PEFT
Safetensors
30b-2e / trainer_state.json
ImagineIt's picture
Upload folder using huggingface_hub
e67dacd verified
Raw
History Blame Contribute Delete
260 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.0066889632107023,
"eval_steps": 500,
"global_step": 1500,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0013377926421404682,
"grad_norm": 15.929469108581543,
"learning_rate": 5e-06,
"loss": 3.4069,
"step": 1
},
{
"epoch": 0.0026755852842809363,
"grad_norm": 15.711925506591797,
"learning_rate": 1e-05,
"loss": 3.5693,
"step": 2
},
{
"epoch": 0.004013377926421404,
"grad_norm": 15.229619979858398,
"learning_rate": 1.5e-05,
"loss": 3.3951,
"step": 3
},
{
"epoch": 0.005351170568561873,
"grad_norm": 9.533045768737793,
"learning_rate": 2e-05,
"loss": 3.2976,
"step": 4
},
{
"epoch": 0.006688963210702341,
"grad_norm": 5.610800266265869,
"learning_rate": 2.5e-05,
"loss": 2.9262,
"step": 5
},
{
"epoch": 0.008026755852842809,
"grad_norm": 3.5882365703582764,
"learning_rate": 3e-05,
"loss": 2.4932,
"step": 6
},
{
"epoch": 0.009364548494983277,
"grad_norm": 3.190317392349243,
"learning_rate": 3.5e-05,
"loss": 2.2106,
"step": 7
},
{
"epoch": 0.010702341137123745,
"grad_norm": 3.090921401977539,
"learning_rate": 4e-05,
"loss": 2.3511,
"step": 8
},
{
"epoch": 0.012040133779264214,
"grad_norm": 1.6608407497406006,
"learning_rate": 4.5e-05,
"loss": 2.0331,
"step": 9
},
{
"epoch": 0.013377926421404682,
"grad_norm": 2.3017241954803467,
"learning_rate": 5e-05,
"loss": 2.4304,
"step": 10
},
{
"epoch": 0.01471571906354515,
"grad_norm": 1.7485525608062744,
"learning_rate": 5.500000000000001e-05,
"loss": 2.4206,
"step": 11
},
{
"epoch": 0.016053511705685617,
"grad_norm": 1.4397761821746826,
"learning_rate": 6e-05,
"loss": 1.9572,
"step": 12
},
{
"epoch": 0.017391304347826087,
"grad_norm": 2.1730759143829346,
"learning_rate": 6.500000000000001e-05,
"loss": 2.0966,
"step": 13
},
{
"epoch": 0.018729096989966554,
"grad_norm": 1.3649462461471558,
"learning_rate": 7e-05,
"loss": 1.8041,
"step": 14
},
{
"epoch": 0.020066889632107024,
"grad_norm": 2.1070213317871094,
"learning_rate": 7.500000000000001e-05,
"loss": 2.1222,
"step": 15
},
{
"epoch": 0.02140468227424749,
"grad_norm": 0.7499298453330994,
"learning_rate": 8e-05,
"loss": 1.9738,
"step": 16
},
{
"epoch": 0.02274247491638796,
"grad_norm": 0.8609057068824768,
"learning_rate": 8.5e-05,
"loss": 1.8804,
"step": 17
},
{
"epoch": 0.024080267558528427,
"grad_norm": 0.9405566453933716,
"learning_rate": 9e-05,
"loss": 1.8886,
"step": 18
},
{
"epoch": 0.025418060200668897,
"grad_norm": 0.6386722922325134,
"learning_rate": 9.5e-05,
"loss": 1.9253,
"step": 19
},
{
"epoch": 0.026755852842809364,
"grad_norm": 0.6493627429008484,
"learning_rate": 0.0001,
"loss": 1.8284,
"step": 20
},
{
"epoch": 0.028093645484949834,
"grad_norm": 0.81246018409729,
"learning_rate": 9.995497523638001e-05,
"loss": 1.6218,
"step": 21
},
{
"epoch": 0.0294314381270903,
"grad_norm": 0.7177069187164307,
"learning_rate": 9.990995047276002e-05,
"loss": 1.8196,
"step": 22
},
{
"epoch": 0.03076923076923077,
"grad_norm": 0.8126530647277832,
"learning_rate": 9.986492570914003e-05,
"loss": 2.1567,
"step": 23
},
{
"epoch": 0.032107023411371234,
"grad_norm": 0.5535350441932678,
"learning_rate": 9.981990094552004e-05,
"loss": 2.0801,
"step": 24
},
{
"epoch": 0.033444816053511704,
"grad_norm": 0.6111394762992859,
"learning_rate": 9.977487618190005e-05,
"loss": 1.8416,
"step": 25
},
{
"epoch": 0.034782608695652174,
"grad_norm": 3.8264172077178955,
"learning_rate": 9.972985141828006e-05,
"loss": 1.5934,
"step": 26
},
{
"epoch": 0.036120401337792644,
"grad_norm": 0.5423453450202942,
"learning_rate": 9.968482665466006e-05,
"loss": 1.7254,
"step": 27
},
{
"epoch": 0.03745819397993311,
"grad_norm": 0.5544487237930298,
"learning_rate": 9.963980189104007e-05,
"loss": 1.9727,
"step": 28
},
{
"epoch": 0.03879598662207358,
"grad_norm": 0.4474441409111023,
"learning_rate": 9.95947771274201e-05,
"loss": 2.0919,
"step": 29
},
{
"epoch": 0.04013377926421405,
"grad_norm": 0.4968682527542114,
"learning_rate": 9.954975236380009e-05,
"loss": 1.8365,
"step": 30
},
{
"epoch": 0.04147157190635452,
"grad_norm": 0.8396462798118591,
"learning_rate": 9.95047276001801e-05,
"loss": 1.5896,
"step": 31
},
{
"epoch": 0.04280936454849498,
"grad_norm": 0.45893028378486633,
"learning_rate": 9.945970283656011e-05,
"loss": 1.8436,
"step": 32
},
{
"epoch": 0.04414715719063545,
"grad_norm": 0.5614059567451477,
"learning_rate": 9.941467807294013e-05,
"loss": 1.8984,
"step": 33
},
{
"epoch": 0.04548494983277592,
"grad_norm": 0.4408206343650818,
"learning_rate": 9.936965330932014e-05,
"loss": 2.1722,
"step": 34
},
{
"epoch": 0.046822742474916385,
"grad_norm": 0.4931108057498932,
"learning_rate": 9.932462854570013e-05,
"loss": 2.0494,
"step": 35
},
{
"epoch": 0.048160535117056855,
"grad_norm": 1.6200001239776611,
"learning_rate": 9.927960378208014e-05,
"loss": 1.8094,
"step": 36
},
{
"epoch": 0.049498327759197325,
"grad_norm": 4.759247303009033,
"learning_rate": 9.923457901846016e-05,
"loss": 1.7833,
"step": 37
},
{
"epoch": 0.050836120401337795,
"grad_norm": 0.5733784437179565,
"learning_rate": 9.918955425484017e-05,
"loss": 1.7082,
"step": 38
},
{
"epoch": 0.05217391304347826,
"grad_norm": 0.6474048495292664,
"learning_rate": 9.914452949122017e-05,
"loss": 1.6784,
"step": 39
},
{
"epoch": 0.05351170568561873,
"grad_norm": 0.47228601574897766,
"learning_rate": 9.909950472760019e-05,
"loss": 1.8055,
"step": 40
},
{
"epoch": 0.0548494983277592,
"grad_norm": 1.4710334539413452,
"learning_rate": 9.90544799639802e-05,
"loss": 1.8068,
"step": 41
},
{
"epoch": 0.05618729096989967,
"grad_norm": 0.467542439699173,
"learning_rate": 9.900945520036021e-05,
"loss": 1.6524,
"step": 42
},
{
"epoch": 0.05752508361204013,
"grad_norm": 0.4020284116268158,
"learning_rate": 9.89644304367402e-05,
"loss": 2.1742,
"step": 43
},
{
"epoch": 0.0588628762541806,
"grad_norm": 0.7365009188652039,
"learning_rate": 9.891940567312022e-05,
"loss": 1.6893,
"step": 44
},
{
"epoch": 0.06020066889632107,
"grad_norm": 0.46295851469039917,
"learning_rate": 9.887438090950023e-05,
"loss": 1.9585,
"step": 45
},
{
"epoch": 0.06153846153846154,
"grad_norm": 0.5056664943695068,
"learning_rate": 9.882935614588024e-05,
"loss": 1.9968,
"step": 46
},
{
"epoch": 0.06287625418060201,
"grad_norm": 0.3968791961669922,
"learning_rate": 9.878433138226025e-05,
"loss": 2.0845,
"step": 47
},
{
"epoch": 0.06421404682274247,
"grad_norm": 0.4480941891670227,
"learning_rate": 9.873930661864026e-05,
"loss": 1.7947,
"step": 48
},
{
"epoch": 0.06555183946488294,
"grad_norm": 0.4318523108959198,
"learning_rate": 9.869428185502027e-05,
"loss": 2.093,
"step": 49
},
{
"epoch": 0.06688963210702341,
"grad_norm": 0.43793708086013794,
"learning_rate": 9.864925709140028e-05,
"loss": 1.7113,
"step": 50
},
{
"epoch": 0.06822742474916388,
"grad_norm": 0.8287019729614258,
"learning_rate": 9.860423232778028e-05,
"loss": 1.6314,
"step": 51
},
{
"epoch": 0.06956521739130435,
"grad_norm": 0.36626288294792175,
"learning_rate": 9.855920756416029e-05,
"loss": 1.9596,
"step": 52
},
{
"epoch": 0.07090301003344482,
"grad_norm": 0.39505261182785034,
"learning_rate": 9.85141828005403e-05,
"loss": 1.9304,
"step": 53
},
{
"epoch": 0.07224080267558529,
"grad_norm": 0.3602500557899475,
"learning_rate": 9.846915803692031e-05,
"loss": 1.9283,
"step": 54
},
{
"epoch": 0.07357859531772576,
"grad_norm": 0.4631737470626831,
"learning_rate": 9.842413327330032e-05,
"loss": 1.88,
"step": 55
},
{
"epoch": 0.07491638795986622,
"grad_norm": 0.8588123321533203,
"learning_rate": 9.837910850968033e-05,
"loss": 1.9562,
"step": 56
},
{
"epoch": 0.07625418060200669,
"grad_norm": 0.4845622181892395,
"learning_rate": 9.833408374606034e-05,
"loss": 1.8439,
"step": 57
},
{
"epoch": 0.07759197324414716,
"grad_norm": 0.3816477060317993,
"learning_rate": 9.828905898244036e-05,
"loss": 1.9877,
"step": 58
},
{
"epoch": 0.07892976588628763,
"grad_norm": 0.4842838943004608,
"learning_rate": 9.824403421882035e-05,
"loss": 1.3894,
"step": 59
},
{
"epoch": 0.0802675585284281,
"grad_norm": 0.5041085481643677,
"learning_rate": 9.819900945520036e-05,
"loss": 2.2597,
"step": 60
},
{
"epoch": 0.08160535117056857,
"grad_norm": 0.30196964740753174,
"learning_rate": 9.815398469158037e-05,
"loss": 1.7422,
"step": 61
},
{
"epoch": 0.08294314381270904,
"grad_norm": 0.3853469491004944,
"learning_rate": 9.810895992796039e-05,
"loss": 1.9314,
"step": 62
},
{
"epoch": 0.08428093645484949,
"grad_norm": 1.472838044166565,
"learning_rate": 9.806393516434039e-05,
"loss": 1.8565,
"step": 63
},
{
"epoch": 0.08561872909698996,
"grad_norm": 0.3896331787109375,
"learning_rate": 9.80189104007204e-05,
"loss": 1.7477,
"step": 64
},
{
"epoch": 0.08695652173913043,
"grad_norm": 0.3785063922405243,
"learning_rate": 9.797388563710042e-05,
"loss": 2.1115,
"step": 65
},
{
"epoch": 0.0882943143812709,
"grad_norm": 0.3835844397544861,
"learning_rate": 9.792886087348043e-05,
"loss": 1.8925,
"step": 66
},
{
"epoch": 0.08963210702341137,
"grad_norm": 0.32654905319213867,
"learning_rate": 9.788383610986042e-05,
"loss": 1.9158,
"step": 67
},
{
"epoch": 0.09096989966555184,
"grad_norm": 0.38295117020606995,
"learning_rate": 9.783881134624043e-05,
"loss": 1.7556,
"step": 68
},
{
"epoch": 0.09230769230769231,
"grad_norm": 0.4461645781993866,
"learning_rate": 9.779378658262045e-05,
"loss": 1.4386,
"step": 69
},
{
"epoch": 0.09364548494983277,
"grad_norm": 0.7319120764732361,
"learning_rate": 9.774876181900046e-05,
"loss": 2.0061,
"step": 70
},
{
"epoch": 0.09498327759197324,
"grad_norm": 0.4447004199028015,
"learning_rate": 9.770373705538046e-05,
"loss": 1.8502,
"step": 71
},
{
"epoch": 0.09632107023411371,
"grad_norm": 0.3706307113170624,
"learning_rate": 9.765871229176046e-05,
"loss": 1.9704,
"step": 72
},
{
"epoch": 0.09765886287625418,
"grad_norm": 2.6014416217803955,
"learning_rate": 9.761368752814049e-05,
"loss": 1.673,
"step": 73
},
{
"epoch": 0.09899665551839465,
"grad_norm": 0.36813560128211975,
"learning_rate": 9.75686627645205e-05,
"loss": 1.9571,
"step": 74
},
{
"epoch": 0.10033444816053512,
"grad_norm": 0.2965801954269409,
"learning_rate": 9.752363800090049e-05,
"loss": 1.8679,
"step": 75
},
{
"epoch": 0.10167224080267559,
"grad_norm": 0.36737844347953796,
"learning_rate": 9.747861323728051e-05,
"loss": 1.914,
"step": 76
},
{
"epoch": 0.10301003344481606,
"grad_norm": 0.5539958477020264,
"learning_rate": 9.743358847366052e-05,
"loss": 1.7668,
"step": 77
},
{
"epoch": 0.10434782608695652,
"grad_norm": 0.9024600982666016,
"learning_rate": 9.738856371004053e-05,
"loss": 1.602,
"step": 78
},
{
"epoch": 0.10568561872909699,
"grad_norm": 0.47630879282951355,
"learning_rate": 9.734353894642053e-05,
"loss": 1.9614,
"step": 79
},
{
"epoch": 0.10702341137123746,
"grad_norm": 0.33962881565093994,
"learning_rate": 9.729851418280055e-05,
"loss": 2.0143,
"step": 80
},
{
"epoch": 0.10836120401337793,
"grad_norm": 0.34463438391685486,
"learning_rate": 9.725348941918056e-05,
"loss": 1.8305,
"step": 81
},
{
"epoch": 0.1096989966555184,
"grad_norm": 0.5177729725837708,
"learning_rate": 9.720846465556056e-05,
"loss": 1.5823,
"step": 82
},
{
"epoch": 0.11103678929765887,
"grad_norm": 0.3883843421936035,
"learning_rate": 9.716343989194057e-05,
"loss": 1.9247,
"step": 83
},
{
"epoch": 0.11237458193979934,
"grad_norm": 0.3565119206905365,
"learning_rate": 9.711841512832058e-05,
"loss": 1.8981,
"step": 84
},
{
"epoch": 0.11371237458193979,
"grad_norm": 0.9169286489486694,
"learning_rate": 9.707339036470059e-05,
"loss": 1.6898,
"step": 85
},
{
"epoch": 0.11505016722408026,
"grad_norm": 2.372044801712036,
"learning_rate": 9.70283656010806e-05,
"loss": 2.0062,
"step": 86
},
{
"epoch": 0.11638795986622073,
"grad_norm": 0.4511963725090027,
"learning_rate": 9.698334083746061e-05,
"loss": 1.9983,
"step": 87
},
{
"epoch": 0.1177257525083612,
"grad_norm": 0.2942259907722473,
"learning_rate": 9.693831607384062e-05,
"loss": 1.9471,
"step": 88
},
{
"epoch": 0.11906354515050167,
"grad_norm": 0.3808842897415161,
"learning_rate": 9.689329131022062e-05,
"loss": 1.9287,
"step": 89
},
{
"epoch": 0.12040133779264214,
"grad_norm": 0.3474765121936798,
"learning_rate": 9.684826654660063e-05,
"loss": 2.166,
"step": 90
},
{
"epoch": 0.12173913043478261,
"grad_norm": 0.4045035243034363,
"learning_rate": 9.680324178298064e-05,
"loss": 1.9044,
"step": 91
},
{
"epoch": 0.12307692307692308,
"grad_norm": 0.3970213532447815,
"learning_rate": 9.675821701936065e-05,
"loss": 2.023,
"step": 92
},
{
"epoch": 0.12441471571906354,
"grad_norm": 0.3404536843299866,
"learning_rate": 9.671319225574066e-05,
"loss": 1.7032,
"step": 93
},
{
"epoch": 0.12575250836120402,
"grad_norm": 0.441057413816452,
"learning_rate": 9.666816749212068e-05,
"loss": 1.5145,
"step": 94
},
{
"epoch": 0.12709030100334448,
"grad_norm": 0.6671501398086548,
"learning_rate": 9.662314272850068e-05,
"loss": 1.7196,
"step": 95
},
{
"epoch": 0.12842809364548494,
"grad_norm": 0.37828826904296875,
"learning_rate": 9.657811796488068e-05,
"loss": 1.3918,
"step": 96
},
{
"epoch": 0.12976588628762542,
"grad_norm": 0.4070121645927429,
"learning_rate": 9.653309320126069e-05,
"loss": 1.7636,
"step": 97
},
{
"epoch": 0.13110367892976588,
"grad_norm": 0.349697470664978,
"learning_rate": 9.648806843764072e-05,
"loss": 2.383,
"step": 98
},
{
"epoch": 0.13244147157190636,
"grad_norm": 0.3957515358924866,
"learning_rate": 9.644304367402071e-05,
"loss": 2.0249,
"step": 99
},
{
"epoch": 0.13377926421404682,
"grad_norm": 0.6649832725524902,
"learning_rate": 9.639801891040072e-05,
"loss": 1.7339,
"step": 100
},
{
"epoch": 0.1351170568561873,
"grad_norm": 0.6055648922920227,
"learning_rate": 9.635299414678074e-05,
"loss": 1.5403,
"step": 101
},
{
"epoch": 0.13645484949832776,
"grad_norm": 0.4690856635570526,
"learning_rate": 9.630796938316075e-05,
"loss": 1.7279,
"step": 102
},
{
"epoch": 0.13779264214046824,
"grad_norm": 0.4584232568740845,
"learning_rate": 9.626294461954074e-05,
"loss": 1.8899,
"step": 103
},
{
"epoch": 0.1391304347826087,
"grad_norm": 0.3669104278087616,
"learning_rate": 9.621791985592075e-05,
"loss": 2.2324,
"step": 104
},
{
"epoch": 0.14046822742474915,
"grad_norm": 0.506993293762207,
"learning_rate": 9.617289509230078e-05,
"loss": 1.7735,
"step": 105
},
{
"epoch": 0.14180602006688964,
"grad_norm": 0.405987948179245,
"learning_rate": 9.612787032868078e-05,
"loss": 1.9026,
"step": 106
},
{
"epoch": 0.1431438127090301,
"grad_norm": 0.8430399298667908,
"learning_rate": 9.608284556506079e-05,
"loss": 1.4946,
"step": 107
},
{
"epoch": 0.14448160535117058,
"grad_norm": 0.38154295086860657,
"learning_rate": 9.603782080144079e-05,
"loss": 2.0693,
"step": 108
},
{
"epoch": 0.14581939799331103,
"grad_norm": 0.43963882327079773,
"learning_rate": 9.599279603782081e-05,
"loss": 1.6948,
"step": 109
},
{
"epoch": 0.14715719063545152,
"grad_norm": 0.46459442377090454,
"learning_rate": 9.594777127420082e-05,
"loss": 1.8412,
"step": 110
},
{
"epoch": 0.14849498327759197,
"grad_norm": 0.409085214138031,
"learning_rate": 9.590274651058083e-05,
"loss": 1.7562,
"step": 111
},
{
"epoch": 0.14983277591973243,
"grad_norm": 0.35461780428886414,
"learning_rate": 9.585772174696084e-05,
"loss": 1.9862,
"step": 112
},
{
"epoch": 0.15117056856187291,
"grad_norm": 0.3632294237613678,
"learning_rate": 9.581269698334084e-05,
"loss": 2.0089,
"step": 113
},
{
"epoch": 0.15250836120401337,
"grad_norm": 0.6132340431213379,
"learning_rate": 9.576767221972085e-05,
"loss": 2.116,
"step": 114
},
{
"epoch": 0.15384615384615385,
"grad_norm": 0.4431364834308624,
"learning_rate": 9.572264745610086e-05,
"loss": 1.76,
"step": 115
},
{
"epoch": 0.1551839464882943,
"grad_norm": 0.3065609037876129,
"learning_rate": 9.567762269248087e-05,
"loss": 1.7882,
"step": 116
},
{
"epoch": 0.1565217391304348,
"grad_norm": 0.6607733964920044,
"learning_rate": 9.563259792886088e-05,
"loss": 1.8994,
"step": 117
},
{
"epoch": 0.15785953177257525,
"grad_norm": 0.5508191585540771,
"learning_rate": 9.558757316524089e-05,
"loss": 2.1155,
"step": 118
},
{
"epoch": 0.1591973244147157,
"grad_norm": 0.3941163420677185,
"learning_rate": 9.55425484016209e-05,
"loss": 2.2935,
"step": 119
},
{
"epoch": 0.1605351170568562,
"grad_norm": 0.6320101618766785,
"learning_rate": 9.54975236380009e-05,
"loss": 1.801,
"step": 120
},
{
"epoch": 0.16187290969899665,
"grad_norm": 0.2693134844303131,
"learning_rate": 9.545249887438091e-05,
"loss": 1.9149,
"step": 121
},
{
"epoch": 0.16321070234113713,
"grad_norm": 0.36482131481170654,
"learning_rate": 9.540747411076092e-05,
"loss": 1.7871,
"step": 122
},
{
"epoch": 0.1645484949832776,
"grad_norm": 0.3653654456138611,
"learning_rate": 9.536244934714093e-05,
"loss": 1.5542,
"step": 123
},
{
"epoch": 0.16588628762541807,
"grad_norm": 0.3543652296066284,
"learning_rate": 9.531742458352094e-05,
"loss": 1.766,
"step": 124
},
{
"epoch": 0.16722408026755853,
"grad_norm": 0.5176847577095032,
"learning_rate": 9.527239981990095e-05,
"loss": 1.8646,
"step": 125
},
{
"epoch": 0.16856187290969898,
"grad_norm": 0.3467628061771393,
"learning_rate": 9.522737505628096e-05,
"loss": 2.217,
"step": 126
},
{
"epoch": 0.16989966555183947,
"grad_norm": 0.3651058077812195,
"learning_rate": 9.518235029266098e-05,
"loss": 1.7298,
"step": 127
},
{
"epoch": 0.17123745819397992,
"grad_norm": 0.3439328968524933,
"learning_rate": 9.513732552904097e-05,
"loss": 2.0967,
"step": 128
},
{
"epoch": 0.1725752508361204,
"grad_norm": 0.367436945438385,
"learning_rate": 9.509230076542098e-05,
"loss": 2.0584,
"step": 129
},
{
"epoch": 0.17391304347826086,
"grad_norm": 0.4682665467262268,
"learning_rate": 9.5047276001801e-05,
"loss": 1.2279,
"step": 130
},
{
"epoch": 0.17525083612040135,
"grad_norm": 0.6343939900398254,
"learning_rate": 9.500225123818101e-05,
"loss": 1.8301,
"step": 131
},
{
"epoch": 0.1765886287625418,
"grad_norm": 0.37340253591537476,
"learning_rate": 9.495722647456101e-05,
"loss": 1.6715,
"step": 132
},
{
"epoch": 0.17792642140468226,
"grad_norm": 0.4212535619735718,
"learning_rate": 9.491220171094102e-05,
"loss": 1.3813,
"step": 133
},
{
"epoch": 0.17926421404682275,
"grad_norm": 0.7796082496643066,
"learning_rate": 9.486717694732104e-05,
"loss": 2.0301,
"step": 134
},
{
"epoch": 0.1806020066889632,
"grad_norm": 0.3705946207046509,
"learning_rate": 9.482215218370105e-05,
"loss": 1.9791,
"step": 135
},
{
"epoch": 0.18193979933110369,
"grad_norm": 1.0462645292282104,
"learning_rate": 9.477712742008104e-05,
"loss": 1.8377,
"step": 136
},
{
"epoch": 0.18327759197324414,
"grad_norm": 0.8452597856521606,
"learning_rate": 9.473210265646106e-05,
"loss": 1.587,
"step": 137
},
{
"epoch": 0.18461538461538463,
"grad_norm": 0.34128281474113464,
"learning_rate": 9.468707789284107e-05,
"loss": 1.7248,
"step": 138
},
{
"epoch": 0.18595317725752508,
"grad_norm": 0.3522973656654358,
"learning_rate": 9.464205312922108e-05,
"loss": 1.4179,
"step": 139
},
{
"epoch": 0.18729096989966554,
"grad_norm": 0.3930247128009796,
"learning_rate": 9.459702836560108e-05,
"loss": 1.8452,
"step": 140
},
{
"epoch": 0.18862876254180602,
"grad_norm": 0.4119817018508911,
"learning_rate": 9.45520036019811e-05,
"loss": 2.0599,
"step": 141
},
{
"epoch": 0.18996655518394648,
"grad_norm": 0.31673747301101685,
"learning_rate": 9.45069788383611e-05,
"loss": 1.9548,
"step": 142
},
{
"epoch": 0.19130434782608696,
"grad_norm": 0.6387179493904114,
"learning_rate": 9.446195407474112e-05,
"loss": 1.8265,
"step": 143
},
{
"epoch": 0.19264214046822742,
"grad_norm": 0.36570000648498535,
"learning_rate": 9.441692931112111e-05,
"loss": 1.9025,
"step": 144
},
{
"epoch": 0.1939799331103679,
"grad_norm": 0.8736979365348816,
"learning_rate": 9.437190454750113e-05,
"loss": 1.7701,
"step": 145
},
{
"epoch": 0.19531772575250836,
"grad_norm": 0.35222160816192627,
"learning_rate": 9.432687978388114e-05,
"loss": 2.1553,
"step": 146
},
{
"epoch": 0.19665551839464884,
"grad_norm": 0.33907803893089294,
"learning_rate": 9.428185502026115e-05,
"loss": 1.9887,
"step": 147
},
{
"epoch": 0.1979933110367893,
"grad_norm": 0.5510755181312561,
"learning_rate": 9.423683025664116e-05,
"loss": 2.0307,
"step": 148
},
{
"epoch": 0.19933110367892976,
"grad_norm": 0.3364030122756958,
"learning_rate": 9.419180549302117e-05,
"loss": 1.7475,
"step": 149
},
{
"epoch": 0.20066889632107024,
"grad_norm": 0.3714922070503235,
"learning_rate": 9.414678072940118e-05,
"loss": 1.8964,
"step": 150
},
{
"epoch": 0.2020066889632107,
"grad_norm": 0.3768271803855896,
"learning_rate": 9.410175596578118e-05,
"loss": 1.8011,
"step": 151
},
{
"epoch": 0.20334448160535118,
"grad_norm": 0.3435843586921692,
"learning_rate": 9.405673120216119e-05,
"loss": 2.368,
"step": 152
},
{
"epoch": 0.20468227424749164,
"grad_norm": 0.3646228313446045,
"learning_rate": 9.40117064385412e-05,
"loss": 1.6072,
"step": 153
},
{
"epoch": 0.20602006688963212,
"grad_norm": 0.35950562357902527,
"learning_rate": 9.396668167492121e-05,
"loss": 1.9689,
"step": 154
},
{
"epoch": 0.20735785953177258,
"grad_norm": 0.40717366337776184,
"learning_rate": 9.392165691130123e-05,
"loss": 1.6078,
"step": 155
},
{
"epoch": 0.20869565217391303,
"grad_norm": 0.6353041529655457,
"learning_rate": 9.387663214768123e-05,
"loss": 2.2688,
"step": 156
},
{
"epoch": 0.21003344481605352,
"grad_norm": 0.4215109050273895,
"learning_rate": 9.383160738406124e-05,
"loss": 1.9392,
"step": 157
},
{
"epoch": 0.21137123745819397,
"grad_norm": 0.37301746010780334,
"learning_rate": 9.378658262044124e-05,
"loss": 1.9112,
"step": 158
},
{
"epoch": 0.21270903010033446,
"grad_norm": 0.36573272943496704,
"learning_rate": 9.374155785682127e-05,
"loss": 1.6324,
"step": 159
},
{
"epoch": 0.2140468227424749,
"grad_norm": 0.5235257148742676,
"learning_rate": 9.369653309320126e-05,
"loss": 1.9713,
"step": 160
},
{
"epoch": 0.2153846153846154,
"grad_norm": 0.34718433022499084,
"learning_rate": 9.365150832958127e-05,
"loss": 2.0838,
"step": 161
},
{
"epoch": 0.21672240802675585,
"grad_norm": 0.39332470297813416,
"learning_rate": 9.360648356596128e-05,
"loss": 2.0534,
"step": 162
},
{
"epoch": 0.2180602006688963,
"grad_norm": 0.7983328104019165,
"learning_rate": 9.35614588023413e-05,
"loss": 1.7208,
"step": 163
},
{
"epoch": 0.2193979933110368,
"grad_norm": 0.4300435185432434,
"learning_rate": 9.35164340387213e-05,
"loss": 1.39,
"step": 164
},
{
"epoch": 0.22073578595317725,
"grad_norm": 0.377658873796463,
"learning_rate": 9.34714092751013e-05,
"loss": 1.5722,
"step": 165
},
{
"epoch": 0.22207357859531773,
"grad_norm": 0.4393126964569092,
"learning_rate": 9.342638451148133e-05,
"loss": 1.5827,
"step": 166
},
{
"epoch": 0.2234113712374582,
"grad_norm": 0.35238978266716003,
"learning_rate": 9.338135974786133e-05,
"loss": 1.6863,
"step": 167
},
{
"epoch": 0.22474916387959867,
"grad_norm": 0.3524191379547119,
"learning_rate": 9.333633498424133e-05,
"loss": 1.8321,
"step": 168
},
{
"epoch": 0.22608695652173913,
"grad_norm": 0.6319050788879395,
"learning_rate": 9.329131022062134e-05,
"loss": 1.7113,
"step": 169
},
{
"epoch": 0.22742474916387959,
"grad_norm": 0.3477349281311035,
"learning_rate": 9.324628545700136e-05,
"loss": 1.8297,
"step": 170
},
{
"epoch": 0.22876254180602007,
"grad_norm": 1.4175788164138794,
"learning_rate": 9.320126069338137e-05,
"loss": 1.7897,
"step": 171
},
{
"epoch": 0.23010033444816053,
"grad_norm": 0.3294921815395355,
"learning_rate": 9.315623592976136e-05,
"loss": 1.9491,
"step": 172
},
{
"epoch": 0.231438127090301,
"grad_norm": 0.35877764225006104,
"learning_rate": 9.311121116614139e-05,
"loss": 2.0432,
"step": 173
},
{
"epoch": 0.23277591973244147,
"grad_norm": 0.3810911476612091,
"learning_rate": 9.30661864025214e-05,
"loss": 1.7374,
"step": 174
},
{
"epoch": 0.23411371237458195,
"grad_norm": 0.39322909712791443,
"learning_rate": 9.30211616389014e-05,
"loss": 1.3096,
"step": 175
},
{
"epoch": 0.2354515050167224,
"grad_norm": 0.3306376039981842,
"learning_rate": 9.29761368752814e-05,
"loss": 2.0244,
"step": 176
},
{
"epoch": 0.23678929765886286,
"grad_norm": 0.35327744483947754,
"learning_rate": 9.293111211166142e-05,
"loss": 1.8494,
"step": 177
},
{
"epoch": 0.23812709030100335,
"grad_norm": 0.3139340579509735,
"learning_rate": 9.288608734804143e-05,
"loss": 2.0111,
"step": 178
},
{
"epoch": 0.2394648829431438,
"grad_norm": 0.29642781615257263,
"learning_rate": 9.284106258442144e-05,
"loss": 1.8517,
"step": 179
},
{
"epoch": 0.2408026755852843,
"grad_norm": 0.6292489171028137,
"learning_rate": 9.279603782080145e-05,
"loss": 1.7926,
"step": 180
},
{
"epoch": 0.24214046822742474,
"grad_norm": 0.5784863233566284,
"learning_rate": 9.275101305718146e-05,
"loss": 1.9574,
"step": 181
},
{
"epoch": 0.24347826086956523,
"grad_norm": 0.5684370398521423,
"learning_rate": 9.270598829356146e-05,
"loss": 1.7677,
"step": 182
},
{
"epoch": 0.24481605351170568,
"grad_norm": 0.435253769159317,
"learning_rate": 9.266096352994147e-05,
"loss": 2.0751,
"step": 183
},
{
"epoch": 0.24615384615384617,
"grad_norm": 0.394359827041626,
"learning_rate": 9.261593876632148e-05,
"loss": 2.0506,
"step": 184
},
{
"epoch": 0.24749163879598662,
"grad_norm": 0.5143195390701294,
"learning_rate": 9.257091400270149e-05,
"loss": 2.0168,
"step": 185
},
{
"epoch": 0.24882943143812708,
"grad_norm": 0.368755966424942,
"learning_rate": 9.25258892390815e-05,
"loss": 1.609,
"step": 186
},
{
"epoch": 0.25016722408026754,
"grad_norm": 3.3827755451202393,
"learning_rate": 9.24808644754615e-05,
"loss": 1.5342,
"step": 187
},
{
"epoch": 0.25150501672240805,
