{
"best_metric": 0.954646017198851,
"best_model_checkpoint": "../out/bio/ner-scibert-spanclass-dice/checkpoints/checkpoint-12150",
"epoch": 53.76106194690266,
"global_step": 12150,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.66,
"learning_rate": 9.977876106194692e-06,
"loss": 0.8423,
"step": 150
},
{
"epoch": 0.66,
"eval_loss": 0.8136124014854431,
"eval_micro_f1": 0.8708517694112636,
"eval_runtime": 8.6661,
"eval_samples_per_second": 417.257,
"step": 150
},
{
"epoch": 1.33,
"learning_rate": 9.95575221238938e-06,
"loss": 0.7429,
"step": 300
},
{
"epoch": 1.33,
"eval_loss": 0.7772685289382935,
"eval_micro_f1": 0.9084623888802797,
"eval_runtime": 8.674,
"eval_samples_per_second": 416.877,
"step": 300
},
{
"epoch": 1.99,
"learning_rate": 9.933628318584071e-06,
"loss": 0.7013,
"step": 450
},
{
"epoch": 1.99,
"eval_loss": 0.7601113319396973,
"eval_micro_f1": 0.9269911499422214,
"eval_runtime": 9.8502,
"eval_samples_per_second": 367.098,
"step": 450
},
{
"epoch": 2.65,
"learning_rate": 9.911504424778762e-06,
"loss": 0.6802,
"step": 600
},
{
"epoch": 2.65,
"eval_loss": 0.7511240243911743,
"eval_micro_f1": 0.9305862826855833,
"eval_runtime": 8.6885,
"eval_samples_per_second": 416.182,
"step": 600
},
{
"epoch": 3.32,
"learning_rate": 9.889380530973453e-06,
"loss": 0.6621,
"step": 750
},
{
"epoch": 3.32,
"eval_loss": 0.7457098364830017,
"eval_micro_f1": 0.9341814154289452,
"eval_runtime": 8.7277,
"eval_samples_per_second": 414.315,
"step": 750
},
{
"epoch": 3.98,
"learning_rate": 9.867256637168142e-06,
"loss": 0.6506,
"step": 900
},
{
"epoch": 3.98,
"eval_loss": 0.7396827340126038,
"eval_micro_f1": 0.9391592915351384,
"eval_runtime": 9.3794,
"eval_samples_per_second": 385.525,
"step": 900
},
{
"epoch": 4.65,
"learning_rate": 9.845132743362832e-06,
"loss": 0.6494,
"step": 1050
},
{
"epoch": 4.65,
"eval_loss": 0.7364537119865417,
"eval_micro_f1": 0.9363938048094755,
"eval_runtime": 9.3845,
"eval_samples_per_second": 385.315,
"step": 1050
},
{
"epoch": 5.31,
"learning_rate": 9.823008849557523e-06,
"loss": 0.6381,
"step": 1200
},
{
"epoch": 5.31,
"eval_loss": 0.7324703335762024,
"eval_micro_f1": 0.9416482295882351,
"eval_runtime": 9.4893,
"eval_samples_per_second": 381.062,
"step": 1200
},
{
"epoch": 5.97,
"learning_rate": 9.800884955752214e-06,
"loss": 0.6266,
"step": 1350
},
{
"epoch": 5.97,
"eval_loss": 0.7316550016403198,
"eval_micro_f1": 0.9347345127740777,
"eval_runtime": 9.5002,
"eval_samples_per_second": 380.623,
"step": 1350
},
{
"epoch": 6.64,
"learning_rate": 9.778761061946903e-06,
"loss": 0.626,
"step": 1500
},
{
"epoch": 6.64,
"eval_loss": 0.7285893559455872,
"eval_micro_f1": 0.9441371676413318,
"eval_runtime": 8.8611,
"eval_samples_per_second": 408.077,
"step": 1500
},
{
"epoch": 7.3,
"learning_rate": 9.756637168141593e-06,
"loss": 0.6256,
"step": 1650
},
{
"epoch": 7.3,
"eval_loss": 0.7262205481529236,
"eval_micro_f1": 0.9491150437475251,
"eval_runtime": 9.627,
"eval_samples_per_second": 375.609,
"step": 1650
},
{
"epoch": 7.96,
"learning_rate": 9.734513274336284e-06,
"loss": 0.6227,
"step": 1800
},
{
"epoch": 7.96,
"eval_loss": 0.7256479859352112,
