llama_3.1_8b_r_1 / trainer_state.json
sam2ai's picture
Add files using upload-large-folder tool
f2f0651 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": 14.739597738910254,
"learning_rate": 1.3513513513513515e-07,
"loss": 0.3301,
"step": 1
},
{
"epoch": 0.008146639511201629,
"grad_norm": 12.452379475537562,
"learning_rate": 2.702702702702703e-07,
"loss": 0.3036,
"step": 2
},
{
"epoch": 0.012219959266802444,
"grad_norm": 12.86836901323053,
"learning_rate": 4.0540540540540546e-07,
"loss": 0.3424,
"step": 3
},
{
"epoch": 0.016293279022403257,
"grad_norm": 13.43090696417895,
"learning_rate": 5.405405405405406e-07,
"loss": 0.3184,
"step": 4
},
{
"epoch": 0.020366598778004074,
"grad_norm": 14.809270827070751,
"learning_rate": 6.756756756756758e-07,
"loss": 0.3138,
"step": 5
},
{
"epoch": 0.024439918533604887,
"grad_norm": 12.082295065079254,
"learning_rate": 8.108108108108109e-07,
"loss": 0.2982,
"step": 6
},
{
"epoch": 0.028513238289205704,
"grad_norm": 8.711027825602685,
"learning_rate": 9.459459459459461e-07,
"loss": 0.3063,
"step": 7
},
{
"epoch": 0.032586558044806514,
"grad_norm": 7.652261107879573,
"learning_rate": 1.0810810810810812e-06,
"loss": 0.2749,
"step": 8
},
{
"epoch": 0.03665987780040733,
"grad_norm": 7.726905307446712,
"learning_rate": 1.2162162162162164e-06,
"loss": 0.2718,
"step": 9
},
{
"epoch": 0.04073319755600815,
"grad_norm": 3.4676017287243446,
"learning_rate": 1.3513513513513515e-06,
"loss": 0.2238,
"step": 10
},
{
"epoch": 0.04480651731160896,
"grad_norm": 4.29492237739491,
"learning_rate": 1.4864864864864868e-06,
"loss": 0.2513,
"step": 11
},
{
"epoch": 0.048879837067209775,
"grad_norm": 3.5844077505877707,
"learning_rate": 1.6216216216216219e-06,
"loss": 0.2327,
"step": 12
},
{
"epoch": 0.05295315682281059,
"grad_norm": 7.031113851112456,
"learning_rate": 1.756756756756757e-06,
"loss": 0.239,
"step": 13
},
{
"epoch": 0.05702647657841141,
"grad_norm": 7.210450292846676,
"learning_rate": 1.8918918918918922e-06,
"loss": 0.277,
"step": 14
},
{
"epoch": 0.06109979633401222,
"grad_norm": 6.526062178388152,
"learning_rate": 2.0270270270270273e-06,
"loss": 0.2495,
"step": 15
},
{
"epoch": 0.06517311608961303,
"grad_norm": 5.157530413977274,
"learning_rate": 2.1621621621621623e-06,
"loss": 0.2425,
"step": 16
},
{
"epoch": 0.06924643584521385,
"grad_norm": 3.242500698401516,
"learning_rate": 2.297297297297298e-06,
"loss": 0.1985,
"step": 17
},
{
"epoch": 0.07331975560081466,
"grad_norm": 2.554097363562634,
"learning_rate": 2.432432432432433e-06,
"loss": 0.1822,
"step": 18
},
{
"epoch": 0.07739307535641547,
"grad_norm": 2.6535341941648047,
"learning_rate": 2.5675675675675675e-06,
"loss": 0.2252,
"step": 19
},
{
"epoch": 0.0814663951120163,
"grad_norm": 2.5516965886789076,
"learning_rate": 2.702702702702703e-06,
"loss": 0.1862,
"step": 20
},
{
"epoch": 0.0855397148676171,
"grad_norm": 2.6017363031504264,
"learning_rate": 2.837837837837838e-06,
"loss": 0.2098,
"step": 21
},
{
"epoch": 0.08961303462321792,
"grad_norm": 2.0129035331912184,
"learning_rate": 2.9729729729729736e-06,
"loss": 0.1699,
"step": 22
},
{
"epoch": 0.09368635437881874,
"grad_norm": 1.849938821231628,
"learning_rate": 3.1081081081081082e-06,
"loss": 0.1868,
"step": 23
},
{
"epoch": 0.09775967413441955,
"grad_norm": 1.764863585345639,
"learning_rate": 3.2432432432432437e-06,
"loss": 0.1669,
"step": 24
},
{
"epoch": 0.10183299389002037,
"grad_norm": 1.8494325730642949,
"learning_rate": 3.3783783783783788e-06,
"loss": 0.1507,
"step": 25
},
{
"epoch": 0.10590631364562118,
"grad_norm": 1.6304914383856781,
"learning_rate": 3.513513513513514e-06,
"loss": 0.1534,
"step": 26
},
{
"epoch": 0.109979633401222,
"grad_norm": 1.4579913516119778,
"learning_rate": 3.648648648648649e-06,
"loss": 0.1387,
"step": 27
},
{
"epoch": 0.11405295315682282,
"grad_norm": 1.6082049119577289,
"learning_rate": 3.7837837837837844e-06,
"loss": 0.1406,
"step": 28
},
{
"epoch": 0.11812627291242363,
"grad_norm": 1.4819461102449234,
"learning_rate": 3.918918918918919e-06,
"loss": 0.1323,
"step": 29
},
{
"epoch": 0.12219959266802444,
"grad_norm": 1.7009165157092279,
"learning_rate": 4.0540540540540545e-06,
"loss": 0.1577,
"step": 30
},
{
"epoch": 0.12627291242362526,
"grad_norm": 1.5721048343370418,
"learning_rate": 4.189189189189189e-06,
"loss": 0.1345,
"step": 31
},
{
"epoch": 0.13034623217922606,
"grad_norm": 1.5868902144082508,
"learning_rate": 4.324324324324325e-06,
"loss": 0.1641,
"step": 32
},
{
"epoch": 0.13441955193482688,
"grad_norm": 1.367409491711825,
"learning_rate": 4.45945945945946e-06,
"loss": 0.1568,
"step": 33
},
{
"epoch": 0.1384928716904277,
"grad_norm": 1.2082341226617432,
"learning_rate": 4.594594594594596e-06,
"loss": 0.1158,
"step": 34
},
{
"epoch": 0.1425661914460285,
"grad_norm": 1.2834846670425744,
"learning_rate": 4.72972972972973e-06,
"loss": 0.1553,
"step": 35
},
{
"epoch": 0.14663951120162932,
"grad_norm": 1.4278045526468992,
"learning_rate": 4.864864864864866e-06,
"loss": 0.1472,
"step": 36
},
{
"epoch": 0.15071283095723015,
"grad_norm": 1.1433863309324082,
"learning_rate": 5e-06,
"loss": 0.1216,
"step": 37
},
{
"epoch": 0.15478615071283094,
"grad_norm": 1.2556861775151085,
"learning_rate": 5.135135135135135e-06,
"loss": 0.1383,
"step": 38
},
{
"epoch": 0.15885947046843177,
"grad_norm": 1.1940515610624718,
"learning_rate": 5.2702702702702705e-06,
"loss": 0.1488,
"step": 39
},
{
"epoch": 0.1629327902240326,
"grad_norm": 1.365361196469323,
"learning_rate": 5.405405405405406e-06,
"loss": 0.1658,
"step": 40
},
{
"epoch": 0.1670061099796334,
"grad_norm": 1.6375597676471598,
"learning_rate": 5.540540540540541e-06,
"loss": 0.1244,
"step": 41
},
{
"epoch": 0.1710794297352342,
"grad_norm": 1.058410205986207,
"learning_rate": 5.675675675675676e-06,
"loss": 0.1129,
"step": 42
},
{
"epoch": 0.17515274949083504,
"grad_norm": 1.468616447672182,
"learning_rate": 5.810810810810811e-06,
"loss": 0.176,
"step": 43
},
{
"epoch": 0.17922606924643583,
"grad_norm": 1.1292066998688302,
"learning_rate": 5.945945945945947e-06,
"loss": 0.1235,
"step": 44
},
{
"epoch": 0.18329938900203666,
"grad_norm": 1.1790440780653373,
"learning_rate": 6.081081081081082e-06,
"loss": 0.1352,
"step": 45
},
{
"epoch": 0.18737270875763748,
"grad_norm": 1.144770740193701,
"learning_rate": 6.2162162162162164e-06,
"loss": 0.1375,
"step": 46
},
{
"epoch": 0.19144602851323828,
"grad_norm": 1.3169675540020822,
"learning_rate": 6.351351351351351e-06,
"loss": 0.1451,
"step": 47
},
{
"epoch": 0.1955193482688391,
"grad_norm": 1.1364743430386761,
"learning_rate": 6.486486486486487e-06,
"loss": 0.1073,
"step": 48
},
{
"epoch": 0.19959266802443992,
"grad_norm": 1.3532964160734307,
"learning_rate": 6.621621621621622e-06,
"loss": 0.1502,
"step": 49
},
{
"epoch": 0.20366598778004075,
"grad_norm": 1.1049371458723167,
"learning_rate": 6.7567567567567575e-06,
"loss": 0.116,
"step": 50
},
{
"epoch": 0.20773930753564154,
"grad_norm": 1.0634720045604809,
"learning_rate": 6.891891891891892e-06,
"loss": 0.1438,
"step": 51
},
{
"epoch": 0.21181262729124237,
"grad_norm": 1.1677623232682453,
"learning_rate": 7.027027027027028e-06,
"loss": 0.1143,
"step": 52
},
{
"epoch": 0.2158859470468432,
"grad_norm": 1.2552603959178812,
"learning_rate": 7.162162162162163e-06,
"loss": 0.1443,
"step": 53
},
{
"epoch": 0.219959266802444,
"grad_norm": 1.1556616828254782,
"learning_rate": 7.297297297297298e-06,
"loss": 0.1341,
"step": 54
},
{
"epoch": 0.2240325865580448,
"grad_norm": 1.1241236805182522,
"learning_rate": 7.4324324324324324e-06,
"loss": 0.1283,
"step": 55
},
{
"epoch": 0.22810590631364563,
"grad_norm": 0.9867741809756463,
"learning_rate": 7.567567567567569e-06,
"loss": 0.1302,
"step": 56
},
{
"epoch": 0.23217922606924643,
"grad_norm": 1.0672327000495885,
"learning_rate": 7.702702702702704e-06,
"loss": 0.113,
"step": 57
},
{
"epoch": 0.23625254582484725,
"grad_norm": 1.0659735135074857,
"learning_rate": 7.837837837837838e-06,
"loss": 0.1293,
"step": 58
},
{
"epoch": 0.24032586558044808,
"grad_norm": 1.2422197356017706,
"learning_rate": 7.972972972972974e-06,
"loss": 0.164,
"step": 59
},
{
"epoch": 0.24439918533604887,
"grad_norm": 1.3538609671806645,
"learning_rate": 8.108108108108109e-06,
"loss": 0.1548,
"step": 60
},
{
"epoch": 0.2484725050916497,
"grad_norm": 1.0759558101958346,
"learning_rate": 8.243243243243245e-06,
"loss": 0.1225,
"step": 61
},
{
"epoch": 0.2525458248472505,
"grad_norm": 1.1244956381449198,
"learning_rate": 8.378378378378378e-06,
"loss": 0.1175,
"step": 62
},
{
"epoch": 0.25661914460285135,
"grad_norm": 1.171629685706723,
"learning_rate": 8.513513513513514e-06,
"loss": 0.1204,
"step": 63
},
{
"epoch": 0.2606924643584521,
"grad_norm": 1.2905585681894916,
"learning_rate": 8.64864864864865e-06,
"loss": 0.1253,
"step": 64
},
{
"epoch": 0.26476578411405294,
"grad_norm": 1.3979008428570314,
"learning_rate": 8.783783783783785e-06,
"loss": 0.191,
"step": 65
},
{
"epoch": 0.26883910386965376,
"grad_norm": 1.226756333773235,
"learning_rate": 8.91891891891892e-06,
"loss": 0.1287,
"step": 66
},
{
"epoch": 0.2729124236252546,
"grad_norm": 1.2835470528218054,
"learning_rate": 9.054054054054054e-06,
"loss": 0.138,
"step": 67
},
{
"epoch": 0.2769857433808554,
"grad_norm": 1.1622195270679896,
"learning_rate": 9.189189189189191e-06,
"loss": 0.1259,
"step": 68
},
{
"epoch": 0.28105906313645623,
"grad_norm": 1.1512666578576678,
"learning_rate": 9.324324324324325e-06,
"loss": 0.1292,
"step": 69
},
{
"epoch": 0.285132382892057,
"grad_norm": 0.9695391815507838,
"learning_rate": 9.45945945945946e-06,
"loss": 0.1142,
"step": 70
},
{
"epoch": 0.2892057026476578,
"grad_norm": 1.1262409828408337,
"learning_rate": 9.594594594594594e-06,
"loss": 0.1188,
"step": 71
},
{
"epoch": 0.29327902240325865,
"grad_norm": 0.9820966211674147,
"learning_rate": 9.729729729729732e-06,
"loss": 0.1052,
"step": 72
},
{
"epoch": 0.2973523421588595,
"grad_norm": 1.1058230077470572,
"learning_rate": 9.864864864864865e-06,
"loss": 0.1246,
"step": 73
},
{
"epoch": 0.3014256619144603,
"grad_norm": 1.3891942844370528,
"learning_rate": 1e-05,
"loss": 0.1651,
"step": 74
},
{
"epoch": 0.3054989816700611,
"grad_norm": 1.1373599847305171,
"learning_rate": 9.99994352762958e-06,
"loss": 0.1259,
"step": 75
},
{
"epoch": 0.3095723014256619,
"grad_norm": 1.0803757941511039,
"learning_rate": 9.999774111793974e-06,
"loss": 0.1485,
"step": 76
},
{
"epoch": 0.3136456211812627,
"grad_norm": 1.509987566205336,
"learning_rate": 9.999491756320105e-06,
"loss": 0.1708,
"step": 77
},
{
"epoch": 0.31771894093686354,
"grad_norm": 1.3769318827034491,
"learning_rate": 9.99909646758609e-06,
"loss": 0.1483,
"step": 78
},
{
"epoch": 0.32179226069246436,
"grad_norm": 0.9995516357476201,
"learning_rate": 9.99858825452108e-06,
"loss": 0.1124,
"step": 79
},
{
"epoch": 0.3258655804480652,
"grad_norm": 1.4328593788226842,
"learning_rate": 9.997967128605078e-06,
"loss": 0.1849,
"step": 80
},
{
"epoch": 0.329938900203666,
"grad_norm": 1.0397129864144867,
"learning_rate": 9.997233103868664e-06,
"loss": 0.1199,
"step": 81
},
{
"epoch": 0.3340122199592668,
"grad_norm": 1.3312975796955133,
"learning_rate": 9.996386196892683e-06,
"loss": 0.1748,
"step": 82
},
{
"epoch": 0.3380855397148676,
"grad_norm": 1.2070448028045222,
"learning_rate": 9.995426426807875e-06,
"loss": 0.1449,
"step": 83
},
{
"epoch": 0.3421588594704684,
"grad_norm": 0.9786604342473315,
"learning_rate": 9.994353815294438e-06,
"loss": 0.1349,
"step": 84
},
{
"epoch": 0.34623217922606925,
"grad_norm": 1.16279378070579,
"learning_rate": 9.993168386581533e-06,
"loss": 0.1111,
"step": 85
},
{
"epoch": 0.35030549898167007,
"grad_norm": 1.0832386326974766,
"learning_rate": 9.991870167446751e-06,
"loss": 0.1271,
"step": 86
},
{
"epoch": 0.3543788187372709,
"grad_norm": 1.076044536856832,
"learning_rate": 9.990459187215498e-06,
"loss": 0.122,
"step": 87
},
{
"epoch": 0.35845213849287166,
"grad_norm": 1.1390626595350608,
"learning_rate": 9.98893547776033e-06,
"loss": 0.1429,
"step": 88
},
{
"epoch": 0.3625254582484725,
"grad_norm": 1.2799324833393828,
"learning_rate": 9.987299073500245e-06,
"loss": 0.1789,
"step": 89
},
{
"epoch": 0.3665987780040733,
"grad_norm": 1.0088789278468007,
"learning_rate": 9.985550011399889e-06,
"loss": 0.1217,
"step": 90
},
{
"epoch": 0.37067209775967414,
"grad_norm": 1.0635380396962304,
"learning_rate": 9.98368833096874e-06,
"loss": 0.1517,
"step": 91
},
{
"epoch": 0.37474541751527496,
"grad_norm": 1.1149195586496816,
"learning_rate": 9.981714074260196e-06,
