|
{
|
|
"best_metric": 0.839046835899353,
|
|
"best_model_checkpoint": "./bert-large-jpn-grammar\\checkpoint-5980",
|
|
"epoch": 10.0,
|
|
"eval_steps": 500,
|
|
"global_step": 5980,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"epoch": 0.0016722408026755853,
|
|
"grad_norm": 4.53647518157959,
|
|
"learning_rate": 2.5083612040133778e-08,
|
|
"loss": 7.8555,
|
|
"step": 1
|
|
},
|
|
{
|
|
"epoch": 0.08361204013377926,
|
|
"grad_norm": 4.600866317749023,
|
|
"learning_rate": 1.254180602006689e-06,
|
|
"loss": 7.8831,
|
|
"step": 50
|
|
},
|
|
{
|
|
"epoch": 0.16722408026755853,
|
|
"grad_norm": 4.168644428253174,
|
|
"learning_rate": 2.508361204013378e-06,
|
|
"loss": 7.8681,
|
|
"step": 100
|
|
},
|
|
{
|
|
"epoch": 0.2508361204013378,
|
|
"grad_norm": 13.43441104888916,
|
|
"learning_rate": 3.762541806020067e-06,
|
|
"loss": 7.8594,
|
|
"step": 150
|
|
},
|
|
{
|
|
"epoch": 0.33444816053511706,
|
|
"grad_norm": 3.9283835887908936,
|
|
"learning_rate": 5.016722408026756e-06,
|
|
"loss": 7.8337,
|
|
"step": 200
|
|
},
|
|
{
|
|
"epoch": 0.4180602006688963,
|
|
"grad_norm": 3.899462938308716,
|
|
"learning_rate": 6.270903010033445e-06,
|
|
"loss": 7.8177,
|
|
"step": 250
|
|
},
|
|
{
|
|
"epoch": 0.5016722408026756,
|
|
"grad_norm": 4.085901737213135,
|
|
"learning_rate": 7.525083612040134e-06,
|
|
"loss": 7.7797,
|
|
"step": 300
|
|
},
|
|
{
|
|
"epoch": 0.5852842809364549,
|
|
"grad_norm": 4.916100025177002,
|
|
"learning_rate": 8.779264214046824e-06,
|
|
"loss": 7.7515,
|
|
"step": 350
|
|
},
|
|
{
|
|
"epoch": 0.6688963210702341,
|
|
"grad_norm": 5.722199440002441,
|
|
"learning_rate": 1.0033444816053512e-05,
|
|
"loss": 7.7095,
|
|
"step": 400
|
|
},
|
|
{
|
|
"epoch": 0.7525083612040134,
|
|
"grad_norm": 6.974526405334473,
|
|
"learning_rate": 1.1287625418060201e-05,
|
|
"loss": 7.6088,
|
|
"step": 450
|
|
},
|
|
{
|
|
"epoch": 0.8361204013377926,
|
|
"grad_norm": 7.1739888191223145,
|
|
"learning_rate": 1.254180602006689e-05,
|
|
"loss": 7.4935,
|
|
"step": 500
|
|
},
|
|
{
|
|
"epoch": 0.919732441471572,
|
|
"grad_norm": 7.732913494110107,
|
|
"learning_rate": 1.379598662207358e-05,
|
|
"loss": 7.3773,
|
|
"step": 550
|
|
},
|
|
{
|
|
"epoch": 1.0,
|
|
"eval_accuracy": 0.1417224109172821,
|
|
"eval_f1": 0.1417224109172821,
|
|
"eval_loss": 7.010268688201904,
|
|
"eval_runtime": 59.6669,
|
|
"eval_samples_per_second": 80.178,
|
|
"eval_steps_per_second": 1.257,
|
|
"step": 598
|
|
},
|
|
{
|
|
"epoch": 1.0033444816053512,
|
|
"grad_norm": 7.402344703674316,
|
|
"learning_rate": 1.5050167224080267e-05,
|
|
"loss": 7.2391,
|
|
"step": 600
|
|
},
|
|
{
|
|
"epoch": 1.0869565217391304,
|
|
"grad_norm": 7.742281913757324,
|
|
"learning_rate": 1.6304347826086955e-05,
|
|
"loss": 6.9797,
|
|
"step": 650
|
|
},
|
|
{
|
|
"epoch": 1.1705685618729098,
|
|
"grad_norm": 7.522943019866943,
|
|
"learning_rate": 1.7558528428093648e-05,
|
|
