josecannete's picture
adding model finetuned on QA (MLQA)
6072f94
raw
history blame
10.5 kB
{
"best_metric": 71.09356770617994,
"best_model_checkpoint": "/data/jcanete/all_results/mlqa/beto_cased/epochs_3_bs_16_lr_3e-5/checkpoint-8100",
"epoch": 3.0,
"global_step": 15390,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.06,
"eval_exact_match": 34.8,
"eval_f1": 58.03970816447246,
"step": 300
},
{
"epoch": 0.1,
"learning_rate": 2.9031189083820663e-05,
"loss": 2.7338,
"step": 500
},
{
"epoch": 0.12,
"eval_exact_match": 36.0,
"eval_f1": 59.841079581289875,
"step": 600
},
{
"epoch": 0.18,
"eval_exact_match": 38.6,
"eval_f1": 64.67169204847333,
"step": 900
},
{
"epoch": 0.19,
"learning_rate": 2.8056530214424954e-05,
"loss": 2.2457,
"step": 1000
},
{
"epoch": 0.23,
"eval_exact_match": 41.0,
"eval_f1": 65.03933016022862,
"step": 1200
},
{
"epoch": 0.29,
"learning_rate": 2.708187134502924e-05,
"loss": 2.1076,
"step": 1500
},
{
"epoch": 0.29,
"eval_exact_match": 38.8,
"eval_f1": 64.6213713578301,
"step": 1500
},
{
"epoch": 0.35,
"eval_exact_match": 41.2,
"eval_f1": 65.84064101398657,
"step": 1800
},
{
"epoch": 0.39,
"learning_rate": 2.6107212475633526e-05,
"loss": 2.0038,
"step": 2000
},
{
"epoch": 0.41,
"eval_exact_match": 41.4,
"eval_f1": 66.8343118103736,
"step": 2100
},
{
"epoch": 0.47,
"eval_exact_match": 44.2,
"eval_f1": 68.74877627633202,
"step": 2400
},
{
"epoch": 0.49,
"learning_rate": 2.5132553606237817e-05,
"loss": 1.9719,
"step": 2500
},
{
"epoch": 0.53,
"eval_exact_match": 41.2,
"eval_f1": 67.75205005784734,
"step": 2700
},
{
"epoch": 0.58,
"learning_rate": 2.4157894736842104e-05,
"loss": 1.9103,
"step": 3000
},
{
"epoch": 0.58,
"eval_exact_match": 44.4,
"eval_f1": 69.00308434085002,
"step": 3000
},
{
"epoch": 0.64,
"eval_exact_match": 41.2,
"eval_f1": 66.19132807246592,
"step": 3300
},
{
"epoch": 0.68,
"learning_rate": 2.318323586744639e-05,
"loss": 1.899,
"step": 3500
},
{
"epoch": 0.7,
"eval_exact_match": 43.6,
"eval_f1": 67.6086413268844,
"step": 3600
},
{
"epoch": 0.76,
"eval_exact_match": 40.0,
"eval_f1": 66.87548187227003,
"step": 3900
},
{
"epoch": 0.78,
"learning_rate": 2.2208576998050682e-05,
"loss": 1.9084,
"step": 4000
},
{
"epoch": 0.82,
"eval_exact_match": 42.4,
"eval_f1": 68.43203629031083,
"step": 4200
},
{
"epoch": 0.88,
"learning_rate": 2.123586744639376e-05,
"loss": 1.8575,
"step": 4500
},
{
"epoch": 0.88,
"eval_exact_match": 41.8,
"eval_f1": 66.57412118606702,
"step": 4500
},
{
"epoch": 0.94,
"eval_exact_match": 42.6,
"eval_f1": 69.08327662197658,
"step": 4800
},
{
"epoch": 0.97,
"learning_rate": 2.026120857699805e-05,
"loss": 1.8016,
"step": 5000
},