"grad_norm": 0.3961702287197113,
"learning_rate": 9.243583971184152e-05,
"loss": 2.0325,
"step": 188
},
{
"epoch": 0.2528428093645485,
"grad_norm": 0.4427662193775177,
"learning_rate": 9.239081494822152e-05,
"loss": 1.8261,
"step": 189
},
{
"epoch": 0.25418060200668896,
"grad_norm": 0.7670131325721741,
"learning_rate": 9.234579018460153e-05,
"loss": 2.0239,
"step": 190
},
{
"epoch": 0.2555183946488294,
"grad_norm": 0.33620205521583557,
"learning_rate": 9.230076542098155e-05,
"loss": 1.8774,
"step": 191
},
{
"epoch": 0.2568561872909699,
"grad_norm": 0.3418707549571991,
"learning_rate": 9.225574065736155e-05,
"loss": 1.8545,
"step": 192
},
{
"epoch": 0.2581939799331104,
"grad_norm": 0.32373929023742676,
"learning_rate": 9.221071589374156e-05,
"loss": 1.4021,
"step": 193
},
{
"epoch": 0.25953177257525084,
"grad_norm": 0.32591140270233154,
"learning_rate": 9.216569113012157e-05,
"loss": 1.9349,
"step": 194
},
{
"epoch": 0.2608695652173913,
"grad_norm": 0.8838204741477966,
"learning_rate": 9.212066636650159e-05,
"loss": 1.9257,
"step": 195
},
{
"epoch": 0.26220735785953175,
"grad_norm": 0.2679453194141388,
"learning_rate": 9.207564160288158e-05,
"loss": 2.2633,
"step": 196
},
{
"epoch": 0.26354515050167227,
"grad_norm": 0.36055904626846313,
"learning_rate": 9.203061683926159e-05,
"loss": 1.9155,
"step": 197
},
{
"epoch": 0.2648829431438127,
"grad_norm": 0.28946951031684875,
"learning_rate": 9.19855920756416e-05,
"loss": 2.0331,
"step": 198
},
{
"epoch": 0.2662207357859532,
"grad_norm": 0.38088685274124146,
"learning_rate": 9.194056731202162e-05,
"loss": 1.9669,
"step": 199
},
{
"epoch": 0.26755852842809363,
"grad_norm": 0.3809536099433899,
"learning_rate": 9.189554254840163e-05,
"loss": 2.1654,
"step": 200
},
{
"epoch": 0.2688963210702341,
"grad_norm": 0.3993472754955292,
"learning_rate": 9.185051778478163e-05,
"loss": 1.4857,
"step": 201
},
{
"epoch": 0.2702341137123746,
"grad_norm": 0.3034169375896454,
"learning_rate": 9.180549302116165e-05,
"loss": 2.2763,
"step": 202
},
{
"epoch": 0.27157190635451506,
"grad_norm": 0.3558633625507355,
"learning_rate": 9.176046825754166e-05,
"loss": 2.0766,
"step": 203
},
{
"epoch": 0.2729096989966555,
"grad_norm": 0.5177504420280457,
"learning_rate": 9.171544349392167e-05,
"loss": 1.7944,
"step": 204
},
{
"epoch": 0.27424749163879597,
"grad_norm": 0.3450024425983429,
"learning_rate": 9.167041873030166e-05,
"loss": 1.4671,
"step": 205
},
{
"epoch": 0.2755852842809365,
"grad_norm": 0.44552841782569885,
"learning_rate": 9.162539396668168e-05,
"loss": 1.6058,
"step": 206
},
{
"epoch": 0.27692307692307694,
"grad_norm": 0.41852515935897827,
"learning_rate": 9.158036920306169e-05,
"loss": 1.9554,
"step": 207
},
{
"epoch": 0.2782608695652174,
"grad_norm": 0.3624606132507324,
"learning_rate": 9.15353444394417e-05,
"loss": 1.6291,
"step": 208
},
{
"epoch": 0.27959866220735785,
"grad_norm": 0.4587150812149048,
"learning_rate": 9.149031967582171e-05,
"loss": 1.9435,
"step": 209
},
{
"epoch": 0.2809364548494983,
"grad_norm": 0.3697861433029175,
"learning_rate": 9.144529491220172e-05,
"loss": 2.0404,
"step": 210
},
{
"epoch": 0.2822742474916388,
"grad_norm": 0.37733131647109985,
"learning_rate": 9.140027014858173e-05,
"loss": 1.8168,
"step": 211
},
{
"epoch": 0.2836120401337793,
"grad_norm": 0.32481226325035095,
"learning_rate": 9.135524538496173e-05,
"loss": 1.8865,
"step": 212
},
{
"epoch": 0.28494983277591973,
"grad_norm": 0.3467252552509308,
"learning_rate": 9.131022062134174e-05,
"loss": 1.5196,
"step": 213
},
{
"epoch": 0.2862876254180602,
"grad_norm": 0.27760034799575806,
"learning_rate": 9.126519585772175e-05,
"loss": 1.6441,
"step": 214
},
{
"epoch": 0.28762541806020064,
"grad_norm": 0.41142746806144714,
"learning_rate": 9.122017109410176e-05,
"loss": 1.8563,
"step": 215
},
{
"epoch": 0.28896321070234116,
"grad_norm": 0.504342257976532,
"learning_rate": 9.117514633048177e-05,
"loss": 1.639,
"step": 216
},
{
"epoch": 0.2903010033444816,
"grad_norm": 0.3190425932407379,
"learning_rate": 9.113012156686178e-05,
"loss": 2.2112,
"step": 217
},
{
"epoch": 0.29163879598662207,
"grad_norm": 0.2690954804420471,
"learning_rate": 9.108509680324179e-05,
"loss": 1.9693,
"step": 218
},
{
"epoch": 0.2929765886287625,
"grad_norm": 0.9039385914802551,
"learning_rate": 9.10400720396218e-05,
"loss": 1.8742,
"step": 219
},
{
"epoch": 0.29431438127090304,
"grad_norm": 0.4258744418621063,
"learning_rate": 9.09950472760018e-05,
"loss": 1.7538,
"step": 220
},
{
"epoch": 0.2956521739130435,
"grad_norm": 0.5416736006736755,
"learning_rate": 9.095002251238181e-05,
"loss": 1.4503,
"step": 221
},
{
"epoch": 0.29698996655518395,
"grad_norm": 0.43486952781677246,
"learning_rate": 9.090499774876182e-05,
"loss": 1.9158,
"step": 222
},
{
"epoch": 0.2983277591973244,
"grad_norm": 0.23655688762664795,
"learning_rate": 9.085997298514183e-05,
"loss": 1.6036,
"step": 223
},
{
"epoch": 0.29966555183946486,
"grad_norm": 0.3448864221572876,
"learning_rate": 9.081494822152185e-05,
"loss": 1.7351,
"step": 224
},
{
"epoch": 0.3010033444816054,
"grad_norm": 0.5123642683029175,
"learning_rate": 9.076992345790185e-05,
"loss": 1.6658,
"step": 225
},
{
"epoch": 0.30234113712374583,
"grad_norm": 0.5703243613243103,
"learning_rate": 9.072489869428186e-05,
"loss": 1.9601,
"step": 226
},
{
"epoch": 0.3036789297658863,
"grad_norm": 0.37126022577285767,
"learning_rate": 9.067987393066188e-05,
"loss": 1.6886,
"step": 227
},
{
"epoch": 0.30501672240802674,
"grad_norm": 0.3115082383155823,
"learning_rate": 9.063484916704189e-05,
"loss": 1.9697,
"step": 228
},
{
"epoch": 0.3063545150501672,
"grad_norm": 0.2590930759906769,
"learning_rate": 9.058982440342188e-05,
"loss": 2.0054,
"step": 229
},
{
"epoch": 0.3076923076923077,
"grad_norm": 0.28678327798843384,
"learning_rate": 9.054479963980189e-05,
"loss": 1.8215,
"step": 230
},
{
"epoch": 0.30903010033444817,
"grad_norm": 0.3587772846221924,
"learning_rate": 9.049977487618191e-05,
"loss": 1.9469,
"step": 231
},
{
"epoch": 0.3103678929765886,
"grad_norm": 0.40721800923347473,
"learning_rate": 9.045475011256192e-05,
"loss": 1.8971,
"step": 232
},
{
"epoch": 0.3117056856187291,
"grad_norm": 0.395236998796463,
"learning_rate": 9.040972534894192e-05,
"loss": 1.7274,
"step": 233
},
{
"epoch": 0.3130434782608696,
"grad_norm": 0.327338308095932,
"learning_rate": 9.036470058532192e-05,
"loss": 1.9656,
"step": 234
},
{
"epoch": 0.31438127090301005,
"grad_norm": 0.35737645626068115,
"learning_rate": 9.031967582170195e-05,
"loss": 1.7269,
"step": 235
},
{
"epoch": 0.3157190635451505,
"grad_norm": 0.376131147146225,
"learning_rate": 9.027465105808195e-05,
"loss": 1.6209,
"step": 236
},
{
"epoch": 0.31705685618729096,
"grad_norm": 1.226188063621521,
"learning_rate": 9.022962629446195e-05,
"loss": 1.7785,
"step": 237
},
{
"epoch": 0.3183946488294314,
"grad_norm": 0.3651161193847656,
"learning_rate": 9.018460153084197e-05,
"loss": 2.0646,
"step": 238
},
{
"epoch": 0.3197324414715719,
"grad_norm": 0.41171953082084656,
"learning_rate": 9.013957676722198e-05,
"loss": 1.8584,
"step": 239
},
{
"epoch": 0.3210702341137124,
"grad_norm": 0.29637786746025085,
"learning_rate": 9.009455200360199e-05,
"loss": 1.9911,
"step": 240
},
{
"epoch": 0.32240802675585284,
"grad_norm": 0.6785081028938293,
"learning_rate": 9.004952723998198e-05,
"loss": 1.9317,
"step": 241
},
{
"epoch": 0.3237458193979933,
"grad_norm": 0.3170675039291382,
"learning_rate": 9.0004502476362e-05,
"loss": 1.6286,
"step": 242
},
{
"epoch": 0.3250836120401338,
"grad_norm": 0.39666345715522766,
"learning_rate": 8.995947771274201e-05,
"loss": 2.0269,
"step": 243
},
{
"epoch": 0.32642140468227426,
"grad_norm": 0.42573514580726624,
"learning_rate": 8.991445294912202e-05,
"loss": 1.8167,
"step": 244
},
{
"epoch": 0.3277591973244147,
"grad_norm": 0.287176251411438,
"learning_rate": 8.986942818550203e-05,
"loss": 1.9762,
"step": 245
},
{
"epoch": 0.3290969899665552,
"grad_norm": 0.328698992729187,
"learning_rate": 8.982440342188204e-05,
"loss": 1.4258,
"step": 246
},
{
"epoch": 0.33043478260869563,
"grad_norm": 0.35098153352737427,
"learning_rate": 8.977937865826205e-05,
"loss": 1.7426,
"step": 247
},
{
"epoch": 0.33177257525083614,
"grad_norm": 0.3326055705547333,
"learning_rate": 8.973435389464206e-05,
"loss": 1.8434,
"step": 248
},
{
"epoch": 0.3331103678929766,
"grad_norm": 0.5356029272079468,
"learning_rate": 8.968932913102207e-05,
"loss": 1.6118,
"step": 249
},
{
"epoch": 0.33444816053511706,
"grad_norm": 0.3759405016899109,
"learning_rate": 8.964430436740207e-05,
"loss": 1.8523,
"step": 250
},
{
"epoch": 0.3357859531772575,
"grad_norm": 0.327251672744751,
"learning_rate": 8.959927960378208e-05,
"loss": 1.9094,
"step": 251
},
{
"epoch": 0.33712374581939797,
"grad_norm": 0.34880220890045166,
"learning_rate": 8.955425484016209e-05,
"loss": 1.9313,
"step": 252
},
{
"epoch": 0.3384615384615385,
"grad_norm": 0.34632259607315063,
"learning_rate": 8.95092300765421e-05,
"loss": 2.0977,
"step": 253
},
{
"epoch": 0.33979933110367894,
"grad_norm": 0.34165191650390625,
"learning_rate": 8.946420531292211e-05,
"loss": 1.7803,
"step": 254
},
{
"epoch": 0.3411371237458194,
"grad_norm": 0.4165712296962738,
"learning_rate": 8.941918054930212e-05,
"loss": 1.9873,
"step": 255
},
{
"epoch": 0.34247491638795985,
"grad_norm": 0.6011919379234314,
"learning_rate": 8.937415578568214e-05,
"loss": 1.9191,
"step": 256
},
{
"epoch": 0.34381270903010036,
"grad_norm": 1.136637568473816,
"learning_rate": 8.932913102206213e-05,
"loss": 1.6417,
"step": 257
},
{
"epoch": 0.3451505016722408,
"grad_norm": 0.3873753845691681,
"learning_rate": 8.928410625844214e-05,
"loss": 1.9526,
"step": 258
},
{
"epoch": 0.3464882943143813,
"grad_norm": 0.38796165585517883,
"learning_rate": 8.923908149482215e-05,
"loss": 1.9183,
"step": 259
},
{
"epoch": 0.34782608695652173,
"grad_norm": 0.37068337202072144,
"learning_rate": 8.919405673120217e-05,
"loss": 1.9615,
"step": 260
},
{
"epoch": 0.3491638795986622,
"grad_norm": 0.41028955578804016,
"learning_rate": 8.914903196758217e-05,
"loss": 1.5558,
"step": 261
},
{
"epoch": 0.3505016722408027,
"grad_norm": 0.6443496942520142,
"learning_rate": 8.910400720396218e-05,
"loss": 1.8737,
"step": 262
},
{
"epoch": 0.35183946488294315,
"grad_norm": 0.3643587827682495,
"learning_rate": 8.90589824403422e-05,
"loss": 1.8952,
"step": 263
},
{
"epoch": 0.3531772575250836,
"grad_norm": 0.3416200578212738,
"learning_rate": 8.901395767672221e-05,
"loss": 2.0677,
"step": 264
},
{
"epoch": 0.35451505016722407,
"grad_norm": 1.0528233051300049,
"learning_rate": 8.89689329131022e-05,
"loss": 2.3153,
"step": 265
},
{
"epoch": 0.3558528428093645,
"grad_norm": 0.49315449595451355,
"learning_rate": 8.892390814948221e-05,
"loss": 1.9844,
"step": 266
},
{
"epoch": 0.35719063545150503,
"grad_norm": 0.3391367793083191,
"learning_rate": 8.887888338586223e-05,
"loss": 1.5993,
"step": 267
},
{
"epoch": 0.3585284280936455,
"grad_norm": 0.3510306477546692,
"learning_rate": 8.883385862224224e-05,
"loss": 1.8076,
"step": 268
},
{
"epoch": 0.35986622073578595,
"grad_norm": 0.36950451135635376,
"learning_rate": 8.878883385862224e-05,
"loss": 1.701,
"step": 269
},
{
"epoch": 0.3612040133779264,
"grad_norm": 0.4125894606113434,
"learning_rate": 8.874380909500225e-05,
"loss": 1.9721,
"step": 270
},
{
"epoch": 0.3625418060200669,
"grad_norm": 0.35017120838165283,
"learning_rate": 8.869878433138227e-05,
"loss": 1.8645,
"step": 271
},
{
"epoch": 0.36387959866220737,
"grad_norm": 0.30292269587516785,
"learning_rate": 8.865375956776228e-05,
"loss": 1.758,
"step": 272
},
{
"epoch": 0.3652173913043478,
"grad_norm": 0.8412917256355286,
"learning_rate": 8.860873480414229e-05,
"loss": 2.1873,
"step": 273
},
{
"epoch": 0.3665551839464883,
"grad_norm": 0.2973708212375641,
"learning_rate": 8.85637100405223e-05,
"loss": 1.9731,
"step": 274
},
{
"epoch": 0.36789297658862874,
"grad_norm": 1.3152915239334106,
"learning_rate": 8.85186852769023e-05,
"loss": 2.067,
"step": 275
},
{
"epoch": 0.36923076923076925,
"grad_norm": 0.30133023858070374,
"learning_rate": 8.847366051328231e-05,
"loss": 1.9556,
"step": 276
},
{
"epoch": 0.3705685618729097,
"grad_norm": 0.3463495671749115,
"learning_rate": 8.842863574966232e-05,
"loss": 1.9353,
"step": 277
},
{
"epoch": 0.37190635451505016,
"grad_norm": 1.1229948997497559,
"learning_rate": 8.838361098604233e-05,
"loss": 1.2522,
"step": 278
},
{
"epoch": 0.3732441471571906,
"grad_norm": 0.38608673214912415,
"learning_rate": 8.833858622242234e-05,
"loss": 1.8587,
"step": 279
},
{
"epoch": 0.3745819397993311,
"grad_norm": 0.31167173385620117,
"learning_rate": 8.829356145880235e-05,
"loss": 1.9185,
"step": 280
},
{
"epoch": 0.3759197324414716,
"grad_norm": 0.5619071125984192,
"learning_rate": 8.824853669518235e-05,
"loss": 1.814,
"step": 281
},
{
"epoch": 0.37725752508361204,
"grad_norm": 0.5451438426971436,
"learning_rate": 8.820351193156236e-05,
"loss": 1.5962,
"step": 282
},
{
"epoch": 0.3785953177257525,
"grad_norm": 0.41030535101890564,
"learning_rate": 8.815848716794237e-05,
"loss": 1.7109,
"step": 283
},
{
"epoch": 0.37993311036789296,
"grad_norm": 0.8977394104003906,
"learning_rate": 8.811346240432238e-05,
"loss": 2.0786,
"step": 284
},
{
"epoch": 0.38127090301003347,
"grad_norm": 0.35019567608833313,
"learning_rate": 8.806843764070239e-05,
"loss": 1.8875,
"step": 285
},
{
"epoch": 0.3826086956521739,
"grad_norm": 0.4626282751560211,
"learning_rate": 8.80234128770824e-05,
"loss": 2.0935,
"step": 286
},
{
"epoch": 0.3839464882943144,
"grad_norm": 0.3663441240787506,
"learning_rate": 8.79783881134624e-05,
"loss": 1.9519,
"step": 287
},
{
"epoch": 0.38528428093645484,
"grad_norm": 0.3869033753871918,
"learning_rate": 8.793336334984241e-05,
"loss": 1.876,
"step": 288
},
{
"epoch": 0.3866220735785953,
"grad_norm": 0.43982234597206116,
"learning_rate": 8.788833858622242e-05,
"loss": 1.9525,
"step": 289
},
{
"epoch": 0.3879598662207358,
"grad_norm": 0.2982971966266632,
"learning_rate": 8.784331382260243e-05,
"loss": 2.0443,
"step": 290
},
{
"epoch": 0.38929765886287626,
"grad_norm": 0.3693241477012634,
"learning_rate": 8.779828905898244e-05,
"loss": 1.9276,
"step": 291
},
{
"epoch": 0.3906354515050167,
"grad_norm": 0.39458420872688293,
"learning_rate": 8.775326429536246e-05,
"loss": 1.6411,
"step": 292
},
{
"epoch": 0.3919732441471572,
"grad_norm": 1.7858718633651733,
"learning_rate": 8.770823953174246e-05,
"loss": 1.8085,
"step": 293
},
{
"epoch": 0.3933110367892977,
"grad_norm": 0.4990810751914978,
"learning_rate": 8.766321476812247e-05,
"loss": 1.9774,
"step": 294
},
{
"epoch": 0.39464882943143814,
"grad_norm": 0.5808725357055664,
"learning_rate": 8.761819000450247e-05,
"loss": 2.1432,
"step": 295
},
{
"epoch": 0.3959866220735786,
"grad_norm": 1.3691102266311646,
"learning_rate": 8.75731652408825e-05,
"loss": 1.869,
"step": 296
},
{
"epoch": 0.39732441471571905,
"grad_norm": 0.4151483476161957,
"learning_rate": 8.75281404772625e-05,
"loss": 2.0903,
"step": 297
},
{
"epoch": 0.3986622073578595,
"grad_norm": 0.34775832295417786,
"learning_rate": 8.74831157136425e-05,
"loss": 2.2351,
"step": 298
},
{
"epoch": 0.4,
"grad_norm": 0.35934117436408997,
"learning_rate": 8.743809095002252e-05,
"loss": 2.1257,
"step": 299
},
{
"epoch": 0.4013377926421405,
"grad_norm": 0.5283246636390686,
"learning_rate": 8.739306618640253e-05,
"loss": 2.0353,
"step": 300
},
{
"epoch": 0.40267558528428093,
"grad_norm": 0.44409871101379395,
"learning_rate": 8.734804142278254e-05,
"loss": 1.7333,
"step": 301
},
{
"epoch": 0.4040133779264214,
"grad_norm": 0.41795071959495544,
"learning_rate": 8.730301665916253e-05,
"loss": 1.9624,
"step": 302
},
{
"epoch": 0.40535117056856185,
"grad_norm": 0.338204950094223,
"learning_rate": 8.725799189554256e-05,
"loss": 1.9867,
"step": 303
},
{
"epoch": 0.40668896321070236,
"grad_norm": 0.3992144763469696,
"learning_rate": 8.721296713192257e-05,
"loss": 1.7873,
"step": 304
},
{
"epoch": 0.4080267558528428,
"grad_norm": 0.41751959919929504,
"learning_rate": 8.716794236830257e-05,
"loss": 2.1863,
"step": 305
},
{
"epoch": 0.40936454849498327,
"grad_norm": 1.1421892642974854,
"learning_rate": 8.712291760468257e-05,
"loss": 2.1859,
"step": 306
},
{
"epoch": 0.4107023411371237,
"grad_norm": 0.35610154271125793,
"learning_rate": 8.707789284106259e-05,
"loss": 1.8319,
"step": 307
},
{
"epoch": 0.41204013377926424,
"grad_norm": 0.5834404826164246,
"learning_rate": 8.70328680774426e-05,
"loss": 1.819,
"step": 308
},
{
"epoch": 0.4133779264214047,
"grad_norm": 0.4020130932331085,
"learning_rate": 8.698784331382261e-05,
"loss": 1.6767,
"step": 309
},
{
"epoch": 0.41471571906354515,
"grad_norm": 0.7889217138290405,
"learning_rate": 8.694281855020262e-05,
"loss": 1.9374,
"step": 310
},
{
"epoch": 0.4160535117056856,
"grad_norm": 0.37698855996131897,
"learning_rate": 8.689779378658263e-05,
"loss": 2.0154,
"step": 311
},
{
"epoch": 0.41739130434782606,
"grad_norm": 0.43415555357933044,
"learning_rate": 8.685276902296263e-05,
"loss": 1.9254,
"step": 312
},
{
"epoch": 0.4187290969899666,
"grad_norm": 0.3221052587032318,
"learning_rate": 8.680774425934264e-05,
"loss": 2.0957,
"step": 313
},
{
"epoch": 0.42006688963210703,
"grad_norm": 0.3111625909805298,
"learning_rate": 8.676271949572265e-05,
"loss": 1.7638,
"step": 314
},
{
"epoch": 0.4214046822742475,
"grad_norm": 0.356253445148468,
"learning_rate": 8.671769473210266e-05,
"loss": 1.9813,
"step": 315
},
{
"epoch": 0.42274247491638794,
"grad_norm": 0.7372254729270935,
"learning_rate": 8.667266996848267e-05,
"loss": 1.8251,
"step": 316
},
{
"epoch": 0.4240802675585284,
"grad_norm": 1.448208212852478,
"learning_rate": 8.662764520486268e-05,
"loss": 1.6077,
"step": 317
},
{
"epoch": 0.4254180602006689,
"grad_norm": 0.653560221195221,
"learning_rate": 8.658262044124269e-05,
"loss": 1.9399,
"step": 318
},
{
"epoch": 0.42675585284280937,
"grad_norm": 0.4218508005142212,
"learning_rate": 8.65375956776227e-05,
"loss": 2.01,
"step": 319
},
{
"epoch": 0.4280936454849498,
"grad_norm": 0.3056946396827698,
"learning_rate": 8.64925709140027e-05,
"loss": 1.8975,
"step": 320
},
{
"epoch": 0.4294314381270903,
"grad_norm": 0.5265414118766785,
"learning_rate": 8.644754615038273e-05,
"loss": 1.406,
"step": 321
},
{
"epoch": 0.4307692307692308,
"grad_norm": 0.34402531385421753,
"learning_rate": 8.640252138676272e-05,
"loss": 1.8997,
"step": 322
},
{
"epoch": 0.43210702341137125,
"grad_norm": 0.46225470304489136,
"learning_rate": 8.635749662314273e-05,
"loss": 2.0228,
"step": 323
},
{
"epoch": 0.4334448160535117,
"grad_norm": 0.4547097682952881,
"learning_rate": 8.631247185952274e-05,
"loss": 2.0001,
"step": 324
},
{
"epoch": 0.43478260869565216,
"grad_norm": 2.0929479598999023,
"learning_rate": 8.626744709590276e-05,
"loss": 1.9063,
"step": 325
},
{
"epoch": 0.4361204013377926,
"grad_norm": 0.7159757614135742,
"learning_rate": 8.622242233228275e-05,
"loss": 1.9047,
"step": 326
},
{
"epoch": 0.43745819397993313,
"grad_norm": 3.7366178035736084,
"learning_rate": 8.617739756866276e-05,
"loss": 2.4446,
"step": 327
},
{
"epoch": 0.4387959866220736,
"grad_norm": 0.5191693902015686,
"learning_rate": 8.613237280504279e-05,
"loss": 2.0927,
"step": 328
},
{
"epoch": 0.44013377926421404,
"grad_norm": 0.49155277013778687,
"learning_rate": 8.60873480414228e-05,
"loss": 1.7654,
"step": 329
},
{
"epoch": 0.4414715719063545,
"grad_norm": 0.3525542616844177,
"learning_rate": 8.604232327780279e-05,
"loss": 1.791,
"step": 330
},
{
"epoch": 0.442809364548495,
"grad_norm": 0.5055109262466431,
"learning_rate": 8.59972985141828e-05,
"loss": 1.7466,
"step": 331
},
{
"epoch": 0.44414715719063547,
"grad_norm": 0.3834606409072876,
"learning_rate": 8.595227375056282e-05,
"loss": 1.9207,
"step": 332
},
{
"epoch": 0.4454849498327759,
"grad_norm": 0.48070740699768066,
"learning_rate": 8.590724898694283e-05,
"loss": 1.899,
"step": 333
},
{
"epoch": 0.4468227424749164,
"grad_norm": 0.30500054359436035,
"learning_rate": 8.586222422332282e-05,
"loss": 2.1356,
"step": 334
},
{
"epoch": 0.44816053511705684,
"grad_norm": 0.402435839176178,
"learning_rate": 8.581719945970285e-05,
"loss": 1.8639,
"step": 335
},
{
"epoch": 0.44949832775919735,
"grad_norm": 0.4309161305427551,
"learning_rate": 8.577217469608285e-05,
"loss": 2.1182,
"step": 336
},
{
"epoch": 0.4508361204013378,
"grad_norm": 0.6321202516555786,
"learning_rate": 8.572714993246286e-05,
"loss": 1.6111,
"step": 337
},
{
"epoch": 0.45217391304347826,
"grad_norm": 0.37674644589424133,
"learning_rate": 8.568212516884286e-05,
"loss": 1.4399,
"step": 338
},
{
"epoch": 0.4535117056856187,
"grad_norm": 0.38387489318847656,
"learning_rate": 8.563710040522288e-05,
"loss": 1.9581,
"step": 339
},
{
"epoch": 0.45484949832775917,
"grad_norm": 0.37002870440483093,
"learning_rate": 8.559207564160289e-05,
"loss": 1.7161,
"step": 340
},
{
"epoch": 0.4561872909698997,
"grad_norm": 0.3412352502346039,
"learning_rate": 8.55470508779829e-05,
"loss": 1.8263,
"step": 341
},
{
"epoch": 0.45752508361204014,
"grad_norm": 0.47680556774139404,
"learning_rate": 8.550202611436289e-05,
"loss": 2.0372,
"step": 342
},
{
"epoch": 0.4588628762541806,
"grad_norm": 0.5688019394874573,
"learning_rate": 8.545700135074291e-05,
"loss": 2.0338,
"step": 343
},
{
"epoch": 0.46020066889632105,
"grad_norm": 0.4578900635242462,
"learning_rate": 8.541197658712292e-05,
"loss": 2.1086,
"step": 344
},
{
"epoch": 0.46153846153846156,
"grad_norm": 0.3213759958744049,
"learning_rate": 8.536695182350293e-05,
"loss": 1.9701,
"step": 345
},
{
"epoch": 0.462876254180602,
"grad_norm": 0.3761240541934967,
"learning_rate": 8.532192705988294e-05,
"loss": 2.042,
"step": 346
},
{
"epoch": 0.4642140468227425,
"grad_norm": 0.6948366761207581,
"learning_rate": 8.527690229626295e-05,
"loss": 1.7776,
"step": 347
},
{
"epoch": 0.46555183946488293,
"grad_norm": 0.38398241996765137,
"learning_rate": 8.523187753264296e-05,
"loss": 1.6168,
"step": 348
},
{
"epoch": 0.4668896321070234,
"grad_norm": 0.4216487407684326,
"learning_rate": 8.518685276902297e-05,
"loss": 1.5656,
"step": 349
},
{
"epoch": 0.4682274247491639,
"grad_norm": 0.3816230297088623,
"learning_rate": 8.514182800540297e-05,
"loss": 1.9222,
"step": 350
},
{
"epoch": 0.46956521739130436,
"grad_norm": 0.5072731971740723,
"learning_rate": 8.509680324178298e-05,
"loss": 2.257,
"step": 351
},
{
"epoch": 0.4709030100334448,
"grad_norm": 0.38990819454193115,
"learning_rate": 8.505177847816299e-05,
"loss": 1.6184,
"step": 352
},
{
"epoch": 0.47224080267558527,
"grad_norm": 0.3647673726081848,
"learning_rate": 8.500675371454301e-05,
"loss": 1.9753,
"step": 353
},
{
"epoch": 0.4735785953177257,
"grad_norm": 0.503564715385437,
"learning_rate": 8.496172895092301e-05,
"loss": 1.88,
"step": 354
},
{
"epoch": 0.47491638795986624,
"grad_norm": 0.4017346501350403,
"learning_rate": 8.491670418730302e-05,
"loss": 1.8751,
"step": 355
},
{
"epoch": 0.4762541806020067,
"grad_norm": 0.3733857274055481,
"learning_rate": 8.487167942368303e-05,
"loss": 1.8511,
"step": 356
},
{
"epoch": 0.47759197324414715,
"grad_norm": 0.35360386967658997,
"learning_rate": 8.482665466006305e-05,
"loss": 1.8221,
"step": 357
},
{
"epoch": 0.4789297658862876,
"grad_norm": 0.38599494099617004,
"learning_rate": 8.478162989644304e-05,
"loss": 1.8883,
"step": 358
},
{
"epoch": 0.4802675585284281,
"grad_norm": 0.3708478808403015,
"learning_rate": 8.473660513282305e-05,
"loss": 1.3802,
"step": 359
},
{
"epoch": 0.4816053511705686,
"grad_norm": 0.903266191482544,
"learning_rate": 8.469158036920306e-05,
"loss": 1.6616,
"step": 360
},
{
"epoch": 0.48294314381270903,
"grad_norm": 0.3742479681968689,
"learning_rate": 8.464655560558308e-05,
"loss": 1.6105,
"step": 361
},
{
"epoch": 0.4842809364548495,
"grad_norm": 0.34880515933036804,
"learning_rate": 8.460153084196308e-05,
"loss": 1.8501,
"step": 362
},
{
"epoch": 0.48561872909698994,
"grad_norm": 0.3260308802127838,
"learning_rate": 8.455650607834309e-05,
"loss": 1.9235,
"step": 363
},
{
"epoch": 0.48695652173913045,
"grad_norm": 0.4487158954143524,
"learning_rate": 8.451148131472311e-05,
"loss": 1.3566,
"step": 364
},
{
"epoch": 0.4882943143812709,
"grad_norm": 0.6587825417518616,
"learning_rate": 8.446645655110312e-05,
"loss": 1.9224,
"step": 365
},
{
"epoch": 0.48963210702341137,
"grad_norm": 0.3895570635795593,
"learning_rate": 8.442143178748311e-05,
"loss": 2.0373,
"step": 366
},
{
"epoch": 0.4909698996655518,
"grad_norm": 0.4009748101234436,
"learning_rate": 8.437640702386312e-05,
"loss": 1.6818,
"step": 367
},
{
"epoch": 0.49230769230769234,
"grad_norm": 0.36814218759536743,
"learning_rate": 8.433138226024314e-05,
"loss": 1.9883,
"step": 368
},
{
"epoch": 0.4936454849498328,
"grad_norm": 0.3416324257850647,
"learning_rate": 8.428635749662315e-05,
"loss": 1.7745,
"step": 369
},
{
"epoch": 0.49498327759197325,
"grad_norm": 0.3150230646133423,
"learning_rate": 8.424133273300316e-05,
"loss": 1.8217,
"step": 370
},
{
"epoch": 0.4963210702341137,
"grad_norm": 0.6296526193618774,
"learning_rate": 8.419630796938317e-05,
"loss": 2.105,
"step": 371
},
{
"epoch": 0.49765886287625416,
"grad_norm": 0.37235918641090393,
"learning_rate": 8.415128320576318e-05,
"loss": 1.5564,
"step": 372
},
{
"epoch": 0.49899665551839467,
"grad_norm": 0.373354434967041,
"learning_rate": 8.410625844214319e-05,
"loss": 1.7034,
"step": 373
},
{
"epoch": 0.5003344481605351,
"grad_norm": 0.3129545748233795,
"learning_rate": 8.40612336785232e-05,
"loss": 1.9642,
"step": 374
},
{
"epoch": 0.5016722408026756,
"grad_norm": 0.30649834871292114,
"learning_rate": 8.40162089149032e-05,