"loss": 6.8021,
|
|
"step": 700
|
|
},
|
|
{
|
|
"epoch": 1.254180602006689,
|
|
"grad_norm": 7.755581378936768,
|
|
"learning_rate": 1.8812709030100337e-05,
|
|
"loss": 6.6368,
|
|
"step": 750
|
|
},
|
|
{
|
|
"epoch": 1.3377926421404682,
|
|
"grad_norm": 7.403460502624512,
|
|
"learning_rate": 2.0066889632107023e-05,
|
|
"loss": 6.4424,
|
|
"step": 800
|
|
},
|
|
{
|
|
"epoch": 1.4214046822742474,
|
|
"grad_norm": 7.275426864624023,
|
|
"learning_rate": 2.1321070234113713e-05,
|
|
"loss": 6.2322,
|
|
"step": 850
|
|
},
|
|
{
|
|
"epoch": 1.5050167224080266,
|
|
"grad_norm": 7.691336631774902,
|
|
"learning_rate": 2.2575250836120402e-05,
|
|
"loss": 6.0312,
|
|
"step": 900
|
|
},
|
|
{
|
|
"epoch": 1.588628762541806,
|
|
"grad_norm": 7.337137699127197,
|
|
"learning_rate": 2.382943143812709e-05,
|
|
"loss": 5.8301,
|
|
"step": 950
|
|
},
|
|
{
|
|
"epoch": 1.6722408026755853,
|
|
"grad_norm": 7.574013710021973,
|
|
"learning_rate": 2.508361204013378e-05,
|
|
"loss": 5.6195,
|
|
"step": 1000
|
|
},
|
|
{
|
|
"epoch": 1.7558528428093645,
|
|
"grad_norm": 7.1535773277282715,
|
|
"learning_rate": 2.6337792642140467e-05,
|
|
"loss": 5.3459,
|
|
"step": 1050
|
|
},
|
|
{
|
|
"epoch": 1.839464882943144,
|
|
"grad_norm": 7.483964443206787,
|
|
"learning_rate": 2.759197324414716e-05,
|
|
"loss": 5.1481,
|
|
"step": 1100
|
|
},
|
|
{
|
|
"epoch": 1.9230769230769231,
|
|
"grad_norm": 7.383283615112305,
|
|
"learning_rate": 2.884615384615385e-05,
|
|
"loss": 4.9299,
|
|
"step": 1150
|
|
},
|
|
{
|
|
"epoch": 2.0,
|
|
"eval_accuracy": 0.6444398164749146,
|
|
"eval_f1": 0.6444398164749146,
|
|
"eval_loss": 4.192617893218994,
|
|
"eval_runtime": 55.4042,
|
|
"eval_samples_per_second": 86.347,
|
|
"eval_steps_per_second": 1.354,
|
|
"step": 1196
|
|
},
|
|
{
|
|
"epoch": 2.0066889632107023,
|
|
"grad_norm": 7.1753668785095215,
|
|
"learning_rate": 2.9974916387959864e-05,
|
|
"loss": 4.6789,
|
|
"step": 1200
|
|
},
|
|
{
|
|
"epoch": 2.0903010033444818,
|
|
"grad_norm": 7.401775360107422,
|
|
"learning_rate": 2.9661371237458197e-05,
|
|
"loss": 4.2448,
|
|
"step": 1250
|
|
},
|
|
{
|
|
"epoch": 2.1739130434782608,
|
|
"grad_norm": 7.254606246948242,
|
|
"learning_rate": 2.9347826086956523e-05,
|
|
"loss": 4.0163,
|
|
"step": 1300
|
|
},
|
|
{
|
|
"epoch": 2.25752508361204,
|
|
"grad_norm": 7.533504962921143,
|
|
"learning_rate": 2.903428093645485e-05,
|
|
"loss": 3.8083,
|
|
"step": 1350
|
|
},
|
|
{
|
|
"epoch": 2.3411371237458196,
|
|
"grad_norm": 6.658289432525635,
|
|
"learning_rate": 2.8720735785953178e-05,
|
|
"loss": 3.5772,
|
|
"step": 1400
|
|
},
|
|
{
|
|
"epoch": 2.4247491638795986,
|
|
"grad_norm": 9.382271766662598,
|
|
"learning_rate": 2.8407190635451504e-05,
|
|
"loss": 3.4262,
|
|
"step": 1450
|
|
},
|
|
{
|
|
"epoch": 2.508361204013378,
|
|
"grad_norm": 6.964470863342285,
|
|
"learning_rate": 2.8093645484949833e-05,
|
|
"loss": 3.2302,
