PEFT
Safetensors
re400 / trainer_state.json
chen
Upload folder using huggingface_hub
6328235 verified
Raw
History Blame Contribute Delete
69.8 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.094488188976378,
"eval_steps": 500,
"global_step": 400,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.005249343832020997,
"grad_norm": 4.7173380851745605,
"learning_rate": 0.0,
"loss": 2.0177,
"step": 1
},
{
"epoch": 0.010498687664041995,
"grad_norm": 4.437033653259277,
"learning_rate": 3.5087719298245615e-06,
"loss": 1.7211,
"step": 2
},
{
"epoch": 0.015748031496062992,
"grad_norm": 4.145833969116211,
"learning_rate": 7.017543859649123e-06,
"loss": 1.6452,
"step": 3
},
{
"epoch": 0.02099737532808399,
"grad_norm": 2.83351731300354,
"learning_rate": 1.0526315789473684e-05,
"loss": 1.4565,
"step": 4
},
{
"epoch": 0.026246719160104987,
"grad_norm": 4.269916534423828,
"learning_rate": 1.4035087719298246e-05,
"loss": 1.8712,
"step": 5
},
{
"epoch": 0.031496062992125984,
"grad_norm": 3.0349326133728027,
"learning_rate": 1.7543859649122806e-05,
"loss": 1.9409,
"step": 6
},
{
"epoch": 0.03674540682414698,
"grad_norm": 4.1207780838012695,
"learning_rate": 2.105263157894737e-05,
"loss": 1.7171,
"step": 7
},
{
"epoch": 0.04199475065616798,
"grad_norm": 3.790914297103882,
"learning_rate": 2.456140350877193e-05,
"loss": 1.7406,
"step": 8
},
{
"epoch": 0.047244094488188976,
"grad_norm": 3.1181528568267822,
"learning_rate": 2.8070175438596492e-05,
"loss": 1.9785,
"step": 9
},
{
"epoch": 0.05249343832020997,
"grad_norm": 3.166116952896118,
"learning_rate": 3.157894736842105e-05,
"loss": 1.3592,
"step": 10
},
{
"epoch": 0.05774278215223097,
"grad_norm": 3.287881851196289,
"learning_rate": 3.508771929824561e-05,
"loss": 1.8857,
"step": 11
},
{
"epoch": 0.06299212598425197,
"grad_norm": 3.340050458908081,
"learning_rate": 3.859649122807018e-05,
"loss": 1.6225,
"step": 12
},
{
"epoch": 0.06824146981627296,
"grad_norm": 3.426025152206421,
"learning_rate": 4.210526315789474e-05,
"loss": 1.5674,
"step": 13
},
{
"epoch": 0.07349081364829396,
"grad_norm": 3.3121087551116943,
"learning_rate": 4.56140350877193e-05,
"loss": 1.4491,
"step": 14
},
{
"epoch": 0.07874015748031496,
"grad_norm": 3.173163890838623,
"learning_rate": 4.912280701754386e-05,
"loss": 1.4164,
"step": 15
},
{
"epoch": 0.08398950131233596,
"grad_norm": 3.2732391357421875,
"learning_rate": 5.2631578947368424e-05,
"loss": 1.5818,
"step": 16
},
{
"epoch": 0.08923884514435695,
"grad_norm": 4.392367362976074,
"learning_rate": 5.6140350877192984e-05,
"loss": 1.5491,
"step": 17
},
{
"epoch": 0.09448818897637795,
"grad_norm": 4.179324626922607,
"learning_rate": 5.9649122807017544e-05,
"loss": 1.6327,
"step": 18
},
{
"epoch": 0.09973753280839895,
"grad_norm": 3.5691120624542236,
"learning_rate": 6.31578947368421e-05,
"loss": 1.6418,
"step": 19
},
{
"epoch": 0.10498687664041995,
"grad_norm": 2.7537472248077393,
"learning_rate": 6.666666666666667e-05,
"loss": 1.3472,
"step": 20
},
{
"epoch": 0.11023622047244094,
"grad_norm": 4.199432373046875,
"learning_rate": 7.017543859649122e-05,
"loss": 1.3632,
"step": 21
},
{
"epoch": 0.11548556430446194,
"grad_norm": 4.307034969329834,
"learning_rate": 7.368421052631579e-05,
"loss": 1.3628,
"step": 22
},
{
"epoch": 0.12073490813648294,
"grad_norm": 3.5497403144836426,
"learning_rate": 7.719298245614036e-05,
"loss": 1.4816,
"step": 23
},
{
"epoch": 0.12598425196850394,
"grad_norm": 3.31937575340271,
"learning_rate": 8.070175438596491e-05,
"loss": 1.3599,
"step": 24
},
{
"epoch": 0.13123359580052493,
"grad_norm": 2.710407257080078,
"learning_rate": 8.421052631578948e-05,
"loss": 1.0672,
"step": 25
},
{
"epoch": 0.13648293963254593,
"grad_norm": 3.317232847213745,
"learning_rate": 8.771929824561403e-05,
"loss": 1.3396,
"step": 26
},
{
"epoch": 0.14173228346456693,
"grad_norm": 4.023993492126465,
"learning_rate": 9.12280701754386e-05,
"loss": 1.6528,
"step": 27
},
{
"epoch": 0.14698162729658792,
"grad_norm": 3.469362735748291,
"learning_rate": 9.473684210526316e-05,
"loss": 1.4297,
"step": 28
},
{
"epoch": 0.15223097112860892,
"grad_norm": 4.806138515472412,
"learning_rate": 9.824561403508771e-05,
"loss": 1.0343,
"step": 29
},
{
"epoch": 0.15748031496062992,
"grad_norm": 3.6015982627868652,
"learning_rate": 0.0001017543859649123,
"loss": 1.6401,
"step": 30
},
{
"epoch": 0.16272965879265092,
"grad_norm": 3.79402232170105,
"learning_rate": 0.00010526315789473685,
"loss": 1.6519,
"step": 31
},
{
"epoch": 0.1679790026246719,
"grad_norm": 4.123412132263184,
"learning_rate": 0.00010877192982456141,
"loss": 1.4302,
"step": 32
},
{
"epoch": 0.1732283464566929,
"grad_norm": 2.3286683559417725,
"learning_rate": 0.00011228070175438597,
"loss": 1.2029,
"step": 33
},
{
"epoch": 0.1784776902887139,
"grad_norm": 2.6430580615997314,
"learning_rate": 0.00011578947368421053,
"loss": 1.4423,
"step": 34
},
{
"epoch": 0.1837270341207349,
"grad_norm": 3.4652087688446045,
"learning_rate": 0.00011929824561403509,
"loss": 1.1598,
"step": 35
},
{
"epoch": 0.1889763779527559,
"grad_norm": 4.649867534637451,
"learning_rate": 0.00012280701754385965,
"loss": 1.3339,
"step": 36
},
{
"epoch": 0.1942257217847769,
"grad_norm": 2.450688123703003,
"learning_rate": 0.0001263157894736842,
"loss": 1.2083,
"step": 37
},
{
"epoch": 0.1994750656167979,
"grad_norm": 2.2536394596099854,
"learning_rate": 0.0001298245614035088,
"loss": 1.2621,
"step": 38
},
{
"epoch": 0.2047244094488189,
"grad_norm": 2.371548652648926,
"learning_rate": 0.00013333333333333334,
"loss": 1.198,
"step": 39
},
{
"epoch": 0.2099737532808399,
"grad_norm": 2.5427513122558594,
"learning_rate": 0.0001368421052631579,
"loss": 1.3013,
"step": 40
},
{
"epoch": 0.2152230971128609,
"grad_norm": 2.5822794437408447,
"learning_rate": 0.00014035087719298245,
"loss": 1.432,
"step": 41
},
{
"epoch": 0.2204724409448819,
"grad_norm": 2.5134873390197754,
"learning_rate": 0.00014385964912280703,
"loss": 1.2649,
"step": 42
},
{
"epoch": 0.22572178477690288,
"grad_norm": 2.4449403285980225,
"learning_rate": 0.00014736842105263158,
"loss": 1.2715,
"step": 43
},
{
"epoch": 0.23097112860892388,
"grad_norm": 2.940073013305664,
"learning_rate": 0.00015087719298245616,
"loss": 1.1609,
"step": 44
},
{
"epoch": 0.23622047244094488,
"grad_norm": 2.619502544403076,
"learning_rate": 0.0001543859649122807,
"loss": 1.5399,
"step": 45
},
{
"epoch": 0.24146981627296588,
"grad_norm": 2.3362364768981934,
"learning_rate": 0.00015789473684210527,
"loss": 1.3032,
"step": 46
},
{
"epoch": 0.24671916010498687,
"grad_norm": 2.1487205028533936,
"learning_rate": 0.00016140350877192982,
"loss": 1.3317,
"step": 47
},
{
"epoch": 0.25196850393700787,
"grad_norm": 3.1290791034698486,
"learning_rate": 0.0001649122807017544,
"loss": 1.5526,
"step": 48
},
{
"epoch": 0.2572178477690289,
"grad_norm": 2.4382166862487793,