"loss": 0.1648,
"step": 92
},
{
"epoch": 0.3788187372708758,
"grad_norm": 0.9770064004740078,
"learning_rate": 9.979627285870644e-06,
"loss": 0.1173,
"step": 93
},
{
"epoch": 0.38289205702647655,
"grad_norm": 1.5786545324573935,
"learning_rate": 9.977428012938437e-06,
"loss": 0.2148,
"step": 94
},
{
"epoch": 0.3869653767820774,
"grad_norm": 0.9445672697637628,
"learning_rate": 9.975116305142836e-06,
"loss": 0.1272,
"step": 95
},
{
"epoch": 0.3910386965376782,
"grad_norm": 0.832092882135511,
"learning_rate": 9.97269221470289e-06,
"loss": 0.1149,
"step": 96
},
{
"epoch": 0.395112016293279,
"grad_norm": 0.8009975217381654,
"learning_rate": 9.97015579637625e-06,
"loss": 0.1081,
"step": 97
},
{
"epoch": 0.39918533604887985,
"grad_norm": 0.909000272396086,
"learning_rate": 9.967507107457942e-06,
"loss": 0.1249,
"step": 98
},
{
"epoch": 0.40325865580448067,
"grad_norm": 0.9894702747295367,
"learning_rate": 9.96474620777906e-06,
"loss": 0.1404,
"step": 99
},
{
"epoch": 0.4073319755600815,
"grad_norm": 1.1517905886733883,
"learning_rate": 9.961873159705426e-06,
"loss": 0.1433,
"step": 100
},
{
"epoch": 0.41140529531568226,
"grad_norm": 1.2806427058824508,
"learning_rate": 9.95888802813617e-06,
"loss": 0.1723,
"step": 101
},
{
"epoch": 0.4154786150712831,
"grad_norm": 0.919332585767889,
"learning_rate": 9.955790880502278e-06,
"loss": 0.1219,
"step": 102
},
{
"epoch": 0.4195519348268839,
"grad_norm": 0.8901964293186232,
"learning_rate": 9.952581786765057e-06,
"loss": 0.1157,
"step": 103
},
{
"epoch": 0.42362525458248473,
"grad_norm": 1.3877972822654616,
"learning_rate": 9.949260819414557e-06,
"loss": 0.1642,
"step": 104
},
{
"epoch": 0.42769857433808556,
"grad_norm": 0.9602184939318458,
"learning_rate": 9.945828053467939e-06,
"loss": 0.1224,
"step": 105
},
{
"epoch": 0.4317718940936864,
"grad_norm": 1.230791876608231,
"learning_rate": 9.942283566467773e-06,
"loss": 0.1596,
"step": 106
},
{
"epoch": 0.43584521384928715,
"grad_norm": 1.1454248942159495,
"learning_rate": 9.938627438480295e-06,
"loss": 0.1541,
"step": 107
},
{
"epoch": 0.439918533604888,
"grad_norm": 1.0873300186194603,
"learning_rate": 9.93485975209359e-06,
"loss": 0.1533,
"step": 108
},
{
"epoch": 0.4439918533604888,
"grad_norm": 0.9668569607934798,
"learning_rate": 9.930980592415728e-06,
"loss": 0.1539,
"step": 109
},
{
"epoch": 0.4480651731160896,
"grad_norm": 1.487429443095859,
"learning_rate": 9.926990047072849e-06,
"loss": 0.2379,
"step": 110
},
{
"epoch": 0.45213849287169044,
"grad_norm": 1.036501582869458,
"learning_rate": 9.922888206207174e-06,
"loss": 0.1181,
"step": 111
},
{
"epoch": 0.45621181262729127,
"grad_norm": 0.9427386345315173,
"learning_rate": 9.918675162474974e-06,
"loss": 0.1157,
"step": 112
},
{
"epoch": 0.46028513238289204,
"grad_norm": 1.1671785006625848,
"learning_rate": 9.914351011044472e-06,
"loss": 0.1671,
"step": 113
},
{
"epoch": 0.46435845213849286,
"grad_norm": 0.8485104800209154,
"learning_rate": 9.909915849593705e-06,
"loss": 0.1094,
"step": 114
},
{
"epoch": 0.4684317718940937,
"grad_norm": 0.895507646361391,
"learning_rate": 9.905369778308304e-06,
"loss": 0.1205,
"step": 115
},
{
"epoch": 0.4725050916496945,
"grad_norm": 1.1024237478073182,
"learning_rate": 9.900712899879237e-06,
"loss": 0.1551,
"step": 116
},
{
"epoch": 0.47657841140529533,
"grad_norm": 1.0811464118865846,
"learning_rate": 9.895945319500488e-06,
"loss": 0.1402,
"step": 117
},
{
"epoch": 0.48065173116089616,
"grad_norm": 0.9829410685047446,
"learning_rate": 9.891067144866687e-06,
"loss": 0.1381,
"step": 118
},
{
"epoch": 0.4847250509164969,
"grad_norm": 0.8855824729064482,
"learning_rate": 9.886078486170665e-06,
"loss": 0.1038,
"step": 119
},
{
"epoch": 0.48879837067209775,
"grad_norm": 1.1091690462920576,
"learning_rate": 9.880979456100974e-06,
"loss": 0.1372,
"step": 120
},
{
"epoch": 0.49287169042769857,
"grad_norm": 0.907049897730717,
"learning_rate": 9.875770169839343e-06,
"loss": 0.1322,
"step": 121
},
{
"epoch": 0.4969450101832994,
"grad_norm": 1.0224824312976686,
"learning_rate": 9.870450745058066e-06,
"loss": 0.1257,
"step": 122
},
{
"epoch": 0.5010183299389002,
"grad_norm": 1.0439109698157967,
"learning_rate": 9.865021301917358e-06,
"loss": 0.1317,
"step": 123
},
{
"epoch": 0.505091649694501,
"grad_norm": 0.8972366065592501,
"learning_rate": 9.859481963062623e-06,
"loss": 0.1104,
"step": 124
},
{
"epoch": 0.5091649694501018,
"grad_norm": 0.916952485621608,
"learning_rate": 9.853832853621703e-06,
"loss": 0.124,
"step": 125
},
{
"epoch": 0.5132382892057027,
"grad_norm": 0.7586835858660547,
"learning_rate": 9.848074101202037e-06,
"loss": 0.1191,
"step": 126
},
{
"epoch": 0.5173116089613035,
"grad_norm": 0.9149593226270635,
"learning_rate": 9.842205835887785e-06,
"loss": 0.1188,
"step": 127
},
{
"epoch": 0.5213849287169042,
"grad_norm": 0.9483144871900878,
"learning_rate": 9.836228190236892e-06,
"loss": 0.1392,
"step": 128
},
{
"epoch": 0.5254582484725051,
"grad_norm": 1.1137009286811568,
"learning_rate": 9.83014129927808e-06,
"loss": 0.1331,
"step": 129
},
{
"epoch": 0.5295315682281059,
"grad_norm": 1.0049886812823983,
"learning_rate": 9.823945300507815e-06,
"loss": 0.1393,
"step": 130
},
{
"epoch": 0.5336048879837068,
"grad_norm": 1.0017821694016227,
"learning_rate": 9.817640333887194e-06,
"loss": 0.1376,
"step": 131
},
{
"epoch": 0.5376782077393075,
"grad_norm": 0.8770993451067021,
"learning_rate": 9.81122654183878e-06,
"loss": 0.1075,
"step": 132
},
{
"epoch": 0.5417515274949084,
"grad_norm": 0.8112662923925413,
"learning_rate": 9.804704069243389e-06,
"loss": 0.1149,
"step": 133
},
{
"epoch": 0.5458248472505092,
"grad_norm": 0.7783508225595258,
"learning_rate": 9.798073063436815e-06,
"loss": 0.1077,
"step": 134
},
{
"epoch": 0.5498981670061099,
"grad_norm": 1.6671316247114485,
"learning_rate": 9.791333674206507e-06,
"loss": 0.1892,
"step": 135
},
{
"epoch": 0.5539714867617108,
"grad_norm": 0.8856245620297392,
"learning_rate": 9.784486053788179e-06,
"loss": 0.1075,
"step": 136
},
{
"epoch": 0.5580448065173116,
"grad_norm": 2.0578900491298824,
"learning_rate": 9.77753035686237e-06,
"loss": 0.1472,
"step": 137
},
{
"epoch": 0.5621181262729125,
"grad_norm": 1.148525636808097,
"learning_rate": 9.770466740550963e-06,
"loss": 0.1598,
"step": 138
},
{
"epoch": 0.5661914460285132,
"grad_norm": 0.8665254831769179,
"learning_rate": 9.763295364413616e-06,
"loss": 0.1186,
"step": 139
},
{
"epoch": 0.570264765784114,
"grad_norm": 1.0970826186220186,
"learning_rate": 9.756016390444174e-06,
"loss": 0.1386,
"step": 140
},
{
"epoch": 0.5743380855397149,
"grad_norm": 0.9530034310899396,
"learning_rate": 9.748629983067004e-06,
"loss": 0.1282,
"step": 141
},
{
"epoch": 0.5784114052953157,
"grad_norm": 1.2706893271757027,
"learning_rate": 9.741136309133279e-06,
"loss": 0.1754,
"step": 142
},
{
"epoch": 0.5824847250509165,
"grad_norm": 0.9703463762849697,
"learning_rate": 9.733535537917211e-06,
"loss": 0.1194,
"step": 143
},
{
"epoch": 0.5865580448065173,
"grad_norm": 0.8038414888371753,
"learning_rate": 9.725827841112226e-06,
"loss": 0.1162,
"step": 144
},
{
"epoch": 0.5906313645621182,
"grad_norm": 0.9411283645508486,
"learning_rate": 9.718013392827087e-06,
"loss": 0.1121,
"step": 145
},
{
"epoch": 0.594704684317719,
"grad_norm": 1.501666156048829,
"learning_rate": 9.710092369581966e-06,
"loss": 0.16,
"step": 146
},
{
"epoch": 0.5987780040733197,
"grad_norm": 0.9141719119872903,
"learning_rate": 9.702064950304442e-06,
"loss": 0.1211,
"step": 147
},
{
"epoch": 0.6028513238289206,
"grad_norm": 0.8652675727574004,
"learning_rate": 9.693931316325473e-06,
"loss": 0.0946,
"step": 148
},
{
"epoch": 0.6069246435845214,
"grad_norm": 0.7377787499846402,
"learning_rate": 9.685691651375297e-06,
"loss": 0.1016,
"step": 149
},
{
"epoch": 0.6109979633401222,
"grad_norm": 0.7630312206018969,
"learning_rate": 9.677346141579277e-06,
"loss": 0.1014,
"step": 150
},
{
"epoch": 0.615071283095723,
"grad_norm": 0.9718289359974593,
"learning_rate": 9.668894975453705e-06,
"loss": 0.1562,
"step": 151
},
{
"epoch": 0.6191446028513238,
"grad_norm": 1.004301729468449,
"learning_rate": 9.66033834390153e-06,
"loss": 0.1372,
"step": 152
},
{
"epoch": 0.6232179226069247,
"grad_norm": 0.9350824611493259,
"learning_rate": 9.65167644020806e-06,
"loss": 0.1254,
"step": 153
},
{
"epoch": 0.6272912423625254,
"grad_norm": 0.7612329276402703,
"learning_rate": 9.64290946003659e-06,
"loss": 0.0989,
"step": 154
},
{
"epoch": 0.6313645621181263,
"grad_norm": 0.7706614538086551,
"learning_rate": 9.63403760142398e-06,
"loss": 0.1013,
"step": 155
},
{
"epoch": 0.6354378818737271,
"grad_norm": 1.0210499034582712,
"learning_rate": 9.625061064776183e-06,
"loss": 0.1134,
"step": 156
},
{
"epoch": 0.639511201629328,
"grad_norm": 0.7560805642981956,
"learning_rate": 9.61598005286372e-06,
"loss": 0.0939,
"step": 157
},
{
"epoch": 0.6435845213849287,
"grad_norm": 1.0834289937869723,
"learning_rate": 9.606794770817102e-06,
"loss": 0.1785,
"step": 158
},
{
"epoch": 0.6476578411405295,
"grad_norm": 1.0611196002268826,
"learning_rate": 9.597505426122184e-06,
"loss": 0.1571,
"step": 159
},
{
"epoch": 0.6517311608961304,
"grad_norm": 1.0914261737532949,
"learning_rate": 9.588112228615495e-06,
"loss": 0.1745,
"step": 160
},
{
"epoch": 0.6558044806517311,
"grad_norm": 0.953948451978483,
"learning_rate": 9.57861539047949e-06,
"loss": 0.1353,
"step": 161
},
{
"epoch": 0.659877800407332,
"grad_norm": 1.2562247665468482,
"learning_rate": 9.569015126237744e-06,
"loss": 0.1521,
"step": 162
},
{
"epoch": 0.6639511201629328,
"grad_norm": 0.8283783602425362,
"learning_rate": 9.559311652750135e-06,
"loss": 0.1161,
"step": 163
},
{
"epoch": 0.6680244399185336,
"grad_norm": 0.7823509791751794,
"learning_rate": 9.549505189207924e-06,
"loss": 0.0976,
"step": 164
},
{
"epoch": 0.6720977596741344,
"grad_norm": 1.118258806444578,
"learning_rate": 9.539595957128803e-06,
"loss": 0.171,
"step": 165
},
{
"epoch": 0.6761710794297352,
"grad_norm": 0.7563799438807557,
"learning_rate": 9.529584180351902e-06,
"loss": 0.1159,
"step": 166
},
{
"epoch": 0.6802443991853361,
"grad_norm": 1.0059732424782886,
"learning_rate": 9.519470085032733e-06,
"loss": 0.1278,
"step": 167
},
{
"epoch": 0.6843177189409368,
"grad_norm": 0.8261325503708756,
"learning_rate": 9.509253899638066e-06,
"loss": 0.104,
"step": 168
},
{
"epoch": 0.6883910386965377,
"grad_norm": 1.1918252125330613,
"learning_rate": 9.498935854940785e-06,
"loss": 0.1682,
"step": 169
},
{
"epoch": 0.6924643584521385,
"grad_norm": 0.7216709177105455,
"learning_rate": 9.488516184014667e-06,
"loss": 0.1089,
"step": 170
},
{
"epoch": 0.6965376782077393,
"grad_norm": 0.8952054280934858,
"learning_rate": 9.477995122229117e-06,
"loss": 0.1521,
"step": 171
},
{
"epoch": 0.7006109979633401,
"grad_norm": 0.6538828419017942,
"learning_rate": 9.467372907243858e-06,
"loss": 0.1012,
"step": 172
},
{
"epoch": 0.7046843177189409,
"grad_norm": 0.840723056036209,
"learning_rate": 9.456649779003548e-06,
"loss": 0.117,
"step": 173
},
{
"epoch": 0.7087576374745418,
"grad_norm": 0.7652580794490056,
"learning_rate": 9.44582597973238e-06,
"loss": 0.1284,
"step": 174
},
{
"epoch": 0.7128309572301426,
"grad_norm": 0.9696904154678632,
"learning_rate": 9.434901753928593e-06,
"loss": 0.1429,
"step": 175
},
{
"epoch": 0.7169042769857433,
"grad_norm": 0.7509027450046076,
"learning_rate": 9.423877348358956e-06,
"loss": 0.1006,
"step": 176
},
{
"epoch": 0.7209775967413442,
"grad_norm": 0.6942112976471692,
"learning_rate": 9.4127530120532e-06,
"loss": 0.1042,
"step": 177
},
{
"epoch": 0.725050916496945,
"grad_norm": 1.4641902043350905,
"learning_rate": 9.401528996298375e-06,
"loss": 0.1676,
"step": 178
},
{
"epoch": 0.7291242362525459,
"grad_norm": 0.7418396518869238,
"learning_rate": 9.390205554633193e-06,
"loss": 0.1082,
"step": 179
},
{
"epoch": 0.7331975560081466,
"grad_norm": 1.2074617530849705,
"learning_rate": 9.378782942842292e-06,
"loss": 0.1401,
"step": 180
},
{
"epoch": 0.7372708757637475,
"grad_norm": 1.2938802390610347,
"learning_rate": 9.367261418950459e-06,
"loss": 0.1855,
"step": 181
},
{
"epoch": 0.7413441955193483,
"grad_norm": 1.225757248706894,
"learning_rate": 9.355641243216798e-06,
"loss": 0.1729,
"step": 182
},
{
"epoch": 0.745417515274949,
"grad_norm": 1.1483380054973364,
"learning_rate": 9.343922678128854e-06,
"loss": 0.1078,
"step": 183
},
{
"epoch": 0.7494908350305499,
"grad_norm": 0.8222440765781929,
"learning_rate": 9.332105988396692e-06,
"loss": 0.1239,
"step": 184
},
{