|
|
"step": 1500
|
|
},
|
|
{
|
|
"epoch": 2.591973244147157,
|
|
"grad_norm": 6.510575294494629,
|
|
"learning_rate": 2.7780100334448163e-05,
|
|
"loss": 3.0355,
|
|
"step": 1550
|
|
},
|
|
{
|
|
"epoch": 2.6755852842809364,
|
|
"grad_norm": 6.38152551651001,
|
|
"learning_rate": 2.746655518394649e-05,
|
|
"loss": 2.8789,
|
|
"step": 1600
|
|
},
|
|
{
|
|
"epoch": 2.759197324414716,
|
|
"grad_norm": 6.31973934173584,
|
|
"learning_rate": 2.7153010033444814e-05,
|
|
"loss": 2.6744,
|
|
"step": 1650
|
|
},
|
|
{
|
|
"epoch": 2.842809364548495,
|
|
"grad_norm": 8.131747245788574,
|
|
"learning_rate": 2.6839464882943144e-05,
|
|
"loss": 2.5437,
|
|
"step": 1700
|
|
},
|
|
{
|
|
"epoch": 2.9264214046822743,
|
|
"grad_norm": 6.337752342224121,
|
|
"learning_rate": 2.6525919732441473e-05,
|
|
"loss": 2.3661,
|
|
"step": 1750
|
|
},
|
|
{
|
|
"epoch": 3.0,
|
|
"eval_accuracy": 0.735785961151123,
|
|
"eval_f1": 0.735785961151123,
|
|
"eval_loss": 1.8872631788253784,
|
|
"eval_runtime": 58.5991,
|
|
"eval_samples_per_second": 81.64,
|
|
"eval_steps_per_second": 1.28,
|
|
"step": 1794
|
|
},
|
|
{
|
|
"epoch": 3.0100334448160537,
|
|
"grad_norm": 6.063199520111084,
|
|
"learning_rate": 2.6212374581939803e-05,
|
|
"loss": 2.2655,
|
|
"step": 1800
|
|
},
|
|
{
|
|
"epoch": 3.0936454849498327,
|
|
"grad_norm": 5.415772914886475,
|
|
"learning_rate": 2.589882943143813e-05,
|
|
"loss": 1.9437,
|
|
"step": 1850
|
|
},
|
|
{
|
|
"epoch": 3.177257525083612,
|
|
"grad_norm": 5.352977752685547,
|
|
"learning_rate": 2.5585284280936454e-05,
|
|
"loss": 1.8229,
|
|
"step": 1900
|
|
},
|
|
{
|
|
"epoch": 3.260869565217391,
|
|
"grad_norm": 6.909513473510742,
|
|
"learning_rate": 2.5271739130434784e-05,
|
|
"loss": 1.7459,
|
|
"step": 1950
|
|
},
|
|
{
|
|
"epoch": 3.3444816053511706,
|
|
"grad_norm": 5.768405914306641,
|
|
"learning_rate": 2.495819397993311e-05,
|
|
"loss": 1.6214,
|
|
"step": 2000
|
|
},
|
|
{
|
|
"epoch": 3.42809364548495,
|
|
"grad_norm": 4.940030574798584,
|
|
"learning_rate": 2.464464882943144e-05,
|
|
"loss": 1.5412,
|
|
"step": 2050
|
|
},
|
|
{
|
|
"epoch": 3.511705685618729,
|
|
"grad_norm": 5.3233323097229,
|
|
"learning_rate": 2.433110367892977e-05,
|
|
"loss": 1.4467,
|
|
"step": 2100
|
|
},
|
|
{
|
|
"epoch": 3.5953177257525084,
|
|
"grad_norm": 5.229661464691162,
|
|
"learning_rate": 2.4017558528428094e-05,
|
|
"loss": 1.3918,
|
|
"step": 2150
|
|
},
|
|
{
|
|
"epoch": 3.678929765886288,
|
|
"grad_norm": 4.888563632965088,
|
|
"learning_rate": 2.370401337792642e-05,
|
|
"loss": 1.3623,
|
|
"step": 2200
|
|
},
|
|
{
|
|
"epoch": 3.762541806020067,
|
|
"grad_norm": 4.426875591278076,
|
|
"learning_rate": 2.339046822742475e-05,
|
|
"loss": 1.2494,
|
|
"step": 2250
|
|
},
|
|
{
|
|
"epoch": 3.8461538461538463,
|
|
"grad_norm": 5.58463191986084,
|
|
"learning_rate": 2.307692307692308e-05,
|
|
"loss": 1.2034,
|
|
"step": 2300
|
|
},
