.0412,
"step": 506
},
{
"epoch": 2.065173116089613,
"grad_norm": 0.6307360193841159,
"learning_rate": 2.659405903376442e-06,
"loss": 0.0488,
"step": 507
},
{
"epoch": 2.069246435845214,
"grad_norm": 0.6330618915300124,
"learning_rate": 2.6384330449850028e-06,
"loss": 0.0502,
"step": 508
},
{
"epoch": 2.0733197556008145,
"grad_norm": 0.5880527544333638,
"learning_rate": 2.617513531907103e-06,
"loss": 0.0494,
"step": 509
},
{
"epoch": 2.0773930753564156,
"grad_norm": 0.7335711218887413,
"learning_rate": 2.5966478366925406e-06,
"loss": 0.052,
"step": 510
},
{
"epoch": 2.0814663951120163,
"grad_norm": 0.8141104911415494,
"learning_rate": 2.5758364306754247e-06,
"loss": 0.0513,
"step": 511
},
{
"epoch": 2.085539714867617,
"grad_norm": 0.7214645518132131,
"learning_rate": 2.5550797839635283e-06,
"loss": 0.0459,
"step": 512
},
{
"epoch": 2.089613034623218,
"grad_norm": 0.5903300427666225,
"learning_rate": 2.5343783654276644e-06,
"loss": 0.0466,
"step": 513
},
{
"epoch": 2.0936863543788187,
"grad_norm": 0.6287638475587176,
"learning_rate": 2.5137326426911067e-06,
"loss": 0.0457,
"step": 514
},
{
"epoch": 2.0977596741344193,
"grad_norm": 0.6016948711437964,
"learning_rate": 2.493143082119013e-06,
"loss": 0.0474,
"step": 515
},
{
"epoch": 2.1018329938900204,
"grad_norm": 0.7019678605348776,
"learning_rate": 2.472610148807903e-06,
"loss": 0.0471,
"step": 516
},
{
"epoch": 2.105906313645621,
"grad_norm": 0.6650222349525898,
"learning_rate": 2.452134306575139e-06,
"loss": 0.047,
"step": 517
},
{
"epoch": 2.109979633401222,
"grad_norm": 0.6319951154070054,
"learning_rate": 2.431716017948462e-06,
"loss": 0.0541,
"step": 518
},
{
"epoch": 2.114052953156823,
"grad_norm": 0.9156490212278536,
"learning_rate": 2.4113557441555384e-06,
"loss": 0.042,
"step": 519
},
{
"epoch": 2.1181262729124235,
"grad_norm": 0.5908911924587066,
"learning_rate": 2.391053945113533e-06,
"loss": 0.0384,
"step": 520
},
{
"epoch": 2.1221995926680246,
"grad_norm": 0.5997079936401813,
"learning_rate": 2.370811079418735e-06,
"loss": 0.049,
"step": 521
},
{
"epoch": 2.1262729124236253,
"grad_norm": 0.8224985458683419,
"learning_rate": 2.350627604336186e-06,
"loss": 0.053,
"step": 522
},
{
"epoch": 2.130346232179226,
"grad_norm": 0.6710057271136974,
"learning_rate": 2.330503975789361e-06,
"loss": 0.0515,
"step": 523
},
{
"epoch": 2.134419551934827,
"grad_norm": 0.6977361731485422,
"learning_rate": 2.3104406483498593e-06,
"loss": 0.0466,
"step": 524
},
{
"epoch": 2.1384928716904277,
"grad_norm": 0.7318716291498116,
"learning_rate": 2.290438075227146e-06,
"loss": 0.0459,
"step": 525
},
{
"epoch": 2.1425661914460283,
"grad_norm": 0.6417195973672337,
"learning_rate": 2.270496708258309e-06,
"loss": 0.0486,
"step": 526
},
{
"epoch": 2.1466395112016294,
"grad_norm": 0.7278961564752641,
"learning_rate": 2.2506169978978543e-06,
"loss": 0.0485,
"step": 527
},
{
"epoch": 2.15071283095723,
"grad_norm": 0.6405057377374997,
"learning_rate": 2.230799393207526e-06,
"loss": 0.047,
"step": 528
},
{
"epoch": 2.1547861507128308,
"grad_norm": 0.6383506826364762,
"learning_rate": 2.2110443418461723e-06,
"loss": 0.0383,
"step": 529
},
{
"epoch": 2.158859470468432,
"grad_norm": 0.6064990723637974,
"learning_rate": 2.191352290059621e-06,
"loss": 0.0437,
"step": 530
},
{
"epoch": 2.1629327902240325,
"grad_norm": 0.772798038718517,
"learning_rate": 2.171723682670613e-06,
"loss": 0.0582,
"step": 531
},
{
"epoch": 2.167006109979633,
"grad_norm": 0.7039586643696689,
"learning_rate": 2.152158963068739e-06,
"loss": 0.0444,
"step": 532
},
{
"epoch": 2.1710794297352343,
"grad_norm": 0.5636185384959056,
"learning_rate": 2.1326585732004384e-06,
"loss": 0.0431,
"step": 533
},
{
"epoch": 2.175152749490835,
"grad_norm": 0.6483499440908992,
"learning_rate": 2.1132229535590092e-06,
"loss": 0.0509,
"step": 534
},
{
"epoch": 2.179226069246436,
"grad_norm": 0.5790468097555252,
"learning_rate": 2.093852543174652e-06,
"loss": 0.045,
