CHAI_SFT_model_8b / trainer_state.json
chancharikm's picture
End of training
2b63ddc verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 741,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.004056795131845842,
"grad_norm": 20.09847465489327,
"learning_rate": 0.0,
"loss": 1.1858,
"step": 1
},
{
"epoch": 0.008113590263691683,
"grad_norm": 19.197153778335704,
"learning_rate": 7.894736842105263e-07,
"loss": 1.1624,
"step": 2
},
{
"epoch": 0.012170385395537525,
"grad_norm": 20.179166909378754,
"learning_rate": 1.5789473684210526e-06,
"loss": 1.1569,
"step": 3
},
{
"epoch": 0.016227180527383367,
"grad_norm": 17.474818335029532,
"learning_rate": 2.368421052631579e-06,
"loss": 1.0849,
"step": 4
},
{
"epoch": 0.02028397565922921,
"grad_norm": 14.444584965019517,
"learning_rate": 3.157894736842105e-06,
"loss": 1.0109,
"step": 5
},
{
"epoch": 0.02434077079107505,
"grad_norm": 7.870161156231951,
"learning_rate": 3.9473684210526315e-06,
"loss": 0.8763,
"step": 6
},
{
"epoch": 0.028397565922920892,
"grad_norm": 6.250576323149918,
"learning_rate": 4.736842105263158e-06,
"loss": 0.8331,
"step": 7
},
{
"epoch": 0.032454361054766734,
"grad_norm": 2.8630410669781923,
"learning_rate": 5.526315789473684e-06,
"loss": 0.6998,
"step": 8
},
{
"epoch": 0.036511156186612576,
"grad_norm": 2.7842834846404165,
"learning_rate": 6.31578947368421e-06,
"loss": 0.7146,
"step": 9
},
{
"epoch": 0.04056795131845842,
"grad_norm": 1.9178922207738491,
"learning_rate": 7.105263157894737e-06,
"loss": 0.6854,
"step": 10
},
{
"epoch": 0.04462474645030426,
"grad_norm": 2.71533640759056,
"learning_rate": 7.894736842105263e-06,
"loss": 0.579,
"step": 11
},
{
"epoch": 0.0486815415821501,
"grad_norm": 1.7283014150811757,
"learning_rate": 8.68421052631579e-06,
"loss": 0.5833,
"step": 12
},
{
"epoch": 0.05273833671399594,
"grad_norm": 1.2126394020230842,
"learning_rate": 9.473684210526315e-06,
"loss": 0.5641,
"step": 13
},
{
"epoch": 0.056795131845841784,
"grad_norm": 0.954640851521375,
"learning_rate": 1.0263157894736843e-05,
"loss": 0.5682,
"step": 14
},
{
"epoch": 0.060851926977687626,
"grad_norm": 1.1771450912772898,
"learning_rate": 1.1052631578947368e-05,
"loss": 0.5534,
"step": 15
},
{
"epoch": 0.06490872210953347,
"grad_norm": 1.154307340565416,
"learning_rate": 1.1842105263157895e-05,
"loss": 0.4819,
"step": 16
},
{
"epoch": 0.06896551724137931,
"grad_norm": 0.8247855133987926,
"learning_rate": 1.263157894736842e-05,
"loss": 0.5029,
"step": 17
},
{
"epoch": 0.07302231237322515,
"grad_norm": 0.928406591537564,
"learning_rate": 1.3421052631578948e-05,
"loss": 0.5187,
"step": 18
},
{
"epoch": 0.07707910750507099,
"grad_norm": 0.7803328831884472,
"learning_rate": 1.4210526315789473e-05,
"loss": 0.4886,
"step": 19
},
{
"epoch": 0.08113590263691683,
"grad_norm": 0.8210326318273155,
"learning_rate": 1.5e-05,
"loss": 0.5158,
"step": 20
},
{
"epoch": 0.08519269776876268,
"grad_norm": 0.7086772704550919,
"learning_rate": 1.5789473684210526e-05,
"loss": 0.4991,
"step": 21
},
{
"epoch": 0.08924949290060852,
"grad_norm": 0.7734193234814034,
"learning_rate": 1.6578947368421053e-05,
"loss": 0.4783,
"step": 22
},
{
"epoch": 0.09330628803245436,
"grad_norm": 0.9918859015816037,
"learning_rate": 1.736842105263158e-05,
"loss": 0.539,
"step": 23
},
{
"epoch": 0.0973630831643002,
"grad_norm": 0.6852451435428935,
"learning_rate": 1.8157894736842107e-05,
"loss": 0.49,
"step": 24
},
{
"epoch": 0.10141987829614604,
"grad_norm": 0.8715903302725302,
"learning_rate": 1.894736842105263e-05,
"loss": 0.5366,
"step": 25
},
{
"epoch": 0.10547667342799188,
"grad_norm": 0.7938531177188995,
"learning_rate": 1.9736842105263158e-05,
"loss": 0.5248,
"step": 26
},
{
"epoch": 0.10953346855983773,
"grad_norm": 0.6618117922357281,
"learning_rate": 2.0526315789473685e-05,
"loss": 0.5048,
"step": 27
},
{
"epoch": 0.11359026369168357,
"grad_norm": 0.6789614073719777,
"learning_rate": 2.1315789473684212e-05,
"loss": 0.4651,
"step": 28
},
{
"epoch": 0.11764705882352941,
"grad_norm": 1.0018540392183581,
"learning_rate": 2.2105263157894736e-05,
"loss": 0.4712,
"step": 29
},
{
"epoch": 0.12170385395537525,
"grad_norm": 0.7265683281685053,
"learning_rate": 2.2894736842105263e-05,
"loss": 0.5025,
"step": 30
},
{
"epoch": 0.1257606490872211,
"grad_norm": 0.68937690405201,
"learning_rate": 2.368421052631579e-05,
"loss": 0.4577,
"step": 31
},
{
"epoch": 0.12981744421906694,
"grad_norm": 0.6540764219807236,
"learning_rate": 2.4473684210526318e-05,
"loss": 0.4556,
"step": 32
},
{
"epoch": 0.13387423935091278,
"grad_norm": 0.7102940446282798,
"learning_rate": 2.526315789473684e-05,
"loss": 0.4571,
"step": 33
},
{
"epoch": 0.13793103448275862,
"grad_norm": 0.8142861240579994,
"learning_rate": 2.605263157894737e-05,
"loss": 0.4565,
"step": 34
},
{
"epoch": 0.14198782961460446,
"grad_norm": 0.7130725343973457,
"learning_rate": 2.6842105263157896e-05,
"loss": 0.4252,
"step": 35
},
{
"epoch": 0.1460446247464503,
"grad_norm": 0.7733358275781269,
"learning_rate": 2.7631578947368423e-05,
"loss": 0.4433,
"step": 36
},
{
"epoch": 0.15010141987829614,
"grad_norm": 0.8227021020177967,
"learning_rate": 2.8421052631578946e-05,
"loss": 0.4772,
"step": 37
},
{
"epoch": 0.15415821501014199,
"grad_norm": 0.8368912992865613,
"learning_rate": 2.9210526315789474e-05,
"loss": 0.4375,
"step": 38
},
{
"epoch": 0.15821501014198783,
"grad_norm": 0.6687445542989592,
"learning_rate": 3e-05,
"loss": 0.4763,
"step": 39
},
{
"epoch": 0.16227180527383367,
"grad_norm": 0.7016093483692153,
"learning_rate": 2.999985022144514e-05,
"loss": 0.4372,
"step": 40
},
{
"epoch": 0.1663286004056795,
"grad_norm": 0.6478467702162415,
"learning_rate": 2.99994008887717e-05,
"loss": 0.4188,
"step": 41
},
{
"epoch": 0.17038539553752535,
"grad_norm": 0.7186030068739142,
"learning_rate": 2.9998652010953068e-05,
"loss": 0.4583,
"step": 42
},
{
"epoch": 0.1744421906693712,
"grad_norm": 0.6220025087141262,
"learning_rate": 2.9997603602944693e-05,
"loss": 0.4265,
"step": 43
},
{
"epoch": 0.17849898580121704,
"grad_norm": 0.6854775470405489,
"learning_rate": 2.999625568568378e-05,
"loss": 0.4385,
"step": 44
},
{
"epoch": 0.18255578093306288,
"grad_norm": 0.6463932357168245,
"learning_rate": 2.9994608286088872e-05,
"loss": 0.4346,
"step": 45
},
{
"epoch": 0.18661257606490872,
"grad_norm": 1.3944614966148847,
"learning_rate": 2.999266143705932e-05,
"loss": 0.4608,
"step": 46
},
{
"epoch": 0.19066937119675456,
"grad_norm": 0.6199390650948697,
"learning_rate": 2.999041517747462e-05,
"loss": 0.4404,
"step": 47
},
{
"epoch": 0.1947261663286004,
"grad_norm": 0.6467646585862439,
"learning_rate": 2.9987869552193653e-05,
"loss": 0.4208,
"step": 48
},
{
"epoch": 0.19878296146044624,
"grad_norm": 0.6807600202518035,
"learning_rate": 2.9985024612053743e-05,
"loss": 0.3992,
"step": 49
},
{
"epoch": 0.2028397565922921,
"grad_norm": 0.587429199431518,
"learning_rate": 2.9981880413869707e-05,
"loss": 0.3999,
"step": 50
},
{
"epoch": 0.20689655172413793,
"grad_norm": 0.543346791231233,
"learning_rate": 2.9978437020432673e-05,
"loss": 0.4206,
"step": 51
},
{
"epoch": 0.21095334685598377,
"grad_norm": 0.6110794893612019,
"learning_rate": 2.997469450050883e-05,
"loss": 0.422,
"step": 52
},
{
"epoch": 0.2150101419878296,
"grad_norm": 0.6418516554131751,
"learning_rate": 2.997065292883808e-05,
"loss": 0.4132,
"step": 53
},
{
"epoch": 0.21906693711967545,
"grad_norm": 0.6240526477144251,
"learning_rate": 2.9966312386132533e-05,
"loss": 0.4014,
"step": 54
},
{
"epoch": 0.2231237322515213,
"grad_norm": 0.8051120973797415,
"learning_rate": 2.9961672959074872e-05,
"loss": 0.4125,
"step": 55
},
{
"epoch": 0.22718052738336714,
"grad_norm": 0.6929169486054979,
"learning_rate": 2.995673474031666e-05,
"loss": 0.4394,
"step": 56
},
{
"epoch": 0.23123732251521298,
"grad_norm": 0.7718108649153836,
"learning_rate": 2.9951497828476462e-05,
"loss": 0.4109,
"step": 57
},
{
"epoch": 0.23529411764705882,
"grad_norm": 0.7605237120120856,
"learning_rate": 2.9945962328137898e-05,
"loss": 0.4078,
"step": 58
},
{
"epoch": 0.23935091277890466,
"grad_norm": 0.6339295966770117,
"learning_rate": 2.9940128349847526e-05,
"loss": 0.3654,
"step": 59
},
{
"epoch": 0.2434077079107505,
"grad_norm": 0.6020493776861426,
"learning_rate": 2.993399601011266e-05,
"loss": 0.378,
"step": 60
},
{
"epoch": 0.24746450304259635,
"grad_norm": 0.6540298023424858,
"learning_rate": 2.992756543139904e-05,
"loss": 0.3977,
"step": 61
},
{
"epoch": 0.2515212981744422,
"grad_norm": 1.240263594935057,
"learning_rate": 2.9920836742128352e-05,
"loss": 0.3893,
"step": 62
},
{
"epoch": 0.25557809330628806,
"grad_norm": 0.644628703604057,
"learning_rate": 2.9913810076675722e-05,
"loss": 0.384,
"step": 63
},
{
"epoch": 0.25963488843813387,
"grad_norm": 0.5971277516452573,
"learning_rate": 2.990648557536699e-05,
"loss": 0.3573,
"step": 64
},
{
"epoch": 0.26369168356997974,
"grad_norm": 0.5937119511367491,
"learning_rate": 2.9898863384475917e-05,
"loss": 0.3329,
"step": 65
},
{
"epoch": 0.26774847870182555,
"grad_norm": 0.597778784125507,
"learning_rate": 2.9890943656221262e-05,
"loss": 0.3359,
"step": 66
},
{
"epoch": 0.2718052738336714,
"grad_norm": 0.7142578966836572,
"learning_rate": 2.9882726548763765e-05,
"loss": 0.3707,
"step": 67
},
{
"epoch": 0.27586206896551724,
"grad_norm": 2.576807678693372,
"learning_rate": 2.9874212226202946e-05,
"loss": 0.3975,
"step": 68
},
{
"epoch": 0.2799188640973631,
"grad_norm": 0.7335503963481053,
"learning_rate": 2.9865400858573866e-05,
"loss": 0.3639,
"step": 69
},
{
"epoch": 0.2839756592292089,
"grad_norm": 0.6240440547685594,
"learning_rate": 2.985629262184371e-05,
"loss": 0.3945,
"step": 70
},
{
"epoch": 0.2880324543610548,
"grad_norm": 1.3946399019773081,
"learning_rate": 2.9846887697908292e-05,
"loss": 0.3875,
"step": 71
},
{
"epoch": 0.2920892494929006,
"grad_norm": 0.6980597507442647,
"learning_rate": 2.9837186274588388e-05,
"loss": 0.3666,
"step": 72
},
{
"epoch": 0.2961460446247465,
"grad_norm": 0.94558927807513,
"learning_rate": 2.9827188545626026e-05,
"loss": 0.3371,
"step": 73
},
{
"epoch": 0.3002028397565923,
"grad_norm": 0.6265421084575312,
"learning_rate": 2.9816894710680595e-05,
"loss": 0.3472,
"step": 74
},
{
"epoch": 0.30425963488843816,
"grad_norm": 0.6480112949037701,
"learning_rate": 2.9806304975324853e-05,
"loss": 0.3787,
"step": 75
},
{
"epoch": 0.30831643002028397,
"grad_norm": 0.6549385961737066,
"learning_rate": 2.9795419551040836e-05,
"loss": 0.3422,
"step": 76
},
{
"epoch": 0.31237322515212984,
"grad_norm": 0.6523717282727239,
"learning_rate": 2.9784238655215627e-05,
"loss": 0.3424,
"step": 77
},
{
"epoch": 0.31643002028397565,
"grad_norm": 0.734607800437923,
"learning_rate": 2.9772762511137018e-05,
"loss": 0.3363,
"step": 78
},
{
"epoch": 0.3204868154158215,
"grad_norm": 0.6192307084985058,
"learning_rate": 2.976099134798904e-05,
"loss": 0.3256,
"step": 79
},
{
"epoch": 0.32454361054766734,
"grad_norm": 0.753915566361977,
"learning_rate": 2.97489254008474e-05,
"loss": 0.3485,
"step": 80
},
{
"epoch": 0.3286004056795132,
"grad_norm": 0.7608348621933458,
"learning_rate": 2.973656491067479e-05,
"loss": 0.3706,
"step": 81
},
{
"epoch": 0.332657200811359,
"grad_norm": 0.844334116903445,
"learning_rate": 2.972391012431605e-05,
"loss": 0.3272,
"step": 82
},
{
"epoch": 0.3367139959432049,
"grad_norm": 0.65086089898772,
"learning_rate": 2.9710961294493258e-05,
"loss": 0.333,
"step": 83
},
{
"epoch": 0.3407707910750507,
"grad_norm": 0.7018751557669878,
"learning_rate": 2.96977186798007e-05,
"loss": 0.347,
"step": 84
},
{
"epoch": 0.3448275862068966,
"grad_norm": 0.677197135043688,
"learning_rate": 2.9684182544699645e-05,
"loss": 0.3006,
"step": 85
},
{
"epoch": 0.3488843813387424,
"grad_norm": 0.7323590164129933,
"learning_rate": 2.9670353159513148e-05,
"loss": 0.3158,
"step": 86
},
{
"epoch": 0.35294117647058826,
"grad_norm": 2.9063859065858226,
"learning_rate": 2.965623080042057e-05,
"loss": 0.3582,
"step": 87
},
{
"epoch": 0.35699797160243407,
"grad_norm": 0.7448699680639598,
"learning_rate": 2.964181574945213e-05,
"loss": 0.3002,
"step": 88
},
{
"epoch": 0.36105476673427994,
"grad_norm": 0.7343903430492267,
"learning_rate": 2.9627108294483213e-05,
"loss": 0.3232,
"step": 89
},
{
"epoch": 0.36511156186612576,
"grad_norm": 0.7643329255268317,
"learning_rate": 2.9612108729228678e-05,
"loss": 0.3568,
"step": 90
},
{
"epoch": 0.3691683569979716,
"grad_norm": 0.7310512434908586,
"learning_rate": 2.9596817353236945e-05,
"loss": 0.3072,
"step": 91
},
{
"epoch": 0.37322515212981744,
"grad_norm": 1.3438138182721553,
"learning_rate": 2.9581234471884047e-05,
"loss": 0.3363,
"step": 92
},
{
"epoch": 0.3772819472616633,
"grad_norm": 1.5864470945264417,