"learning_rate": 0.00016842105263157895,
"loss": 1.2371,
"step": 49
},
{
"epoch": 0.26246719160104987,
"grad_norm": 2.218122959136963,
"learning_rate": 0.00017192982456140353,
"loss": 1.2424,
"step": 50
},
{
"epoch": 0.2677165354330709,
"grad_norm": 2.6388792991638184,
"learning_rate": 0.00017543859649122806,
"loss": 1.2306,
"step": 51
},
{
"epoch": 0.27296587926509186,
"grad_norm": 2.736288547515869,
"learning_rate": 0.00017894736842105264,
"loss": 1.3486,
"step": 52
},
{
"epoch": 0.2782152230971129,
"grad_norm": 2.465235948562622,
"learning_rate": 0.0001824561403508772,
"loss": 1.4959,
"step": 53
},
{
"epoch": 0.28346456692913385,
"grad_norm": 2.587756872177124,
"learning_rate": 0.00018596491228070177,
"loss": 1.6561,
"step": 54
},
{
"epoch": 0.2887139107611549,
"grad_norm": 2.873962163925171,
"learning_rate": 0.00018947368421052632,
"loss": 1.154,
"step": 55
},
{
"epoch": 0.29396325459317585,
"grad_norm": 2.857832431793213,
"learning_rate": 0.00019298245614035088,
"loss": 1.4398,
"step": 56
},
{
"epoch": 0.2992125984251969,
"grad_norm": 2.450753688812256,
"learning_rate": 0.00019649122807017543,
"loss": 1.3935,
"step": 57
},
{
"epoch": 0.30446194225721784,
"grad_norm": 3.0704116821289062,
"learning_rate": 0.0002,
"loss": 1.4604,
"step": 58
},
{
"epoch": 0.30971128608923887,
"grad_norm": 2.362713098526001,
"learning_rate": 0.00019999812486015523,
"loss": 1.5173,
"step": 59
},
{
"epoch": 0.31496062992125984,
"grad_norm": 2.8491404056549072,
"learning_rate": 0.00019999249951094388,
"loss": 1.5408,
"step": 60
},
{
"epoch": 0.32020997375328086,
"grad_norm": 2.454577684402466,
"learning_rate": 0.00019998312416333227,
"loss": 1.4086,
"step": 61
},
{
"epoch": 0.32545931758530183,
"grad_norm": 2.2537734508514404,
"learning_rate": 0.0001999699991689222,
"loss": 1.1157,
"step": 62
},
{
"epoch": 0.33070866141732286,
"grad_norm": 2.300297260284424,
"learning_rate": 0.00019995312501993765,
"loss": 1.1927,
"step": 63
},
{
"epoch": 0.3359580052493438,
"grad_norm": 2.0972819328308105,
"learning_rate": 0.00019993250234920636,
"loss": 1.1443,
"step": 64
},
{
"epoch": 0.34120734908136485,
"grad_norm": 2.440847396850586,
"learning_rate": 0.00019990813193013625,
"loss": 1.7633,
"step": 65
},
{
"epoch": 0.3464566929133858,
"grad_norm": 2.1753475666046143,
"learning_rate": 0.0001998800146766861,
"loss": 1.2409,
"step": 66
},
{
"epoch": 0.35170603674540685,
"grad_norm": 2.196859836578369,
"learning_rate": 0.00019984815164333163,
"loss": 1.1315,
"step": 67
},
{
"epoch": 0.3569553805774278,
"grad_norm": 2.589327335357666,
"learning_rate": 0.00019981254402502566,
"loss": 1.4154,
"step": 68
},
{
"epoch": 0.36220472440944884,
"grad_norm": 2.040648937225342,
"learning_rate": 0.0001997731931571535,
"loss": 1.1035,
"step": 69
},
{
"epoch": 0.3674540682414698,
"grad_norm": 3.925663709640503,
"learning_rate": 0.00019973010051548275,
"loss": 1.562,
"step": 70
},
{
"epoch": 0.37270341207349084,
"grad_norm": 2.7915618419647217,
"learning_rate": 0.00019968326771610797,
"loss": 1.1201,
"step": 71
},
{
"epoch": 0.3779527559055118,
"grad_norm": 2.5829081535339355,
"learning_rate": 0.00019963269651539017,
"loss": 1.3134,
"step": 72
},
{
"epoch": 0.38320209973753283,
"grad_norm": 3.7920446395874023,
"learning_rate": 0.00019957838880989078,
"loss": 1.4021,
"step": 73
},
{
"epoch": 0.3884514435695538,
"grad_norm": 2.2051122188568115,
"learning_rate": 0.00019952034663630062,
"loss": 1.1945,
"step": 74
},
{
"epoch": 0.3937007874015748,
"grad_norm": 2.5238990783691406,
"learning_rate": 0.00019945857217136363,
"loss": 1.3494,
"step": 75
},
{
"epoch": 0.3989501312335958,
"grad_norm": 2.7515692710876465,
"learning_rate": 0.00019939306773179497,
"loss": 1.3467,
"step": 76
},
{
"epoch": 0.4041994750656168,
"grad_norm": 2.8522536754608154,
"learning_rate": 0.00019932383577419432,
"loss": 1.57,
"step": 77
},
{
"epoch": 0.4094488188976378,
"grad_norm": 1.9163004159927368,
"learning_rate": 0.00019925087889495374,
"loss": 1.3051,
"step": 78
},
{
"epoch": 0.4146981627296588,
"grad_norm": 1.9789259433746338,
"learning_rate": 0.00019917419983016025,
"loss": 0.9476,
"step": 79
},
{
"epoch": 0.4199475065616798,
"grad_norm": 2.6120073795318604,
"learning_rate": 0.00019909380145549324,
"loss": 1.3412,
"step": 80
},
{
"epoch": 0.4251968503937008,
"grad_norm": 2.627711772918701,
"learning_rate": 0.00019900968678611666,
"loss": 1.424,
"step": 81
},
{
"epoch": 0.4304461942257218,
"grad_norm": 2.3083677291870117,
"learning_rate": 0.00019892185897656578,
"loss": 1.137,
"step": 82
},
{
"epoch": 0.4356955380577428,
"grad_norm": 2.250309705734253,
"learning_rate": 0.00019883032132062925,
"loss": 1.0466,
"step": 83
},
{
"epoch": 0.4409448818897638,
"grad_norm": 2.2926323413848877,
"learning_rate": 0.00019873507725122504,
"loss": 1.4918,
"step": 84
},
{
"epoch": 0.4461942257217848,
"grad_norm": 2.8048651218414307,
"learning_rate": 0.00019863613034027224,
"loss": 1.2737,
"step": 85
},
{
"epoch": 0.45144356955380577,
"grad_norm": 2.108257532119751,
"learning_rate": 0.00019853348429855672,
"loss": 1.0448,
"step": 86
},
{
"epoch": 0.4566929133858268,
"grad_norm": 2.1626484394073486,
"learning_rate": 0.00019842714297559213,
"loss": 1.1422,
"step": 87
},
{
"epoch": 0.46194225721784776,
"grad_norm": 2.1590943336486816,
"learning_rate": 0.0001983171103594755,
"loss": 1.073,
"step": 88
},
{
"epoch": 0.4671916010498688,
"grad_norm": 2.6183314323425293,
"learning_rate": 0.0001982033905767377,
"loss": 1.1683,
"step": 89
},
{
"epoch": 0.47244094488188976,
"grad_norm": 2.6423206329345703,
"learning_rate": 0.00019808598789218865,
"loss": 1.284,
"step": 90
},
{
"epoch": 0.4776902887139108,
"grad_norm": 2.9666941165924072,
"learning_rate": 0.0001979649067087574,
"loss": 1.4623,
"step": 91
},
{
"epoch": 0.48293963254593175,
"grad_norm": 2.3356740474700928,
"learning_rate": 0.00019784015156732693,
"loss": 1.4087,
"step": 92
},
{
"epoch": 0.4881889763779528,
"grad_norm": 2.55161190032959,
"learning_rate": 0.000197711727146564,
"loss": 1.2149,
"step": 93
},
{
"epoch": 0.49343832020997375,
"grad_norm": 2.503931760787964,
"learning_rate": 0.00019757963826274357,
"loss": 1.0572,
"step": 94
},
{
"epoch": 0.49868766404199477,
"grad_norm": 2.113316297531128,
"learning_rate": 0.00019744388986956822,
"loss": 1.3952,
"step": 95
},
{
"epoch": 0.5039370078740157,
"grad_norm": 2.413565158843994,
"learning_rate": 0.00019730448705798239,
"loss": 1.0974,
"step": 96
},
{
"epoch": 0.5091863517060368,
"grad_norm": 2.6888630390167236,
"learning_rate": 0.0001971614350559814,
"loss": 1.5181,
"step": 97
},
{
"epoch": 0.5144356955380578,
"grad_norm": 2.208359718322754,
"learning_rate": 0.0001970147392284154,
"loss": 1.1508,
"step": 98
},
{
"epoch": 0.5196850393700787,
"grad_norm": 2.3736019134521484,
"learning_rate": 0.00019686440507678824,
"loss": 1.4067,
"step": 99