"step": 535
},
{
"epoch": 2.1832993890020367,
"grad_norm": 0.5662346010696749,
"learning_rate": 2.0745477796045664e-06,
"loss": 0.0421,
"step": 536
},
{
"epoch": 2.1873727087576373,
"grad_norm": 0.9206343975668372,
"learning_rate": 2.0553090989230527e-06,
"loss": 0.0479,
"step": 537
},
{
"epoch": 2.1914460285132384,
"grad_norm": 0.657682562867677,
"learning_rate": 2.036136935711674e-06,
"loss": 0.0393,
"step": 538
},
{
"epoch": 2.195519348268839,
"grad_norm": 0.5162086561290653,
"learning_rate": 2.017031723049432e-06,
"loss": 0.0348,
"step": 539
},
{
"epoch": 2.1995926680244398,
"grad_norm": 0.7166136712411706,
"learning_rate": 1.997993892502983e-06,
"loss": 0.0478,
"step": 540
},
{
"epoch": 2.203665987780041,
"grad_norm": 0.6420851301612974,
"learning_rate": 1.979023874116895e-06,
"loss": 0.0504,
"step": 541
},
{
"epoch": 2.2077393075356415,
"grad_norm": 0.5674151016939644,
"learning_rate": 1.9601220964039324e-06,
"loss": 0.0418,
"step": 542
},
{
"epoch": 2.211812627291242,
"grad_norm": 0.6248094181782853,
"learning_rate": 1.9412889863353683e-06,
"loss": 0.0464,
"step": 543
},
{
"epoch": 2.2158859470468433,
"grad_norm": 0.6115409983253244,
"learning_rate": 1.9225249693313547e-06,
"loss": 0.0422,
"step": 544
},
{
"epoch": 2.219959266802444,
"grad_norm": 0.5373674294031344,
"learning_rate": 1.9038304692512943e-06,
"loss": 0.0447,
"step": 545
},
{
"epoch": 2.224032586558045,
"grad_norm": 0.6095403003841355,
"learning_rate": 1.8852059083842838e-06,
"loss": 0.0447,
"step": 546
},
{
"epoch": 2.2281059063136457,
"grad_norm": 0.6108027242445478,
"learning_rate": 1.8666517074395607e-06,
"loss": 0.0517,
"step": 547
},
{
"epoch": 2.2321792260692463,
"grad_norm": 0.47923271839833664,
"learning_rate": 1.8481682855370098e-06,
"loss": 0.0342,
"step": 548
},
{
"epoch": 2.2362525458248474,
"grad_norm": 0.6016755510602564,
"learning_rate": 1.829756060197692e-06,
"loss": 0.0417,
"step": 549
},
{
"epoch": 2.240325865580448,
"grad_norm": 0.7022430834159812,
"learning_rate": 1.8114154473344081e-06,
"loss": 0.0481,
"step": 550
},
{
"epoch": 2.2443991853360488,
"grad_norm": 0.5768106404539921,
"learning_rate": 1.7931468612423142e-06,
"loss": 0.039,
"step": 551
},
{
"epoch": 2.24847250509165,
"grad_norm": 0.6323933887994607,
"learning_rate": 1.7749507145895518e-06,
"loss": 0.0415,
"step": 552
},
{
"epoch": 2.2525458248472505,
"grad_norm": 0.706898611038952,
"learning_rate": 1.756827418407936e-06,
"loss": 0.0447,
"step": 553
},
{
"epoch": 2.256619144602851,
"grad_norm": 0.6753433476547962,
"learning_rate": 1.7387773820836668e-06,
"loss": 0.0435,
"step": 554
},
{
"epoch": 2.2606924643584523,
"grad_norm": 0.6345110906473638,
"learning_rate": 1.7208010133480751e-06,
"loss": 0.0408,
"step": 555
},
{
"epoch": 2.264765784114053,
"grad_norm": 0.6349996250849919,
"learning_rate": 1.7028987182684248e-06,
"loss": 0.0467,
"step": 556
},
{
"epoch": 2.2688391038696536,
"grad_norm": 0.5478171278806082,
"learning_rate": 1.6850709012387328e-06,
"loss": 0.0421,
"step": 557
},
{
"epoch": 2.2729124236252547,
"grad_norm": 0.5570382415949389,
"learning_rate": 1.6673179649706312e-06,
"loss": 0.0412,
"step": 558
},
{
"epoch": 2.2769857433808554,
"grad_norm": 0.6704892257058723,
"learning_rate": 1.64964031048428e-06,
"loss": 0.0432,
"step": 559
},
{
"epoch": 2.281059063136456,
"grad_norm": 0.5353427142605596,
"learning_rate": 1.632038337099297e-06,
"loss": 0.0421,
"step": 560
},
{
"epoch": 2.285132382892057,
"grad_norm": 0.5675614565902064,
"learning_rate": 1.6145124424257497e-06,
"loss": 0.0408,
"step": 561
},
{
"epoch": 2.2892057026476578,
"grad_norm": 0.6215446356434404,
"learning_rate": 1.5970630223551614e-06,
"loss": 0.0441,
"step": 562
},
{
"epoch": 2.293279022403259,
"grad_norm": 0.5312552367132775,
"learning_rate": 1.5796904710515792e-06,
"loss": 0.0374,
"step": 563
},
{
"epoch": 2.2973523421588595,
"grad_norm": 0.6711448627336817,
"learning_rate": 1.5623951809426663e-06,