|
|
{
|
|
"epoch": 3.9297658862876252,
|
|
"grad_norm": 4.331861972808838,
|
|
"learning_rate": 2.2763377926421405e-05,
|
|
"loss": 1.1543,
|
|
"step": 2350
|
|
},
|
|
{
|
|
"epoch": 4.0,
|
|
"eval_accuracy": 0.7759197354316711,
|
|
"eval_f1": 0.7759197354316711,
|
|
"eval_loss": 0.9816091060638428,
|
|
"eval_runtime": 56.0644,
|
|
"eval_samples_per_second": 85.33,
|
|
"eval_steps_per_second": 1.338,
|
|
"step": 2392
|
|
},
|
|
{
|
|
"epoch": 4.013377926421405,
|
|
"grad_norm": 4.542367935180664,
|
|
"learning_rate": 2.2449832775919734e-05,
|
|
"loss": 1.0647,
|
|
"step": 2400
|
|
},
|
|
{
|
|
"epoch": 4.096989966555184,
|
|
"grad_norm": 4.851340293884277,
|
|
"learning_rate": 2.213628762541806e-05,
|
|
"loss": 0.9377,
|
|
"step": 2450
|
|
},
|
|
{
|
|
"epoch": 4.1806020066889635,
|
|
"grad_norm": 4.438965797424316,
|
|
"learning_rate": 2.1822742474916386e-05,
|
|
"loss": 0.9529,
|
|
"step": 2500
|
|
},
|
|
{
|
|
"epoch": 4.264214046822742,
|
|
"grad_norm": 4.728175163269043,
|
|
"learning_rate": 2.150919732441472e-05,
|
|
"loss": 0.8806,
|
|
"step": 2550
|
|
},
|
|
{
|
|
"epoch": 4.3478260869565215,
|
|
"grad_norm": 4.2397284507751465,
|
|
"learning_rate": 2.1195652173913045e-05,
|
|
"loss": 0.8575,
|
|
"step": 2600
|
|
},
|
|
{
|
|
"epoch": 4.431438127090301,
|
|
"grad_norm": 3.8949732780456543,
|
|
"learning_rate": 2.088210702341137e-05,
|
|
"loss": 0.8447,
|
|
"step": 2650
|
|
},
|
|
{
|
|
"epoch": 4.51505016722408,
|
|
"grad_norm": 4.523480415344238,
|
|
"learning_rate": 2.05685618729097e-05,
|
|
"loss": 0.8167,
|
|
"step": 2700
|
|
},
|
|
{
|
|
"epoch": 4.59866220735786,
|
|
"grad_norm": 5.011022090911865,
|
|
"learning_rate": 2.0255016722408026e-05,
|
|
"loss": 0.7993,
|
|
"step": 2750
|
|
},
|
|
{
|
|
"epoch": 4.682274247491639,
|
|
"grad_norm": 4.6291399002075195,
|
|
"learning_rate": 1.9941471571906352e-05,
|
|
"loss": 0.7959,
|
|
"step": 2800
|
|
},
|
|
{
|
|
"epoch": 4.765886287625418,
|
|
"grad_norm": 4.647440433502197,
|
|
"learning_rate": 1.9627926421404685e-05,
|
|
"loss": 0.7582,
|
|
"step": 2850
|
|
},
|
|
{
|
|
"epoch": 4.849498327759197,
|
|
"grad_norm": 3.1524765491485596,
|
|
"learning_rate": 1.931438127090301e-05,
|
|
"loss": 0.7298,
|
|
"step": 2900
|
|
},
|
|
{
|
|
"epoch": 4.933110367892977,
|
|
"grad_norm": 5.388834476470947,
|
|
"learning_rate": 1.9000836120401337e-05,
|
|
"loss": 0.6885,
|
|
"step": 2950
|
|
},
|
|
{
|
|
"epoch": 5.0,
|
|
"eval_accuracy": 0.8018394708633423,
|
|
"eval_f1": 0.8018394708633423,
|
|
"eval_loss": 0.6958563327789307,
|
|
"eval_runtime": 57.0066,
|
|
"eval_samples_per_second": 83.92,
|
|
"eval_steps_per_second": 1.316,
|
|
"step": 2990
|
|
},
|
|
{
|
|
"epoch": 5.016722408026756,
|
|
"grad_norm": 2.930847644805908,
|
|
"learning_rate": 1.8687290969899666e-05,
|
|
"loss": 0.6759,
|
|
"step": 3000
|
|
},
|
|
{
|
|
"epoch": 5.1003344481605355,
|
|
"grad_norm": 4.471439361572266,
|
|