"loss": 1.449,
"step": 375
},
{
"epoch": 0.5030100334448161,
"grad_norm": 0.2958093285560608,
"learning_rate": 8.397118415128321e-05,
"loss": 1.9685,
"step": 376
},
{
"epoch": 0.5043478260869565,
"grad_norm": 0.336683064699173,
"learning_rate": 8.392615938766322e-05,
"loss": 1.9817,
"step": 377
},
{
"epoch": 0.505685618729097,
"grad_norm": 0.9705008268356323,
"learning_rate": 8.388113462404323e-05,
"loss": 1.5965,
"step": 378
},
{
"epoch": 0.5070234113712374,
"grad_norm": 0.4110693037509918,
"learning_rate": 8.383610986042324e-05,
"loss": 1.7417,
"step": 379
},
{
"epoch": 0.5083612040133779,
"grad_norm": 0.8143429160118103,
"learning_rate": 8.379108509680325e-05,
"loss": 2.2159,
"step": 380
},
{
"epoch": 0.5096989966555184,
"grad_norm": 0.3616657555103302,
"learning_rate": 8.374606033318325e-05,
"loss": 2.0832,
"step": 381
},
{
"epoch": 0.5110367892976588,
"grad_norm": 0.500678300857544,
"learning_rate": 8.370103556956326e-05,
"loss": 1.5084,
"step": 382
},
{
"epoch": 0.5123745819397993,
"grad_norm": 0.4309229254722595,
"learning_rate": 8.365601080594327e-05,
"loss": 1.1731,
"step": 383
},
{
"epoch": 0.5137123745819397,
"grad_norm": 0.34316661953926086,
"learning_rate": 8.361098604232328e-05,
"loss": 1.7091,
"step": 384
},
{
"epoch": 0.5150501672240803,
"grad_norm": 0.38358014822006226,
"learning_rate": 8.356596127870329e-05,
"loss": 2.0595,
"step": 385
},
{
"epoch": 0.5163879598662208,
"grad_norm": 0.3089703321456909,
"learning_rate": 8.35209365150833e-05,
"loss": 1.6416,
"step": 386
},
{
"epoch": 0.5177257525083612,
"grad_norm": 0.3062998056411743,
"learning_rate": 8.34759117514633e-05,
"loss": 1.4986,
"step": 387
},
{
"epoch": 0.5190635451505017,
"grad_norm": 0.5895267724990845,
"learning_rate": 8.343088698784331e-05,
"loss": 1.8914,
"step": 388
},
{
"epoch": 0.5204013377926422,
"grad_norm": 0.46645572781562805,
"learning_rate": 8.338586222422334e-05,
"loss": 1.6489,
"step": 389
},
{
"epoch": 0.5217391304347826,
"grad_norm": 0.7197579741477966,
"learning_rate": 8.334083746060334e-05,
"loss": 2.0879,
"step": 390
},
{
"epoch": 0.5230769230769231,
"grad_norm": 0.38837894797325134,
"learning_rate": 8.329581269698334e-05,
"loss": 1.4096,
"step": 391
},
{
"epoch": 0.5244147157190635,
"grad_norm": 0.3910644054412842,
"learning_rate": 8.325078793336335e-05,
"loss": 1.8055,
"step": 392
},
{
"epoch": 0.525752508361204,
"grad_norm": 0.5699204206466675,
"learning_rate": 8.320576316974337e-05,
"loss": 1.9215,
"step": 393
},
{
"epoch": 0.5270903010033445,
"grad_norm": 0.4684099853038788,
"learning_rate": 8.316073840612338e-05,
"loss": 1.6515,
"step": 394
},
{
"epoch": 0.5284280936454849,
"grad_norm": 0.3977600634098053,
"learning_rate": 8.311571364250337e-05,
"loss": 1.8285,
"step": 395
},
{
"epoch": 0.5297658862876254,
"grad_norm": 0.3171498775482178,
"learning_rate": 8.307068887888338e-05,
"loss": 1.8305,
"step": 396
},
{
"epoch": 0.5311036789297658,
"grad_norm": 0.321153849363327,
"learning_rate": 8.30256641152634e-05,
"loss": 1.582,
"step": 397
},
{
"epoch": 0.5324414715719064,
"grad_norm": 0.8034740090370178,
"learning_rate": 8.298063935164341e-05,
"loss": 1.816,
"step": 398
},
{
"epoch": 0.5337792642140469,
"grad_norm": 0.3740721642971039,
"learning_rate": 8.293561458802341e-05,
"loss": 1.9171,
"step": 399
},
{
"epoch": 0.5351170568561873,
"grad_norm": 0.35908418893814087,
"learning_rate": 8.289058982440343e-05,
"loss": 1.9799,
"step": 400
},
{
"epoch": 0.5364548494983278,
"grad_norm": 0.3337550163269043,
"learning_rate": 8.284556506078344e-05,
"loss": 1.9837,
"step": 401
},
{
"epoch": 0.5377926421404682,
"grad_norm": 0.47816741466522217,
"learning_rate": 8.280054029716345e-05,
"loss": 1.4683,
"step": 402
},
{
"epoch": 0.5391304347826087,
"grad_norm": 0.3333340585231781,
"learning_rate": 8.275551553354344e-05,
"loss": 2.0083,
"step": 403
},
{
"epoch": 0.5404682274247492,
"grad_norm": 0.5132210850715637,
"learning_rate": 8.271049076992347e-05,
"loss": 1.9655,
"step": 404
},
{
"epoch": 0.5418060200668896,
"grad_norm": 0.49101531505584717,
"learning_rate": 8.266546600630347e-05,
"loss": 1.9655,
"step": 405
},
{
"epoch": 0.5431438127090301,
"grad_norm": 0.4403785765171051,
"learning_rate": 8.262044124268348e-05,
"loss": 2.1054,
"step": 406
},
{
"epoch": 0.5444816053511705,
"grad_norm": 0.4216044843196869,
"learning_rate": 8.257541647906349e-05,
"loss": 1.763,
"step": 407
},
{
"epoch": 0.545819397993311,
"grad_norm": 0.36043408513069153,
"learning_rate": 8.25303917154435e-05,
"loss": 1.7495,
"step": 408
},
{
"epoch": 0.5471571906354515,
"grad_norm": 0.5740343332290649,
"learning_rate": 8.248536695182351e-05,
"loss": 2.0926,
"step": 409
},
{
"epoch": 0.5484949832775919,
"grad_norm": 0.3019751012325287,
"learning_rate": 8.244034218820352e-05,
"loss": 2.029,
"step": 410
},
{
"epoch": 0.5498327759197325,
"grad_norm": 0.3443738520145416,
"learning_rate": 8.239531742458353e-05,
"loss": 1.9477,
"step": 411
},
{
"epoch": 0.551170568561873,
"grad_norm": 0.6188628673553467,
"learning_rate": 8.235029266096353e-05,
"loss": 1.6712,
"step": 412
},
{
"epoch": 0.5525083612040134,
"grad_norm": 0.39438313245773315,
"learning_rate": 8.230526789734354e-05,
"loss": 1.8374,
"step": 413
},
{
"epoch": 0.5538461538461539,
"grad_norm": 0.2995496988296509,
"learning_rate": 8.226024313372355e-05,
"loss": 1.8306,
"step": 414
},
{
"epoch": 0.5551839464882943,
"grad_norm": 0.3863040804862976,
"learning_rate": 8.221521837010356e-05,
"loss": 1.175,
"step": 415
},
{
"epoch": 0.5565217391304348,
"grad_norm": 0.41024214029312134,
"learning_rate": 8.217019360648357e-05,
"loss": 1.6772,
"step": 416
},
{
"epoch": 0.5578595317725753,
"grad_norm": 1.2178558111190796,
"learning_rate": 8.212516884286358e-05,
"loss": 1.5956,
"step": 417
},
{
"epoch": 0.5591973244147157,
"grad_norm": 0.656810998916626,
"learning_rate": 8.20801440792436e-05,
"loss": 1.6988,
"step": 418
},
{
"epoch": 0.5605351170568562,
"grad_norm": 0.36660322546958923,
"learning_rate": 8.20351193156236e-05,
"loss": 2.0613,
"step": 419
},
{
"epoch": 0.5618729096989966,
"grad_norm": 0.4018813967704773,
"learning_rate": 8.19900945520036e-05,
"loss": 1.7731,
"step": 420
},
{
"epoch": 0.5632107023411371,
"grad_norm": 0.3366681635379791,
"learning_rate": 8.194506978838361e-05,
"loss": 1.8276,
"step": 421
},
{
"epoch": 0.5645484949832776,
"grad_norm": 0.615424394607544,
"learning_rate": 8.190004502476363e-05,
"loss": 1.4514,
"step": 422
},
{
"epoch": 0.565886287625418,
"grad_norm": 0.6425555944442749,
"learning_rate": 8.185502026114363e-05,
"loss": 1.6658,
"step": 423
},
{
"epoch": 0.5672240802675586,
"grad_norm": 0.3678185045719147,
"learning_rate": 8.180999549752364e-05,
"loss": 1.9313,
"step": 424
},
{
"epoch": 0.568561872909699,
"grad_norm": 0.43040040135383606,
"learning_rate": 8.176497073390366e-05,
"loss": 1.8427,
"step": 425
},
{
"epoch": 0.5698996655518395,
"grad_norm": 0.3130514323711395,
"learning_rate": 8.171994597028367e-05,
"loss": 1.8946,
"step": 426
},
{
"epoch": 0.57123745819398,
"grad_norm": 0.8794007301330566,
"learning_rate": 8.167492120666366e-05,
"loss": 1.8651,
"step": 427
},
{
"epoch": 0.5725752508361204,
"grad_norm": 0.3628743588924408,
"learning_rate": 8.162989644304367e-05,
"loss": 2.0068,
"step": 428
},
{
"epoch": 0.5739130434782609,
"grad_norm": 0.6228901743888855,
"learning_rate": 8.15848716794237e-05,
"loss": 1.7479,
"step": 429
},
{
"epoch": 0.5752508361204013,
"grad_norm": 0.3421112298965454,
"learning_rate": 8.15398469158037e-05,
"loss": 1.92,
"step": 430
},
{
"epoch": 0.5765886287625418,
"grad_norm": 0.40324121713638306,
"learning_rate": 8.14948221521837e-05,
"loss": 1.8626,
"step": 431
},
{
"epoch": 0.5779264214046823,
"grad_norm": 0.3942056894302368,
"learning_rate": 8.14497973885637e-05,
"loss": 1.8359,
"step": 432
},
{
"epoch": 0.5792642140468227,
"grad_norm": 0.3615321218967438,
"learning_rate": 8.140477262494373e-05,
"loss": 1.8924,
"step": 433
},
{
"epoch": 0.5806020066889632,
"grad_norm": 0.41227877140045166,
"learning_rate": 8.135974786132374e-05,
"loss": 1.6984,
"step": 434
},
{
"epoch": 0.5819397993311036,
"grad_norm": 1.1059563159942627,
"learning_rate": 8.131472309770373e-05,
"loss": 1.4266,
"step": 435
},
{
"epoch": 0.5832775919732441,
"grad_norm": 0.3503919541835785,
"learning_rate": 8.126969833408375e-05,
"loss": 2.1598,
"step": 436
},
{
"epoch": 0.5846153846153846,
"grad_norm": 0.5328121781349182,
"learning_rate": 8.122467357046376e-05,
"loss": 1.9038,
"step": 437
},
{
"epoch": 0.585953177257525,
"grad_norm": 0.3714490234851837,
"learning_rate": 8.117964880684377e-05,
"loss": 1.7892,
"step": 438
},
{
"epoch": 0.5872909698996656,
"grad_norm": 0.3813803493976593,
"learning_rate": 8.113462404322378e-05,
"loss": 1.7479,
"step": 439
},
{
"epoch": 0.5886287625418061,
"grad_norm": 0.3410722017288208,
"learning_rate": 8.108959927960379e-05,
"loss": 1.94,
"step": 440
},
{
"epoch": 0.5899665551839465,
"grad_norm": 0.3214760422706604,
"learning_rate": 8.10445745159838e-05,
"loss": 1.8526,
"step": 441
},
{
"epoch": 0.591304347826087,
"grad_norm": 0.6665300726890564,
"learning_rate": 8.09995497523638e-05,
"loss": 1.951,
"step": 442
},
{
"epoch": 0.5926421404682274,
"grad_norm": 0.33864709734916687,
"learning_rate": 8.095452498874381e-05,
"loss": 1.9193,
"step": 443
},
{
"epoch": 0.5939799331103679,
"grad_norm": 0.3851359188556671,
"learning_rate": 8.090950022512382e-05,
"loss": 1.8382,
"step": 444
},
{
"epoch": 0.5953177257525084,
"grad_norm": 0.3829728960990906,
"learning_rate": 8.086447546150383e-05,
"loss": 1.7381,
"step": 445
},
{
"epoch": 0.5966555183946488,
"grad_norm": 0.41027361154556274,
"learning_rate": 8.081945069788384e-05,
"loss": 1.9236,
"step": 446
},
{
"epoch": 0.5979933110367893,
"grad_norm": 0.4161706864833832,
"learning_rate": 8.077442593426385e-05,
"loss": 1.7539,
"step": 447
},
{
"epoch": 0.5993311036789297,
"grad_norm": 0.4396213889122009,
"learning_rate": 8.072940117064386e-05,
"loss": 1.7225,
"step": 448
},
{
"epoch": 0.6006688963210702,
"grad_norm": 0.36475786566734314,
"learning_rate": 8.068437640702387e-05,
"loss": 1.63,
"step": 449
},
{
"epoch": 0.6020066889632107,
"grad_norm": 0.5781269073486328,
"learning_rate": 8.063935164340387e-05,
"loss": 1.7179,
"step": 450
},
{
"epoch": 0.6033444816053511,
"grad_norm": 1.0035388469696045,
"learning_rate": 8.059432687978388e-05,
"loss": 1.7692,
"step": 451
},
{
"epoch": 0.6046822742474917,
"grad_norm": 0.3682602643966675,
"learning_rate": 8.054930211616389e-05,
"loss": 1.9889,
"step": 452
},
{
"epoch": 0.6060200668896321,
"grad_norm": 0.49924179911613464,
"learning_rate": 8.05042773525439e-05,
"loss": 1.5572,
"step": 453
},
{
"epoch": 0.6073578595317726,
"grad_norm": 0.34571748971939087,
"learning_rate": 8.045925258892392e-05,
"loss": 2.0112,
"step": 454
},
{
"epoch": 0.6086956521739131,
"grad_norm": 0.35000765323638916,
"learning_rate": 8.041422782530392e-05,
"loss": 1.852,
"step": 455
},
{
"epoch": 0.6100334448160535,
"grad_norm": 0.3942291736602783,
"learning_rate": 8.036920306168393e-05,
"loss": 1.7856,
"step": 456
},
{
"epoch": 0.611371237458194,
"grad_norm": 0.5689213275909424,
"learning_rate": 8.032417829806393e-05,
"loss": 1.9164,
"step": 457
},
{
"epoch": 0.6127090301003344,
"grad_norm": 0.4623394310474396,
"learning_rate": 8.027915353444396e-05,
"loss": 2.0554,
"step": 458
},
{
"epoch": 0.6140468227424749,
"grad_norm": 0.5661097168922424,
"learning_rate": 8.023412877082395e-05,
"loss": 1.6894,
"step": 459
},
{
"epoch": 0.6153846153846154,
"grad_norm": 0.41809728741645813,
"learning_rate": 8.018910400720396e-05,
"loss": 1.7757,
"step": 460
},
{
"epoch": 0.6167224080267558,
"grad_norm": 0.4237455427646637,
"learning_rate": 8.014407924358398e-05,
"loss": 2.1467,
"step": 461
},
{
"epoch": 0.6180602006688963,
"grad_norm": 0.3313245177268982,
"learning_rate": 8.009905447996399e-05,
"loss": 2.072,
"step": 462
},
{
"epoch": 0.6193979933110368,
"grad_norm": 0.3133883476257324,
"learning_rate": 8.0054029716344e-05,
"loss": 2.0608,
"step": 463
},
{
"epoch": 0.6207357859531772,
"grad_norm": 0.3357526659965515,
"learning_rate": 8.0009004952724e-05,
"loss": 1.5915,
"step": 464
},
{
"epoch": 0.6220735785953178,
"grad_norm": 0.39856886863708496,
"learning_rate": 7.996398018910402e-05,
"loss": 2.015,
"step": 465
},
{
"epoch": 0.6234113712374582,
"grad_norm": 0.3311580717563629,
"learning_rate": 7.991895542548402e-05,
"loss": 2.2078,
"step": 466
},
{
"epoch": 0.6247491638795987,
"grad_norm": 0.28546395897865295,
"learning_rate": 7.987393066186403e-05,
"loss": 1.7635,
"step": 467
},
{
"epoch": 0.6260869565217392,
"grad_norm": 0.8754998445510864,
"learning_rate": 7.982890589824403e-05,
"loss": 1.9734,
"step": 468
},
{
"epoch": 0.6274247491638796,
"grad_norm": 0.39375200867652893,
"learning_rate": 7.978388113462405e-05,
"loss": 1.3764,
"step": 469
},
{
"epoch": 0.6287625418060201,
"grad_norm": 0.3225364089012146,
"learning_rate": 7.973885637100406e-05,
"loss": 2.1192,
"step": 470
},
{
"epoch": 0.6301003344481605,
"grad_norm": 0.34391409158706665,
"learning_rate": 7.969383160738407e-05,
"loss": 1.9341,
"step": 471
},
{
"epoch": 0.631438127090301,
"grad_norm": 0.37655723094940186,
"learning_rate": 7.964880684376408e-05,
"loss": 1.9184,
"step": 472
},
{
"epoch": 0.6327759197324415,
"grad_norm": 0.32347163558006287,
"learning_rate": 7.960378208014408e-05,
"loss": 1.9153,
"step": 473
},
{
"epoch": 0.6341137123745819,
"grad_norm": 0.31293460726737976,
"learning_rate": 7.95587573165241e-05,
"loss": 1.9334,
"step": 474
},
{
"epoch": 0.6354515050167224,
"grad_norm": 0.29669955372810364,
"learning_rate": 7.95137325529041e-05,
"loss": 2.0717,
"step": 475
},
{
"epoch": 0.6367892976588628,
"grad_norm": 0.5345836281776428,
"learning_rate": 7.946870778928411e-05,
"loss": 1.6862,
"step": 476
},
{
"epoch": 0.6381270903010033,
"grad_norm": 0.3699991703033447,
"learning_rate": 7.942368302566412e-05,
"loss": 2.0417,
"step": 477
},
{
"epoch": 0.6394648829431439,
"grad_norm": 0.502368152141571,
"learning_rate": 7.937865826204413e-05,
"loss": 1.8139,
"step": 478
},
{
"epoch": 0.6408026755852843,
"grad_norm": 0.339144766330719,
"learning_rate": 7.933363349842414e-05,
"loss": 1.9849,
"step": 479
},
{
"epoch": 0.6421404682274248,
"grad_norm": 0.3838237226009369,
"learning_rate": 7.928860873480414e-05,
"loss": 1.7248,
"step": 480
},
{
"epoch": 0.6434782608695652,
"grad_norm": 0.3984738886356354,
"learning_rate": 7.924358397118415e-05,
"loss": 1.7374,
"step": 481
},
{
"epoch": 0.6448160535117057,
"grad_norm": 1.3994791507720947,
"learning_rate": 7.919855920756416e-05,
"loss": 1.3882,
"step": 482
},
{
"epoch": 0.6461538461538462,
"grad_norm": 0.41736477613449097,
"learning_rate": 7.915353444394417e-05,
"loss": 1.8299,
"step": 483
},
{
"epoch": 0.6474916387959866,
"grad_norm": 0.2818329930305481,
"learning_rate": 7.910850968032418e-05,
"loss": 1.94,
"step": 484
},
{
"epoch": 0.6488294314381271,
"grad_norm": 0.27325278520584106,
"learning_rate": 7.906348491670419e-05,
"loss": 2.001,
"step": 485
},
{
"epoch": 0.6501672240802676,
"grad_norm": 0.642238199710846,
"learning_rate": 7.90184601530842e-05,
"loss": 1.9489,
"step": 486
},
{
"epoch": 0.651505016722408,
"grad_norm": 0.38445815443992615,
"learning_rate": 7.897343538946422e-05,
"loss": 1.7597,
"step": 487
},
{
"epoch": 0.6528428093645485,
"grad_norm": 0.2907230257987976,
"learning_rate": 7.892841062584421e-05,
"loss": 1.65,
"step": 488
},
{
"epoch": 0.6541806020066889,
"grad_norm": 0.3835679888725281,
"learning_rate": 7.888338586222422e-05,
"loss": 1.8158,
"step": 489
},
{
"epoch": 0.6555183946488294,
"grad_norm": 0.44675248861312866,
"learning_rate": 7.883836109860424e-05,
"loss": 1.4216,
"step": 490
},
{
"epoch": 0.65685618729097,
"grad_norm": 0.385631263256073,
"learning_rate": 7.879333633498425e-05,
"loss": 1.8526,
"step": 491
},
{
"epoch": 0.6581939799331104,
"grad_norm": 0.3172449469566345,
"learning_rate": 7.874831157136425e-05,
"loss": 1.9237,
"step": 492
},
{
"epoch": 0.6595317725752509,
"grad_norm": 0.41748154163360596,
"learning_rate": 7.870328680774426e-05,
"loss": 2.026,
"step": 493
},
{
"epoch": 0.6608695652173913,
"grad_norm": 0.4081355035305023,
"learning_rate": 7.865826204412428e-05,
"loss": 1.4274,
"step": 494
},
{
"epoch": 0.6622073578595318,
"grad_norm": 0.338257759809494,
"learning_rate": 7.861323728050429e-05,
"loss": 1.8443,
"step": 495
},
{
"epoch": 0.6635451505016723,
"grad_norm": 0.40070194005966187,
"learning_rate": 7.856821251688428e-05,
"loss": 1.9438,
"step": 496
},
{
"epoch": 0.6648829431438127,
"grad_norm": 0.2751341462135315,
"learning_rate": 7.85231877532643e-05,
"loss": 1.9885,
"step": 497
},
{
"epoch": 0.6662207357859532,
"grad_norm": 0.3681652843952179,
"learning_rate": 7.847816298964431e-05,
"loss": 1.9029,
"step": 498
},
{
"epoch": 0.6675585284280936,
"grad_norm": 0.317685067653656,
"learning_rate": 7.843313822602432e-05,
"loss": 1.8493,
"step": 499
},
{
"epoch": 0.6688963210702341,
"grad_norm": 0.5167484283447266,
"learning_rate": 7.838811346240432e-05,
"loss": 2.0053,
"step": 500
},
{
"epoch": 0.6702341137123746,
"grad_norm": 0.283053457736969,
"learning_rate": 7.834308869878434e-05,
"loss": 1.9312,
"step": 501
},
{
"epoch": 0.671571906354515,
"grad_norm": 0.3423631191253662,
"learning_rate": 7.829806393516435e-05,
"loss": 2.0661,
"step": 502
},
{
"epoch": 0.6729096989966555,
"grad_norm": 0.41216039657592773,
"learning_rate": 7.825303917154436e-05,
"loss": 1.6058,
"step": 503
},
{
"epoch": 0.6742474916387959,
"grad_norm": 0.40254294872283936,
"learning_rate": 7.820801440792435e-05,
"loss": 2.0609,
"step": 504
},
{
"epoch": 0.6755852842809364,
"grad_norm": 0.35236918926239014,
"learning_rate": 7.816298964430437e-05,
"loss": 1.8236,
"step": 505
},
{
"epoch": 0.676923076923077,
"grad_norm": 0.7102991342544556,
"learning_rate": 7.811796488068438e-05,
"loss": 1.3943,
"step": 506
},
{
"epoch": 0.6782608695652174,
"grad_norm": 0.4407871961593628,
"learning_rate": 7.807294011706439e-05,
"loss": 1.7709,
"step": 507
},
{
"epoch": 0.6795986622073579,
"grad_norm": 0.29523321986198425,
"learning_rate": 7.80279153534444e-05,
"loss": 1.8408,
"step": 508
},
{
"epoch": 0.6809364548494983,
"grad_norm": 0.4053889811038971,
"learning_rate": 7.798289058982441e-05,
"loss": 1.9194,
"step": 509
},
{
"epoch": 0.6822742474916388,
"grad_norm": 0.4437839686870575,
"learning_rate": 7.793786582620442e-05,
"loss": 1.7094,
"step": 510
},
{
"epoch": 0.6836120401337793,
"grad_norm": 0.3048234283924103,
"learning_rate": 7.789284106258442e-05,
"loss": 1.7877,
"step": 511
},
{
"epoch": 0.6849498327759197,
"grad_norm": 0.3253330588340759,
"learning_rate": 7.784781629896443e-05,
"loss": 1.7939,
"step": 512
},
{
"epoch": 0.6862876254180602,
"grad_norm": 0.4025883674621582,
"learning_rate": 7.780279153534444e-05,
"loss": 1.3158,
"step": 513
},
{
"epoch": 0.6876254180602007,
"grad_norm": 0.3705989122390747,
"learning_rate": 7.775776677172445e-05,
"loss": 1.9503,
"step": 514
},
{
"epoch": 0.6889632107023411,
"grad_norm": 1.0391372442245483,
"learning_rate": 7.771274200810447e-05,
"loss": 1.9089,
"step": 515
},
{
"epoch": 0.6903010033444816,
"grad_norm": 0.33640894293785095,
"learning_rate": 7.766771724448447e-05,
"loss": 1.6701,
"step": 516
},
{
"epoch": 0.691638795986622,
"grad_norm": 0.34906384348869324,
"learning_rate": 7.762269248086448e-05,
"loss": 1.4419,
"step": 517
},
{
"epoch": 0.6929765886287625,
"grad_norm": 0.4226134121417999,
"learning_rate": 7.757766771724448e-05,
"loss": 1.4917,
"step": 518
},
{
"epoch": 0.6943143812709031,
"grad_norm": 0.3549327552318573,
"learning_rate": 7.753264295362451e-05,
"loss": 1.4032,
"step": 519
},
{
"epoch": 0.6956521739130435,
"grad_norm": 0.4137127995491028,
"learning_rate": 7.74876181900045e-05,
"loss": 1.8731,
"step": 520
},
{
"epoch": 0.696989966555184,
"grad_norm": 0.3058634102344513,
"learning_rate": 7.744259342638451e-05,
"loss": 1.8576,
"step": 521
},
{
"epoch": 0.6983277591973244,
"grad_norm": 0.38258033990859985,
"learning_rate": 7.739756866276452e-05,
"loss": 1.9838,
"step": 522
},
{
"epoch": 0.6996655518394649,
"grad_norm": 0.7790876626968384,
"learning_rate": 7.735254389914454e-05,
"loss": 2.1705,
"step": 523
},
{
"epoch": 0.7010033444816054,
"grad_norm": 0.3393210768699646,
"learning_rate": 7.730751913552454e-05,
"loss": 1.949,
"step": 524
},
{
"epoch": 0.7023411371237458,
"grad_norm": 0.3894656300544739,
"learning_rate": 7.726249437190454e-05,
"loss": 1.7554,
"step": 525
},
{
"epoch": 0.7036789297658863,
"grad_norm": 0.44569316506385803,
"learning_rate": 7.721746960828457e-05,
"loss": 1.9891,
"step": 526
},
{
"epoch": 0.7050167224080267,
"grad_norm": 0.40504804253578186,
"learning_rate": 7.717244484466458e-05,
"loss": 1.7219,
"step": 527
},
{
"epoch": 0.7063545150501672,
"grad_norm": 0.40453100204467773,
"learning_rate": 7.712742008104457e-05,
"loss": 1.8641,
"step": 528
},
{
"epoch": 0.7076923076923077,
"grad_norm": 0.36321282386779785,
"learning_rate": 7.708239531742458e-05,
"loss": 1.9582,
"step": 529
},
{
"epoch": 0.7090301003344481,
"grad_norm": 0.588949978351593,
"learning_rate": 7.70373705538046e-05,
"loss": 1.4981,
"step": 530
},
{
"epoch": 0.7103678929765886,
"grad_norm": 0.6625288128852844,
"learning_rate": 7.699234579018461e-05,
"loss": 1.5321,
"step": 531
},
{
"epoch": 0.711705685618729,
"grad_norm": 0.3740290403366089,
"learning_rate": 7.69473210265646e-05,
"loss": 1.9171,
"step": 532
},
{
"epoch": 0.7130434782608696,
"grad_norm": 0.30846625566482544,
"learning_rate": 7.690229626294463e-05,
"loss": 2.1682,
"step": 533
},
{
"epoch": 0.7143812709030101,
"grad_norm": 0.34462276101112366,
"learning_rate": 7.685727149932464e-05,
"loss": 2.2179,
"step": 534
},
{
"epoch": 0.7157190635451505,
"grad_norm": 0.3851476013660431,
"learning_rate": 7.681224673570464e-05,
"loss": 1.9401,
"step": 535
},
{
"epoch": 0.717056856187291,
"grad_norm": 0.3371794521808624,
"learning_rate": 7.676722197208465e-05,
"loss": 1.8142,
"step": 536
},
{
"epoch": 0.7183946488294315,
"grad_norm": 0.34345126152038574,
"learning_rate": 7.672219720846466e-05,
"loss": 1.8962,
"step": 537
},
{
"epoch": 0.7197324414715719,
"grad_norm": 0.3379805386066437,
"learning_rate": 7.667717244484467e-05,
"loss": 2.1848,
"step": 538
},
{
"epoch": 0.7210702341137124,
"grad_norm": 0.3815709054470062,
"learning_rate": 7.663214768122468e-05,
"loss": 1.4092,
"step": 539
},
{
"epoch": 0.7224080267558528,
"grad_norm": 0.3566444516181946,
"learning_rate": 7.658712291760469e-05,
"loss": 1.9942,
"step": 540
},
{
"epoch": 0.7237458193979933,
"grad_norm": 0.42823100090026855,
"learning_rate": 7.65420981539847e-05,
"loss": 1.9884,
"step": 541
},
{
"epoch": 0.7250836120401338,
"grad_norm": 0.5952591896057129,
"learning_rate": 7.64970733903647e-05,
"loss": 1.6506,
"step": 542
},
{
"epoch": 0.7264214046822742,
"grad_norm": 0.321796715259552,
"learning_rate": 7.645204862674471e-05,
"loss": 1.8377,
"step": 543
},
{
"epoch": 0.7277591973244147,
"grad_norm": 0.3382924497127533,
"learning_rate": 7.640702386312472e-05,
"loss": 1.8229,
"step": 544
},
{
"epoch": 0.7290969899665551,
"grad_norm": 0.6358975768089294,
"learning_rate": 7.636199909950473e-05,
"loss": 1.3559,
"step": 545
},
{
"epoch": 0.7304347826086957,
"grad_norm": 2.565281629562378,
"learning_rate": 7.631697433588474e-05,
"loss": 2.0329,
"step": 546
},
{
"epoch": 0.7317725752508362,
"grad_norm": 0.3889695405960083,
"learning_rate": 7.627194957226475e-05,
"loss": 1.7389,
"step": 547
},
{
"epoch": 0.7331103678929766,
"grad_norm": 0.3816166818141937,
"learning_rate": 7.622692480864476e-05,
"loss": 2.3092,
"step": 548
},
{
"epoch": 0.7344481605351171,
"grad_norm": 0.46524176001548767,
"learning_rate": 7.618190004502476e-05,
"loss": 1.7616,
"step": 549
},
{
"epoch": 0.7357859531772575,
"grad_norm": 0.36591479182243347,
"learning_rate": 7.613687528140477e-05,
"loss": 2.0946,
"step": 550
},
{
"epoch": 0.737123745819398,
"grad_norm": 0.34079721570014954,
"learning_rate": 7.60918505177848e-05,
"loss": 2.0089,
"step": 551
},
{
"epoch": 0.7384615384615385,
"grad_norm": 0.3858109712600708,
"learning_rate": 7.604682575416479e-05,
"loss": 1.9751,
"step": 552
},
{
"epoch": 0.7397993311036789,
"grad_norm": 0.3411719799041748,
"learning_rate": 7.60018009905448e-05,
"loss": 0.8972,
"step": 553
},
{
"epoch": 0.7411371237458194,
"grad_norm": 0.3922092914581299,
"learning_rate": 7.595677622692481e-05,
"loss": 1.2931,
"step": 554
},
{
"epoch": 0.7424749163879598,
"grad_norm": 0.27630677819252014,
"learning_rate": 7.591175146330483e-05,
"loss": 1.8692,
"step": 555
},
{
"epoch": 0.7438127090301003,
"grad_norm": 0.3395419716835022,
"learning_rate": 7.586672669968482e-05,
"loss": 1.967,
"step": 556
},
{
"epoch": 0.7451505016722408,
"grad_norm": 0.34704309701919556,
"learning_rate": 7.582170193606483e-05,
"loss": 1.7707,
"step": 557
},
{
"epoch": 0.7464882943143812,
"grad_norm": 0.7625495791435242,
"learning_rate": 7.577667717244484e-05,
"loss": 1.336,
"step": 558
},
{
"epoch": 0.7478260869565218,
"grad_norm": 0.33455583453178406,
"learning_rate": 7.573165240882486e-05,
"loss": 1.7007,
"step": 559
},
{
"epoch": 0.7491638795986622,
"grad_norm": 0.3844684064388275,
"learning_rate": 7.568662764520487e-05,
"loss": 2.0887,
"step": 560
},
{
"epoch": 0.7505016722408027,
"grad_norm": 0.4520154595375061,
"learning_rate": 7.564160288158487e-05,
"loss": 1.5324,
"step": 561
},
{
"epoch": 0.7518394648829432,
"grad_norm": 0.2970287799835205,
"learning_rate": 7.559657811796489e-05,
"loss": 1.842,
"step": 562
},
{
"epoch": 0.7531772575250836,
"grad_norm": 0.6159781813621521,
"learning_rate": 7.55515533543449e-05,
"loss": 1.9882,
"step": 563
},
{