"loss": 0.0466,
"step": 564
},
{
"epoch": 2.30142566191446,
"grad_norm": 0.7355915752823369,
"learning_rate": 1.5451775427108302e-06,
"loss": 0.0409,
"step": 565
},
{
"epoch": 2.3054989816700613,
"grad_norm": 0.5492792440607698,
"learning_rate": 1.5280379452844124e-06,
"loss": 0.0406,
"step": 566
},
{
"epoch": 2.309572301425662,
"grad_norm": 0.5646796515852444,
"learning_rate": 1.510976775828887e-06,
"loss": 0.0384,
"step": 567
},
{
"epoch": 2.3136456211812626,
"grad_norm": 0.5825549184664148,
"learning_rate": 1.493994419738129e-06,
"loss": 0.041,
"step": 568
},
{
"epoch": 2.3177189409368637,
"grad_norm": 0.5817560681286719,
"learning_rate": 1.4770912606257003e-06,
"loss": 0.0499,
"step": 569
},
{
"epoch": 2.3217922606924644,
"grad_norm": 0.9844440093209132,
"learning_rate": 1.4602676803161842e-06,
"loss": 0.0517,
"step": 570
},
{
"epoch": 2.325865580448065,
"grad_norm": 0.6142006283022792,
"learning_rate": 1.4435240588365645e-06,
"loss": 0.0406,
"step": 571
},
{
"epoch": 2.329938900203666,
"grad_norm": 0.5693932954149746,
"learning_rate": 1.4268607744076419e-06,
"loss": 0.0407,
"step": 572
},
{
"epoch": 2.3340122199592668,
"grad_norm": 0.5950772597911713,
"learning_rate": 1.41027820343548e-06,
"loss": 0.048,
"step": 573
},
{
"epoch": 2.3380855397148674,
"grad_norm": 0.7237495768701298,
"learning_rate": 1.3937767205029196e-06,
"loss": 0.0549,
"step": 574
},
{
"epoch": 2.3421588594704685,
"grad_norm": 0.67988274483473,
"learning_rate": 1.3773566983610992e-06,
"loss": 0.0523,
"step": 575
},
{
"epoch": 2.346232179226069,
"grad_norm": 0.8056627345317472,
"learning_rate": 1.3610185079210514e-06,
"loss": 0.0426,
"step": 576
},
{
"epoch": 2.35030549898167,
"grad_norm": 0.6115391073948011,
"learning_rate": 1.34476251824531e-06,
"loss": 0.0419,
"step": 577
},
{
"epoch": 2.354378818737271,
"grad_norm": 0.6092462729328992,
"learning_rate": 1.3285890965395853e-06,
"loss": 0.0451,
"step": 578
},
{
"epoch": 2.3584521384928716,
"grad_norm": 0.5776725050326151,
"learning_rate": 1.3124986081444625e-06,
"loss": 0.0471,
"step": 579
},
{
"epoch": 2.3625254582484727,
"grad_norm": 0.6555280480435673,
"learning_rate": 1.296491416527147e-06,
"loss": 0.0449,
"step": 580
},
{
"epoch": 2.3665987780040734,
"grad_norm": 0.6569585284725156,
"learning_rate": 1.2805678832732627e-06,
"loss": 0.0546,
"step": 581
},
{
"epoch": 2.370672097759674,
"grad_norm": 0.7843231218862423,
"learning_rate": 1.264728368078678e-06,
"loss": 0.0447,
"step": 582
},
{
"epoch": 2.374745417515275,
"grad_norm": 0.5190101550736153,
"learning_rate": 1.248973228741378e-06,
"loss": 0.0444,
"step": 583
},
{
"epoch": 2.378818737270876,
"grad_norm": 0.5731375070374565,
"learning_rate": 1.2333028211533916e-06,
"loss": 0.0402,
"step": 584
},
{
"epoch": 2.3828920570264764,
"grad_norm": 0.7216183443132362,
"learning_rate": 1.21771749929274e-06,
"loss": 0.047,
"step": 585
},
{
"epoch": 2.3869653767820775,
"grad_norm": 0.6574547850481611,
"learning_rate": 1.2022176152154525e-06,
"loss": 0.0452,
"step": 586
},
{
"epoch": 2.391038696537678,
"grad_norm": 0.6995633619657223,
"learning_rate": 1.1868035190476085e-06,
"loss": 0.0437,
"step": 587
},
{
"epoch": 2.395112016293279,
"grad_norm": 0.5381295030205419,
"learning_rate": 1.1714755589774252e-06,
"loss": 0.0374,
"step": 588
},
{
"epoch": 2.39918533604888,
"grad_norm": 0.5193316596822176,
"learning_rate": 1.1562340812474004e-06,
"loss": 0.0335,
"step": 589
},
{
"epoch": 2.4032586558044806,
"grad_norm": 0.558050549263979,
"learning_rate": 1.1410794301464817e-06,
"loss": 0.04,
"step": 590
},
{
"epoch": 2.4073319755600817,
"grad_norm": 0.6124096552105914,
"learning_rate": 1.1260119480023008e-06,
"loss": 0.0365,
"step": 591
},
{
"epoch": 2.4114052953156824,
"grad_norm": 0.5294199833428392,
"learning_rate": 1.1110319751734271e-06,
"loss": 0.0386,
"step": 592
},
{
"epoch": 2.415478615071283,
"grad_norm": 0.6469614215212903,
"learning_rate": 1.0961398500416926e-06,