"learning_rate": 1.8373745819397992e-05,
|
|
"loss": 0.6077,
|
|
"step": 3050
|
|
},
|
|
{
|
|
"epoch": 5.183946488294314,
|
|
"grad_norm": 3.0460543632507324,
|
|
"learning_rate": 1.806020066889632e-05,
|
|
"loss": 0.6214,
|
|
"step": 3100
|
|
},
|
|
{
|
|
"epoch": 5.2675585284280935,
|
|
"grad_norm": 3.273836851119995,
|
|
"learning_rate": 1.774665551839465e-05,
|
|
"loss": 0.6055,
|
|
"step": 3150
|
|
},
|
|
{
|
|
"epoch": 5.351170568561873,
|
|
"grad_norm": 3.214381456375122,
|
|
"learning_rate": 1.7433110367892977e-05,
|
|
"loss": 0.5718,
|
|
"step": 3200
|
|
},
|
|
{
|
|
"epoch": 5.434782608695652,
|
|
"grad_norm": 2.87959361076355,
|
|
"learning_rate": 1.7119565217391303e-05,
|
|
"loss": 0.565,
|
|
"step": 3250
|
|
},
|
|
{
|
|
"epoch": 5.518394648829432,
|
|
"grad_norm": 4.1827521324157715,
|
|
"learning_rate": 1.6806020066889632e-05,
|
|
"loss": 0.5923,
|
|
"step": 3300
|
|
},
|
|
{
|
|
"epoch": 5.602006688963211,
|
|
"grad_norm": 2.584045886993408,
|
|
"learning_rate": 1.649247491638796e-05,
|
|
"loss": 0.5553,
|
|
"step": 3350
|
|
},
|
|
{
|
|
"epoch": 5.68561872909699,
|
|
"grad_norm": 2.441519260406494,
|
|
"learning_rate": 1.6178929765886288e-05,
|
|
"loss": 0.557,
|
|
"step": 3400
|
|
},
|
|
{
|
|
"epoch": 5.769230769230769,
|
|
"grad_norm": 3.489903688430786,
|
|
"learning_rate": 1.5865384615384617e-05,
|
|
"loss": 0.5212,
|
|
"step": 3450
|
|
},
|
|
{
|
|
"epoch": 5.852842809364549,
|
|
"grad_norm": 3.7490415573120117,
|
|
"learning_rate": 1.5551839464882943e-05,
|
|
"loss": 0.5246,
|
|
"step": 3500
|
|
},
|
|
{
|
|
"epoch": 5.936454849498328,
|
|
"grad_norm": 3.0013935565948486,
|
|
"learning_rate": 1.523829431438127e-05,
|
|
"loss": 0.5341,
|
|
"step": 3550
|
|
},
|
|
{
|
|
"epoch": 6.0,
|
|
"eval_accuracy": 0.8160535097122192,
|
|
"eval_f1": 0.8160535097122192,
|
|
"eval_loss": 0.610401451587677,
|
|
"eval_runtime": 57.2156,
|
|
"eval_samples_per_second": 83.614,
|
|
"eval_steps_per_second": 1.311,
|
|
"step": 3588
|
|
},
|
|
{
|
|
"epoch": 6.0200668896321075,
|
|
"grad_norm": 3.0697216987609863,
|
|
"learning_rate": 1.49247491638796e-05,
|
|
"loss": 0.5124,
|
|
"step": 3600
|
|
},
|
|
{
|
|
"epoch": 6.103678929765886,
|
|
"grad_norm": 3.071810722351074,
|
|
"learning_rate": 1.4611204013377926e-05,
|
|
"loss": 0.4665,
|
|
"step": 3650
|
|
},
|
|
{
|
|
"epoch": 6.187290969899665,
|
|
"grad_norm": 2.360440969467163,
|
|
"learning_rate": 1.4297658862876255e-05,
|
|
"loss": 0.4523,
|
|
"step": 3700
|
|
},
|
|
{
|
|
"epoch": 6.270903010033445,
|
|
"grad_norm": 3.7450270652770996,
|
|
"learning_rate": 1.3984113712374583e-05,
|
|
"loss": 0.4557,
|
|
"step": 3750
|
|
},
|
|
{
|
|
"epoch": 6.354515050167224,
|
|
"grad_norm": 2.299044370651245,
|
|
"learning_rate": 1.3670568561872909e-05,
|
|
"loss": 0.4486,
|
|
"step": 3800
|
|
},
|
|
{
|
|
"epoch": 6.438127090301004,
|
|
"grad_norm": 3.2113373279571533,
|
|
