stefan-it's picture
model+tokenizer: add initial version
9deb821
{
"best_metric": 0.006780553143471479,
"best_model_checkpoint": "./span_marker-gelectra-large-bs16-lr5e-05-1/checkpoint-3208",
"epoch": 3.0,
"global_step": 4812,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"learning_rate": 1.037344398340249e-07,
"loss": 3.4599,
"step": 1
},
{
"epoch": 0.03,
"learning_rate": 4.564315352697096e-06,
"loss": 2.5081,
"step": 50
},
{
"epoch": 0.06,
"learning_rate": 9.647302904564317e-06,
"loss": 0.1807,
"step": 100
},
{
"epoch": 0.09,
"learning_rate": 1.4834024896265561e-05,
"loss": 0.056,
"step": 150
},
{
"epoch": 0.12,
"learning_rate": 2.0020746887966806e-05,
"loss": 0.0411,
"step": 200
},
{
"epoch": 0.16,
"learning_rate": 2.520746887966805e-05,
"loss": 0.0309,
"step": 250
},
{
"epoch": 0.19,
"learning_rate": 3.0394190871369292e-05,
"loss": 0.0239,
"step": 300
},
{
"epoch": 0.22,
"learning_rate": 3.558091286307054e-05,
"loss": 0.0192,
"step": 350
},
{
"epoch": 0.25,
"learning_rate": 4.0767634854771785e-05,
"loss": 0.0165,
"step": 400
},
{
"epoch": 0.28,
"learning_rate": 4.5954356846473026e-05,
"loss": 0.0158,
"step": 450
},
{
"epoch": 0.31,
"learning_rate": 4.9872979214780604e-05,
"loss": 0.0147,
"step": 500
},
{
"epoch": 0.34,
"learning_rate": 4.929561200923788e-05,
"loss": 0.0134,
"step": 550
},
{
"epoch": 0.37,
"learning_rate": 4.871824480369515e-05,
"loss": 0.0128,
"step": 600
},
{
"epoch": 0.41,
"learning_rate": 4.814087759815243e-05,
"loss": 0.0112,
"step": 650
},
{
"epoch": 0.44,
"learning_rate": 4.75635103926097e-05,
"loss": 0.0115,
"step": 700
},
{
"epoch": 0.47,
"learning_rate": 4.698614318706698e-05,
"loss": 0.0123,
"step": 750
},
{
"epoch": 0.5,
"learning_rate": 4.640877598152425e-05,
"loss": 0.0102,
"step": 800
},
{
"epoch": 0.53,
"learning_rate": 4.5831408775981525e-05,
"loss": 0.0089,
"step": 850
},
{
"epoch": 0.56,
"learning_rate": 4.52540415704388e-05,
"loss": 0.0104,
"step": 900
},
{
"epoch": 0.59,
"learning_rate": 4.4676674364896075e-05,
"loss": 0.0107,
"step": 950
},
{
"epoch": 0.62,
"learning_rate": 4.409930715935335e-05,
"loss": 0.01,
"step": 1000
},
{
"epoch": 0.65,
"learning_rate": 4.3521939953810626e-05,
"loss": 0.0096,
"step": 1050
},
{
"epoch": 0.69,
"learning_rate": 4.29445727482679e-05,
"loss": 0.0108,
"step": 1100
},
{
"epoch": 0.72,
"learning_rate": 4.236720554272518e-05,
"loss": 0.0093,
"step": 1150
},
{
"epoch": 0.75,
"learning_rate": 4.178983833718245e-05,
"loss": 0.0096,
"step": 1200
},
{
"epoch": 0.78,
"learning_rate": 4.121247113163972e-05,
"loss": 0.0098,
"step": 1250
},
{
"epoch": 0.81,
"learning_rate": 4.0635103926097e-05,
"loss": 0.0085,
"step": 1300
},
{
"epoch": 0.84,
"learning_rate": 4.005773672055427e-05,
"loss": 0.0079,
"step": 1350
},
{
"epoch": 0.87,
"learning_rate": 3.9480369515011554e-05,
"loss": 0.0098,
"step": 1400
},
{
"epoch": 0.9,
"learning_rate": 3.890300230946882e-05,
"loss": 0.0089,
"step": 1450
},
{
"epoch": 0.94,
"learning_rate": 3.83256351039261e-05,
"loss": 0.0095,
"step": 1500
},
{
"epoch": 0.97,
"learning_rate": 3.774826789838337e-05,
"loss": 0.008,
"step": 1550
},
{
"epoch": 1.0,
"learning_rate": 3.717090069284065e-05,
"loss": 0.0076,
"step": 1600
},
{
"epoch": 1.0,
"eval_loss": 0.0075011407025158405,
"eval_overall_accuracy": 0.9820421098120183,
"eval_overall_f1": 0.880326046683957,
"eval_overall_precision": 0.8693743139407245,