"learning_rate": 2.9565360396367506e-05,
"loss": 0.2996,
"step": 93
},
{
"epoch": 0.3813387423935091,
"grad_norm": 0.7964568719444062,
"learning_rate": 2.954919544370014e-05,
"loss": 0.347,
"step": 94
},
{
"epoch": 0.385395537525355,
"grad_norm": 0.7064106279304047,
"learning_rate": 2.9532739936703703e-05,
"loss": 0.3332,
"step": 95
},
{
"epoch": 0.3894523326572008,
"grad_norm": 0.6490703458390841,
"learning_rate": 2.9515994204002485e-05,
"loss": 0.2828,
"step": 96
},
{
"epoch": 0.3935091277890467,
"grad_norm": 0.6631023061002581,
"learning_rate": 2.9498958580016692e-05,
"loss": 0.3293,
"step": 97
},
{
"epoch": 0.3975659229208925,
"grad_norm": 1.0292699955420868,
"learning_rate": 2.9481633404955816e-05,
"loss": 0.3071,
"step": 98
},
{
"epoch": 0.40162271805273836,
"grad_norm": 0.7524386552279533,
"learning_rate": 2.9464019024811817e-05,
"loss": 0.293,
"step": 99
},
{
"epoch": 0.4056795131845842,
"grad_norm": 0.6516481911899922,
"learning_rate": 2.9446115791352207e-05,
"loss": 0.3201,
"step": 100
},
{
"epoch": 0.40973630831643004,
"grad_norm": 0.683164430845498,
"learning_rate": 2.9427924062113057e-05,
"loss": 0.2989,
"step": 101
},
{
"epoch": 0.41379310344827586,
"grad_norm": 0.6907076175572305,
"learning_rate": 2.9409444200391807e-05,
"loss": 0.2965,
"step": 102
},
{
"epoch": 0.4178498985801217,
"grad_norm": 0.6354065180379413,
"learning_rate": 2.939067657524007e-05,
"loss": 0.2853,
"step": 103
},
{
"epoch": 0.42190669371196754,
"grad_norm": 0.7712222390874938,
"learning_rate": 2.9371621561456205e-05,
"loss": 0.2996,
"step": 104
},
{
"epoch": 0.4259634888438134,
"grad_norm": 0.7281372221391285,
"learning_rate": 2.935227953957787e-05,
"loss": 0.2882,
"step": 105
},
{
"epoch": 0.4300202839756592,
"grad_norm": 0.6209716277230989,
"learning_rate": 2.9332650895874413e-05,
"loss": 0.2542,
"step": 106
},
{
"epoch": 0.4340770791075051,
"grad_norm": 0.691865325929656,
"learning_rate": 2.9312736022339152e-05,
"loss": 0.2641,
"step": 107
},
{
"epoch": 0.4381338742393509,
"grad_norm": 0.6590079207511579,
"learning_rate": 2.929253531668155e-05,
"loss": 0.2876,
"step": 108
},
{
"epoch": 0.4421906693711968,
"grad_norm": 0.7713287216103389,
"learning_rate": 2.9272049182319276e-05,
"loss": 0.2971,
"step": 109
},
{
"epoch": 0.4462474645030426,
"grad_norm": 0.6020654700963447,
"learning_rate": 2.9251278028370137e-05,
"loss": 0.2656,
"step": 110
},
{
"epoch": 0.45030425963488846,
"grad_norm": 0.7933746678999056,
"learning_rate": 2.923022226964393e-05,
"loss": 0.2929,
"step": 111
},
{
"epoch": 0.4543610547667343,
"grad_norm": 0.7365666480876217,
"learning_rate": 2.920888232663413e-05,
"loss": 0.2905,
"step": 112
},
{
"epoch": 0.45841784989858014,
"grad_norm": 0.7426222606192678,
"learning_rate": 2.9187258625509518e-05,
"loss": 0.2974,
"step": 113
},
{
"epoch": 0.46247464503042596,
"grad_norm": 0.6635176120922496,
"learning_rate": 2.9165351598105657e-05,
"loss": 0.2805,
"step": 114
},
{
"epoch": 0.4665314401622718,
"grad_norm": 0.7359042862527405,
"learning_rate": 2.9143161681916264e-05,
"loss": 0.2666,
"step": 115
},
{
"epoch": 0.47058823529411764,
"grad_norm": 0.5833980665928568,
"learning_rate": 2.912068932008448e-05,
"loss": 0.2437,
"step": 116
},
{
"epoch": 0.4746450304259635,
"grad_norm": 0.7121307261752945,
"learning_rate": 2.9097934961394028e-05,
"loss": 0.2624,
"step": 117
},
{
"epoch": 0.4787018255578093,
"grad_norm": 0.628140379275478,
"learning_rate": 2.9074899060260237e-05,
"loss": 0.2679,
"step": 118
},
{
"epoch": 0.4827586206896552,
"grad_norm": 0.6155641296091218,
"learning_rate": 2.905158207672096e-05,
"loss": 0.2408,
"step": 119
},
{
"epoch": 0.486815415821501,
"grad_norm": 0.669882023645336,
"learning_rate": 2.9027984476427427e-05,
"loss": 0.2681,
"step": 120
},
{
"epoch": 0.4908722109533469,
"grad_norm": 0.5998818630794119,
"learning_rate": 2.9004106730634888e-05,
"loss": 0.2469,
"step": 121
},
{
"epoch": 0.4949290060851927,
"grad_norm": 0.8004135025605319,
"learning_rate": 2.897994931619325e-05,
"loss": 0.2126,
"step": 122
},
{
"epoch": 0.49898580121703856,
"grad_norm": 0.7284465238344607,
"learning_rate": 2.8955512715537523e-05,
"loss": 0.2537,
"step": 123
},
{
"epoch": 0.5030425963488844,
"grad_norm": 0.7338390729083896,
"learning_rate": 2.893079741667821e-05,
"loss": 0.2773,
"step": 124
},
{
"epoch": 0.5070993914807302,
"grad_norm": 0.6486605797373974,
"learning_rate": 2.8905803913191545e-05,
"loss": 0.2264,
"step": 125
},
{
"epoch": 0.5111561866125761,
"grad_norm": 0.6776351745411153,
"learning_rate": 2.8880532704209636e-05,
"loss": 0.2503,
"step": 126
},
{
"epoch": 0.5152129817444219,
"grad_norm": 0.6538874763108758,
"learning_rate": 2.8854984294410513e-05,
"loss": 0.2259,
"step": 127
},
{
"epoch": 0.5192697768762677,
"grad_norm": 0.6108311197354839,
"learning_rate": 2.8829159194008014e-05,
"loss": 0.25,
"step": 128
},
{
"epoch": 0.5233265720081136,
"grad_norm": 0.6002147965900707,
"learning_rate": 2.8803057918741643e-05,
"loss": 0.2893,
"step": 129
},
{
"epoch": 0.5273833671399595,
"grad_norm": 0.6762470121404814,
"learning_rate": 2.877668098986624e-05,
"loss": 0.2466,
"step": 130
},
{
"epoch": 0.5314401622718052,
"grad_norm": 0.6974139790167463,
"learning_rate": 2.8750028934141572e-05,
"loss": 0.2629,
"step": 131
},
{
"epoch": 0.5354969574036511,
"grad_norm": 1.236891646338022,
"learning_rate": 2.8723102283821826e-05,
"loss": 0.2334,
"step": 132
},
{
"epoch": 0.539553752535497,
"grad_norm": 0.7563708424985823,
"learning_rate": 2.8695901576644977e-05,
"loss": 0.2274,
"step": 133
},
{
"epoch": 0.5436105476673428,
"grad_norm": 0.5992720935663637,
"learning_rate": 2.8668427355822036e-05,
"loss": 0.2269,
"step": 134
},
{
"epoch": 0.5476673427991886,
"grad_norm": 0.5919029812932312,
"learning_rate": 2.864068017002621e-05,
"loss": 0.2292,
"step": 135
},
{
"epoch": 0.5517241379310345,
"grad_norm": 0.7568845058840511,
"learning_rate": 2.8612660573381964e-05,
"loss": 0.2354,
"step": 136
},
{
"epoch": 0.5557809330628803,
"grad_norm": 0.6446514196916809,
"learning_rate": 2.8584369125453913e-05,
"loss": 0.2451,
"step": 137
},
{
"epoch": 0.5598377281947262,
"grad_norm": 0.604214552142815,
"learning_rate": 2.855580639123568e-05,
"loss": 0.2107,
"step": 138
},
{
"epoch": 0.563894523326572,
"grad_norm": 0.6522826920802496,
"learning_rate": 2.8526972941138612e-05,
"loss": 0.2237,
"step": 139
},
{
"epoch": 0.5679513184584178,
"grad_norm": 0.563202224656028,
"learning_rate": 2.8497869350980376e-05,
"loss": 0.2003,
"step": 140
},
{
"epoch": 0.5720081135902637,
"grad_norm": 0.8061973371002982,
"learning_rate": 2.8468496201973454e-05,
"loss": 0.2276,
"step": 141
},
{
"epoch": 0.5760649087221096,
"grad_norm": 0.6208681068302313,
"learning_rate": 2.843885408071356e-05,
"loss": 0.2344,
"step": 142
},
{
"epoch": 0.5801217038539553,
"grad_norm": 0.6055109887630478,
"learning_rate": 2.8408943579167905e-05,
"loss": 0.2294,
"step": 143
},
{
"epoch": 0.5841784989858012,
"grad_norm": 0.67040596145082,
"learning_rate": 2.8378765294663373e-05,
"loss": 0.2167,
"step": 144
},
{
"epoch": 0.5882352941176471,
"grad_norm": 0.781769940390108,
"learning_rate": 2.834831982987462e-05,
"loss": 0.21,
"step": 145
},
{
"epoch": 0.592292089249493,
"grad_norm": 0.6780636258565765,
"learning_rate": 2.831760779281201e-05,
"loss": 0.2249,
"step": 146
},
{
"epoch": 0.5963488843813387,
"grad_norm": 0.625921083157255,
"learning_rate": 2.828662979680947e-05,
"loss": 0.2219,
"step": 147
},
{
"epoch": 0.6004056795131846,
"grad_norm": 0.5893412180647384,
"learning_rate": 2.8255386460512273e-05,
"loss": 0.2067,
"step": 148
},
{
"epoch": 0.6044624746450304,
"grad_norm": 0.5705751125220077,
"learning_rate": 2.8223878407864642e-05,
"loss": 0.1833,
"step": 149
},
{
"epoch": 0.6085192697768763,
"grad_norm": 0.5700901015886054,
"learning_rate": 2.8192106268097336e-05,
"loss": 0.1848,
"step": 150
},
{
"epoch": 0.6125760649087221,
"grad_norm": 0.6000865550277297,
"learning_rate": 2.8160070675715038e-05,
"loss": 0.1938,
"step": 151
},
{
"epoch": 0.6166328600405679,
"grad_norm": 0.5791492060921085,
"learning_rate": 2.8127772270483714e-05,
"loss": 0.2085,
"step": 152
},
{
"epoch": 0.6206896551724138,
"grad_norm": 0.6023954274654707,
"learning_rate": 2.8095211697417823e-05,
"loss": 0.2237,
"step": 153
},
{
"epoch": 0.6247464503042597,
"grad_norm": 0.6536097341613256,
"learning_rate": 2.8062389606767445e-05,
"loss": 0.2123,
"step": 154
},
{
"epoch": 0.6288032454361054,
"grad_norm": 0.5127334450984861,
"learning_rate": 2.8029306654005288e-05,
"loss": 0.1829,
"step": 155
},
{
"epoch": 0.6328600405679513,
"grad_norm": 0.5624075392241792,
"learning_rate": 2.7995963499813596e-05,
"loss": 0.2025,
"step": 156
},
{
"epoch": 0.6369168356997972,
"grad_norm": 0.6449150568592729,
"learning_rate": 2.796236081007096e-05,
"loss": 0.1949,
"step": 157
},
{
"epoch": 0.640973630831643,
"grad_norm": 0.701187654484129,
"learning_rate": 2.7928499255839033e-05,
"loss": 0.2151,
"step": 158
},
{
"epoch": 0.6450304259634888,
"grad_norm": 0.6431071185310107,
"learning_rate": 2.789437951334909e-05,
"loss": 0.1695,
"step": 159
},
{
"epoch": 0.6490872210953347,
"grad_norm": 0.883066522041192,
"learning_rate": 2.7860002263988565e-05,
"loss": 0.2035,
"step": 160
},
{
"epoch": 0.6531440162271805,
"grad_norm": 0.8635932764225648,
"learning_rate": 2.7825368194287425e-05,
"loss": 0.1695,
"step": 161
},
{
"epoch": 0.6572008113590264,
"grad_norm": 0.6782443532910216,
"learning_rate": 2.7790477995904454e-05,
"loss": 0.1865,
"step": 162
},
{
"epoch": 0.6612576064908722,
"grad_norm": 0.5924664424714287,
"learning_rate": 2.7755332365613453e-05,
"loss": 0.202,
"step": 163
},
{
"epoch": 0.665314401622718,
"grad_norm": 0.519362457725557,
"learning_rate": 2.7719932005289317e-05,
"loss": 0.1877,
"step": 164
},
{
"epoch": 0.6693711967545639,
"grad_norm": 0.567689209781307,
"learning_rate": 2.768427762189402e-05,
"loss": 0.1943,
"step": 165
},
{
"epoch": 0.6734279918864098,
"grad_norm": 0.5265646643108667,
"learning_rate": 2.7648369927462503e-05,
"loss": 0.1953,
"step": 166
},
{
"epoch": 0.6774847870182555,
"grad_norm": 0.6457809658154948,
"learning_rate": 2.7612209639088433e-05,
"loss": 0.2075,
"step": 167
},
{
"epoch": 0.6815415821501014,
"grad_norm": 0.6522533222938539,
"learning_rate": 2.7575797478909914e-05,
"loss": 0.2209,
"step": 168
},
{
"epoch": 0.6855983772819473,
"grad_norm": 0.5926208068621334,
"learning_rate": 2.7539134174095047e-05,
"loss": 0.1833,
"step": 169
},
{
"epoch": 0.6896551724137931,
"grad_norm": 0.6215684101485875,
"learning_rate": 2.75022204568274e-05,
"loss": 0.1849,
"step": 170
},
{
"epoch": 0.6937119675456389,
"grad_norm": 0.7621052736863643,
"learning_rate": 2.7465057064291405e-05,
"loss": 0.1666,
"step": 171
},
{
"epoch": 0.6977687626774848,
"grad_norm": 0.5424544870247899,
"learning_rate": 2.7427644738657635e-05,
"loss": 0.1672,
"step": 172
},
{
"epoch": 0.7018255578093306,
"grad_norm": 0.5762207841432611,
"learning_rate": 2.738998422706795e-05,
"loss": 0.1725,
"step": 173
},
{
"epoch": 0.7058823529411765,
"grad_norm": 0.5414406787384946,
"learning_rate": 2.735207628162063e-05,
"loss": 0.1631,
"step": 174
},
{
"epoch": 0.7099391480730223,
"grad_norm": 0.6368334896233788,
"learning_rate": 2.7313921659355305e-05,
"loss": 0.1774,
"step": 175
},
{
"epoch": 0.7139959432048681,
"grad_norm": 0.5101994046433813,
"learning_rate": 2.727552112223787e-05,
"loss": 0.1776,
"step": 176
},
{
"epoch": 0.718052738336714,
"grad_norm": 0.6797950370510993,
"learning_rate": 2.7236875437145256e-05,
"loss": 0.165,
"step": 177
},
{
"epoch": 0.7221095334685599,
"grad_norm": 0.6280142554651347,
"learning_rate": 2.71979853758501e-05,
"loss": 0.1904,
"step": 178
},
{
"epoch": 0.7261663286004056,
"grad_norm": 0.5597196718491728,
"learning_rate": 2.715885171500537e-05,
"loss": 0.1679,
"step": 179
},
{
"epoch": 0.7302231237322515,
"grad_norm": 0.49009801560261895,
"learning_rate": 2.7119475236128816e-05,
"loss": 0.1567,
"step": 180
},
{
"epoch": 0.7342799188640974,
"grad_norm": 0.6563071312139116,
"learning_rate": 2.707985672558739e-05,
"loss": 0.1891,
"step": 181
},
{
"epoch": 0.7383367139959433,
"grad_norm": 0.5376302977824341,
"learning_rate": 2.7039996974581524e-05,
"loss": 0.156,
"step": 182
},
{
"epoch": 0.742393509127789,
"grad_norm": 0.5105263903567433,
"learning_rate": 2.6999896779129333e-05,
"loss": 0.157,
"step": 183
},
{
"epoch": 0.7464503042596349,
"grad_norm": 0.5799467332504898,
"learning_rate": 2.6959556940050736e-05,
"loss": 0.169,
"step": 184
},
{
"epoch": 0.7505070993914807,
"grad_norm": 0.59461838466469,
"learning_rate": 2.691897826295143e-05,
"loss": 0.1666,
"step": 185
},
{
"epoch": 0.7545638945233266,
"grad_norm": 0.5812562725844442,