"epoch": 0.7535641547861507,
"grad_norm": 0.9655962832595171,
"learning_rate": 9.3201914409469e-06,
"loss": 0.1309,
"step": 185
},
{
"epoch": 0.7576374745417516,
"grad_norm": 0.8060791719318856,
"learning_rate": 9.308179304916573e-06,
"loss": 0.1159,
"step": 186
},
{
"epoch": 0.7617107942973523,
"grad_norm": 0.7357782726661909,
"learning_rate": 9.29606985164723e-06,
"loss": 0.1052,
"step": 187
},
{
"epoch": 0.7657841140529531,
"grad_norm": 0.9536045205176826,
"learning_rate": 9.283863354678683e-06,
"loss": 0.1351,
"step": 188
},
{
"epoch": 0.769857433808554,
"grad_norm": 0.8771938059672718,
"learning_rate": 9.27156008974286e-06,
"loss": 0.1304,
"step": 189
},
{
"epoch": 0.7739307535641547,
"grad_norm": 0.7232888469506753,
"learning_rate": 9.259160334757575e-06,
"loss": 0.1054,
"step": 190
},
{
"epoch": 0.7780040733197556,
"grad_norm": 0.8295211262810136,
"learning_rate": 9.246664369820249e-06,
"loss": 0.1323,
"step": 191
},
{
"epoch": 0.7820773930753564,
"grad_norm": 1.546126242212441,
"learning_rate": 9.234072477201588e-06,
"loss": 0.2385,
"step": 192
},
{
"epoch": 0.7861507128309573,
"grad_norm": 1.3189210288828541,
"learning_rate": 9.2213849413392e-06,
"loss": 0.1312,
"step": 193
},
{
"epoch": 0.790224032586558,
"grad_norm": 0.6640416710388396,
"learning_rate": 9.208602048831176e-06,
"loss": 0.1032,
"step": 194
},
{
"epoch": 0.7942973523421588,
"grad_norm": 0.7975892776697048,
"learning_rate": 9.195724088429611e-06,
"loss": 0.1089,
"step": 195
},
{
"epoch": 0.7983706720977597,
"grad_norm": 0.706905690575772,
"learning_rate": 9.18275135103409e-06,
"loss": 0.1166,
"step": 196
},
{
"epoch": 0.8024439918533605,
"grad_norm": 0.8769448196441653,
"learning_rate": 9.169684129685099e-06,
"loss": 0.1317,
"step": 197
},
{
"epoch": 0.8065173116089613,
"grad_norm": 1.3681899543939136,
"learning_rate": 9.156522719557428e-06,
"loss": 0.1892,
"step": 198
},
{
"epoch": 0.8105906313645621,
"grad_norm": 1.0165895452906009,
"learning_rate": 9.143267417953486e-06,
"loss": 0.1526,
"step": 199
},
{
"epoch": 0.814663951120163,
"grad_norm": 0.9252869599364745,
"learning_rate": 9.129918524296596e-06,
"loss": 0.1791,
"step": 200
},
{
"epoch": 0.8187372708757638,
"grad_norm": 0.7566289195807724,
"learning_rate": 9.11647634012422e-06,
"loss": 0.1018,
"step": 201
},
{
"epoch": 0.8228105906313645,
"grad_norm": 0.7097020344942068,
"learning_rate": 9.102941169081167e-06,
"loss": 0.1174,
"step": 202
},
{
"epoch": 0.8268839103869654,
"grad_norm": 0.8335131746923946,
"learning_rate": 9.089313316912708e-06,
"loss": 0.14,
"step": 203
},
{
"epoch": 0.8309572301425662,
"grad_norm": 0.7934600650652943,
"learning_rate": 9.075593091457692e-06,
"loss": 0.1208,
"step": 204
},
{
"epoch": 0.835030549898167,
"grad_norm": 0.7614374059129773,
"learning_rate": 9.061780802641582e-06,
"loss": 0.1166,
"step": 205
},
{
"epoch": 0.8391038696537678,
"grad_norm": 0.7158974362347166,
"learning_rate": 9.047876762469451e-06,
"loss": 0.1046,
"step": 206
},
{
"epoch": 0.8431771894093686,
"grad_norm": 0.676023527010282,
"learning_rate": 9.033881285018945e-06,
"loss": 0.1049,
"step": 207
},
{
"epoch": 0.8472505091649695,
"grad_norm": 1.0542817712970116,
"learning_rate": 9.019794686433174e-06,
"loss": 0.1605,
"step": 208
},
{
"epoch": 0.8513238289205702,
"grad_norm": 0.791238316768574,
"learning_rate": 9.005617284913586e-06,
"loss": 0.1008,
"step": 209
},
{
"epoch": 0.8553971486761711,
"grad_norm": 1.3679274286147247,
"learning_rate": 8.991349400712772e-06,
"loss": 0.1174,
"step": 210
},
{
"epoch": 0.8594704684317719,
"grad_norm": 0.8904165376343479,
"learning_rate": 8.976991356127225e-06,
"loss": 0.1252,
"step": 211
},
{
"epoch": 0.8635437881873728,
"grad_norm": 0.6365058101639782,
"learning_rate": 8.962543475490068e-06,
"loss": 0.1054,
"step": 212
},
{
"epoch": 0.8676171079429735,
"grad_norm": 0.6899915324730952,
"learning_rate": 8.948006085163735e-06,
"loss": 0.1059,
"step": 213
},
{
"epoch": 0.8716904276985743,
"grad_norm": 0.7033665303348221,
"learning_rate": 8.933379513532575e-06,
"loss": 0.1055,
"step": 214
},
{
"epoch": 0.8757637474541752,
"grad_norm": 0.7051229848942461,
"learning_rate": 8.91866409099546e-06,
"loss": 0.1047,
"step": 215
},
{
"epoch": 0.879837067209776,
"grad_norm": 0.7365152922519815,
"learning_rate": 8.903860149958308e-06,
"loss": 0.1028,
"step": 216
},
{
"epoch": 0.8839103869653768,
"grad_norm": 0.8798834115379963,
"learning_rate": 8.888968024826575e-06,
"loss": 0.131,
"step": 217
},
{
"epoch": 0.8879837067209776,
"grad_norm": 0.8127281754244611,
"learning_rate": 8.873988051997702e-06,
"loss": 0.1014,
"step": 218
},
{
"epoch": 0.8920570264765784,
"grad_norm": 0.841292566312256,
"learning_rate": 8.85892056985352e-06,
"loss": 0.1335,
"step": 219
},
{
"epoch": 0.8961303462321792,
"grad_norm": 1.3435689868107352,
"learning_rate": 8.8437659187526e-06,
"loss": 0.2286,
"step": 220
},
{
"epoch": 0.90020366598778,
"grad_norm": 1.8444300521677208,
"learning_rate": 8.828524441022575e-06,
"loss": 0.1827,
"step": 221
},
{
"epoch": 0.9042769857433809,
"grad_norm": 0.7545922474592645,
"learning_rate": 8.813196480952393e-06,
"loss": 0.1027,
"step": 222
},
{
"epoch": 0.9083503054989817,
"grad_norm": 0.75537983489465,
"learning_rate": 8.797782384784549e-06,
"loss": 0.1198,
"step": 223
},
{
"epoch": 0.9124236252545825,
"grad_norm": 0.8104999041705286,
"learning_rate": 8.782282500707262e-06,
"loss": 0.1029,
"step": 224
},
{
"epoch": 0.9164969450101833,
"grad_norm": 0.8405282400775482,
"learning_rate": 8.766697178846611e-06,
"loss": 0.1241,
"step": 225
},
{
"epoch": 0.9205702647657841,
"grad_norm": 1.013551552697806,
"learning_rate": 8.751026771258622e-06,
"loss": 0.1343,
"step": 226
},
{
"epoch": 0.924643584521385,
"grad_norm": 0.6728989996123187,
"learning_rate": 8.735271631921322e-06,
"loss": 0.1058,
"step": 227
},
{
"epoch": 0.9287169042769857,
"grad_norm": 0.8690442261224494,
"learning_rate": 8.719432116726738e-06,
"loss": 0.1332,
"step": 228
},
{
"epoch": 0.9327902240325866,
"grad_norm": 0.9449187305589617,
"learning_rate": 8.703508583472855e-06,
"loss": 0.1451,
"step": 229
},
{
"epoch": 0.9368635437881874,
"grad_norm": 0.8067318337898685,
"learning_rate": 8.68750139185554e-06,
"loss": 0.1248,
"step": 230
},
{
"epoch": 0.9409368635437881,
"grad_norm": 0.7905017587261095,
"learning_rate": 8.671410903460416e-06,
"loss": 0.119,
"step": 231
},
{
"epoch": 0.945010183299389,
"grad_norm": 1.1238154965476772,
"learning_rate": 8.65523748175469e-06,
"loss": 0.1559,
"step": 232
},
{
"epoch": 0.9490835030549898,
"grad_norm": 1.1027211644152675,
"learning_rate": 8.63898149207895e-06,
"loss": 0.1693,
"step": 233
},
{
"epoch": 0.9531568228105907,
"grad_norm": 0.9411765578825619,
"learning_rate": 8.622643301638902e-06,
"loss": 0.1346,
"step": 234
},
{
"epoch": 0.9572301425661914,
"grad_norm": 0.6884466751221227,
"learning_rate": 8.606223279497081e-06,
"loss": 0.0968,
"step": 235
},
{
"epoch": 0.9613034623217923,
"grad_norm": 0.7219918781543078,
"learning_rate": 8.589721796564521e-06,
"loss": 0.0966,
"step": 236
},
{
"epoch": 0.9653767820773931,
"grad_norm": 0.7967809896092082,
"learning_rate": 8.57313922559236e-06,
"loss": 0.1201,
"step": 237
},
{
"epoch": 0.9694501018329938,
"grad_norm": 0.8113807921190012,
"learning_rate": 8.556475941163436e-06,
"loss": 0.1097,
"step": 238
},
{
"epoch": 0.9735234215885947,
"grad_norm": 1.0943551126152973,
"learning_rate": 8.539732319683817e-06,
"loss": 0.1552,
"step": 239
},
{
"epoch": 0.9775967413441955,
"grad_norm": 0.7854046329247982,
"learning_rate": 8.5229087393743e-06,
"loss": 0.1138,
"step": 240
},
{
"epoch": 0.9816700610997964,
"grad_norm": 1.1720562073286809,
"learning_rate": 8.506005580261872e-06,
"loss": 0.1525,
"step": 241
},
{
"epoch": 0.9857433808553971,
"grad_norm": 0.718895289386658,
"learning_rate": 8.489023224171114e-06,
"loss": 0.1082,
"step": 242
},
{
"epoch": 0.9898167006109979,
"grad_norm": 0.613834884154541,
"learning_rate": 8.47196205471559e-06,
"loss": 0.0877,
"step": 243
},
{
"epoch": 0.9938900203665988,
"grad_norm": 0.9789990123927295,
"learning_rate": 8.45482245728917e-06,
"loss": 0.1675,
"step": 244
},
{
"epoch": 0.9979633401221996,
"grad_norm": 1.5580291175140415,
"learning_rate": 8.437604819057336e-06,
"loss": 0.15,
"step": 245
},
{
"epoch": 1.0020366598778003,
"grad_norm": 0.7685763736473359,
"learning_rate": 8.420309528948422e-06,
"loss": 0.1072,
"step": 246
},
{
"epoch": 1.0061099796334012,
"grad_norm": 0.6434124354999965,
"learning_rate": 8.40293697764484e-06,
"loss": 0.0844,
"step": 247
},
{
"epoch": 1.010183299389002,
"grad_norm": 0.5841852692369695,
"learning_rate": 8.385487557574253e-06,
"loss": 0.0859,
"step": 248
},
{
"epoch": 1.0142566191446027,
"grad_norm": 0.6061435282600086,
"learning_rate": 8.367961662900704e-06,
"loss": 0.0809,
"step": 249
},
{
"epoch": 1.0183299389002036,
"grad_norm": 0.8866327026089017,
"learning_rate": 8.35035968951572e-06,
"loss": 0.0996,
"step": 250
},
{
"epoch": 1.0224032586558045,
"grad_norm": 0.789311514275454,
"learning_rate": 8.33268203502937e-06,
"loss": 0.0999,
"step": 251
},
{
"epoch": 1.0264765784114054,
"grad_norm": 0.7470915493623619,
"learning_rate": 8.314929098761268e-06,
"loss": 0.0836,
"step": 252
},
{
"epoch": 1.030549898167006,
"grad_norm": 0.7275329446393578,
"learning_rate": 8.297101281731576e-06,
"loss": 0.0866,
"step": 253
},
{
"epoch": 1.034623217922607,
"grad_norm": 0.7227258514093932,
"learning_rate": 8.279198986651925e-06,
"loss": 0.0901,
"step": 254
},
{
"epoch": 1.0386965376782078,
"grad_norm": 0.6146754288814568,
"learning_rate": 8.261222617916335e-06,
"loss": 0.0789,
"step": 255
},
{
"epoch": 1.0427698574338085,
"grad_norm": 0.8514917967475527,
"learning_rate": 8.243172581592066e-06,
"loss": 0.1017,
"step": 256
},
{
"epoch": 1.0468431771894093,
"grad_norm": 0.7579530053794002,
"learning_rate": 8.22504928541045e-06,
"loss": 0.085,
"step": 257
},
{
"epoch": 1.0509164969450102,
"grad_norm": 0.6252945360785674,
"learning_rate": 8.206853138757687e-06,
"loss": 0.0777,
"step": 258
},
{
"epoch": 1.054989816700611,
"grad_norm": 0.7706416070190195,
"learning_rate": 8.188584552665592e-06,
"loss": 0.0833,
"step": 259
},
{
"epoch": 1.0590631364562118,
"grad_norm": 0.7980656602534597,
"learning_rate": 8.17024393980231e-06,
"loss": 0.0968,
"step": 260
},
{
"epoch": 1.0631364562118126,
"grad_norm": 0.6040157250127779,
"learning_rate": 8.15183171446299e-06,
"loss": 0.0632,
"step": 261
},
{
"epoch": 1.0672097759674135,
"grad_norm": 0.6247841753185668,
"learning_rate": 8.133348292560442e-06,
"loss": 0.073,
"step": 262
},
{
"epoch": 1.0712830957230142,
"grad_norm": 0.9705375898534241,
"learning_rate": 8.114794091615718e-06,
"loss": 0.1158,
"step": 263
},
{
"epoch": 1.075356415478615,
"grad_norm": 0.5868673627641846,
"learning_rate": 8.096169530748708e-06,
"loss": 0.0616,
"step": 264
},
{
"epoch": 1.079429735234216,
"grad_norm": 0.6511676677923491,
"learning_rate": 8.077475030668647e-06,
"loss": 0.0924,
"step": 265
},
{
"epoch": 1.0835030549898166,
"grad_norm": 0.8628274720451822,
"learning_rate": 8.058711013664633e-06,
"loss": 0.0841,
"step": 266
},
{
"epoch": 1.0875763747454175,
"grad_norm": 0.6366922861725464,
"learning_rate": 8.039877903596069e-06,
"loss": 0.0781,
"step": 267
},
{
"epoch": 1.0916496945010183,
"grad_norm": 0.6679013936452773,
"learning_rate": 8.020976125883105e-06,
"loss": 0.074,
"step": 268
},
{
"epoch": 1.0957230142566192,
"grad_norm": 0.7115218188251841,
"learning_rate": 8.002006107497018e-06,
"loss": 0.0909,
"step": 269
},
{
"epoch": 1.0997963340122199,
"grad_norm": 0.6044721294592456,
"learning_rate": 7.982968276950568e-06,
"loss": 0.0682,
"step": 270
},
{
"epoch": 1.1038696537678208,
"grad_norm": 1.1198941616515987,
"learning_rate": 7.963863064288326e-06,
"loss": 0.1067,
"step": 271
},
{
"epoch": 1.1079429735234216,
"grad_norm": 0.5209754431453612,
"learning_rate": 7.944690901076949e-06,
"loss": 0.0743,
"step": 272
},
{
"epoch": 1.1120162932790225,
"grad_norm": 0.7595155294604947,
"learning_rate": 7.925452220395436e-06,
"loss": 0.0903,
"step": 273
},
{
"epoch": 1.1160896130346232,
"grad_norm": 0.692986942378061,
"learning_rate": 7.906147456825349e-06,
"loss": 0.0835,
"step": 274
},
{
"epoch": 1.120162932790224,
"grad_norm": 0.6090247595888676,
"learning_rate": 7.886777046440993e-06,
"loss": 0.0739,
"step": 275
},
{
"epoch": 1.124236252545825,
"grad_norm": 1.042879657433014,
"learning_rate": 7.867341426799562e-06,
"loss": 0.1219,
"step": 276
},
{
"epoch": 1.1283095723014256,
"grad_norm": 0.7996350912985092,