"learning_rate": 1.3357023411371238e-05,
|
|
"loss": 0.4441,
|
|
"step": 3850
|
|
},
|
|
{
|
|
"epoch": 6.521739130434782,
|
|
"grad_norm": 3.265704393386841,
|
|
"learning_rate": 1.3043478260869566e-05,
|
|
"loss": 0.4476,
|
|
"step": 3900
|
|
},
|
|
{
|
|
"epoch": 6.605351170568562,
|
|
"grad_norm": 2.365684747695923,
|
|
"learning_rate": 1.2729933110367893e-05,
|
|
"loss": 0.4326,
|
|
"step": 3950
|
|
},
|
|
{
|
|
"epoch": 6.688963210702341,
|
|
"grad_norm": 2.623091697692871,
|
|
"learning_rate": 1.2416387959866221e-05,
|
|
"loss": 0.4493,
|
|
"step": 4000
|
|
},
|
|
{
|
|
"epoch": 6.7725752508361206,
|
|
"grad_norm": 4.2816362380981445,
|
|
"learning_rate": 1.2102842809364549e-05,
|
|
"loss": 0.4374,
|
|
"step": 4050
|
|
},
|
|
{
|
|
"epoch": 6.8561872909699,
|
|
"grad_norm": 2.976745367050171,
|
|
"learning_rate": 1.1789297658862876e-05,
|
|
"loss": 0.4246,
|
|
"step": 4100
|
|
},
|
|
{
|
|
"epoch": 6.9397993311036785,
|
|
"grad_norm": 2.723569631576538,
|
|
"learning_rate": 1.1475752508361204e-05,
|
|
"loss": 0.4104,
|
|
"step": 4150
|
|
},
|
|
{
|
|
"epoch": 7.0,
|
|
"eval_accuracy": 0.8235785961151123,
|
|
"eval_f1": 0.8235785961151123,
|
|
"eval_loss": 0.5483660697937012,
|
|
"eval_runtime": 56.6001,
|
|
"eval_samples_per_second": 84.523,
|
|
"eval_steps_per_second": 1.325,
|
|
"step": 4186
|
|
},
|
|
{
|
|
"epoch": 7.023411371237458,
|
|
"grad_norm": 2.4986026287078857,
|
|
"learning_rate": 1.1162207357859532e-05,
|
|
"loss": 0.413,
|
|
"step": 4200
|
|
},
|
|
{
|
|
"epoch": 7.107023411371237,
|
|
"grad_norm": 3.6669883728027344,
|
|
"learning_rate": 1.084866220735786e-05,
|
|
"loss": 0.3636,
|
|
"step": 4250
|
|
},
|
|
{
|
|
"epoch": 7.190635451505017,
|
|
"grad_norm": 2.7657742500305176,
|
|
"learning_rate": 1.0535117056856187e-05,
|
|
"loss": 0.3762,
|
|
"step": 4300
|
|
},
|
|
{
|
|
"epoch": 7.274247491638796,
|
|
"grad_norm": 2.6917202472686768,
|
|
"learning_rate": 1.0221571906354516e-05,
|
|
"loss": 0.3688,
|
|
"step": 4350
|
|
},
|
|
{
|
|
"epoch": 7.357859531772576,
|
|
"grad_norm": 2.6124022006988525,
|
|
"learning_rate": 9.908026755852842e-06,
|
|
"loss": 0.3534,
|
|
"step": 4400
|
|
},
|
|
{
|
|
"epoch": 7.441471571906354,
|
|
"grad_norm": 2.3492910861968994,
|
|
"learning_rate": 9.59448160535117e-06,
|
|
"loss": 0.3648,
|
|
"step": 4450
|
|
},
|
|
{
|
|
"epoch": 7.525083612040134,
|
|
"grad_norm": 1.842259168624878,
|
|
"learning_rate": 9.2809364548495e-06,
|
|
"loss": 0.3457,
|
|
"step": 4500
|
|
},
|
|
{
|
|
"epoch": 7.608695652173913,
|
|
"grad_norm": 2.880143404006958,
|
|
"learning_rate": 8.967391304347827e-06,
|
|
"loss": 0.3747,
|
|
"step": 4550
|
|
},
|
|
{
|
|
"epoch": 7.6923076923076925,
|
|
"grad_norm": 2.546765089035034,
|
|
"learning_rate": 8.653846153846153e-06,
|
|
"loss": 0.349,
|
|
"step": 4600
|
|
},
|
|
{
|
|
"epoch": 7.775919732441472,
|
|
"grad_norm": 2.572425127029419,
|
|
"learning_rate": 8.340301003344482e-06,