"epoch": 0.7545150501672241,
"grad_norm": 0.3633442521095276,
"learning_rate": 7.550652859072491e-05,
"loss": 1.7449,
"step": 564
},
{
"epoch": 0.7558528428093646,
"grad_norm": 0.33036932349205017,
"learning_rate": 7.54615038271049e-05,
"loss": 1.4392,
"step": 565
},
{
"epoch": 0.757190635451505,
"grad_norm": 0.3699447214603424,
"learning_rate": 7.541647906348492e-05,
"loss": 1.9728,
"step": 566
},
{
"epoch": 0.7585284280936455,
"grad_norm": 0.4786345362663269,
"learning_rate": 7.537145429986493e-05,
"loss": 1.3553,
"step": 567
},
{
"epoch": 0.7598662207357859,
"grad_norm": 0.3800508677959442,
"learning_rate": 7.532642953624494e-05,
"loss": 1.6109,
"step": 568
},
{
"epoch": 0.7612040133779264,
"grad_norm": 0.3128887414932251,
"learning_rate": 7.528140477262495e-05,
"loss": 1.5856,
"step": 569
},
{
"epoch": 0.7625418060200669,
"grad_norm": 0.35543105006217957,
"learning_rate": 7.523638000900496e-05,
"loss": 1.8994,
"step": 570
},
{
"epoch": 0.7638795986622073,
"grad_norm": 0.33077099919319153,
"learning_rate": 7.519135524538497e-05,
"loss": 1.6525,
"step": 571
},
{
"epoch": 0.7652173913043478,
"grad_norm": 0.3991314470767975,
"learning_rate": 7.514633048176498e-05,
"loss": 1.9248,
"step": 572
},
{
"epoch": 0.7665551839464882,
"grad_norm": 0.36139121651649475,
"learning_rate": 7.510130571814498e-05,
"loss": 2.0326,
"step": 573
},
{
"epoch": 0.7678929765886288,
"grad_norm": 0.35255613923072815,
"learning_rate": 7.505628095452499e-05,
"loss": 1.8889,
"step": 574
},
{
"epoch": 0.7692307692307693,
"grad_norm": 0.3487282693386078,
"learning_rate": 7.5011256190905e-05,
"loss": 1.4981,
"step": 575
},
{
"epoch": 0.7705685618729097,
"grad_norm": 0.6345880627632141,
"learning_rate": 7.496623142728501e-05,
"loss": 2.1821,
"step": 576
},
{
"epoch": 0.7719063545150502,
"grad_norm": 0.5673653483390808,
"learning_rate": 7.492120666366502e-05,
"loss": 2.1171,
"step": 577
},
{
"epoch": 0.7732441471571906,
"grad_norm": 0.290094256401062,
"learning_rate": 7.487618190004503e-05,
"loss": 1.8362,
"step": 578
},
{
"epoch": 0.7745819397993311,
"grad_norm": 0.33212703466415405,
"learning_rate": 7.483115713642504e-05,
"loss": 1.9094,
"step": 579
},
{
"epoch": 0.7759197324414716,
"grad_norm": 0.6631553769111633,
"learning_rate": 7.478613237280504e-05,
"loss": 2.1565,
"step": 580
},
{
"epoch": 0.777257525083612,
"grad_norm": 0.33475595712661743,
"learning_rate": 7.474110760918505e-05,
"loss": 1.4291,
"step": 581
},
{
"epoch": 0.7785953177257525,
"grad_norm": 0.3770703673362732,
"learning_rate": 7.469608284556506e-05,
"loss": 2.0621,
"step": 582
},
{
"epoch": 0.7799331103678929,
"grad_norm": 0.7004806995391846,
"learning_rate": 7.465105808194507e-05,
"loss": 2.0331,
"step": 583
},
{
"epoch": 0.7812709030100334,
"grad_norm": 0.38309189677238464,
"learning_rate": 7.460603331832509e-05,
"loss": 1.8102,
"step": 584
},
{
"epoch": 0.782608695652174,
"grad_norm": 0.33871129155158997,
"learning_rate": 7.456100855470509e-05,
"loss": 2.1456,
"step": 585
},
{
"epoch": 0.7839464882943143,
"grad_norm": 0.4187791347503662,
"learning_rate": 7.45159837910851e-05,
"loss": 1.7323,
"step": 586
},
{
"epoch": 0.7852842809364549,
"grad_norm": 0.2464730590581894,
"learning_rate": 7.447095902746512e-05,
"loss": 1.9111,
"step": 587
},
{
"epoch": 0.7866220735785954,
"grad_norm": 0.37835320830345154,
"learning_rate": 7.442593426384513e-05,
"loss": 1.396,
"step": 588
},
{
"epoch": 0.7879598662207358,
"grad_norm": 0.363475501537323,
"learning_rate": 7.438090950022512e-05,
"loss": 1.9542,
"step": 589
},
{
"epoch": 0.7892976588628763,
"grad_norm": 0.37294596433639526,
"learning_rate": 7.433588473660513e-05,
"loss": 1.4258,
"step": 590
},
{
"epoch": 0.7906354515050167,
"grad_norm": 0.3721500039100647,
"learning_rate": 7.429085997298515e-05,
"loss": 1.987,
"step": 591
},
{
"epoch": 0.7919732441471572,
"grad_norm": 0.32486671209335327,
"learning_rate": 7.424583520936516e-05,
"loss": 1.8621,
"step": 592
},
{
"epoch": 0.7933110367892977,
"grad_norm": 0.3256564736366272,
"learning_rate": 7.420081044574516e-05,
"loss": 1.7656,
"step": 593
},
{
"epoch": 0.7946488294314381,
"grad_norm": 0.3214690089225769,
"learning_rate": 7.415578568212516e-05,
"loss": 2.0624,
"step": 594
},
{
"epoch": 0.7959866220735786,
"grad_norm": 0.6959064602851868,
"learning_rate": 7.411076091850519e-05,
"loss": 1.1767,
"step": 595
},
{
"epoch": 0.797324414715719,
"grad_norm": 0.36852458119392395,
"learning_rate": 7.40657361548852e-05,
"loss": 2.0539,
"step": 596
},
{
"epoch": 0.7986622073578595,
"grad_norm": 0.2969950735569,
"learning_rate": 7.402071139126519e-05,
"loss": 1.5776,
"step": 597
},
{
"epoch": 0.8,
"grad_norm": 0.450792521238327,
"learning_rate": 7.397568662764521e-05,
"loss": 1.5855,
"step": 598
},
{
"epoch": 0.8013377926421404,
"grad_norm": 0.5402222871780396,
"learning_rate": 7.393066186402522e-05,
"loss": 1.8693,
"step": 599
},
{
"epoch": 0.802675585284281,
"grad_norm": 0.406260222196579,
"learning_rate": 7.388563710040523e-05,
"loss": 1.8339,
"step": 600
},
{
"epoch": 0.8040133779264214,
"grad_norm": 0.3280527591705322,
"learning_rate": 7.384061233678522e-05,
"loss": 1.469,
"step": 601
},
{
"epoch": 0.8053511705685619,
"grad_norm": 0.3896856904029846,
"learning_rate": 7.379558757316525e-05,
"loss": 1.7457,
"step": 602
},
{
"epoch": 0.8066889632107024,
"grad_norm": 0.3198883831501007,
"learning_rate": 7.375056280954526e-05,
"loss": 1.8016,
"step": 603
},
{
"epoch": 0.8080267558528428,
"grad_norm": 0.3065289258956909,
"learning_rate": 7.370553804592526e-05,
"loss": 2.0356,
"step": 604
},
{
"epoch": 0.8093645484949833,
"grad_norm": 0.37292978167533875,
"learning_rate": 7.366051328230527e-05,
"loss": 1.8988,
"step": 605
},
{
"epoch": 0.8107023411371237,
"grad_norm": 0.3623843193054199,
"learning_rate": 7.361548851868528e-05,
"loss": 1.5998,
"step": 606
},
{
"epoch": 0.8120401337792642,
"grad_norm": 0.43376150727272034,
"learning_rate": 7.357046375506529e-05,
"loss": 1.4865,
"step": 607
},
{
"epoch": 0.8133779264214047,
"grad_norm": 0.3577169179916382,
"learning_rate": 7.35254389914453e-05,
"loss": 1.8154,
"step": 608
},
{
"epoch": 0.8147157190635451,
"grad_norm": 0.33039671182632446,
"learning_rate": 7.348041422782531e-05,
"loss": 1.9667,
"step": 609
},
{
"epoch": 0.8160535117056856,
"grad_norm": 0.3932124674320221,
"learning_rate": 7.343538946420532e-05,
"loss": 1.7125,
"step": 610
},
{
"epoch": 0.8173913043478261,
"grad_norm": 0.3652454912662506,
"learning_rate": 7.339036470058532e-05,
"loss": 1.4416,
"step": 611
},
{
"epoch": 0.8187290969899665,
"grad_norm": 0.6105794310569763,
"learning_rate": 7.334533993696533e-05,
"loss": 1.7701,
"step": 612
},
{
"epoch": 0.820066889632107,
"grad_norm": 0.3876180946826935,
"learning_rate": 7.330031517334534e-05,
"loss": 1.6549,
"step": 613
},
{
"epoch": 0.8214046822742475,
"grad_norm": 0.29518094658851624,
"learning_rate": 7.325529040972535e-05,
"loss": 1.8002,
"step": 614
},
{
"epoch": 0.822742474916388,
"grad_norm": 0.7063753604888916,
"learning_rate": 7.321026564610536e-05,
"loss": 1.6917,
"step": 615
},
{
"epoch": 0.8240802675585285,
"grad_norm": 0.5985257625579834,
"learning_rate": 7.316524088248538e-05,
"loss": 1.9754,
"step": 616
},
{
"epoch": 0.8254180602006689,
"grad_norm": 0.36610302329063416,
"learning_rate": 7.312021611886538e-05,
"loss": 2.1499,
"step": 617
},
{
"epoch": 0.8267558528428094,
"grad_norm": 0.32885974645614624,
"learning_rate": 7.307519135524538e-05,
"loss": 1.8947,
"step": 618
},
{
"epoch": 0.8280936454849498,
"grad_norm": 0.449900358915329,
"learning_rate": 7.303016659162539e-05,
"loss": 1.257,
"step": 619
},
{
"epoch": 0.8294314381270903,
"grad_norm": 0.37102293968200684,
"learning_rate": 7.298514182800542e-05,
"loss": 1.5899,
"step": 620
},
{
"epoch": 0.8307692307692308,
"grad_norm": 0.31158074736595154,
"learning_rate": 7.294011706438541e-05,
"loss": 1.7936,
"step": 621
},
{
"epoch": 0.8321070234113712,
"grad_norm": 0.3557308614253998,
"learning_rate": 7.289509230076542e-05,
"loss": 1.6656,
"step": 622
},
{
"epoch": 0.8334448160535117,
"grad_norm": 0.7624556422233582,
"learning_rate": 7.285006753714544e-05,
"loss": 1.8276,
"step": 623
},
{
"epoch": 0.8347826086956521,
"grad_norm": 0.3486712872982025,
"learning_rate": 7.280504277352545e-05,
"loss": 2.2463,
"step": 624
},
{
"epoch": 0.8361204013377926,
"grad_norm": 0.5724909901618958,
"learning_rate": 7.276001800990544e-05,
"loss": 1.8355,
"step": 625
},
{
"epoch": 0.8374581939799332,
"grad_norm": 0.8517188429832458,
"learning_rate": 7.271499324628545e-05,
"loss": 0.8871,
"step": 626
},
{
"epoch": 0.8387959866220736,
"grad_norm": 0.4913051128387451,
"learning_rate": 7.266996848266548e-05,
"loss": 1.856,
"step": 627
},
{
"epoch": 0.8401337792642141,
"grad_norm": 0.45800983905792236,
"learning_rate": 7.262494371904548e-05,
"loss": 1.4221,
"step": 628
},
{
"epoch": 0.8414715719063545,
"grad_norm": 0.6228715777397156,
"learning_rate": 7.257991895542548e-05,
"loss": 2.224,
"step": 629
},
{
"epoch": 0.842809364548495,
"grad_norm": 0.5099749565124512,
"learning_rate": 7.253489419180549e-05,
"loss": 1.702,
"step": 630
},
{
"epoch": 0.8441471571906355,
"grad_norm": 0.39042428135871887,
"learning_rate": 7.248986942818551e-05,
"loss": 1.5742,
"step": 631
},
{
"epoch": 0.8454849498327759,
"grad_norm": 0.7996957898139954,
"learning_rate": 7.244484466456552e-05,
"loss": 1.328,
"step": 632
},
{
"epoch": 0.8468227424749164,
"grad_norm": 0.35331302881240845,
"learning_rate": 7.239981990094553e-05,
"loss": 1.6865,
"step": 633
},
{
"epoch": 0.8481605351170568,
"grad_norm": 0.4120136499404907,
"learning_rate": 7.235479513732554e-05,
"loss": 1.684,
"step": 634
},
{
"epoch": 0.8494983277591973,
"grad_norm": 0.4208261966705322,
"learning_rate": 7.230977037370554e-05,
"loss": 1.8001,
"step": 635
},
{
"epoch": 0.8508361204013378,
"grad_norm": 0.41124269366264343,
"learning_rate": 7.226474561008555e-05,
"loss": 1.7426,
"step": 636
},
{
"epoch": 0.8521739130434782,
"grad_norm": 0.5735759139060974,
"learning_rate": 7.221972084646556e-05,
"loss": 1.6861,
"step": 637
},
{
"epoch": 0.8535117056856187,
"grad_norm": 0.3232871890068054,
"learning_rate": 7.217469608284557e-05,
"loss": 1.9313,
"step": 638
},
{
"epoch": 0.8548494983277592,
"grad_norm": 0.315378338098526,
"learning_rate": 7.212967131922558e-05,
"loss": 1.7994,
"step": 639
},
{
"epoch": 0.8561872909698997,
"grad_norm": 0.5437686443328857,
"learning_rate": 7.208464655560559e-05,
"loss": 1.761,
"step": 640
},
{
"epoch": 0.8575250836120402,
"grad_norm": 0.4012024998664856,
"learning_rate": 7.20396217919856e-05,
"loss": 1.8082,
"step": 641
},
{
"epoch": 0.8588628762541806,
"grad_norm": 0.35168206691741943,
"learning_rate": 7.19945970283656e-05,
"loss": 1.8141,
"step": 642
},
{
"epoch": 0.8602006688963211,
"grad_norm": 0.29326000809669495,
"learning_rate": 7.194957226474561e-05,
"loss": 2.0852,
"step": 643
},
{
"epoch": 0.8615384615384616,
"grad_norm": 0.5200529098510742,
"learning_rate": 7.190454750112562e-05,
"loss": 1.7179,
"step": 644
},
{
"epoch": 0.862876254180602,
"grad_norm": 0.419899046421051,
"learning_rate": 7.185952273750563e-05,
"loss": 1.8952,
"step": 645
},
{
"epoch": 0.8642140468227425,
"grad_norm": 0.3791028559207916,
"learning_rate": 7.181449797388564e-05,
"loss": 1.4306,
"step": 646
},
{
"epoch": 0.8655518394648829,
"grad_norm": 0.3782893121242523,
"learning_rate": 7.176947321026565e-05,
"loss": 1.8628,
"step": 647
},
{
"epoch": 0.8668896321070234,
"grad_norm": 0.3860854208469391,
"learning_rate": 7.172444844664566e-05,
"loss": 1.6255,
"step": 648
},
{
"epoch": 0.8682274247491639,
"grad_norm": 0.361833393573761,
"learning_rate": 7.167942368302566e-05,
"loss": 1.7462,
"step": 649
},
{
"epoch": 0.8695652173913043,
"grad_norm": 0.400254487991333,
"learning_rate": 7.163439891940567e-05,
"loss": 1.3854,
"step": 650
},
{
"epoch": 0.8709030100334448,
"grad_norm": 0.34192296862602234,
"learning_rate": 7.158937415578568e-05,
"loss": 2.0707,
"step": 651
},
{
"epoch": 0.8722408026755852,
"grad_norm": 0.5104194283485413,
"learning_rate": 7.15443493921657e-05,
"loss": 1.6693,
"step": 652
},
{
"epoch": 0.8735785953177257,
"grad_norm": 0.3650829792022705,
"learning_rate": 7.149932462854571e-05,
"loss": 0.9515,
"step": 653
},
{
"epoch": 0.8749163879598663,
"grad_norm": 0.42598962783813477,
"learning_rate": 7.145429986492571e-05,
"loss": 1.6728,
"step": 654
},
{
"epoch": 0.8762541806020067,
"grad_norm": 0.4158768057823181,
"learning_rate": 7.140927510130572e-05,
"loss": 1.2007,
"step": 655
},
{
"epoch": 0.8775919732441472,
"grad_norm": 0.34016454219818115,
"learning_rate": 7.136425033768574e-05,
"loss": 2.1394,
"step": 656
},
{
"epoch": 0.8789297658862876,
"grad_norm": 0.34736478328704834,
"learning_rate": 7.131922557406575e-05,
"loss": 1.566,
"step": 657
},
{
"epoch": 0.8802675585284281,
"grad_norm": 0.47918611764907837,
"learning_rate": 7.127420081044574e-05,
"loss": 1.7246,
"step": 658
},
{
"epoch": 0.8816053511705686,
"grad_norm": 0.41045960783958435,
"learning_rate": 7.122917604682576e-05,
"loss": 1.7283,
"step": 659
},
{
"epoch": 0.882943143812709,
"grad_norm": 0.3634040057659149,
"learning_rate": 7.118415128320577e-05,
"loss": 1.6028,
"step": 660
},
{
"epoch": 0.8842809364548495,
"grad_norm": 0.3791142404079437,
"learning_rate": 7.113912651958578e-05,
"loss": 1.7992,
"step": 661
},
{
"epoch": 0.88561872909699,
"grad_norm": 0.356406569480896,
"learning_rate": 7.109410175596578e-05,
"loss": 1.7343,
"step": 662
},
{
"epoch": 0.8869565217391304,
"grad_norm": 0.36963218450546265,
"learning_rate": 7.10490769923458e-05,
"loss": 1.8745,
"step": 663
},
{
"epoch": 0.8882943143812709,
"grad_norm": 0.5904031991958618,
"learning_rate": 7.10040522287258e-05,
"loss": 1.6621,
"step": 664
},
{
"epoch": 0.8896321070234113,
"grad_norm": 0.40875113010406494,
"learning_rate": 7.095902746510582e-05,
"loss": 1.826,
"step": 665
},
{
"epoch": 0.8909698996655518,
"grad_norm": 0.6849672198295593,
"learning_rate": 7.091400270148581e-05,
"loss": 1.7498,
"step": 666
},
{
"epoch": 0.8923076923076924,
"grad_norm": 0.8226297497749329,
"learning_rate": 7.086897793786583e-05,
"loss": 1.8779,
"step": 667
},
{
"epoch": 0.8936454849498328,
"grad_norm": 0.39494749903678894,
"learning_rate": 7.082395317424584e-05,
"loss": 1.3091,
"step": 668
},
{
"epoch": 0.8949832775919733,
"grad_norm": 0.36774224042892456,
"learning_rate": 7.077892841062585e-05,
"loss": 2.0947,
"step": 669
},
{
"epoch": 0.8963210702341137,
"grad_norm": 0.38955172896385193,
"learning_rate": 7.073390364700586e-05,
"loss": 1.8579,
"step": 670
},
{
"epoch": 0.8976588628762542,
"grad_norm": 0.3742164671421051,
"learning_rate": 7.068887888338587e-05,
"loss": 1.6201,
"step": 671
},
{
"epoch": 0.8989966555183947,
"grad_norm": 0.3513646721839905,
"learning_rate": 7.064385411976588e-05,
"loss": 1.9323,
"step": 672
},
{
"epoch": 0.9003344481605351,
"grad_norm": 0.3952341079711914,
"learning_rate": 7.059882935614588e-05,
"loss": 1.6883,
"step": 673
},
{
"epoch": 0.9016722408026756,
"grad_norm": 0.42206764221191406,
"learning_rate": 7.055380459252589e-05,
"loss": 1.9177,
"step": 674
},
{
"epoch": 0.903010033444816,
"grad_norm": 0.36615172028541565,
"learning_rate": 7.05087798289059e-05,
"loss": 1.7579,
"step": 675
},
{
"epoch": 0.9043478260869565,
"grad_norm": 0.3826829791069031,
"learning_rate": 7.046375506528591e-05,
"loss": 2.0081,
"step": 676
},
{
"epoch": 0.905685618729097,
"grad_norm": 0.3107741177082062,
"learning_rate": 7.041873030166593e-05,
"loss": 2.1041,
"step": 677
},
{
"epoch": 0.9070234113712374,
"grad_norm": 0.33083420991897583,
"learning_rate": 7.037370553804593e-05,
"loss": 1.8097,
"step": 678
},
{
"epoch": 0.9083612040133779,
"grad_norm": 0.3041648864746094,
"learning_rate": 7.032868077442594e-05,
"loss": 1.0782,
"step": 679
},
{
"epoch": 0.9096989966555183,
"grad_norm": 0.5441336035728455,
"learning_rate": 7.028365601080594e-05,
"loss": 1.6318,
"step": 680
},
{
"epoch": 0.9110367892976589,
"grad_norm": 0.363393634557724,
"learning_rate": 7.023863124718597e-05,
"loss": 1.7169,
"step": 681
},
{
"epoch": 0.9123745819397994,
"grad_norm": 0.6789020299911499,
"learning_rate": 7.019360648356596e-05,
"loss": 1.9805,
"step": 682
},
{
"epoch": 0.9137123745819398,
"grad_norm": 0.3467765748500824,
"learning_rate": 7.014858171994597e-05,
"loss": 1.8415,
"step": 683
},
{
"epoch": 0.9150501672240803,
"grad_norm": 0.7270785570144653,
"learning_rate": 7.010355695632598e-05,
"loss": 1.3206,
"step": 684
},
{
"epoch": 0.9163879598662207,
"grad_norm": 0.5416232943534851,
"learning_rate": 7.0058532192706e-05,
"loss": 1.8754,
"step": 685
},
{
"epoch": 0.9177257525083612,
"grad_norm": 0.3280985653400421,
"learning_rate": 7.0013507429086e-05,
"loss": 1.8428,
"step": 686
},
{
"epoch": 0.9190635451505017,
"grad_norm": 0.3461412787437439,
"learning_rate": 6.9968482665466e-05,
"loss": 1.5161,
"step": 687
},
{
"epoch": 0.9204013377926421,
"grad_norm": 0.3746154308319092,
"learning_rate": 6.992345790184603e-05,
"loss": 1.8658,
"step": 688
},
{
"epoch": 0.9217391304347826,
"grad_norm": 0.3138766288757324,
"learning_rate": 6.987843313822603e-05,
"loss": 1.4987,
"step": 689
},
{
"epoch": 0.9230769230769231,
"grad_norm": 0.34391963481903076,
"learning_rate": 6.983340837460603e-05,
"loss": 1.7672,
"step": 690
},
{
"epoch": 0.9244147157190635,
"grad_norm": 0.7809226512908936,
"learning_rate": 6.978838361098604e-05,
"loss": 1.6137,
"step": 691
},
{
"epoch": 0.925752508361204,
"grad_norm": 0.3945732116699219,
"learning_rate": 6.974335884736606e-05,
"loss": 1.7833,
"step": 692
},
{
"epoch": 0.9270903010033444,
"grad_norm": 0.41453108191490173,
"learning_rate": 6.969833408374607e-05,
"loss": 1.5949,
"step": 693
},
{
"epoch": 0.928428093645485,
"grad_norm": 0.3755939304828644,
"learning_rate": 6.965330932012606e-05,
"loss": 2.3068,
"step": 694
},
{
"epoch": 0.9297658862876255,
"grad_norm": 0.345302015542984,
"learning_rate": 6.960828455650609e-05,
"loss": 1.7721,
"step": 695
},
{
"epoch": 0.9311036789297659,
"grad_norm": 0.3220181465148926,
"learning_rate": 6.95632597928861e-05,
"loss": 1.6998,
"step": 696
},
{
"epoch": 0.9324414715719064,
"grad_norm": 0.366615891456604,
"learning_rate": 6.95182350292661e-05,
"loss": 2.0303,
"step": 697
},
{
"epoch": 0.9337792642140468,
"grad_norm": 0.4134201407432556,
"learning_rate": 6.94732102656461e-05,
"loss": 1.3428,
"step": 698
},
{
"epoch": 0.9351170568561873,
"grad_norm": 0.38845616579055786,
"learning_rate": 6.942818550202612e-05,
"loss": 2.0773,
"step": 699
},
{
"epoch": 0.9364548494983278,
"grad_norm": 0.4948274493217468,
"learning_rate": 6.938316073840613e-05,
"loss": 1.8458,
"step": 700
},
{
"epoch": 0.9377926421404682,
"grad_norm": 0.43470853567123413,
"learning_rate": 6.933813597478614e-05,
"loss": 1.9182,
"step": 701
},
{
"epoch": 0.9391304347826087,
"grad_norm": 0.503555417060852,
"learning_rate": 6.929311121116615e-05,
"loss": 2.1079,
"step": 702
},
{
"epoch": 0.9404682274247491,
"grad_norm": 0.31471186876296997,
"learning_rate": 6.924808644754615e-05,
"loss": 1.8166,
"step": 703
},
{
"epoch": 0.9418060200668896,
"grad_norm": 0.5391733646392822,
"learning_rate": 6.920306168392616e-05,
"loss": 1.6998,
"step": 704
},
{
"epoch": 0.9431438127090301,
"grad_norm": 0.3019644618034363,
"learning_rate": 6.915803692030617e-05,
"loss": 2.0563,
"step": 705
},
{
"epoch": 0.9444816053511705,
"grad_norm": 0.5476101040840149,
"learning_rate": 6.911301215668618e-05,
"loss": 1.5407,
"step": 706
},
{
"epoch": 0.945819397993311,
"grad_norm": 0.5264822244644165,
"learning_rate": 6.906798739306619e-05,
"loss": 1.7866,
"step": 707
},
{
"epoch": 0.9471571906354515,
"grad_norm": 0.2899741232395172,
"learning_rate": 6.90229626294462e-05,
"loss": 1.8482,
"step": 708
},
{
"epoch": 0.948494983277592,
"grad_norm": 0.45578619837760925,
"learning_rate": 6.89779378658262e-05,
"loss": 1.3086,
"step": 709
},
{
"epoch": 0.9498327759197325,
"grad_norm": 0.3755444586277008,
"learning_rate": 6.893291310220622e-05,
"loss": 1.9272,
"step": 710
},
{
"epoch": 0.9511705685618729,
"grad_norm": 0.5911352634429932,
"learning_rate": 6.888788833858622e-05,
"loss": 1.9175,
"step": 711
},
{
"epoch": 0.9525083612040134,
"grad_norm": 0.2892666757106781,
"learning_rate": 6.884286357496623e-05,
"loss": 1.3744,
"step": 712
},
{
"epoch": 0.9538461538461539,
"grad_norm": 0.4964825212955475,
"learning_rate": 6.879783881134625e-05,
"loss": 1.8149,
"step": 713
},
{
"epoch": 0.9551839464882943,
"grad_norm": 0.3442051410675049,
"learning_rate": 6.875281404772625e-05,
"loss": 1.9712,
"step": 714
},
{
"epoch": 0.9565217391304348,
"grad_norm": 0.5018149614334106,
"learning_rate": 6.870778928410626e-05,
"loss": 1.2699,
"step": 715
},
{
"epoch": 0.9578595317725752,
"grad_norm": 0.45120319724082947,
"learning_rate": 6.866276452048627e-05,
"loss": 1.871,
"step": 716
},
{
"epoch": 0.9591973244147157,
"grad_norm": 0.3946061134338379,
"learning_rate": 6.861773975686629e-05,
"loss": 1.749,
"step": 717
},
{
"epoch": 0.9605351170568562,
"grad_norm": 0.7921074628829956,
"learning_rate": 6.857271499324628e-05,
"loss": 1.9881,
"step": 718
},
{
"epoch": 0.9618729096989966,
"grad_norm": 0.29882094264030457,
"learning_rate": 6.852769022962629e-05,
"loss": 1.8386,
"step": 719
},
{
"epoch": 0.9632107023411371,
"grad_norm": 0.38971996307373047,
"learning_rate": 6.84826654660063e-05,
"loss": 1.8224,
"step": 720
},
{
"epoch": 0.9645484949832775,
"grad_norm": 0.435881644487381,
"learning_rate": 6.843764070238632e-05,
"loss": 1.5976,
"step": 721
},
{
"epoch": 0.9658862876254181,
"grad_norm": 0.3308292329311371,
"learning_rate": 6.839261593876632e-05,
"loss": 1.9759,
"step": 722
},
{
"epoch": 0.9672240802675586,
"grad_norm": 0.38128048181533813,
"learning_rate": 6.834759117514633e-05,
"loss": 2.0457,
"step": 723
},
{
"epoch": 0.968561872909699,
"grad_norm": 0.37816083431243896,
"learning_rate": 6.830256641152635e-05,
"loss": 1.7379,
"step": 724
},
{
"epoch": 0.9698996655518395,
"grad_norm": 0.3599185049533844,
"learning_rate": 6.825754164790636e-05,
"loss": 1.3529,
"step": 725
},
{
"epoch": 0.9712374581939799,
"grad_norm": 0.34246405959129333,
"learning_rate": 6.821251688428637e-05,
"loss": 2.2177,
"step": 726
},
{
"epoch": 0.9725752508361204,
"grad_norm": 0.36853083968162537,
"learning_rate": 6.816749212066636e-05,
"loss": 1.8784,
"step": 727
},
{
"epoch": 0.9739130434782609,
"grad_norm": 0.39494502544403076,
"learning_rate": 6.812246735704638e-05,
"loss": 2.0329,
"step": 728
},
{
"epoch": 0.9752508361204013,
"grad_norm": 1.9131022691726685,
"learning_rate": 6.807744259342639e-05,
"loss": 1.391,
"step": 729
},
{
"epoch": 0.9765886287625418,
"grad_norm": 0.4040629267692566,
"learning_rate": 6.80324178298064e-05,
"loss": 2.0042,
"step": 730
},
{
"epoch": 0.9779264214046822,
"grad_norm": 0.38302701711654663,
"learning_rate": 6.798739306618641e-05,
"loss": 2.1064,
"step": 731
},
{
"epoch": 0.9792642140468227,
"grad_norm": 0.3989606201648712,
"learning_rate": 6.794236830256642e-05,
"loss": 1.6208,
"step": 732
},
{
"epoch": 0.9806020066889632,
"grad_norm": 0.5049397945404053,
"learning_rate": 6.789734353894643e-05,
"loss": 1.9809,
"step": 733
},
{
"epoch": 0.9819397993311036,
"grad_norm": 0.5564031004905701,
"learning_rate": 6.785231877532643e-05,
"loss": 1.9769,
"step": 734
},
{
"epoch": 0.9832775919732442,
"grad_norm": 1.37478506565094,
"learning_rate": 6.780729401170644e-05,
"loss": 1.6798,
"step": 735
},
{
"epoch": 0.9846153846153847,
"grad_norm": 0.2812560498714447,
"learning_rate": 6.776226924808645e-05,
"loss": 1.4855,
"step": 736
},
{
"epoch": 0.9859531772575251,
"grad_norm": 0.5258631706237793,
"learning_rate": 6.771724448446646e-05,
"loss": 2.1563,
"step": 737
},
{
"epoch": 0.9872909698996656,
"grad_norm": 0.7295839190483093,
"learning_rate": 6.767221972084647e-05,
"loss": 2.0624,
"step": 738
},
{
"epoch": 0.988628762541806,
"grad_norm": 0.4331284463405609,
"learning_rate": 6.762719495722648e-05,
"loss": 1.5767,
"step": 739
},
{
"epoch": 0.9899665551839465,
"grad_norm": 0.38064879179000854,
"learning_rate": 6.758217019360649e-05,
"loss": 1.8441,
"step": 740
},
{
"epoch": 0.991304347826087,
"grad_norm": 0.3991696834564209,
"learning_rate": 6.75371454299865e-05,
"loss": 1.769,
"step": 741
},
{
"epoch": 0.9926421404682274,
"grad_norm": 0.8858866691589355,
"learning_rate": 6.74921206663665e-05,
"loss": 1.6091,
"step": 742
},
{
"epoch": 0.9939799331103679,
"grad_norm": 0.2829599380493164,
"learning_rate": 6.744709590274651e-05,
"loss": 1.4218,
"step": 743
},
{
"epoch": 0.9953177257525083,
"grad_norm": 0.3447091579437256,
"learning_rate": 6.740207113912652e-05,
"loss": 2.08,
"step": 744
},
{
"epoch": 0.9966555183946488,
"grad_norm": 0.5587038397789001,
"learning_rate": 6.735704637550653e-05,
"loss": 1.9443,
"step": 745
},
{
"epoch": 0.9979933110367893,
"grad_norm": 0.29927799105644226,
"learning_rate": 6.731202161188654e-05,
"loss": 2.1269,
"step": 746
},
{
"epoch": 0.9993311036789297,
"grad_norm": 0.3965597152709961,
"learning_rate": 6.726699684826655e-05,
"loss": 1.7172,
"step": 747
},
{
"epoch": 1.0006688963210701,
"grad_norm": 0.36948877573013306,
"learning_rate": 6.722197208464655e-05,
"loss": 1.3854,
"step": 748
},
{
"epoch": 1.0020066889632107,
"grad_norm": 0.36516493558883667,
"learning_rate": 6.717694732102658e-05,
"loss": 1.4223,
"step": 749
},
{
"epoch": 1.0033444816053512,
"grad_norm": 0.4343159794807434,
"learning_rate": 6.713192255740659e-05,
"loss": 1.8986,
"step": 750
},
{
"epoch": 1.0046822742474917,
"grad_norm": 0.4338577091693878,
"learning_rate": 6.708689779378658e-05,
"loss": 0.5871,
"step": 751
},
{
"epoch": 1.0060200668896322,
"grad_norm": 0.41223862767219543,