"learning_rate": 7.847841036931263e-06,
"loss": 0.0913,
"step": 277
},
{
"epoch": 1.1323828920570265,
"grad_norm": 0.5332223023774978,
"learning_rate": 7.828276317329388e-06,
"loss": 0.0757,
"step": 278
},
{
"epoch": 1.1364562118126273,
"grad_norm": 0.5635521535562537,
"learning_rate": 7.80864770994038e-06,
"loss": 0.0641,
"step": 279
},
{
"epoch": 1.140529531568228,
"grad_norm": 0.6987390220817623,
"learning_rate": 7.788955658153829e-06,
"loss": 0.0821,
"step": 280
},
{
"epoch": 1.1446028513238289,
"grad_norm": 0.6962375028972171,
"learning_rate": 7.769200606792476e-06,
"loss": 0.0686,
"step": 281
},
{
"epoch": 1.1486761710794298,
"grad_norm": 1.3654151723987282,
"learning_rate": 7.749383002102147e-06,
"loss": 0.0959,
"step": 282
},
{
"epoch": 1.1527494908350306,
"grad_norm": 0.8307371553295552,
"learning_rate": 7.72950329174169e-06,
"loss": 0.0859,
"step": 283
},
{
"epoch": 1.1568228105906313,
"grad_norm": 0.9349410963502993,
"learning_rate": 7.709561924772855e-06,
"loss": 0.0874,
"step": 284
},
{
"epoch": 1.1608961303462322,
"grad_norm": 1.1074351510282887,
"learning_rate": 7.689559351650142e-06,
"loss": 0.1012,
"step": 285
},
{
"epoch": 1.164969450101833,
"grad_norm": 0.9416925430977862,
"learning_rate": 7.66949602421064e-06,
"loss": 0.0923,
"step": 286
},
{
"epoch": 1.1690427698574337,
"grad_norm": 0.6905374196650557,
"learning_rate": 7.649372395663816e-06,
"loss": 0.0676,
"step": 287
},
{
"epoch": 1.1731160896130346,
"grad_norm": 0.7723265248527,
"learning_rate": 7.629188920581267e-06,
"loss": 0.0812,
"step": 288
},
{
"epoch": 1.1771894093686355,
"grad_norm": 0.815712697347911,
"learning_rate": 7.608946054886468e-06,
"loss": 0.0755,
"step": 289
},
{
"epoch": 1.1812627291242364,
"grad_norm": 0.8142685931991146,
"learning_rate": 7.588644255844464e-06,
"loss": 0.097,
"step": 290
},
{
"epoch": 1.185336048879837,
"grad_norm": 0.6448741497851783,
"learning_rate": 7.568283982051538e-06,
"loss": 0.0719,
"step": 291
},
{
"epoch": 1.189409368635438,
"grad_norm": 0.5839738850266732,
"learning_rate": 7.5478656934248626e-06,
"loss": 0.0715,
"step": 292
},
{
"epoch": 1.1934826883910388,
"grad_norm": 0.6469439075383099,
"learning_rate": 7.527389851192099e-06,
"loss": 0.0745,
"step": 293
},
{
"epoch": 1.1975560081466394,
"grad_norm": 0.7083801028247904,
"learning_rate": 7.506856917880989e-06,
"loss": 0.077,
"step": 294
},
{
"epoch": 1.2016293279022403,
"grad_norm": 0.5584165288788803,
"learning_rate": 7.486267357308896e-06,
"loss": 0.0757,
"step": 295
},
{
"epoch": 1.2057026476578412,
"grad_norm": 0.7093198997979762,
"learning_rate": 7.465621634572336e-06,
"loss": 0.0821,
"step": 296
},
{
"epoch": 1.2097759674134418,
"grad_norm": 0.8170300802168915,
"learning_rate": 7.444920216036473e-06,
"loss": 0.0857,
"step": 297
},
{
"epoch": 1.2138492871690427,
"grad_norm": 0.6052440895883001,
"learning_rate": 7.4241635693245766e-06,
"loss": 0.0706,
"step": 298
},
{
"epoch": 1.2179226069246436,
"grad_norm": 0.5358799947933865,
"learning_rate": 7.40335216330746e-06,
"loss": 0.0698,
"step": 299
},
{
"epoch": 1.2219959266802445,
"grad_norm": 0.57379599749216,
"learning_rate": 7.382486468092899e-06,
"loss": 0.0778,
"step": 300
},
{
"epoch": 1.2260692464358451,
"grad_norm": 1.9428087979785686,
"learning_rate": 7.361566955014999e-06,
"loss": 0.0881,
"step": 301
},
{
"epoch": 1.230142566191446,
"grad_norm": 0.8052892598402892,
"learning_rate": 7.340594096623559e-06,
"loss": 0.0807,
"step": 302
},
{
"epoch": 1.234215885947047,
"grad_norm": 0.5697887043869344,
"learning_rate": 7.319568366673389e-06,
"loss": 0.0761,
"step": 303
},
{
"epoch": 1.2382892057026478,
"grad_norm": 0.8103909158109461,
"learning_rate": 7.2984902401136115e-06,
"loss": 0.084,
"step": 304
},
{
"epoch": 1.2423625254582484,
"grad_norm": 0.7842393785164873,
"learning_rate": 7.277360193076936e-06,
"loss": 0.0762,
"step": 305
},
{
"epoch": 1.2464358452138493,
"grad_norm": 0.5722519256754022,
"learning_rate": 7.256178702868899e-06,
"loss": 0.0723,
"step": 306
},
{
"epoch": 1.2505091649694502,
"grad_norm": 0.5772908994522038,
"learning_rate": 7.234946247957087e-06,
"loss": 0.0809,
"step": 307
},
{
"epoch": 1.2545824847250509,
"grad_norm": 0.7436038482880873,
"learning_rate": 7.213663307960321e-06,
"loss": 0.0822,
"step": 308
},
{
"epoch": 1.2586558044806517,
"grad_norm": 0.9056276676473372,
"learning_rate": 7.192330363637832e-06,
"loss": 0.1005,
"step": 309
},
{
"epoch": 1.2627291242362526,
"grad_norm": 0.6358405282433213,
"learning_rate": 7.170947896878392e-06,
"loss": 0.0737,
"step": 310
},
{
"epoch": 1.2668024439918533,
"grad_norm": 0.5954677442268389,
"learning_rate": 7.149516390689433e-06,
"loss": 0.0694,
"step": 311
},
{
"epoch": 1.2708757637474541,
"grad_norm": 0.6088987642638558,
"learning_rate": 7.12803632918614e-06,
"loss": 0.0712,
"step": 312
},
{
"epoch": 1.274949083503055,
"grad_norm": 0.6327020826012894,
"learning_rate": 7.1065081975805086e-06,
"loss": 0.0709,
"step": 313
},
{
"epoch": 1.2790224032586557,
"grad_norm": 0.8052023527305602,
"learning_rate": 7.084932482170385e-06,
"loss": 0.0884,
"step": 314
},
{
"epoch": 1.2830957230142566,
"grad_norm": 0.6196057921350728,
"learning_rate": 7.063309670328491e-06,
"loss": 0.0791,
"step": 315
},
{
"epoch": 1.2871690427698574,
"grad_norm": 0.6874689974891509,
"learning_rate": 7.041640250491398e-06,
"loss": 0.0725,
"step": 316
},
{
"epoch": 1.2912423625254583,
"grad_norm": 0.5750839664752618,
"learning_rate": 7.019924712148511e-06,
"loss": 0.0676,
"step": 317
},
{
"epoch": 1.2953156822810592,
"grad_norm": 0.7742197007819996,
"learning_rate": 6.998163545830998e-06,
"loss": 0.0734,
"step": 318
},
{
"epoch": 1.2993890020366599,
"grad_norm": 0.6195282494860945,
"learning_rate": 6.976357243100718e-06,
"loss": 0.0775,
"step": 319
},
{
"epoch": 1.3034623217922607,
"grad_norm": 0.6875577568472137,
"learning_rate": 6.954506296539112e-06,
"loss": 0.0878,
"step": 320
},
{
"epoch": 1.3075356415478616,
"grad_norm": 0.7320126566140536,
"learning_rate": 6.9326111997360775e-06,
"loss": 0.081,
"step": 321
},
{
"epoch": 1.3116089613034623,
"grad_norm": 0.843149661393896,
"learning_rate": 6.910672447278827e-06,
"loss": 0.0918,
"step": 322
},
{
"epoch": 1.3156822810590632,
"grad_norm": 0.6249054893219663,
"learning_rate": 6.8886905347406985e-06,
"loss": 0.0859,
"step": 323
},
{
"epoch": 1.319755600814664,
"grad_norm": 0.6921145932385908,
"learning_rate": 6.866665958669976e-06,
"loss": 0.0916,
"step": 324
},
{
"epoch": 1.3238289205702647,
"grad_norm": 0.5550915898105676,
"learning_rate": 6.844599216578667e-06,
"loss": 0.0755,
"step": 325
},
{
"epoch": 1.3279022403258656,
"grad_norm": 0.8923390816851975,
"learning_rate": 6.822490806931262e-06,
"loss": 0.0903,
"step": 326
},
{
"epoch": 1.3319755600814664,
"grad_norm": 0.6246266762679324,
"learning_rate": 6.800341229133486e-06,
"loss": 0.0833,
"step": 327
},
{
"epoch": 1.336048879837067,
"grad_norm": 0.8145107114030108,
"learning_rate": 6.778150983520999e-06,
"loss": 0.1131,
"step": 328
},
{
"epoch": 1.340122199592668,
"grad_norm": 0.6593824730687659,
"learning_rate": 6.755920571348111e-06,
"loss": 0.0783,
"step": 329
},
{
"epoch": 1.3441955193482689,
"grad_norm": 0.7871250490461288,
"learning_rate": 6.73365049477645e-06,
"loss": 0.0791,
"step": 330
},
{
"epoch": 1.3482688391038695,
"grad_norm": 1.5668561976011524,
"learning_rate": 6.711341256863623e-06,
"loss": 0.116,
"step": 331
},
{
"epoch": 1.3523421588594704,
"grad_norm": 0.6238740583787221,
"learning_rate": 6.688993361551847e-06,
"loss": 0.079,
"step": 332
},
{
"epoch": 1.3564154786150713,
"grad_norm": 0.7619408457347728,
"learning_rate": 6.66660731365657e-06,
"loss": 0.0876,
"step": 333
},
{
"epoch": 1.3604887983706722,
"grad_norm": 0.5260527244857047,
"learning_rate": 6.64418361885507e-06,
"loss": 0.08,
"step": 334
},
{
"epoch": 1.364562118126273,
"grad_norm": 0.7273285421050715,
"learning_rate": 6.621722783675024e-06,
"loss": 0.0803,
"step": 335
},
{
"epoch": 1.3686354378818737,
"grad_norm": 0.5976232642003504,
"learning_rate": 6.599225315483076e-06,
"loss": 0.0773,
"step": 336
},
{
"epoch": 1.3727087576374746,
"grad_norm": 0.5354838304438007,
"learning_rate": 6.576691722473368e-06,
"loss": 0.0699,
"step": 337
},
{
"epoch": 1.3767820773930755,
"grad_norm": 0.7164176676979538,
"learning_rate": 6.554122513656065e-06,
"loss": 0.0898,
"step": 338
},
{
"epoch": 1.3808553971486761,
"grad_norm": 0.694148331721139,
"learning_rate": 6.531518198845854e-06,
"loss": 0.0796,
"step": 339
},
{
"epoch": 1.384928716904277,
"grad_norm": 0.8668103406807447,
"learning_rate": 6.508879288650431e-06,
"loss": 0.0864,
"step": 340
},
{
"epoch": 1.3890020366598779,
"grad_norm": 0.905281439053181,
"learning_rate": 6.486206294458966e-06,
"loss": 0.1032,
"step": 341
},
{
"epoch": 1.3930753564154785,
"grad_norm": 0.6090466800578089,
"learning_rate": 6.463499728430549e-06,
"loss": 0.0739,
"step": 342
},
{
"epoch": 1.3971486761710794,
"grad_norm": 0.870879688990723,
"learning_rate": 6.4407601034826225e-06,
"loss": 0.0911,
"step": 343
},
{
"epoch": 1.4012219959266803,
"grad_norm": 0.5520211139425468,
"learning_rate": 6.417987933279397e-06,
"loss": 0.0763,
"step": 344
},
{
"epoch": 1.405295315682281,
"grad_norm": 0.567715626333358,
"learning_rate": 6.395183732220242e-06,
"loss": 0.0685,
"step": 345
},
{
"epoch": 1.4093686354378818,
"grad_norm": 0.6211209151590639,
"learning_rate": 6.372348015428077e-06,
"loss": 0.0763,
"step": 346
},
{
"epoch": 1.4134419551934827,
"grad_norm": 0.6300074755034071,
"learning_rate": 6.349481298737723e-06,
"loss": 0.0811,
"step": 347
},
{
"epoch": 1.4175152749490836,
"grad_norm": 0.5995929997342367,
"learning_rate": 6.32658409868426e-06,
"loss": 0.0651,
"step": 348
},
{
"epoch": 1.4215885947046842,
"grad_norm": 0.6049082049048853,
"learning_rate": 6.303656932491349e-06,
"loss": 0.0763,
"step": 349
},
{
"epoch": 1.4256619144602851,
"grad_norm": 0.9822040378006914,
"learning_rate": 6.280700318059563e-06,
"loss": 0.1054,
"step": 350
},
{
"epoch": 1.429735234215886,
"grad_norm": 0.8449953491667241,
"learning_rate": 6.257714773954674e-06,
"loss": 0.102,
"step": 351
},
{
"epoch": 1.4338085539714869,
"grad_norm": 0.5999494361508966,
"learning_rate": 6.234700819395946e-06,
"loss": 0.0813,
"step": 352
},
{
"epoch": 1.4378818737270875,
"grad_norm": 0.6512734204457252,
"learning_rate": 6.211658974244407e-06,
"loss": 0.0829,
"step": 353
},
{
"epoch": 1.4419551934826884,
"grad_norm": 0.616189473053879,
"learning_rate": 6.1885897589911e-06,
"loss": 0.0782,
"step": 354
},
{
"epoch": 1.4460285132382893,
"grad_norm": 0.9515071645483372,
"learning_rate": 6.1654936947453355e-06,
"loss": 0.0975,
"step": 355
},
{
"epoch": 1.45010183299389,
"grad_norm": 0.6021477200299695,
"learning_rate": 6.142371303222909e-06,
"loss": 0.071,
"step": 356
},
{
"epoch": 1.4541751527494908,
"grad_norm": 0.678653399667039,
"learning_rate": 6.119223106734328e-06,
"loss": 0.0812,
"step": 357
},
{
"epoch": 1.4582484725050917,
"grad_norm": 0.6291449966527708,
"learning_rate": 6.0960496281729995e-06,
"loss": 0.0689,
"step": 358
},
{
"epoch": 1.4623217922606924,
"grad_norm": 0.5932315331898775,
"learning_rate": 6.072851391003432e-06,
"loss": 0.075,
"step": 359
},
{
"epoch": 1.4663951120162932,
"grad_norm": 0.6492759903008403,
"learning_rate": 6.0496289192494e-06,
"loss": 0.0851,
"step": 360
},
{
"epoch": 1.4704684317718941,
"grad_norm": 0.8227230707733793,
"learning_rate": 6.026382737482116e-06,
"loss": 0.0939,
"step": 361
},
{
"epoch": 1.4745417515274948,
"grad_norm": 0.7553836500415501,
"learning_rate": 6.003113370808375e-06,
"loss": 0.089,
"step": 362
},
{
"epoch": 1.4786150712830957,
"grad_norm": 0.8341775236686235,
"learning_rate": 5.979821344858695e-06,
"loss": 0.1087,
"step": 363
},
{
"epoch": 1.4826883910386965,
"grad_norm": 0.7407455393175063,
"learning_rate": 5.956507185775441e-06,
"loss": 0.0875,
"step": 364
},
{
"epoch": 1.4867617107942974,
"grad_norm": 0.6302976593180085,
"learning_rate": 5.933171420200946e-06,
"loss": 0.074,
"step": 365
},
{
"epoch": 1.4908350305498983,
"grad_norm": 0.6017971359417691,
"learning_rate": 5.909814575265609e-06,
"loss": 0.0771,
"step": 366
},
{
"epoch": 1.494908350305499,
"grad_norm": 0.5327101770243312,
"learning_rate": 5.88643717857599e-06,
"loss": 0.064,
"step": 367
},
{
"epoch": 1.4989816700610998,
"grad_norm": 0.5530940520217251,
"learning_rate": 5.863039758202889e-06,
"loss": 0.0732,
"step": 368
},
{
"epoch": 1.5030549898167007,
"grad_norm": 0.575697348750185,
"learning_rate": 5.839622842669423e-06,
"loss": 0.0794,