},
{
"epoch": 0.5249343832020997,
"grad_norm": 2.40409255027771,
"learning_rate": 0.0001967104382390511,
"loss": 1.1971,
"step": 100
},
{
"epoch": 0.5301837270341208,
"grad_norm": 2.203507661819458,
"learning_rate": 0.00019655284448939094,
"loss": 1.3298,
"step": 101
},
{
"epoch": 0.5354330708661418,
"grad_norm": 2.1737711429595947,
"learning_rate": 0.00019639162973801426,
"loss": 1.4618,
"step": 102
},
{
"epoch": 0.5406824146981627,
"grad_norm": 3.172531843185425,
"learning_rate": 0.00019622680003092503,
"loss": 1.6789,
"step": 103
},
{
"epoch": 0.5459317585301837,
"grad_norm": 2.5439305305480957,
"learning_rate": 0.0001960583615496984,
"loss": 1.6033,
"step": 104
},
{
"epoch": 0.5511811023622047,
"grad_norm": 2.414440870285034,
"learning_rate": 0.00019588632061124837,
"loss": 1.1086,
"step": 105
},
{
"epoch": 0.5564304461942258,
"grad_norm": 2.828094244003296,
"learning_rate": 0.00019571068366759143,
"loss": 1.2589,
"step": 106
},
{
"epoch": 0.5616797900262467,
"grad_norm": 2.0109658241271973,
"learning_rate": 0.00019553145730560415,
"loss": 0.8927,
"step": 107
},
{
"epoch": 0.5669291338582677,
"grad_norm": 2.842677116394043,
"learning_rate": 0.0001953486482467764,
"loss": 1.4432,
"step": 108
},
{
"epoch": 0.5721784776902887,
"grad_norm": 2.2482569217681885,
"learning_rate": 0.0001951622633469592,
"loss": 1.1888,
"step": 109
},
{
"epoch": 0.5774278215223098,
"grad_norm": 2.6185362339019775,
"learning_rate": 0.00019497230959610756,
"loss": 1.1135,
"step": 110
},
{
"epoch": 0.5826771653543307,
"grad_norm": 2.3009023666381836,
"learning_rate": 0.00019477879411801844,
"loss": 1.1068,
"step": 111
},
{
"epoch": 0.5879265091863517,
"grad_norm": 2.750929832458496,
"learning_rate": 0.00019458172417006347,
"loss": 1.0964,
"step": 112
},
{
"epoch": 0.5931758530183727,
"grad_norm": 2.6601922512054443,
"learning_rate": 0.00019438110714291694,
"loss": 1.4321,
"step": 113
},
{
"epoch": 0.5984251968503937,
"grad_norm": 2.598353624343872,
"learning_rate": 0.00019417695056027844,
"loss": 1.1159,
"step": 114
},
{
"epoch": 0.6036745406824147,
"grad_norm": 2.5931055545806885,
"learning_rate": 0.00019396926207859084,
"loss": 1.5328,
"step": 115
},
{
"epoch": 0.6089238845144357,
"grad_norm": 2.612820625305176,
"learning_rate": 0.00019375804948675306,
"loss": 1.5006,
"step": 116
},
{
"epoch": 0.6141732283464567,
"grad_norm": 2.5003819465637207,
"learning_rate": 0.0001935433207058281,
"loss": 1.5221,
"step": 117
},
{
"epoch": 0.6194225721784777,
"grad_norm": 2.438344955444336,
"learning_rate": 0.0001933250837887457,
"loss": 1.1499,
"step": 118
},
{
"epoch": 0.6246719160104987,
"grad_norm": 2.2599291801452637,
"learning_rate": 0.00019310334692000075,
"loss": 1.1137,
"step": 119
},
{
"epoch": 0.6299212598425197,
"grad_norm": 2.45780086517334,
"learning_rate": 0.00019287811841534595,
"loss": 1.363,
"step": 120
},
{
"epoch": 0.6351706036745407,
"grad_norm": 2.098870038986206,
"learning_rate": 0.00019264940672148018,
"loss": 1.2201,
"step": 121
},
{
"epoch": 0.6404199475065617,
"grad_norm": 1.8870902061462402,
"learning_rate": 0.00019241722041573166,
"loss": 1.3615,
"step": 122
},
{
"epoch": 0.6456692913385826,
"grad_norm": 1.971105694770813,
"learning_rate": 0.0001921815682057362,
"loss": 1.1174,
"step": 123
},
{
"epoch": 0.6509186351706037,
"grad_norm": 2.195482015609741,
"learning_rate": 0.0001919424589291108,
"loss": 1.2562,
"step": 124
},
{
"epoch": 0.6561679790026247,
"grad_norm": 2.429697036743164,
"learning_rate": 0.0001916999015531221,
"loss": 1.3243,
"step": 125
},
{
"epoch": 0.6614173228346457,
"grad_norm": 2.208289623260498,
"learning_rate": 0.00019145390517435012,
"loss": 1.0649,
"step": 126
},
{
"epoch": 0.6666666666666666,
"grad_norm": 2.1772406101226807,
"learning_rate": 0.00019120447901834706,
"loss": 1.1999,
"step": 127
},
{
"epoch": 0.6719160104986877,
"grad_norm": 2.6156198978424072,
"learning_rate": 0.00019095163243929142,
"loss": 1.2446,
"step": 128
},
{
"epoch": 0.6771653543307087,
"grad_norm": 2.6300153732299805,
"learning_rate": 0.0001906953749196371,
"loss": 1.3671,
"step": 129
},
{
"epoch": 0.6824146981627297,
"grad_norm": 2.3485264778137207,
"learning_rate": 0.00019043571606975777,
"loss": 0.9951,
"step": 130
},
{
"epoch": 0.6876640419947506,
"grad_norm": 2.8363163471221924,
"learning_rate": 0.00019017266562758659,
"loss": 1.3425,
"step": 131
},
{
"epoch": 0.6929133858267716,
"grad_norm": 2.223158121109009,
"learning_rate": 0.00018990623345825083,
"loss": 1.3456,
"step": 132
},
{
"epoch": 0.6981627296587927,
"grad_norm": 2.4961116313934326,
"learning_rate": 0.00018963642955370201,
"loss": 1.4691,
"step": 133
},
{
"epoch": 0.7034120734908137,
"grad_norm": 2.3813393115997314,
"learning_rate": 0.00018936326403234125,
"loss": 1.2141,
"step": 134
},
{
"epoch": 0.7086614173228346,
"grad_norm": 2.0415079593658447,
"learning_rate": 0.00018908674713863952,
"loss": 1.1995,
"step": 135
},
{
"epoch": 0.7139107611548556,
"grad_norm": 2.443024158477783,
"learning_rate": 0.00018880688924275378,
"loss": 1.5299,
"step": 136
},
{
"epoch": 0.7191601049868767,
"grad_norm": 2.8351380825042725,
"learning_rate": 0.0001885237008401378,
"loss": 1.2206,
"step": 137
},
{
"epoch": 0.7244094488188977,
"grad_norm": 2.696510076522827,
"learning_rate": 0.0001882371925511488,
"loss": 1.0794,
"step": 138
},
{
"epoch": 0.7296587926509186,
"grad_norm": 2.9665584564208984,
"learning_rate": 0.0001879473751206489,
"loss": 1.575,
"step": 139
},
{
"epoch": 0.7349081364829396,
"grad_norm": 2.1545960903167725,
"learning_rate": 0.00018765425941760238,
"loss": 1.0347,
"step": 140
},
{
"epoch": 0.7401574803149606,
"grad_norm": 2.3312506675720215,
"learning_rate": 0.00018735785643466784,
"loss": 1.3836,
"step": 141
},
{
"epoch": 0.7454068241469817,
"grad_norm": 2.3382620811462402,
"learning_rate": 0.00018705817728778624,
"loss": 1.1358,
"step": 142
},
{
"epoch": 0.7506561679790026,
"grad_norm": 2.4452474117279053,
"learning_rate": 0.00018675523321576371,
"loss": 1.1749,
"step": 143
},
{
"epoch": 0.7559055118110236,
"grad_norm": 2.0249741077423096,
"learning_rate": 0.00018644903557985025,
"loss": 0.9222,
"step": 144
},
{
"epoch": 0.7611548556430446,
"grad_norm": 2.624455451965332,
"learning_rate": 0.00018613959586331362,
"loss": 1.0862,
"step": 145
},
{
"epoch": 0.7664041994750657,
"grad_norm": 2.175856351852417,
"learning_rate": 0.00018582692567100867,
"loss": 1.1019,
"step": 146
},
{
"epoch": 0.7716535433070866,
"grad_norm": 2.5203380584716797,
"learning_rate": 0.00018551103672894206,
"loss": 1.5635,
"step": 147
},
{
"epoch": 0.7769028871391076,
"grad_norm": 2.405241012573242,
"learning_rate": 0.00018519194088383273,
"loss": 1.2331,
"step": 148
},
{
"epoch": 0.7821522309711286,
"grad_norm": 2.598151206970215,
"learning_rate": 0.00018486965010266725,
"loss": 1.7277,
"step": 149
},
{