"learning_rate": 6.704187303016659e-05,
"loss": 1.5983,
"step": 752
},
{
"epoch": 1.0073578595317725,
"grad_norm": 0.7298925518989563,
"learning_rate": 6.699684826654661e-05,
"loss": 1.4322,
"step": 753
},
{
"epoch": 1.008695652173913,
"grad_norm": 1.1361639499664307,
"learning_rate": 6.695182350292662e-05,
"loss": 1.4163,
"step": 754
},
{
"epoch": 1.0100334448160535,
"grad_norm": 0.5715144276618958,
"learning_rate": 6.690679873930662e-05,
"loss": 1.498,
"step": 755
},
{
"epoch": 1.011371237458194,
"grad_norm": 0.6086817383766174,
"learning_rate": 6.686177397568662e-05,
"loss": 1.7753,
"step": 756
},
{
"epoch": 1.0127090301003345,
"grad_norm": 0.8653783202171326,
"learning_rate": 6.681674921206665e-05,
"loss": 1.5424,
"step": 757
},
{
"epoch": 1.0140468227424748,
"grad_norm": 0.5191454291343689,
"learning_rate": 6.677172444844665e-05,
"loss": 1.7054,
"step": 758
},
{
"epoch": 1.0153846153846153,
"grad_norm": 0.5097001791000366,
"learning_rate": 6.672669968482665e-05,
"loss": 1.4023,
"step": 759
},
{
"epoch": 1.0167224080267558,
"grad_norm": 0.5936307311058044,
"learning_rate": 6.668167492120667e-05,
"loss": 1.1395,
"step": 760
},
{
"epoch": 1.0180602006688964,
"grad_norm": 0.5675538778305054,
"learning_rate": 6.663665015758668e-05,
"loss": 1.3809,
"step": 761
},
{
"epoch": 1.0193979933110369,
"grad_norm": 0.407277911901474,
"learning_rate": 6.659162539396669e-05,
"loss": 1.7331,
"step": 762
},
{
"epoch": 1.0207357859531772,
"grad_norm": 1.353188157081604,
"learning_rate": 6.654660063034668e-05,
"loss": 1.409,
"step": 763
},
{
"epoch": 1.0220735785953177,
"grad_norm": 0.3952758312225342,
"learning_rate": 6.65015758667267e-05,
"loss": 1.6219,
"step": 764
},
{
"epoch": 1.0234113712374582,
"grad_norm": 0.6428177356719971,
"learning_rate": 6.645655110310671e-05,
"loss": 1.1543,
"step": 765
},
{
"epoch": 1.0247491638795987,
"grad_norm": 1.0934861898422241,
"learning_rate": 6.641152633948672e-05,
"loss": 0.9668,
"step": 766
},
{
"epoch": 1.0260869565217392,
"grad_norm": 0.4791591763496399,
"learning_rate": 6.636650157586673e-05,
"loss": 1.4271,
"step": 767
},
{
"epoch": 1.0274247491638795,
"grad_norm": 0.5504617094993591,
"learning_rate": 6.632147681224674e-05,
"loss": 1.4648,
"step": 768
},
{
"epoch": 1.02876254180602,
"grad_norm": 0.45765021443367004,
"learning_rate": 6.627645204862675e-05,
"loss": 1.3732,
"step": 769
},
{
"epoch": 1.0301003344481605,
"grad_norm": 0.46277210116386414,
"learning_rate": 6.623142728500676e-05,
"loss": 1.0407,
"step": 770
},
{
"epoch": 1.031438127090301,
"grad_norm": 0.5813681483268738,
"learning_rate": 6.618640252138677e-05,
"loss": 1.2612,
"step": 771
},
{
"epoch": 1.0327759197324415,
"grad_norm": 0.408017635345459,
"learning_rate": 6.614137775776677e-05,
"loss": 1.5765,
"step": 772
},
{
"epoch": 1.034113712374582,
"grad_norm": 0.6632124185562134,
"learning_rate": 6.609635299414678e-05,
"loss": 1.6674,
"step": 773
},
{
"epoch": 1.0354515050167223,
"grad_norm": 0.563066303730011,
"learning_rate": 6.605132823052679e-05,
"loss": 1.4702,
"step": 774
},
{
"epoch": 1.0367892976588629,
"grad_norm": 0.4824022948741913,
"learning_rate": 6.60063034669068e-05,
"loss": 1.4301,
"step": 775
},
{
"epoch": 1.0381270903010034,
"grad_norm": 0.5674872398376465,
"learning_rate": 6.596127870328681e-05,
"loss": 1.1038,
"step": 776
},
{
"epoch": 1.0394648829431439,
"grad_norm": 0.5925496816635132,
"learning_rate": 6.591625393966682e-05,
"loss": 1.3891,
"step": 777
},
{
"epoch": 1.0408026755852844,
"grad_norm": 0.45445653796195984,
"learning_rate": 6.587122917604684e-05,
"loss": 1.4653,
"step": 778
},
{
"epoch": 1.0421404682274247,
"grad_norm": 0.44357824325561523,
"learning_rate": 6.582620441242683e-05,
"loss": 1.4065,
"step": 779
},
{
"epoch": 1.0434782608695652,
"grad_norm": 0.4146457612514496,
"learning_rate": 6.578117964880684e-05,
"loss": 1.6831,
"step": 780
},
{
"epoch": 1.0448160535117057,
"grad_norm": 0.48196661472320557,
"learning_rate": 6.573615488518685e-05,
"loss": 1.5082,
"step": 781
},
{
"epoch": 1.0461538461538462,
"grad_norm": 0.4578336179256439,
"learning_rate": 6.569113012156687e-05,
"loss": 1.7227,
"step": 782
},
{
"epoch": 1.0474916387959867,
"grad_norm": 0.5198684334754944,
"learning_rate": 6.564610535794687e-05,
"loss": 1.1455,
"step": 783
},
{
"epoch": 1.048829431438127,
"grad_norm": 0.497288316488266,
"learning_rate": 6.560108059432688e-05,
"loss": 1.6872,
"step": 784
},
{
"epoch": 1.0501672240802675,
"grad_norm": 0.36458608508110046,
"learning_rate": 6.55560558307069e-05,
"loss": 2.2226,
"step": 785
},
{
"epoch": 1.051505016722408,
"grad_norm": 0.8122197985649109,
"learning_rate": 6.551103106708691e-05,
"loss": 1.3132,
"step": 786
},
{
"epoch": 1.0528428093645485,
"grad_norm": 0.4065374732017517,
"learning_rate": 6.54660063034669e-05,
"loss": 1.7364,
"step": 787
},
{
"epoch": 1.054180602006689,
"grad_norm": 0.5058987140655518,
"learning_rate": 6.542098153984691e-05,
"loss": 1.6037,
"step": 788
},
{
"epoch": 1.0555183946488294,
"grad_norm": 0.5093656778335571,
"learning_rate": 6.537595677622693e-05,
"loss": 1.4635,
"step": 789
},
{
"epoch": 1.0568561872909699,
"grad_norm": 0.4434761106967926,
"learning_rate": 6.533093201260694e-05,
"loss": 1.7289,
"step": 790
},
{
"epoch": 1.0581939799331104,
"grad_norm": 0.8123733997344971,
"learning_rate": 6.528590724898694e-05,
"loss": 1.3977,
"step": 791
},
{
"epoch": 1.0595317725752509,
"grad_norm": 0.419659286737442,
"learning_rate": 6.524088248536695e-05,
"loss": 1.9813,
"step": 792
},
{
"epoch": 1.0608695652173914,
"grad_norm": 0.4884895980358124,
"learning_rate": 6.519585772174697e-05,
"loss": 1.388,
"step": 793
},
{
"epoch": 1.0622073578595317,
"grad_norm": 0.6140890717506409,
"learning_rate": 6.515083295812698e-05,
"loss": 1.0599,
"step": 794
},
{
"epoch": 1.0635451505016722,
"grad_norm": 0.45217812061309814,
"learning_rate": 6.510580819450697e-05,
"loss": 1.5021,
"step": 795
},
{
"epoch": 1.0648829431438127,
"grad_norm": 0.45027461647987366,
"learning_rate": 6.5060783430887e-05,
"loss": 1.4057,
"step": 796
},
{
"epoch": 1.0662207357859532,
"grad_norm": 0.6308950185775757,
"learning_rate": 6.5015758667267e-05,
"loss": 1.3602,
"step": 797
},
{
"epoch": 1.0675585284280937,
"grad_norm": 1.535835862159729,
"learning_rate": 6.497073390364701e-05,
"loss": 1.6195,
"step": 798
},
{
"epoch": 1.068896321070234,
"grad_norm": 0.4812445640563965,
"learning_rate": 6.492570914002702e-05,
"loss": 1.6331,
"step": 799
},
{
"epoch": 1.0702341137123745,
"grad_norm": 0.5334980487823486,
"learning_rate": 6.488068437640703e-05,
"loss": 1.1138,
"step": 800
},
{
"epoch": 1.071571906354515,
"grad_norm": 0.6437579989433289,
"learning_rate": 6.483565961278704e-05,
"loss": 1.4453,
"step": 801
},
{
"epoch": 1.0729096989966556,
"grad_norm": 1.0662591457366943,
"learning_rate": 6.479063484916705e-05,
"loss": 0.603,
"step": 802
},
{
"epoch": 1.074247491638796,
"grad_norm": 0.6081445217132568,
"learning_rate": 6.474561008554705e-05,
"loss": 1.4157,
"step": 803
},
{
"epoch": 1.0755852842809364,
"grad_norm": 0.5122383832931519,
"learning_rate": 6.470058532192706e-05,
"loss": 1.6912,
"step": 804
},
{
"epoch": 1.0769230769230769,
"grad_norm": 0.5406622290611267,
"learning_rate": 6.465556055830707e-05,
"loss": 0.8931,
"step": 805
},
{
"epoch": 1.0782608695652174,
"grad_norm": 0.6800340414047241,
"learning_rate": 6.461053579468708e-05,
"loss": 1.2054,
"step": 806
},
{
"epoch": 1.079598662207358,
"grad_norm": 0.891855001449585,
"learning_rate": 6.456551103106709e-05,
"loss": 1.2156,
"step": 807
},
{
"epoch": 1.0809364548494984,
"grad_norm": 0.4451705515384674,
"learning_rate": 6.45204862674471e-05,
"loss": 1.7379,
"step": 808
},
{
"epoch": 1.0822742474916387,
"grad_norm": 1.2329106330871582,
"learning_rate": 6.44754615038271e-05,
"loss": 1.0769,
"step": 809
},
{
"epoch": 1.0836120401337792,
"grad_norm": 0.46149975061416626,
"learning_rate": 6.443043674020711e-05,
"loss": 1.7287,
"step": 810
},
{
"epoch": 1.0849498327759197,
"grad_norm": 0.4032789170742035,
"learning_rate": 6.438541197658712e-05,
"loss": 1.6798,
"step": 811
},
{
"epoch": 1.0862876254180602,
"grad_norm": 0.5288923978805542,
"learning_rate": 6.434038721296713e-05,
"loss": 1.3158,
"step": 812
},
{
"epoch": 1.0876254180602007,
"grad_norm": 0.4517413377761841,
"learning_rate": 6.429536244934714e-05,
"loss": 1.7064,
"step": 813
},
{
"epoch": 1.088963210702341,
"grad_norm": 0.5029974579811096,
"learning_rate": 6.425033768572716e-05,
"loss": 1.0985,
"step": 814
},
{
"epoch": 1.0903010033444815,
"grad_norm": 0.649246096611023,
"learning_rate": 6.420531292210716e-05,
"loss": 1.3266,
"step": 815
},
{
"epoch": 1.091638795986622,
"grad_norm": 0.8336004614830017,
"learning_rate": 6.416028815848717e-05,
"loss": 0.9194,
"step": 816
},
{
"epoch": 1.0929765886287626,
"grad_norm": 0.3941144347190857,
"learning_rate": 6.411526339486717e-05,
"loss": 1.7471,
"step": 817
},
{
"epoch": 1.094314381270903,
"grad_norm": 0.38524457812309265,
"learning_rate": 6.40702386312472e-05,
"loss": 1.3742,
"step": 818
},
{
"epoch": 1.0956521739130434,
"grad_norm": 0.45431381464004517,
"learning_rate": 6.402521386762719e-05,
"loss": 0.9257,
"step": 819
},
{
"epoch": 1.0969899665551839,
"grad_norm": 0.5474141240119934,
"learning_rate": 6.39801891040072e-05,
"loss": 1.437,
"step": 820
},
{
"epoch": 1.0983277591973244,
"grad_norm": 0.5354152321815491,
"learning_rate": 6.393516434038722e-05,
"loss": 1.4539,
"step": 821
},
{
"epoch": 1.099665551839465,
"grad_norm": 0.6111787557601929,
"learning_rate": 6.389013957676723e-05,
"loss": 1.3616,
"step": 822
},
{
"epoch": 1.1010033444816054,
"grad_norm": 0.5158681273460388,
"learning_rate": 6.384511481314724e-05,
"loss": 1.4201,
"step": 823
},
{
"epoch": 1.1023411371237457,
"grad_norm": 0.6790710687637329,
"learning_rate": 6.380009004952723e-05,
"loss": 1.4573,
"step": 824
},
{
"epoch": 1.1036789297658862,
"grad_norm": 0.6283074021339417,
"learning_rate": 6.375506528590726e-05,
"loss": 1.0444,
"step": 825
},
{
"epoch": 1.1050167224080267,
"grad_norm": 0.41498756408691406,
"learning_rate": 6.371004052228727e-05,
"loss": 1.3567,
"step": 826
},
{
"epoch": 1.1063545150501672,
"grad_norm": 0.60197913646698,
"learning_rate": 6.366501575866727e-05,
"loss": 1.1496,
"step": 827
},
{
"epoch": 1.1076923076923078,
"grad_norm": 0.4228304922580719,
"learning_rate": 6.361999099504727e-05,
"loss": 1.3917,
"step": 828
},
{
"epoch": 1.109030100334448,
"grad_norm": 0.5080685019493103,
"learning_rate": 6.357496623142729e-05,
"loss": 1.8623,
"step": 829
},
{
"epoch": 1.1103678929765886,
"grad_norm": 0.49745386838912964,
"learning_rate": 6.35299414678073e-05,
"loss": 1.421,
"step": 830
},
{
"epoch": 1.111705685618729,
"grad_norm": 0.8473727107048035,
"learning_rate": 6.348491670418731e-05,
"loss": 0.7793,
"step": 831
},
{
"epoch": 1.1130434782608696,
"grad_norm": 0.5693299174308777,
"learning_rate": 6.343989194056732e-05,
"loss": 1.0338,
"step": 832
},
{
"epoch": 1.11438127090301,
"grad_norm": 0.5039275884628296,
"learning_rate": 6.339486717694733e-05,
"loss": 1.0676,
"step": 833
},
{
"epoch": 1.1157190635451506,
"grad_norm": 0.5785884261131287,
"learning_rate": 6.334984241332733e-05,
"loss": 1.4532,
"step": 834
},
{
"epoch": 1.117056856187291,
"grad_norm": 0.4384928047657013,
"learning_rate": 6.330481764970734e-05,
"loss": 1.9992,
"step": 835
},
{
"epoch": 1.1183946488294314,
"grad_norm": 0.6519571542739868,
"learning_rate": 6.325979288608735e-05,
"loss": 1.8105,
"step": 836
},
{
"epoch": 1.119732441471572,
"grad_norm": 0.5303328633308411,
"learning_rate": 6.321476812246736e-05,
"loss": 1.5492,
"step": 837
},
{
"epoch": 1.1210702341137124,
"grad_norm": 0.5411662459373474,
"learning_rate": 6.316974335884737e-05,
"loss": 1.4728,
"step": 838
},
{
"epoch": 1.122408026755853,
"grad_norm": 0.8510624170303345,
"learning_rate": 6.312471859522738e-05,
"loss": 1.0466,
"step": 839
},
{
"epoch": 1.1237458193979932,
"grad_norm": 0.3836539685726166,
"learning_rate": 6.307969383160739e-05,
"loss": 1.1762,
"step": 840
},
{
"epoch": 1.1250836120401337,
"grad_norm": 0.4980509877204895,
"learning_rate": 6.30346690679874e-05,
"loss": 1.7758,
"step": 841
},
{
"epoch": 1.1264214046822743,
"grad_norm": 0.364563524723053,
"learning_rate": 6.29896443043674e-05,
"loss": 1.2367,
"step": 842
},
{
"epoch": 1.1277591973244148,
"grad_norm": 0.7302637100219727,
"learning_rate": 6.294461954074741e-05,
"loss": 1.5548,
"step": 843
},
{
"epoch": 1.1290969899665553,
"grad_norm": 0.5207593441009521,
"learning_rate": 6.289959477712742e-05,
"loss": 1.3097,
"step": 844
},
{
"epoch": 1.1304347826086956,
"grad_norm": 0.519754946231842,
"learning_rate": 6.285457001350743e-05,
"loss": 1.5021,
"step": 845
},
{
"epoch": 1.131772575250836,
"grad_norm": 0.5610724687576294,
"learning_rate": 6.280954524988744e-05,
"loss": 1.4954,
"step": 846
},
{
"epoch": 1.1331103678929766,
"grad_norm": 0.44261258840560913,
"learning_rate": 6.276452048626746e-05,
"loss": 1.6946,
"step": 847
},
{
"epoch": 1.134448160535117,
"grad_norm": 0.5406861305236816,
"learning_rate": 6.271949572264745e-05,
"loss": 1.5805,
"step": 848
},
{
"epoch": 1.1357859531772576,
"grad_norm": 0.6939743757247925,
"learning_rate": 6.267447095902746e-05,
"loss": 1.3863,
"step": 849
},
{
"epoch": 1.137123745819398,
"grad_norm": 0.7074916362762451,
"learning_rate": 6.262944619540749e-05,
"loss": 0.8737,
"step": 850
},
{
"epoch": 1.1384615384615384,
"grad_norm": 0.5969648957252502,
"learning_rate": 6.25844214317875e-05,
"loss": 1.5225,
"step": 851
},
{
"epoch": 1.139799331103679,
"grad_norm": 0.542970597743988,
"learning_rate": 6.253939666816749e-05,
"loss": 1.5206,
"step": 852
},
{
"epoch": 1.1411371237458194,
"grad_norm": 0.4956679344177246,
"learning_rate": 6.24943719045475e-05,
"loss": 1.6725,
"step": 853
},
{
"epoch": 1.14247491638796,
"grad_norm": 0.620997965335846,
"learning_rate": 6.244934714092752e-05,
"loss": 1.3671,
"step": 854
},
{
"epoch": 1.1438127090301002,
"grad_norm": 0.5618900060653687,
"learning_rate": 6.240432237730753e-05,
"loss": 1.4067,
"step": 855
},
{
"epoch": 1.1451505016722408,
"grad_norm": 0.847754716873169,
"learning_rate": 6.235929761368752e-05,
"loss": 1.1148,
"step": 856
},
{
"epoch": 1.1464882943143813,
"grad_norm": 0.5353214144706726,
"learning_rate": 6.231427285006755e-05,
"loss": 1.5326,
"step": 857
},
{
"epoch": 1.1478260869565218,
"grad_norm": 0.6438226699829102,
"learning_rate": 6.226924808644755e-05,
"loss": 1.651,
"step": 858
},
{
"epoch": 1.1491638795986623,
"grad_norm": 0.5264010429382324,
"learning_rate": 6.222422332282756e-05,
"loss": 1.3338,
"step": 859
},
{
"epoch": 1.1505016722408028,
"grad_norm": 0.4603655934333801,
"learning_rate": 6.217919855920756e-05,
"loss": 1.1312,
"step": 860
},
{
"epoch": 1.151839464882943,
"grad_norm": 0.850045919418335,
"learning_rate": 6.213417379558758e-05,
"loss": 1.1957,
"step": 861
},
{
"epoch": 1.1531772575250836,
"grad_norm": 0.5338031053543091,
"learning_rate": 6.208914903196759e-05,
"loss": 1.9606,
"step": 862
},
{
"epoch": 1.154515050167224,
"grad_norm": 0.7752164006233215,
"learning_rate": 6.20441242683476e-05,
"loss": 1.2606,
"step": 863
},
{
"epoch": 1.1558528428093646,
"grad_norm": 0.532210111618042,
"learning_rate": 6.199909950472759e-05,
"loss": 1.7095,
"step": 864
},
{
"epoch": 1.1571906354515051,
"grad_norm": 0.5509344339370728,
"learning_rate": 6.195407474110761e-05,
"loss": 1.047,
"step": 865
},
{
"epoch": 1.1585284280936454,
"grad_norm": 0.43656522035598755,
"learning_rate": 6.190904997748762e-05,
"loss": 1.5622,
"step": 866
},
{
"epoch": 1.159866220735786,
"grad_norm": 0.5709621906280518,
"learning_rate": 6.186402521386763e-05,
"loss": 1.425,
"step": 867
},
{
"epoch": 1.1612040133779264,
"grad_norm": 0.4696739912033081,
"learning_rate": 6.181900045024764e-05,
"loss": 1.4856,
"step": 868
},
{
"epoch": 1.162541806020067,
"grad_norm": 0.4259323179721832,
"learning_rate": 6.177397568662765e-05,
"loss": 1.1235,
"step": 869
},
{
"epoch": 1.1638795986622075,
"grad_norm": 0.5247973799705505,
"learning_rate": 6.172895092300766e-05,
"loss": 1.5293,
"step": 870
},
{
"epoch": 1.1652173913043478,
"grad_norm": 0.510625422000885,
"learning_rate": 6.168392615938767e-05,
"loss": 1.5001,
"step": 871
},
{
"epoch": 1.1665551839464883,
"grad_norm": 0.5341862440109253,
"learning_rate": 6.163890139576767e-05,
"loss": 1.4726,
"step": 872
},
{
"epoch": 1.1678929765886288,
"grad_norm": 0.42538025975227356,
"learning_rate": 6.159387663214768e-05,
"loss": 1.1921,
"step": 873
},
{
"epoch": 1.1692307692307693,
"grad_norm": 0.5730955600738525,
"learning_rate": 6.154885186852769e-05,
"loss": 1.2907,
"step": 874
},
{
"epoch": 1.1705685618729098,
"grad_norm": 0.3825736939907074,
"learning_rate": 6.150382710490771e-05,
"loss": 1.7252,
"step": 875
},
{
"epoch": 1.17190635451505,
"grad_norm": 0.4969604015350342,
"learning_rate": 6.145880234128771e-05,
"loss": 1.4712,
"step": 876
},
{
"epoch": 1.1732441471571906,
"grad_norm": 0.4480053782463074,
"learning_rate": 6.141377757766772e-05,
"loss": 1.5548,
"step": 877
},
{
"epoch": 1.1745819397993311,
"grad_norm": 1.1541557312011719,
"learning_rate": 6.136875281404773e-05,
"loss": 1.3099,
"step": 878
},
{
"epoch": 1.1759197324414716,
"grad_norm": 0.5004495978355408,
"learning_rate": 6.132372805042775e-05,
"loss": 1.7068,
"step": 879
},
{
"epoch": 1.1772575250836121,
"grad_norm": 0.7697363495826721,
"learning_rate": 6.127870328680774e-05,
"loss": 1.5432,
"step": 880
},
{
"epoch": 1.1785953177257524,
"grad_norm": 0.9075251221656799,
"learning_rate": 6.123367852318775e-05,
"loss": 1.0909,
"step": 881
},
{
"epoch": 1.179933110367893,
"grad_norm": 0.509414553642273,
"learning_rate": 6.118865375956776e-05,
"loss": 1.3679,
"step": 882
},
{
"epoch": 1.1812709030100335,
"grad_norm": 0.924888014793396,
"learning_rate": 6.114362899594778e-05,
"loss": 1.5587,
"step": 883
},
{
"epoch": 1.182608695652174,
"grad_norm": 0.6355368494987488,
"learning_rate": 6.109860423232778e-05,
"loss": 1.1649,
"step": 884
},
{
"epoch": 1.1839464882943145,
"grad_norm": 0.550225019454956,
"learning_rate": 6.105357946870779e-05,
"loss": 1.5418,
"step": 885
},
{
"epoch": 1.1852842809364548,
"grad_norm": 0.5702582001686096,
"learning_rate": 6.100855470508781e-05,
"loss": 0.8423,
"step": 886
},
{
"epoch": 1.1866220735785953,
"grad_norm": 0.4942728877067566,
"learning_rate": 6.096352994146781e-05,
"loss": 0.9389,
"step": 887
},
{
"epoch": 1.1879598662207358,
"grad_norm": 0.5035423636436462,
"learning_rate": 6.091850517784782e-05,
"loss": 1.659,
"step": 888
},
{
"epoch": 1.1892976588628763,
"grad_norm": 0.3537347614765167,
"learning_rate": 6.087348041422783e-05,
"loss": 1.6727,
"step": 889
},
{
"epoch": 1.1906354515050168,
"grad_norm": 0.4293750822544098,
"learning_rate": 6.082845565060784e-05,
"loss": 1.9057,
"step": 890
},
{
"epoch": 1.191973244147157,
"grad_norm": 0.6442123055458069,
"learning_rate": 6.0783430886987844e-05,
"loss": 1.2067,
"step": 891
},
{
"epoch": 1.1933110367892976,
"grad_norm": 0.48610013723373413,
"learning_rate": 6.073840612336785e-05,
"loss": 1.6491,
"step": 892
},
{
"epoch": 1.1946488294314381,
"grad_norm": 0.5059003233909607,
"learning_rate": 6.069338135974787e-05,
"loss": 1.5232,
"step": 893
},
{
"epoch": 1.1959866220735786,
"grad_norm": 0.4608552157878876,
"learning_rate": 6.0648356596127876e-05,
"loss": 1.652,
"step": 894
},
{
"epoch": 1.1973244147157192,
"grad_norm": 0.5745627284049988,
"learning_rate": 6.060333183250788e-05,
"loss": 1.3177,
"step": 895
},
{
"epoch": 1.1986622073578594,
"grad_norm": 0.5850013494491577,
"learning_rate": 6.055830706888789e-05,
"loss": 0.9728,
"step": 896
},
{
"epoch": 1.2,
"grad_norm": 0.595733642578125,
"learning_rate": 6.05132823052679e-05,
"loss": 1.4109,
"step": 897
},
{
"epoch": 1.2013377926421405,
"grad_norm": 0.5972495675086975,
"learning_rate": 6.046825754164791e-05,
"loss": 1.762,
"step": 898
},
{
"epoch": 1.202675585284281,
"grad_norm": 0.7657443881034851,
"learning_rate": 6.042323277802792e-05,
"loss": 1.2932,
"step": 899
},
{
"epoch": 1.2040133779264215,
"grad_norm": 0.5386178493499756,
"learning_rate": 6.037820801440792e-05,
"loss": 0.8722,
"step": 900
},
{
"epoch": 1.2053511705685618,
"grad_norm": 0.4514966607093811,
"learning_rate": 6.033318325078794e-05,
"loss": 1.2226,
"step": 901
},
{
"epoch": 1.2066889632107023,
"grad_norm": 0.561952531337738,
"learning_rate": 6.0288158487167945e-05,
"loss": 1.3083,
"step": 902
},
{
"epoch": 1.2080267558528428,
"grad_norm": 0.46882307529449463,
"learning_rate": 6.0243133723547954e-05,
"loss": 1.6229,
"step": 903
},
{
"epoch": 1.2093645484949833,
"grad_norm": 0.41895994544029236,
"learning_rate": 6.019810895992797e-05,
"loss": 1.6748,
"step": 904
},
{
"epoch": 1.2107023411371238,
"grad_norm": 0.4929127097129822,
"learning_rate": 6.015308419630797e-05,
"loss": 1.7661,
"step": 905
},
{
"epoch": 1.2120401337792641,
"grad_norm": 0.5943595170974731,
"learning_rate": 6.010805943268798e-05,
"loss": 1.3203,
"step": 906
},
{
"epoch": 1.2133779264214046,
"grad_norm": 0.5031788945198059,
"learning_rate": 6.006303466906799e-05,
"loss": 1.5386,
"step": 907
},
{
"epoch": 1.2147157190635451,
"grad_norm": 0.5768786072731018,
"learning_rate": 6.0018009905448004e-05,
"loss": 1.1965,
"step": 908
},
{
"epoch": 1.2160535117056857,
"grad_norm": 0.3699272871017456,
"learning_rate": 5.9972985141828005e-05,
"loss": 1.2933,
"step": 909
},
{
"epoch": 1.2173913043478262,
"grad_norm": 0.7275379300117493,
"learning_rate": 5.9927960378208014e-05,
"loss": 1.3549,
"step": 910
},
{
"epoch": 1.2187290969899665,
"grad_norm": 0.48716142773628235,
"learning_rate": 5.988293561458803e-05,
"loss": 1.0621,
"step": 911
},
{
"epoch": 1.220066889632107,
"grad_norm": 0.8821597695350647,
"learning_rate": 5.983791085096804e-05,
"loss": 1.3591,
"step": 912
},
{
"epoch": 1.2214046822742475,
"grad_norm": 0.48447397351264954,
"learning_rate": 5.9792886087348046e-05,
"loss": 1.1019,
"step": 913
},
{
"epoch": 1.222742474916388,
"grad_norm": 0.5043441653251648,
"learning_rate": 5.974786132372805e-05,
"loss": 1.7618,
"step": 914
},
{
"epoch": 1.2240802675585285,
"grad_norm": 0.6771908402442932,
"learning_rate": 5.9702836560108064e-05,
"loss": 1.3763,
"step": 915
},
{
"epoch": 1.2254180602006688,
"grad_norm": 0.4996180832386017,
"learning_rate": 5.965781179648807e-05,
"loss": 1.7422,
"step": 916
},
{
"epoch": 1.2267558528428093,
"grad_norm": 0.32585346698760986,
"learning_rate": 5.961278703286808e-05,
"loss": 1.9686,
"step": 917
},
{
"epoch": 1.2280936454849498,
"grad_norm": 0.4633427560329437,
"learning_rate": 5.956776226924808e-05,
"loss": 1.619,
"step": 918
},
{
"epoch": 1.2294314381270903,
"grad_norm": 0.47458308935165405,
"learning_rate": 5.95227375056281e-05,
"loss": 1.8054,
"step": 919
},
{
"epoch": 1.2307692307692308,
"grad_norm": 0.5091044902801514,
"learning_rate": 5.9477712742008107e-05,
"loss": 1.7261,
"step": 920
},
{
"epoch": 1.2321070234113711,
"grad_norm": 0.5245359539985657,
"learning_rate": 5.9432687978388115e-05,
"loss": 1.9436,
"step": 921
},
{
"epoch": 1.2334448160535116,
"grad_norm": 0.6491963267326355,
"learning_rate": 5.938766321476813e-05,
"loss": 1.4365,
"step": 922
},
{
"epoch": 1.2347826086956522,
"grad_norm": 0.3472338616847992,
"learning_rate": 5.934263845114814e-05,
"loss": 1.1998,
"step": 923
},
{
"epoch": 1.2361204013377927,
"grad_norm": 0.520018458366394,
"learning_rate": 5.929761368752814e-05,
"loss": 1.409,
"step": 924
},
{
"epoch": 1.2374581939799332,
"grad_norm": 0.3718913197517395,
"learning_rate": 5.925258892390815e-05,
"loss": 1.1047,
"step": 925
},
{
"epoch": 1.2387959866220735,
"grad_norm": 0.5046524405479431,
"learning_rate": 5.9207564160288165e-05,
"loss": 1.6533,
"step": 926
},
{
"epoch": 1.240133779264214,
"grad_norm": 0.49469617009162903,
"learning_rate": 5.9162539396668173e-05,
"loss": 1.3681,
"step": 927
},
{
"epoch": 1.2414715719063545,
"grad_norm": 0.5294374227523804,
"learning_rate": 5.9117514633048175e-05,
"loss": 1.8672,
"step": 928
},
{
"epoch": 1.242809364548495,
"grad_norm": 0.44988906383514404,
"learning_rate": 5.907248986942819e-05,
"loss": 1.3822,
"step": 929
},
{
"epoch": 1.2441471571906355,
"grad_norm": 0.4916761517524719,
"learning_rate": 5.90274651058082e-05,
"loss": 1.3785,
"step": 930
},
{
"epoch": 1.2454849498327758,
"grad_norm": 0.7246065139770508,
"learning_rate": 5.898244034218821e-05,
"loss": 1.0528,
"step": 931
},
{
"epoch": 1.2468227424749163,
"grad_norm": 0.5096299648284912,
"learning_rate": 5.893741557856821e-05,
"loss": 1.5648,
"step": 932
},
{
"epoch": 1.2481605351170568,
"grad_norm": 0.5564544796943665,
"learning_rate": 5.889239081494823e-05,
"loss": 1.4249,
"step": 933
},
{
"epoch": 1.2494983277591973,
"grad_norm": 0.4242939352989197,
"learning_rate": 5.8847366051328234e-05,
"loss": 1.6266,
"step": 934
},
{
"epoch": 1.2508361204013378,
"grad_norm": 0.965453565120697,
"learning_rate": 5.880234128770824e-05,
"loss": 1.1993,
"step": 935
},
{
"epoch": 1.2521739130434781,
"grad_norm": 0.8625237941741943,
"learning_rate": 5.8757316524088244e-05,
"loss": 1.2209,
"step": 936
},
{
"epoch": 1.2535117056856186,
"grad_norm": 0.542377769947052,
"learning_rate": 5.8712291760468266e-05,
"loss": 1.7549,
"step": 937
},
{
"epoch": 1.2548494983277592,
"grad_norm": 0.5045947432518005,
"learning_rate": 5.866726699684827e-05,
"loss": 1.6812,
"step": 938
},
{
"epoch": 1.2561872909698997,
"grad_norm": 0.5573652982711792,
"learning_rate": 5.8622242233228277e-05,
"loss": 1.5869,
"step": 939
},
{
"epoch": 1.2575250836120402,
"grad_norm": 0.771754264831543,
"learning_rate": 5.857721746960829e-05,
"loss": 1.502,
"step": 940
},
{