"learning_rate": 2.6878161558206845e-05,
"loss": 0.1757,
"step": 186
},
{
"epoch": 0.7586206896551724,
"grad_norm": 0.5570290569403675,
"learning_rate": 2.6837107640945904e-05,
"loss": 0.1557,
"step": 187
},
{
"epoch": 0.7626774847870182,
"grad_norm": 0.5210782779050981,
"learning_rate": 2.6795817331034803e-05,
"loss": 0.1681,
"step": 188
},
{
"epoch": 0.7667342799188641,
"grad_norm": 0.8814030705223431,
"learning_rate": 2.6754291453060598e-05,
"loss": 0.1892,
"step": 189
},
{
"epoch": 0.77079107505071,
"grad_norm": 0.5918031831634013,
"learning_rate": 2.6712530836314763e-05,
"loss": 0.1574,
"step": 190
},
{
"epoch": 0.7748478701825557,
"grad_norm": 0.7326389750531642,
"learning_rate": 2.6670536314776593e-05,
"loss": 0.1544,
"step": 191
},
{
"epoch": 0.7789046653144016,
"grad_norm": 0.5086355741172542,
"learning_rate": 2.66283087270966e-05,
"loss": 0.1572,
"step": 192
},
{
"epoch": 0.7829614604462475,
"grad_norm": 0.5965212460678235,
"learning_rate": 2.6585848916579723e-05,
"loss": 0.1497,
"step": 193
},
{
"epoch": 0.7870182555780934,
"grad_norm": 0.5692689362307511,
"learning_rate": 2.6543157731168502e-05,
"loss": 0.1618,
"step": 194
},
{
"epoch": 0.7910750507099391,
"grad_norm": 0.516188907315308,
"learning_rate": 2.650023602342614e-05,
"loss": 0.1456,
"step": 195
},
{
"epoch": 0.795131845841785,
"grad_norm": 0.7593371898283678,
"learning_rate": 2.645708465051949e-05,
"loss": 0.1809,
"step": 196
},
{
"epoch": 0.7991886409736308,
"grad_norm": 0.5366328407002237,
"learning_rate": 2.6413704474201924e-05,
"loss": 0.1572,
"step": 197
},
{
"epoch": 0.8032454361054767,
"grad_norm": 0.5277744342808298,
"learning_rate": 2.6370096360796116e-05,
"loss": 0.1593,
"step": 198
},
{
"epoch": 0.8073022312373225,
"grad_norm": 0.5609600455428916,
"learning_rate": 2.6326261181176772e-05,
"loss": 0.1428,
"step": 199
},
{
"epoch": 0.8113590263691683,
"grad_norm": 0.5420590413429403,
"learning_rate": 2.6282199810753192e-05,
"loss": 0.1407,
"step": 200
},
{
"epoch": 0.8154158215010142,
"grad_norm": 0.5956145475058654,
"learning_rate": 2.623791312945184e-05,
"loss": 0.1727,
"step": 201
},
{
"epoch": 0.8194726166328601,
"grad_norm": 0.6915730266004114,
"learning_rate": 2.6193402021698732e-05,
"loss": 0.1514,
"step": 202
},
{
"epoch": 0.8235294117647058,
"grad_norm": 0.5360960125723859,
"learning_rate": 2.6148667376401773e-05,
"loss": 0.1311,
"step": 203
},
{
"epoch": 0.8275862068965517,
"grad_norm": 0.602399284640832,
"learning_rate": 2.610371008693305e-05,
"loss": 0.1511,
"step": 204
},
{
"epoch": 0.8316430020283976,
"grad_norm": 0.5611575786025276,
"learning_rate": 2.6058531051110933e-05,
"loss": 0.1487,
"step": 205
},
{
"epoch": 0.8356997971602435,
"grad_norm": 0.5164559085496964,
"learning_rate": 2.6013131171182185e-05,
"loss": 0.1575,
"step": 206
},
{
"epoch": 0.8397565922920892,
"grad_norm": 0.5445898147983643,
"learning_rate": 2.5967511353803926e-05,
"loss": 0.1569,
"step": 207
},
{
"epoch": 0.8438133874239351,
"grad_norm": 0.4804010780955451,
"learning_rate": 2.592167251002553e-05,
"loss": 0.1236,
"step": 208
},
{
"epoch": 0.847870182555781,
"grad_norm": 0.6503245044438516,
"learning_rate": 2.587561555527044e-05,
"loss": 0.1639,
"step": 209
},
{
"epoch": 0.8519269776876268,
"grad_norm": 0.5951176833693469,
"learning_rate": 2.5829341409317866e-05,
"loss": 0.1674,
"step": 210
},
{
"epoch": 0.8559837728194726,
"grad_norm": 0.49598023403627806,
"learning_rate": 2.578285099628444e-05,
"loss": 0.1497,
"step": 211
},
{
"epoch": 0.8600405679513184,
"grad_norm": 0.5742271108265362,
"learning_rate": 2.5736145244605746e-05,
"loss": 0.154,
"step": 212
},
{
"epoch": 0.8640973630831643,
"grad_norm": 0.5082796879110931,
"learning_rate": 2.5689225087017775e-05,
"loss": 0.1286,
"step": 213
},
{
"epoch": 0.8681541582150102,
"grad_norm": 0.5510606523552836,
"learning_rate": 2.5642091460538324e-05,
"loss": 0.1439,
"step": 214
},
{
"epoch": 0.8722109533468559,
"grad_norm": 0.5447078949028836,
"learning_rate": 2.559474530644825e-05,
"loss": 0.1538,
"step": 215
},
{
"epoch": 0.8762677484787018,
"grad_norm": 0.6167285401328151,
"learning_rate": 2.5547187570272687e-05,
"loss": 0.1545,
"step": 216
},
{
"epoch": 0.8803245436105477,
"grad_norm": 0.50394073309322,
"learning_rate": 2.5499419201762174e-05,
"loss": 0.1191,
"step": 217
},
{
"epoch": 0.8843813387423936,
"grad_norm": 0.5673254976030906,
"learning_rate": 2.5451441154873666e-05,
"loss": 0.137,
"step": 218
},
{
"epoch": 0.8884381338742393,
"grad_norm": 0.5387470774448104,
"learning_rate": 2.5403254387751512e-05,
"loss": 0.129,
"step": 219
},
{
"epoch": 0.8924949290060852,
"grad_norm": 0.4608513157666502,
"learning_rate": 2.5354859862708276e-05,
"loss": 0.1158,
"step": 220
},
{
"epoch": 0.896551724137931,
"grad_norm": 0.5256344466688219,
"learning_rate": 2.530625854620557e-05,
"loss": 0.1222,
"step": 221
},
{
"epoch": 0.9006085192697769,
"grad_norm": 0.5137885015037381,
"learning_rate": 2.5257451408834715e-05,
"loss": 0.1353,
"step": 222
},
{
"epoch": 0.9046653144016227,
"grad_norm": 0.5360332627787981,
"learning_rate": 2.5208439425297385e-05,
"loss": 0.1337,
"step": 223
},
{
"epoch": 0.9087221095334685,
"grad_norm": 0.4975091772962004,
"learning_rate": 2.5159223574386117e-05,
"loss": 0.1563,
"step": 224
},
{
"epoch": 0.9127789046653144,
"grad_norm": 0.5099800307673038,
"learning_rate": 2.5109804838964785e-05,
"loss": 0.143,
"step": 225
},
{
"epoch": 0.9168356997971603,
"grad_norm": 0.46703357661804473,
"learning_rate": 2.5060184205948953e-05,
"loss": 0.1241,
"step": 226
},
{
"epoch": 0.920892494929006,
"grad_norm": 0.5140596431464904,
"learning_rate": 2.5010362666286185e-05,
"loss": 0.1405,
"step": 227
},
{
"epoch": 0.9249492900608519,
"grad_norm": 0.5271654527108036,
"learning_rate": 2.4960341214936244e-05,
"loss": 0.1321,
"step": 228
},
{
"epoch": 0.9290060851926978,
"grad_norm": 0.5768457462148251,
"learning_rate": 2.491012085085122e-05,
"loss": 0.1234,
"step": 229
},
{
"epoch": 0.9330628803245437,
"grad_norm": 0.5454185697970022,
"learning_rate": 2.4859702576955595e-05,
"loss": 0.1349,
"step": 230
},
{
"epoch": 0.9371196754563894,
"grad_norm": 0.5179952256745091,
"learning_rate": 2.480908740012618e-05,
"loss": 0.1193,
"step": 231
},
{
"epoch": 0.9411764705882353,
"grad_norm": 0.4991739609059931,
"learning_rate": 2.4758276331172065e-05,
"loss": 0.1208,
"step": 232
},
{
"epoch": 0.9452332657200812,
"grad_norm": 0.6073345070204801,
"learning_rate": 2.470727038481437e-05,
"loss": 0.114,
"step": 233
},
{
"epoch": 0.949290060851927,
"grad_norm": 0.4556948904301903,
"learning_rate": 2.4656070579666022e-05,
"loss": 0.1129,
"step": 234
},
{
"epoch": 0.9533468559837728,
"grad_norm": 0.6204458532609,
"learning_rate": 2.4604677938211394e-05,
"loss": 0.1411,
"step": 235
},
{
"epoch": 0.9574036511156186,
"grad_norm": 0.46490020601621357,
"learning_rate": 2.45530934867859e-05,
"loss": 0.1103,
"step": 236
},
{
"epoch": 0.9614604462474645,
"grad_norm": 0.5058495256544994,
"learning_rate": 2.4501318255555483e-05,
"loss": 0.1045,
"step": 237
},
{
"epoch": 0.9655172413793104,
"grad_norm": 0.5847707043545994,
"learning_rate": 2.4449353278496052e-05,
"loss": 0.1301,
"step": 238
},
{
"epoch": 0.9695740365111561,
"grad_norm": 0.5156881183238065,
"learning_rate": 2.439719959337283e-05,
"loss": 0.1237,
"step": 239
},
{
"epoch": 0.973630831643002,
"grad_norm": 0.47559822285660797,
"learning_rate": 2.4344858241719625e-05,
"loss": 0.1122,
"step": 240
},
{
"epoch": 0.9776876267748479,
"grad_norm": 0.629693543735342,
"learning_rate": 2.429233026881805e-05,
"loss": 0.1163,
"step": 241
},
{
"epoch": 0.9817444219066938,
"grad_norm": 0.4631203986746975,
"learning_rate": 2.4239616723676603e-05,
"loss": 0.1071,
"step": 242
},
{
"epoch": 0.9858012170385395,
"grad_norm": 0.48166221209061655,
"learning_rate": 2.4186718659009776e-05,
"loss": 0.1106,
"step": 243
},
{
"epoch": 0.9898580121703854,
"grad_norm": 0.4722397090336021,
"learning_rate": 2.4133637131216994e-05,
"loss": 0.1099,
"step": 244
},
{
"epoch": 0.9939148073022313,
"grad_norm": 0.48081357836920197,
"learning_rate": 2.4080373200361527e-05,
"loss": 0.1133,
"step": 245
},
{
"epoch": 0.9979716024340771,
"grad_norm": 0.4827455819559188,
"learning_rate": 2.4026927930149318e-05,
"loss": 0.1201,
"step": 246
},
{
"epoch": 1.0,
"grad_norm": 0.6591980404499224,
"learning_rate": 2.3973302387907742e-05,
"loss": 0.1198,
"step": 247
},
{
"epoch": 1.0040567951318458,
"grad_norm": 0.6168338300228252,
"learning_rate": 2.3919497644564302e-05,
"loss": 0.0973,
"step": 248
},
{
"epoch": 1.0081135902636917,
"grad_norm": 0.43650673897931774,
"learning_rate": 2.3865514774625215e-05,
"loss": 0.0581,
"step": 249
},
{
"epoch": 1.0121703853955375,
"grad_norm": 0.48170811195224444,
"learning_rate": 2.3811354856153996e-05,
"loss": 0.0779,
"step": 250
},
{
"epoch": 1.0162271805273835,
"grad_norm": 0.4159998966040527,
"learning_rate": 2.3757018970749874e-05,
"loss": 0.0693,
"step": 251
},
{
"epoch": 1.0202839756592292,
"grad_norm": 0.534637865441219,
"learning_rate": 2.3702508203526236e-05,
"loss": 0.0743,
"step": 252
},
{
"epoch": 1.024340770791075,
"grad_norm": 0.5428780017401106,
"learning_rate": 2.3647823643088952e-05,
"loss": 0.0824,
"step": 253
},
{
"epoch": 1.028397565922921,
"grad_norm": 0.5058889770865802,
"learning_rate": 2.35929663815146e-05,
"loss": 0.0621,
"step": 254
},
{
"epoch": 1.0324543610547667,
"grad_norm": 0.6657583405724089,
"learning_rate": 2.3537937514328704e-05,
"loss": 0.0735,
"step": 255
},
{
"epoch": 1.0365111561866125,
"grad_norm": 0.43782441423036716,
"learning_rate": 2.3482738140483815e-05,
"loss": 0.074,
"step": 256
},
{
"epoch": 1.0405679513184585,
"grad_norm": 0.5190186796325933,
"learning_rate": 2.3427369362337602e-05,
"loss": 0.0725,
"step": 257
},
{
"epoch": 1.0446247464503042,
"grad_norm": 0.4154897947758653,
"learning_rate": 2.3371832285630805e-05,
"loss": 0.0725,
"step": 258
},
{
"epoch": 1.04868154158215,
"grad_norm": 0.4875026178896022,
"learning_rate": 2.3316128019465165e-05,
"loss": 0.076,
"step": 259
},
{
"epoch": 1.052738336713996,
"grad_norm": 0.4880205996397405,
"learning_rate": 2.326025767628129e-05,
"loss": 0.0875,
"step": 260
},
{
"epoch": 1.0567951318458417,
"grad_norm": 0.457507116532244,
"learning_rate": 2.320422237183641e-05,
"loss": 0.0746,
"step": 261
},
{
"epoch": 1.0608519269776877,
"grad_norm": 0.5011298451992371,
"learning_rate": 2.3148023225182102e-05,
"loss": 0.0865,
"step": 262
},
{
"epoch": 1.0649087221095335,
"grad_norm": 0.4310865407957085,
"learning_rate": 2.309166135864198e-05,
"loss": 0.0657,
"step": 263
},
{
"epoch": 1.0689655172413792,
"grad_norm": 0.46794467781250354,
"learning_rate": 2.303513789778923e-05,
"loss": 0.0662,
"step": 264
},
{
"epoch": 1.0730223123732252,
"grad_norm": 0.4228722860949894,
"learning_rate": 2.297845397142415e-05,
"loss": 0.0644,
"step": 265
},
{
"epoch": 1.077079107505071,
"grad_norm": 0.5300905761078608,
"learning_rate": 2.2921610711551618e-05,
"loss": 0.0805,
"step": 266
},
{
"epoch": 1.081135902636917,
"grad_norm": 0.42302112184214735,
"learning_rate": 2.2864609253358477e-05,
"loss": 0.0654,
"step": 267
},
{
"epoch": 1.0851926977687627,
"grad_norm": 0.4833732847143345,
"learning_rate": 2.2807450735190864e-05,
"loss": 0.0807,
"step": 268
},
{
"epoch": 1.0892494929006085,
"grad_norm": 0.546877371515379,
"learning_rate": 2.2750136298531486e-05,
"loss": 0.0683,
"step": 269
},
{
"epoch": 1.0933062880324544,
"grad_norm": 0.4255827063856378,
"learning_rate": 2.2692667087976803e-05,
"loss": 0.0694,
"step": 270
},
{
"epoch": 1.0973630831643002,
"grad_norm": 0.5310767204910947,
"learning_rate": 2.263504425121419e-05,
"loss": 0.0882,
"step": 271
},
{
"epoch": 1.101419878296146,
"grad_norm": 0.4741282639389029,
"learning_rate": 2.2577268938999008e-05,
"loss": 0.0709,
"step": 272
},
{
"epoch": 1.105476673427992,
"grad_norm": 0.46558806246697076,
"learning_rate": 2.2519342305131627e-05,
"loss": 0.0743,
"step": 273
},
{
"epoch": 1.1095334685598377,
"grad_norm": 0.4693076664752292,
"learning_rate": 2.246126550643438e-05,
"loss": 0.083,
"step": 274
},
{
"epoch": 1.1135902636916835,
"grad_norm": 0.46565319997402443,
"learning_rate": 2.2403039702728466e-05,
"loss": 0.0651,
"step": 275
},
{
"epoch": 1.1176470588235294,
"grad_norm": 0.4222193938799318,
"learning_rate": 2.2344666056810788e-05,
"loss": 0.0623,
"step": 276
},
{
"epoch": 1.1217038539553752,
"grad_norm": 0.43899014503310846,
"learning_rate": 2.2286145734430715e-05,
"loss": 0.0655,
"step": 277
},
{
"epoch": 1.1257606490872212,
"grad_norm": 0.7568799917810151,
"learning_rate": 2.2227479904266827e-05,
"loss": 0.0703,
"step": 278
},
{
"epoch": 1.129817444219067,
"grad_norm": 0.4689401911791941,