"step": 369
},
{
"epoch": 1.5071283095723014,
"grad_norm": 0.7879263375060825,
"learning_rate": 5.816186960939084e-06,
"loss": 0.0873,
"step": 370
},
{
"epoch": 1.5112016293279023,
"grad_norm": 0.6477317557359421,
"learning_rate": 5.7927326424037875e-06,
"loss": 0.0808,
"step": 371
},
{
"epoch": 1.5152749490835031,
"grad_norm": 0.6424367902201409,
"learning_rate": 5.7692604168719225e-06,
"loss": 0.0743,
"step": 372
},
{
"epoch": 1.5193482688391038,
"grad_norm": 0.5205522605427637,
"learning_rate": 5.745770814556373e-06,
"loss": 0.0737,
"step": 373
},
{
"epoch": 1.5234215885947047,
"grad_norm": 0.8437328046734178,
"learning_rate": 5.722264366062549e-06,
"loss": 0.1075,
"step": 374
},
{
"epoch": 1.5274949083503055,
"grad_norm": 0.705710140903758,
"learning_rate": 5.698741602376395e-06,
"loss": 0.0854,
"step": 375
},
{
"epoch": 1.5315682281059062,
"grad_norm": 0.5479078071724787,
"learning_rate": 5.675203054852403e-06,
"loss": 0.0735,
"step": 376
},
{
"epoch": 1.535641547861507,
"grad_norm": 0.628783685599969,
"learning_rate": 5.651649255201603e-06,
"loss": 0.0893,
"step": 377
},
{
"epoch": 1.539714867617108,
"grad_norm": 0.6393301256468628,
"learning_rate": 5.628080735479553e-06,
"loss": 0.0808,
"step": 378
},
{
"epoch": 1.5437881873727086,
"grad_norm": 0.5419627427127232,
"learning_rate": 5.604498028074323e-06,
"loss": 0.0693,
"step": 379
},
{
"epoch": 1.5478615071283097,
"grad_norm": 0.5585947140971957,
"learning_rate": 5.580901665694471e-06,
"loss": 0.0708,
"step": 380
},
{
"epoch": 1.5519348268839104,
"grad_norm": 0.7630225067115134,
"learning_rate": 5.557292181357003e-06,
"loss": 0.0916,
"step": 381
},
{
"epoch": 1.556008146639511,
"grad_norm": 0.7132672152748122,
"learning_rate": 5.533670108375334e-06,
"loss": 0.075,
"step": 382
},
{
"epoch": 1.5600814663951121,
"grad_norm": 0.6317003602484598,
"learning_rate": 5.510035980347249e-06,
"loss": 0.0629,
"step": 383
},
{
"epoch": 1.5641547861507128,
"grad_norm": 1.0965759213987818,
"learning_rate": 5.486390331142841e-06,
"loss": 0.129,
"step": 384
},
{
"epoch": 1.5682281059063137,
"grad_norm": 0.9961574441023341,
"learning_rate": 5.462733694892452e-06,
"loss": 0.0994,
"step": 385
},
{
"epoch": 1.5723014256619146,
"grad_norm": 0.55138988687004,
"learning_rate": 5.439066605974615e-06,
"loss": 0.0884,
"step": 386
},
{
"epoch": 1.5763747454175152,
"grad_norm": 0.642013542309746,
"learning_rate": 5.415389599003972e-06,
"loss": 0.0758,
"step": 387
},
{
"epoch": 1.580448065173116,
"grad_norm": 0.7003801753473788,
"learning_rate": 5.391703208819209e-06,
"loss": 0.0822,
"step": 388
},
{
"epoch": 1.584521384928717,
"grad_norm": 0.641290834723853,
"learning_rate": 5.368007970470964e-06,
"loss": 0.0794,
"step": 389
},
{
"epoch": 1.5885947046843176,
"grad_norm": 0.7220121980133926,
"learning_rate": 5.344304419209748e-06,
"loss": 0.0908,
"step": 390
},
{
"epoch": 1.5926680244399185,
"grad_norm": 0.5338391641253294,
"learning_rate": 5.3205930904738544e-06,
"loss": 0.065,
"step": 391
},
{
"epoch": 1.5967413441955194,
"grad_norm": 0.8220572260271829,
"learning_rate": 5.296874519877256e-06,
"loss": 0.1063,
"step": 392
},
{
"epoch": 1.60081466395112,
"grad_norm": 0.6948752546108359,
"learning_rate": 5.273149243197517e-06,
"loss": 0.092,
"step": 393
},
{
"epoch": 1.6048879837067211,
"grad_norm": 0.5973741327673737,
"learning_rate": 5.2494177963636785e-06,
"loss": 0.0781,
"step": 394
},
{
"epoch": 1.6089613034623218,
"grad_norm": 0.7082285511834933,
"learning_rate": 5.225680715444168e-06,
"loss": 0.0844,
"step": 395
},
{
"epoch": 1.6130346232179225,
"grad_norm": 0.5776337469174156,
"learning_rate": 5.201938536634674e-06,
"loss": 0.0755,
"step": 396
},
{
"epoch": 1.6171079429735236,
"grad_norm": 0.6439617862748772,
"learning_rate": 5.178191796246043e-06,
"loss": 0.0692,
"step": 397
},
{
"epoch": 1.6211812627291242,
"grad_norm": 0.693976310427531,
"learning_rate": 5.154441030692162e-06,
"loss": 0.0844,
"step": 398
},
{
"epoch": 1.625254582484725,
"grad_norm": 0.6310937269228004,
"learning_rate": 5.1306867764778445e-06,
"loss": 0.0605,
"step": 399
},
{
"epoch": 1.629327902240326,
"grad_norm": 0.5557882712930212,
"learning_rate": 5.106929570186706e-06,
"loss": 0.0616,
"step": 400
},
{
"epoch": 1.6334012219959266,
"grad_norm": 0.7105865341669582,
"learning_rate": 5.083169948469049e-06,
"loss": 0.0888,
"step": 401
},
{
"epoch": 1.6374745417515275,
"grad_norm": 0.785914280291214,
"learning_rate": 5.059408448029737e-06,
"loss": 0.0924,
"step": 402
},
{
"epoch": 1.6415478615071284,
"grad_norm": 0.760854086300454,
"learning_rate": 5.0356456056160715e-06,
"loss": 0.0899,
"step": 403
},
{
"epoch": 1.645621181262729,
"grad_norm": 0.5160083716123763,
"learning_rate": 5.0118819580056686e-06,
"loss": 0.0676,
"step": 404
},
{
"epoch": 1.64969450101833,
"grad_norm": 0.5926381494692319,
"learning_rate": 4.988118041994332e-06,
"loss": 0.0719,
"step": 405
},
{
"epoch": 1.6537678207739308,
"grad_norm": 0.5387161542722511,
"learning_rate": 4.964354394383929e-06,
"loss": 0.0757,
"step": 406
},
{
"epoch": 1.6578411405295315,
"grad_norm": 0.5028703565858573,
"learning_rate": 4.940591551970264e-06,
"loss": 0.066,
"step": 407
},
{
"epoch": 1.6619144602851323,
"grad_norm": 0.5736039189545704,
"learning_rate": 4.9168300515309515e-06,
"loss": 0.0724,
"step": 408
},
{
"epoch": 1.6659877800407332,
"grad_norm": 0.984691517709554,
"learning_rate": 4.8930704298132965e-06,
"loss": 0.1022,
"step": 409
},
{
"epoch": 1.6700610997963339,
"grad_norm": 0.5927643518511072,
"learning_rate": 4.869313223522159e-06,
"loss": 0.0718,
"step": 410
},
{
"epoch": 1.674134419551935,
"grad_norm": 0.566899954027869,
"learning_rate": 4.845558969307839e-06,
"loss": 0.0707,
"step": 411
},
{
"epoch": 1.6782077393075356,
"grad_norm": 0.5788045225856518,
"learning_rate": 4.821808203753959e-06,
"loss": 0.0785,
"step": 412
},
{
"epoch": 1.6822810590631363,
"grad_norm": 0.5224303081188956,
"learning_rate": 4.798061463365327e-06,
"loss": 0.0755,
"step": 413
},
{
"epoch": 1.6863543788187374,
"grad_norm": 0.5256289426946346,
"learning_rate": 4.774319284555833e-06,
"loss": 0.0725,
"step": 414
},
{
"epoch": 1.690427698574338,
"grad_norm": 0.5514388613040209,
"learning_rate": 4.7505822036363214e-06,
"loss": 0.0698,
"step": 415
},
{
"epoch": 1.694501018329939,
"grad_norm": 0.7507069842501244,
"learning_rate": 4.726850756802486e-06,
"loss": 0.0779,
"step": 416
},
{
"epoch": 1.6985743380855398,
"grad_norm": 0.5535879129510451,
"learning_rate": 4.703125480122747e-06,
"loss": 0.0677,
"step": 417
},
{
"epoch": 1.7026476578411405,
"grad_norm": 0.7586101007852933,
"learning_rate": 4.679406909526147e-06,
"loss": 0.0959,
"step": 418
},
{
"epoch": 1.7067209775967414,
"grad_norm": 0.5235003242969455,
"learning_rate": 4.655695580790254e-06,
"loss": 0.0782,
"step": 419
},
{
"epoch": 1.7107942973523422,
"grad_norm": 0.8957116693381422,
"learning_rate": 4.631992029529037e-06,
"loss": 0.104,
"step": 420
},
{
"epoch": 1.7148676171079429,
"grad_norm": 0.5397767746056776,
"learning_rate": 4.608296791180793e-06,
"loss": 0.0794,
"step": 421
},
{
"epoch": 1.7189409368635438,
"grad_norm": 0.5510200443982937,
"learning_rate": 4.584610400996028e-06,
"loss": 0.0727,
"step": 422
},
{
"epoch": 1.7230142566191446,
"grad_norm": 0.7218539657297133,
"learning_rate": 4.560933394025386e-06,
"loss": 0.0812,
"step": 423
},
{
"epoch": 1.7270875763747453,
"grad_norm": 0.6450718218339647,
"learning_rate": 4.537266305107549e-06,
"loss": 0.0857,
"step": 424
},
{
"epoch": 1.7311608961303462,
"grad_norm": 0.5240562658243174,
"learning_rate": 4.513609668857162e-06,
"loss": 0.0658,
"step": 425
},
{
"epoch": 1.735234215885947,
"grad_norm": 0.5203639231556227,
"learning_rate": 4.489964019652752e-06,
"loss": 0.0637,
"step": 426
},
{
"epoch": 1.7393075356415477,
"grad_norm": 0.7157158022653508,
"learning_rate": 4.4663298916246665e-06,
"loss": 0.0878,
"step": 427
},
{
"epoch": 1.7433808553971488,
"grad_norm": 0.6919816756435726,
"learning_rate": 4.442707818642999e-06,
"loss": 0.0801,
"step": 428
},
{
"epoch": 1.7474541751527495,
"grad_norm": 0.7381380434766637,
"learning_rate": 4.419098334305529e-06,
"loss": 0.0831,
"step": 429
},
{
"epoch": 1.7515274949083504,
"grad_norm": 0.720540642558599,
"learning_rate": 4.395501971925677e-06,
"loss": 0.0782,
"step": 430
},
{
"epoch": 1.7556008146639512,
"grad_norm": 0.6824501618323201,
"learning_rate": 4.371919264520449e-06,
"loss": 0.0795,
"step": 431
},
{
"epoch": 1.759674134419552,
"grad_norm": 0.7572127655566931,
"learning_rate": 4.348350744798399e-06,
"loss": 0.0798,
"step": 432
},
{
"epoch": 1.7637474541751528,
"grad_norm": 0.5634906535431049,
"learning_rate": 4.324796945147598e-06,
"loss": 0.0728,
"step": 433
},
{
"epoch": 1.7678207739307537,
"grad_norm": 0.7686272783650896,
"learning_rate": 4.301258397623606e-06,
"loss": 0.0782,
"step": 434
},
{
"epoch": 1.7718940936863543,
"grad_norm": 0.712370232236961,
"learning_rate": 4.2777356339374526e-06,
"loss": 0.0856,
"step": 435
},
{
"epoch": 1.7759674134419552,
"grad_norm": 0.9290726298620688,
"learning_rate": 4.254229185443628e-06,
"loss": 0.0783,
"step": 436
},
{
"epoch": 1.780040733197556,
"grad_norm": 0.5716455342190312,
"learning_rate": 4.230739583128078e-06,
"loss": 0.0701,
"step": 437
},
{
"epoch": 1.7841140529531567,
"grad_norm": 1.1820087104502222,
"learning_rate": 4.2072673575962125e-06,
"loss": 0.0977,
"step": 438
},
{
"epoch": 1.7881873727087576,
"grad_norm": 0.935231069436012,
"learning_rate": 4.183813039060919e-06,
"loss": 0.1103,
"step": 439
},
{
"epoch": 1.7922606924643585,
"grad_norm": 0.5110831181130548,
"learning_rate": 4.160377157330579e-06,
"loss": 0.0787,
"step": 440
},
{
"epoch": 1.7963340122199591,
"grad_norm": 0.6036766363233732,
"learning_rate": 4.136960241797113e-06,
"loss": 0.0648,
"step": 441
},
{
"epoch": 1.8004073319755602,
"grad_norm": 0.6253574899226054,
"learning_rate": 4.113562821424012e-06,
"loss": 0.0856,
"step": 442
},
{
"epoch": 1.804480651731161,
"grad_norm": 0.5737047935293089,
"learning_rate": 4.090185424734392e-06,
"loss": 0.0769,
"step": 443
},
{
"epoch": 1.8085539714867616,
"grad_norm": 0.6675899229386297,
"learning_rate": 4.066828579799054e-06,
"loss": 0.0761,
"step": 444
},
{
"epoch": 1.8126272912423627,
"grad_norm": 0.5292920337446484,
"learning_rate": 4.043492814224559e-06,
"loss": 0.0684,
"step": 445
},
{
"epoch": 1.8167006109979633,
"grad_norm": 0.7551251294419339,
"learning_rate": 4.020178655141307e-06,
"loss": 0.0792,
"step": 446
},
{
"epoch": 1.8207739307535642,
"grad_norm": 0.6230513981956748,
"learning_rate": 3.9968866291916254e-06,
"loss": 0.082,
"step": 447
},
{
"epoch": 1.824847250509165,
"grad_norm": 0.5934762051634166,
"learning_rate": 3.973617262517886e-06,
"loss": 0.0638,
"step": 448
},
{
"epoch": 1.8289205702647657,
"grad_norm": 0.675289075760882,
"learning_rate": 3.950371080750602e-06,
"loss": 0.0795,
"step": 449
},
{
"epoch": 1.8329938900203666,
"grad_norm": 0.9147897851217298,
"learning_rate": 3.927148608996569e-06,
"loss": 0.1063,
"step": 450
},
{
"epoch": 1.8370672097759675,
"grad_norm": 1.1105124935695379,
"learning_rate": 3.903950371827001e-06,
"loss": 0.087,
"step": 451
},
{
"epoch": 1.8411405295315681,
"grad_norm": 0.5889228060416761,
"learning_rate": 3.880776893265673e-06,
"loss": 0.0767,
"step": 452
},
{
"epoch": 1.845213849287169,
"grad_norm": 0.5032013275393701,
"learning_rate": 3.85762869677709e-06,
"loss": 0.0575,
"step": 453
},
{
"epoch": 1.84928716904277,
"grad_norm": 0.6559816525725005,
"learning_rate": 3.834506305254667e-06,
"loss": 0.0896,
"step": 454
},
{
"epoch": 1.8533604887983706,
"grad_norm": 0.7532212211298266,
"learning_rate": 3.811410241008902e-06,
"loss": 0.0856,
"step": 455
},
{
"epoch": 1.8574338085539714,
"grad_norm": 0.5738261899311171,
"learning_rate": 3.788341025755595e-06,
"loss": 0.0733,
"step": 456
},
{
"epoch": 1.8615071283095723,
"grad_norm": 0.7208121405309541,
"learning_rate": 3.765299180604055e-06,
"loss": 0.0829,
"step": 457
},
{
"epoch": 1.865580448065173,
"grad_norm": 0.5296431964709345,
"learning_rate": 3.7422852260453274e-06,
"loss": 0.0704,
"step": 458
},
{
"epoch": 1.869653767820774,
"grad_norm": 0.6719188744341698,
"learning_rate": 3.719299681940437e-06,
"loss": 0.09,
"step": 459
},
{
"epoch": 1.8737270875763747,
"grad_norm": 0.5867743083933351,
"learning_rate": 3.696343067508651e-06,
"loss": 0.0816,
"step": 460
},
{
"epoch": 1.8778004073319754,
"grad_norm": 0.49214441808140924,
"learning_rate": 3.673415901315743e-06,
"loss": 0.0648,
"step": 461
},
{