"epoch": 1.2588628762541805,
"grad_norm": 0.5458933711051941,
"learning_rate": 5.85321927059883e-05,
"loss": 1.7175,
"step": 941
},
{
"epoch": 1.2602006688963212,
"grad_norm": 0.4187781810760498,
"learning_rate": 5.84871679423683e-05,
"loss": 1.7161,
"step": 942
},
{
"epoch": 1.2615384615384615,
"grad_norm": 0.5237634778022766,
"learning_rate": 5.844214317874831e-05,
"loss": 0.8686,
"step": 943
},
{
"epoch": 1.262876254180602,
"grad_norm": 0.32764026522636414,
"learning_rate": 5.8397118415128326e-05,
"loss": 1.4472,
"step": 944
},
{
"epoch": 1.2642140468227425,
"grad_norm": 0.7707194685935974,
"learning_rate": 5.8352093651508335e-05,
"loss": 1.1732,
"step": 945
},
{
"epoch": 1.2655518394648828,
"grad_norm": 0.4281066656112671,
"learning_rate": 5.830706888788834e-05,
"loss": 1.6155,
"step": 946
},
{
"epoch": 1.2668896321070235,
"grad_norm": 0.5570145845413208,
"learning_rate": 5.826204412426836e-05,
"loss": 1.1536,
"step": 947
},
{
"epoch": 1.2682274247491638,
"grad_norm": 0.7634395360946655,
"learning_rate": 5.821701936064836e-05,
"loss": 1.1469,
"step": 948
},
{
"epoch": 1.2695652173913043,
"grad_norm": 0.5208802819252014,
"learning_rate": 5.817199459702837e-05,
"loss": 1.382,
"step": 949
},
{
"epoch": 1.2709030100334449,
"grad_norm": 0.5326187610626221,
"learning_rate": 5.812696983340837e-05,
"loss": 1.5838,
"step": 950
},
{
"epoch": 1.2722408026755851,
"grad_norm": 0.506121814250946,
"learning_rate": 5.808194506978839e-05,
"loss": 1.5555,
"step": 951
},
{
"epoch": 1.2735785953177259,
"grad_norm": 0.6669555902481079,
"learning_rate": 5.8036920306168395e-05,
"loss": 1.3553,
"step": 952
},
{
"epoch": 1.2749163879598662,
"grad_norm": 0.7633807063102722,
"learning_rate": 5.7991895542548404e-05,
"loss": 1.3975,
"step": 953
},
{
"epoch": 1.2762541806020067,
"grad_norm": 0.6958717107772827,
"learning_rate": 5.7946870778928405e-05,
"loss": 1.3524,
"step": 954
},
{
"epoch": 1.2775919732441472,
"grad_norm": 0.688834011554718,
"learning_rate": 5.790184601530843e-05,
"loss": 1.2323,
"step": 955
},
{
"epoch": 1.2789297658862877,
"grad_norm": 0.8541237115859985,
"learning_rate": 5.785682125168843e-05,
"loss": 1.172,
"step": 956
},
{
"epoch": 1.2802675585284282,
"grad_norm": 0.4427731931209564,
"learning_rate": 5.781179648806844e-05,
"loss": 1.6397,
"step": 957
},
{
"epoch": 1.2816053511705685,
"grad_norm": 0.5415353775024414,
"learning_rate": 5.776677172444845e-05,
"loss": 1.3547,
"step": 958
},
{
"epoch": 1.282943143812709,
"grad_norm": 0.6246786117553711,
"learning_rate": 5.772174696082846e-05,
"loss": 1.523,
"step": 959
},
{
"epoch": 1.2842809364548495,
"grad_norm": 0.477402001619339,
"learning_rate": 5.7676722197208464e-05,
"loss": 1.652,
"step": 960
},
{
"epoch": 1.28561872909699,
"grad_norm": 0.4590649902820587,
"learning_rate": 5.763169743358847e-05,
"loss": 1.7355,
"step": 961
},
{
"epoch": 1.2869565217391306,
"grad_norm": 0.43109437823295593,
"learning_rate": 5.758667266996849e-05,
"loss": 1.8118,
"step": 962
},
{
"epoch": 1.2882943143812708,
"grad_norm": 0.7176820635795593,
"learning_rate": 5.7541647906348496e-05,
"loss": 0.9641,
"step": 963
},
{
"epoch": 1.2896321070234114,
"grad_norm": 0.5545156598091125,
"learning_rate": 5.74966231427285e-05,
"loss": 1.3134,
"step": 964
},
{
"epoch": 1.2909698996655519,
"grad_norm": 0.7610632181167603,
"learning_rate": 5.745159837910852e-05,
"loss": 1.5356,
"step": 965
},
{
"epoch": 1.2923076923076924,
"grad_norm": 0.5686385035514832,
"learning_rate": 5.740657361548852e-05,
"loss": 1.4585,
"step": 966
},
{
"epoch": 1.293645484949833,
"grad_norm": 0.7597028613090515,
"learning_rate": 5.736154885186853e-05,
"loss": 1.5304,
"step": 967
},
{
"epoch": 1.2949832775919732,
"grad_norm": 0.5620704889297485,
"learning_rate": 5.731652408824853e-05,
"loss": 1.4065,
"step": 968
},
{
"epoch": 1.2963210702341137,
"grad_norm": 0.46487486362457275,
"learning_rate": 5.7271499324628554e-05,
"loss": 1.6403,
"step": 969
},
{
"epoch": 1.2976588628762542,
"grad_norm": 0.6293439865112305,
"learning_rate": 5.7226474561008556e-05,
"loss": 1.4635,
"step": 970
},
{
"epoch": 1.2989966555183947,
"grad_norm": 0.9160875678062439,
"learning_rate": 5.7181449797388565e-05,
"loss": 1.0431,
"step": 971
},
{
"epoch": 1.3003344481605352,
"grad_norm": 0.8708162903785706,
"learning_rate": 5.7136425033768573e-05,
"loss": 1.5403,
"step": 972
},
{
"epoch": 1.3016722408026755,
"grad_norm": 0.9074984788894653,
"learning_rate": 5.709140027014859e-05,
"loss": 1.2068,
"step": 973
},
{
"epoch": 1.303010033444816,
"grad_norm": 0.6425647139549255,
"learning_rate": 5.704637550652859e-05,
"loss": 1.2269,
"step": 974
},
{
"epoch": 1.3043478260869565,
"grad_norm": 0.57787024974823,
"learning_rate": 5.70013507429086e-05,
"loss": 1.5209,
"step": 975
},
{
"epoch": 1.305685618729097,
"grad_norm": 0.6588714122772217,
"learning_rate": 5.6956325979288615e-05,
"loss": 1.4408,
"step": 976
},
{
"epoch": 1.3070234113712376,
"grad_norm": 0.4849056899547577,
"learning_rate": 5.691130121566862e-05,
"loss": 1.416,
"step": 977
},
{
"epoch": 1.3083612040133779,
"grad_norm": 0.5900976061820984,
"learning_rate": 5.6866276452048625e-05,
"loss": 1.3661,
"step": 978
},
{
"epoch": 1.3096989966555184,
"grad_norm": 0.528537392616272,
"learning_rate": 5.6821251688428634e-05,
"loss": 1.4337,
"step": 979
},
{
"epoch": 1.3110367892976589,
"grad_norm": 0.6017236113548279,
"learning_rate": 5.677622692480865e-05,
"loss": 1.3365,
"step": 980
},
{
"epoch": 1.3123745819397994,
"grad_norm": 0.5581938624382019,
"learning_rate": 5.673120216118866e-05,
"loss": 1.0921,
"step": 981
},
{
"epoch": 1.31371237458194,
"grad_norm": 0.4984164535999298,
"learning_rate": 5.6686177397568666e-05,
"loss": 1.7687,
"step": 982
},
{
"epoch": 1.3150501672240802,
"grad_norm": 0.6263704895973206,
"learning_rate": 5.664115263394868e-05,
"loss": 0.8377,
"step": 983
},
{
"epoch": 1.3163879598662207,
"grad_norm": 0.5864318013191223,
"learning_rate": 5.659612787032868e-05,
"loss": 1.2958,
"step": 984
},
{
"epoch": 1.3177257525083612,
"grad_norm": 0.5671847462654114,
"learning_rate": 5.655110310670869e-05,
"loss": 1.4972,
"step": 985
},
{
"epoch": 1.3190635451505017,
"grad_norm": 1.1298816204071045,
"learning_rate": 5.65060783430887e-05,
"loss": 0.9593,
"step": 986
},
{
"epoch": 1.3204013377926422,
"grad_norm": 0.5613873600959778,
"learning_rate": 5.6461053579468716e-05,
"loss": 1.7714,
"step": 987
},
{
"epoch": 1.3217391304347825,
"grad_norm": 0.4778936207294464,
"learning_rate": 5.641602881584872e-05,
"loss": 1.4784,
"step": 988
},
{
"epoch": 1.323076923076923,
"grad_norm": 0.5867298245429993,
"learning_rate": 5.6371004052228726e-05,
"loss": 1.6855,
"step": 989
},
{
"epoch": 1.3244147157190636,
"grad_norm": 0.67825847864151,
"learning_rate": 5.6325979288608735e-05,
"loss": 1.3786,
"step": 990
},
{
"epoch": 1.325752508361204,
"grad_norm": 0.566182017326355,
"learning_rate": 5.628095452498875e-05,
"loss": 1.5857,
"step": 991
},
{
"epoch": 1.3270903010033446,
"grad_norm": 0.6108747124671936,
"learning_rate": 5.623592976136875e-05,
"loss": 0.9231,
"step": 992
},
{
"epoch": 1.3284280936454849,
"grad_norm": 0.3762817680835724,
"learning_rate": 5.619090499774876e-05,
"loss": 1.8333,
"step": 993
},
{
"epoch": 1.3297658862876254,
"grad_norm": 0.42532092332839966,
"learning_rate": 5.6145880234128776e-05,
"loss": 1.4002,
"step": 994
},
{
"epoch": 1.3311036789297659,
"grad_norm": 0.409871369600296,
"learning_rate": 5.6100855470508785e-05,
"loss": 1.5429,
"step": 995
},
{
"epoch": 1.3324414715719064,
"grad_norm": 0.4747501313686371,
"learning_rate": 5.605583070688879e-05,
"loss": 1.2592,
"step": 996
},
{
"epoch": 1.333779264214047,
"grad_norm": 0.4586211144924164,
"learning_rate": 5.6010805943268795e-05,
"loss": 1.5638,
"step": 997
},
{
"epoch": 1.3351170568561872,
"grad_norm": 0.48024147748947144,
"learning_rate": 5.596578117964881e-05,
"loss": 1.3406,
"step": 998
},
{
"epoch": 1.3364548494983277,
"grad_norm": 0.5290399193763733,
"learning_rate": 5.592075641602882e-05,
"loss": 1.4827,
"step": 999
},
{
"epoch": 1.3377926421404682,
"grad_norm": 0.6649457812309265,
"learning_rate": 5.587573165240883e-05,
"loss": 1.3084,
"step": 1000
},
{
"epoch": 1.3391304347826087,
"grad_norm": 0.5224519371986389,
"learning_rate": 5.583070688878884e-05,
"loss": 1.4575,
"step": 1001
},
{
"epoch": 1.3404682274247492,
"grad_norm": 0.5149902105331421,
"learning_rate": 5.5785682125168845e-05,
"loss": 1.2122,
"step": 1002
},
{
"epoch": 1.3418060200668895,
"grad_norm": 0.5006117820739746,
"learning_rate": 5.574065736154885e-05,
"loss": 1.6049,
"step": 1003
},
{
"epoch": 1.34314381270903,
"grad_norm": 1.1320693492889404,
"learning_rate": 5.569563259792886e-05,
"loss": 1.076,
"step": 1004
},
{
"epoch": 1.3444816053511706,
"grad_norm": 0.5004568099975586,
"learning_rate": 5.565060783430888e-05,
"loss": 1.7695,
"step": 1005
},
{
"epoch": 1.345819397993311,
"grad_norm": 0.7288807034492493,
"learning_rate": 5.5605583070688886e-05,
"loss": 1.5639,
"step": 1006
},
{
"epoch": 1.3471571906354516,
"grad_norm": 0.5516156554222107,
"learning_rate": 5.556055830706889e-05,
"loss": 1.2222,
"step": 1007
},
{
"epoch": 1.3484949832775919,
"grad_norm": 0.5309849381446838,
"learning_rate": 5.5515533543448896e-05,
"loss": 1.451,
"step": 1008
},
{
"epoch": 1.3498327759197324,
"grad_norm": 0.5378895401954651,
"learning_rate": 5.547050877982891e-05,
"loss": 1.4105,
"step": 1009
},
{
"epoch": 1.351170568561873,
"grad_norm": 0.5446533560752869,
"learning_rate": 5.542548401620892e-05,
"loss": 1.6093,
"step": 1010
},
{
"epoch": 1.3525083612040134,
"grad_norm": 0.4849577844142914,
"learning_rate": 5.538045925258892e-05,
"loss": 1.4458,
"step": 1011
},
{
"epoch": 1.353846153846154,
"grad_norm": 0.7450141906738281,
"learning_rate": 5.533543448896894e-05,
"loss": 0.6454,
"step": 1012
},
{
"epoch": 1.3551839464882942,
"grad_norm": 0.4997408986091614,
"learning_rate": 5.5290409725348946e-05,
"loss": 1.5501,
"step": 1013
},
{
"epoch": 1.3565217391304347,
"grad_norm": 0.49663230776786804,
"learning_rate": 5.5245384961728954e-05,
"loss": 1.2957,
"step": 1014
},
{
"epoch": 1.3578595317725752,
"grad_norm": 0.47518983483314514,
"learning_rate": 5.5200360198108956e-05,
"loss": 1.5983,
"step": 1015
},
{
"epoch": 1.3591973244147157,
"grad_norm": 0.3959956765174866,
"learning_rate": 5.515533543448897e-05,
"loss": 1.5998,
"step": 1016
},
{
"epoch": 1.3605351170568563,
"grad_norm": 0.5632640719413757,
"learning_rate": 5.511031067086898e-05,
"loss": 1.4659,
"step": 1017
},
{
"epoch": 1.3618729096989965,
"grad_norm": 0.6305011510848999,
"learning_rate": 5.506528590724899e-05,
"loss": 1.1801,
"step": 1018
},
{
"epoch": 1.363210702341137,
"grad_norm": 0.6637885570526123,
"learning_rate": 5.5020261143629004e-05,
"loss": 1.137,
"step": 1019
},
{
"epoch": 1.3645484949832776,
"grad_norm": 0.5782528519630432,
"learning_rate": 5.497523638000901e-05,
"loss": 1.4629,
"step": 1020
},
{
"epoch": 1.365886287625418,
"grad_norm": 0.5602366924285889,
"learning_rate": 5.4930211616389015e-05,
"loss": 1.3326,
"step": 1021
},
{
"epoch": 1.3672240802675586,
"grad_norm": 0.6684134006500244,
"learning_rate": 5.488518685276902e-05,
"loss": 1.3521,
"step": 1022
},
{
"epoch": 1.3685618729096989,
"grad_norm": 0.7985564470291138,
"learning_rate": 5.484016208914904e-05,
"loss": 0.8858,
"step": 1023
},
{
"epoch": 1.3698996655518394,
"grad_norm": 0.5301754474639893,
"learning_rate": 5.479513732552905e-05,
"loss": 1.5861,
"step": 1024
},
{
"epoch": 1.37123745819398,
"grad_norm": 0.5510058403015137,
"learning_rate": 5.475011256190905e-05,
"loss": 1.4559,
"step": 1025
},
{
"epoch": 1.3725752508361204,
"grad_norm": 0.8531842827796936,
"learning_rate": 5.470508779828906e-05,
"loss": 1.4512,
"step": 1026
},
{
"epoch": 1.373913043478261,
"grad_norm": 0.5832063555717468,
"learning_rate": 5.466006303466907e-05,
"loss": 1.392,
"step": 1027
},
{
"epoch": 1.3752508361204012,
"grad_norm": 0.587691068649292,
"learning_rate": 5.461503827104908e-05,
"loss": 1.5835,
"step": 1028
},
{
"epoch": 1.3765886287625417,
"grad_norm": 0.5541633367538452,
"learning_rate": 5.457001350742908e-05,
"loss": 1.5491,
"step": 1029
},
{
"epoch": 1.3779264214046822,
"grad_norm": 0.46723368763923645,
"learning_rate": 5.4524988743809105e-05,
"loss": 1.5469,
"step": 1030
},
{
"epoch": 1.3792642140468228,
"grad_norm": 0.5442748069763184,
"learning_rate": 5.447996398018911e-05,
"loss": 1.4914,
"step": 1031
},
{
"epoch": 1.3806020066889633,
"grad_norm": 0.5219207406044006,
"learning_rate": 5.4434939216569116e-05,
"loss": 1.7056,
"step": 1032
},
{
"epoch": 1.3819397993311036,
"grad_norm": 0.5391073822975159,
"learning_rate": 5.438991445294912e-05,
"loss": 1.1345,
"step": 1033
},
{
"epoch": 1.383277591973244,
"grad_norm": 0.5431692004203796,
"learning_rate": 5.434488968932914e-05,
"loss": 1.6605,
"step": 1034
},
{
"epoch": 1.3846153846153846,
"grad_norm": 0.5454650521278381,
"learning_rate": 5.429986492570914e-05,
"loss": 1.6891,
"step": 1035
},
{
"epoch": 1.385953177257525,
"grad_norm": 0.5413758754730225,
"learning_rate": 5.425484016208915e-05,
"loss": 1.5395,
"step": 1036
},
{
"epoch": 1.3872909698996656,
"grad_norm": 0.44604915380477905,
"learning_rate": 5.4209815398469166e-05,
"loss": 1.6368,
"step": 1037
},
{
"epoch": 1.388628762541806,
"grad_norm": 0.6125640869140625,
"learning_rate": 5.4164790634849174e-05,
"loss": 1.2193,
"step": 1038
},
{
"epoch": 1.3899665551839464,
"grad_norm": 0.5223219990730286,
"learning_rate": 5.4119765871229176e-05,
"loss": 1.511,
"step": 1039
},
{
"epoch": 1.391304347826087,
"grad_norm": 0.7312715649604797,
"learning_rate": 5.4074741107609185e-05,
"loss": 1.4429,
"step": 1040
},
{
"epoch": 1.3926421404682274,
"grad_norm": 0.5728110671043396,
"learning_rate": 5.40297163439892e-05,
"loss": 1.2849,
"step": 1041
},
{
"epoch": 1.393979933110368,
"grad_norm": 0.475210040807724,
"learning_rate": 5.398469158036921e-05,
"loss": 1.5608,
"step": 1042
},
{
"epoch": 1.3953177257525082,
"grad_norm": 0.4216032326221466,
"learning_rate": 5.393966681674921e-05,
"loss": 1.4798,
"step": 1043
},
{
"epoch": 1.396655518394649,
"grad_norm": 0.45767804980278015,
"learning_rate": 5.389464205312922e-05,
"loss": 1.8409,
"step": 1044
},
{
"epoch": 1.3979933110367893,
"grad_norm": 0.46793442964553833,
"learning_rate": 5.3849617289509234e-05,
"loss": 1.5628,
"step": 1045
},
{
"epoch": 1.3993311036789298,
"grad_norm": 0.5342912673950195,
"learning_rate": 5.380459252588924e-05,
"loss": 1.6847,
"step": 1046
},
{
"epoch": 1.4006688963210703,
"grad_norm": 0.6303462386131287,
"learning_rate": 5.3759567762269245e-05,
"loss": 1.4752,
"step": 1047
},
{
"epoch": 1.4020066889632106,
"grad_norm": 0.4938918650150299,
"learning_rate": 5.371454299864927e-05,
"loss": 1.8291,
"step": 1048
},
{
"epoch": 1.4033444816053513,
"grad_norm": 0.6580408215522766,
"learning_rate": 5.366951823502927e-05,
"loss": 0.8986,
"step": 1049
},
{
"epoch": 1.4046822742474916,
"grad_norm": 0.600391685962677,
"learning_rate": 5.362449347140928e-05,
"loss": 1.4318,
"step": 1050
},
{
"epoch": 1.406020066889632,
"grad_norm": 0.4521050453186035,
"learning_rate": 5.357946870778928e-05,
"loss": 1.4837,
"step": 1051
},
{
"epoch": 1.4073578595317726,
"grad_norm": 0.6474264860153198,
"learning_rate": 5.35344439441693e-05,
"loss": 1.3424,
"step": 1052
},
{
"epoch": 1.4086956521739131,
"grad_norm": 0.500771701335907,
"learning_rate": 5.34894191805493e-05,
"loss": 1.8707,
"step": 1053
},
{
"epoch": 1.4100334448160536,
"grad_norm": 0.5102487206459045,
"learning_rate": 5.344439441692931e-05,
"loss": 1.8188,
"step": 1054
},
{
"epoch": 1.411371237458194,
"grad_norm": 0.406031996011734,
"learning_rate": 5.339936965330933e-05,
"loss": 1.854,
"step": 1055
},
{
"epoch": 1.4127090301003344,
"grad_norm": 0.415464848279953,
"learning_rate": 5.3354344889689335e-05,
"loss": 1.8053,
"step": 1056
},
{
"epoch": 1.414046822742475,
"grad_norm": 0.5517300963401794,
"learning_rate": 5.330932012606934e-05,
"loss": 1.227,
"step": 1057
},
{
"epoch": 1.4153846153846155,
"grad_norm": 0.6167062520980835,
"learning_rate": 5.3264295362449346e-05,
"loss": 1.31,
"step": 1058
},
{
"epoch": 1.416722408026756,
"grad_norm": 0.5428354740142822,
"learning_rate": 5.321927059882936e-05,
"loss": 1.2271,
"step": 1059
},
{
"epoch": 1.4180602006688963,
"grad_norm": 0.5453240871429443,
"learning_rate": 5.317424583520937e-05,
"loss": 1.759,
"step": 1060
},
{
"epoch": 1.4193979933110368,
"grad_norm": 0.4308525323867798,
"learning_rate": 5.312922107158937e-05,
"loss": 1.8145,
"step": 1061
},
{
"epoch": 1.4207357859531773,
"grad_norm": 0.5427666902542114,
"learning_rate": 5.308419630796938e-05,
"loss": 1.4224,
"step": 1062
},
{
"epoch": 1.4220735785953178,
"grad_norm": 0.46834635734558105,
"learning_rate": 5.3039171544349396e-05,
"loss": 1.4914,
"step": 1063
},
{
"epoch": 1.4234113712374583,
"grad_norm": 0.615035355091095,
"learning_rate": 5.2994146780729404e-05,
"loss": 1.2526,
"step": 1064
},
{
"epoch": 1.4247491638795986,
"grad_norm": 0.7380486130714417,
"learning_rate": 5.2949122017109406e-05,
"loss": 1.4013,
"step": 1065
},
{
"epoch": 1.4260869565217391,
"grad_norm": 0.5666106939315796,
"learning_rate": 5.290409725348943e-05,
"loss": 1.1257,
"step": 1066
},
{
"epoch": 1.4274247491638796,
"grad_norm": 0.3648187220096588,
"learning_rate": 5.285907248986943e-05,
"loss": 1.1758,
"step": 1067
},
{
"epoch": 1.4287625418060201,
"grad_norm": 0.4655036926269531,
"learning_rate": 5.281404772624944e-05,
"loss": 1.0849,
"step": 1068
},
{
"epoch": 1.4301003344481606,
"grad_norm": 0.4767707586288452,
"learning_rate": 5.276902296262945e-05,
"loss": 1.4827,
"step": 1069
},
{
"epoch": 1.431438127090301,
"grad_norm": 0.5286468267440796,
"learning_rate": 5.272399819900946e-05,
"loss": 1.4783,
"step": 1070
},
{
"epoch": 1.4327759197324414,
"grad_norm": 0.47978368401527405,
"learning_rate": 5.2678973435389464e-05,
"loss": 1.4588,
"step": 1071
},
{
"epoch": 1.434113712374582,
"grad_norm": 0.6320868730545044,
"learning_rate": 5.263394867176947e-05,
"loss": 1.4445,
"step": 1072
},
{
"epoch": 1.4354515050167225,
"grad_norm": 0.7280179262161255,
"learning_rate": 5.258892390814949e-05,
"loss": 1.33,
"step": 1073
},
{
"epoch": 1.436789297658863,
"grad_norm": 0.5280917286872864,
"learning_rate": 5.25438991445295e-05,
"loss": 1.3857,
"step": 1074
},
{
"epoch": 1.4381270903010033,
"grad_norm": 0.6721616387367249,
"learning_rate": 5.24988743809095e-05,
"loss": 1.5877,
"step": 1075
},
{
"epoch": 1.4394648829431438,
"grad_norm": 0.5173552632331848,
"learning_rate": 5.245384961728951e-05,
"loss": 0.9419,
"step": 1076
},
{
"epoch": 1.4408026755852843,
"grad_norm": 0.6093307137489319,
"learning_rate": 5.240882485366952e-05,
"loss": 0.8124,
"step": 1077
},
{
"epoch": 1.4421404682274248,
"grad_norm": 0.8187419772148132,
"learning_rate": 5.236380009004953e-05,
"loss": 1.5013,
"step": 1078
},
{
"epoch": 1.4434782608695653,
"grad_norm": 0.5983606576919556,
"learning_rate": 5.231877532642954e-05,
"loss": 1.2655,
"step": 1079
},
{
"epoch": 1.4448160535117056,
"grad_norm": 0.5444653630256653,
"learning_rate": 5.227375056280954e-05,
"loss": 1.2253,
"step": 1080
},
{
"epoch": 1.4461538461538461,
"grad_norm": 0.563248336315155,
"learning_rate": 5.222872579918956e-05,
"loss": 1.414,
"step": 1081
},
{
"epoch": 1.4474916387959866,
"grad_norm": 0.6191748976707458,
"learning_rate": 5.2183701035569566e-05,
"loss": 1.1588,
"step": 1082
},
{
"epoch": 1.4488294314381271,
"grad_norm": 0.604679524898529,
"learning_rate": 5.2138676271949574e-05,
"loss": 1.4095,
"step": 1083
},
{
"epoch": 1.4501672240802677,
"grad_norm": 0.5420505404472351,
"learning_rate": 5.209365150832959e-05,
"loss": 1.1935,
"step": 1084
},
{
"epoch": 1.451505016722408,
"grad_norm": 0.651980996131897,
"learning_rate": 5.204862674470959e-05,
"loss": 1.4503,
"step": 1085
},
{
"epoch": 1.4528428093645485,
"grad_norm": 0.6438544988632202,
"learning_rate": 5.20036019810896e-05,
"loss": 1.5319,
"step": 1086
},
{
"epoch": 1.454180602006689,
"grad_norm": 0.5814422965049744,
"learning_rate": 5.195857721746961e-05,
"loss": 0.9881,
"step": 1087
},
{
"epoch": 1.4555183946488295,
"grad_norm": 0.4369351863861084,
"learning_rate": 5.1913552453849624e-05,
"loss": 1.2893,
"step": 1088
},
{
"epoch": 1.45685618729097,
"grad_norm": 0.8631730079650879,
"learning_rate": 5.186852769022963e-05,
"loss": 1.6095,
"step": 1089
},
{
"epoch": 1.4581939799331103,
"grad_norm": 0.4449801743030548,
"learning_rate": 5.1823502926609634e-05,
"loss": 1.5782,
"step": 1090
},
{
"epoch": 1.4595317725752508,
"grad_norm": 0.6127569079399109,
"learning_rate": 5.177847816298965e-05,
"loss": 1.1066,
"step": 1091
},
{
"epoch": 1.4608695652173913,
"grad_norm": 0.5010789632797241,
"learning_rate": 5.173345339936966e-05,
"loss": 1.103,
"step": 1092
},
{
"epoch": 1.4622073578595318,
"grad_norm": 0.6753023862838745,
"learning_rate": 5.168842863574967e-05,
"loss": 0.9127,
"step": 1093
},
{
"epoch": 1.4635451505016723,
"grad_norm": 0.6370785236358643,
"learning_rate": 5.164340387212967e-05,
"loss": 1.4349,
"step": 1094
},
{
"epoch": 1.4648829431438126,
"grad_norm": 0.5292558073997498,
"learning_rate": 5.1598379108509684e-05,
"loss": 1.0429,
"step": 1095
},
{
"epoch": 1.4662207357859531,
"grad_norm": 0.6851668953895569,
"learning_rate": 5.155335434488969e-05,
"loss": 1.0644,
"step": 1096
},
{
"epoch": 1.4675585284280936,
"grad_norm": 0.5523104071617126,
"learning_rate": 5.15083295812697e-05,
"loss": 1.436,
"step": 1097
},
{
"epoch": 1.4688963210702342,
"grad_norm": 0.44476109743118286,
"learning_rate": 5.14633048176497e-05,
"loss": 1.2197,
"step": 1098
},
{
"epoch": 1.4702341137123747,
"grad_norm": 0.6210783123970032,
"learning_rate": 5.141828005402972e-05,
"loss": 1.1594,
"step": 1099
},
{
"epoch": 1.471571906354515,
"grad_norm": 0.5576224327087402,
"learning_rate": 5.137325529040973e-05,
"loss": 1.379,
"step": 1100
},
{
"epoch": 1.4729096989966555,
"grad_norm": 0.7876810431480408,
"learning_rate": 5.1328230526789736e-05,
"loss": 1.3052,
"step": 1101
},
{
"epoch": 1.474247491638796,
"grad_norm": 0.5204209685325623,
"learning_rate": 5.128320576316975e-05,
"loss": 1.5625,
"step": 1102
},
{
"epoch": 1.4755852842809365,
"grad_norm": 0.6800068020820618,
"learning_rate": 5.123818099954976e-05,
"loss": 1.1098,
"step": 1103
},
{
"epoch": 1.476923076923077,
"grad_norm": 0.5767846703529358,
"learning_rate": 5.119315623592976e-05,
"loss": 1.3146,
"step": 1104
},
{
"epoch": 1.4782608695652173,
"grad_norm": 0.6460472941398621,
"learning_rate": 5.114813147230977e-05,
"loss": 1.5605,
"step": 1105
},
{
"epoch": 1.4795986622073578,
"grad_norm": 0.5904108285903931,
"learning_rate": 5.1103106708689785e-05,
"loss": 1.5344,
"step": 1106
},
{
"epoch": 1.4809364548494983,
"grad_norm": 0.5445310473442078,
"learning_rate": 5.1058081945069794e-05,
"loss": 1.4232,
"step": 1107
},
{
"epoch": 1.4822742474916388,
"grad_norm": 0.642361044883728,
"learning_rate": 5.1013057181449796e-05,
"loss": 1.259,
"step": 1108
},
{
"epoch": 1.4836120401337793,
"grad_norm": 0.5225979089736938,
"learning_rate": 5.096803241782981e-05,
"loss": 1.6267,
"step": 1109
},
{
"epoch": 1.4849498327759196,
"grad_norm": 0.467866450548172,
"learning_rate": 5.092300765420982e-05,
"loss": 1.5308,
"step": 1110
},
{
"epoch": 1.4862876254180601,
"grad_norm": 0.5182499289512634,
"learning_rate": 5.087798289058983e-05,
"loss": 1.6366,
"step": 1111
},
{
"epoch": 1.4876254180602007,
"grad_norm": 0.4528636932373047,
"learning_rate": 5.083295812696983e-05,
"loss": 1.7337,
"step": 1112
},
{
"epoch": 1.4889632107023412,
"grad_norm": 0.6056678295135498,
"learning_rate": 5.078793336334985e-05,
"loss": 1.6116,
"step": 1113
},
{
"epoch": 1.4903010033444817,
"grad_norm": 0.7801780700683594,
"learning_rate": 5.0742908599729854e-05,
"loss": 1.3719,
"step": 1114
},
{
"epoch": 1.491638795986622,
"grad_norm": 0.37030741572380066,
"learning_rate": 5.069788383610986e-05,
"loss": 1.3142,
"step": 1115
},
{
"epoch": 1.4929765886287625,
"grad_norm": 0.43180251121520996,
"learning_rate": 5.0652859072489864e-05,
"loss": 1.6445,
"step": 1116
},
{
"epoch": 1.494314381270903,
"grad_norm": 0.5785216689109802,
"learning_rate": 5.0607834308869886e-05,
"loss": 1.4886,
"step": 1117
},
{
"epoch": 1.4956521739130435,
"grad_norm": 0.6844804286956787,
"learning_rate": 5.056280954524989e-05,
"loss": 1.5056,
"step": 1118
},
{
"epoch": 1.496989966555184,
"grad_norm": 0.5532545447349548,
"learning_rate": 5.05177847816299e-05,
"loss": 1.4305,
"step": 1119
},
{
"epoch": 1.4983277591973243,
"grad_norm": 1.0775946378707886,
"learning_rate": 5.047276001800991e-05,
"loss": 1.1778,
"step": 1120
},
{
"epoch": 1.4996655518394648,
"grad_norm": 0.6081984043121338,
"learning_rate": 5.042773525438992e-05,
"loss": 1.4213,
"step": 1121
},
{
"epoch": 1.5010033444816053,
"grad_norm": 0.664418637752533,
"learning_rate": 5.038271049076992e-05,
"loss": 0.9045,
"step": 1122
},
{
"epoch": 1.5023411371237458,
"grad_norm": 0.7059425711631775,
"learning_rate": 5.033768572714993e-05,
"loss": 1.2277,
"step": 1123
},
{
"epoch": 1.5036789297658864,
"grad_norm": 0.5413299202919006,
"learning_rate": 5.0292660963529947e-05,
"loss": 1.5393,
"step": 1124
},
{
"epoch": 1.5050167224080266,
"grad_norm": 0.47964879870414734,
"learning_rate": 5.0247636199909955e-05,
"loss": 1.7589,
"step": 1125
},
{
"epoch": 1.5063545150501674,
"grad_norm": 0.6963114142417908,
"learning_rate": 5.020261143628996e-05,
"loss": 1.5982,
"step": 1126
},
{
"epoch": 1.5076923076923077,
"grad_norm": 0.567061722278595,
"learning_rate": 5.015758667266998e-05,
"loss": 0.9174,
"step": 1127
},
{
"epoch": 1.5090301003344482,