"learning_rate": 2.2168669737903555e-05,
"loss": 0.0787,
"step": 279
},
{
"epoch": 1.1338742393509127,
"grad_norm": 0.4628308683329603,
"learning_rate": 2.2109716409807804e-05,
"loss": 0.0648,
"step": 280
},
{
"epoch": 1.1379310344827587,
"grad_norm": 0.43149649261110756,
"learning_rate": 2.205062109730547e-05,
"loss": 0.0575,
"step": 281
},
{
"epoch": 1.1419878296146044,
"grad_norm": 0.44715734359024867,
"learning_rate": 2.1991384980557958e-05,
"loss": 0.0714,
"step": 282
},
{
"epoch": 1.1460446247464504,
"grad_norm": 0.45975170848693486,
"learning_rate": 2.1932009242538595e-05,
"loss": 0.0782,
"step": 283
},
{
"epoch": 1.1501014198782962,
"grad_norm": 0.44707442495336097,
"learning_rate": 2.187249506900902e-05,
"loss": 0.0663,
"step": 284
},
{
"epoch": 1.154158215010142,
"grad_norm": 0.4143954800504552,
"learning_rate": 2.1812843648495465e-05,
"loss": 0.0593,
"step": 285
},
{
"epoch": 1.158215010141988,
"grad_norm": 0.43790846727578037,
"learning_rate": 2.17530561722651e-05,
"loss": 0.0706,
"step": 286
},
{
"epoch": 1.1622718052738337,
"grad_norm": 0.4592105560143304,
"learning_rate": 2.1693133834302148e-05,
"loss": 0.0697,
"step": 287
},
{
"epoch": 1.1663286004056794,
"grad_norm": 0.40947554363650396,
"learning_rate": 2.16330778312841e-05,
"loss": 0.0631,
"step": 288
},
{
"epoch": 1.1703853955375254,
"grad_norm": 0.4415229839300465,
"learning_rate": 2.15728893625578e-05,
"loss": 0.0726,
"step": 289
},
{
"epoch": 1.1744421906693712,
"grad_norm": 0.46497173706211997,
"learning_rate": 2.1512569630115507e-05,
"loss": 0.0742,
"step": 290
},
{
"epoch": 1.178498985801217,
"grad_norm": 0.410904105419549,
"learning_rate": 2.1452119838570857e-05,
"loss": 0.0672,
"step": 291
},
{
"epoch": 1.182555780933063,
"grad_norm": 0.415458681872725,
"learning_rate": 2.1391541195134845e-05,
"loss": 0.0725,
"step": 292
},
{
"epoch": 1.1866125760649087,
"grad_norm": 0.4062317481190398,
"learning_rate": 2.1330834909591696e-05,
"loss": 0.0646,
"step": 293
},
{
"epoch": 1.1906693711967546,
"grad_norm": 0.43806610932754925,
"learning_rate": 2.1270002194274698e-05,
"loss": 0.0723,
"step": 294
},
{
"epoch": 1.1947261663286004,
"grad_norm": 0.43543768126538734,
"learning_rate": 2.1209044264042016e-05,
"loss": 0.0653,
"step": 295
},
{
"epoch": 1.1987829614604462,
"grad_norm": 0.4051994881674093,
"learning_rate": 2.114796233625241e-05,
"loss": 0.0575,
"step": 296
},
{
"epoch": 1.2028397565922921,
"grad_norm": 0.4819124551084273,
"learning_rate": 2.1086757630740924e-05,
"loss": 0.0613,
"step": 297
},
{
"epoch": 1.206896551724138,
"grad_norm": 1.0299389777613355,
"learning_rate": 2.1025431369794546e-05,
"loss": 0.0703,
"step": 298
},
{
"epoch": 1.2109533468559839,
"grad_norm": 0.3820733350489595,
"learning_rate": 2.0963984778127763e-05,
"loss": 0.0569,
"step": 299
},
{
"epoch": 1.2150101419878296,
"grad_norm": 0.4252407347521453,
"learning_rate": 2.0902419082858142e-05,
"loss": 0.055,
"step": 300
},
{
"epoch": 1.2190669371196754,
"grad_norm": 0.3963130915735559,
"learning_rate": 2.08407355134818e-05,
"loss": 0.0605,
"step": 301
},
{
"epoch": 1.2231237322515214,
"grad_norm": 0.3740655420714341,
"learning_rate": 2.0778935301848852e-05,
"loss": 0.0591,
"step": 302
},
{
"epoch": 1.2271805273833671,
"grad_norm": 0.3633249347339507,
"learning_rate": 2.0717019682138814e-05,
"loss": 0.0621,
"step": 303
},
{
"epoch": 1.231237322515213,
"grad_norm": 0.44545868345146,
"learning_rate": 2.065498989083596e-05,
"loss": 0.0631,
"step": 304
},
{
"epoch": 1.2352941176470589,
"grad_norm": 0.715477713356301,
"learning_rate": 2.059284716670463e-05,
"loss": 0.0694,
"step": 305
},
{
"epoch": 1.2393509127789046,
"grad_norm": 0.3819969581826308,
"learning_rate": 2.053059275076447e-05,
"loss": 0.0668,
"step": 306
},
{
"epoch": 1.2434077079107504,
"grad_norm": 0.3797185410468661,
"learning_rate": 2.0468227886265683e-05,
"loss": 0.0661,
"step": 307
},
{
"epoch": 1.2474645030425964,
"grad_norm": 0.465835802610691,
"learning_rate": 2.0405753818664162e-05,
"loss": 0.063,
"step": 308
},
{
"epoch": 1.2515212981744421,
"grad_norm": 0.3682170190485834,
"learning_rate": 2.0343171795596654e-05,
"loss": 0.058,
"step": 309
},
{
"epoch": 1.2555780933062881,
"grad_norm": 0.3659098877003393,
"learning_rate": 2.0280483066855823e-05,
"loss": 0.0575,
"step": 310
},
{
"epoch": 1.2596348884381339,
"grad_norm": 0.5023156710027872,
"learning_rate": 2.0217688884365293e-05,
"loss": 0.0783,
"step": 311
},
{
"epoch": 1.2636916835699799,
"grad_norm": 0.5216416280425438,
"learning_rate": 2.0154790502154655e-05,
"loss": 0.0613,
"step": 312
},
{
"epoch": 1.2677484787018256,
"grad_norm": 0.4735363083812559,
"learning_rate": 2.0091789176334414e-05,
"loss": 0.079,
"step": 313
},
{
"epoch": 1.2718052738336714,
"grad_norm": 0.4652822824613967,
"learning_rate": 2.00286861650709e-05,
"loss": 0.0767,
"step": 314
},
{
"epoch": 1.2758620689655173,
"grad_norm": 0.4190192878483682,
"learning_rate": 1.996548272856117e-05,
"loss": 0.0585,
"step": 315
},
{
"epoch": 1.279918864097363,
"grad_norm": 0.3693884498803365,
"learning_rate": 1.9902180129007806e-05,
"loss": 0.0532,
"step": 316
},
{
"epoch": 1.2839756592292089,
"grad_norm": 0.3852707702368636,
"learning_rate": 1.983877963059372e-05,
"loss": 0.0584,
"step": 317
},
{
"epoch": 1.2880324543610548,
"grad_norm": 0.37753306339799014,
"learning_rate": 1.9775282499456915e-05,
"loss": 0.0582,
"step": 318
},
{
"epoch": 1.2920892494929006,
"grad_norm": 0.3758348932846382,
"learning_rate": 1.9711690003665202e-05,
"loss": 0.0586,
"step": 319
},
{
"epoch": 1.2961460446247464,
"grad_norm": 0.4194871916934118,
"learning_rate": 1.9648003413190858e-05,
"loss": 0.0469,
"step": 320
},
{
"epoch": 1.3002028397565923,
"grad_norm": 0.46048009587877814,
"learning_rate": 1.9584223999885288e-05,
"loss": 0.0789,
"step": 321
},
{
"epoch": 1.304259634888438,
"grad_norm": 0.39213100039505067,
"learning_rate": 1.95203530374536e-05,
"loss": 0.0608,
"step": 322
},
{
"epoch": 1.3083164300202839,
"grad_norm": 0.3993708308666585,
"learning_rate": 1.945639180142919e-05,
"loss": 0.0605,
"step": 323
},
{
"epoch": 1.3123732251521298,
"grad_norm": 0.37984652567355076,
"learning_rate": 1.9392341569148254e-05,
"loss": 0.0626,
"step": 324
},
{
"epoch": 1.3164300202839756,
"grad_norm": 0.4280714515322399,
"learning_rate": 1.9328203619724297e-05,
"loss": 0.0485,
"step": 325
},
{
"epoch": 1.3204868154158216,
"grad_norm": 0.3917723604992393,
"learning_rate": 1.9263979234022565e-05,
"loss": 0.0518,
"step": 326
},
{
"epoch": 1.3245436105476673,
"grad_norm": 0.37974066794354555,
"learning_rate": 1.919966969463448e-05,
"loss": 0.0582,
"step": 327
},
{
"epoch": 1.3286004056795133,
"grad_norm": 0.3687296850227283,
"learning_rate": 1.9135276285852028e-05,
"loss": 0.0489,
"step": 328
},
{
"epoch": 1.332657200811359,
"grad_norm": 0.3882503420638247,
"learning_rate": 1.90708002936421e-05,
"loss": 0.0714,
"step": 329
},
{
"epoch": 1.3367139959432048,
"grad_norm": 0.3969234950017993,
"learning_rate": 1.9006243005620825e-05,
"loss": 0.0608,
"step": 330
},
{
"epoch": 1.3407707910750508,
"grad_norm": 0.6237926409920694,
"learning_rate": 1.894160571102784e-05,
"loss": 0.057,
"step": 331
},
{
"epoch": 1.3448275862068966,
"grad_norm": 0.3624371856816983,
"learning_rate": 1.887688970070056e-05,
"loss": 0.0593,
"step": 332
},
{
"epoch": 1.3488843813387423,
"grad_norm": 0.43439892379173217,
"learning_rate": 1.8812096267048375e-05,
"loss": 0.0518,
"step": 333
},
{
"epoch": 1.3529411764705883,
"grad_norm": 0.38596267591475464,
"learning_rate": 1.874722670402688e-05,
"loss": 0.0621,
"step": 334
},
{
"epoch": 1.356997971602434,
"grad_norm": 0.4649551687377741,
"learning_rate": 1.8682282307111988e-05,
"loss": 0.0599,
"step": 335
},
{
"epoch": 1.3610547667342798,
"grad_norm": 0.4395548596766693,
"learning_rate": 1.861726437327409e-05,
"loss": 0.0693,
"step": 336
},
{
"epoch": 1.3651115618661258,
"grad_norm": 0.4862258593203981,
"learning_rate": 1.8552174200952137e-05,
"loss": 0.0612,
"step": 337
},
{
"epoch": 1.3691683569979716,
"grad_norm": 0.44291191177485717,
"learning_rate": 1.8487013090027727e-05,
"loss": 0.0744,
"step": 338
},
{
"epoch": 1.3732251521298173,
"grad_norm": 0.38563581033707284,
"learning_rate": 1.8421782341799127e-05,
"loss": 0.0565,
"step": 339
},
{
"epoch": 1.3772819472616633,
"grad_norm": 0.41423708346018334,
"learning_rate": 1.8356483258955302e-05,
"loss": 0.0751,
"step": 340
},
{
"epoch": 1.381338742393509,
"grad_norm": 0.35924533141908455,
"learning_rate": 1.829111714554988e-05,
"loss": 0.0587,
"step": 341
},
{
"epoch": 1.385395537525355,
"grad_norm": 0.4220818939935522,
"learning_rate": 1.8225685306975135e-05,
"loss": 0.0544,
"step": 342
},
{
"epoch": 1.3894523326572008,
"grad_norm": 0.42380504949224374,
"learning_rate": 1.8160189049935895e-05,
"loss": 0.0675,
"step": 343
},
{
"epoch": 1.3935091277890468,
"grad_norm": 0.42903121805073235,
"learning_rate": 1.8094629682423455e-05,
"loss": 0.0553,
"step": 344
},
{
"epoch": 1.3975659229208925,
"grad_norm": 0.3212183188904997,
"learning_rate": 1.802900851368946e-05,
"loss": 0.0454,
"step": 345
},
{
"epoch": 1.4016227180527383,
"grad_norm": 0.36796684826856213,
"learning_rate": 1.796332685421975e-05,
"loss": 0.0522,
"step": 346
},
{
"epoch": 1.4056795131845843,
"grad_norm": 0.3859937827585194,
"learning_rate": 1.78975860157082e-05,
"loss": 0.0502,
"step": 347
},
{
"epoch": 1.40973630831643,
"grad_norm": 0.48175546838276356,
"learning_rate": 1.7831787311030514e-05,
"loss": 0.0672,
"step": 348
},
{
"epoch": 1.4137931034482758,
"grad_norm": 0.35741222694673913,
"learning_rate": 1.7765932054218013e-05,
"loss": 0.0587,
"step": 349
},
{
"epoch": 1.4178498985801218,
"grad_norm": 0.33476461422223275,
"learning_rate": 1.770002156043138e-05,
"loss": 0.0524,
"step": 350
},
{
"epoch": 1.4219066937119675,
"grad_norm": 0.4375594713384405,
"learning_rate": 1.7634057145934436e-05,
"loss": 0.0734,
"step": 351
},
{
"epoch": 1.4259634888438133,
"grad_norm": 0.6125411034762981,
"learning_rate": 1.756804012806778e-05,
"loss": 0.054,
"step": 352
},
{
"epoch": 1.4300202839756593,
"grad_norm": 0.3969621592412149,
"learning_rate": 1.7501971825222574e-05,
"loss": 0.0607,
"step": 353
},
{
"epoch": 1.434077079107505,
"grad_norm": 0.40912007576868703,
"learning_rate": 1.7435853556814128e-05,
"loss": 0.0508,
"step": 354
},
{
"epoch": 1.4381338742393508,
"grad_norm": 0.4028347679550677,
"learning_rate": 1.73696866432556e-05,
"loss": 0.0625,
"step": 355
},
{
"epoch": 1.4421906693711968,
"grad_norm": 0.35121176777719804,
"learning_rate": 1.7303472405931614e-05,
"loss": 0.0423,
"step": 356
},
{
"epoch": 1.4462474645030425,
"grad_norm": 0.3734997556745032,
"learning_rate": 1.723721216717188e-05,
"loss": 0.0508,
"step": 357
},
{
"epoch": 1.4503042596348885,
"grad_norm": 0.39957427660410166,
"learning_rate": 1.7170907250224772e-05,
"loss": 0.0611,
"step": 358
},
{
"epoch": 1.4543610547667343,
"grad_norm": 0.4061196583559151,
"learning_rate": 1.7104558979230908e-05,
"loss": 0.0653,
"step": 359
},
{
"epoch": 1.4584178498985803,
"grad_norm": 0.42066306297323564,
"learning_rate": 1.70381686791967e-05,
"loss": 0.0544,
"step": 360
},
{
"epoch": 1.462474645030426,
"grad_norm": 0.3414243495462988,
"learning_rate": 1.697173767596792e-05,
"loss": 0.0402,
"step": 361
},
{
"epoch": 1.4665314401622718,
"grad_norm": 0.3745448318503202,
"learning_rate": 1.6905267296203182e-05,
"loss": 0.055,
"step": 362
},
{
"epoch": 1.4705882352941178,
"grad_norm": 0.3979687411753602,
"learning_rate": 1.6838758867347478e-05,
"loss": 0.0587,
"step": 363
},
{
"epoch": 1.4746450304259635,
"grad_norm": 0.4280060924115683,
"learning_rate": 1.677221371760565e-05,
"loss": 0.0546,
"step": 364
},
{
"epoch": 1.4787018255578093,
"grad_norm": 0.38953339017944766,
"learning_rate": 1.670563317591589e-05,
"loss": 0.0532,
"step": 365
},
{
"epoch": 1.4827586206896552,
"grad_norm": 0.33687703339250086,
"learning_rate": 1.6639018571923165e-05,
"loss": 0.0427,
"step": 366
},
{
"epoch": 1.486815415821501,
"grad_norm": 0.36071368662447184,
"learning_rate": 1.6572371235952698e-05,
"loss": 0.049,
"step": 367
},
{
"epoch": 1.4908722109533468,
"grad_norm": 0.4388319382689881,
"learning_rate": 1.6505692498983367e-05,
"loss": 0.0624,
"step": 368
},
{
"epoch": 1.4949290060851927,
"grad_norm": 0.4240499829869781,
"learning_rate": 1.643898369262117e-05,
"loss": 0.0664,
"step": 369
},
{
"epoch": 1.4989858012170385,
"grad_norm": 0.39531241733876216,
"learning_rate": 1.6372246149072572e-05,
"loss": 0.0647,
"step": 370
},
{
"epoch": 1.5030425963488843,
"grad_norm": 0.3866849983999559,
"learning_rate": 1.6305481201117958e-05,
"loss": 0.0519,
"step": 371
},
{
"epoch": 1.5070993914807302,
"grad_norm": 0.3932432735778696,