"eval_overall_recall": 0.8915572232645403,
"eval_runtime": 25.2918,
"eval_samples_per_second": 93.153,
"eval_steps_per_second": 5.852,
"step": 1604
},
{
"epoch": 1.03,
"learning_rate": 3.6593533487297924e-05,
"loss": 0.0068,
"step": 1650
},
{
"epoch": 1.06,
"learning_rate": 3.60161662817552e-05,
"loss": 0.0062,
"step": 1700
},
{
"epoch": 1.09,
"learning_rate": 3.5438799076212475e-05,
"loss": 0.0054,
"step": 1750
},
{
"epoch": 1.12,
"learning_rate": 3.486143187066975e-05,
"loss": 0.0062,
"step": 1800
},
{
"epoch": 1.15,
"learning_rate": 3.4284064665127025e-05,
"loss": 0.0063,
"step": 1850
},
{
"epoch": 1.18,
"learning_rate": 3.3706697459584294e-05,
"loss": 0.0053,
"step": 1900
},
{
"epoch": 1.22,
"learning_rate": 3.3129330254041576e-05,
"loss": 0.0056,
"step": 1950
},
{
"epoch": 1.25,
"learning_rate": 3.2551963048498845e-05,
"loss": 0.0066,
"step": 2000
},
{
"epoch": 1.28,
"learning_rate": 3.197459584295612e-05,
"loss": 0.0054,
"step": 2050
},
{
"epoch": 1.31,
"learning_rate": 3.1397228637413396e-05,
"loss": 0.0061,
"step": 2100
},
{
"epoch": 1.34,
"learning_rate": 3.081986143187067e-05,
"loss": 0.0063,
"step": 2150
},
{
"epoch": 1.37,
"learning_rate": 3.0242494226327943e-05,
"loss": 0.0062,
"step": 2200
},
{
"epoch": 1.4,
"learning_rate": 2.9665127020785222e-05,
"loss": 0.0063,
"step": 2250
},
{
"epoch": 1.43,
"learning_rate": 2.9087759815242494e-05,
"loss": 0.005,
"step": 2300
},
{
"epoch": 1.47,
"learning_rate": 2.8510392609699772e-05,
"loss": 0.0049,
"step": 2350
},
{
"epoch": 1.5,
"learning_rate": 2.793302540415704e-05,
"loss": 0.006,
"step": 2400
},
{
"epoch": 1.53,
"learning_rate": 2.735565819861432e-05,
"loss": 0.0059,
"step": 2450
},
{
"epoch": 1.56,
"learning_rate": 2.67782909930716e-05,
"loss": 0.006,
"step": 2500
},
{
"epoch": 1.59,
"learning_rate": 2.620092378752887e-05,
"loss": 0.0057,
"step": 2550
},
{
"epoch": 1.62,
"learning_rate": 2.5623556581986146e-05,
"loss": 0.0051,
"step": 2600
},
{
"epoch": 1.65,
"learning_rate": 2.5046189376443418e-05,
"loss": 0.005,
"step": 2650
},
{
"epoch": 1.68,
"learning_rate": 2.4468822170900693e-05,
"loss": 0.0056,
"step": 2700
},
{
"epoch": 1.71,
"learning_rate": 2.389145496535797e-05,
"loss": 0.0054,
"step": 2750
},
{
"epoch": 1.75,
"learning_rate": 2.3314087759815244e-05,
"loss": 0.0053,
"step": 2800
},
{
"epoch": 1.78,
"learning_rate": 2.273672055427252e-05,
"loss": 0.0056,
"step": 2850
},
{
"epoch": 1.81,
"learning_rate": 2.2159353348729795e-05,
"loss": 0.0053,
"step": 2900
},
{
"epoch": 1.84,
"learning_rate": 2.158198614318707e-05,
"loss": 0.0044,
"step": 2950
},
{
"epoch": 1.87,
"learning_rate": 2.1004618937644342e-05,
"loss": 0.0044,
"step": 3000
},
{
"epoch": 1.9,
"learning_rate": 2.0427251732101618e-05,
"loss": 0.0057,
"step": 3050
},
{
"epoch": 1.93,
"learning_rate": 1.9849884526558893e-05,
"loss": 0.0058,
"step": 3100
},
{
"epoch": 1.96,
"learning_rate": 1.927251732101617e-05,
"loss": 0.0056,
"step": 3150
},
{
"epoch": 2.0,
"learning_rate": 1.8695150115473444e-05,
"loss": 0.0059,
"step": 3200
},
{
"epoch": 2.0,
"eval_loss": 0.006780553143471479,
"eval_overall_accuracy": 0.983866708280316,
"eval_overall_f1": 0.8873682263732198,
"eval_overall_precision": 0.87490882567469,
"eval_overall_recall": 0.900187617260788,
"eval_runtime": 25.1096,
"eval_samples_per_second": 93.829,
"eval_steps_per_second": 5.894,