"epoch": 0.7874015748031497,
"grad_norm": 2.171656370162964,
"learning_rate": 0.0001845441764722514,
"loss": 1.2162,
"step": 150
},
{
"epoch": 0.7926509186351706,
"grad_norm": 2.485642433166504,
"learning_rate": 0.00018421553219875658,
"loss": 1.5379,
"step": 151
},
{
"epoch": 0.7979002624671916,
"grad_norm": 2.297564744949341,
"learning_rate": 0.00018388372960726228,
"loss": 1.168,
"step": 152
},
{
"epoch": 0.8031496062992126,
"grad_norm": 1.9809898138046265,
"learning_rate": 0.00018354878114129367,
"loss": 1.1138,
"step": 153
},
{
"epoch": 0.8083989501312336,
"grad_norm": 2.277679920196533,
"learning_rate": 0.00018321069936235503,
"loss": 0.9873,
"step": 154
},
{
"epoch": 0.8136482939632546,
"grad_norm": 2.324462890625,
"learning_rate": 0.00018286949694945866,
"loss": 1.2728,
"step": 155
},
{
"epoch": 0.8188976377952756,
"grad_norm": 2.318861722946167,
"learning_rate": 0.00018252518669864936,
"loss": 1.2905,
"step": 156
},
{
"epoch": 0.8241469816272966,
"grad_norm": 2.5826478004455566,
"learning_rate": 0.0001821777815225245,
"loss": 1.2632,
"step": 157
},
{
"epoch": 0.8293963254593176,
"grad_norm": 1.9897890090942383,
"learning_rate": 0.00018182729444974992,
"loss": 1.1829,
"step": 158
},
{
"epoch": 0.8346456692913385,
"grad_norm": 2.1801273822784424,
"learning_rate": 0.00018147373862457107,
"loss": 1.1657,
"step": 159
},
{
"epoch": 0.8398950131233596,
"grad_norm": 2.332425117492676,
"learning_rate": 0.00018111712730632022,
"loss": 1.1415,
"step": 160
},
{
"epoch": 0.8451443569553806,
"grad_norm": 2.0971760749816895,
"learning_rate": 0.0001807574738689193,
"loss": 1.3374,
"step": 161
},
{
"epoch": 0.8503937007874016,
"grad_norm": 2.436295509338379,
"learning_rate": 0.000180394791800378,
"loss": 1.2387,
"step": 162
},
{
"epoch": 0.8556430446194225,
"grad_norm": 2.6877710819244385,
"learning_rate": 0.00018002909470228842,
"loss": 1.5381,
"step": 163
},
{
"epoch": 0.8608923884514436,
"grad_norm": 2.4646594524383545,
"learning_rate": 0.00017966039628931446,
"loss": 1.4384,
"step": 164
},
{
"epoch": 0.8661417322834646,
"grad_norm": 2.020935297012329,
"learning_rate": 0.00017928871038867784,
"loss": 1.2144,
"step": 165
},
{
"epoch": 0.8713910761154856,
"grad_norm": 2.4887311458587646,
"learning_rate": 0.00017891405093963938,
"loss": 1.3464,
"step": 166
},
{
"epoch": 0.8766404199475065,
"grad_norm": 2.3591365814208984,
"learning_rate": 0.00017853643199297633,
"loss": 1.1275,
"step": 167
},
{
"epoch": 0.8818897637795275,
"grad_norm": 1.8852261304855347,
"learning_rate": 0.00017815586771045535,
"loss": 1.1718,
"step": 168
},
{
"epoch": 0.8871391076115486,
"grad_norm": 2.501147747039795,
"learning_rate": 0.0001777723723643014,
"loss": 1.5136,
"step": 169
},
{
"epoch": 0.8923884514435696,
"grad_norm": 2.3164873123168945,
"learning_rate": 0.0001773859603366626,
"loss": 1.2471,
"step": 170
},
{
"epoch": 0.8976377952755905,
"grad_norm": 2.3213908672332764,
"learning_rate": 0.00017699664611907072,
"loss": 1.5152,
"step": 171
},
{
"epoch": 0.9028871391076115,
"grad_norm": 2.225059986114502,
"learning_rate": 0.0001766044443118978,
"loss": 0.9982,
"step": 172
},
{
"epoch": 0.9081364829396326,
"grad_norm": 2.617565631866455,
"learning_rate": 0.00017620936962380856,
"loss": 1.1797,
"step": 173
},
{
"epoch": 0.9133858267716536,
"grad_norm": 2.1258366107940674,
"learning_rate": 0.00017581143687120875,
"loss": 1.2575,
"step": 174
},
{
"epoch": 0.9186351706036745,
"grad_norm": 2.1051361560821533,
"learning_rate": 0.00017541066097768963,
"loss": 0.8311,
"step": 175
},
{
"epoch": 0.9238845144356955,
"grad_norm": 2.1657958030700684,
"learning_rate": 0.0001750070569734681,
"loss": 1.093,
"step": 176
},
{
"epoch": 0.9291338582677166,
"grad_norm": 2.4127931594848633,
"learning_rate": 0.00017460063999482316,
"loss": 1.0833,
"step": 177
},
{
"epoch": 0.9343832020997376,
"grad_norm": 2.753276824951172,
"learning_rate": 0.00017419142528352817,
"loss": 1.178,
"step": 178
},
{
"epoch": 0.9396325459317585,
"grad_norm": 2.67309832572937,
"learning_rate": 0.00017377942818627942,
"loss": 1.303,
"step": 179
},
{
"epoch": 0.9448818897637795,
"grad_norm": 2.5962181091308594,
"learning_rate": 0.00017336466415412028,
"loss": 1.5608,
"step": 180
},
{
"epoch": 0.9501312335958005,
"grad_norm": 3.0257773399353027,
"learning_rate": 0.0001729471487418621,
"loss": 1.3775,
"step": 181
},
{
"epoch": 0.9553805774278216,
"grad_norm": 1.9228885173797607,
"learning_rate": 0.0001725268976075005,
"loss": 1.0905,
"step": 182
},
{
"epoch": 0.9606299212598425,
"grad_norm": 3.0711910724639893,
"learning_rate": 0.0001721039265116285,
"loss": 1.1671,
"step": 183
},
{
"epoch": 0.9658792650918635,
"grad_norm": 2.53415846824646,
"learning_rate": 0.00017167825131684513,
"loss": 1.458,
"step": 184
},
{
"epoch": 0.9711286089238845,
"grad_norm": 2.325265407562256,
"learning_rate": 0.00017124988798716083,
"loss": 1.3166,
"step": 185
},
{
"epoch": 0.9763779527559056,
"grad_norm": 1.972551941871643,
"learning_rate": 0.00017081885258739846,
"loss": 1.1434,
"step": 186
},
{
"epoch": 0.9816272965879265,
"grad_norm": 2.308837890625,
"learning_rate": 0.00017038516128259115,
"loss": 1.25,
"step": 187
},
{
"epoch": 0.9868766404199475,
"grad_norm": 2.6584460735321045,
"learning_rate": 0.00016994883033737582,
"loss": 1.3522,
"step": 188
},
{
"epoch": 0.9921259842519685,
"grad_norm": 2.0960097312927246,
"learning_rate": 0.00016950987611538324,
"loss": 1.0266,
"step": 189
},
{
"epoch": 0.9973753280839895,
"grad_norm": 2.1752665042877197,
"learning_rate": 0.00016906831507862443,
"loss": 1.3627,
"step": 190
},
{
"epoch": 1.0,
"grad_norm": 2.6606860160827637,
"learning_rate": 0.0001686241637868734,
"loss": 0.8187,
"step": 191
},
{
"epoch": 1.005249343832021,
"grad_norm": 2.049098491668701,
"learning_rate": 0.00016817743889704565,
"loss": 1.0819,
"step": 192
},
{
"epoch": 1.010498687664042,
"grad_norm": 1.9727576971054077,
"learning_rate": 0.00016772815716257412,
"loss": 1.0079,
"step": 193
},
{
"epoch": 1.015748031496063,
"grad_norm": 1.9842270612716675,
"learning_rate": 0.0001672763354327804,
"loss": 0.8639,
"step": 194
},
{
"epoch": 1.020997375328084,
"grad_norm": 1.639641523361206,
"learning_rate": 0.00016682199065224307,
"loss": 0.6213,
"step": 195
},
{
"epoch": 1.026246719160105,
"grad_norm": 2.2830586433410645,
"learning_rate": 0.00016636513986016213,
"loss": 0.7467,
"step": 196
},
{
"epoch": 1.031496062992126,
"grad_norm": 3.062472105026245,
"learning_rate": 0.0001659058001897201,
"loss": 0.9837,
"step": 197
},
{
"epoch": 1.036745406824147,
"grad_norm": 2.6546707153320312,
"learning_rate": 0.00016544398886743933,
"loss": 1.0166,
"step": 198
},
{
"epoch": 1.041994750656168,
"grad_norm": 2.1115078926086426,
"learning_rate": 0.000164979723212536,
"loss": 0.8896,
"step": 199
},
{
"epoch": 1.047244094488189,
"grad_norm": 2.4062306880950928,