"learning_rate": 1.623869018208499e-05,
"loss": 0.0574,
"step": 372
},
{
"epoch": 1.5111561866125762,
"grad_norm": 0.3648065777660632,
"learning_rate": 1.617187442582196e-05,
"loss": 0.0536,
"step": 373
},
{
"epoch": 1.5152129817444218,
"grad_norm": 0.3191973621717305,
"learning_rate": 1.61050352666712e-05,
"loss": 0.0357,
"step": 374
},
{
"epoch": 1.5192697768762677,
"grad_norm": 0.36611928703171565,
"learning_rate": 1.6038174039442408e-05,
"loss": 0.049,
"step": 375
},
{
"epoch": 1.5233265720081137,
"grad_norm": 0.3382368716620762,
"learning_rate": 1.5971292079385963e-05,
"loss": 0.0491,
"step": 376
},
{
"epoch": 1.5273833671399595,
"grad_norm": 0.4357138284943321,
"learning_rate": 1.590439072216632e-05,
"loss": 0.0534,
"step": 377
},
{
"epoch": 1.5314401622718052,
"grad_norm": 0.4239454871704157,
"learning_rate": 1.58374713038353e-05,
"loss": 0.0574,
"step": 378
},
{
"epoch": 1.5354969574036512,
"grad_norm": 0.4366363374367421,
"learning_rate": 1.5770535160805383e-05,
"loss": 0.0578,
"step": 379
},
{
"epoch": 1.539553752535497,
"grad_norm": 0.6084387765008901,
"learning_rate": 1.570358362982309e-05,
"loss": 0.0425,
"step": 380
},
{
"epoch": 1.5436105476673427,
"grad_norm": 0.33997122436540295,
"learning_rate": 1.5636618047942225e-05,
"loss": 0.0467,
"step": 381
},
{
"epoch": 1.5476673427991887,
"grad_norm": 0.35089831421268103,
"learning_rate": 1.55696397524972e-05,
"loss": 0.055,
"step": 382
},
{
"epoch": 1.5517241379310345,
"grad_norm": 0.3798972719987602,
"learning_rate": 1.5502650081076316e-05,
"loss": 0.0563,
"step": 383
},
{
"epoch": 1.5557809330628802,
"grad_norm": 0.32510731462108816,
"learning_rate": 1.5435650371495066e-05,
"loss": 0.0467,
"step": 384
},
{
"epoch": 1.5598377281947262,
"grad_norm": 0.33402496992983266,
"learning_rate": 1.5368641961769404e-05,
"loss": 0.0484,
"step": 385
},
{
"epoch": 1.563894523326572,
"grad_norm": 0.36474864088922704,
"learning_rate": 1.5301626190089037e-05,
"loss": 0.0465,
"step": 386
},
{
"epoch": 1.5679513184584177,
"grad_norm": 0.36295702849052125,
"learning_rate": 1.523460439479069e-05,
"loss": 0.0492,
"step": 387
},
{
"epoch": 1.5720081135902637,
"grad_norm": 0.39047953624252674,
"learning_rate": 1.5167577914331383e-05,
"loss": 0.0528,
"step": 388
},
{
"epoch": 1.5760649087221097,
"grad_norm": 0.37194589071658085,
"learning_rate": 1.5100548087261692e-05,
"loss": 0.0637,
"step": 389
},
{
"epoch": 1.5801217038539552,
"grad_norm": 0.3810548350728733,
"learning_rate": 1.5033516252199047e-05,
"loss": 0.0555,
"step": 390
},
{
"epoch": 1.5841784989858012,
"grad_norm": 0.40043981849846777,
"learning_rate": 1.4966483747800957e-05,
"loss": 0.0561,
"step": 391
},
{
"epoch": 1.5882352941176472,
"grad_norm": 0.5220754123514092,
"learning_rate": 1.4899451912738307e-05,
"loss": 0.0612,
"step": 392
},
{
"epoch": 1.592292089249493,
"grad_norm": 0.43468404049599646,
"learning_rate": 1.483242208566862e-05,
"loss": 0.0612,
"step": 393
},
{
"epoch": 1.5963488843813387,
"grad_norm": 0.3042175270465621,
"learning_rate": 1.4765395605209309e-05,
"loss": 0.0429,
"step": 394
},
{
"epoch": 1.6004056795131847,
"grad_norm": 0.3721980766639612,
"learning_rate": 1.4698373809910964e-05,
"loss": 0.053,
"step": 395
},
{
"epoch": 1.6044624746450304,
"grad_norm": 0.3792114270784363,
"learning_rate": 1.4631358038230602e-05,
"loss": 0.0608,
"step": 396
},
{
"epoch": 1.6085192697768762,
"grad_norm": 0.558961664405718,
"learning_rate": 1.4564349628504938e-05,
"loss": 0.057,
"step": 397
},
{
"epoch": 1.6125760649087222,
"grad_norm": 0.38762889149568547,
"learning_rate": 1.4497349918923693e-05,
"loss": 0.0483,
"step": 398
},
{
"epoch": 1.616632860040568,
"grad_norm": 0.41842317615428753,
"learning_rate": 1.4430360247502802e-05,
"loss": 0.0549,
"step": 399
},
{
"epoch": 1.6206896551724137,
"grad_norm": 0.3356947240634165,
"learning_rate": 1.4363381952057779e-05,
"loss": 0.0444,
"step": 400
},
{
"epoch": 1.6247464503042597,
"grad_norm": 0.3385969976968215,
"learning_rate": 1.4296416370176911e-05,
"loss": 0.0506,
"step": 401
},
{
"epoch": 1.6288032454361054,
"grad_norm": 0.34833330430761994,
"learning_rate": 1.422946483919462e-05,
"loss": 0.0433,
"step": 402
},
{
"epoch": 1.6328600405679512,
"grad_norm": 0.3547317090371702,
"learning_rate": 1.4162528696164708e-05,
"loss": 0.0383,
"step": 403
},
{
"epoch": 1.6369168356997972,
"grad_norm": 0.3572464802981435,
"learning_rate": 1.409560927783368e-05,
"loss": 0.0471,
"step": 404
},
{
"epoch": 1.6409736308316432,
"grad_norm": 0.4078511085972787,
"learning_rate": 1.402870792061404e-05,
"loss": 0.0479,
"step": 405
},
{
"epoch": 1.6450304259634887,
"grad_norm": 0.45053367326742416,
"learning_rate": 1.3961825960557598e-05,
"loss": 0.0488,
"step": 406
},
{
"epoch": 1.6490872210953347,
"grad_norm": 0.391437337418899,
"learning_rate": 1.38949647333288e-05,
"loss": 0.0514,
"step": 407
},
{
"epoch": 1.6531440162271807,
"grad_norm": 0.408732102689673,
"learning_rate": 1.382812557417804e-05,
"loss": 0.0431,
"step": 408
},
{
"epoch": 1.6572008113590264,
"grad_norm": 0.4038960351294187,
"learning_rate": 1.3761309817915017e-05,
"loss": 0.0503,
"step": 409
},
{
"epoch": 1.6612576064908722,
"grad_norm": 0.34726808950461435,
"learning_rate": 1.3694518798882041e-05,
"loss": 0.043,
"step": 410
},
{
"epoch": 1.6653144016227182,
"grad_norm": 0.3569286082484257,
"learning_rate": 1.3627753850927434e-05,
"loss": 0.0514,
"step": 411
},
{
"epoch": 1.669371196754564,
"grad_norm": 2.5156687118725736,
"learning_rate": 1.356101630737883e-05,
"loss": 0.0466,
"step": 412
},
{
"epoch": 1.6734279918864097,
"grad_norm": 0.3305227447341641,
"learning_rate": 1.3494307501016634e-05,
"loss": 0.0435,
"step": 413
},
{
"epoch": 1.6774847870182557,
"grad_norm": 0.3368406729262284,
"learning_rate": 1.3427628764047308e-05,
"loss": 0.0412,
"step": 414
},
{
"epoch": 1.6815415821501014,
"grad_norm": 0.36346054994322113,
"learning_rate": 1.3360981428076837e-05,
"loss": 0.0419,
"step": 415
},
{
"epoch": 1.6855983772819472,
"grad_norm": 0.35727350467498614,
"learning_rate": 1.3294366824084114e-05,
"loss": 0.0579,
"step": 416
},
{
"epoch": 1.6896551724137931,
"grad_norm": 0.2969626674288336,
"learning_rate": 1.322778628239435e-05,
"loss": 0.0382,
"step": 417
},
{
"epoch": 1.693711967545639,
"grad_norm": 0.4115522039939296,
"learning_rate": 1.3161241132652526e-05,
"loss": 0.0524,
"step": 418
},
{
"epoch": 1.6977687626774847,
"grad_norm": 0.3226088009987354,
"learning_rate": 1.3094732703796819e-05,
"loss": 0.0435,
"step": 419
},
{
"epoch": 1.7018255578093306,
"grad_norm": 0.4166243770414429,
"learning_rate": 1.3028262324032083e-05,
"loss": 0.0529,
"step": 420
},
{
"epoch": 1.7058823529411766,
"grad_norm": 0.5768808567980608,
"learning_rate": 1.2961831320803298e-05,
"loss": 0.0507,
"step": 421
},
{
"epoch": 1.7099391480730222,
"grad_norm": 0.3115823293971135,
"learning_rate": 1.2895441020769096e-05,
"loss": 0.0356,
"step": 422
},
{
"epoch": 1.7139959432048681,
"grad_norm": 0.3525413888420736,
"learning_rate": 1.2829092749775229e-05,
"loss": 0.0438,
"step": 423
},
{
"epoch": 1.7180527383367141,
"grad_norm": 0.5771695307719267,
"learning_rate": 1.276278783282812e-05,
"loss": 0.044,
"step": 424
},
{
"epoch": 1.7221095334685599,
"grad_norm": 0.32856795722956933,
"learning_rate": 1.269652759406839e-05,
"loss": 0.0376,
"step": 425
},
{
"epoch": 1.7261663286004056,
"grad_norm": 0.40237568439445387,
"learning_rate": 1.2630313356744404e-05,
"loss": 0.0535,
"step": 426
},
{
"epoch": 1.7302231237322516,
"grad_norm": 0.3457471850990833,
"learning_rate": 1.256414644318588e-05,
"loss": 0.0455,
"step": 427
},
{
"epoch": 1.7342799188640974,
"grad_norm": 0.2958982436386828,
"learning_rate": 1.2498028174777428e-05,
"loss": 0.0324,
"step": 428
},
{
"epoch": 1.7383367139959431,
"grad_norm": 0.3636255930835156,
"learning_rate": 1.243195987193222e-05,
"loss": 0.0445,
"step": 429
},
{
"epoch": 1.7423935091277891,
"grad_norm": 0.38524609276341326,
"learning_rate": 1.236594285406557e-05,
"loss": 0.041,
"step": 430
},
{
"epoch": 1.7464503042596349,
"grad_norm": 0.5336365912189125,
"learning_rate": 1.229997843956862e-05,
"loss": 0.0395,
"step": 431
},
{
"epoch": 1.7505070993914806,
"grad_norm": 0.34460744048253517,
"learning_rate": 1.2234067945781991e-05,
"loss": 0.0344,
"step": 432
},
{
"epoch": 1.7545638945233266,
"grad_norm": 0.40294802503464133,
"learning_rate": 1.2168212688969489e-05,
"loss": 0.0439,
"step": 433
},
{
"epoch": 1.7586206896551724,
"grad_norm": 0.45454193803533016,
"learning_rate": 1.2102413984291804e-05,
"loss": 0.0491,
"step": 434
},
{
"epoch": 1.7626774847870181,
"grad_norm": 0.3317346567653446,
"learning_rate": 1.2036673145780251e-05,
"loss": 0.0368,
"step": 435
},
{
"epoch": 1.7667342799188641,
"grad_norm": 0.4450615911596974,
"learning_rate": 1.1970991486310543e-05,
"loss": 0.0461,
"step": 436
},
{
"epoch": 1.77079107505071,
"grad_norm": 0.3566006140898405,
"learning_rate": 1.1905370317576544e-05,
"loss": 0.0361,
"step": 437
},
{
"epoch": 1.7748478701825556,
"grad_norm": 0.39635531600963336,
"learning_rate": 1.183981095006411e-05,
"loss": 0.0469,
"step": 438
},
{
"epoch": 1.7789046653144016,
"grad_norm": 0.39314677552680516,
"learning_rate": 1.1774314693024863e-05,
"loss": 0.0412,
"step": 439
},
{
"epoch": 1.7829614604462476,
"grad_norm": 0.3404742231149006,
"learning_rate": 1.1708882854450123e-05,
"loss": 0.0401,
"step": 440
},
{
"epoch": 1.7870182555780934,
"grad_norm": 0.4018473444675262,
"learning_rate": 1.1643516741044697e-05,
"loss": 0.0568,
"step": 441
},
{
"epoch": 1.791075050709939,
"grad_norm": 0.42524465351455154,
"learning_rate": 1.1578217658200875e-05,
"loss": 0.0515,
"step": 442
},
{
"epoch": 1.795131845841785,
"grad_norm": 0.38868002112425293,
"learning_rate": 1.1512986909972279e-05,
"loss": 0.0514,
"step": 443
},
{
"epoch": 1.7991886409736308,
"grad_norm": 0.35122549166081785,
"learning_rate": 1.1447825799047868e-05,
"loss": 0.0483,
"step": 444
},
{
"epoch": 1.8032454361054766,
"grad_norm": 0.3439279884475484,
"learning_rate": 1.1382735626725916e-05,
"loss": 0.0449,
"step": 445
},
{
"epoch": 1.8073022312373226,
"grad_norm": 0.317289330890349,
"learning_rate": 1.1317717692888014e-05,
"loss": 0.0389,
"step": 446
},
{
"epoch": 1.8113590263691683,
"grad_norm": 0.30112048560120525,
"learning_rate": 1.1252773295973122e-05,
"loss": 0.038,
"step": 447
},
{
"epoch": 1.815415821501014,
"grad_norm": 0.3519313377042473,
"learning_rate": 1.1187903732951624e-05,
"loss": 0.0387,
"step": 448
},
{
"epoch": 1.81947261663286,
"grad_norm": 0.2964711727898209,
"learning_rate": 1.1123110299299445e-05,
"loss": 0.0345,
"step": 449
},
{
"epoch": 1.8235294117647058,
"grad_norm": 0.32468825235424914,
"learning_rate": 1.105839428897216e-05,
"loss": 0.0341,
"step": 450
},
{
"epoch": 1.8275862068965516,
"grad_norm": 0.337431774628168,
"learning_rate": 1.0993756994379178e-05,
"loss": 0.0453,
"step": 451
},
{
"epoch": 1.8316430020283976,
"grad_norm": 0.36965782178837847,
"learning_rate": 1.0929199706357906e-05,
"loss": 0.0419,
"step": 452
},
{
"epoch": 1.8356997971602436,
"grad_norm": 0.33955413410890395,
"learning_rate": 1.0864723714147975e-05,
"loss": 0.0371,
"step": 453
},
{
"epoch": 1.839756592292089,
"grad_norm": 0.37702859177437276,
"learning_rate": 1.0800330305365526e-05,
"loss": 0.0447,
"step": 454
},
{
"epoch": 1.843813387423935,
"grad_norm": 0.43380541102915904,
"learning_rate": 1.0736020765977434e-05,
"loss": 0.0525,
"step": 455
},
{
"epoch": 1.847870182555781,
"grad_norm": 0.3308232636624218,
"learning_rate": 1.0671796380275705e-05,
"loss": 0.0275,
"step": 456
},
{
"epoch": 1.8519269776876268,
"grad_norm": 0.30675776077260075,
"learning_rate": 1.0607658430851745e-05,
"loss": 0.0387,
"step": 457
},
{
"epoch": 1.8559837728194726,
"grad_norm": 0.3441545600958111,
"learning_rate": 1.0543608198570815e-05,
"loss": 0.0351,
"step": 458
},
{
"epoch": 1.8600405679513186,
"grad_norm": 0.3574675624451796,
"learning_rate": 1.0479646962546401e-05,
"loss": 0.0395,
"step": 459
},
{
"epoch": 1.8640973630831643,
"grad_norm": 0.41572429982747455,
"learning_rate": 1.0415776000114716e-05,
"loss": 0.0488,
"step": 460
},
{
"epoch": 1.86815415821501,
"grad_norm": 0.39369896529868187,
"learning_rate": 1.0351996586809145e-05,
"loss": 0.0418,
"step": 461
},
{
"epoch": 1.872210953346856,
"grad_norm": 0.34363426124475344,
"learning_rate": 1.0288309996334802e-05,
"loss": 0.0394,
"step": 462
},
{
"epoch": 1.8762677484787018,
"grad_norm": 0.327289268468903,
"learning_rate": 1.0224717500543087e-05,
"loss": 0.0407,
"step": 463
},
{
"epoch": 1.8803245436105476,
"grad_norm": 0.504229258734072,
"learning_rate": 1.0161220369406283e-05,
"loss": 0.0339,
"step": 464
},
{
"epoch": 1.8843813387423936,
"grad_norm": 0.32336927283271066,