|
|
"loss": 0.3479,
|
|
"step": 4650
|
|
},
|
|
{
|
|
"epoch": 7.8595317725752505,
|
|
"grad_norm": 3.774538278579712,
|
|
"learning_rate": 8.02675585284281e-06,
|
|
"loss": 0.3458,
|
|
"step": 4700
|
|
},
|
|
{
|
|
"epoch": 7.94314381270903,
|
|
"grad_norm": 2.3983170986175537,
|
|
"learning_rate": 7.713210702341138e-06,
|
|
"loss": 0.3549,
|
|
"step": 4750
|
|
},
|
|
{
|
|
"epoch": 8.0,
|
|
"eval_accuracy": 0.8317307829856873,
|
|
"eval_f1": 0.8317307829856873,
|
|
"eval_loss": 0.5173951983451843,
|
|
"eval_runtime": 57.872,
|
|
"eval_samples_per_second": 82.665,
|
|
"eval_steps_per_second": 1.296,
|
|
"step": 4784
|
|
},
|
|
{
|
|
"epoch": 8.02675585284281,
|
|
"grad_norm": 2.7807955741882324,
|
|
"learning_rate": 7.399665551839465e-06,
|
|
"loss": 0.3324,
|
|
"step": 4800
|
|
},
|
|
{
|
|
"epoch": 8.110367892976589,
|
|
"grad_norm": 3.530358076095581,
|
|
"learning_rate": 7.086120401337793e-06,
|
|
"loss": 0.3212,
|
|
"step": 4850
|
|
},
|
|
{
|
|
"epoch": 8.193979933110368,
|
|
"grad_norm": 2.1531357765197754,
|
|
"learning_rate": 6.772575250836121e-06,
|
|
"loss": 0.3141,
|
|
"step": 4900
|
|
},
|
|
{
|
|
"epoch": 8.277591973244148,
|
|
"grad_norm": 1.996618628501892,
|
|
"learning_rate": 6.459030100334448e-06,
|
|
"loss": 0.288,
|
|
"step": 4950
|
|
},
|
|
{
|
|
"epoch": 8.361204013377927,
|
|
"grad_norm": 2.6897034645080566,
|
|
"learning_rate": 6.145484949832776e-06,
|
|
"loss": 0.2943,
|
|
"step": 5000
|
|
},
|
|
{
|
|
"epoch": 8.444816053511706,
|
|
"grad_norm": 2.446686267852783,
|
|
"learning_rate": 5.831939799331104e-06,
|
|
"loss": 0.308,
|
|
"step": 5050
|
|
},
|
|
{
|
|
"epoch": 8.528428093645484,
|
|
"grad_norm": 2.7985997200012207,
|
|
"learning_rate": 5.518394648829431e-06,
|
|
"loss": 0.3014,
|
|
"step": 5100
|
|
},
|
|
{
|
|
"epoch": 8.612040133779264,
|
|
"grad_norm": 2.4453961849212646,
|
|
"learning_rate": 5.204849498327759e-06,
|
|
"loss": 0.2996,
|
|
"step": 5150
|
|
},
|
|
{
|
|
"epoch": 8.695652173913043,
|
|
"grad_norm": 3.4949986934661865,
|
|
"learning_rate": 4.891304347826087e-06,
|
|
"loss": 0.3024,
|
|
"step": 5200
|
|
},
|
|
{
|
|
"epoch": 8.779264214046822,
|
|
"grad_norm": 3.0933334827423096,
|
|
"learning_rate": 4.577759197324415e-06,
|
|
"loss": 0.318,
|
|
"step": 5250
|
|
},
|
|
{
|
|
"epoch": 8.862876254180602,
|
|
"grad_norm": 2.353951930999756,
|
|
"learning_rate": 4.264214046822742e-06,
|
|
"loss": 0.3042,
|
|
"step": 5300
|
|
},
|
|
{
|
|
"epoch": 8.946488294314381,
|
|
"grad_norm": 3.0331270694732666,
|
|
"learning_rate": 3.95066889632107e-06,
|
|
"loss": 0.299,
|
|
"step": 5350
|
|
},
|
|
{
|
|
"epoch": 9.0,
|
|
"eval_accuracy": 0.8377926349639893,
|
|
"eval_f1": 0.8377926349639893,
|
|
"eval_loss": 0.4919103980064392,
|
|
"eval_runtime": 57.5334,
|
|
"eval_samples_per_second": 83.152,
|
|
"eval_steps_per_second": 1.304,
|
|
"step": 5382
|
|
},
|
|
{
|
|
"epoch": 9.03010033444816,
|