"loss": 0.042,
"step": 593
},
{
"epoch": 2.4195519348268837,
"grad_norm": 0.5557240307053494,
"learning_rate": 1.0813359090045412e-06,
"loss": 0.0399,
"step": 594
},
{
"epoch": 2.423625254582485,
"grad_norm": 0.6300256880608739,
"learning_rate": 1.0666204864674263e-06,
"loss": 0.0388,
"step": 595
},
{
"epoch": 2.4276985743380854,
"grad_norm": 0.655959102130783,
"learning_rate": 1.0519939148362667e-06,
"loss": 0.0478,
"step": 596
},
{
"epoch": 2.4317718940936865,
"grad_norm": 0.6094559864136577,
"learning_rate": 1.0374565245099328e-06,
"loss": 0.0388,
"step": 597
},
{
"epoch": 2.435845213849287,
"grad_norm": 0.7411603450389928,
"learning_rate": 1.0230086438727771e-06,
"loss": 0.0504,
"step": 598
},
{
"epoch": 2.439918533604888,
"grad_norm": 0.5681006917233394,
"learning_rate": 1.0086505992872304e-06,
"loss": 0.0427,
"step": 599
},
{
"epoch": 2.443991853360489,
"grad_norm": 0.609417026917768,
"learning_rate": 9.943827150864143e-07,
"loss": 0.0396,
"step": 600
},
{
"epoch": 2.4480651731160896,
"grad_norm": 0.7715933182445011,
"learning_rate": 9.80205313566827e-07,
"loss": 0.0479,
"step": 601
},
{
"epoch": 2.4521384928716903,
"grad_norm": 0.5588281033029895,
"learning_rate": 9.66118714981058e-07,
"loss": 0.0468,
"step": 602
},
{
"epoch": 2.4562118126272914,
"grad_norm": 0.6202480044275774,
"learning_rate": 9.521232375305494e-07,
"loss": 0.0428,
"step": 603
},
{
"epoch": 2.460285132382892,
"grad_norm": 0.5549990136472819,
"learning_rate": 9.382191973584193e-07,
"loss": 0.0321,
"step": 604
},
{
"epoch": 2.4643584521384927,
"grad_norm": 0.7125183614456646,
"learning_rate": 9.244069085423074e-07,
"loss": 0.052,
"step": 605
},
{
"epoch": 2.468431771894094,
"grad_norm": 0.5565747141010096,
"learning_rate": 9.106866830872929e-07,
"loss": 0.0366,
"step": 606
},
{
"epoch": 2.4725050916496945,
"grad_norm": 0.616895795551855,
"learning_rate": 8.970588309188343e-07,
"loss": 0.0438,
"step": 607
},
{
"epoch": 2.4765784114052956,
"grad_norm": 0.7071703538896843,
"learning_rate": 8.835236598757796e-07,
"loss": 0.0437,
"step": 608
},
{
"epoch": 2.480651731160896,
"grad_norm": 0.550733977240549,
"learning_rate": 8.70081475703406e-07,
"loss": 0.0355,
"step": 609
},
{
"epoch": 2.484725050916497,
"grad_norm": 0.9168151315445275,
"learning_rate": 8.567325820465156e-07,
"loss": 0.0536,
"step": 610
},
{
"epoch": 2.4887983706720975,
"grad_norm": 0.5877013322142949,
"learning_rate": 8.434772804425734e-07,
"loss": 0.0401,
"step": 611
},
{
"epoch": 2.4928716904276986,
"grad_norm": 0.6525278969423508,
"learning_rate": 8.303158703149023e-07,
"loss": 0.0434,
"step": 612
},
{
"epoch": 2.4969450101832993,
"grad_norm": 0.5668632582217858,
"learning_rate": 8.172486489659115e-07,
"loss": 0.0405,
"step": 613
},
{
"epoch": 2.5010183299389004,
"grad_norm": 0.5767158406225441,
"learning_rate": 8.042759115703891e-07,
"loss": 0.0439,
"step": 614
},
{
"epoch": 2.505091649694501,
"grad_norm": 0.7263909107216523,
"learning_rate": 7.913979511688252e-07,
"loss": 0.039,
"step": 615
},
{
"epoch": 2.5091649694501017,
"grad_norm": 0.6388219751227069,
"learning_rate": 7.78615058660801e-07,
"loss": 0.0489,
"step": 616
},
{
"epoch": 2.513238289205703,
"grad_norm": 0.5854521751719799,
"learning_rate": 7.659275227984142e-07,
"loss": 0.0389,
"step": 617
},
{
"epoch": 2.5173116089613035,
"grad_norm": 0.5305304944305677,
"learning_rate": 7.533356301797523e-07,
"loss": 0.0411,
"step": 618
},
{
"epoch": 2.521384928716904,
"grad_norm": 0.557310110228571,
"learning_rate": 7.408396652424271e-07,
"loss": 0.0353,
"step": 619
},
{
"epoch": 2.525458248472505,
"grad_norm": 0.7836187943265139,
"learning_rate": 7.28439910257141e-07,
"loss": 0.0564,
"step": 620
},
{
"epoch": 2.529531568228106,
"grad_norm": 0.6076346237905078,
"learning_rate": 7.161366453213181e-07,
"loss": 0.0388,
"step": 621
},
{
"epoch": 2.5336048879837065,
"grad_norm": 0.5905401181549683,
"learning_rate": 7.03930148352771e-07,
"loss": 0.0426,