"learning_rate": 1.0097819870992196e-05,
"loss": 0.0399,
"step": 465
},
{
"epoch": 1.8884381338742393,
"grad_norm": 0.30534018999460294,
"learning_rate": 1.0034517271438826e-05,
"loss": 0.0363,
"step": 466
},
{
"epoch": 1.892494929006085,
"grad_norm": 0.3363030481693883,
"learning_rate": 9.9713138349291e-06,
"loss": 0.0424,
"step": 467
},
{
"epoch": 1.896551724137931,
"grad_norm": 0.3940025228623559,
"learning_rate": 9.908210823665587e-06,
"loss": 0.0475,
"step": 468
},
{
"epoch": 1.900608519269777,
"grad_norm": 0.3465688297316223,
"learning_rate": 9.84520949784535e-06,
"loss": 0.0339,
"step": 469
},
{
"epoch": 1.9046653144016226,
"grad_norm": 0.3304003434324526,
"learning_rate": 9.782311115634706e-06,
"loss": 0.0431,
"step": 470
},
{
"epoch": 1.9087221095334685,
"grad_norm": 0.36499605332027524,
"learning_rate": 9.719516933144183e-06,
"loss": 0.0378,
"step": 471
},
{
"epoch": 1.9127789046653145,
"grad_norm": 0.40038829914160073,
"learning_rate": 9.656828204403352e-06,
"loss": 0.0326,
"step": 472
},
{
"epoch": 1.9168356997971603,
"grad_norm": 0.2998673014346592,
"learning_rate": 9.594246181335844e-06,
"loss": 0.0304,
"step": 473
},
{
"epoch": 1.920892494929006,
"grad_norm": 0.3233930312542325,
"learning_rate": 9.531772113734325e-06,
"loss": 0.0337,
"step": 474
},
{
"epoch": 1.924949290060852,
"grad_norm": 0.30650567980290994,
"learning_rate": 9.469407249235532e-06,
"loss": 0.0305,
"step": 475
},
{
"epoch": 1.9290060851926978,
"grad_norm": 0.4514180300508764,
"learning_rate": 9.407152833295373e-06,
"loss": 0.0337,
"step": 476
},
{
"epoch": 1.9330628803245435,
"grad_norm": 0.4313574733470663,
"learning_rate": 9.345010109164038e-06,
"loss": 0.047,
"step": 477
},
{
"epoch": 1.9371196754563895,
"grad_norm": 0.36322481811838864,
"learning_rate": 9.282980317861188e-06,
"loss": 0.0411,
"step": 478
},
{
"epoch": 1.9411764705882353,
"grad_norm": 0.31572498922793873,
"learning_rate": 9.22106469815115e-06,
"loss": 0.0302,
"step": 479
},
{
"epoch": 1.945233265720081,
"grad_norm": 0.3623773378997451,
"learning_rate": 9.1592644865182e-06,
"loss": 0.043,
"step": 480
},
{
"epoch": 1.949290060851927,
"grad_norm": 0.35603531168253105,
"learning_rate": 9.097580917141862e-06,
"loss": 0.0394,
"step": 481
},
{
"epoch": 1.9533468559837728,
"grad_norm": 0.3298984695320802,
"learning_rate": 9.03601522187224e-06,
"loss": 0.0382,
"step": 482
},
{
"epoch": 1.9574036511156185,
"grad_norm": 0.34985337949550965,
"learning_rate": 8.974568630205462e-06,
"loss": 0.0305,
"step": 483
},
{
"epoch": 1.9614604462474645,
"grad_norm": 0.37580676659704776,
"learning_rate": 8.913242369259076e-06,
"loss": 0.0369,
"step": 484
},
{
"epoch": 1.9655172413793105,
"grad_norm": 0.3578031464742577,
"learning_rate": 8.852037663747595e-06,
"loss": 0.041,
"step": 485
},
{
"epoch": 1.969574036511156,
"grad_norm": 0.3055635144638437,
"learning_rate": 8.790955735957984e-06,
"loss": 0.0298,
"step": 486
},
{
"epoch": 1.973630831643002,
"grad_norm": 0.4410780953526682,
"learning_rate": 8.729997805725305e-06,
"loss": 0.0342,
"step": 487
},
{
"epoch": 1.977687626774848,
"grad_norm": 0.36929913708419226,
"learning_rate": 8.669165090408303e-06,
"loss": 0.0364,
"step": 488
},
{
"epoch": 1.9817444219066938,
"grad_norm": 0.32405392390526616,
"learning_rate": 8.60845880486516e-06,
"loss": 0.0343,
"step": 489
},
{
"epoch": 1.9858012170385395,
"grad_norm": 0.38038844185366744,
"learning_rate": 8.54788016142915e-06,
"loss": 0.039,
"step": 490
},
{
"epoch": 1.9898580121703855,
"grad_norm": 0.3097549740680631,
"learning_rate": 8.487430369884497e-06,
"loss": 0.0265,
"step": 491
},
{
"epoch": 1.9939148073022313,
"grad_norm": 0.33837296723283344,
"learning_rate": 8.427110637442202e-06,
"loss": 0.0423,
"step": 492
},
{
"epoch": 1.997971602434077,
"grad_norm": 0.7548004728332394,
"learning_rate": 8.366922168715907e-06,
"loss": 0.0436,
"step": 493
},
{
"epoch": 2.0,
"grad_norm": 0.3588161803270943,
"learning_rate": 8.30686616569786e-06,
"loss": 0.0218,
"step": 494
},
{
"epoch": 2.004056795131846,
"grad_norm": 0.3166650742352506,
"learning_rate": 8.246943827734899e-06,
"loss": 0.0145,
"step": 495
},
{
"epoch": 2.0081135902636915,
"grad_norm": 0.29379020109122056,
"learning_rate": 8.18715635150453e-06,
"loss": 0.0122,
"step": 496
},
{
"epoch": 2.0121703853955375,
"grad_norm": 0.21926970559190412,
"learning_rate": 8.127504930990986e-06,
"loss": 0.0128,
"step": 497
},
{
"epoch": 2.0162271805273835,
"grad_norm": 0.228510859709003,
"learning_rate": 8.067990757461406e-06,
"loss": 0.0122,
"step": 498
},
{
"epoch": 2.020283975659229,
"grad_norm": 0.376799062256713,
"learning_rate": 8.008615019442046e-06,
"loss": 0.0114,
"step": 499
},
{
"epoch": 2.024340770791075,
"grad_norm": 0.24226554064647018,
"learning_rate": 7.94937890269453e-06,
"loss": 0.015,
"step": 500
},
{
"epoch": 2.028397565922921,
"grad_norm": 0.24487658318343053,
"learning_rate": 7.890283590192203e-06,
"loss": 0.0134,
"step": 501
},
{
"epoch": 2.032454361054767,
"grad_norm": 0.2831459815506862,
"learning_rate": 7.831330262096444e-06,
"loss": 0.0139,
"step": 502
},
{
"epoch": 2.0365111561866125,
"grad_norm": 0.24261217304937002,
"learning_rate": 7.772520095733177e-06,
"loss": 0.013,
"step": 503
},
{
"epoch": 2.0405679513184585,
"grad_norm": 0.25184431686328385,
"learning_rate": 7.713854265569283e-06,
"loss": 0.011,
"step": 504
},
{
"epoch": 2.0446247464503045,
"grad_norm": 0.30101462279040575,
"learning_rate": 7.655333943189218e-06,
"loss": 0.0135,
"step": 505
},
{
"epoch": 2.04868154158215,
"grad_norm": 0.2606959867535079,
"learning_rate": 7.596960297271531e-06,
"loss": 0.0103,
"step": 506
},
{
"epoch": 2.052738336713996,
"grad_norm": 0.24820046887000718,
"learning_rate": 7.538734493565621e-06,
"loss": 0.0123,
"step": 507
},
{
"epoch": 2.056795131845842,
"grad_norm": 0.2918905563444268,
"learning_rate": 7.480657694868378e-06,
"loss": 0.0109,
"step": 508
},
{
"epoch": 2.0608519269776875,
"grad_norm": 0.26284895212730003,
"learning_rate": 7.422731061000998e-06,
"loss": 0.0143,
"step": 509
},
{
"epoch": 2.0649087221095335,
"grad_norm": 0.2299243190180256,
"learning_rate": 7.364955748785817e-06,
"loss": 0.0105,
"step": 510
},
{
"epoch": 2.0689655172413794,
"grad_norm": 0.2780394702085014,
"learning_rate": 7.307332912023197e-06,
"loss": 0.0123,
"step": 511
},
{
"epoch": 2.073022312373225,
"grad_norm": 0.372944451393654,
"learning_rate": 7.2498637014685144e-06,
"loss": 0.0131,
"step": 512
},
{
"epoch": 2.077079107505071,
"grad_norm": 0.2485004068898688,
"learning_rate": 7.192549264809136e-06,
"loss": 0.0105,
"step": 513
},
{
"epoch": 2.081135902636917,
"grad_norm": 0.23050371072210507,
"learning_rate": 7.135390746641527e-06,
"loss": 0.0093,
"step": 514
},
{
"epoch": 2.0851926977687625,
"grad_norm": 0.2626322620347206,
"learning_rate": 7.078389288448384e-06,
"loss": 0.0158,
"step": 515
},
{
"epoch": 2.0892494929006085,
"grad_norm": 0.2491223084332659,
"learning_rate": 7.021546028575854e-06,
"loss": 0.0105,
"step": 516
},
{
"epoch": 2.0933062880324544,
"grad_norm": 0.2914558707105371,
"learning_rate": 6.964862102210773e-06,
"loss": 0.0109,
"step": 517
},
{
"epoch": 2.0973630831643,
"grad_norm": 0.2879693675651145,
"learning_rate": 6.908338641358021e-06,
"loss": 0.0131,
"step": 518
},
{
"epoch": 2.101419878296146,
"grad_norm": 0.2458165550394172,
"learning_rate": 6.8519767748179005e-06,
"loss": 0.0094,
"step": 519
},
{
"epoch": 2.105476673427992,
"grad_norm": 0.22406613482172252,
"learning_rate": 6.795777628163599e-06,
"loss": 0.01,
"step": 520
},
{
"epoch": 2.109533468559838,
"grad_norm": 0.2356332206172294,
"learning_rate": 6.739742323718716e-06,
"loss": 0.0088,
"step": 521
},
{
"epoch": 2.1135902636916835,
"grad_norm": 0.24668005158070347,
"learning_rate": 6.683871980534833e-06,
"loss": 0.01,
"step": 522
},
{
"epoch": 2.1176470588235294,
"grad_norm": 0.2518605659887813,
"learning_rate": 6.628167714369197e-06,
"loss": 0.0111,
"step": 523
},
{
"epoch": 2.1217038539553754,
"grad_norm": 0.2705553856669013,
"learning_rate": 6.5726306376624e-06,
"loss": 0.0122,
"step": 524
},
{
"epoch": 2.125760649087221,
"grad_norm": 0.24625652105303836,
"learning_rate": 6.517261859516188e-06,
"loss": 0.0119,
"step": 525
},
{
"epoch": 2.129817444219067,
"grad_norm": 0.2621098976362248,
"learning_rate": 6.462062485671299e-06,
"loss": 0.0141,
"step": 526
},
{
"epoch": 2.133874239350913,
"grad_norm": 0.22374065082170833,
"learning_rate": 6.407033618485401e-06,
"loss": 0.0107,
"step": 527
},
{
"epoch": 2.1379310344827585,
"grad_norm": 0.2630409589185763,
"learning_rate": 6.352176356911056e-06,
"loss": 0.012,
"step": 528
},
{
"epoch": 2.1419878296146044,
"grad_norm": 0.28222343712803627,
"learning_rate": 6.297491796473765e-06,
"loss": 0.0139,
"step": 529
},
{
"epoch": 2.1460446247464504,
"grad_norm": 0.317176893149287,
"learning_rate": 6.242981029250131e-06,
"loss": 0.011,
"step": 530
},
{
"epoch": 2.150101419878296,
"grad_norm": 0.19869118994266702,
"learning_rate": 6.188645143846005e-06,
"loss": 0.0077,
"step": 531
},
{
"epoch": 2.154158215010142,
"grad_norm": 0.319593585799946,
"learning_rate": 6.134485225374787e-06,
"loss": 0.01,
"step": 532
},
{
"epoch": 2.158215010141988,
"grad_norm": 0.2366052906120083,
"learning_rate": 6.080502355435702e-06,
"loss": 0.0116,
"step": 533
},
{
"epoch": 2.162271805273834,
"grad_norm": 0.1913014899550726,
"learning_rate": 6.026697612092259e-06,
"loss": 0.008,
"step": 534
},
{
"epoch": 2.1663286004056794,
"grad_norm": 0.23493412563901295,
"learning_rate": 5.973072069850686e-06,
"loss": 0.0096,
"step": 535
},
{
"epoch": 2.1703853955375254,
"grad_norm": 0.2609102661845724,
"learning_rate": 5.919626799638477e-06,
"loss": 0.0122,
"step": 536
},
{
"epoch": 2.1744421906693714,
"grad_norm": 0.28503067190982867,
"learning_rate": 5.866362868783008e-06,
"loss": 0.0112,
"step": 537
},
{
"epoch": 2.178498985801217,
"grad_norm": 0.2640652554619867,
"learning_rate": 5.8132813409902226e-06,
"loss": 0.0122,
"step": 538
},
{
"epoch": 2.182555780933063,
"grad_norm": 0.2391357216466815,
"learning_rate": 5.760383276323398e-06,
"loss": 0.0126,
"step": 539
},
{
"epoch": 2.186612576064909,
"grad_norm": 0.3103681231035885,
"learning_rate": 5.707669731181955e-06,
"loss": 0.0111,
"step": 540
},
{
"epoch": 2.1906693711967544,
"grad_norm": 0.21657582935164016,
"learning_rate": 5.655141758280374e-06,
"loss": 0.0101,
"step": 541
},
{
"epoch": 2.1947261663286004,
"grad_norm": 0.21854371695303926,
"learning_rate": 5.6028004066271686e-06,
"loss": 0.0105,
"step": 542
},
{
"epoch": 2.1987829614604464,
"grad_norm": 0.24975655642710679,
"learning_rate": 5.550646721503947e-06,
"loss": 0.0115,
"step": 543
},
{
"epoch": 2.202839756592292,
"grad_norm": 0.25634458122766723,
"learning_rate": 5.498681744444518e-06,
"loss": 0.0125,
"step": 544
},
{
"epoch": 2.206896551724138,
"grad_norm": 0.45332168206537427,
"learning_rate": 5.4469065132141025e-06,
"loss": 0.011,
"step": 545
},
{
"epoch": 2.210953346855984,
"grad_norm": 0.22621788005120613,
"learning_rate": 5.39532206178861e-06,
"loss": 0.011,
"step": 546
},
{
"epoch": 2.2150101419878294,
"grad_norm": 0.26240485222820514,
"learning_rate": 5.3439294203339844e-06,
"loss": 0.0134,
"step": 547
},
{
"epoch": 2.2190669371196754,
"grad_norm": 0.21029205760374123,
"learning_rate": 5.292729615185637e-06,
"loss": 0.0099,
"step": 548
},
{
"epoch": 2.2231237322515214,
"grad_norm": 0.2792807255391835,
"learning_rate": 5.241723668827937e-06,
"loss": 0.0129,
"step": 549
},
{
"epoch": 2.227180527383367,
"grad_norm": 0.2795152742456593,
"learning_rate": 5.190912599873818e-06,
"loss": 0.0126,
"step": 550
},
{
"epoch": 2.231237322515213,
"grad_norm": 0.24802505785116624,
"learning_rate": 5.14029742304441e-06,
"loss": 0.0108,
"step": 551
},
{
"epoch": 2.235294117647059,
"grad_norm": 0.2595006809811974,
"learning_rate": 5.089879149148781e-06,
"loss": 0.0117,
"step": 552
},
{
"epoch": 2.239350912778905,
"grad_norm": 0.25411628526886854,
"learning_rate": 5.0396587850637554e-06,
"loss": 0.0091,
"step": 553
},
{
"epoch": 2.2434077079107504,
"grad_norm": 0.2612915952356164,
"learning_rate": 4.989637333713814e-06,
"loss": 0.01,
"step": 554
},
{
"epoch": 2.2474645030425964,
"grad_norm": 0.26843182159434964,
"learning_rate": 4.93981579405105e-06,
"loss": 0.0111,
"step": 555
},
{
"epoch": 2.2515212981744424,
"grad_norm": 0.21565821566578344,
"learning_rate": 4.89019516103522e-06,
"loss": 0.0093,
"step": 556
},
{
"epoch": 2.255578093306288,
"grad_norm": 0.257015977515393,
"learning_rate": 4.840776425613887e-06,
"loss": 0.014,
"step": 557
},
{
"epoch": 2.259634888438134,
"grad_norm": 0.20799379451848776,
"learning_rate": 4.791560574702614e-06,
"loss": 0.0082,
"step": 558
},
{