|
"grad_norm": 1.635090947151184,
|
|
"learning_rate": 3.637123745819398e-06,
|
|
"loss": 0.27,
|
|
"step": 5400
|
|
},
|
|
{
|
|
"epoch": 9.11371237458194,
|
|
"grad_norm": 2.002584934234619,
|
|
"learning_rate": 3.3235785953177257e-06,
|
|
"loss": 0.2604,
|
|
"step": 5450
|
|
},
|
|
{
|
|
"epoch": 9.19732441471572,
|
|
"grad_norm": 2.2020089626312256,
|
|
"learning_rate": 3.0100334448160537e-06,
|
|
"loss": 0.2643,
|
|
"step": 5500
|
|
},
|
|
{
|
|
"epoch": 9.280936454849499,
|
|
"grad_norm": 2.3574795722961426,
|
|
"learning_rate": 2.6964882943143814e-06,
|
|
"loss": 0.2754,
|
|
"step": 5550
|
|
},
|
|
{
|
|
"epoch": 9.364548494983278,
|
|
"grad_norm": 2.37418794631958,
|
|
"learning_rate": 2.382943143812709e-06,
|
|
"loss": 0.2833,
|
|
"step": 5600
|
|
},
|
|
{
|
|
"epoch": 9.448160535117056,
|
|
"grad_norm": 2.4970650672912598,
|
|
"learning_rate": 2.0693979933110367e-06,
|
|
"loss": 0.2598,
|
|
"step": 5650
|
|
},
|
|
{
|
|
"epoch": 9.531772575250836,
|
|
"grad_norm": 2.6536877155303955,
|
|
"learning_rate": 1.7558528428093646e-06,
|
|
"loss": 0.2685,
|
|
"step": 5700
|
|
},
|
|
{
|
|
"epoch": 9.615384615384615,
|
|
"grad_norm": 2.722958564758301,
|
|
"learning_rate": 1.4423076923076924e-06,
|
|
"loss": 0.2588,
|
|
"step": 5750
|
|
},
|
|
{
|
|
"epoch": 9.698996655518394,
|
|
"grad_norm": 2.4485061168670654,
|
|
"learning_rate": 1.12876254180602e-06,
|
|
"loss": 0.2728,
|
|
"step": 5800
|
|
},
|
|
{
|
|
"epoch": 9.782608695652174,
|
|
"grad_norm": 2.3278067111968994,
|
|
"learning_rate": 8.152173913043479e-07,
|
|
"loss": 0.2752,
|
|
"step": 5850
|
|
},
|
|
{
|
|
"epoch": 9.866220735785953,
|
|
"grad_norm": 1.9324153661727905,
|
|
"learning_rate": 5.016722408026756e-07,
|
|
"loss": 0.2604,
|
|
"step": 5900
|
|
},
|
|
{
|
|
"epoch": 9.949832775919733,
|
|
"grad_norm": 1.9075779914855957,
|
|
"learning_rate": 1.8812709030100336e-07,
|
|
"loss": 0.2599,
|
|
"step": 5950
|
|
},
|
|
{
|
|
"epoch": 10.0,
|
|
"eval_accuracy": 0.839046835899353,
|
|
"eval_f1": 0.839046835899353,
|
|
"eval_loss": 0.49083635210990906,
|
|
"eval_runtime": 59.1954,
|
|
"eval_samples_per_second": 80.817,
|
|
"eval_steps_per_second": 1.267,
|
|
"step": 5980
|
|
}
|
|
],
|
|
"logging_steps": 50,
|
|
"max_steps": 5980,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 10,
|
|
"save_steps": 500,
|
|
"stateful_callbacks": {
|
|
"EarlyStoppingCallback": {
|
|
"args": {
|
|
"early_stopping_patience": 3,
|
|
"early_stopping_threshold": 0.0
|
|
},
|
|
"attributes": {
|
|
"early_stopping_patience_counter": 0
|
|
}
|
|
},
|
|
"TrainerControl": {
|
|
"args": {
|
|
"should_epoch_stop": false,
|
|
"should_evaluate": false,
|
|
"should_log": false,
|
|
"should_save": true,
|
|
"should_training_stop": true
|
|
},
|
|
"attributes": {}
|
|
}
|
|
},
|
|
"total_flos": 1.6022352226511616e+16,
|
|
"train_batch_size": 64,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|
|
|