"learning_rate": 0.00016451302063627066,
"loss": 0.9431,
"step": 200
},
{
"epoch": 1.05249343832021,
"grad_norm": 2.4070167541503906,
"learning_rate": 0.00016404389864129533,
"loss": 0.7534,
"step": 201
},
{
"epoch": 1.057742782152231,
"grad_norm": 2.7025625705718994,
"learning_rate": 0.00016357237482099684,
"loss": 0.6826,
"step": 202
},
{
"epoch": 1.0629921259842519,
"grad_norm": 2.4860000610351562,
"learning_rate": 0.00016309846685883726,
"loss": 0.7801,
"step": 203
},
{
"epoch": 1.068241469816273,
"grad_norm": 2.442067861557007,
"learning_rate": 0.00016262219252769064,
"loss": 0.8237,
"step": 204
},
{
"epoch": 1.073490813648294,
"grad_norm": 2.510655164718628,
"learning_rate": 0.00016214356968917648,
"loss": 0.8001,
"step": 205
},
{
"epoch": 1.078740157480315,
"grad_norm": 3.1672093868255615,
"learning_rate": 0.00016166261629298995,
"loss": 0.9451,
"step": 206
},
{
"epoch": 1.083989501312336,
"grad_norm": 3.2933852672576904,
"learning_rate": 0.0001611793503762285,
"loss": 0.9388,
"step": 207
},
{
"epoch": 1.0892388451443569,
"grad_norm": 3.1465041637420654,
"learning_rate": 0.00016069379006271566,
"loss": 1.0714,
"step": 208
},
{
"epoch": 1.094488188976378,
"grad_norm": 2.6741809844970703,
"learning_rate": 0.00016020595356232135,
"loss": 0.7402,
"step": 209
},
{
"epoch": 1.099737532808399,
"grad_norm": 2.4382576942443848,
"learning_rate": 0.00015971585917027862,
"loss": 0.6643,
"step": 210
},
{
"epoch": 1.10498687664042,
"grad_norm": 2.5925192832946777,
"learning_rate": 0.00015922352526649803,
"loss": 0.8338,
"step": 211
},
{
"epoch": 1.110236220472441,
"grad_norm": 2.8813819885253906,
"learning_rate": 0.00015872897031487791,
"loss": 0.9946,
"step": 212
},
{
"epoch": 1.1154855643044619,
"grad_norm": 2.19948673248291,
"learning_rate": 0.00015823221286261215,
"loss": 0.916,
"step": 213
},
{
"epoch": 1.120734908136483,
"grad_norm": 2.330357551574707,
"learning_rate": 0.00015773327153949465,
"loss": 0.9889,
"step": 214
},
{
"epoch": 1.125984251968504,
"grad_norm": 2.321606397628784,
"learning_rate": 0.0001572321650572205,
"loss": 0.9533,
"step": 215
},
{
"epoch": 1.1312335958005248,
"grad_norm": 2.219377040863037,
"learning_rate": 0.00015672891220868432,
"loss": 0.6138,
"step": 216
},
{
"epoch": 1.136482939632546,
"grad_norm": 2.7651844024658203,
"learning_rate": 0.00015622353186727544,
"loss": 0.8008,
"step": 217
},
{
"epoch": 1.141732283464567,
"grad_norm": 2.5043694972991943,
"learning_rate": 0.0001557160429861702,
"loss": 0.9846,
"step": 218
},
{
"epoch": 1.1469816272965878,
"grad_norm": 2.6043448448181152,
"learning_rate": 0.000155206464597621,
"loss": 0.932,
"step": 219
},
{
"epoch": 1.152230971128609,
"grad_norm": 2.5369505882263184,
"learning_rate": 0.00015469481581224272,
"loss": 0.8897,
"step": 220
},
{
"epoch": 1.1574803149606299,
"grad_norm": 2.6622097492218018,
"learning_rate": 0.00015418111581829574,
"loss": 0.952,
"step": 221
},
{
"epoch": 1.162729658792651,
"grad_norm": 2.4234066009521484,
"learning_rate": 0.0001536653838809667,
"loss": 0.8107,
"step": 222
},
{
"epoch": 1.167979002624672,
"grad_norm": 2.463465452194214,
"learning_rate": 0.0001531476393416456,
"loss": 0.7944,
"step": 223
},
{
"epoch": 1.1732283464566928,
"grad_norm": 2.748023271560669,
"learning_rate": 0.0001526279016172008,
"loss": 0.8127,
"step": 224
},
{
"epoch": 1.178477690288714,
"grad_norm": 2.4598255157470703,
"learning_rate": 0.00015210619019925066,
"loss": 0.7711,
"step": 225
},
{
"epoch": 1.1837270341207349,
"grad_norm": 2.580259323120117,
"learning_rate": 0.00015158252465343242,
"loss": 0.8544,
"step": 226
},
{
"epoch": 1.188976377952756,
"grad_norm": 2.5548136234283447,
"learning_rate": 0.00015105692461866874,
"loss": 0.7007,
"step": 227
},
{
"epoch": 1.194225721784777,
"grad_norm": 2.474823236465454,
"learning_rate": 0.000150529409806431,
"loss": 0.7501,
"step": 228
},
{
"epoch": 1.1994750656167978,
"grad_norm": 2.46974515914917,
"learning_rate": 0.00015000000000000001,
"loss": 0.6326,
"step": 229
},
{
"epoch": 1.204724409448819,
"grad_norm": 2.542745590209961,
"learning_rate": 0.00014946871505372425,
"loss": 0.63,
"step": 230
},
{
"epoch": 1.20997375328084,
"grad_norm": 2.988248109817505,
"learning_rate": 0.00014893557489227517,
"loss": 1.0908,
"step": 231
},
{
"epoch": 1.2152230971128608,
"grad_norm": 3.3131489753723145,
"learning_rate": 0.0001484005995098999,
"loss": 1.2089,
"step": 232
},
{
"epoch": 1.220472440944882,
"grad_norm": 3.060359477996826,
"learning_rate": 0.0001478638089696716,
"loss": 0.7317,
"step": 233
},
{
"epoch": 1.2257217847769029,
"grad_norm": 3.0129809379577637,
"learning_rate": 0.00014732522340273684,
"loss": 0.6776,
"step": 234
},
{
"epoch": 1.2309711286089238,
"grad_norm": 2.892338752746582,
"learning_rate": 0.0001467848630075608,
"loss": 0.731,
"step": 235
},
{
"epoch": 1.236220472440945,
"grad_norm": 3.0470635890960693,
"learning_rate": 0.00014624274804916958,
"loss": 0.9027,
"step": 236
},
{
"epoch": 1.2414698162729658,
"grad_norm": 2.4284868240356445,
"learning_rate": 0.00014569889885839037,
"loss": 0.7888,
"step": 237
},
{
"epoch": 1.246719160104987,
"grad_norm": 2.565948486328125,
"learning_rate": 0.00014515333583108896,
"loss": 0.6491,
"step": 238
},
{
"epoch": 1.2519685039370079,
"grad_norm": 2.1786141395568848,
"learning_rate": 0.00014460607942740468,
"loss": 0.7064,
"step": 239
},
{
"epoch": 1.257217847769029,
"grad_norm": 2.6910786628723145,
"learning_rate": 0.00014405715017098335,
"loss": 0.6241,
"step": 240
},
{
"epoch": 1.26246719160105,
"grad_norm": 2.9195802211761475,
"learning_rate": 0.00014350656864820733,
"loss": 0.815,
"step": 241
},
{
"epoch": 1.2677165354330708,
"grad_norm": 2.84618878364563,
"learning_rate": 0.0001429543555074237,
"loss": 0.8676,
"step": 242
},
{
"epoch": 1.272965879265092,
"grad_norm": 2.5274624824523926,
"learning_rate": 0.00014240053145816967,
"loss": 0.8162,
"step": 243
},
{
"epoch": 1.2782152230971129,
"grad_norm": 2.549936294555664,
"learning_rate": 0.00014184511727039612,
"loss": 0.7318,
"step": 244
},
{
"epoch": 1.2834645669291338,
"grad_norm": 2.6315500736236572,
"learning_rate": 0.0001412881337736885,
"loss": 0.828,
"step": 245
},
{
"epoch": 1.288713910761155,
"grad_norm": 2.9797654151916504,
"learning_rate": 0.00014072960185648577,
"loss": 0.8592,
"step": 246
},
{
"epoch": 1.2939632545931758,
"grad_norm": 3.380535364151001,
"learning_rate": 0.00014016954246529696,
"loss": 1.2413,
"step": 247
},
{
"epoch": 1.2992125984251968,
"grad_norm": 2.557527780532837,
"learning_rate": 0.0001396079766039157,
"loss": 0.7001,
"step": 248
},
{
"epoch": 1.304461942257218,
"grad_norm": 2.8971340656280518,
"learning_rate": 0.00013904492533263244,
"loss": 0.7283,
"step": 249
},
{
"epoch": 1.3097112860892388,
"grad_norm": 2.7849743366241455,
"learning_rate": 0.00013848040976744457,
"loss": 0.7282,
"step": 250
},
{