{
"epoch": 0.99,
"eval_exact_match": 43.8,
"eval_f1": 69.83716582639477,
"step": 5100
},
{
"epoch": 1.05,
"eval_exact_match": 43.8,
"eval_f1": 69.87892753163113,
"step": 5400
},
{
"epoch": 1.07,
"learning_rate": 1.928654970760234e-05,
"loss": 1.6069,
"step": 5500
},
{
"epoch": 1.11,
"eval_exact_match": 42.2,
"eval_f1": 68.22307993614402,
"step": 5700
},
{
"epoch": 1.17,
"learning_rate": 1.8311890838206626e-05,
"loss": 1.4879,
"step": 6000
},
{
"epoch": 1.17,
"eval_exact_match": 43.2,
"eval_f1": 68.58203389587109,
"step": 6000
},
{
"epoch": 1.23,
"eval_exact_match": 43.2,
"eval_f1": 68.83200756284295,
"step": 6300
},
{
"epoch": 1.27,
"learning_rate": 1.7339181286549707e-05,
"loss": 1.4833,
"step": 6500
},
{
"epoch": 1.29,
"eval_exact_match": 43.6,
"eval_f1": 68.2076220533459,
"step": 6600
},
{
"epoch": 1.35,
"eval_exact_match": 43.4,
"eval_f1": 68.79536487676586,
"step": 6900
},
{
"epoch": 1.36,
"learning_rate": 1.6364522417153995e-05,
"loss": 1.4904,
"step": 7000
},
{
"epoch": 1.4,
"eval_exact_match": 42.4,
"eval_f1": 69.97669950942714,
"step": 7200
},
{
"epoch": 1.46,
"learning_rate": 1.5389863547758286e-05,
"loss": 1.4762,
"step": 7500
},
{
"epoch": 1.46,
"eval_exact_match": 43.8,
"eval_f1": 69.861214423741,
"step": 7500
},
{
"epoch": 1.52,
"eval_exact_match": 43.0,
"eval_f1": 69.21209743481882,
"step": 7800
},
{
"epoch": 1.56,
"learning_rate": 1.4415204678362573e-05,
"loss": 1.4892,
"step": 8000
},
{
"epoch": 1.58,
"eval_exact_match": 44.6,
"eval_f1": 71.09356770617994,
"step": 8100
},
{
"epoch": 1.64,
"eval_exact_match": 43.4,
"eval_f1": 70.64992293047675,
"step": 8400
},
{
"epoch": 1.66,
"learning_rate": 1.3442495126705655e-05,
"loss": 1.4634,
"step": 8500
},
{
"epoch": 1.7,
"eval_exact_match": 43.8,
"eval_f1": 68.71836001461945,
"step": 8700
},
{
"epoch": 1.75,
"learning_rate": 1.2467836257309942e-05,
"loss": 1.4689,
"step": 9000
},
{
"epoch": 1.75,
"eval_exact_match": 43.2,
"eval_f1": 69.3302964310732,
"step": 9000
},
{
"epoch": 1.81,
"eval_exact_match": 43.0,
"eval_f1": 68.57716516263318,
"step": 9300
},
{
"epoch": 1.85,
"learning_rate": 1.1493177387914231e-05,
"loss": 1.4355,
"step": 9500
},
{
"epoch": 1.87,
"eval_exact_match": 42.6,
"eval_f1": 67.68193670391558,
"step": 9600
},
{
"epoch": 1.93,
"eval_exact_match": 43.4,
"eval_f1": 68.5709862176482,
"step": 9900
},
{
"epoch": 1.95,
"learning_rate": 1.051851851851852e-05,
"loss": 1.4671,
"step": 10000
},
{
"epoch": 1.99,
"eval_exact_match": 42.2,
"eval_f1": 67.86405792646175,
"step": 10200
},
{
"epoch": 2.05,
"learning_rate": 9.543859649122808e-06,
"loss": 1.3096,
"step": 10500
},
{
"epoch": 2.05,
"eval_exact_match": 42.6,