"grad_norm": 1.0022151470184326,
"learning_rate": 5.011256190904998e-05,
"loss": 1.0001,
"step": 1128
},
{
"epoch": 1.5103678929765887,
"grad_norm": 0.49206629395484924,
"learning_rate": 5.006753714542999e-05,
"loss": 1.7532,
"step": 1129
},
{
"epoch": 1.511705685618729,
"grad_norm": 0.6257492899894714,
"learning_rate": 5.002251238180999e-05,
"loss": 1.6988,
"step": 1130
},
{
"epoch": 1.5130434782608697,
"grad_norm": 0.4502313435077667,
"learning_rate": 4.997748761819001e-05,
"loss": 1.5575,
"step": 1131
},
{
"epoch": 1.51438127090301,
"grad_norm": 0.7328041195869446,
"learning_rate": 4.9932462854570015e-05,
"loss": 1.3456,
"step": 1132
},
{
"epoch": 1.5157190635451505,
"grad_norm": 0.5006700158119202,
"learning_rate": 4.9887438090950024e-05,
"loss": 1.962,
"step": 1133
},
{
"epoch": 1.517056856187291,
"grad_norm": 0.6119201183319092,
"learning_rate": 4.984241332733003e-05,
"loss": 1.2587,
"step": 1134
},
{
"epoch": 1.5183946488294313,
"grad_norm": 0.4837782382965088,
"learning_rate": 4.979738856371005e-05,
"loss": 1.5549,
"step": 1135
},
{
"epoch": 1.519732441471572,
"grad_norm": 0.7072848677635193,
"learning_rate": 4.975236380009005e-05,
"loss": 1.1933,
"step": 1136
},
{
"epoch": 1.5210702341137123,
"grad_norm": 0.9050494432449341,
"learning_rate": 4.9707339036470065e-05,
"loss": 0.9648,
"step": 1137
},
{
"epoch": 1.5224080267558529,
"grad_norm": 0.9395582675933838,
"learning_rate": 4.966231427285007e-05,
"loss": 1.3409,
"step": 1138
},
{
"epoch": 1.5237458193979934,
"grad_norm": 0.6518849730491638,
"learning_rate": 4.961728950923008e-05,
"loss": 1.3857,
"step": 1139
},
{
"epoch": 1.5250836120401337,
"grad_norm": 0.6363665461540222,
"learning_rate": 4.9572264745610084e-05,
"loss": 1.6827,
"step": 1140
},
{
"epoch": 1.5264214046822744,
"grad_norm": 0.6692220568656921,
"learning_rate": 4.95272399819901e-05,
"loss": 1.2984,
"step": 1141
},
{
"epoch": 1.5277591973244147,
"grad_norm": 0.4842371940612793,
"learning_rate": 4.94822152183701e-05,
"loss": 1.8894,
"step": 1142
},
{
"epoch": 1.5290969899665552,
"grad_norm": 0.6643278002738953,
"learning_rate": 4.9437190454750117e-05,
"loss": 1.4444,
"step": 1143
},
{
"epoch": 1.5304347826086957,
"grad_norm": 0.3972679674625397,
"learning_rate": 4.9392165691130125e-05,
"loss": 1.1412,
"step": 1144
},
{
"epoch": 1.531772575250836,
"grad_norm": 0.5554966926574707,
"learning_rate": 4.9347140927510134e-05,
"loss": 0.9472,
"step": 1145
},
{
"epoch": 1.5331103678929767,
"grad_norm": 0.6864800453186035,
"learning_rate": 4.930211616389014e-05,
"loss": 0.91,
"step": 1146
},
{
"epoch": 1.534448160535117,
"grad_norm": 0.45306283235549927,
"learning_rate": 4.925709140027015e-05,
"loss": 1.8113,
"step": 1147
},
{
"epoch": 1.5357859531772575,
"grad_norm": 0.6085485219955444,
"learning_rate": 4.921206663665016e-05,
"loss": 1.1981,
"step": 1148
},
{
"epoch": 1.537123745819398,
"grad_norm": 0.5183812975883484,
"learning_rate": 4.916704187303017e-05,
"loss": 1.6182,
"step": 1149
},
{
"epoch": 1.5384615384615383,
"grad_norm": 0.5374182462692261,
"learning_rate": 4.912201710941018e-05,
"loss": 1.467,
"step": 1150
},
{
"epoch": 1.539799331103679,
"grad_norm": 0.5360819101333618,
"learning_rate": 4.9076992345790185e-05,
"loss": 1.5612,
"step": 1151
},
{
"epoch": 1.5411371237458193,
"grad_norm": 0.5663394927978516,
"learning_rate": 4.9031967582170194e-05,
"loss": 0.9437,
"step": 1152
},
{
"epoch": 1.5424749163879599,
"grad_norm": 0.8458664417266846,
"learning_rate": 4.898694281855021e-05,
"loss": 1.1492,
"step": 1153
},
{
"epoch": 1.5438127090301004,
"grad_norm": 0.8276277184486389,
"learning_rate": 4.894191805493021e-05,
"loss": 1.2647,
"step": 1154
},
{
"epoch": 1.5451505016722407,
"grad_norm": 0.9039729833602905,
"learning_rate": 4.8896893291310226e-05,
"loss": 1.2929,
"step": 1155
},
{
"epoch": 1.5464882943143814,
"grad_norm": 0.63777095079422,
"learning_rate": 4.885186852769023e-05,
"loss": 1.3589,
"step": 1156
},
{
"epoch": 1.5478260869565217,
"grad_norm": 0.825036883354187,
"learning_rate": 4.8806843764070244e-05,
"loss": 1.233,
"step": 1157
},
{
"epoch": 1.5491638795986622,
"grad_norm": 0.5583025217056274,
"learning_rate": 4.8761819000450245e-05,
"loss": 1.8935,
"step": 1158
},
{
"epoch": 1.5505016722408027,
"grad_norm": 0.4850707948207855,
"learning_rate": 4.871679423683026e-05,
"loss": 1.5355,
"step": 1159
},
{
"epoch": 1.551839464882943,
"grad_norm": 0.5671193599700928,
"learning_rate": 4.867176947321026e-05,
"loss": 1.3592,
"step": 1160
},
{
"epoch": 1.5531772575250837,
"grad_norm": 0.6245211362838745,
"learning_rate": 4.862674470959028e-05,
"loss": 1.1835,
"step": 1161
},
{
"epoch": 1.554515050167224,
"grad_norm": 0.5358390808105469,
"learning_rate": 4.8581719945970286e-05,
"loss": 1.4703,
"step": 1162
},
{
"epoch": 1.5558528428093645,
"grad_norm": 0.7337758541107178,
"learning_rate": 4.8536695182350295e-05,
"loss": 1.0353,
"step": 1163
},
{
"epoch": 1.557190635451505,
"grad_norm": 0.5729077458381653,
"learning_rate": 4.8491670418730304e-05,
"loss": 1.495,
"step": 1164
},
{
"epoch": 1.5585284280936453,
"grad_norm": 0.5456716418266296,
"learning_rate": 4.844664565511031e-05,
"loss": 1.3311,
"step": 1165
},
{
"epoch": 1.559866220735786,
"grad_norm": 0.5501260757446289,
"learning_rate": 4.840162089149032e-05,
"loss": 0.8144,
"step": 1166
},
{
"epoch": 1.5612040133779264,
"grad_norm": 0.5962331295013428,
"learning_rate": 4.835659612787033e-05,
"loss": 1.5534,
"step": 1167
},
{
"epoch": 1.5625418060200669,
"grad_norm": 0.6607121825218201,
"learning_rate": 4.831157136425034e-05,
"loss": 1.2124,
"step": 1168
},
{
"epoch": 1.5638795986622074,
"grad_norm": 0.46615132689476013,
"learning_rate": 4.8266546600630347e-05,
"loss": 1.846,
"step": 1169
},
{
"epoch": 1.5652173913043477,
"grad_norm": 0.5132423639297485,
"learning_rate": 4.8221521837010355e-05,
"loss": 1.5602,
"step": 1170
},
{
"epoch": 1.5665551839464884,
"grad_norm": 0.5841774344444275,
"learning_rate": 4.817649707339037e-05,
"loss": 1.4619,
"step": 1171
},
{
"epoch": 1.5678929765886287,
"grad_norm": 0.5244891047477722,
"learning_rate": 4.813147230977037e-05,
"loss": 1.54,
"step": 1172
},
{
"epoch": 1.5692307692307692,
"grad_norm": 0.5597742199897766,
"learning_rate": 4.808644754615039e-05,
"loss": 1.6573,
"step": 1173
},
{
"epoch": 1.5705685618729097,
"grad_norm": 0.7487589716911316,
"learning_rate": 4.8041422782530396e-05,
"loss": 1.5697,
"step": 1174
},
{
"epoch": 1.57190635451505,
"grad_norm": 0.3977908790111542,
"learning_rate": 4.7996398018910405e-05,
"loss": 1.7967,
"step": 1175
},
{
"epoch": 1.5732441471571907,
"grad_norm": 0.5278039574623108,
"learning_rate": 4.7951373255290413e-05,
"loss": 1.4762,
"step": 1176
},
{
"epoch": 1.574581939799331,
"grad_norm": 0.6334072947502136,
"learning_rate": 4.790634849167042e-05,
"loss": 1.527,
"step": 1177
},
{
"epoch": 1.5759197324414715,
"grad_norm": 0.4393516182899475,
"learning_rate": 4.786132372805043e-05,
"loss": 1.1867,
"step": 1178
},
{
"epoch": 1.577257525083612,
"grad_norm": 1.0068846940994263,
"learning_rate": 4.781629896443044e-05,
"loss": 1.5931,
"step": 1179
},
{
"epoch": 1.5785953177257523,
"grad_norm": 0.53587406873703,
"learning_rate": 4.777127420081045e-05,
"loss": 0.9545,
"step": 1180
},
{
"epoch": 1.579933110367893,
"grad_norm": 0.39035940170288086,
"learning_rate": 4.7726249437190456e-05,
"loss": 1.255,
"step": 1181
},
{
"epoch": 1.5812709030100334,
"grad_norm": 0.4179837703704834,
"learning_rate": 4.7681224673570465e-05,
"loss": 1.7384,
"step": 1182
},
{
"epoch": 1.5826086956521739,
"grad_norm": 0.5482913255691528,
"learning_rate": 4.7636199909950474e-05,
"loss": 1.6497,
"step": 1183
},
{
"epoch": 1.5839464882943144,
"grad_norm": 0.7347548007965088,
"learning_rate": 4.759117514633049e-05,
"loss": 0.603,
"step": 1184
},
{
"epoch": 1.585284280936455,
"grad_norm": 0.6300942897796631,
"learning_rate": 4.754615038271049e-05,
"loss": 1.6236,
"step": 1185
},
{
"epoch": 1.5866220735785954,
"grad_norm": 0.39102619886398315,
"learning_rate": 4.7501125619090506e-05,
"loss": 1.2623,
"step": 1186
},
{
"epoch": 1.5879598662207357,
"grad_norm": 0.6081557273864746,
"learning_rate": 4.745610085547051e-05,
"loss": 1.4035,
"step": 1187
},
{
"epoch": 1.5892976588628762,
"grad_norm": 0.7682766318321228,
"learning_rate": 4.741107609185052e-05,
"loss": 1.4293,
"step": 1188
},
{
"epoch": 1.5906354515050167,
"grad_norm": 0.6367042660713196,
"learning_rate": 4.736605132823053e-05,
"loss": 0.9255,
"step": 1189
},
{
"epoch": 1.5919732441471572,
"grad_norm": 0.5380322337150574,
"learning_rate": 4.732102656461054e-05,
"loss": 1.5305,
"step": 1190
},
{
"epoch": 1.5933110367892978,
"grad_norm": 0.4204621911048889,
"learning_rate": 4.727600180099055e-05,
"loss": 1.7735,
"step": 1191
},
{
"epoch": 1.594648829431438,
"grad_norm": 0.6201006770133972,
"learning_rate": 4.723097703737056e-05,
"loss": 1.2954,
"step": 1192
},
{
"epoch": 1.5959866220735786,
"grad_norm": 0.5950494408607483,
"learning_rate": 4.7185952273750566e-05,
"loss": 1.3848,
"step": 1193
},
{
"epoch": 1.597324414715719,
"grad_norm": 0.5504522919654846,
"learning_rate": 4.7140927510130575e-05,
"loss": 1.5958,
"step": 1194
},
{
"epoch": 1.5986622073578596,
"grad_norm": 0.4658716320991516,
"learning_rate": 4.7095902746510583e-05,
"loss": 1.5919,
"step": 1195
},
{
"epoch": 1.6,
"grad_norm": 0.7538139820098877,
"learning_rate": 4.705087798289059e-05,
"loss": 1.4238,
"step": 1196
},
{
"epoch": 1.6013377926421404,
"grad_norm": 0.6065263152122498,
"learning_rate": 4.70058532192706e-05,
"loss": 1.4358,
"step": 1197
},
{
"epoch": 1.602675585284281,
"grad_norm": 0.481280118227005,
"learning_rate": 4.6960828455650616e-05,
"loss": 1.5438,
"step": 1198
},
{
"epoch": 1.6040133779264214,
"grad_norm": 0.5536462664604187,
"learning_rate": 4.691580369203062e-05,
"loss": 1.1397,
"step": 1199
},
{
"epoch": 1.605351170568562,
"grad_norm": 0.6622151732444763,
"learning_rate": 4.687077892841063e-05,
"loss": 0.8969,
"step": 1200
},
{
"epoch": 1.6066889632107024,
"grad_norm": 0.758965015411377,
"learning_rate": 4.6825754164790635e-05,
"loss": 1.1739,
"step": 1201
},
{
"epoch": 1.6080267558528427,
"grad_norm": 0.7237421274185181,
"learning_rate": 4.678072940117065e-05,
"loss": 1.1649,
"step": 1202
},
{
"epoch": 1.6093645484949832,
"grad_norm": 0.7297394871711731,
"learning_rate": 4.673570463755065e-05,
"loss": 1.9839,
"step": 1203
},
{
"epoch": 1.6107023411371237,
"grad_norm": 0.6325135231018066,
"learning_rate": 4.669067987393067e-05,
"loss": 1.4049,
"step": 1204
},
{
"epoch": 1.6120401337792643,
"grad_norm": 0.46899211406707764,
"learning_rate": 4.664565511031067e-05,
"loss": 1.7795,
"step": 1205
},
{
"epoch": 1.6133779264214048,
"grad_norm": 0.6146332025527954,
"learning_rate": 4.6600630346690685e-05,
"loss": 1.4186,
"step": 1206
},
{
"epoch": 1.614715719063545,
"grad_norm": 0.6527016758918762,
"learning_rate": 4.655560558307069e-05,
"loss": 1.5647,
"step": 1207
},
{
"epoch": 1.6160535117056856,
"grad_norm": 0.6520981192588806,
"learning_rate": 4.65105808194507e-05,
"loss": 1.5652,
"step": 1208
},
{
"epoch": 1.617391304347826,
"grad_norm": 0.6226683855056763,
"learning_rate": 4.646555605583071e-05,
"loss": 0.9986,
"step": 1209
},
{
"epoch": 1.6187290969899666,
"grad_norm": 0.6789349317550659,
"learning_rate": 4.642053129221072e-05,
"loss": 1.499,
"step": 1210
},
{
"epoch": 1.620066889632107,
"grad_norm": 0.5669140815734863,
"learning_rate": 4.637550652859073e-05,
"loss": 1.2055,
"step": 1211
},
{
"epoch": 1.6214046822742474,
"grad_norm": 0.5172103047370911,
"learning_rate": 4.6330481764970736e-05,
"loss": 1.4732,
"step": 1212
},
{
"epoch": 1.6227424749163881,
"grad_norm": 0.6144703030586243,
"learning_rate": 4.6285457001350745e-05,
"loss": 1.6759,
"step": 1213
},
{
"epoch": 1.6240802675585284,
"grad_norm": 0.5319096446037292,
"learning_rate": 4.624043223773075e-05,
"loss": 1.5875,
"step": 1214
},
{
"epoch": 1.625418060200669,
"grad_norm": 0.37053182721138,
"learning_rate": 4.619540747411076e-05,
"loss": 1.8184,
"step": 1215
},
{
"epoch": 1.6267558528428094,
"grad_norm": 0.5792490243911743,
"learning_rate": 4.615038271049078e-05,
"loss": 1.0232,
"step": 1216
},
{
"epoch": 1.6280936454849497,
"grad_norm": 0.4057672619819641,
"learning_rate": 4.610535794687078e-05,
"loss": 1.5673,
"step": 1217
},
{
"epoch": 1.6294314381270905,
"grad_norm": 0.49478086829185486,
"learning_rate": 4.6060333183250794e-05,
"loss": 1.7108,
"step": 1218
},
{
"epoch": 1.6307692307692307,
"grad_norm": 0.4574749171733856,
"learning_rate": 4.6015308419630796e-05,
"loss": 1.7347,
"step": 1219
},
{
"epoch": 1.6321070234113713,
"grad_norm": 0.6416162252426147,
"learning_rate": 4.597028365601081e-05,
"loss": 1.2321,
"step": 1220
},
{
"epoch": 1.6334448160535118,
"grad_norm": 0.6849283576011658,
"learning_rate": 4.5925258892390813e-05,
"loss": 1.7808,
"step": 1221
},
{
"epoch": 1.634782608695652,
"grad_norm": 0.639849841594696,
"learning_rate": 4.588023412877083e-05,
"loss": 1.4418,
"step": 1222
},
{
"epoch": 1.6361204013377928,
"grad_norm": 0.5444992780685425,
"learning_rate": 4.583520936515083e-05,
"loss": 1.3752,
"step": 1223
},
{
"epoch": 1.637458193979933,
"grad_norm": 0.47308918833732605,
"learning_rate": 4.5790184601530846e-05,
"loss": 1.396,
"step": 1224
},
{
"epoch": 1.6387959866220736,
"grad_norm": 0.5986432433128357,
"learning_rate": 4.5745159837910855e-05,
"loss": 1.548,
"step": 1225
},
{
"epoch": 1.640133779264214,
"grad_norm": 0.499828577041626,
"learning_rate": 4.570013507429086e-05,
"loss": 1.0529,
"step": 1226
},
{
"epoch": 1.6414715719063544,
"grad_norm": 1.0466457605361938,
"learning_rate": 4.565511031067087e-05,
"loss": 1.0108,
"step": 1227
},
{
"epoch": 1.6428093645484951,
"grad_norm": 0.5736595988273621,
"learning_rate": 4.561008554705088e-05,
"loss": 1.1951,
"step": 1228
},
{
"epoch": 1.6441471571906354,
"grad_norm": 0.5176533460617065,
"learning_rate": 4.556506078343089e-05,
"loss": 1.6408,
"step": 1229
},
{
"epoch": 1.645484949832776,
"grad_norm": 0.823546290397644,
"learning_rate": 4.55200360198109e-05,
"loss": 0.8908,
"step": 1230
},
{
"epoch": 1.6468227424749164,
"grad_norm": 0.6029638051986694,
"learning_rate": 4.5475011256190906e-05,
"loss": 1.3513,
"step": 1231
},
{
"epoch": 1.6481605351170567,
"grad_norm": 0.5313596129417419,
"learning_rate": 4.5429986492570915e-05,
"loss": 1.3504,
"step": 1232
},
{
"epoch": 1.6494983277591975,
"grad_norm": 0.4931012988090515,
"learning_rate": 4.538496172895092e-05,
"loss": 1.5954,
"step": 1233
},
{
"epoch": 1.6508361204013378,
"grad_norm": 0.6142311692237854,
"learning_rate": 4.533993696533094e-05,
"loss": 1.4441,
"step": 1234
},
{
"epoch": 1.6521739130434783,
"grad_norm": 0.5163131952285767,
"learning_rate": 4.529491220171094e-05,
"loss": 1.3505,
"step": 1235
},
{
"epoch": 1.6535117056856188,
"grad_norm": 0.4623430669307709,
"learning_rate": 4.5249887438090956e-05,
"loss": 1.5548,
"step": 1236
},
{
"epoch": 1.654849498327759,
"grad_norm": 0.7020666599273682,
"learning_rate": 4.520486267447096e-05,
"loss": 1.3642,
"step": 1237
},
{
"epoch": 1.6561872909698998,
"grad_norm": 0.6045846343040466,
"learning_rate": 4.515983791085097e-05,
"loss": 1.4035,
"step": 1238
},
{
"epoch": 1.65752508361204,
"grad_norm": 0.5127092003822327,
"learning_rate": 4.5114813147230975e-05,
"loss": 1.4297,
"step": 1239
},
{
"epoch": 1.6588628762541806,
"grad_norm": 0.37600478529930115,
"learning_rate": 4.506978838361099e-05,
"loss": 1.2362,
"step": 1240
},
{
"epoch": 1.6602006688963211,
"grad_norm": 0.8099076747894287,
"learning_rate": 4.502476361999099e-05,
"loss": 1.3902,
"step": 1241
},
{
"epoch": 1.6615384615384614,
"grad_norm": 0.6966046690940857,
"learning_rate": 4.497973885637101e-05,
"loss": 1.63,
"step": 1242
},
{
"epoch": 1.6628762541806021,
"grad_norm": 0.607402503490448,
"learning_rate": 4.4934714092751016e-05,
"loss": 1.5215,
"step": 1243
},
{
"epoch": 1.6642140468227424,
"grad_norm": 0.5723284482955933,
"learning_rate": 4.4889689329131025e-05,
"loss": 1.3004,
"step": 1244
},
{
"epoch": 1.665551839464883,
"grad_norm": 0.5554037690162659,
"learning_rate": 4.484466456551103e-05,
"loss": 1.3593,
"step": 1245
},
{
"epoch": 1.6668896321070235,
"grad_norm": 0.7105173468589783,
"learning_rate": 4.479963980189104e-05,
"loss": 1.2843,
"step": 1246
},
{
"epoch": 1.6682274247491637,
"grad_norm": 0.604043185710907,
"learning_rate": 4.475461503827105e-05,
"loss": 1.6447,
"step": 1247
},
{
"epoch": 1.6695652173913045,
"grad_norm": 0.6090781688690186,
"learning_rate": 4.470959027465106e-05,
"loss": 1.6954,
"step": 1248
},
{
"epoch": 1.6709030100334448,
"grad_norm": 0.5769763588905334,
"learning_rate": 4.466456551103107e-05,
"loss": 1.3212,
"step": 1249
},
{
"epoch": 1.6722408026755853,
"grad_norm": 0.5302979350090027,
"learning_rate": 4.4619540747411076e-05,
"loss": 1.6354,
"step": 1250
},
{
"epoch": 1.6735785953177258,
"grad_norm": 0.5180411338806152,
"learning_rate": 4.4574515983791085e-05,
"loss": 1.55,
"step": 1251
},
{
"epoch": 1.674916387959866,
"grad_norm": 0.8196110725402832,
"learning_rate": 4.45294912201711e-05,
"loss": 1.4945,
"step": 1252
},
{
"epoch": 1.6762541806020068,
"grad_norm": 0.4334547519683838,
"learning_rate": 4.44844664565511e-05,
"loss": 1.2096,
"step": 1253
},
{
"epoch": 1.677591973244147,
"grad_norm": 0.4448462128639221,
"learning_rate": 4.443944169293112e-05,
"loss": 1.1928,
"step": 1254
},
{
"epoch": 1.6789297658862876,
"grad_norm": 0.4100378155708313,
"learning_rate": 4.439441692931112e-05,
"loss": 1.8475,
"step": 1255
},
{
"epoch": 1.6802675585284281,
"grad_norm": 0.5713425278663635,
"learning_rate": 4.4349392165691134e-05,
"loss": 1.5135,
"step": 1256
},
{
"epoch": 1.6816053511705684,
"grad_norm": 0.42964500188827515,
"learning_rate": 4.430436740207114e-05,
"loss": 1.522,
"step": 1257
},
{
"epoch": 1.6829431438127092,
"grad_norm": 0.7717701196670532,
"learning_rate": 4.425934263845115e-05,
"loss": 1.3358,
"step": 1258
},
{
"epoch": 1.6842809364548494,
"grad_norm": 0.5887270569801331,
"learning_rate": 4.421431787483116e-05,
"loss": 1.272,
"step": 1259
},
{
"epoch": 1.68561872909699,
"grad_norm": 0.8065038323402405,
"learning_rate": 4.416929311121117e-05,
"loss": 0.8129,
"step": 1260
},
{
"epoch": 1.6869565217391305,
"grad_norm": 0.6046693325042725,
"learning_rate": 4.412426834759118e-05,
"loss": 1.5241,
"step": 1261
},
{
"epoch": 1.6882943143812708,
"grad_norm": 0.4960588812828064,
"learning_rate": 4.4079243583971186e-05,
"loss": 1.4725,
"step": 1262
},
{
"epoch": 1.6896321070234115,
"grad_norm": 0.6210783123970032,
"learning_rate": 4.4034218820351194e-05,
"loss": 1.2133,
"step": 1263
},
{
"epoch": 1.6909698996655518,
"grad_norm": 0.5066959857940674,
"learning_rate": 4.39891940567312e-05,
"loss": 0.7267,
"step": 1264
},
{
"epoch": 1.6923076923076923,
"grad_norm": 0.3616136610507965,
"learning_rate": 4.394416929311121e-05,
"loss": 1.4818,
"step": 1265
},
{
"epoch": 1.6936454849498328,
"grad_norm": 0.6788431406021118,
"learning_rate": 4.389914452949122e-05,
"loss": 1.4515,
"step": 1266
},
{
"epoch": 1.694983277591973,
"grad_norm": 0.5391234755516052,
"learning_rate": 4.385411976587123e-05,
"loss": 1.5275,
"step": 1267
},
{
"epoch": 1.6963210702341138,
"grad_norm": 0.6781142354011536,
"learning_rate": 4.380909500225124e-05,
"loss": 1.1785,
"step": 1268
},
{
"epoch": 1.6976588628762541,
"grad_norm": 0.45886972546577454,
"learning_rate": 4.376407023863125e-05,
"loss": 1.0886,
"step": 1269
},
{
"epoch": 1.6989966555183946,
"grad_norm": 0.6447398662567139,
"learning_rate": 4.371904547501126e-05,
"loss": 1.5353,
"step": 1270
},
{
"epoch": 1.7003344481605351,
"grad_norm": 0.6216405034065247,
"learning_rate": 4.367402071139127e-05,
"loss": 1.5214,
"step": 1271
},
{
"epoch": 1.7016722408026754,
"grad_norm": 0.6292665600776672,
"learning_rate": 4.362899594777128e-05,
"loss": 1.1899,
"step": 1272
},
{
"epoch": 1.7030100334448162,
"grad_norm": 0.7419617176055908,
"learning_rate": 4.358397118415129e-05,
"loss": 1.2945,
"step": 1273
},
{
"epoch": 1.7043478260869565,
"grad_norm": 0.49025335907936096,
"learning_rate": 4.3538946420531296e-05,
"loss": 1.7496,
"step": 1274
},
{
"epoch": 1.705685618729097,
"grad_norm": 0.659555196762085,
"learning_rate": 4.3493921656911304e-05,
"loss": 1.5394,
"step": 1275
},
{
"epoch": 1.7070234113712375,
"grad_norm": 0.7264639735221863,
"learning_rate": 4.344889689329131e-05,
"loss": 1.5763,
"step": 1276
},
{
"epoch": 1.7083612040133778,
"grad_norm": 0.9294725656509399,
"learning_rate": 4.340387212967132e-05,
"loss": 1.3261,
"step": 1277
},
{
"epoch": 1.7096989966555185,
"grad_norm": 0.6634767651557922,
"learning_rate": 4.335884736605133e-05,
"loss": 1.1431,
"step": 1278
},
{
"epoch": 1.7110367892976588,
"grad_norm": 1.48710036277771,
"learning_rate": 4.331382260243134e-05,
"loss": 1.3194,
"step": 1279
},
{
"epoch": 1.7123745819397993,
"grad_norm": 0.4782906770706177,
"learning_rate": 4.326879783881135e-05,
"loss": 1.8644,
"step": 1280
},
{
"epoch": 1.7137123745819398,
"grad_norm": 0.47573214769363403,
"learning_rate": 4.322377307519136e-05,
"loss": 1.7005,
"step": 1281
},
{
"epoch": 1.71505016722408,
"grad_norm": 0.5692914724349976,
"learning_rate": 4.3178748311571364e-05,
"loss": 1.1717,
"step": 1282
},
{
"epoch": 1.7163879598662208,
"grad_norm": 0.6902774572372437,
"learning_rate": 4.313372354795138e-05,
"loss": 0.8005,
"step": 1283
},
{
"epoch": 1.7177257525083611,
"grad_norm": 0.4654804766178131,
"learning_rate": 4.308869878433138e-05,
"loss": 1.8657,
"step": 1284
},
{
"epoch": 1.7190635451505016,
"grad_norm": 0.525980532169342,
"learning_rate": 4.30436740207114e-05,
"loss": 1.7472,
"step": 1285
},
{
"epoch": 1.7204013377926421,
"grad_norm": 3.5633068084716797,
"learning_rate": 4.29986492570914e-05,
"loss": 1.1038,
"step": 1286
},
{
"epoch": 1.7217391304347827,
"grad_norm": 0.6150617599487305,
"learning_rate": 4.2953624493471414e-05,
"loss": 1.5504,
"step": 1287
},
{
"epoch": 1.7230769230769232,
"grad_norm": 0.8399518728256226,
"learning_rate": 4.290859972985142e-05,
"loss": 0.5775,
"step": 1288
},
{
"epoch": 1.7244147157190635,
"grad_norm": 0.47936901450157166,
"learning_rate": 4.286357496623143e-05,
"loss": 1.5882,
"step": 1289
},
{
"epoch": 1.725752508361204,
"grad_norm": 0.7554563879966736,
"learning_rate": 4.281855020261144e-05,
"loss": 1.2668,
"step": 1290
},
{
"epoch": 1.7270903010033445,
"grad_norm": 0.6163820624351501,
"learning_rate": 4.277352543899145e-05,
"loss": 1.7666,
"step": 1291
},
{
"epoch": 1.728428093645485,
"grad_norm": 0.5814427733421326,
"learning_rate": 4.272850067537146e-05,
"loss": 1.3781,
"step": 1292
},
{
"epoch": 1.7297658862876255,
"grad_norm": 1.035375714302063,
"learning_rate": 4.2683475911751466e-05,
"loss": 1.128,
"step": 1293
},
{
"epoch": 1.7311036789297658,
"grad_norm": 0.4975506067276001,
"learning_rate": 4.2638451148131474e-05,
"loss": 1.2066,
"step": 1294
},
{
"epoch": 1.7324414715719063,
"grad_norm": 0.533572793006897,
"learning_rate": 4.259342638451148e-05,
"loss": 1.076,
"step": 1295
},
{
"epoch": 1.7337792642140468,
"grad_norm": 0.5955677628517151,
"learning_rate": 4.254840162089149e-05,
"loss": 1.6552,
"step": 1296
},
{
"epoch": 1.7351170568561873,
"grad_norm": 0.7461806535720825,
"learning_rate": 4.250337685727151e-05,
"loss": 1.2706,
"step": 1297
},
{
"epoch": 1.7364548494983278,
"grad_norm": 0.5620076060295105,
"learning_rate": 4.245835209365151e-05,
"loss": 1.8955,
"step": 1298
},
{
"epoch": 1.7377926421404681,
"grad_norm": 0.5752491354942322,
"learning_rate": 4.2413327330031524e-05,
"loss": 1.3457,
"step": 1299
},
{
"epoch": 1.7391304347826086,
"grad_norm": 0.6129872798919678,
"learning_rate": 4.2368302566411526e-05,
"loss": 1.2544,
"step": 1300
},
{
"epoch": 1.7404682274247492,
"grad_norm": 0.43887007236480713,
"learning_rate": 4.232327780279154e-05,
"loss": 1.8382,
"step": 1301
},
{
"epoch": 1.7418060200668897,
"grad_norm": 0.7264611124992371,
"learning_rate": 4.227825303917154e-05,
"loss": 1.0174,
"step": 1302
},
{
"epoch": 1.7431438127090302,
"grad_norm": 0.570402979850769,
"learning_rate": 4.223322827555156e-05,
"loss": 1.4562,
"step": 1303
},
{
"epoch": 1.7444816053511705,
"grad_norm": 0.6455773115158081,
"learning_rate": 4.218820351193156e-05,
"loss": 1.4974,
"step": 1304
},
{
"epoch": 1.745819397993311,
"grad_norm": 0.5429325699806213,
"learning_rate": 4.2143178748311576e-05,
"loss": 1.4276,
"step": 1305
},
{
"epoch": 1.7471571906354515,
"grad_norm": 0.5509850978851318,
"learning_rate": 4.2098153984691584e-05,
"loss": 1.436,
"step": 1306
},
{
"epoch": 1.748494983277592,
"grad_norm": 0.7787244915962219,
"learning_rate": 4.205312922107159e-05,
"loss": 0.9241,
"step": 1307
},
{
"epoch": 1.7498327759197325,
"grad_norm": 0.7184525728225708,
"learning_rate": 4.20081044574516e-05,
"loss": 1.0473,
"step": 1308
},
{
"epoch": 1.7511705685618728,
"grad_norm": 0.6182901859283447,
"learning_rate": 4.196307969383161e-05,
"loss": 1.6895,
"step": 1309
},
{
"epoch": 1.7525083612040135,
"grad_norm": 0.6682157516479492,
"learning_rate": 4.191805493021162e-05,
"loss": 0.8872,
"step": 1310
},
{
"epoch": 1.7538461538461538,
"grad_norm": 0.5342267155647278,
"learning_rate": 4.187303016659163e-05,
"loss": 1.5721,
"step": 1311
},
{
"epoch": 1.7551839464882943,
"grad_norm": 0.8530276417732239,
"learning_rate": 4.1828005402971636e-05,
"loss": 1.1826,
"step": 1312
},
{
"epoch": 1.7565217391304349,
"grad_norm": 0.5931107997894287,
"learning_rate": 4.1782980639351644e-05,
"loss": 1.4665,
"step": 1313
},
{
"epoch": 1.7578595317725751,
"grad_norm": 0.5947359204292297,
"learning_rate": 4.173795587573165e-05,
"loss": 0.9617,
"step": 1314
},
{
"epoch": 1.7591973244147159,