"step": 3208
},
{
"epoch": 2.03,
"learning_rate": 1.8117782909930716e-05,
"loss": 0.0037,
"step": 3250
},
{
"epoch": 2.06,
"learning_rate": 1.754041570438799e-05,
"loss": 0.0027,
"step": 3300
},
{
"epoch": 2.09,
"learning_rate": 1.6963048498845266e-05,
"loss": 0.003,
"step": 3350
},
{
"epoch": 2.12,
"learning_rate": 1.6385681293302542e-05,
"loss": 0.003,
"step": 3400
},
{
"epoch": 2.15,
"learning_rate": 1.5808314087759814e-05,
"loss": 0.003,
"step": 3450
},
{
"epoch": 2.18,
"learning_rate": 1.523094688221709e-05,
"loss": 0.0036,
"step": 3500
},
{
"epoch": 2.21,
"learning_rate": 1.4653579676674365e-05,
"loss": 0.0024,
"step": 3550
},
{
"epoch": 2.24,
"learning_rate": 1.4076212471131642e-05,
"loss": 0.0031,
"step": 3600
},
{
"epoch": 2.28,
"learning_rate": 1.3498845265588917e-05,
"loss": 0.003,
"step": 3650
},
{
"epoch": 2.31,
"learning_rate": 1.292147806004619e-05,
"loss": 0.003,
"step": 3700
},
{
"epoch": 2.34,
"learning_rate": 1.2344110854503464e-05,
"loss": 0.0027,
"step": 3750
},
{
"epoch": 2.37,
"learning_rate": 1.176674364896074e-05,
"loss": 0.0028,
"step": 3800
},
{
"epoch": 2.4,
"learning_rate": 1.1189376443418015e-05,
"loss": 0.0025,
"step": 3850
},
{
"epoch": 2.43,
"learning_rate": 1.0612009237875289e-05,
"loss": 0.0026,
"step": 3900
},
{
"epoch": 2.46,
"learning_rate": 1.0034642032332564e-05,
"loss": 0.0029,
"step": 3950
},
{
"epoch": 2.49,
"learning_rate": 9.457274826789838e-06,
"loss": 0.0025,
"step": 4000
},
{
"epoch": 2.52,
"learning_rate": 8.879907621247113e-06,
"loss": 0.0022,
"step": 4050
},
{
"epoch": 2.56,
"learning_rate": 8.302540415704389e-06,
"loss": 0.0025,
"step": 4100
},
{
"epoch": 2.59,
"learning_rate": 7.725173210161662e-06,
"loss": 0.0026,
"step": 4150
},
{
"epoch": 2.62,
"learning_rate": 7.147806004618939e-06,
"loss": 0.0026,
"step": 4200
},
{
"epoch": 2.65,
"learning_rate": 6.570438799076213e-06,
"loss": 0.0019,
"step": 4250
},
{
"epoch": 2.68,
"learning_rate": 5.993071593533488e-06,
"loss": 0.0031,
"step": 4300
},
{
"epoch": 2.71,
"learning_rate": 5.415704387990762e-06,
"loss": 0.0027,
"step": 4350
},
{
"epoch": 2.74,
"learning_rate": 4.838337182448037e-06,
"loss": 0.003,
"step": 4400
},
{
"epoch": 2.77,
"learning_rate": 4.260969976905312e-06,
"loss": 0.0023,
"step": 4450
},
{
"epoch": 2.81,
"learning_rate": 3.683602771362587e-06,
"loss": 0.0031,
"step": 4500
},
{
"epoch": 2.84,
"learning_rate": 3.1062355658198616e-06,
"loss": 0.0033,
"step": 4550
},
{
"epoch": 2.87,
"learning_rate": 2.5288683602771365e-06,
"loss": 0.0022,
"step": 4600
},
{
"epoch": 2.9,
"learning_rate": 1.951501154734411e-06,
"loss": 0.0023,
"step": 4650
},
{
"epoch": 2.93,
"learning_rate": 1.374133949191686e-06,
"loss": 0.0022,
"step": 4700
},
{
"epoch": 2.96,
"learning_rate": 7.967667436489609e-07,
"loss": 0.0024,
"step": 4750
},
{
"epoch": 2.99,
"learning_rate": 2.1939953810623557e-07,
"loss": 0.003,
"step": 4800
},
{
"epoch": 3.0,
"eval_loss": 0.007023950573056936,
"eval_overall_accuracy": 0.9848990468873791,
"eval_overall_f1": 0.8998883513211762,
"eval_overall_precision": 0.8925802879291251,
"eval_overall_recall": 0.9073170731707317,
"eval_runtime": 25.1076,
"eval_samples_per_second": 93.836,
"eval_steps_per_second": 5.895,
"step": 4812
}
],
"max_steps": 4812,
"num_train_epochs": 3,
"total_flos": 7.147260152258458e+16,
"trial_name": null,
"trial_params": null
}