"eval_f1": 68.10115237113098,
"step": 10500
},
{
"epoch": 2.11,
"eval_exact_match": 42.6,
"eval_f1": 68.4050539633749,
"step": 10800
},
{
"epoch": 2.14,
"learning_rate": 8.571150097465888e-06,
"loss": 1.1729,
"step": 11000
},
{
"epoch": 2.16,
"eval_exact_match": 42.8,
"eval_f1": 69.14531712828095,
"step": 11100
},
{
"epoch": 2.22,
"eval_exact_match": 42.0,
"eval_f1": 67.95727828313431,
"step": 11400
},
{
"epoch": 2.24,
"learning_rate": 7.596491228070175e-06,
"loss": 1.1765,
"step": 11500
},
{
"epoch": 2.28,
"eval_exact_match": 42.6,
"eval_f1": 68.51480513144087,
"step": 11700
},
{
"epoch": 2.34,
"learning_rate": 6.621832358674464e-06,
"loss": 1.1663,
"step": 12000
},
{
"epoch": 2.34,
"eval_exact_match": 42.6,
"eval_f1": 68.06559800304674,
"step": 12000
},
{
"epoch": 2.4,
"eval_exact_match": 42.4,
"eval_f1": 68.1790742027127,
"step": 12300
},
{
"epoch": 2.44,
"learning_rate": 5.647173489278752e-06,
"loss": 1.1738,
"step": 12500
},
{
"epoch": 2.46,
"eval_exact_match": 42.4,
"eval_f1": 68.06432275532964,
"step": 12600
},
{
"epoch": 2.51,
"eval_exact_match": 43.4,
"eval_f1": 68.94404068079874,
"step": 12900
},
{
"epoch": 2.53,
"learning_rate": 4.674463937621833e-06,
"loss": 1.1522,
"step": 13000
},
{
"epoch": 2.57,
"eval_exact_match": 43.0,
"eval_f1": 68.0807601536333,
"step": 13200
},
{
"epoch": 2.63,
"learning_rate": 3.699805068226121e-06,
"loss": 1.1268,
"step": 13500
},
{
"epoch": 2.63,
"eval_exact_match": 43.6,
"eval_f1": 68.44401279047236,
"step": 13500
},
{
"epoch": 2.69,
"eval_exact_match": 43.4,
"eval_f1": 68.29713540149096,
"step": 13800
},
{
"epoch": 2.73,
"learning_rate": 2.7251461988304095e-06,
"loss": 1.1324,
"step": 14000
},
{
"epoch": 2.75,
"eval_exact_match": 42.8,
"eval_f1": 67.98091672699975,
"step": 14100
},
{
"epoch": 2.81,
"eval_exact_match": 43.0,
"eval_f1": 67.89319270154854,
"step": 14400
},
{
"epoch": 2.83,
"learning_rate": 1.7504873294346978e-06,
"loss": 1.1632,
"step": 14500
},
{
"epoch": 2.87,
"eval_exact_match": 43.2,
"eval_f1": 68.04489013422236,
"step": 14700
},
{
"epoch": 2.92,
"learning_rate": 7.777777777777778e-07,
"loss": 1.146,
"step": 15000
},
{
"epoch": 2.92,
"eval_exact_match": 43.4,
"eval_f1": 68.39650380908567,
"step": 15000
},
{
"epoch": 2.98,
"eval_exact_match": 43.2,
"eval_f1": 68.13623191723482,
"step": 15300
},
{
"epoch": 3.0,
"step": 15390,
"total_flos": 4.556938681553549e+16,
"train_loss": 1.556812895637894,
"train_runtime": 24717.2821,
"train_samples_per_second": 9.961,
"train_steps_per_second": 0.623
}
],
"max_steps": 15390,
"num_train_epochs": 3,
"total_flos": 4.556938681553549e+16,
"trial_name": null,
"trial_params": null
}