"epoch": 1.3149606299212597,
"grad_norm": 2.8634750843048096,
"learning_rate": 0.00013791445107926478,
"loss": 0.8419,
"step": 251
},
{
"epoch": 1.3202099737532809,
"grad_norm": 2.310309410095215,
"learning_rate": 0.00013734707049312673,
"loss": 0.5901,
"step": 252
},
{
"epoch": 1.3254593175853018,
"grad_norm": 2.7387428283691406,
"learning_rate": 0.00013677828928738934,
"loss": 0.7358,
"step": 253
},
{
"epoch": 1.330708661417323,
"grad_norm": 2.7614364624023438,
"learning_rate": 0.00013620812879293863,
"loss": 0.792,
"step": 254
},
{
"epoch": 1.3359580052493438,
"grad_norm": 2.8637635707855225,
"learning_rate": 0.00013563661039238785,
"loss": 0.8998,
"step": 255
},
{
"epoch": 1.341207349081365,
"grad_norm": 2.455495595932007,
"learning_rate": 0.00013506375551927547,
"loss": 0.6392,
"step": 256
},
{
"epoch": 1.3464566929133859,
"grad_norm": 3.1629257202148438,
"learning_rate": 0.00013448958565726144,
"loss": 0.8548,
"step": 257
},
{
"epoch": 1.3517060367454068,
"grad_norm": 2.6103389263153076,
"learning_rate": 0.00013391412233932149,
"loss": 0.6498,
"step": 258
},
{
"epoch": 1.356955380577428,
"grad_norm": 2.9760587215423584,
"learning_rate": 0.00013333738714693956,
"loss": 0.7394,
"step": 259
},
{
"epoch": 1.3622047244094488,
"grad_norm": 2.5022411346435547,
"learning_rate": 0.00013275940170929843,
"loss": 0.6005,
"step": 260
},
{
"epoch": 1.3674540682414698,
"grad_norm": 2.8095810413360596,
"learning_rate": 0.00013218018770246858,
"loss": 0.6823,
"step": 261
},
{
"epoch": 1.372703412073491,
"grad_norm": 2.7416610717773438,
"learning_rate": 0.00013159976684859527,
"loss": 0.7922,
"step": 262
},
{
"epoch": 1.3779527559055118,
"grad_norm": 2.2850701808929443,
"learning_rate": 0.00013101816091508388,
"loss": 0.7069,
"step": 263
},
{
"epoch": 1.3832020997375327,
"grad_norm": 2.835042953491211,
"learning_rate": 0.0001304353917137836,
"loss": 0.7351,
"step": 264
},
{
"epoch": 1.3884514435695539,
"grad_norm": 2.9981067180633545,
"learning_rate": 0.00012985148110016947,
"loss": 0.8191,
"step": 265
},
{
"epoch": 1.3937007874015748,
"grad_norm": 2.754315137863159,
"learning_rate": 0.0001292664509725226,
"loss": 0.6671,
"step": 266
},
{
"epoch": 1.3989501312335957,
"grad_norm": 2.780714750289917,
"learning_rate": 0.00012868032327110904,
"loss": 0.6961,
"step": 267
},
{
"epoch": 1.4041994750656168,
"grad_norm": 2.5875256061553955,
"learning_rate": 0.00012809311997735696,
"loss": 0.5871,
"step": 268
},
{
"epoch": 1.4094488188976377,
"grad_norm": 2.6932876110076904,
"learning_rate": 0.00012750486311303218,
"loss": 0.7657,
"step": 269
},
{
"epoch": 1.4146981627296589,
"grad_norm": 2.3105850219726562,
"learning_rate": 0.00012691557473941243,
"loss": 0.4689,
"step": 270
},
{
"epoch": 1.4199475065616798,
"grad_norm": 2.749490261077881,
"learning_rate": 0.00012632527695645993,
"loss": 0.9078,
"step": 271
},
{
"epoch": 1.425196850393701,
"grad_norm": 2.5074801445007324,
"learning_rate": 0.0001257339919019925,
"loss": 0.6867,
"step": 272
},
{
"epoch": 1.4304461942257218,
"grad_norm": 2.0776736736297607,
"learning_rate": 0.00012514174175085345,
"loss": 0.6132,
"step": 273
},
{
"epoch": 1.4356955380577427,
"grad_norm": 2.6735877990722656,
"learning_rate": 0.00012454854871407994,
"loss": 0.7334,
"step": 274
},
{
"epoch": 1.4409448818897639,
"grad_norm": 2.643948554992676,
"learning_rate": 0.0001239544350380699,
"loss": 0.6834,
"step": 275
},
{
"epoch": 1.4461942257217848,
"grad_norm": 3.092463254928589,
"learning_rate": 0.00012335942300374788,
"loss": 1.0018,
"step": 276
},
{
"epoch": 1.4514435695538057,
"grad_norm": 2.8050155639648438,
"learning_rate": 0.00012276353492572935,
"loss": 0.8914,
"step": 277
},
{
"epoch": 1.4566929133858268,
"grad_norm": 2.794189214706421,
"learning_rate": 0.00012216679315148386,
"loss": 0.8057,
"step": 278
},
{
"epoch": 1.4619422572178478,
"grad_norm": 2.8576409816741943,
"learning_rate": 0.00012156922006049702,
"loss": 0.9259,
"step": 279
},
{
"epoch": 1.4671916010498687,
"grad_norm": 3.4657177925109863,
"learning_rate": 0.00012097083806343103,
"loss": 1.0509,
"step": 280
},
{
"epoch": 1.4724409448818898,
"grad_norm": 2.971736431121826,
"learning_rate": 0.00012037166960128443,
"loss": 1.2074,
"step": 281
},
{
"epoch": 1.4776902887139107,
"grad_norm": 2.635852336883545,
"learning_rate": 0.00011977173714455034,
"loss": 0.739,
"step": 282
},
{
"epoch": 1.4829396325459316,
"grad_norm": 2.4697046279907227,
"learning_rate": 0.00011917106319237386,
"loss": 0.8156,
"step": 283
},
{
"epoch": 1.4881889763779528,
"grad_norm": 2.8160102367401123,
"learning_rate": 0.00011856967027170818,
"loss": 0.7248,
"step": 284
},
{
"epoch": 1.4934383202099737,
"grad_norm": 2.7745792865753174,
"learning_rate": 0.00011796758093646989,
"loss": 0.7826,
"step": 285
},
{
"epoch": 1.4986876640419948,
"grad_norm": 2.684284210205078,
"learning_rate": 0.00011736481776669306,
"loss": 0.6299,
"step": 286
},
{
"epoch": 1.5039370078740157,
"grad_norm": 2.7349674701690674,
"learning_rate": 0.00011676140336768236,
"loss": 0.8248,
"step": 287
},
{
"epoch": 1.5091863517060369,
"grad_norm": 3.207608938217163,
"learning_rate": 0.00011615736036916549,
"loss": 1.0109,
"step": 288
},
{
"epoch": 1.5144356955380578,
"grad_norm": 2.8411524295806885,
"learning_rate": 0.00011555271142444433,
"loss": 0.8515,
"step": 289
},
{
"epoch": 1.5196850393700787,
"grad_norm": 2.936953544616699,
"learning_rate": 0.00011494747920954545,
"loss": 0.7542,
"step": 290
},
{
"epoch": 1.5249343832020998,
"grad_norm": 2.1223931312561035,
"learning_rate": 0.00011434168642236964,
"loss": 0.5709,
"step": 291
},
{
"epoch": 1.5301837270341208,
"grad_norm": 3.362982988357544,
"learning_rate": 0.00011373535578184082,
"loss": 0.9206,
"step": 292
},
{
"epoch": 1.5354330708661417,
"grad_norm": 3.006317138671875,
"learning_rate": 0.00011312851002705383,
"loss": 1.017,
"step": 293
},
{
"epoch": 1.5406824146981628,
"grad_norm": 3.088223934173584,
"learning_rate": 0.00011252117191642175,
"loss": 1.025,
"step": 294
},
{
"epoch": 1.5459317585301837,
"grad_norm": 2.5200142860412598,
"learning_rate": 0.00011191336422682237,
"loss": 0.5996,
"step": 295
},
{
"epoch": 1.5511811023622046,
"grad_norm": 2.9995408058166504,
"learning_rate": 0.00011130510975274409,
"loss": 0.6813,
"step": 296
},
{
"epoch": 1.5564304461942258,
"grad_norm": 2.932790994644165,
"learning_rate": 0.00011069643130543084,
"loss": 0.9218,
"step": 297
},
{
"epoch": 1.5616797900262467,
"grad_norm": 2.309981346130371,
"learning_rate": 0.00011008735171202684,
"loss": 0.7003,
"step": 298
},
{
"epoch": 1.5669291338582676,
"grad_norm": 2.602471113204956,
"learning_rate": 0.00010947789381472035,
"loss": 0.7051,
"step": 299
},
{
"epoch": 1.5721784776902887,
"grad_norm": 2.633842945098877,
"learning_rate": 0.00010886808046988717,
"loss": 0.6228,
"step": 300
},
{
"epoch": 1.5774278215223099,