"epoch": 2.26369168356998,
"grad_norm": 0.31771401066766947,
"learning_rate": 4.742548591165289e-06,
"loss": 0.0145,
"step": 559
},
{
"epoch": 2.2677484787018254,
"grad_norm": 0.2951694121169533,
"learning_rate": 4.693741453794433e-06,
"loss": 0.0133,
"step": 560
},
{
"epoch": 2.2718052738336714,
"grad_norm": 0.2628861243112249,
"learning_rate": 4.6451401372917275e-06,
"loss": 0.0099,
"step": 561
},
{
"epoch": 2.2758620689655173,
"grad_norm": 0.22269435944973373,
"learning_rate": 4.596745612248488e-06,
"loss": 0.01,
"step": 562
},
{
"epoch": 2.279918864097363,
"grad_norm": 0.28704577522446245,
"learning_rate": 4.548558845126334e-06,
"loss": 0.0133,
"step": 563
},
{
"epoch": 2.283975659229209,
"grad_norm": 0.26309253454993753,
"learning_rate": 4.500580798237831e-06,
"loss": 0.0112,
"step": 564
},
{
"epoch": 2.288032454361055,
"grad_norm": 0.36204770339604975,
"learning_rate": 4.452812429727313e-06,
"loss": 0.0102,
"step": 565
},
{
"epoch": 2.292089249492901,
"grad_norm": 0.27129867464793533,
"learning_rate": 4.405254693551754e-06,
"loss": 0.0129,
"step": 566
},
{
"epoch": 2.2961460446247464,
"grad_norm": 0.22000281062189,
"learning_rate": 4.357908539461679e-06,
"loss": 0.008,
"step": 567
},
{
"epoch": 2.3002028397565923,
"grad_norm": 0.29624602171723224,
"learning_rate": 4.310774912982227e-06,
"loss": 0.0144,
"step": 568
},
{
"epoch": 2.3042596348884383,
"grad_norm": 0.22414337252422092,
"learning_rate": 4.263854755394256e-06,
"loss": 0.0089,
"step": 569
},
{
"epoch": 2.308316430020284,
"grad_norm": 0.28474418070215113,
"learning_rate": 4.21714900371556e-06,
"loss": 0.012,
"step": 570
},
{
"epoch": 2.31237322515213,
"grad_norm": 0.210326196860165,
"learning_rate": 4.170658590682134e-06,
"loss": 0.0082,
"step": 571
},
{
"epoch": 2.316430020283976,
"grad_norm": 0.25084743918519153,
"learning_rate": 4.124384444729561e-06,
"loss": 0.0089,
"step": 572
},
{
"epoch": 2.3204868154158214,
"grad_norm": 0.28537259628253975,
"learning_rate": 4.078327489974466e-06,
"loss": 0.0113,
"step": 573
},
{
"epoch": 2.3245436105476673,
"grad_norm": 0.22908276842060815,
"learning_rate": 4.032488646196077e-06,
"loss": 0.0101,
"step": 574
},
{
"epoch": 2.3286004056795133,
"grad_norm": 0.23845424830115683,
"learning_rate": 3.986868828817818e-06,
"loss": 0.0082,
"step": 575
},
{
"epoch": 2.332657200811359,
"grad_norm": 0.184143153098664,
"learning_rate": 3.941468948889067e-06,
"loss": 0.0068,
"step": 576
},
{
"epoch": 2.336713995943205,
"grad_norm": 0.6122313601455306,
"learning_rate": 3.8962899130669525e-06,
"loss": 0.0108,
"step": 577
},
{
"epoch": 2.340770791075051,
"grad_norm": 0.28166906523725416,
"learning_rate": 3.851332623598227e-06,
"loss": 0.0123,
"step": 578
},
{
"epoch": 2.344827586206897,
"grad_norm": 0.25383591722677934,
"learning_rate": 3.8065979783012746e-06,
"loss": 0.0112,
"step": 579
},
{
"epoch": 2.3488843813387423,
"grad_norm": 0.19603119074642442,
"learning_rate": 3.7620868705481586e-06,
"loss": 0.0079,
"step": 580
},
{
"epoch": 2.3529411764705883,
"grad_norm": 0.6430341184106714,
"learning_rate": 3.717800189246807e-06,
"loss": 0.0079,
"step": 581
},
{
"epoch": 2.356997971602434,
"grad_norm": 0.18811787242514347,
"learning_rate": 3.6737388188232305e-06,
"loss": 0.0083,
"step": 582
},
{
"epoch": 2.36105476673428,
"grad_norm": 0.20181404498766184,
"learning_rate": 3.629903639203884e-06,
"loss": 0.0083,
"step": 583
},
{
"epoch": 2.365111561866126,
"grad_norm": 0.277342849633144,
"learning_rate": 3.5862955257980813e-06,
"loss": 0.0117,
"step": 584
},
{
"epoch": 2.369168356997972,
"grad_norm": 0.33333350892821556,
"learning_rate": 3.5429153494805087e-06,
"loss": 0.0123,
"step": 585
},
{
"epoch": 2.3732251521298173,
"grad_norm": 0.23513638472728224,
"learning_rate": 3.499763976573866e-06,
"loss": 0.0108,
"step": 586
},
{
"epoch": 2.3772819472616633,
"grad_norm": 0.9093270963617072,
"learning_rate": 3.4568422688315027e-06,
"loss": 0.0142,
"step": 587
},
{
"epoch": 2.3813387423935093,
"grad_norm": 0.34746001689384526,
"learning_rate": 3.41415108342028e-06,
"loss": 0.0091,
"step": 588
},
{
"epoch": 2.385395537525355,
"grad_norm": 0.2173220084157316,
"learning_rate": 3.371691272903398e-06,
"loss": 0.0072,
"step": 589
},
{
"epoch": 2.389452332657201,
"grad_norm": 0.24334697959441584,
"learning_rate": 3.329463685223411e-06,
"loss": 0.01,
"step": 590
},
{
"epoch": 2.393509127789047,
"grad_norm": 0.17003427758882395,
"learning_rate": 3.287469163685241e-06,
"loss": 0.0078,
"step": 591
},
{
"epoch": 2.3975659229208923,
"grad_norm": 0.275983232519286,
"learning_rate": 3.2457085469394015e-06,
"loss": 0.0098,
"step": 592
},
{
"epoch": 2.4016227180527383,
"grad_norm": 0.2634629301055357,
"learning_rate": 3.204182668965198e-06,
"loss": 0.0093,
"step": 593
},
{
"epoch": 2.4056795131845843,
"grad_norm": 0.22325919432645464,
"learning_rate": 3.162892359054098e-06,
"loss": 0.0102,
"step": 594
},
{
"epoch": 2.40973630831643,
"grad_norm": 0.18144358371062383,
"learning_rate": 3.12183844179316e-06,
"loss": 0.0072,
"step": 595
},
{
"epoch": 2.413793103448276,
"grad_norm": 0.19726131763241822,
"learning_rate": 3.081021737048565e-06,
"loss": 0.0079,
"step": 596
},
{
"epoch": 2.417849898580122,
"grad_norm": 0.2681563404805776,
"learning_rate": 3.040443059949264e-06,
"loss": 0.0124,
"step": 597
},
{
"epoch": 2.4219066937119678,
"grad_norm": 2.0846202190030847,
"learning_rate": 3.0001032208706653e-06,
"loss": 0.0108,
"step": 598
},
{
"epoch": 2.4259634888438133,
"grad_norm": 0.20806969883961424,
"learning_rate": 2.960003025418478e-06,
"loss": 0.0082,
"step": 599
},
{
"epoch": 2.4300202839756593,
"grad_norm": 0.21726326567282914,
"learning_rate": 2.9201432744126074e-06,
"loss": 0.0114,
"step": 600
},
{
"epoch": 2.4340770791075053,
"grad_norm": 0.23506065145671068,
"learning_rate": 2.8805247638711814e-06,
"loss": 0.0084,
"step": 601
},
{
"epoch": 2.438133874239351,
"grad_norm": 0.18812525391216506,
"learning_rate": 2.8411482849946352e-06,
"loss": 0.0067,
"step": 602
},
{
"epoch": 2.4421906693711968,
"grad_norm": 0.22631798439472142,
"learning_rate": 2.802014624149902e-06,
"loss": 0.0083,
"step": 603
},
{
"epoch": 2.4462474645030428,
"grad_norm": 0.23945298074449045,
"learning_rate": 2.7631245628547496e-06,
"loss": 0.008,
"step": 604
},
{
"epoch": 2.4503042596348883,
"grad_norm": 0.25259941453485113,
"learning_rate": 2.7244788777621326e-06,
"loss": 0.0084,
"step": 605
},
{
"epoch": 2.4543610547667343,
"grad_norm": 0.2652122618762015,
"learning_rate": 2.6860783406446986e-06,
"loss": 0.0083,
"step": 606
},
{
"epoch": 2.4584178498985803,
"grad_norm": 0.22664519985433124,
"learning_rate": 2.6479237183793716e-06,
"loss": 0.0081,
"step": 607
},
{
"epoch": 2.462474645030426,
"grad_norm": 0.27386433835548185,
"learning_rate": 2.6100157729320497e-06,
"loss": 0.0084,
"step": 608
},
{
"epoch": 2.4665314401622718,
"grad_norm": 0.24744848716097542,
"learning_rate": 2.572355261342369e-06,
"loss": 0.0103,
"step": 609
},
{
"epoch": 2.4705882352941178,
"grad_norm": 0.29076814152760166,
"learning_rate": 2.5349429357085952e-06,
"loss": 0.0115,
"step": 610
},
{
"epoch": 2.4746450304259637,
"grad_norm": 0.246208068848628,
"learning_rate": 2.497779543172603e-06,
"loss": 0.0112,
"step": 611
},
{
"epoch": 2.4787018255578093,
"grad_norm": 0.18929036981063926,
"learning_rate": 2.4608658259049544e-06,
"loss": 0.0074,
"step": 612
},
{
"epoch": 2.4827586206896552,
"grad_norm": 0.3125249504834572,
"learning_rate": 2.424202521090091e-06,
"loss": 0.0091,
"step": 613
},
{
"epoch": 2.486815415821501,
"grad_norm": 0.23707086469052593,
"learning_rate": 2.3877903609115697e-06,
"loss": 0.0088,
"step": 614
},
{
"epoch": 2.4908722109533468,
"grad_norm": 0.32146587243757385,
"learning_rate": 2.3516300725375016e-06,
"loss": 0.0118,
"step": 615
},
{
"epoch": 2.4949290060851927,
"grad_norm": 0.25434522572220347,
"learning_rate": 2.3157223781059776e-06,
"loss": 0.0079,
"step": 616
},
{
"epoch": 2.4989858012170387,
"grad_norm": 0.24238745433424433,
"learning_rate": 2.2800679947106858e-06,
"loss": 0.0109,
"step": 617
},
{
"epoch": 2.5030425963488843,
"grad_norm": 0.25094583442033846,
"learning_rate": 2.2446676343865464e-06,
"loss": 0.0118,
"step": 618
},
{
"epoch": 2.5070993914807302,
"grad_norm": 0.16936601125564227,
"learning_rate": 2.209522004095547e-06,
"loss": 0.0054,
"step": 619
},
{
"epoch": 2.5111561866125762,
"grad_norm": 0.24921797976060717,
"learning_rate": 2.174631805712576e-06,
"loss": 0.0097,
"step": 620
},
{
"epoch": 2.5152129817444218,
"grad_norm": 0.2993762969900287,
"learning_rate": 2.1399977360114376e-06,
"loss": 0.0106,
"step": 621
},
{
"epoch": 2.5192697768762677,
"grad_norm": 0.19797722360809955,
"learning_rate": 2.1056204866509145e-06,
"loss": 0.0085,
"step": 622
},
{
"epoch": 2.5233265720081137,
"grad_norm": 0.282509193504894,
"learning_rate": 2.0715007441609694e-06,
"loss": 0.0108,
"step": 623
},
{
"epoch": 2.5273833671399597,
"grad_norm": 0.22730726091766484,
"learning_rate": 2.0376391899290397e-06,
"loss": 0.0082,
"step": 624
},
{
"epoch": 2.5314401622718052,
"grad_norm": 0.22036323180624318,
"learning_rate": 2.0040365001864074e-06,
"loss": 0.0079,
"step": 625
},
{
"epoch": 2.535496957403651,
"grad_norm": 0.24907228575266155,
"learning_rate": 1.9706933459947156e-06,
"loss": 0.0115,
"step": 626
},
{
"epoch": 2.5395537525354968,
"grad_norm": 0.29345103112240617,
"learning_rate": 1.9376103932325552e-06,
"loss": 0.0119,
"step": 627
},
{
"epoch": 2.5436105476673427,
"grad_norm": 0.2276489727076261,
"learning_rate": 1.9047883025821777e-06,
"loss": 0.0099,
"step": 628
},
{
"epoch": 2.5476673427991887,
"grad_norm": 0.1682439024937179,
"learning_rate": 1.8722277295162881e-06,
"loss": 0.0055,
"step": 629
},
{
"epoch": 2.5517241379310347,
"grad_norm": 0.29245117392862996,
"learning_rate": 1.8399293242849635e-06,
"loss": 0.0114,
"step": 630
},
{
"epoch": 2.5557809330628802,
"grad_norm": 0.25775324221398893,
"learning_rate": 1.8078937319026655e-06,
"loss": 0.0101,
"step": 631
},
{
"epoch": 2.559837728194726,
"grad_norm": 0.20713597802336198,
"learning_rate": 1.776121592135358e-06,
"loss": 0.0075,
"step": 632
},
{
"epoch": 2.5638945233265718,
"grad_norm": 0.24027964794024184,
"learning_rate": 1.7446135394877327e-06,
"loss": 0.0086,
"step": 633
},
{
"epoch": 2.5679513184584177,
"grad_norm": 0.2272031647767097,
"learning_rate": 1.7133702031905307e-06,
"loss": 0.0109,
"step": 634
},
{
"epoch": 2.5720081135902637,
"grad_norm": 0.37511064492446994,
"learning_rate": 1.6823922071879932e-06,
"loss": 0.0103,
"step": 635
},
{
"epoch": 2.5760649087221097,
"grad_norm": 0.18921052381613496,
"learning_rate": 1.6516801701253802e-06,
"loss": 0.0061,
"step": 636
},
{
"epoch": 2.5801217038539552,
"grad_norm": 0.3363009906887839,
"learning_rate": 1.621234705336629e-06,
"loss": 0.008,
"step": 637
},
{
"epoch": 2.584178498985801,
"grad_norm": 0.34887166835747263,
"learning_rate": 1.5910564208320993e-06,
"loss": 0.0099,
"step": 638
},
{
"epoch": 2.588235294117647,
"grad_norm": 0.258451194057935,
"learning_rate": 1.561145919286442e-06,
"loss": 0.0103,
"step": 639
},
{
"epoch": 2.5922920892494927,
"grad_norm": 0.3295393253593759,
"learning_rate": 1.5315037980265478e-06,
"loss": 0.0103,
"step": 640
},
{
"epoch": 2.5963488843813387,
"grad_norm": 0.19642120190004772,
"learning_rate": 1.5021306490196273e-06,
"loss": 0.0072,
"step": 641
},
{
"epoch": 2.6004056795131847,
"grad_norm": 0.18844082881675525,
"learning_rate": 1.4730270588613876e-06,
"loss": 0.0076,
"step": 642
},
{
"epoch": 2.6044624746450307,
"grad_norm": 0.2334747249976257,
"learning_rate": 1.4441936087643182e-06,
"loss": 0.0085,
"step": 643
},
{
"epoch": 2.608519269776876,
"grad_norm": 0.21454076957961676,
"learning_rate": 1.4156308745460928e-06,
"loss": 0.0079,
"step": 644
},
{
"epoch": 2.612576064908722,
"grad_norm": 0.23373233938323304,
"learning_rate": 1.3873394266180377e-06,
"loss": 0.009,
"step": 645
},
{
"epoch": 2.6166328600405677,
"grad_norm": 0.24633588532171205,
"learning_rate": 1.3593198299737869e-06,
"loss": 0.0065,
"step": 646
},
{
"epoch": 2.6206896551724137,
"grad_norm": 0.26000762104813135,
"learning_rate": 1.331572644177963e-06,
"loss": 0.0123,
"step": 647
},
{
"epoch": 2.6247464503042597,
"grad_norm": 0.2915566393938057,
"learning_rate": 1.3040984233550242e-06,
"loss": 0.0101,
"step": 648
},
{
"epoch": 2.6288032454361057,
"grad_norm": 0.20647014289947155,
"learning_rate": 1.2768977161781736e-06,
"loss": 0.0098,
"step": 649
},
{
"epoch": 2.632860040567951,
"grad_norm": 0.24731975614208263,
"learning_rate": 1.2499710658584296e-06,
"loss": 0.0108,
"step": 650
},
{
"epoch": 2.636916835699797,
"grad_norm": 0.31247533393817645,
"learning_rate": 1.2233190101337644e-06,
"loss": 0.0126,
"step": 651