"step": 622
},
{
"epoch": 2.5376782077393076,
"grad_norm": 0.5845080037540452,
"learning_rate": 6.918206950834283e-07,
"loss": 0.0439,
"step": 623
},
{
"epoch": 2.5417515274949083,
"grad_norm": 0.6528843977830857,
"learning_rate": 6.798085590531012e-07,
"loss": 0.0424,
"step": 624
},
{
"epoch": 2.5458248472505094,
"grad_norm": 0.5060089875025451,
"learning_rate": 6.678940116033095e-07,
"loss": 0.0306,
"step": 625
},
{
"epoch": 2.54989816700611,
"grad_norm": 0.6444888573719301,
"learning_rate": 6.560773218711458e-07,
"loss": 0.0405,
"step": 626
},
{
"epoch": 2.5539714867617107,
"grad_norm": 0.6910140407804498,
"learning_rate": 6.443587567832044e-07,
"loss": 0.0395,
"step": 627
},
{
"epoch": 2.5580448065173114,
"grad_norm": 0.7112890892772754,
"learning_rate": 6.327385810495423e-07,
"loss": 0.0454,
"step": 628
},
{
"epoch": 2.5621181262729125,
"grad_norm": 0.5927299095854747,
"learning_rate": 6.212170571577087e-07,
"loss": 0.0444,
"step": 629
},
{
"epoch": 2.566191446028513,
"grad_norm": 0.5535609388587451,
"learning_rate": 6.097944453668081e-07,
"loss": 0.0406,
"step": 630
},
{
"epoch": 2.5702647657841142,
"grad_norm": 0.5451818995222956,
"learning_rate": 5.984710037016267e-07,
"loss": 0.0374,
"step": 631
},
{
"epoch": 2.574338085539715,
"grad_norm": 0.7054245569688469,
"learning_rate": 5.872469879468024e-07,
"loss": 0.055,
"step": 632
},
{
"epoch": 2.5784114052953155,
"grad_norm": 0.5518787075176712,
"learning_rate": 5.761226516410434e-07,
"loss": 0.0347,
"step": 633
},
{
"epoch": 2.5824847250509166,
"grad_norm": 0.796243379976555,
"learning_rate": 5.650982460714083e-07,
"loss": 0.0481,
"step": 634
},
{
"epoch": 2.5865580448065173,
"grad_norm": 0.7188252233855069,
"learning_rate": 5.54174020267621e-07,
"loss": 0.0443,
"step": 635
},
{
"epoch": 2.5906313645621184,
"grad_norm": 0.5526783617745519,
"learning_rate": 5.433502209964531e-07,
"loss": 0.0396,
"step": 636
},
{
"epoch": 2.594704684317719,
"grad_norm": 0.5825176845240135,
"learning_rate": 5.326270927561444e-07,
"loss": 0.0455,
"step": 637
},
{
"epoch": 2.5987780040733197,
"grad_norm": 0.571850531305161,
"learning_rate": 5.22004877770883e-07,
"loss": 0.041,
"step": 638
},
{
"epoch": 2.6028513238289204,
"grad_norm": 0.6483190905465236,
"learning_rate": 5.114838159853336e-07,
"loss": 0.0438,
"step": 639
},
{
"epoch": 2.6069246435845215,
"grad_norm": 0.6037537099679922,
"learning_rate": 5.010641450592158e-07,
"loss": 0.0437,
"step": 640
},
{
"epoch": 2.610997963340122,
"grad_norm": 0.7252652851456096,
"learning_rate": 4.907461003619346e-07,
"loss": 0.048,
"step": 641
},
{
"epoch": 2.6150712830957232,
"grad_norm": 0.7432983808595361,
"learning_rate": 4.805299149672682e-07,
"loss": 0.0448,
"step": 642
},
{
"epoch": 2.619144602851324,
"grad_norm": 0.5876266367931727,
"learning_rate": 4.7041581964809733e-07,
"loss": 0.0377,
"step": 643
},
{
"epoch": 2.6232179226069245,
"grad_norm": 0.5798950545458074,
"learning_rate": 4.6040404287119924e-07,
"loss": 0.0438,
"step": 644
},
{
"epoch": 2.627291242362525,
"grad_norm": 0.5827291901167612,
"learning_rate": 4.504948107920781e-07,
"loss": 0.0386,
"step": 645
},
{
"epoch": 2.6313645621181263,
"grad_norm": 0.5454690396247462,
"learning_rate": 4.4068834724986466e-07,
"loss": 0.0404,
"step": 646
},
{
"epoch": 2.635437881873727,
"grad_norm": 0.591772305720897,
"learning_rate": 4.309848737622568e-07,
"loss": 0.0411,
"step": 647
},
{
"epoch": 2.639511201629328,
"grad_norm": 0.6030630084243364,
"learning_rate": 4.213846095205126e-07,
"loss": 0.0378,
"step": 648
},
{
"epoch": 2.6435845213849287,
"grad_norm": 0.9035073373124242,
"learning_rate": 4.1188777138450487e-07,
"loss": 0.0637,
"step": 649
},
{
"epoch": 2.6476578411405294,
"grad_norm": 0.6306988274701608,
"learning_rate": 4.024945738778163e-07,
"loss": 0.0455,
"step": 650
},
{
"epoch": 2.6517311608961305,
"grad_norm": 0.5571148701616122,
"learning_rate": 3.9320522918289973e-07,
"loss": 0.0418,
"step": 651
},