"grad_norm": 3.1037299633026123,
"learning_rate": 0.00010825793454723325,
"loss": 0.7738,
"step": 301
},
{
"epoch": 1.5826771653543306,
"grad_norm": 2.92518949508667,
"learning_rate": 0.00010764747892893723,
"loss": 1.1082,
"step": 302
},
{
"epoch": 1.5879265091863517,
"grad_norm": 2.7783684730529785,
"learning_rate": 0.00010703673650879218,
"loss": 0.7418,
"step": 303
},
{
"epoch": 1.5931758530183728,
"grad_norm": 2.6031315326690674,
"learning_rate": 0.00010642573019134703,
"loss": 0.6829,
"step": 304
},
{
"epoch": 1.5984251968503937,
"grad_norm": 2.9734861850738525,
"learning_rate": 0.00010581448289104758,
"loss": 0.7531,
"step": 305
},
{
"epoch": 1.6036745406824147,
"grad_norm": 2.526393175125122,
"learning_rate": 0.00010520301753137724,
"loss": 0.7303,
"step": 306
},
{
"epoch": 1.6089238845144358,
"grad_norm": 3.016721487045288,
"learning_rate": 0.00010459135704399718,
"loss": 0.9378,
"step": 307
},
{
"epoch": 1.6141732283464567,
"grad_norm": 2.5446736812591553,
"learning_rate": 0.00010397952436788642,
"loss": 0.6229,
"step": 308
},
{
"epoch": 1.6194225721784776,
"grad_norm": 2.3816916942596436,
"learning_rate": 0.00010336754244848157,
"loss": 0.6167,
"step": 309
},
{
"epoch": 1.6246719160104988,
"grad_norm": 2.27657413482666,
"learning_rate": 0.00010275543423681621,
"loss": 0.5536,
"step": 310
},
{
"epoch": 1.6299212598425197,
"grad_norm": 2.585496664047241,
"learning_rate": 0.00010214322268866032,
"loss": 0.5935,
"step": 311
},
{
"epoch": 1.6351706036745406,
"grad_norm": 2.376311779022217,
"learning_rate": 0.00010153093076365923,
"loss": 0.6658,
"step": 312
},
{
"epoch": 1.6404199475065617,
"grad_norm": 2.830921173095703,
"learning_rate": 0.00010091858142447265,
"loss": 0.7428,
"step": 313
},
{
"epoch": 1.6456692913385826,
"grad_norm": 2.712291717529297,
"learning_rate": 0.00010030619763591347,
"loss": 0.7036,
"step": 314
},
{
"epoch": 1.6509186351706036,
"grad_norm": 3.999314546585083,
"learning_rate": 9.969380236408656e-05,
"loss": 1.0524,
"step": 315
},
{
"epoch": 1.6561679790026247,
"grad_norm": 3.0849058628082275,
"learning_rate": 9.908141857552737e-05,
"loss": 0.6273,
"step": 316
},
{
"epoch": 1.6614173228346458,
"grad_norm": 2.9424805641174316,
"learning_rate": 9.846906923634079e-05,
"loss": 0.6581,
"step": 317
},
{
"epoch": 1.6666666666666665,
"grad_norm": 2.7930872440338135,
"learning_rate": 9.78567773113397e-05,
"loss": 0.7108,
"step": 318
},
{
"epoch": 1.6719160104986877,
"grad_norm": 2.6098508834838867,
"learning_rate": 9.724456576318381e-05,
"loss": 0.7179,
"step": 319
},
{
"epoch": 1.6771653543307088,
"grad_norm": 3.0556881427764893,
"learning_rate": 9.663245755151846e-05,
"loss": 0.7449,
"step": 320
},
{
"epoch": 1.6824146981627297,
"grad_norm": 2.536299705505371,
"learning_rate": 9.602047563211359e-05,
"loss": 0.5627,
"step": 321
},
{
"epoch": 1.6876640419947506,
"grad_norm": 2.9197168350219727,
"learning_rate": 9.540864295600283e-05,
"loss": 0.6701,
"step": 322
},
{
"epoch": 1.6929133858267718,
"grad_norm": 2.77664852142334,
"learning_rate": 9.479698246862276e-05,
"loss": 0.8544,
"step": 323
},
{
"epoch": 1.6981627296587927,
"grad_norm": 2.833603858947754,
"learning_rate": 9.418551710895243e-05,
"loss": 0.7597,
"step": 324
},
{
"epoch": 1.7034120734908136,
"grad_norm": 2.527200937271118,
"learning_rate": 9.357426980865301e-05,
"loss": 0.7198,
"step": 325
},
{
"epoch": 1.7086614173228347,
"grad_norm": 2.7248785495758057,
"learning_rate": 9.296326349120785e-05,
"loss": 0.7121,
"step": 326
},
{
"epoch": 1.7139107611548556,
"grad_norm": 2.6824331283569336,
"learning_rate": 9.235252107106279e-05,
"loss": 0.884,
"step": 327
},
{
"epoch": 1.7191601049868765,
"grad_norm": 2.580566167831421,
"learning_rate": 9.174206545276677e-05,
"loss": 0.6406,
"step": 328
},
{
"epoch": 1.7244094488188977,
"grad_norm": 2.741162061691284,
"learning_rate": 9.113191953011287e-05,
"loss": 0.8105,
"step": 329
},
{
"epoch": 1.7296587926509186,
"grad_norm": 2.2699320316314697,
"learning_rate": 9.052210618527966e-05,
"loss": 0.598,
"step": 330
},
{
"epoch": 1.7349081364829395,
"grad_norm": 2.4851629734039307,
"learning_rate": 8.991264828797319e-05,
"loss": 0.7384,
"step": 331
},
{
"epoch": 1.7401574803149606,
"grad_norm": 2.476698637008667,
"learning_rate": 8.930356869456919e-05,
"loss": 0.6595,
"step": 332
},
{
"epoch": 1.7454068241469818,
"grad_norm": 2.465303897857666,
"learning_rate": 8.869489024725595e-05,
"loss": 0.7266,
"step": 333
},
{
"epoch": 1.7506561679790025,
"grad_norm": 2.6506006717681885,
"learning_rate": 8.808663577317764e-05,
"loss": 0.8465,
"step": 334
},
{
"epoch": 1.7559055118110236,
"grad_norm": 2.6736481189727783,
"learning_rate": 8.747882808357828e-05,
"loss": 0.7328,
"step": 335
},
{
"epoch": 1.7611548556430447,
"grad_norm": 2.2621066570281982,
"learning_rate": 8.687148997294621e-05,
"loss": 0.5939,
"step": 336
},
{
"epoch": 1.7664041994750657,
"grad_norm": 2.6498119831085205,
"learning_rate": 8.626464421815919e-05,
"loss": 0.6821,
"step": 337
},
{
"epoch": 1.7716535433070866,
"grad_norm": 2.5098793506622314,
"learning_rate": 8.565831357763039e-05,
"loss": 0.6737,
"step": 338
},
{
"epoch": 1.7769028871391077,
"grad_norm": 2.26190447807312,
"learning_rate": 8.505252079045458e-05,
"loss": 0.6358,
"step": 339
},
{
"epoch": 1.7821522309711286,
"grad_norm": 2.6021599769592285,
"learning_rate": 8.444728857555572e-05,
"loss": 0.764,
"step": 340
},
{
"epoch": 1.7874015748031495,
"grad_norm": 2.805920362472534,
"learning_rate": 8.384263963083453e-05,
"loss": 0.7641,
"step": 341
},
{
"epoch": 1.7926509186351707,
"grad_norm": 3.371159791946411,
"learning_rate": 8.323859663231768e-05,
"loss": 0.8458,
"step": 342
},
{
"epoch": 1.7979002624671916,
"grad_norm": 3.0499484539031982,
"learning_rate": 8.263518223330697e-05,
"loss": 0.7696,
"step": 343
},
{
"epoch": 1.8031496062992125,
"grad_norm": 3.3315682411193848,
"learning_rate": 8.203241906353014e-05,
"loss": 0.8452,
"step": 344
},
{
"epoch": 1.8083989501312336,
"grad_norm": 2.8910834789276123,
"learning_rate": 8.143032972829183e-05,
"loss": 0.7281,
"step": 345
},
{
"epoch": 1.8136482939632546,
"grad_norm": 2.424577236175537,
"learning_rate": 8.082893680762619e-05,
"loss": 0.5982,
"step": 346
},
{
"epoch": 1.8188976377952755,
"grad_norm": 2.4802539348602295,
"learning_rate": 8.022826285544968e-05,
"loss": 0.4913,
"step": 347
},
{
"epoch": 1.8241469816272966,
"grad_norm": 3.608314037322998,
"learning_rate": 7.96283303987156e-05,
"loss": 0.965,
"step": 348
},
{
"epoch": 1.8293963254593177,
"grad_norm": 2.9640915393829346,
"learning_rate": 7.902916193656898e-05,
"loss": 0.8494,
"step": 349
},
{
"epoch": 1.8346456692913384,
"grad_norm": 3.1008963584899902,
"learning_rate": 7.843077993950302e-05,
"loss": 0.9431,
"step": 350
},
{
"epoch": 1.8398950131233596,
"grad_norm": 3.1320395469665527,