"epoch": 1.8818737270875765,
"grad_norm": 0.5871371037925425,
"learning_rate": 3.650518701262278e-06,
"loss": 0.0732,
"step": 462
},
{
"epoch": 1.8859470468431772,
"grad_norm": 0.6415787500205045,
"learning_rate": 3.6276519845719237e-06,
"loss": 0.0731,
"step": 463
},
{
"epoch": 1.890020366598778,
"grad_norm": 0.7505406685838056,
"learning_rate": 3.6048162677797595e-06,
"loss": 0.0847,
"step": 464
},
{
"epoch": 1.894093686354379,
"grad_norm": 0.5203285332468817,
"learning_rate": 3.582012066720605e-06,
"loss": 0.0666,
"step": 465
},
{
"epoch": 1.8981670061099796,
"grad_norm": 0.5664388898818371,
"learning_rate": 3.559239896517379e-06,
"loss": 0.0814,
"step": 466
},
{
"epoch": 1.9022403258655805,
"grad_norm": 0.9911644175743987,
"learning_rate": 3.536500271569452e-06,
"loss": 0.0766,
"step": 467
},
{
"epoch": 1.9063136456211813,
"grad_norm": 0.5541020704419577,
"learning_rate": 3.5137937055410343e-06,
"loss": 0.0677,
"step": 468
},
{
"epoch": 1.910386965376782,
"grad_norm": 0.7733672524931661,
"learning_rate": 3.4911207113495703e-06,
"loss": 0.0883,
"step": 469
},
{
"epoch": 1.9144602851323829,
"grad_norm": 0.5994378228763769,
"learning_rate": 3.4684818011541484e-06,
"loss": 0.0779,
"step": 470
},
{
"epoch": 1.9185336048879837,
"grad_norm": 0.7925341371043554,
"learning_rate": 3.4458774863439366e-06,
"loss": 0.0882,
"step": 471
},
{
"epoch": 1.9226069246435844,
"grad_norm": 0.8007592704584606,
"learning_rate": 3.423308277526633e-06,
"loss": 0.088,
"step": 472
},
{
"epoch": 1.9266802443991853,
"grad_norm": 0.6130809497009315,
"learning_rate": 3.4007746845169253e-06,
"loss": 0.0836,
"step": 473
},
{
"epoch": 1.9307535641547862,
"grad_norm": 0.6462482272234705,
"learning_rate": 3.3782772163249767e-06,
"loss": 0.0671,
"step": 474
},
{
"epoch": 1.9348268839103868,
"grad_norm": 0.4781126271398999,
"learning_rate": 3.3558163811449317e-06,
"loss": 0.0694,
"step": 475
},
{
"epoch": 1.938900203665988,
"grad_norm": 0.7328378092623428,
"learning_rate": 3.3333926863434317e-06,
"loss": 0.0847,
"step": 476
},
{
"epoch": 1.9429735234215886,
"grad_norm": 0.6002434419865161,
"learning_rate": 3.311006638448155e-06,
"loss": 0.0724,
"step": 477
},
{
"epoch": 1.9470468431771895,
"grad_norm": 0.6276269450961816,
"learning_rate": 3.288658743136378e-06,
"loss": 0.066,
"step": 478
},
{
"epoch": 1.9511201629327903,
"grad_norm": 0.8586874267355604,
"learning_rate": 3.2663495052235505e-06,
"loss": 0.1077,
"step": 479
},
{
"epoch": 1.955193482688391,
"grad_norm": 0.6160197884131686,
"learning_rate": 3.2440794286518896e-06,
"loss": 0.085,
"step": 480
},
{
"epoch": 1.9592668024439919,
"grad_norm": 0.5494876667673338,
"learning_rate": 3.2218490164790015e-06,
"loss": 0.0656,
"step": 481
},
{
"epoch": 1.9633401221995928,
"grad_norm": 0.5953585828338879,
"learning_rate": 3.199658770866515e-06,
"loss": 0.0754,
"step": 482
},
{
"epoch": 1.9674134419551934,
"grad_norm": 0.5401626625339145,
"learning_rate": 3.1775091930687374e-06,
"loss": 0.0668,
"step": 483
},
{
"epoch": 1.9714867617107943,
"grad_norm": 0.5111554952361316,
"learning_rate": 3.1554007834213357e-06,
"loss": 0.0686,
"step": 484
},
{
"epoch": 1.9755600814663952,
"grad_norm": 0.722625027505199,
"learning_rate": 3.1333340413300263e-06,
"loss": 0.0848,
"step": 485
},
{
"epoch": 1.9796334012219958,
"grad_norm": 0.5959140201333295,
"learning_rate": 3.1113094652593023e-06,
"loss": 0.0701,
"step": 486
},
{
"epoch": 1.9837067209775967,
"grad_norm": 0.8208975505384392,
"learning_rate": 3.0893275527211742e-06,
"loss": 0.1013,
"step": 487
},
{
"epoch": 1.9877800407331976,
"grad_norm": 0.6224896038998268,
"learning_rate": 3.067388800263923e-06,
"loss": 0.0832,
"step": 488
},
{
"epoch": 1.9918533604887982,
"grad_norm": 0.6241176624569225,
"learning_rate": 3.04549370346089e-06,
"loss": 0.0678,
"step": 489
},
{
"epoch": 1.9959266802443993,
"grad_norm": 0.6061475364205144,
"learning_rate": 3.0236427568992845e-06,
"loss": 0.0768,
"step": 490
},
{
"epoch": 2.0,
"grad_norm": 0.7233218000939327,
"learning_rate": 3.0018364541690048e-06,
"loss": 0.0861,
"step": 491
},
{
"epoch": 2.0040733197556007,
"grad_norm": 0.4042479382149737,
"learning_rate": 2.9800752878514903e-06,
"loss": 0.0466,
"step": 492
},
{
"epoch": 2.0081466395112018,
"grad_norm": 0.5139407106351861,
"learning_rate": 2.958359749508603e-06,
"loss": 0.0515,
"step": 493
},
{
"epoch": 2.0122199592668024,
"grad_norm": 0.46115082125552953,
"learning_rate": 2.936690329671511e-06,
"loss": 0.0435,
"step": 494
},
{
"epoch": 2.016293279022403,
"grad_norm": 0.4268534345955416,
"learning_rate": 2.915067517829615e-06,
"loss": 0.0455,
"step": 495
},
{
"epoch": 2.020366598778004,
"grad_norm": 0.4635470622803733,
"learning_rate": 2.893491802419492e-06,
"loss": 0.0476,
"step": 496
},
{
"epoch": 2.024439918533605,
"grad_norm": 0.411133073158324,
"learning_rate": 2.871963670813861e-06,
"loss": 0.0383,
"step": 497
},
{
"epoch": 2.0285132382892055,
"grad_norm": 0.5300632118826186,
"learning_rate": 2.850483609310567e-06,
"loss": 0.0429,
"step": 498
},
{
"epoch": 2.0325865580448066,
"grad_norm": 0.5473407298179113,
"learning_rate": 2.829052103121611e-06,
"loss": 0.0461,
"step": 499
},
{
"epoch": 2.0366598778004072,
"grad_norm": 0.5221255489838956,
"learning_rate": 2.807669636362169e-06,
"loss": 0.0468,
"step": 500
},
{
"epoch": 2.0407331975560083,
"grad_norm": 1.1657623948850715,
"learning_rate": 2.7863366920396805e-06,
"loss": 0.0661,
"step": 501
},
{
"epoch": 2.044806517311609,
"grad_norm": 0.6686933034288756,
"learning_rate": 2.765053752042915e-06,
"loss": 0.0528,
"step": 502
},
{
"epoch": 2.0488798370672097,
"grad_norm": 0.570328535871806,
"learning_rate": 2.7438212971311016e-06,
"loss": 0.0445,
"step": 503
},
{
"epoch": 2.0529531568228108,
"grad_norm": 0.514036454560942,
"learning_rate": 2.722639806923066e-06,
"loss": 0.0377,
"step": 504
},
{
"epoch": 2.0570264765784114,
"grad_norm": 0.7454897039779705,
"learning_rate": 2.7015097598863906e-06,
"loss": 0.0493,
"step": 505
},
{
"epoch": 2.061099796334012,
"grad_norm": 0.6127243032686627,
"learning_rate": 2.680431633326614e-06,
"loss": 0.0397,
"step": 506
},
{
"epoch": 2.065173116089613,
"grad_norm": 0.5755519569317435,
"learning_rate": 2.659405903376442e-06,
"loss": 0.0485,
"step": 507
},
{
"epoch": 2.069246435845214,
"grad_norm": 0.5575385047600512,
"learning_rate": 2.6384330449850028e-06,
"loss": 0.0479,
"step": 508
},
{
"epoch": 2.0733197556008145,
"grad_norm": 0.5392806277095619,
"learning_rate": 2.617513531907103e-06,
"loss": 0.0466,
"step": 509
},
{
"epoch": 2.0773930753564156,
"grad_norm": 0.671579294280409,
"learning_rate": 2.5966478366925406e-06,
"loss": 0.0472,
"step": 510
},
{
"epoch": 2.0814663951120163,
"grad_norm": 0.6268395862195464,
"learning_rate": 2.5758364306754247e-06,
"loss": 0.0448,
"step": 511
},
{
"epoch": 2.085539714867617,
"grad_norm": 0.7629937873111097,
"learning_rate": 2.5550797839635283e-06,
"loss": 0.0448,
"step": 512
},
{
"epoch": 2.089613034623218,
"grad_norm": 0.5361361289284211,
"learning_rate": 2.5343783654276644e-06,
"loss": 0.0446,
"step": 513
},
{
"epoch": 2.0936863543788187,
"grad_norm": 0.5890396253795798,
"learning_rate": 2.5137326426911067e-06,
"loss": 0.0433,
"step": 514
},
{
"epoch": 2.0977596741344193,
"grad_norm": 0.6330309017862886,
"learning_rate": 2.493143082119013e-06,
"loss": 0.048,
"step": 515
},
{
"epoch": 2.1018329938900204,
"grad_norm": 0.6503134273714366,
"learning_rate": 2.472610148807903e-06,
"loss": 0.0478,
"step": 516
},
{
"epoch": 2.105906313645621,
"grad_norm": 0.6422107457849697,
"learning_rate": 2.452134306575139e-06,
"loss": 0.0464,
"step": 517
},
{
"epoch": 2.109979633401222,
"grad_norm": 0.5877663930261545,
"learning_rate": 2.431716017948462e-06,
"loss": 0.0535,
"step": 518
},
{
"epoch": 2.114052953156823,
"grad_norm": 0.5447537397982213,
"learning_rate": 2.4113557441555384e-06,
"loss": 0.0374,
"step": 519
},
{
"epoch": 2.1181262729124235,
"grad_norm": 0.5207073923132541,
"learning_rate": 2.391053945113533e-06,
"loss": 0.0344,
"step": 520
},
{
"epoch": 2.1221995926680246,
"grad_norm": 0.5610363989876352,
"learning_rate": 2.370811079418735e-06,
"loss": 0.0452,
"step": 521
},
{
"epoch": 2.1262729124236253,
"grad_norm": 0.6673691208587358,
"learning_rate": 2.350627604336186e-06,
"loss": 0.0491,
"step": 522
},
{
"epoch": 2.130346232179226,
"grad_norm": 0.6073812740579083,
"learning_rate": 2.330503975789361e-06,
"loss": 0.0522,
"step": 523
},
{
"epoch": 2.134419551934827,
"grad_norm": 0.5565972854137461,
"learning_rate": 2.3104406483498593e-06,
"loss": 0.0433,
"step": 524
},
{
"epoch": 2.1384928716904277,
"grad_norm": 0.6165041221340056,
"learning_rate": 2.290438075227146e-06,
"loss": 0.0431,
"step": 525
},
{
"epoch": 2.1425661914460283,
"grad_norm": 0.5010290577858647,
"learning_rate": 2.270496708258309e-06,
"loss": 0.045,
"step": 526
},
{
"epoch": 2.1466395112016294,
"grad_norm": 0.55185695474535,
"learning_rate": 2.2506169978978543e-06,
"loss": 0.041,
"step": 527
},
{
"epoch": 2.15071283095723,
"grad_norm": 0.610147889264061,
"learning_rate": 2.230799393207526e-06,
"loss": 0.0453,
"step": 528
},
{
"epoch": 2.1547861507128308,
"grad_norm": 0.503191638472138,
"learning_rate": 2.2110443418461723e-06,
"loss": 0.0333,
"step": 529
},
{
"epoch": 2.158859470468432,
"grad_norm": 0.5012816515718463,
"learning_rate": 2.191352290059621e-06,
"loss": 0.041,
"step": 530
},
{
"epoch": 2.1629327902240325,
"grad_norm": 0.6375299907541431,
"learning_rate": 2.171723682670613e-06,
"loss": 0.0518,
"step": 531
},
{
"epoch": 2.167006109979633,
"grad_norm": 0.6386194779538906,
"learning_rate": 2.152158963068739e-06,
"loss": 0.0413,
"step": 532
},
{
"epoch": 2.1710794297352343,
"grad_norm": 0.5082368802837912,
"learning_rate": 2.1326585732004384e-06,
"loss": 0.0407,
"step": 533
},
{
"epoch": 2.175152749490835,
"grad_norm": 0.6010063459701015,
"learning_rate": 2.1132229535590092e-06,
"loss": 0.047,
"step": 534
},
{
"epoch": 2.179226069246436,
"grad_norm": 0.5342559384940562,
"learning_rate": 2.093852543174652e-06,
"loss": 0.0442,
"step": 535
},
{
"epoch": 2.1832993890020367,
"grad_norm": 0.577105364648582,
"learning_rate": 2.0745477796045664e-06,
"loss": 0.0409,
"step": 536
},
{
"epoch": 2.1873727087576373,
"grad_norm": 0.5567485551750603,
"learning_rate": 2.0553090989230527e-06,
"loss": 0.0455,
"step": 537
},
{
"epoch": 2.1914460285132384,
"grad_norm": 0.5438386031817741,
"learning_rate": 2.036136935711674e-06,
"loss": 0.0362,
"step": 538
},
{
"epoch": 2.195519348268839,
"grad_norm": 0.5315263327187112,
"learning_rate": 2.017031723049432e-06,
"loss": 0.034,
"step": 539
},
{
"epoch": 2.1995926680244398,
"grad_norm": 0.673612549921327,
"learning_rate": 1.997993892502983e-06,
"loss": 0.0456,
"step": 540
},
{
"epoch": 2.203665987780041,
"grad_norm": 0.5852681942169125,
"learning_rate": 1.979023874116895e-06,
"loss": 0.0472,
"step": 541
},
{
"epoch": 2.2077393075356415,
"grad_norm": 0.5437231347482441,
"learning_rate": 1.9601220964039324e-06,
"loss": 0.0417,
"step": 542
},
{
"epoch": 2.211812627291242,
"grad_norm": 0.5680092663795543,
"learning_rate": 1.9412889863353683e-06,
"loss": 0.0439,
"step": 543
},
{
"epoch": 2.2158859470468433,
"grad_norm": 0.5311280129610669,
"learning_rate": 1.9225249693313547e-06,
"loss": 0.0403,
"step": 544
},
{
"epoch": 2.219959266802444,
"grad_norm": 0.5412826132559947,
"learning_rate": 1.9038304692512943e-06,
"loss": 0.0447,
"step": 545
},
{
"epoch": 2.224032586558045,
"grad_norm": 0.5452801648176083,
"learning_rate": 1.8852059083842838e-06,
"loss": 0.043,
"step": 546
},
{
"epoch": 2.2281059063136457,
"grad_norm": 0.5596216040559803,
"learning_rate": 1.8666517074395607e-06,
"loss": 0.0503,
"step": 547
},
{
"epoch": 2.2321792260692463,
"grad_norm": 0.45337689088312033,
"learning_rate": 1.8481682855370098e-06,
"loss": 0.034,
"step": 548
},
{
"epoch": 2.2362525458248474,
"grad_norm": 0.5750769024644237,
"learning_rate": 1.829756060197692e-06,
"loss": 0.0405,
"step": 549
},
{
"epoch": 2.240325865580448,
"grad_norm": 0.6084440583738112,
"learning_rate": 1.8114154473344081e-06,
"loss": 0.0473,
"step": 550
},
{
"epoch": 2.2443991853360488,
"grad_norm": 0.4803812820375662,
"learning_rate": 1.7931468612423142e-06,
"loss": 0.0378,
"step": 551
},
{
"epoch": 2.24847250509165,
"grad_norm": 0.51784716618696,
"learning_rate": 1.7749507145895518e-06,
"loss": 0.0379,
"step": 552
},
{
"epoch": 2.2525458248472505,
"grad_norm": 0.7122683699927368,
"learning_rate": 1.756827418407936e-06,
"loss": 0.0461,
"step": 553
},
{
"epoch": 2.256619144602851,