{
"epoch": 2.655804480651731,
"grad_norm": 0.6025888229873192,
"learning_rate": 3.8401994713628044e-07,
"loss": 0.0447,
"step": 652
},
{
"epoch": 2.6598778004073322,
"grad_norm": 0.5348586235615737,
"learning_rate": 3.7493893522381866e-07,
"loss": 0.0389,
"step": 653
},
{
"epoch": 2.663951120162933,
"grad_norm": 0.6516858785582209,
"learning_rate": 3.6596239857602136e-07,
"loss": 0.0488,
"step": 654
},
{
"epoch": 2.6680244399185336,
"grad_norm": 0.5588332726110832,
"learning_rate": 3.570905399634111e-07,
"loss": 0.0354,
"step": 655
},
{
"epoch": 2.672097759674134,
"grad_norm": 0.5451393645035304,
"learning_rate": 3.483235597919404e-07,
"loss": 0.0382,
"step": 656
},
{
"epoch": 2.6761710794297353,
"grad_norm": 0.5736021789458674,
"learning_rate": 3.396616560984711e-07,
"loss": 0.0361,
"step": 657
},
{
"epoch": 2.680244399185336,
"grad_norm": 0.7760132004231503,
"learning_rate": 3.31105024546296e-07,
"loss": 0.0376,
"step": 658
},
{
"epoch": 2.684317718940937,
"grad_norm": 0.7396815329059558,
"learning_rate": 3.226538584207228e-07,
"loss": 0.0473,
"step": 659
},
{
"epoch": 2.6883910386965377,
"grad_norm": 0.7964726796467382,
"learning_rate": 3.1430834862470395e-07,
"loss": 0.0603,
"step": 660
},
{
"epoch": 2.6924643584521384,
"grad_norm": 1.5980896143918564,
"learning_rate": 3.0606868367452746e-07,
"loss": 0.0643,
"step": 661
},
{
"epoch": 2.696537678207739,
"grad_norm": 0.6085849621464253,
"learning_rate": 2.9793504969555965e-07,
"loss": 0.0447,
"step": 662
},
{
"epoch": 2.70061099796334,
"grad_norm": 0.5833066393844298,
"learning_rate": 2.899076304180348e-07,
"loss": 0.0412,
"step": 663
},
{
"epoch": 2.704684317718941,
"grad_norm": 0.819223247703981,
"learning_rate": 2.819866071729127e-07,
"loss": 0.0594,
"step": 664
},
{
"epoch": 2.708757637474542,
"grad_norm": 0.6560481778570941,
"learning_rate": 2.7417215888777493e-07,
"loss": 0.0391,
"step": 665
},
{
"epoch": 2.7128309572301426,
"grad_norm": 0.7882666613871305,
"learning_rate": 2.6646446208279054e-07,
"loss": 0.0442,
"step": 666
},
{
"epoch": 2.716904276985743,
"grad_norm": 0.5727739809065823,
"learning_rate": 2.5886369086672193e-07,
"loss": 0.0431,
"step": 667
},
{
"epoch": 2.7209775967413443,
"grad_norm": 0.5899312889497369,
"learning_rate": 2.513700169329963e-07,
"loss": 0.0378,
"step": 668
},
{
"epoch": 2.725050916496945,
"grad_norm": 0.8321164992248397,
"learning_rate": 2.439836095558262e-07,
"loss": 0.052,
"step": 669
},
{
"epoch": 2.729124236252546,
"grad_norm": 0.5571226051042548,
"learning_rate": 2.3670463558638556e-07,
"loss": 0.0392,
"step": 670
},
{
"epoch": 2.7331975560081467,
"grad_norm": 0.6530488652200966,
"learning_rate": 2.2953325944903848e-07,
"loss": 0.0395,
"step": 671
},
{
"epoch": 2.7372708757637474,
"grad_norm": 0.6113145873412146,
"learning_rate": 2.2246964313763053e-07,
"loss": 0.041,
"step": 672
},
{
"epoch": 2.741344195519348,
"grad_norm": 0.7095148513537615,
"learning_rate": 2.1551394621182277e-07,
"loss": 0.0397,
"step": 673
},
{
"epoch": 2.745417515274949,
"grad_norm": 0.6033544892162519,
"learning_rate": 2.08666325793494e-07,
"loss": 0.0443,
"step": 674
},
{
"epoch": 2.74949083503055,
"grad_norm": 0.5446102953157629,
"learning_rate": 2.0192693656318597e-07,
"loss": 0.0408,
"step": 675
},
{
"epoch": 2.753564154786151,
"grad_norm": 0.6143476593669657,
"learning_rate": 1.9529593075661267e-07,
"loss": 0.0431,
"step": 676
},
{
"epoch": 2.7576374745417516,
"grad_norm": 0.5937558119917979,
"learning_rate": 1.8877345816122162e-07,
"loss": 0.0389,
"step": 677
},
{
"epoch": 2.7617107942973522,
"grad_norm": 0.6700927045976569,
"learning_rate": 1.8235966611280687e-07,
"loss": 0.0427,
"step": 678
},
{
"epoch": 2.765784114052953,
"grad_norm": 0.6396378419079983,
"learning_rate": 1.760546994921858e-07,
"loss": 0.041,
"step": 679
},
{
"epoch": 2.769857433808554,
"grad_norm": 0.7273252196426618,
"learning_rate": 1.6985870072192156e-07,
"loss": 0.0486,
"step": 680
},
{