"learning_rate": 7.783320684851614e-05,
"loss": 1.0122,
"step": 351
},
{
"epoch": 1.8451443569553807,
"grad_norm": 2.527387857437134,
"learning_rate": 7.72364650742707e-05,
"loss": 0.7018,
"step": 352
},
{
"epoch": 1.8503937007874016,
"grad_norm": 2.8600051403045654,
"learning_rate": 7.664057699625214e-05,
"loss": 0.7482,
"step": 353
},
{
"epoch": 1.8556430446194225,
"grad_norm": 2.6719634532928467,
"learning_rate": 7.604556496193015e-05,
"loss": 0.7669,
"step": 354
},
{
"epoch": 1.8608923884514437,
"grad_norm": 2.4626736640930176,
"learning_rate": 7.54514512859201e-05,
"loss": 0.6617,
"step": 355
},
{
"epoch": 1.8661417322834646,
"grad_norm": 3.1328530311584473,
"learning_rate": 7.485825824914659e-05,
"loss": 0.9866,
"step": 356
},
{
"epoch": 1.8713910761154855,
"grad_norm": 2.8781819343566895,
"learning_rate": 7.426600809800752e-05,
"loss": 0.8311,
"step": 357
},
{
"epoch": 1.8766404199475066,
"grad_norm": 2.889805793762207,
"learning_rate": 7.36747230435401e-05,
"loss": 0.6937,
"step": 358
},
{
"epoch": 1.8818897637795275,
"grad_norm": 3.0408389568328857,
"learning_rate": 7.308442526058756e-05,
"loss": 0.6836,
"step": 359
},
{
"epoch": 1.8871391076115485,
"grad_norm": 3.536181926727295,
"learning_rate": 7.249513688696786e-05,
"loss": 1.0847,
"step": 360
},
{
"epoch": 1.8923884514435696,
"grad_norm": 3.20278263092041,
"learning_rate": 7.190688002264308e-05,
"loss": 1.0111,
"step": 361
},
{
"epoch": 1.8976377952755905,
"grad_norm": 2.5212113857269287,
"learning_rate": 7.131967672889101e-05,
"loss": 0.5622,
"step": 362
},
{
"epoch": 1.9028871391076114,
"grad_norm": 2.54811429977417,
"learning_rate": 7.073354902747741e-05,
"loss": 0.6633,
"step": 363
},
{
"epoch": 1.9081364829396326,
"grad_norm": 2.259267568588257,
"learning_rate": 7.014851889983057e-05,
"loss": 0.5843,
"step": 364
},
{
"epoch": 1.9133858267716537,
"grad_norm": 2.4579734802246094,
"learning_rate": 6.95646082862164e-05,
"loss": 0.573,
"step": 365
},
{
"epoch": 1.9186351706036744,
"grad_norm": 3.317549228668213,
"learning_rate": 6.898183908491617e-05,
"loss": 0.9837,
"step": 366
},
{
"epoch": 1.9238845144356955,
"grad_norm": 2.377362012863159,
"learning_rate": 6.840023315140475e-05,
"loss": 0.7062,
"step": 367
},
{
"epoch": 1.9291338582677167,
"grad_norm": 2.8858208656311035,
"learning_rate": 6.781981229753145e-05,
"loss": 0.7657,
"step": 368
},
{
"epoch": 1.9343832020997376,
"grad_norm": 2.407649278640747,
"learning_rate": 6.724059829070158e-05,
"loss": 0.6103,
"step": 369
},
{
"epoch": 1.9396325459317585,
"grad_norm": 2.943190813064575,
"learning_rate": 6.666261285306047e-05,
"loss": 0.9676,
"step": 370
},
{
"epoch": 1.9448818897637796,
"grad_norm": 2.791794538497925,
"learning_rate": 6.608587766067852e-05,
"loss": 0.7191,
"step": 371
},
{
"epoch": 1.9501312335958005,
"grad_norm": 2.5075485706329346,
"learning_rate": 6.551041434273861e-05,
"loss": 0.634,
"step": 372
},
{
"epoch": 1.9553805774278215,
"grad_norm": 2.816528797149658,
"learning_rate": 6.493624448072457e-05,
"loss": 0.7335,
"step": 373
},
{
"epoch": 1.9606299212598426,
"grad_norm": 2.5737648010253906,
"learning_rate": 6.43633896076122e-05,
"loss": 0.6795,
"step": 374
},
{
"epoch": 1.9658792650918635,
"grad_norm": 2.7753782272338867,
"learning_rate": 6.379187120706138e-05,
"loss": 0.721,
"step": 375
},
{
"epoch": 1.9711286089238844,
"grad_norm": 2.586721181869507,
"learning_rate": 6.322171071261071e-05,
"loss": 0.6044,
"step": 376
},
{
"epoch": 1.9763779527559056,
"grad_norm": 2.821720838546753,
"learning_rate": 6.26529295068733e-05,
"loss": 0.687,
"step": 377
},
{
"epoch": 1.9816272965879265,
"grad_norm": 2.5189385414123535,
"learning_rate": 6.208554892073528e-05,
"loss": 0.6973,
"step": 378
},
{
"epoch": 1.9868766404199474,
"grad_norm": 2.7033824920654297,
"learning_rate": 6.151959023255545e-05,
"loss": 0.5675,
"step": 379
},
{
"epoch": 1.9921259842519685,
"grad_norm": 2.864698886871338,
"learning_rate": 6.095507466736763e-05,
"loss": 0.8994,
"step": 380
},
{
"epoch": 1.9973753280839897,
"grad_norm": 2.771103858947754,
"learning_rate": 6.039202339608432e-05,
"loss": 0.7414,
"step": 381
},
{
"epoch": 2.0,
"grad_norm": 4.474254131317139,
"learning_rate": 5.983045753470308e-05,
"loss": 0.8273,
"step": 382
},
{
"epoch": 2.005249343832021,
"grad_norm": 1.8631569147109985,
"learning_rate": 5.927039814351426e-05,
"loss": 0.3455,
"step": 383
},
{
"epoch": 2.010498687664042,
"grad_norm": 1.857398271560669,
"learning_rate": 5.8711866226311553e-05,
"loss": 0.3454,
"step": 384
},
{
"epoch": 2.015748031496063,
"grad_norm": 2.315157651901245,
"learning_rate": 5.8154882729603876e-05,
"loss": 0.4925,
"step": 385
},
{
"epoch": 2.020997375328084,
"grad_norm": 2.0346195697784424,
"learning_rate": 5.7599468541830356e-05,
"loss": 0.3541,
"step": 386
},
{
"epoch": 2.026246719160105,
"grad_norm": 2.2926833629608154,
"learning_rate": 5.7045644492576346e-05,
"loss": 0.3705,
"step": 387
},
{
"epoch": 2.031496062992126,
"grad_norm": 1.8985930681228638,
"learning_rate": 5.64934313517927e-05,
"loss": 0.3665,
"step": 388
},
{
"epoch": 2.036745406824147,
"grad_norm": 1.6224257946014404,
"learning_rate": 5.5942849829016695e-05,
"loss": 0.2719,
"step": 389
},
{
"epoch": 2.041994750656168,
"grad_norm": 1.9239685535430908,
"learning_rate": 5.5393920572595356e-05,
"loss": 0.2669,
"step": 390
},
{
"epoch": 2.047244094488189,
"grad_norm": 1.7690950632095337,
"learning_rate": 5.484666416891109e-05,
"loss": 0.2509,
"step": 391
},
{
"epoch": 2.05249343832021,
"grad_norm": 2.04697322845459,
"learning_rate": 5.430110114160964e-05,
"loss": 0.2676,
"step": 392
},
{
"epoch": 2.057742782152231,
"grad_norm": 2.4531140327453613,
"learning_rate": 5.375725195083046e-05,
"loss": 0.3482,
"step": 393
},
{
"epoch": 2.062992125984252,
"grad_norm": 2.427957057952881,
"learning_rate": 5.321513699243924e-05,
"loss": 0.3182,
"step": 394
},
{
"epoch": 2.068241469816273,
"grad_norm": 2.123382091522217,
"learning_rate": 5.2674776597263186e-05,
"loss": 0.2888,
"step": 395
},
{
"epoch": 2.073490813648294,
"grad_norm": 3.2138147354125977,
"learning_rate": 5.2136191030328455e-05,
"loss": 0.3561,
"step": 396
},
{
"epoch": 2.078740157480315,
"grad_norm": 2.332437753677368,
"learning_rate": 5.159940049010015e-05,
"loss": 0.3039,
"step": 397
},
{
"epoch": 2.083989501312336,
"grad_norm": 2.8219337463378906,
"learning_rate": 5.106442510772489e-05,
"loss": 0.3129,
"step": 398
},
{
"epoch": 2.089238845144357,
"grad_norm": 2.9364748001098633,
"learning_rate": 5.0531284946275784e-05,
"loss": 0.2937,
"step": 399
},
{
"epoch": 2.094488188976378,
"grad_norm": 2.8431618213653564,
"learning_rate": 5.000000000000002e-05,
"loss": 0.3048,
"step": 400
}
],
"logging_steps": 1,
"max_steps": 570,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 9.535375299019562e+16,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}