"grad_norm": 0.6243772081300375,
"learning_rate": 1.7387773820836668e-06,
"loss": 0.0405,
"step": 554
},
{
"epoch": 2.2606924643584523,
"grad_norm": 0.6228491364800817,
"learning_rate": 1.7208010133480751e-06,
"loss": 0.0398,
"step": 555
},
{
"epoch": 2.264765784114053,
"grad_norm": 0.5651946448346165,
"learning_rate": 1.7028987182684248e-06,
"loss": 0.0445,
"step": 556
},
{
"epoch": 2.2688391038696536,
"grad_norm": 0.503547616132502,
"learning_rate": 1.6850709012387328e-06,
"loss": 0.0407,
"step": 557
},
{
"epoch": 2.2729124236252547,
"grad_norm": 0.5555445714555036,
"learning_rate": 1.6673179649706312e-06,
"loss": 0.04,
"step": 558
},
{
"epoch": 2.2769857433808554,
"grad_norm": 0.5874178542752004,
"learning_rate": 1.64964031048428e-06,
"loss": 0.0399,
"step": 559
},
{
"epoch": 2.281059063136456,
"grad_norm": 0.5336372195671713,
"learning_rate": 1.632038337099297e-06,
"loss": 0.042,
"step": 560
},
{
"epoch": 2.285132382892057,
"grad_norm": 0.5383594089044018,
"learning_rate": 1.6145124424257497e-06,
"loss": 0.0396,
"step": 561
},
{
"epoch": 2.2892057026476578,
"grad_norm": 0.5590885541277442,
"learning_rate": 1.5970630223551614e-06,
"loss": 0.0411,
"step": 562
},
{
"epoch": 2.293279022403259,
"grad_norm": 0.5009898528517284,
"learning_rate": 1.5796904710515792e-06,
"loss": 0.0371,
"step": 563
},
{
"epoch": 2.2973523421588595,
"grad_norm": 0.548643352778303,
"learning_rate": 1.5623951809426663e-06,
"loss": 0.0443,
"step": 564
},
{
"epoch": 2.30142566191446,
"grad_norm": 0.6669365745322423,
"learning_rate": 1.5451775427108302e-06,
"loss": 0.0377,
"step": 565
},
{
"epoch": 2.3054989816700613,
"grad_norm": 0.5016854380779077,
"learning_rate": 1.5280379452844124e-06,
"loss": 0.0392,
"step": 566
},
{
"epoch": 2.309572301425662,
"grad_norm": 0.5047140084164453,
"learning_rate": 1.510976775828887e-06,
"loss": 0.0371,
"step": 567
},
{
"epoch": 2.3136456211812626,
"grad_norm": 0.579908909538212,
"learning_rate": 1.493994419738129e-06,
"loss": 0.0414,
"step": 568
},
{
"epoch": 2.3177189409368637,
"grad_norm": 0.5387115270421546,
"learning_rate": 1.4770912606257003e-06,
"loss": 0.0481,
"step": 569
},
{
"epoch": 2.3217922606924644,
"grad_norm": 0.7385149697772013,
"learning_rate": 1.4602676803161842e-06,
"loss": 0.045,
"step": 570
},
{
"epoch": 2.325865580448065,
"grad_norm": 0.5947860407063731,
"learning_rate": 1.4435240588365645e-06,
"loss": 0.0382,
"step": 571
},
{
"epoch": 2.329938900203666,
"grad_norm": 0.5512237269241814,
"learning_rate": 1.4268607744076419e-06,
"loss": 0.0388,
"step": 572
},
{
"epoch": 2.3340122199592668,
"grad_norm": 0.5946964883265247,
"learning_rate": 1.41027820343548e-06,
"loss": 0.0453,
"step": 573
},
{
"epoch": 2.3380855397148674,
"grad_norm": 0.6664453909137862,
"learning_rate": 1.3937767205029196e-06,
"loss": 0.0506,
"step": 574
},
{
"epoch": 2.3421588594704685,
"grad_norm": 0.5836986263456211,
"learning_rate": 1.3773566983610992e-06,
"loss": 0.048,
"step": 575
},
{
"epoch": 2.346232179226069,
"grad_norm": 0.6465467628151139,
"learning_rate": 1.3610185079210514e-06,
"loss": 0.0372,
"step": 576
},
{
"epoch": 2.35030549898167,
"grad_norm": 0.6252291597352545,
"learning_rate": 1.34476251824531e-06,
"loss": 0.0418,
"step": 577
},
{
"epoch": 2.354378818737271,
"grad_norm": 0.6247058605044977,
"learning_rate": 1.3285890965395853e-06,
"loss": 0.0448,
"step": 578
},
{
"epoch": 2.3584521384928716,
"grad_norm": 0.5412768116067983,
"learning_rate": 1.3124986081444625e-06,
"loss": 0.0459,
"step": 579
},
{
"epoch": 2.3625254582484727,
"grad_norm": 0.5791251838362595,
"learning_rate": 1.296491416527147e-06,
"loss": 0.0424,
"step": 580
},
{
"epoch": 2.3665987780040734,
"grad_norm": 0.593745342304442,
"learning_rate": 1.2805678832732627e-06,
"loss": 0.0522,
"step": 581
},
{
"epoch": 2.370672097759674,
"grad_norm": 0.8494211302879079,
"learning_rate": 1.264728368078678e-06,
"loss": 0.0403,
"step": 582
},
{
"epoch": 2.374745417515275,
"grad_norm": 0.5337776817823944,
"learning_rate": 1.248973228741378e-06,
"loss": 0.0435,
"step": 583
},
{
"epoch": 2.378818737270876,
"grad_norm": 0.585415744482421,
"learning_rate": 1.2333028211533916e-06,
"loss": 0.0399,
"step": 584
},
{
"epoch": 2.3828920570264764,
"grad_norm": 0.6610414798056414,
"learning_rate": 1.21771749929274e-06,
"loss": 0.0457,
"step": 585
},
{
"epoch": 2.3869653767820775,
"grad_norm": 0.542889748912829,
"learning_rate": 1.2022176152154525e-06,
"loss": 0.0422,
"step": 586
},
{
"epoch": 2.391038696537678,
"grad_norm": 0.5789981235953507,
"learning_rate": 1.1868035190476085e-06,
"loss": 0.0408,
"step": 587
},
{
"epoch": 2.395112016293279,
"grad_norm": 0.5350956934163298,
"learning_rate": 1.1714755589774252e-06,
"loss": 0.0374,
"step": 588
},
{
"epoch": 2.39918533604888,
"grad_norm": 0.49474503763151906,
"learning_rate": 1.1562340812474004e-06,
"loss": 0.0331,
"step": 589
},
{
"epoch": 2.4032586558044806,
"grad_norm": 0.49516175923003436,
"learning_rate": 1.1410794301464817e-06,
"loss": 0.0381,
"step": 590
},
{
"epoch": 2.4073319755600817,
"grad_norm": 0.512692253376879,
"learning_rate": 1.1260119480023008e-06,
"loss": 0.0338,
"step": 591
},
{
"epoch": 2.4114052953156824,
"grad_norm": 0.5273075654992884,
"learning_rate": 1.1110319751734271e-06,
"loss": 0.0382,
"step": 592
},
{
"epoch": 2.415478615071283,
"grad_norm": 0.5525265566643786,
"learning_rate": 1.0961398500416926e-06,
"loss": 0.0388,
"step": 593
},
{
"epoch": 2.4195519348268837,
"grad_norm": 0.5075084748842459,
"learning_rate": 1.0813359090045412e-06,
"loss": 0.039,
"step": 594
},
{
"epoch": 2.423625254582485,
"grad_norm": 0.5542848034343792,
"learning_rate": 1.0666204864674263e-06,
"loss": 0.0353,
"step": 595
},
{
"epoch": 2.4276985743380854,
"grad_norm": 0.5987619949589811,
"learning_rate": 1.0519939148362667e-06,
"loss": 0.0439,
"step": 596
},
{
"epoch": 2.4317718940936865,
"grad_norm": 0.5077117076298409,
"learning_rate": 1.0374565245099328e-06,
"loss": 0.0365,
"step": 597
},
{
"epoch": 2.435845213849287,
"grad_norm": 0.5983086948318251,
"learning_rate": 1.0230086438727771e-06,
"loss": 0.0426,
"step": 598
},
{
"epoch": 2.439918533604888,
"grad_norm": 0.531303529426792,
"learning_rate": 1.0086505992872304e-06,
"loss": 0.0425,
"step": 599
},
{
"epoch": 2.443991853360489,
"grad_norm": 0.5287789149426023,
"learning_rate": 9.943827150864143e-07,
"loss": 0.0387,
"step": 600
},
{
"epoch": 2.4480651731160896,
"grad_norm": 0.5413705197589561,
"learning_rate": 9.80205313566827e-07,
"loss": 0.0405,
"step": 601
},
{
"epoch": 2.4521384928716903,
"grad_norm": 0.5254335377173296,
"learning_rate": 9.66118714981058e-07,
"loss": 0.0446,
"step": 602
},
{
"epoch": 2.4562118126272914,
"grad_norm": 0.4973949287110427,
"learning_rate": 9.521232375305494e-07,
"loss": 0.0399,
"step": 603
},
{
"epoch": 2.460285132382892,
"grad_norm": 0.5774952931364324,
"learning_rate": 9.382191973584193e-07,
"loss": 0.03,
"step": 604
},
{
"epoch": 2.4643584521384927,
"grad_norm": 0.6662418655190413,
"learning_rate": 9.244069085423074e-07,
"loss": 0.0498,
"step": 605
},
{
"epoch": 2.468431771894094,
"grad_norm": 0.5094732488489113,
"learning_rate": 9.106866830872929e-07,
"loss": 0.0339,
"step": 606
},
{
"epoch": 2.4725050916496945,
"grad_norm": 0.5411020588710564,
"learning_rate": 8.970588309188343e-07,
"loss": 0.043,
"step": 607
},
{
"epoch": 2.4765784114052956,
"grad_norm": 0.7698681919411557,
"learning_rate": 8.835236598757796e-07,
"loss": 0.0428,
"step": 608
},
{
"epoch": 2.480651731160896,
"grad_norm": 0.46481272931573697,
"learning_rate": 8.70081475703406e-07,
"loss": 0.0347,
"step": 609
},
{
"epoch": 2.484725050916497,
"grad_norm": 0.8419803303836032,
"learning_rate": 8.567325820465156e-07,
"loss": 0.0495,
"step": 610
},
{
"epoch": 2.4887983706720975,
"grad_norm": 0.5599968776361592,
"learning_rate": 8.434772804425734e-07,
"loss": 0.0408,
"step": 611
},
{
"epoch": 2.4928716904276986,
"grad_norm": 0.5857539950345456,
"learning_rate": 8.303158703149023e-07,
"loss": 0.0411,
"step": 612
},
{
"epoch": 2.4969450101832993,
"grad_norm": 0.5261597215598826,
"learning_rate": 8.172486489659115e-07,
"loss": 0.0393,
"step": 613
},
{
"epoch": 2.5010183299389004,
"grad_norm": 0.5676439893059966,
"learning_rate": 8.042759115703891e-07,
"loss": 0.0412,
"step": 614
},
{
"epoch": 2.505091649694501,
"grad_norm": 0.6917881182955395,
"learning_rate": 7.913979511688252e-07,
"loss": 0.0409,
"step": 615
},
{
"epoch": 2.5091649694501017,
"grad_norm": 0.6445730147266382,
"learning_rate": 7.78615058660801e-07,
"loss": 0.0477,
"step": 616
},
{
"epoch": 2.513238289205703,
"grad_norm": 0.51653497568893,
"learning_rate": 7.659275227984142e-07,
"loss": 0.0379,
"step": 617
},
{
"epoch": 2.5173116089613035,
"grad_norm": 0.49200602033604846,
"learning_rate": 7.533356301797523e-07,
"loss": 0.0398,
"step": 618
},
{
"epoch": 2.521384928716904,
"grad_norm": 0.5668774352962834,
"learning_rate": 7.408396652424271e-07,
"loss": 0.0345,
"step": 619
},
{
"epoch": 2.525458248472505,
"grad_norm": 0.6916666283952536,
"learning_rate": 7.28439910257141e-07,
"loss": 0.053,
"step": 620
},
{
"epoch": 2.529531568228106,
"grad_norm": 0.5926967728678049,
"learning_rate": 7.161366453213181e-07,
"loss": 0.0382,
"step": 621
},
{
"epoch": 2.5336048879837065,
"grad_norm": 0.5473156833255498,
"learning_rate": 7.03930148352771e-07,
"loss": 0.041,
"step": 622
},
{
"epoch": 2.5376782077393076,
"grad_norm": 0.5346824810255936,
"learning_rate": 6.918206950834283e-07,
"loss": 0.0433,
"step": 623
},
{
"epoch": 2.5417515274949083,
"grad_norm": 0.5930343664045405,
"learning_rate": 6.798085590531012e-07,
"loss": 0.0403,
"step": 624
},
{
"epoch": 2.5458248472505094,
"grad_norm": 0.6064677518833592,
"learning_rate": 6.678940116033095e-07,
"loss": 0.0307,
"step": 625
},
{
"epoch": 2.54989816700611,
"grad_norm": 0.5224836625520722,
"learning_rate": 6.560773218711458e-07,
"loss": 0.0387,
"step": 626
},
{
"epoch": 2.5539714867617107,
"grad_norm": 0.59765107128643,
"learning_rate": 6.443587567832044e-07,
"loss": 0.0364,
"step": 627
},
{
"epoch": 2.5580448065173114,
"grad_norm": 0.6354038537233366,
"learning_rate": 6.327385810495423e-07,
"loss": 0.0396,
"step": 628
},
{
"epoch": 2.5621181262729125,
"grad_norm": 0.5543781150256897,
"learning_rate": 6.212170571577087e-07,
"loss": 0.0433,
"step": 629
},
{
"epoch": 2.566191446028513,
"grad_norm": 0.5478340831952899,
"learning_rate": 6.097944453668081e-07,
"loss": 0.0398,
"step": 630
},
{
"epoch": 2.5702647657841142,
"grad_norm": 0.4730479933130945,
"learning_rate": 5.984710037016267e-07,
"loss": 0.036,
"step": 631
},
{
"epoch": 2.574338085539715,
"grad_norm": 0.6518340293036718,
"learning_rate": 5.872469879468024e-07,
"loss": 0.051,
"step": 632
},
{
"epoch": 2.5784114052953155,
"grad_norm": 0.525283427975801,
"learning_rate": 5.761226516410434e-07,
"loss": 0.0336,
"step": 633
},
{
"epoch": 2.5824847250509166,
"grad_norm": 0.6611576813326997,
"learning_rate": 5.650982460714083e-07,
"loss": 0.0424,
"step": 634
},
{
"epoch": 2.5865580448065173,
"grad_norm": 0.6300996132757367,
"learning_rate": 5.54174020267621e-07,
"loss": 0.0416,
"step": 635
},
{
"epoch": 2.5906313645621184,
"grad_norm": 0.48987769259317776,
"learning_rate": 5.433502209964531e-07,
"loss": 0.0386,
"step": 636
},
{
"epoch": 2.594704684317719,
"grad_norm": 0.550269517656816,
"learning_rate": 5.326270927561444e-07,
"loss": 0.0433,
"step": 637
},
{
"epoch": 2.5987780040733197,
"grad_norm": 0.5416789018992982,
"learning_rate": 5.22004877770883e-07,
"loss": 0.0406,
"step": 638
},
{
"epoch": 2.6028513238289204,
"grad_norm": 0.5636474353072044,
"learning_rate": 5.114838159853336e-07,
"loss": 0.0413,
"step": 639
},
{
"epoch": 2.6069246435845215,
"grad_norm": 0.5254864379504877,
"learning_rate": 5.010641450592158e-07,
"loss": 0.0424,
"step": 640
},
{
"epoch": 2.610997963340122,
"grad_norm": 0.6691497595163663,
"learning_rate": 4.907461003619346e-07,
"loss": 0.0457,
"step": 641
},
{
"epoch": 2.6150712830957232,
"grad_norm": 0.6033224692953435,
"learning_rate": 4.805299149672682e-07,
"loss": 0.0409,
"step": 642
},
{
"epoch": 2.619144602851324,
"grad_norm": 0.5548154773351458,
"learning_rate": 4.7041581964809733e-07,
"loss": 0.0375,
"step": 643
},
{
"epoch": 2.6232179226069245,
"grad_norm": 0.5302610063387178,
"learning_rate": 4.6040404287119924e-07,
"loss": 0.0425,
"step": 644
},
{
"epoch": 2.627291242362525,
"grad_norm": 0.5475760159123029,
"learning_rate": 4.504948107920781e-07,
"loss": 0.037,
"step": 645
},
{
"epoch": 2.6313645621181263,
"grad_norm": 0.5384431678073536,