"epoch": 2.7739307535641546,
"grad_norm": 0.6092698543049133,
"learning_rate": 1.6377180976310968e-07,
"loss": 0.0414,
"step": 681
},
{
"epoch": 2.7780040733197557,
"grad_norm": 0.6384474489803748,
"learning_rate": 1.5779416411221437e-07,
"loss": 0.0409,
"step": 682
},
{
"epoch": 2.7820773930753564,
"grad_norm": 0.7062056069710516,
"learning_rate": 1.5192589879796383e-07,
"loss": 0.0361,
"step": 683
},
{
"epoch": 2.786150712830957,
"grad_norm": 0.6060235266872488,
"learning_rate": 1.4616714637829822e-07,
"loss": 0.0416,
"step": 684
},
{
"epoch": 2.790224032586558,
"grad_norm": 0.6872805545772854,
"learning_rate": 1.4051803693737876e-07,
"loss": 0.0447,
"step": 685
},
{
"epoch": 2.794297352342159,
"grad_norm": 1.1080142515005726,
"learning_rate": 1.3497869808264453e-07,
"loss": 0.0572,
"step": 686
},
{
"epoch": 2.79837067209776,
"grad_norm": 0.576061423457441,
"learning_rate": 1.2954925494193472e-07,
"loss": 0.0407,
"step": 687
},
{
"epoch": 2.8024439918533606,
"grad_norm": 0.5287732197801528,
"learning_rate": 1.2422983016065816e-07,
"loss": 0.037,
"step": 688
},
{
"epoch": 2.8065173116089612,
"grad_norm": 1.0183232038437686,
"learning_rate": 1.1902054389902662e-07,
"loss": 0.0519,
"step": 689
},
{
"epoch": 2.810590631364562,
"grad_norm": 0.7757556677352597,
"learning_rate": 1.1392151382933647e-07,
"loss": 0.0393,
"step": 690
},
{
"epoch": 2.814663951120163,
"grad_norm": 0.6719286279824974,
"learning_rate": 1.0893285513331353e-07,
"loss": 0.0466,
"step": 691
},
{
"epoch": 2.8187372708757636,
"grad_norm": 0.6944618367685592,
"learning_rate": 1.0405468049951184e-07,
"loss": 0.0413,
"step": 692
},
{
"epoch": 2.8228105906313647,
"grad_norm": 0.61552802643045,
"learning_rate": 9.928710012076404e-08,
"loss": 0.0471,
"step": 693
},
{
"epoch": 2.8268839103869654,
"grad_norm": 0.5410095050369199,
"learning_rate": 9.463022169169666e-08,
"loss": 0.0405,
"step": 694
},
{
"epoch": 2.830957230142566,
"grad_norm": 0.6222514939239194,
"learning_rate": 9.008415040629548e-08,
"loss": 0.0381,
"step": 695
},
{
"epoch": 2.835030549898167,
"grad_norm": 0.716237190221308,
"learning_rate": 8.564898895552843e-08,
"loss": 0.0483,
"step": 696
},
{
"epoch": 2.839103869653768,
"grad_norm": 0.5438713399396514,
"learning_rate": 8.132483752502806e-08,
"loss": 0.0307,
"step": 697
},
{
"epoch": 2.8431771894093685,
"grad_norm": 0.5546226611008334,
"learning_rate": 7.711179379282674e-08,
"loss": 0.0436,
"step": 698
},
{
"epoch": 2.8472505091649696,
"grad_norm": 0.7795946084989639,
"learning_rate": 7.300995292715107e-08,
"loss": 0.0544,
"step": 699
},
{
"epoch": 2.8513238289205702,
"grad_norm": 0.7271455444954849,
"learning_rate": 6.901940758427206e-08,
"loss": 0.0506,
"step": 700
},
{
"epoch": 2.855397148676171,
"grad_norm": 0.6167947815542473,
"learning_rate": 6.514024790641116e-08,
"loss": 0.0338,
"step": 701
},
{
"epoch": 2.859470468431772,
"grad_norm": 0.9028615937159957,
"learning_rate": 6.137256151970583e-08,
"loss": 0.0384,
"step": 702
},
{
"epoch": 2.8635437881873727,
"grad_norm": 0.5586903994455261,
"learning_rate": 5.771643353222778e-08,
"loss": 0.0417,
"step": 703
},
{
"epoch": 2.8676171079429738,
"grad_norm": 0.5744286666789391,
"learning_rate": 5.417194653206337e-08,
"loss": 0.0398,
"step": 704
},
{
"epoch": 2.8716904276985744,
"grad_norm": 0.5715192295451592,
"learning_rate": 5.073918058544458e-08,
"loss": 0.0418,
"step": 705
},
{
"epoch": 2.875763747454175,
"grad_norm": 0.6253865033754882,
"learning_rate": 4.741821323494489e-08,
"loss": 0.039,
"step": 706
},
{
"epoch": 2.8798370672097757,
"grad_norm": 0.5416303891716191,
"learning_rate": 4.4209119497722883e-08,
"loss": 0.0354,
"step": 707
},
{
"epoch": 2.883910386965377,
"grad_norm": 0.7147732942247966,
"learning_rate": 4.1111971863830866e-08,
"loss": 0.0429,
"step": 708
},
{
"epoch": 2.8879837067209775,
"grad_norm": 0.6143363731337814,
"learning_rate": 3.812684029457614e-08,
"loss": 0.0409,
"step": 709
},
{