},
{
"epoch": 2.640973630831643,
"grad_norm": 0.23703695879461487,
"learning_rate": 1.1969420812583603e-06,
"loss": 0.0113,
"step": 652
},
{
"epoch": 2.6450304259634887,
"grad_norm": 0.29797798103465356,
"learning_rate": 1.1708408059919901e-06,
"loss": 0.0139,
"step": 653
},
{
"epoch": 2.6490872210953347,
"grad_norm": 3.139650970793164,
"learning_rate": 1.145015705589491e-06,
"loss": 0.0112,
"step": 654
},
{
"epoch": 2.6531440162271807,
"grad_norm": 0.2405246638894348,
"learning_rate": 1.119467295790363e-06,
"loss": 0.011,
"step": 655
},
{
"epoch": 2.6572008113590266,
"grad_norm": 0.24093581404167375,
"learning_rate": 1.0941960868084549e-06,
"loss": 0.0108,
"step": 656
},
{
"epoch": 2.661257606490872,
"grad_norm": 1.4166321141255014,
"learning_rate": 1.06920258332179e-06,
"loss": 0.0092,
"step": 657
},
{
"epoch": 2.665314401622718,
"grad_norm": 0.2516080027548331,
"learning_rate": 1.0444872844624804e-06,
"loss": 0.0091,
"step": 658
},
{
"epoch": 2.6693711967545637,
"grad_norm": 0.27372507155820414,
"learning_rate": 1.0200506838067558e-06,
"loss": 0.0161,
"step": 659
},
{
"epoch": 2.6734279918864097,
"grad_norm": 0.22462772010565904,
"learning_rate": 9.958932693651168e-07,
"loss": 0.0085,
"step": 660
},
{
"epoch": 2.6774847870182557,
"grad_norm": 0.19482170746221347,
"learning_rate": 9.720155235725753e-07,
"loss": 0.0077,
"step": 661
},
{
"epoch": 2.6815415821501016,
"grad_norm": 0.21706494083610625,
"learning_rate": 9.484179232790391e-07,
"loss": 0.0071,
"step": 662
},
{
"epoch": 2.685598377281947,
"grad_norm": 0.2700405708098266,
"learning_rate": 9.251009397397663e-07,
"loss": 0.0096,
"step": 663
},
{
"epoch": 2.689655172413793,
"grad_norm": 0.20416894288766446,
"learning_rate": 9.020650386059731e-07,
"loss": 0.0084,
"step": 664
},
{
"epoch": 2.6937119675456387,
"grad_norm": 0.2977765225079465,
"learning_rate": 8.793106799155187e-07,
"loss": 0.0122,
"step": 665
},
{
"epoch": 2.6977687626774847,
"grad_norm": 0.39111551936936895,
"learning_rate": 8.568383180837369e-07,
"loss": 0.0061,
"step": 666
},
{
"epoch": 2.7018255578093306,
"grad_norm": 0.23369929762822697,
"learning_rate": 8.346484018943429e-07,
"loss": 0.0071,
"step": 667
},
{
"epoch": 2.7058823529411766,
"grad_norm": 0.2529908334984344,
"learning_rate": 8.127413744904805e-07,
"loss": 0.0109,
"step": 668
},
{
"epoch": 2.709939148073022,
"grad_norm": 0.19978914969994244,
"learning_rate": 7.911176733658709e-07,
"loss": 0.0071,
"step": 669
},
{
"epoch": 2.713995943204868,
"grad_norm": 0.35516946776094027,
"learning_rate": 7.697777303560699e-07,
"loss": 0.0076,
"step": 670
},
{
"epoch": 2.718052738336714,
"grad_norm": 0.32064994090406973,
"learning_rate": 7.487219716298643e-07,
"loss": 0.0078,
"step": 671
},
{
"epoch": 2.7221095334685597,
"grad_norm": 0.17610583979954028,
"learning_rate": 7.279508176807259e-07,
"loss": 0.007,
"step": 672
},
{
"epoch": 2.7261663286004056,
"grad_norm": 0.1879019239049559,
"learning_rate": 7.074646833184495e-07,
"loss": 0.0077,
"step": 673
},
{
"epoch": 2.7302231237322516,
"grad_norm": 0.2508655898096489,
"learning_rate": 6.872639776608464e-07,
"loss": 0.012,
"step": 674
},
{
"epoch": 2.7342799188640976,
"grad_norm": 0.20939981063074065,
"learning_rate": 6.673491041255874e-07,
"loss": 0.0072,
"step": 675
},
{
"epoch": 2.738336713995943,
"grad_norm": 0.2524661562844755,
"learning_rate": 6.477204604221298e-07,
"loss": 0.0089,
"step": 676
},
{
"epoch": 2.742393509127789,
"grad_norm": 0.18123227676882245,
"learning_rate": 6.283784385437963e-07,
"loss": 0.0074,
"step": 677
},
{
"epoch": 2.7464503042596347,
"grad_norm": 0.28169632716010695,
"learning_rate": 6.093234247599317e-07,
"loss": 0.0115,
"step": 678
},
{
"epoch": 2.7505070993914806,
"grad_norm": 0.24112096399796779,
"learning_rate": 5.905557996081906e-07,
"loss": 0.0104,
"step": 679
},
{
"epoch": 2.7545638945233266,
"grad_norm": 1.891213001719549,
"learning_rate": 5.720759378869455e-07,
"loss": 0.0108,
"step": 680
},
{
"epoch": 2.7586206896551726,
"grad_norm": 0.2143920645162132,
"learning_rate": 5.538842086477891e-07,
"loss": 0.0073,
"step": 681
},
{
"epoch": 2.762677484787018,
"grad_norm": 0.3298995677537989,
"learning_rate": 5.359809751881834e-07,
"loss": 0.0114,
"step": 682
},
{
"epoch": 2.766734279918864,
"grad_norm": 0.18127821937547067,
"learning_rate": 5.18366595044184e-07,
"loss": 0.0059,
"step": 683
},
{
"epoch": 2.77079107505071,
"grad_norm": 0.3580659253774967,
"learning_rate": 5.010414199833097e-07,
"loss": 0.0134,
"step": 684
},
{
"epoch": 2.7748478701825556,
"grad_norm": 0.20714170407676127,
"learning_rate": 4.840057959975169e-07,
"loss": 0.0072,
"step": 685
},
{
"epoch": 2.7789046653144016,
"grad_norm": 0.2277337899414502,
"learning_rate": 4.6726006329629645e-07,
"loss": 0.0101,
"step": 686
},
{
"epoch": 2.7829614604462476,
"grad_norm": 0.19613471713041244,
"learning_rate": 4.5080455629986475e-07,
"loss": 0.0047,
"step": 687
},
{
"epoch": 2.7870182555780936,
"grad_norm": 0.2823039689966462,
"learning_rate": 4.3463960363249286e-07,
"loss": 0.0115,
"step": 688
},
{
"epoch": 2.791075050709939,
"grad_norm": 0.20345163681925613,
"learning_rate": 4.187655281159536e-07,
"loss": 0.0074,
"step": 689
},
{
"epoch": 2.795131845841785,
"grad_norm": 0.29840307095638663,
"learning_rate": 4.031826467630545e-07,
"loss": 0.01,
"step": 690
},
{
"epoch": 2.7991886409736306,
"grad_norm": 0.1860689908066093,
"learning_rate": 3.8789127077132336e-07,
"loss": 0.0055,
"step": 691
},
{
"epoch": 2.8032454361054766,
"grad_norm": 0.3478181689879201,
"learning_rate": 3.7289170551678633e-07,
"loss": 0.0118,
"step": 692
},
{
"epoch": 2.8073022312373226,
"grad_norm": 0.22939108780939063,
"learning_rate": 3.581842505478744e-07,
"loss": 0.0112,
"step": 693
},
{
"epoch": 2.8113590263691686,
"grad_norm": 0.3032587745992294,
"learning_rate": 3.437691995794301e-07,
"loss": 0.0124,
"step": 694
},
{
"epoch": 2.815415821501014,
"grad_norm": 0.2094386636776182,
"learning_rate": 3.2964684048685523e-07,
"loss": 0.0073,
"step": 695
},
{
"epoch": 2.81947261663286,
"grad_norm": 0.24925291522480858,
"learning_rate": 3.15817455300354e-07,
"loss": 0.01,
"step": 696
},
{
"epoch": 2.8235294117647056,
"grad_norm": 0.22927354124709365,
"learning_rate": 3.022813201993041e-07,
"loss": 0.0103,
"step": 697
},
{
"epoch": 2.8275862068965516,
"grad_norm": 0.24523657497451284,
"learning_rate": 2.890387055067412e-07,
"loss": 0.0092,
"step": 698
},
{
"epoch": 2.8316430020283976,
"grad_norm": 0.2375204245523175,
"learning_rate": 2.7608987568395306e-07,
"loss": 0.0116,
"step": 699
},
{
"epoch": 2.8356997971602436,
"grad_norm": 0.2066202436946509,
"learning_rate": 2.6343508932521243e-07,
"loss": 0.0086,
"step": 700
},
{
"epoch": 2.839756592292089,
"grad_norm": 0.1847332447135413,
"learning_rate": 2.510745991525992e-07,
"loss": 0.0069,
"step": 701
},
{
"epoch": 2.843813387423935,
"grad_norm": 0.21938966536504567,
"learning_rate": 2.39008652010963e-07,
"loss": 0.0086,
"step": 702
},
{
"epoch": 2.847870182555781,
"grad_norm": 0.23134459803966625,
"learning_rate": 2.2723748886298523e-07,
"loss": 0.009,
"step": 703
},
{
"epoch": 2.8519269776876266,
"grad_norm": 0.26561602719544714,
"learning_rate": 2.1576134478437315e-07,
"loss": 0.0102,
"step": 704
},
{
"epoch": 2.8559837728194726,
"grad_norm": 1.1071488735425152,
"learning_rate": 2.0458044895916516e-07,
"loss": 0.0084,
"step": 705
},
{
"epoch": 2.8600405679513186,
"grad_norm": 0.27265875970319914,
"learning_rate": 1.9369502467514788e-07,
"loss": 0.0111,
"step": 706
},
{
"epoch": 2.8640973630831645,
"grad_norm": 0.22724233458925192,
"learning_rate": 1.831052893194063e-07,
"loss": 0.0089,
"step": 707
},
{
"epoch": 2.86815415821501,
"grad_norm": 0.26572957456388885,
"learning_rate": 1.7281145437397394e-07,
"loss": 0.0102,
"step": 708
},
{
"epoch": 2.872210953346856,
"grad_norm": 0.21131153496136418,
"learning_rate": 1.62813725411613e-07,
"loss": 0.0081,
"step": 709
},
{
"epoch": 2.8762677484787016,
"grad_norm": 0.2164022001956463,
"learning_rate": 1.5311230209171078e-07,
"loss": 0.0115,
"step": 710
},
{
"epoch": 2.8803245436105476,
"grad_norm": 0.18362776733451797,
"learning_rate": 1.4370737815628809e-07,
"loss": 0.0068,
"step": 711
},
{
"epoch": 2.8843813387423936,
"grad_norm": 0.22201535643027853,
"learning_rate": 1.3459914142613384e-07,
"loss": 0.0093,
"step": 712
},
{
"epoch": 2.8884381338742395,
"grad_norm": 0.23084858527126895,
"learning_rate": 1.2578777379705476e-07,
"loss": 0.0084,
"step": 713
},
{
"epoch": 2.892494929006085,
"grad_norm": 0.33392265560180945,
"learning_rate": 1.1727345123623667e-07,
"loss": 0.0108,
"step": 714
},
{
"epoch": 2.896551724137931,
"grad_norm": 0.2883283240758792,
"learning_rate": 1.0905634377873563e-07,
"loss": 0.0092,
"step": 715
},
{
"epoch": 2.900608519269777,
"grad_norm": 0.2686572168243743,
"learning_rate": 1.011366155240856e-07,
"loss": 0.0131,
"step": 716
},
{
"epoch": 2.9046653144016226,
"grad_norm": 0.2101145436417912,
"learning_rate": 9.351442463301274e-08,
"loss": 0.0073,
"step": 717
},
{
"epoch": 2.9087221095334685,
"grad_norm": 0.18153151909382642,
"learning_rate": 8.618992332427966e-08,
"loss": 0.0054,
"step": 718
},
{
"epoch": 2.9127789046653145,
"grad_norm": 0.1970262952306368,
"learning_rate": 7.916325787164947e-08,
"loss": 0.0073,
"step": 719
},
{
"epoch": 2.9168356997971605,
"grad_norm": 0.23074904705177698,
"learning_rate": 7.243456860096476e-08,
"loss": 0.0079,
"step": 720
},
{
"epoch": 2.920892494929006,
"grad_norm": 0.2628324630369001,
"learning_rate": 6.600398988733824e-08,
"loss": 0.0085,
"step": 721
},
{
"epoch": 2.924949290060852,
"grad_norm": 0.28909539190720107,
"learning_rate": 5.98716501524732e-08,
"loss": 0.0126,
"step": 722
},
{
"epoch": 2.9290060851926976,
"grad_norm": 0.27286922538855823,
"learning_rate": 5.403767186210218e-08,
"loss": 0.0093,
"step": 723
},
{
"epoch": 2.9330628803245435,
"grad_norm": 0.2568428715793184,
"learning_rate": 4.850217152353731e-08,
"loss": 0.0074,
"step": 724
},
{
"epoch": 2.9371196754563895,
"grad_norm": 0.4731785556263277,
"learning_rate": 4.326525968334216e-08,
"loss": 0.0076,
"step": 725
},
{
"epoch": 2.9411764705882355,
"grad_norm": 0.2031164746992896,
"learning_rate": 3.8327040925130175e-08,
"loss": 0.0074,
"step": 726
},
{
"epoch": 2.945233265720081,
"grad_norm": 0.4807857991510011,
"learning_rate": 3.368761386746966e-08,
"loss": 0.0072,
"step": 727
},
{
"epoch": 2.949290060851927,
"grad_norm": 0.66364903613515,
"learning_rate": 2.9347071161918703e-08,
"loss": 0.0097,
"step": 728
},
{
"epoch": 2.9533468559837726,
"grad_norm": 0.3120023778508287,
"learning_rate": 2.530549949117167e-08,
"loss": 0.0103,
"step": 729
},
{
"epoch": 2.9574036511156185,
"grad_norm": 0.20907205106461102,
"learning_rate": 2.1562979567330554e-08,
"loss": 0.0069,
"step": 730
},
{
"epoch": 2.9614604462474645,
"grad_norm": 0.23184300955403422,
"learning_rate": 1.8119586130292964e-08,
"loss": 0.0062,
"step": 731
},
{
"epoch": 2.9655172413793105,
"grad_norm": 0.2240122101684402,
"learning_rate": 1.4975387946256634e-08,
"loss": 0.0083,
"step": 732
},
{
"epoch": 2.969574036511156,
"grad_norm": 0.23638952694105592,
"learning_rate": 1.213044780635053e-08,
"loss": 0.0078,
"step": 733
},
{
"epoch": 2.973630831643002,
"grad_norm": 0.23626023739586408,
"learning_rate": 9.584822525377512e-09,
"loss": 0.0081,
"step": 734
},
{
"epoch": 2.977687626774848,
"grad_norm": 0.19999524138260236,
"learning_rate": 7.338562940680249e-09,
"loss": 0.0074,
"step": 735
},
{
"epoch": 2.9817444219066935,
"grad_norm": 0.21364040585606983,
"learning_rate": 5.391713911128693e-09,
"loss": 0.0069,
"step": 736
},
{
"epoch": 2.9858012170385395,
"grad_norm": 0.20996179998709147,
"learning_rate": 3.744314316220798e-09,
"loss": 0.0072,
"step": 737
},
{
"epoch": 2.9898580121703855,
"grad_norm": 0.19358339065466926,
"learning_rate": 2.396397055306476e-09,
"loss": 0.0061,
"step": 738
},
{
"epoch": 2.9939148073022315,
"grad_norm": 0.18667044909228617,
"learning_rate": 1.3479890469314527e-09,
"loss": 0.0075,
"step": 739
},
{
"epoch": 2.997971602434077,
"grad_norm": 0.1920159799469308,
"learning_rate": 5.991112283026956e-10,
"loss": 0.0074,
"step": 740
},
{
"epoch": 3.0,
"grad_norm": 0.1960028788559094,
"learning_rate": 1.4977855486209002e-10,
"loss": 0.0037,
"step": 741
},
{
"epoch": 3.0,
"step": 741,
"total_flos": 5016006509985792.0,
"train_loss": 0.12115606438027954,
"train_runtime": 234868.3901,
"train_samples_per_second": 4.024,
"train_steps_per_second": 0.003
}
],
"logging_steps": 1,
"max_steps": 741,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 50,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 5016006509985792.0,
"train_batch_size": 10,
"trial_name": null,
"trial_params": null
}