"eval_micro_f1": 0.9485619464023926,
"eval_runtime": 9.26,
"eval_samples_per_second": 390.497,
"step": 1800
},
{
"epoch": 8.63,
"learning_rate": 9.712389380530975e-06,
"loss": 0.6212,
"step": 1950
},
{
"epoch": 8.63,
"eval_loss": 0.7240573763847351,
"eval_micro_f1": 0.9485619464023926,
"eval_runtime": 9.5068,
"eval_samples_per_second": 380.358,
"step": 1950
},
{
"epoch": 9.29,
"learning_rate": 9.690265486725664e-06,
"loss": 0.6139,
"step": 2100
},
{
"epoch": 9.29,
"eval_loss": 0.723069965839386,
"eval_micro_f1": 0.9466261056944284,
"eval_runtime": 9.5375,
"eval_samples_per_second": 379.133,
"step": 2100
},
{
"epoch": 9.96,
"learning_rate": 9.668141592920355e-06,
"loss": 0.6171,
"step": 2250
},
{
"epoch": 9.96,
"eval_loss": 0.7234218716621399,
"eval_micro_f1": 0.9496681410926577,
"eval_runtime": 9.3125,
"eval_samples_per_second": 388.295,
"step": 2250
},
{
"epoch": 10.62,
"learning_rate": 9.646017699115045e-06,
"loss": 0.6093,
"step": 2400
},
{
"epoch": 10.62,
"eval_loss": 0.7221475839614868,
"eval_micro_f1": 0.9469026543669947,
"eval_runtime": 9.5515,
"eval_samples_per_second": 378.58,
"step": 2400
},
{
"epoch": 11.28,
"learning_rate": 9.623893805309736e-06,
"loss": 0.6078,
"step": 2550
},
{
"epoch": 11.28,
"eval_loss": 0.7231661081314087,
"eval_micro_f1": 0.947179203039561,
"eval_runtime": 9.4735,
"eval_samples_per_second": 381.697,
"step": 2550
},
{
"epoch": 11.95,
"learning_rate": 9.601769911504427e-06,
"loss": 0.6122,
"step": 2700
},
{
"epoch": 11.95,
"eval_loss": 0.7217335104942322,
"eval_micro_f1": 0.9446902649864644,
"eval_runtime": 9.4638,
"eval_samples_per_second": 382.089,
"step": 2700
},
{
"epoch": 12.61,
"learning_rate": 9.579646017699116e-06,
"loss": 0.612,
"step": 2850
},
{
"epoch": 12.61,
"eval_loss": 0.7212682366371155,
"eval_micro_f1": 0.9466261056944284,
"eval_runtime": 9.4758,
"eval_samples_per_second": 381.603,
"step": 2850
},
{
"epoch": 13.27,
"learning_rate": 9.557522123893806e-06,
"loss": 0.6134,
"step": 3000
},
{
"epoch": 13.27,
"eval_loss": 0.7197656035423279,
"eval_micro_f1": 0.9474557517121274,
"eval_runtime": 9.1216,
"eval_samples_per_second": 396.422,
"step": 3000
},
{
"epoch": 13.94,
"learning_rate": 9.535398230088495e-06,
"loss": 0.607,
"step": 3150
},
{
"epoch": 13.94,
"eval_loss": 0.7211037278175354,
"eval_micro_f1": 0.9430309729510666,
"eval_runtime": 9.5003,
"eval_samples_per_second": 380.62,
"step": 3150
},
{
"epoch": 14.6,
"learning_rate": 9.513274336283188e-06,
"loss": 0.6023,
"step": 3300
},
{
"epoch": 14.6,
"eval_loss": 0.7198509573936462,
"eval_micro_f1": 0.9477323003846936,
"eval_runtime": 9.7084,
"eval_samples_per_second": 372.461,
"step": 3300
},
{
"epoch": 15.27,
"learning_rate": 9.491150442477877e-06,
"loss": 0.6084,
"step": 3450
},
{
"epoch": 15.27,
"eval_loss": 0.7201675772666931,
"eval_micro_f1": 0.9460730083492959,
"eval_runtime": 8.7893,
"eval_samples_per_second": 411.412,
"step": 3450
},
{
"epoch": 15.93,
"learning_rate": 9.469026548672568e-06,
"loss": 0.6111,
"step": 3600
},
{
"epoch": 15.93,
"eval_loss": 0.7195026874542236,
"eval_micro_f1": 0.9463495570218622,
"eval_runtime": 9.5062,
"eval_samples_per_second": 380.382,