"grad_norm": 1.004927396774292,
"learning_rate": 4.169293111211167e-05,
"loss": 1.2389,
"step": 1315
},
{
"epoch": 1.7605351170568562,
"grad_norm": 0.5625318288803101,
"learning_rate": 4.164790634849167e-05,
"loss": 0.8856,
"step": 1316
},
{
"epoch": 1.7618729096989967,
"grad_norm": 0.6681520342826843,
"learning_rate": 4.1602881584871685e-05,
"loss": 1.0649,
"step": 1317
},
{
"epoch": 1.7632107023411372,
"grad_norm": 0.45618459582328796,
"learning_rate": 4.155785682125169e-05,
"loss": 1.3875,
"step": 1318
},
{
"epoch": 1.7645484949832775,
"grad_norm": 0.5873126983642578,
"learning_rate": 4.15128320576317e-05,
"loss": 1.1558,
"step": 1319
},
{
"epoch": 1.7658862876254182,
"grad_norm": 0.5614002346992493,
"learning_rate": 4.1467807294011704e-05,
"loss": 1.4631,
"step": 1320
},
{
"epoch": 1.7672240802675585,
"grad_norm": 0.5961179733276367,
"learning_rate": 4.142278253039172e-05,
"loss": 1.3621,
"step": 1321
},
{
"epoch": 1.768561872909699,
"grad_norm": 0.6985737085342407,
"learning_rate": 4.137775776677172e-05,
"loss": 0.8297,
"step": 1322
},
{
"epoch": 1.7698996655518395,
"grad_norm": 0.5106129050254822,
"learning_rate": 4.133273300315174e-05,
"loss": 1.7857,
"step": 1323
},
{
"epoch": 1.7712374581939798,
"grad_norm": 0.6078426837921143,
"learning_rate": 4.1287708239531745e-05,
"loss": 0.6776,
"step": 1324
},
{
"epoch": 1.7725752508361206,
"grad_norm": 0.7135141491889954,
"learning_rate": 4.1242683475911754e-05,
"loss": 1.3022,
"step": 1325
},
{
"epoch": 1.7739130434782608,
"grad_norm": 0.5969687104225159,
"learning_rate": 4.119765871229176e-05,
"loss": 1.2663,
"step": 1326
},
{
"epoch": 1.7752508361204014,
"grad_norm": 0.519609272480011,
"learning_rate": 4.115263394867177e-05,
"loss": 1.3102,
"step": 1327
},
{
"epoch": 1.7765886287625419,
"grad_norm": 0.907187283039093,
"learning_rate": 4.110760918505178e-05,
"loss": 0.7346,
"step": 1328
},
{
"epoch": 1.7779264214046822,
"grad_norm": 0.5560885667800903,
"learning_rate": 4.106258442143179e-05,
"loss": 1.4642,
"step": 1329
},
{
"epoch": 1.779264214046823,
"grad_norm": 0.701776921749115,
"learning_rate": 4.10175596578118e-05,
"loss": 1.2355,
"step": 1330
},
{
"epoch": 1.7806020066889632,
"grad_norm": 0.5565826892852783,
"learning_rate": 4.0972534894191806e-05,
"loss": 1.3008,
"step": 1331
},
{
"epoch": 1.7819397993311037,
"grad_norm": 0.6748178005218506,
"learning_rate": 4.0927510130571814e-05,
"loss": 1.2313,
"step": 1332
},
{
"epoch": 1.7832775919732442,
"grad_norm": 0.9739630222320557,
"learning_rate": 4.088248536695183e-05,
"loss": 1.4087,
"step": 1333
},
{
"epoch": 1.7846153846153845,
"grad_norm": 0.7524449825286865,
"learning_rate": 4.083746060333183e-05,
"loss": 1.4471,
"step": 1334
},
{
"epoch": 1.7859531772575252,
"grad_norm": 0.5693138241767883,
"learning_rate": 4.079243583971185e-05,
"loss": 1.7757,
"step": 1335
},
{
"epoch": 1.7872909698996655,
"grad_norm": 0.4918235242366791,
"learning_rate": 4.074741107609185e-05,
"loss": 1.5435,
"step": 1336
},
{
"epoch": 1.788628762541806,
"grad_norm": 0.39673417806625366,
"learning_rate": 4.0702386312471864e-05,
"loss": 1.6191,
"step": 1337
},
{
"epoch": 1.7899665551839465,
"grad_norm": 0.487466424703598,
"learning_rate": 4.0657361548851866e-05,
"loss": 1.6805,
"step": 1338
},
{
"epoch": 1.7913043478260868,
"grad_norm": 0.5584820508956909,
"learning_rate": 4.061233678523188e-05,
"loss": 0.7761,
"step": 1339
},
{
"epoch": 1.7926421404682276,
"grad_norm": 0.5846370458602905,
"learning_rate": 4.056731202161189e-05,
"loss": 1.2145,
"step": 1340
},
{
"epoch": 1.7939799331103679,
"grad_norm": 0.7206812500953674,
"learning_rate": 4.05222872579919e-05,
"loss": 1.1692,
"step": 1341
},
{
"epoch": 1.7953177257525084,
"grad_norm": 0.5454712510108948,
"learning_rate": 4.047726249437191e-05,
"loss": 1.0397,
"step": 1342
},
{
"epoch": 1.7966555183946489,
"grad_norm": 1.1169034242630005,
"learning_rate": 4.0432237730751915e-05,
"loss": 1.1179,
"step": 1343
},
{
"epoch": 1.7979933110367892,
"grad_norm": 0.5023759007453918,
"learning_rate": 4.0387212967131924e-05,
"loss": 1.4789,
"step": 1344
},
{
"epoch": 1.79933110367893,
"grad_norm": 1.1506235599517822,
"learning_rate": 4.034218820351193e-05,
"loss": 1.2992,
"step": 1345
},
{
"epoch": 1.8006688963210702,
"grad_norm": 0.44863954186439514,
"learning_rate": 4.029716343989194e-05,
"loss": 1.1604,
"step": 1346
},
{
"epoch": 1.8020066889632107,
"grad_norm": 0.6827251315116882,
"learning_rate": 4.025213867627195e-05,
"loss": 0.756,
"step": 1347
},
{
"epoch": 1.8033444816053512,
"grad_norm": 0.688870906829834,
"learning_rate": 4.020711391265196e-05,
"loss": 1.17,
"step": 1348
},
{
"epoch": 1.8046822742474915,
"grad_norm": 0.5064193606376648,
"learning_rate": 4.016208914903197e-05,
"loss": 1.2382,
"step": 1349
},
{
"epoch": 1.8060200668896322,
"grad_norm": 0.4319298565387726,
"learning_rate": 4.0117064385411976e-05,
"loss": 1.8041,
"step": 1350
},
{
"epoch": 1.8073578595317725,
"grad_norm": 0.5849002003669739,
"learning_rate": 4.007203962179199e-05,
"loss": 1.5873,
"step": 1351
},
{
"epoch": 1.808695652173913,
"grad_norm": 0.6112298369407654,
"learning_rate": 4.0027014858172e-05,
"loss": 1.5176,
"step": 1352
},
{
"epoch": 1.8100334448160535,
"grad_norm": 0.839145302772522,
"learning_rate": 3.998199009455201e-05,
"loss": 0.8495,
"step": 1353
},
{
"epoch": 1.8113712374581938,
"grad_norm": 0.6453105807304382,
"learning_rate": 3.993696533093202e-05,
"loss": 1.3811,
"step": 1354
},
{
"epoch": 1.8127090301003346,
"grad_norm": 0.5794005990028381,
"learning_rate": 3.9891940567312025e-05,
"loss": 1.411,
"step": 1355
},
{
"epoch": 1.8140468227424749,
"grad_norm": 0.6016961932182312,
"learning_rate": 3.9846915803692034e-05,
"loss": 1.2842,
"step": 1356
},
{
"epoch": 1.8153846153846154,
"grad_norm": 0.4260030686855316,
"learning_rate": 3.980189104007204e-05,
"loss": 1.8403,
"step": 1357
},
{
"epoch": 1.8167224080267559,
"grad_norm": 0.8129619359970093,
"learning_rate": 3.975686627645205e-05,
"loss": 1.4538,
"step": 1358
},
{
"epoch": 1.8180602006688962,
"grad_norm": 0.7668693661689758,
"learning_rate": 3.971184151283206e-05,
"loss": 1.329,
"step": 1359
},
{
"epoch": 1.819397993311037,
"grad_norm": 0.9522974491119385,
"learning_rate": 3.966681674921207e-05,
"loss": 1.4439,
"step": 1360
},
{
"epoch": 1.8207357859531772,
"grad_norm": 0.6404708027839661,
"learning_rate": 3.962179198559208e-05,
"loss": 1.2675,
"step": 1361
},
{
"epoch": 1.8220735785953177,
"grad_norm": 0.5133899450302124,
"learning_rate": 3.9576767221972085e-05,
"loss": 1.8803,
"step": 1362
},
{
"epoch": 1.8234113712374582,
"grad_norm": 0.5757606625556946,
"learning_rate": 3.9531742458352094e-05,
"loss": 1.4274,
"step": 1363
},
{
"epoch": 1.8247491638795985,
"grad_norm": 0.5684844255447388,
"learning_rate": 3.948671769473211e-05,
"loss": 1.5116,
"step": 1364
},
{
"epoch": 1.8260869565217392,
"grad_norm": 0.5964487791061401,
"learning_rate": 3.944169293111211e-05,
"loss": 1.0454,
"step": 1365
},
{
"epoch": 1.8274247491638795,
"grad_norm": 0.6252050399780273,
"learning_rate": 3.9396668167492126e-05,
"loss": 1.5453,
"step": 1366
},
{
"epoch": 1.82876254180602,
"grad_norm": 1.144151210784912,
"learning_rate": 3.935164340387213e-05,
"loss": 1.0064,
"step": 1367
},
{
"epoch": 1.8301003344481606,
"grad_norm": 0.716434121131897,
"learning_rate": 3.9306618640252144e-05,
"loss": 1.2945,
"step": 1368
},
{
"epoch": 1.8314381270903008,
"grad_norm": 0.9308441281318665,
"learning_rate": 3.926159387663215e-05,
"loss": 1.5683,
"step": 1369
},
{
"epoch": 1.8327759197324416,
"grad_norm": 0.5083184838294983,
"learning_rate": 3.921656911301216e-05,
"loss": 1.859,
"step": 1370
},
{
"epoch": 1.8341137123745819,
"grad_norm": 0.5718291997909546,
"learning_rate": 3.917154434939217e-05,
"loss": 1.5812,
"step": 1371
},
{
"epoch": 1.8354515050167224,
"grad_norm": 1.204251766204834,
"learning_rate": 3.912651958577218e-05,
"loss": 0.9265,
"step": 1372
},
{
"epoch": 1.836789297658863,
"grad_norm": 0.5010527968406677,
"learning_rate": 3.9081494822152187e-05,
"loss": 1.5297,
"step": 1373
},
{
"epoch": 1.8381270903010032,
"grad_norm": 0.5558527708053589,
"learning_rate": 3.9036470058532195e-05,
"loss": 1.5666,
"step": 1374
},
{
"epoch": 1.839464882943144,
"grad_norm": 0.66669762134552,
"learning_rate": 3.8991445294912204e-05,
"loss": 1.2004,
"step": 1375
},
{
"epoch": 1.8408026755852842,
"grad_norm": 0.5700245499610901,
"learning_rate": 3.894642053129221e-05,
"loss": 1.5689,
"step": 1376
},
{
"epoch": 1.8421404682274247,
"grad_norm": 0.4422018826007843,
"learning_rate": 3.890139576767222e-05,
"loss": 1.7777,
"step": 1377
},
{
"epoch": 1.8434782608695652,
"grad_norm": 0.6519413590431213,
"learning_rate": 3.8856371004052236e-05,
"loss": 1.372,
"step": 1378
},
{
"epoch": 1.8448160535117055,
"grad_norm": 0.591915488243103,
"learning_rate": 3.881134624043224e-05,
"loss": 1.1428,
"step": 1379
},
{
"epoch": 1.8461538461538463,
"grad_norm": 0.5624032020568848,
"learning_rate": 3.8766321476812253e-05,
"loss": 1.3378,
"step": 1380
},
{
"epoch": 1.8474916387959865,
"grad_norm": 0.5132156014442444,
"learning_rate": 3.8721296713192255e-05,
"loss": 1.5724,
"step": 1381
},
{
"epoch": 1.848829431438127,
"grad_norm": 0.49696585536003113,
"learning_rate": 3.867627194957227e-05,
"loss": 1.2503,
"step": 1382
},
{
"epoch": 1.8501672240802676,
"grad_norm": 0.6636870503425598,
"learning_rate": 3.863124718595227e-05,
"loss": 1.4385,
"step": 1383
},
{
"epoch": 1.851505016722408,
"grad_norm": 0.4480172097682953,
"learning_rate": 3.858622242233229e-05,
"loss": 1.7326,
"step": 1384
},
{
"epoch": 1.8528428093645486,
"grad_norm": 0.893785297870636,
"learning_rate": 3.854119765871229e-05,
"loss": 1.7338,
"step": 1385
},
{
"epoch": 1.8541806020066889,
"grad_norm": 0.41819554567337036,
"learning_rate": 3.8496172895092305e-05,
"loss": 1.1574,
"step": 1386
},
{
"epoch": 1.8555183946488294,
"grad_norm": 1.0737574100494385,
"learning_rate": 3.8451148131472314e-05,
"loss": 0.7989,
"step": 1387
},
{
"epoch": 1.85685618729097,
"grad_norm": 0.5694720149040222,
"learning_rate": 3.840612336785232e-05,
"loss": 1.5935,
"step": 1388
},
{
"epoch": 1.8581939799331104,
"grad_norm": 0.5654177069664001,
"learning_rate": 3.836109860423233e-05,
"loss": 0.8953,
"step": 1389
},
{
"epoch": 1.859531772575251,
"grad_norm": 0.5609742403030396,
"learning_rate": 3.831607384061234e-05,
"loss": 1.5239,
"step": 1390
},
{
"epoch": 1.8608695652173912,
"grad_norm": 0.6583299040794373,
"learning_rate": 3.827104907699235e-05,
"loss": 1.2912,
"step": 1391
},
{
"epoch": 1.8622073578595317,
"grad_norm": 0.5440747141838074,
"learning_rate": 3.8226024313372357e-05,
"loss": 1.6956,
"step": 1392
},
{
"epoch": 1.8635451505016722,
"grad_norm": 0.6128625869750977,
"learning_rate": 3.8180999549752365e-05,
"loss": 1.3703,
"step": 1393
},
{
"epoch": 1.8648829431438128,
"grad_norm": 0.643354058265686,
"learning_rate": 3.8135974786132374e-05,
"loss": 1.2531,
"step": 1394
},
{
"epoch": 1.8662207357859533,
"grad_norm": 0.6170260310173035,
"learning_rate": 3.809095002251238e-05,
"loss": 1.5198,
"step": 1395
},
{
"epoch": 1.8675585284280936,
"grad_norm": 0.4028455317020416,
"learning_rate": 3.80459252588924e-05,
"loss": 1.8615,
"step": 1396
},
{
"epoch": 1.868896321070234,
"grad_norm": 0.7505510449409485,
"learning_rate": 3.80009004952724e-05,
"loss": 1.2079,
"step": 1397
},
{
"epoch": 1.8702341137123746,
"grad_norm": 0.6865562796592712,
"learning_rate": 3.7955875731652415e-05,
"loss": 1.1446,
"step": 1398
},
{
"epoch": 1.871571906354515,
"grad_norm": 0.5812001824378967,
"learning_rate": 3.791085096803242e-05,
"loss": 1.4853,
"step": 1399
},
{
"epoch": 1.8729096989966556,
"grad_norm": 0.6272473335266113,
"learning_rate": 3.786582620441243e-05,
"loss": 1.1798,
"step": 1400
},
{
"epoch": 1.874247491638796,
"grad_norm": 0.5115850567817688,
"learning_rate": 3.7820801440792434e-05,
"loss": 1.2559,
"step": 1401
},
{
"epoch": 1.8755852842809364,
"grad_norm": 0.43357017636299133,
"learning_rate": 3.777577667717245e-05,
"loss": 0.9326,
"step": 1402
},
{
"epoch": 1.876923076923077,
"grad_norm": 0.5632050633430481,
"learning_rate": 3.773075191355245e-05,
"loss": 1.388,
"step": 1403
},
{
"epoch": 1.8782608695652174,
"grad_norm": 0.6570121049880981,
"learning_rate": 3.7685727149932466e-05,
"loss": 1.1719,
"step": 1404
},
{
"epoch": 1.879598662207358,
"grad_norm": 0.480442613363266,
"learning_rate": 3.7640702386312475e-05,
"loss": 1.6521,
"step": 1405
},
{
"epoch": 1.8809364548494982,
"grad_norm": 0.7593757510185242,
"learning_rate": 3.7595677622692484e-05,
"loss": 1.0407,
"step": 1406
},
{
"epoch": 1.8822742474916387,
"grad_norm": 0.6240037679672241,
"learning_rate": 3.755065285907249e-05,
"loss": 1.1501,
"step": 1407
},
{
"epoch": 1.8836120401337793,
"grad_norm": 0.48465749621391296,
"learning_rate": 3.75056280954525e-05,
"loss": 1.6513,
"step": 1408
},
{
"epoch": 1.8849498327759198,
"grad_norm": 0.7941508889198303,
"learning_rate": 3.746060333183251e-05,
"loss": 1.3305,
"step": 1409
},
{
"epoch": 1.8862876254180603,
"grad_norm": 0.574839174747467,
"learning_rate": 3.741557856821252e-05,
"loss": 1.4869,
"step": 1410
},
{
"epoch": 1.8876254180602006,
"grad_norm": 0.6008003354072571,
"learning_rate": 3.7370553804592526e-05,
"loss": 1.5017,
"step": 1411
},
{
"epoch": 1.8889632107023413,
"grad_norm": 0.5762261152267456,
"learning_rate": 3.7325529040972535e-05,
"loss": 0.7915,
"step": 1412
},
{
"epoch": 1.8903010033444816,
"grad_norm": 0.6438276767730713,
"learning_rate": 3.7280504277352544e-05,
"loss": 1.5426,
"step": 1413
},
{
"epoch": 1.891638795986622,
"grad_norm": 0.5164512395858765,
"learning_rate": 3.723547951373256e-05,
"loss": 1.1212,
"step": 1414
},
{
"epoch": 1.8929765886287626,
"grad_norm": 0.5601268410682678,
"learning_rate": 3.719045475011256e-05,
"loss": 1.6185,
"step": 1415
},
{
"epoch": 1.894314381270903,
"grad_norm": 0.6093222498893738,
"learning_rate": 3.7145429986492576e-05,
"loss": 1.4573,
"step": 1416
},
{
"epoch": 1.8956521739130436,
"grad_norm": 0.6153028607368469,
"learning_rate": 3.710040522287258e-05,
"loss": 1.4644,
"step": 1417
},
{
"epoch": 1.896989966555184,
"grad_norm": 0.5739652514457703,
"learning_rate": 3.705538045925259e-05,
"loss": 1.4324,
"step": 1418
},
{
"epoch": 1.8983277591973244,
"grad_norm": 0.6016227602958679,
"learning_rate": 3.7010355695632595e-05,
"loss": 1.7096,
"step": 1419
},
{
"epoch": 1.899665551839465,
"grad_norm": 0.9013569355010986,
"learning_rate": 3.696533093201261e-05,
"loss": 1.0067,
"step": 1420
},
{
"epoch": 1.9010033444816052,
"grad_norm": 0.7423526644706726,
"learning_rate": 3.692030616839261e-05,
"loss": 1.1336,
"step": 1421
},
{
"epoch": 1.902341137123746,
"grad_norm": 0.5514466762542725,
"learning_rate": 3.687528140477263e-05,
"loss": 1.097,
"step": 1422
},
{
"epoch": 1.9036789297658863,
"grad_norm": 0.6179841160774231,
"learning_rate": 3.6830256641152636e-05,
"loss": 1.5022,
"step": 1423
},
{
"epoch": 1.9050167224080268,
"grad_norm": 0.48712772130966187,
"learning_rate": 3.6785231877532645e-05,
"loss": 1.0286,
"step": 1424
},
{
"epoch": 1.9063545150501673,
"grad_norm": 0.6644452214241028,
"learning_rate": 3.6740207113912653e-05,
"loss": 1.4727,
"step": 1425
},
{
"epoch": 1.9076923076923076,
"grad_norm": 0.6408343315124512,
"learning_rate": 3.669518235029266e-05,
"loss": 1.5816,
"step": 1426
},
{
"epoch": 1.9090301003344483,
"grad_norm": 0.6295036673545837,
"learning_rate": 3.665015758667267e-05,
"loss": 1.5896,
"step": 1427
},
{
"epoch": 1.9103678929765886,
"grad_norm": 0.6161107420921326,
"learning_rate": 3.660513282305268e-05,
"loss": 1.397,
"step": 1428
},
{
"epoch": 1.9117056856187291,
"grad_norm": 0.8119416832923889,
"learning_rate": 3.656010805943269e-05,
"loss": 1.6998,
"step": 1429
},
{
"epoch": 1.9130434782608696,
"grad_norm": 0.504453718662262,
"learning_rate": 3.6515083295812696e-05,
"loss": 1.4708,
"step": 1430
},
{
"epoch": 1.91438127090301,
"grad_norm": 0.5680826902389526,
"learning_rate": 3.6470058532192705e-05,
"loss": 1.416,
"step": 1431
},
{
"epoch": 1.9157190635451506,
"grad_norm": 0.5165196657180786,
"learning_rate": 3.642503376857272e-05,
"loss": 1.153,
"step": 1432
},
{
"epoch": 1.917056856187291,
"grad_norm": 0.640468418598175,
"learning_rate": 3.638000900495272e-05,
"loss": 1.3228,
"step": 1433
},
{
"epoch": 1.9183946488294314,
"grad_norm": 0.6464949250221252,
"learning_rate": 3.633498424133274e-05,
"loss": 1.5836,
"step": 1434
},
{
"epoch": 1.919732441471572,
"grad_norm": 0.48071908950805664,
"learning_rate": 3.628995947771274e-05,
"loss": 1.7179,
"step": 1435
},
{
"epoch": 1.9210702341137122,
"grad_norm": 0.3788856565952301,
"learning_rate": 3.6244934714092755e-05,
"loss": 1.8167,
"step": 1436
},
{
"epoch": 1.922408026755853,
"grad_norm": 0.563204824924469,
"learning_rate": 3.619990995047276e-05,
"loss": 1.5081,
"step": 1437
},
{
"epoch": 1.9237458193979933,
"grad_norm": 0.4034525156021118,
"learning_rate": 3.615488518685277e-05,
"loss": 1.2595,
"step": 1438
},
{
"epoch": 1.9250836120401338,
"grad_norm": 0.7248520255088806,
"learning_rate": 3.610986042323278e-05,
"loss": 1.7668,
"step": 1439
},
{
"epoch": 1.9264214046822743,
"grad_norm": 0.612653911113739,
"learning_rate": 3.606483565961279e-05,
"loss": 1.294,
"step": 1440
},
{
"epoch": 1.9277591973244146,
"grad_norm": 0.49420690536499023,
"learning_rate": 3.60198108959928e-05,
"loss": 1.5397,
"step": 1441
},
{
"epoch": 1.9290969899665553,
"grad_norm": 0.679892897605896,
"learning_rate": 3.5974786132372806e-05,
"loss": 1.0175,
"step": 1442
},
{
"epoch": 1.9304347826086956,
"grad_norm": 0.44175776839256287,
"learning_rate": 3.5929761368752815e-05,
"loss": 1.7368,
"step": 1443
},
{
"epoch": 1.9317725752508361,
"grad_norm": 0.634681224822998,
"learning_rate": 3.5884736605132823e-05,
"loss": 1.4975,
"step": 1444
},
{
"epoch": 1.9331103678929766,
"grad_norm": 0.5663192272186279,
"learning_rate": 3.583971184151283e-05,
"loss": 1.4068,
"step": 1445
},
{
"epoch": 1.934448160535117,
"grad_norm": 0.6057150959968567,
"learning_rate": 3.579468707789284e-05,
"loss": 1.2817,
"step": 1446
},
{
"epoch": 1.9357859531772577,
"grad_norm": 0.595527708530426,
"learning_rate": 3.5749662314272856e-05,
"loss": 1.6711,
"step": 1447
},
{
"epoch": 1.937123745819398,
"grad_norm": 0.5907280445098877,
"learning_rate": 3.570463755065286e-05,
"loss": 1.5177,
"step": 1448
},
{
"epoch": 1.9384615384615385,
"grad_norm": 0.39169877767562866,
"learning_rate": 3.565961278703287e-05,
"loss": 2.0893,
"step": 1449
},
{
"epoch": 1.939799331103679,
"grad_norm": 0.5973325967788696,
"learning_rate": 3.561458802341288e-05,
"loss": 1.421,
"step": 1450
},
{
"epoch": 1.9411371237458193,
"grad_norm": 0.6645547151565552,
"learning_rate": 3.556956325979289e-05,
"loss": 1.3091,
"step": 1451
},
{
"epoch": 1.94247491638796,
"grad_norm": 0.7206888198852539,
"learning_rate": 3.55245384961729e-05,
"loss": 1.5102,
"step": 1452
},
{
"epoch": 1.9438127090301003,
"grad_norm": 0.5224019289016724,
"learning_rate": 3.547951373255291e-05,
"loss": 1.5029,
"step": 1453
},
{
"epoch": 1.9451505016722408,
"grad_norm": 0.8365053534507751,
"learning_rate": 3.5434488968932916e-05,
"loss": 1.2303,
"step": 1454
},
{
"epoch": 1.9464882943143813,
"grad_norm": 0.6158286929130554,
"learning_rate": 3.5389464205312925e-05,
"loss": 1.2886,
"step": 1455
},
{
"epoch": 1.9478260869565216,
"grad_norm": 0.7675470113754272,
"learning_rate": 3.534443944169293e-05,
"loss": 1.0725,
"step": 1456
},
{
"epoch": 1.9491638795986623,
"grad_norm": 1.0587267875671387,
"learning_rate": 3.529941467807294e-05,
"loss": 1.0416,
"step": 1457
},
{
"epoch": 1.9505016722408026,
"grad_norm": 0.7863774299621582,
"learning_rate": 3.525438991445295e-05,
"loss": 1.0111,
"step": 1458
},
{
"epoch": 1.9518394648829431,
"grad_norm": 0.47847363352775574,
"learning_rate": 3.5209365150832966e-05,
"loss": 1.6936,
"step": 1459
},
{
"epoch": 1.9531772575250836,
"grad_norm": 0.5850685834884644,
"learning_rate": 3.516434038721297e-05,
"loss": 1.5939,
"step": 1460
},
{
"epoch": 1.954515050167224,
"grad_norm": 0.47978919744491577,
"learning_rate": 3.511931562359298e-05,
"loss": 1.8567,
"step": 1461
},
{
"epoch": 1.9558528428093647,
"grad_norm": 0.5951332449913025,
"learning_rate": 3.5074290859972985e-05,
"loss": 1.0255,
"step": 1462
},
{
"epoch": 1.957190635451505,
"grad_norm": 0.8558800220489502,
"learning_rate": 3.5029266096353e-05,
"loss": 1.1657,
"step": 1463
},
{
"epoch": 1.9585284280936455,
"grad_norm": 0.4947897791862488,
"learning_rate": 3.4984241332733e-05,
"loss": 1.6925,
"step": 1464
},
{
"epoch": 1.959866220735786,
"grad_norm": 0.48866933584213257,
"learning_rate": 3.493921656911302e-05,
"loss": 1.6474,
"step": 1465
},
{
"epoch": 1.9612040133779263,
"grad_norm": 0.6075807213783264,
"learning_rate": 3.489419180549302e-05,
"loss": 1.3366,
"step": 1466
},
{
"epoch": 1.962541806020067,
"grad_norm": 0.6691599488258362,
"learning_rate": 3.4849167041873035e-05,
"loss": 1.3393,
"step": 1467
},
{
"epoch": 1.9638795986622073,
"grad_norm": 0.656430721282959,
"learning_rate": 3.480414227825304e-05,
"loss": 1.2173,
"step": 1468
},
{
"epoch": 1.9652173913043478,
"grad_norm": 0.540453314781189,
"learning_rate": 3.475911751463305e-05,
"loss": 1.3623,
"step": 1469
},
{
"epoch": 1.9665551839464883,
"grad_norm": 0.6606508493423462,
"learning_rate": 3.471409275101306e-05,
"loss": 1.2194,
"step": 1470
},
{
"epoch": 1.9678929765886286,
"grad_norm": 0.6008053421974182,
"learning_rate": 3.466906798739307e-05,
"loss": 1.3245,
"step": 1471
},
{
"epoch": 1.9692307692307693,
"grad_norm": 0.7815794348716736,
"learning_rate": 3.462404322377308e-05,
"loss": 1.3401,
"step": 1472
},
{
"epoch": 1.9705685618729096,
"grad_norm": 0.4332338869571686,
"learning_rate": 3.4579018460153086e-05,
"loss": 1.6547,
"step": 1473
},
{
"epoch": 1.9719063545150501,
"grad_norm": 0.7275242209434509,
"learning_rate": 3.4533993696533095e-05,
"loss": 1.5228,
"step": 1474
},
{
"epoch": 1.9732441471571907,
"grad_norm": 0.5694137811660767,
"learning_rate": 3.44889689329131e-05,
"loss": 1.3479,
"step": 1475
},
{
"epoch": 1.974581939799331,
"grad_norm": 0.529672384262085,
"learning_rate": 3.444394416929311e-05,
"loss": 1.4186,
"step": 1476
},
{
"epoch": 1.9759197324414717,
"grad_norm": 0.6662195920944214,
"learning_rate": 3.439891940567313e-05,
"loss": 1.3579,
"step": 1477
},
{
"epoch": 1.977257525083612,
"grad_norm": 0.47519707679748535,
"learning_rate": 3.435389464205313e-05,
"loss": 1.2978,
"step": 1478
},
{
"epoch": 1.9785953177257525,
"grad_norm": 1.7963173389434814,
"learning_rate": 3.4308869878433144e-05,
"loss": 1.3871,
"step": 1479
},
{
"epoch": 1.979933110367893,
"grad_norm": 0.5129130482673645,
"learning_rate": 3.4263845114813146e-05,
"loss": 1.5933,
"step": 1480
},
{
"epoch": 1.9812709030100333,
"grad_norm": 0.5007113218307495,
"learning_rate": 3.421882035119316e-05,
"loss": 1.1155,
"step": 1481
},
{
"epoch": 1.982608695652174,
"grad_norm": 0.46495723724365234,
"learning_rate": 3.417379558757316e-05,
"loss": 1.408,
"step": 1482
},
{
"epoch": 1.9839464882943143,
"grad_norm": 0.5745284557342529,
"learning_rate": 3.412877082395318e-05,
"loss": 1.6946,
"step": 1483
},
{
"epoch": 1.9852842809364548,
"grad_norm": 0.48399198055267334,
"learning_rate": 3.408374606033318e-05,
"loss": 1.5874,
"step": 1484
},
{
"epoch": 1.9866220735785953,
"grad_norm": 0.5783731937408447,
"learning_rate": 3.4038721296713196e-05,
"loss": 1.537,
"step": 1485
},
{
"epoch": 1.9879598662207358,
"grad_norm": 0.6687440276145935,
"learning_rate": 3.3993696533093204e-05,
"loss": 1.5465,
"step": 1486
},
{
"epoch": 1.9892976588628764,
"grad_norm": 0.5151572227478027,
"learning_rate": 3.394867176947321e-05,
"loss": 1.471,
"step": 1487
},
{
"epoch": 1.9906354515050166,
"grad_norm": 0.5424696207046509,
"learning_rate": 3.390364700585322e-05,
"loss": 1.5837,
"step": 1488
},
{
"epoch": 1.9919732441471572,
"grad_norm": 0.6752240061759949,
"learning_rate": 3.385862224223323e-05,
"loss": 1.3542,
"step": 1489
},
{
"epoch": 1.9933110367892977,
"grad_norm": 0.9526898264884949,
"learning_rate": 3.381359747861324e-05,
"loss": 1.255,
"step": 1490
},
{
"epoch": 1.9946488294314382,
"grad_norm": 0.5398581624031067,
"learning_rate": 3.376857271499325e-05,
"loss": 1.468,
"step": 1491
},
{
"epoch": 1.9959866220735787,
"grad_norm": 0.5538991093635559,
"learning_rate": 3.3723547951373256e-05,
"loss": 1.5927,
"step": 1492
},
{
"epoch": 1.997324414715719,
"grad_norm": 0.5554574728012085,
"learning_rate": 3.3678523187753265e-05,
"loss": 1.3998,
"step": 1493
},
{
"epoch": 1.9986622073578595,
"grad_norm": 0.7692258358001709,
"learning_rate": 3.363349842413327e-05,
"loss": 1.0331,
"step": 1494
},
{
"epoch": 2.0,
"grad_norm": 0.6130909323692322,
"learning_rate": 3.358847366051329e-05,
"loss": 1.4259,
"step": 1495
},
{
"epoch": 2.0013377926421403,
"grad_norm": 0.8359543085098267,
"learning_rate": 3.354344889689329e-05,
"loss": 0.8013,
"step": 1496
},
{
"epoch": 2.002675585284281,
"grad_norm": 0.4600941240787506,
"learning_rate": 3.3498424133273306e-05,
"loss": 0.3069,
"step": 1497
},
{
"epoch": 2.0040133779264213,
"grad_norm": 0.6617846488952637,
"learning_rate": 3.345339936965331e-05,
"loss": 0.604,
"step": 1498
},
{
"epoch": 2.005351170568562,
"grad_norm": 0.7170037627220154,
"learning_rate": 3.340837460603332e-05,
"loss": 0.7151,
"step": 1499
},
{
"epoch": 2.0066889632107023,
"grad_norm": 0.6578344106674194,
"learning_rate": 3.3363349842413325e-05,
"loss": 1.1236,
"step": 1500
}
],
"logging_steps": 1,
"max_steps": 2241,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 5.694569146514203e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}