"epoch": 2.8920570264765786,
"grad_norm": 0.5571624531996129,
"learning_rate": 3.525379222094061e-08,
"loss": 0.0378,
"step": 710
},
{
"epoch": 2.8961303462321792,
"grad_norm": 0.5623398083462782,
"learning_rate": 3.249289254205867e-08,
"loss": 0.0384,
"step": 711
},
{
"epoch": 2.90020366598778,
"grad_norm": 0.620822915325178,
"learning_rate": 2.984420362375007e-08,
"loss": 0.0389,
"step": 712
},
{
"epoch": 2.904276985743381,
"grad_norm": 0.6323804053973247,
"learning_rate": 2.7307785297111533e-08,
"loss": 0.0382,
"step": 713
},
{
"epoch": 2.9083503054989817,
"grad_norm": 0.577630536988922,
"learning_rate": 2.488369485716513e-08,
"loss": 0.0424,
"step": 714
},
{
"epoch": 2.9124236252545828,
"grad_norm": 0.5908644464528476,
"learning_rate": 2.2571987061564827e-08,
"loss": 0.0374,
"step": 715
},
{
"epoch": 2.9164969450101834,
"grad_norm": 0.5819680044309804,
"learning_rate": 2.0372714129356375e-08,
"loss": 0.0418,
"step": 716
},
{
"epoch": 2.920570264765784,
"grad_norm": 0.5998719655116901,
"learning_rate": 1.8285925739803812e-08,
"loss": 0.0373,
"step": 717
},
{
"epoch": 2.9246435845213847,
"grad_norm": 0.6039864647607079,
"learning_rate": 1.631166903126147e-08,
"loss": 0.0459,
"step": 718
},
{
"epoch": 2.928716904276986,
"grad_norm": 0.762786479746698,
"learning_rate": 1.4449988600111486e-08,
"loss": 0.0515,
"step": 719
},
{
"epoch": 2.9327902240325865,
"grad_norm": 0.6018854526331516,
"learning_rate": 1.2700926499756295e-08,
"loss": 0.0414,
"step": 720
},
{
"epoch": 2.9368635437881876,
"grad_norm": 0.5961101580877264,
"learning_rate": 1.1064522239669916e-08,
"loss": 0.0354,
"step": 721
},
{
"epoch": 2.9409368635437882,
"grad_norm": 1.7052061344176328,
"learning_rate": 9.54081278450314e-09,
"loss": 0.056,
"step": 722
},
{
"epoch": 2.945010183299389,
"grad_norm": 0.6291245374019063,
"learning_rate": 8.129832553249173e-09,
"loss": 0.0396,
"step": 723
},
{
"epoch": 2.9490835030549896,
"grad_norm": 0.6127187852232583,
"learning_rate": 6.831613418468163e-09,
"loss": 0.0459,
"step": 724
},
{
"epoch": 2.9531568228105907,
"grad_norm": 0.7310466052458664,
"learning_rate": 5.646184705563884e-09,
"loss": 0.0375,
"step": 725
},
{
"epoch": 2.9572301425661913,
"grad_norm": 0.5622139337651338,
"learning_rate": 4.573573192125369e-09,
"loss": 0.0351,
"step": 726
},
{
"epoch": 2.9613034623217924,
"grad_norm": 0.5994897822486772,
"learning_rate": 3.613803107317959e-09,
"loss": 0.0483,
"step": 727
},
{
"epoch": 2.965376782077393,
"grad_norm": 0.6310995479806456,
"learning_rate": 2.7668961313376263e-09,
"loss": 0.0414,
"step": 728
},
{
"epoch": 2.9694501018329937,
"grad_norm": 0.6948629856817712,
"learning_rate": 2.0328713949230304e-09,
"loss": 0.0495,
"step": 729
},
{
"epoch": 2.973523421588595,
"grad_norm": 0.7318123074015203,
"learning_rate": 1.4117454789208673e-09,
"loss": 0.046,
"step": 730
},
{
"epoch": 2.9775967413441955,
"grad_norm": 0.6523465210659862,
"learning_rate": 9.03532413911723e-10,
"loss": 0.0442,
"step": 731
},
{
"epoch": 2.9816700610997966,
"grad_norm": 0.5390724631100189,
"learning_rate": 5.08243679894771e-10,
"loss": 0.0332,
"step": 732
},
{
"epoch": 2.9857433808553973,
"grad_norm": 0.6407224408398684,
"learning_rate": 2.2588820602631457e-10,
"loss": 0.0489,
"step": 733
},
{
"epoch": 2.989816700610998,
"grad_norm": 0.7098372871776469,
"learning_rate": 5.6472370419391464e-11,
"loss": 0.0476,
"step": 734
},
{
"epoch": 2.9938900203665986,
"grad_norm": 0.6576944358362913,
"learning_rate": 0.0,
"loss": 0.048,
"step": 735
},
{
"epoch": 2.9938900203665986,
"step": 735,
"total_flos": 48064094208000.0,
"train_loss": 0.0952699331224573,
"train_runtime": 5965.7023,
"train_samples_per_second": 1.975,
"train_steps_per_second": 0.123
}
],
"logging_steps": 1,
"max_steps": 735,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"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": 48064094208000.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}