"step": 3600
},
{
"epoch": 16.59,
"learning_rate": 9.446902654867257e-06,
"loss": 0.5994,
"step": 3750
},
{
"epoch": 16.59,
"eval_loss": 0.7199171185493469,
"eval_micro_f1": 0.9449668136590308,
"eval_runtime": 9.3322,
"eval_samples_per_second": 387.476,
"step": 3750
},
{
"epoch": 17.26,
"learning_rate": 9.424778761061947e-06,
"loss": 0.6087,
"step": 3900
},
{
"epoch": 17.26,
"eval_loss": 0.7187968492507935,
"eval_micro_f1": 0.9477323003846936,
"eval_runtime": 10.5814,
"eval_samples_per_second": 341.733,
"step": 3900
},
{
"epoch": 17.92,
"learning_rate": 9.402654867256638e-06,
"loss": 0.604,
"step": 4050
},
{
"epoch": 17.92,
"eval_loss": 0.7195943593978882,
"eval_micro_f1": 0.9435840702961992,
"eval_runtime": 9.4668,
"eval_samples_per_second": 381.967,
"step": 4050
},
{
"epoch": 18.58,
"learning_rate": 9.380530973451329e-06,
"loss": 0.6044,
"step": 4200
},
{
"epoch": 18.58,
"eval_loss": 0.718863844871521,
"eval_micro_f1": 0.9358407074643429,
"eval_runtime": 8.7769,
"eval_samples_per_second": 411.992,
"step": 4200
},
{
"epoch": 19.25,
"learning_rate": 9.358407079646018e-06,
"loss": 0.6096,
"step": 4350
},
{
"epoch": 19.25,
"eval_loss": 0.7180087566375732,
"eval_micro_f1": 0.9438606189687655,
"eval_runtime": 9.581,
"eval_samples_per_second": 377.415,
"step": 4350
},
{
"epoch": 19.91,
"learning_rate": 9.336283185840708e-06,
"loss": 0.6046,
"step": 4500
},
{
"epoch": 19.91,
"eval_loss": 0.7181408405303955,
"eval_micro_f1": 0.9457964596767295,
"eval_runtime": 9.2502,
"eval_samples_per_second": 390.908,
"step": 4500
},
{
"epoch": 20.58,
"learning_rate": 9.314159292035399e-06,
"loss": 0.6028,
"step": 4650
},
{
"epoch": 20.58,
"eval_loss": 0.7181771397590637,
"eval_micro_f1": 0.9496681410926577,
"eval_runtime": 9.4887,
"eval_samples_per_second": 381.084,
"step": 4650
},
{
"epoch": 21.24,
"learning_rate": 9.29203539823009e-06,
"loss": 0.6066,
"step": 4800
},
{
"epoch": 21.24,
"eval_loss": 0.7180553674697876,
"eval_micro_f1": 0.9496681410926577,
"eval_runtime": 9.5643,
"eval_samples_per_second": 378.074,
"step": 4800
},
{
"epoch": 21.9,
"learning_rate": 9.26991150442478e-06,
"loss": 0.5981,
"step": 4950
},
{
"epoch": 21.9,
"eval_loss": 0.7184288501739502,
"eval_micro_f1": 0.9474557517121274,
"eval_runtime": 9.5293,
"eval_samples_per_second": 379.463,
"step": 4950
},
{
"epoch": 22.57,
"learning_rate": 9.24778761061947e-06,
"loss": 0.6051,
"step": 5100
},
{
"epoch": 22.57,
"eval_loss": 0.7198731899261475,
"eval_micro_f1": 0.9460730083492959,
"eval_runtime": 9.5484,
"eval_samples_per_second": 378.701,
"step": 5100
},
{
"epoch": 23.23,
"learning_rate": 9.22566371681416e-06,
"loss": 0.606,
"step": 5250
},
{
"epoch": 23.23,
"eval_loss": 0.719587504863739,
"eval_micro_f1": 0.9480088490572599,
"eval_runtime": 9.1658,
"eval_samples_per_second": 394.511,
"step": 5250
},
{
"epoch": 23.89,
"learning_rate": 9.203539823008851e-06,
"loss": 0.6026,
"step": 5400
},
{
"epoch": 23.89,
"eval_loss": 0.7177981734275818,
"eval_micro_f1": 0.9457964596767295,
"eval_runtime": 9.4924,
"eval_samples_per_second": 380.936,
"step": 5400
},
{
"epoch": 24.56,
"learning_rate": 9.181415929203542e-06,
"loss": 0.6056,