"learning_rate": 4.4068834724986466e-07,
"loss": 0.0404,
"step": 646
},
{
"epoch": 2.635437881873727,
"grad_norm": 0.5960889282149573,
"learning_rate": 4.309848737622568e-07,
"loss": 0.0405,
"step": 647
},
{
"epoch": 2.639511201629328,
"grad_norm": 0.5392965352716715,
"learning_rate": 4.213846095205126e-07,
"loss": 0.037,
"step": 648
},
{
"epoch": 2.6435845213849287,
"grad_norm": 0.8274526363011494,
"learning_rate": 4.1188777138450487e-07,
"loss": 0.0635,
"step": 649
},
{
"epoch": 2.6476578411405294,
"grad_norm": 0.5633117328220467,
"learning_rate": 4.024945738778163e-07,
"loss": 0.045,
"step": 650
},
{
"epoch": 2.6517311608961305,
"grad_norm": 0.5325073134146674,
"learning_rate": 3.9320522918289973e-07,
"loss": 0.0427,
"step": 651
},
{
"epoch": 2.655804480651731,
"grad_norm": 0.5506079099112636,
"learning_rate": 3.8401994713628044e-07,
"loss": 0.0444,
"step": 652
},
{
"epoch": 2.6598778004073322,
"grad_norm": 0.5182068970747546,
"learning_rate": 3.7493893522381866e-07,
"loss": 0.0368,
"step": 653
},
{
"epoch": 2.663951120162933,
"grad_norm": 0.5922018474747655,
"learning_rate": 3.6596239857602136e-07,
"loss": 0.0448,
"step": 654
},
{
"epoch": 2.6680244399185336,
"grad_norm": 0.6568887443509075,
"learning_rate": 3.570905399634111e-07,
"loss": 0.0381,
"step": 655
},
{
"epoch": 2.672097759674134,
"grad_norm": 0.5388479916650084,
"learning_rate": 3.483235597919404e-07,
"loss": 0.0377,
"step": 656
},
{
"epoch": 2.6761710794297353,
"grad_norm": 0.5081067207399704,
"learning_rate": 3.396616560984711e-07,
"loss": 0.0351,
"step": 657
},
{
"epoch": 2.680244399185336,
"grad_norm": 0.48450121917494543,
"learning_rate": 3.31105024546296e-07,
"loss": 0.036,
"step": 658
},
{
"epoch": 2.684317718940937,
"grad_norm": 0.647050948858888,
"learning_rate": 3.226538584207228e-07,
"loss": 0.0459,
"step": 659
},
{
"epoch": 2.6883910386965377,
"grad_norm": 0.6890279836440739,
"learning_rate": 3.1430834862470395e-07,
"loss": 0.0554,
"step": 660
},
{
"epoch": 2.6924643584521384,
"grad_norm": 1.8140245687384717,
"learning_rate": 3.0606868367452746e-07,
"loss": 0.0561,
"step": 661
},
{
"epoch": 2.696537678207739,
"grad_norm": 0.512535359667785,
"learning_rate": 2.9793504969555965e-07,
"loss": 0.0433,
"step": 662
},
{
"epoch": 2.70061099796334,
"grad_norm": 0.5146434179799297,
"learning_rate": 2.899076304180348e-07,
"loss": 0.0393,
"step": 663
},
{
"epoch": 2.704684317718941,
"grad_norm": 0.7613608910074268,
"learning_rate": 2.819866071729127e-07,
"loss": 0.0571,
"step": 664
},
{
"epoch": 2.708757637474542,
"grad_norm": 0.5788433559178309,
"learning_rate": 2.7417215888777493e-07,
"loss": 0.0367,
"step": 665
},
{
"epoch": 2.7128309572301426,
"grad_norm": 0.6619235135428994,
"learning_rate": 2.6646446208279054e-07,
"loss": 0.0373,
"step": 666
},
{
"epoch": 2.716904276985743,
"grad_norm": 0.5169219759499485,
"learning_rate": 2.5886369086672193e-07,
"loss": 0.0423,
"step": 667
},
{
"epoch": 2.7209775967413443,
"grad_norm": 0.5379720171061586,
"learning_rate": 2.513700169329963e-07,
"loss": 0.0365,
"step": 668
},
{
"epoch": 2.725050916496945,
"grad_norm": 0.6756425751790065,
"learning_rate": 2.439836095558262e-07,
"loss": 0.045,
"step": 669
},
{
"epoch": 2.729124236252546,
"grad_norm": 0.48551183048653324,
"learning_rate": 2.3670463558638556e-07,
"loss": 0.0369,
"step": 670
},
{
"epoch": 2.7331975560081467,
"grad_norm": 0.5897842837912828,
"learning_rate": 2.2953325944903848e-07,
"loss": 0.0368,
"step": 671
},
{
"epoch": 2.7372708757637474,
"grad_norm": 0.5831457921590835,
"learning_rate": 2.2246964313763053e-07,
"loss": 0.0406,
"step": 672
},
{
"epoch": 2.741344195519348,
"grad_norm": 0.6668669980963122,
"learning_rate": 2.1551394621182277e-07,
"loss": 0.0413,
"step": 673
},
{
"epoch": 2.745417515274949,
"grad_norm": 0.5505672609048925,
"learning_rate": 2.08666325793494e-07,
"loss": 0.0414,
"step": 674
},
{
"epoch": 2.74949083503055,
"grad_norm": 0.550600720968359,
"learning_rate": 2.0192693656318597e-07,
"loss": 0.0416,
"step": 675
},
{
"epoch": 2.753564154786151,
"grad_norm": 0.5310931391199994,
"learning_rate": 1.9529593075661267e-07,
"loss": 0.0412,
"step": 676
},
{
"epoch": 2.7576374745417516,
"grad_norm": 0.5363558016582138,
"learning_rate": 1.8877345816122162e-07,
"loss": 0.038,
"step": 677
},
{
"epoch": 2.7617107942973522,
"grad_norm": 0.5178984255141141,
"learning_rate": 1.8235966611280687e-07,
"loss": 0.039,
"step": 678
},
{
"epoch": 2.765784114052953,
"grad_norm": 0.5466402434707056,
"learning_rate": 1.760546994921858e-07,
"loss": 0.0389,
"step": 679
},
{
"epoch": 2.769857433808554,
"grad_norm": 0.6121457657463409,
"learning_rate": 1.6985870072192156e-07,
"loss": 0.0452,
"step": 680
},
{
"epoch": 2.7739307535641546,
"grad_norm": 0.5349893671952921,
"learning_rate": 1.6377180976310968e-07,
"loss": 0.0419,
"step": 681
},
{
"epoch": 2.7780040733197557,
"grad_norm": 0.5604090248089841,
"learning_rate": 1.5779416411221437e-07,
"loss": 0.0384,
"step": 682
},
{
"epoch": 2.7820773930753564,
"grad_norm": 0.564686751441713,
"learning_rate": 1.5192589879796383e-07,
"loss": 0.0356,
"step": 683
},
{
"epoch": 2.786150712830957,
"grad_norm": 0.5005018586429554,
"learning_rate": 1.4616714637829822e-07,
"loss": 0.0389,
"step": 684
},
{
"epoch": 2.790224032586558,
"grad_norm": 0.6015625122582622,
"learning_rate": 1.4051803693737876e-07,
"loss": 0.0403,
"step": 685
},
{
"epoch": 2.794297352342159,
"grad_norm": 1.2111120323286364,
"learning_rate": 1.3497869808264453e-07,
"loss": 0.0626,
"step": 686
},
{
"epoch": 2.79837067209776,
"grad_norm": 0.5396431590816398,
"learning_rate": 1.2954925494193472e-07,
"loss": 0.0395,
"step": 687
},
{
"epoch": 2.8024439918533606,
"grad_norm": 0.4817069481887003,
"learning_rate": 1.2422983016065816e-07,
"loss": 0.0371,
"step": 688
},
{
"epoch": 2.8065173116089612,
"grad_norm": 1.2000698848409261,
"learning_rate": 1.1902054389902662e-07,
"loss": 0.0606,
"step": 689
},
{
"epoch": 2.810590631364562,
"grad_norm": 0.5941877180307404,
"learning_rate": 1.1392151382933647e-07,
"loss": 0.0368,
"step": 690
},
{
"epoch": 2.814663951120163,
"grad_norm": 0.6266160796772987,
"learning_rate": 1.0893285513331353e-07,
"loss": 0.0447,
"step": 691
},
{
"epoch": 2.8187372708757636,
"grad_norm": 0.6046216067369392,
"learning_rate": 1.0405468049951184e-07,
"loss": 0.0392,
"step": 692
},
{
"epoch": 2.8228105906313647,
"grad_norm": 0.5575790775752935,
"learning_rate": 9.928710012076404e-08,
"loss": 0.0451,
"step": 693
},
{
"epoch": 2.8268839103869654,
"grad_norm": 0.5473539803468005,
"learning_rate": 9.463022169169666e-08,
"loss": 0.0399,
"step": 694
},
{
"epoch": 2.830957230142566,
"grad_norm": 0.6009599586325383,
"learning_rate": 9.008415040629548e-08,
"loss": 0.0353,
"step": 695
},
{
"epoch": 2.835030549898167,
"grad_norm": 0.6489298732282478,
"learning_rate": 8.564898895552843e-08,
"loss": 0.0464,
"step": 696
},
{
"epoch": 2.839103869653768,
"grad_norm": 0.5963225875320323,
"learning_rate": 8.132483752502806e-08,
"loss": 0.0309,
"step": 697
},
{
"epoch": 2.8431771894093685,
"grad_norm": 0.5218917037293123,
"learning_rate": 7.711179379282674e-08,
"loss": 0.0427,
"step": 698
},
{
"epoch": 2.8472505091649696,
"grad_norm": 0.6978736014217868,
"learning_rate": 7.300995292715107e-08,
"loss": 0.0503,
"step": 699
},
{
"epoch": 2.8513238289205702,
"grad_norm": 0.9455337732227677,
"learning_rate": 6.901940758427206e-08,
"loss": 0.0496,
"step": 700
},
{
"epoch": 2.855397148676171,
"grad_norm": 0.5306662874239818,
"learning_rate": 6.514024790641116e-08,
"loss": 0.0324,
"step": 701
},
{
"epoch": 2.859470468431772,
"grad_norm": 0.5226514754428617,
"learning_rate": 6.137256151970583e-08,
"loss": 0.0358,
"step": 702
},
{
"epoch": 2.8635437881873727,
"grad_norm": 0.5460041457771847,
"learning_rate": 5.771643353222778e-08,
"loss": 0.0399,
"step": 703
},
{
"epoch": 2.8676171079429738,
"grad_norm": 0.5162398669868244,
"learning_rate": 5.417194653206337e-08,
"loss": 0.0396,
"step": 704
},
{
"epoch": 2.8716904276985744,
"grad_norm": 0.5368972754214257,
"learning_rate": 5.073918058544458e-08,
"loss": 0.0406,
"step": 705
},
{
"epoch": 2.875763747454175,
"grad_norm": 0.588105419732977,
"learning_rate": 4.741821323494489e-08,
"loss": 0.0371,
"step": 706
},
{
"epoch": 2.8798370672097757,
"grad_norm": 0.4500501700097903,
"learning_rate": 4.4209119497722883e-08,
"loss": 0.0347,
"step": 707
},
{
"epoch": 2.883910386965377,
"grad_norm": 0.5696365953710792,
"learning_rate": 4.1111971863830866e-08,
"loss": 0.0373,
"step": 708
},
{
"epoch": 2.8879837067209775,
"grad_norm": 0.5792607166669516,
"learning_rate": 3.812684029457614e-08,
"loss": 0.0385,
"step": 709
},
{
"epoch": 2.8920570264765786,
"grad_norm": 0.5055106261358345,
"learning_rate": 3.525379222094061e-08,
"loss": 0.0368,
"step": 710
},
{
"epoch": 2.8961303462321792,
"grad_norm": 0.541427455199586,
"learning_rate": 3.249289254205867e-08,
"loss": 0.0387,
"step": 711
},
{
"epoch": 2.90020366598778,
"grad_norm": 0.5454327791955479,
"learning_rate": 2.984420362375007e-08,
"loss": 0.0374,
"step": 712
},
{
"epoch": 2.904276985743381,
"grad_norm": 0.5647837269627647,
"learning_rate": 2.7307785297111533e-08,
"loss": 0.0361,
"step": 713
},
{
"epoch": 2.9083503054989817,
"grad_norm": 0.4987542173953722,
"learning_rate": 2.488369485716513e-08,
"loss": 0.0396,
"step": 714
},
{
"epoch": 2.9124236252545828,
"grad_norm": 0.5027481564306886,
"learning_rate": 2.2571987061564827e-08,
"loss": 0.0349,
"step": 715
},
{
"epoch": 2.9164969450101834,
"grad_norm": 0.5256678599998655,
"learning_rate": 2.0372714129356375e-08,
"loss": 0.0405,
"step": 716
},
{
"epoch": 2.920570264765784,
"grad_norm": 0.49502069413096106,
"learning_rate": 1.8285925739803812e-08,
"loss": 0.036,
"step": 717
},
{
"epoch": 2.9246435845213847,
"grad_norm": 0.5467110241512876,
"learning_rate": 1.631166903126147e-08,
"loss": 0.0434,
"step": 718
},
{
"epoch": 2.928716904276986,
"grad_norm": 0.7022157113889199,
"learning_rate": 1.4449988600111486e-08,
"loss": 0.0495,
"step": 719
},
{
"epoch": 2.9327902240325865,
"grad_norm": 0.5499848006034249,
"learning_rate": 1.2700926499756295e-08,
"loss": 0.0405,
"step": 720
},
{
"epoch": 2.9368635437881876,
"grad_norm": 0.5059098241990838,
"learning_rate": 1.1064522239669916e-08,
"loss": 0.0339,
"step": 721
},
{
"epoch": 2.9409368635437882,
"grad_norm": 1.192165546566186,
"learning_rate": 9.54081278450314e-09,
"loss": 0.0497,
"step": 722
},
{
"epoch": 2.945010183299389,
"grad_norm": 0.5876747884158724,
"learning_rate": 8.129832553249173e-09,
"loss": 0.0393,
"step": 723
},
{
"epoch": 2.9490835030549896,
"grad_norm": 0.564386085643717,
"learning_rate": 6.831613418468163e-09,
"loss": 0.0445,
"step": 724
},
{
"epoch": 2.9531568228105907,
"grad_norm": 0.5352000628276116,
"learning_rate": 5.646184705563884e-09,
"loss": 0.0325,
"step": 725
},
{
"epoch": 2.9572301425661913,
"grad_norm": 0.5408383588017381,
"learning_rate": 4.573573192125369e-09,
"loss": 0.0356,
"step": 726
},
{
"epoch": 2.9613034623217924,
"grad_norm": 0.5942181511771125,
"learning_rate": 3.613803107317959e-09,
"loss": 0.0463,
"step": 727
},
{
"epoch": 2.965376782077393,
"grad_norm": 0.5365629409093408,
"learning_rate": 2.7668961313376263e-09,
"loss": 0.0377,
"step": 728
},
{
"epoch": 2.9694501018329937,
"grad_norm": 0.5977649244620887,
"learning_rate": 2.0328713949230304e-09,
"loss": 0.0466,
"step": 729
},
{
"epoch": 2.973523421588595,
"grad_norm": 0.5879417626438596,
"learning_rate": 1.4117454789208673e-09,
"loss": 0.0427,
"step": 730
},
{
"epoch": 2.9775967413441955,
"grad_norm": 0.5682231856966399,
"learning_rate": 9.03532413911723e-10,
"loss": 0.0424,
"step": 731
},
{
"epoch": 2.9816700610997966,
"grad_norm": 0.48337158082961007,
"learning_rate": 5.08243679894771e-10,
"loss": 0.0305,
"step": 732
},
{
"epoch": 2.9857433808553973,
"grad_norm": 0.5700756040994207,
"learning_rate": 2.2588820602631457e-10,
"loss": 0.047,
"step": 733
},
{
"epoch": 2.989816700610998,
"grad_norm": 0.6339798377809833,
"learning_rate": 5.6472370419391464e-11,
"loss": 0.045,
"step": 734
},
{
"epoch": 2.9938900203665986,
"grad_norm": 0.5348219812753423,
"learning_rate": 0.0,
"loss": 0.044,
"step": 735
},
{
"epoch": 2.9938900203665986,
"step": 735,
"total_flos": 48064094208000.0,
"train_loss": 0.08951896556025865,
"train_runtime": 6057.9419,
"train_samples_per_second": 1.945,
"train_steps_per_second": 0.121
}
],
"logging_steps": 1,
"max_steps": 735,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 250,
"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
}