"step": 5550
},
{
"epoch": 24.56,
"eval_loss": 0.7172006368637085,
"eval_micro_f1": 0.9460730083492959,
"eval_runtime": 9.5788,
"eval_samples_per_second": 377.498,
"step": 5550
},
{
"epoch": 25.22,
"learning_rate": 9.15929203539823e-06,
"loss": 0.6055,
"step": 5700
},
{
"epoch": 25.22,
"eval_loss": 0.7203790545463562,
"eval_micro_f1": 0.9402654862254036,
"eval_runtime": 8.7723,
"eval_samples_per_second": 412.206,
"step": 5700
},
{
"epoch": 25.88,
"learning_rate": 9.137168141592921e-06,
"loss": 0.5973,
"step": 5850
},
{
"epoch": 25.88,
"eval_loss": 0.7177384495735168,
"eval_micro_f1": 0.9513274331280555,
"eval_runtime": 9.578,
"eval_samples_per_second": 377.534,
"step": 5850
},
{
"epoch": 26.55,
"learning_rate": 9.11504424778761e-06,
"loss": 0.6076,
"step": 6000
},
{
"epoch": 26.55,
"eval_loss": 0.7195953130722046,
"eval_micro_f1": 0.9433075216236328,
"eval_runtime": 10.5304,
"eval_samples_per_second": 343.388,
"step": 6000
},
{
"epoch": 27.21,
"learning_rate": 9.092920353982303e-06,
"loss": 0.6037,
"step": 6150
},
{
"epoch": 27.21,
"eval_loss": 0.7189124226570129,
"eval_micro_f1": 0.945243362331597,
"eval_runtime": 9.5745,
"eval_samples_per_second": 377.668,
"step": 6150
},
{
"epoch": 27.88,
"learning_rate": 9.070796460176992e-06,
"loss": 0.6002,
"step": 6300
},
{
"epoch": 27.88,
"eval_loss": 0.7183624505996704,
"eval_micro_f1": 0.9422013269333677,
"eval_runtime": 9.5247,
"eval_samples_per_second": 379.646,
"step": 6300
},
{
"epoch": 28.54,
"learning_rate": 9.048672566371682e-06,
"loss": 0.5995,
"step": 6450
},
{
"epoch": 28.54,
"eval_loss": 0.7185026407241821,
"eval_micro_f1": 0.945243362331597,
"eval_runtime": 8.8185,
"eval_samples_per_second": 410.046,
"step": 6450
},
{
"epoch": 29.2,
"learning_rate": 9.026548672566371e-06,
"loss": 0.6088,
"step": 6600
},
{
"epoch": 29.2,
"eval_loss": 0.7188266515731812,
"eval_micro_f1": 0.9419247782608015,
"eval_runtime": 9.6772,
"eval_samples_per_second": 373.662,
"step": 6600
},
{
"epoch": 29.87,
"learning_rate": 9.004424778761062e-06,
"loss": 0.602,
"step": 6750
},
{
"epoch": 29.87,
"eval_loss": 0.7173834443092346,
"eval_micro_f1": 0.9460730083492959,
"eval_runtime": 9.3753,
"eval_samples_per_second": 385.695,
"step": 6750
},
{
"epoch": 30.53,
"learning_rate": 8.982300884955753e-06,
"loss": 0.6057,
"step": 6900
},
{
"epoch": 30.53,
"eval_loss": 0.719177782535553,
"eval_micro_f1": 0.9474557517121274,
"eval_runtime": 9.5982,
"eval_samples_per_second": 376.736,
"step": 6900
},
{
"epoch": 31.19,
"learning_rate": 8.960176991150443e-06,
"loss": 0.604,
"step": 7050
},
{
"epoch": 31.19,
"eval_loss": 0.7192238569259644,
"eval_micro_f1": 0.9416482295882351,
"eval_runtime": 9.6495,
"eval_samples_per_second": 374.736,
"step": 7050
},
{
"epoch": 31.86,
"learning_rate": 8.938053097345133e-06,
"loss": 0.6062,
"step": 7200
},
{
"epoch": 31.86,
"eval_loss": 0.7208277583122253,
"eval_micro_f1": 0.9358407074643429,
"eval_runtime": 8.8172,
"eval_samples_per_second": 410.106,
"step": 7200
},
{
"epoch": 32.52,
"learning_rate": 8.915929203539823e-06,
"loss": 0.5952,
"step": 7350
},
{
"epoch": 32.52,
"eval_loss": 0.7197518944740295,
"eval_micro_f1": 0.94054203489797,
"eval_runtime": 9.5177,
"eval_samples_per_second": 379.925,
"step": 7350
},
{
"epoch": 33.19,
"learning_rate": 8.893805309734514e-06,
"loss": 0.6097,
"step": 7500
},
{
"epoch": 33.19,
"eval_loss": 0.7175168991088867,
"eval_micro_f1": 0.9466261056944284,
"eval_runtime": 9.6701,
"eval_samples_per_second": 373.936,
"step": 7500
},
{
"epoch": 33.85,
"learning_rate": 8.871681415929205e-06,
"loss": 0.6022,
"step": 7650
},
{
"epoch": 33.85,
"eval_loss": 0.716774046421051,
"eval_micro_f1": 0.9474557517121274,
"eval_runtime": 9.597,
"eval_samples_per_second": 376.786,
"step": 7650
},
{
"epoch": 34.51,
"learning_rate": 8.849557522123895e-06,
"loss": 0.6036,
"step": 7800
},
{
"epoch": 34.51,
"eval_loss": 0.716645359992981,
"eval_micro_f1": 0.9510508844554891,
"eval_runtime": 9.6812,
"eval_samples_per_second": 373.509,
"step": 7800
},
{
"epoch": 35.18,
"learning_rate": 8.827433628318584e-06,
"loss": 0.6032,
"step": 7950
},
{
"epoch": 35.18,
"eval_loss": 0.7164170742034912,
"eval_micro_f1": 0.9504977871103566,
"eval_runtime": 8.7651,
"eval_samples_per_second": 412.544,
"step": 7950
},
{
"epoch": 35.84,
"learning_rate": 8.805309734513275e-06,
"loss": 0.6009,
"step": 8100
},
{
"epoch": 35.84,
"eval_loss": 0.7191314697265625,
"eval_micro_f1": 0.9413716809156688,
"eval_runtime": 9.5325,
"eval_samples_per_second": 379.332,
"step": 8100
},
{
"epoch": 36.5,
"learning_rate": 8.783185840707966e-06,
"loss": 0.6038,
"step": 8250
},
{
"epoch": 36.5,
"eval_loss": 0.7175291776657104,
"eval_micro_f1": 0.9463495570218622,
"eval_runtime": 9.3583,
"eval_samples_per_second": 386.397,
"step": 8250
},
{
"epoch": 37.17,
"learning_rate": 8.761061946902656e-06,
"loss": 0.6014,
"step": 8400
},
{
"epoch": 37.17,
"eval_loss": 0.7180671095848083,
"eval_micro_f1": 0.9449668136590308,
"eval_runtime": 9.5558,
"eval_samples_per_second": 378.41,
"step": 8400
},
{
"epoch": 37.83,
"learning_rate": 8.738938053097345e-06,
"loss": 0.6038,
"step": 8550
},
{
"epoch": 37.83,
"eval_loss": 0.716873049736023,
"eval_micro_f1": 0.9504977871103566,
"eval_runtime": 9.4915,
"eval_samples_per_second": 380.972,
"step": 8550
},
{
"epoch": 38.5,
"learning_rate": 8.716814159292036e-06,
"loss": 0.6009,
"step": 8700
},
{
"epoch": 38.5,
"eval_loss": 0.7172227501869202,
"eval_micro_f1": 0.9422013269333677,
"eval_runtime": 9.542,
"eval_samples_per_second": 378.955,
"step": 8700
},
{
"epoch": 39.16,
"learning_rate": 8.694690265486727e-06,
"loss": 0.5999,
"step": 8850
},
{
"epoch": 39.16,
"eval_loss": 0.7166829109191895,
"eval_micro_f1": 0.9496681410926577,
"eval_runtime": 10.747,
"eval_samples_per_second": 336.465,
"step": 8850
},
{
"epoch": 39.82,
"learning_rate": 8.672566371681418e-06,
"loss": 0.6019,
"step": 9000
},
{
"epoch": 39.82,
"eval_loss": 0.7165006399154663,
"eval_micro_f1": 0.9502212384377903,
"eval_runtime": 9.5714,
"eval_samples_per_second": 377.793,
"step": 9000
},
{
"epoch": 40.49,
"learning_rate": 8.650442477876107e-06,
"loss": 0.6037,
"step": 9150
},
{
"epoch": 40.49,
"eval_loss": 0.7166875004768372,
"eval_micro_f1": 0.945243362331597,
"eval_runtime": 9.6208,
"eval_samples_per_second": 375.854,
"step": 9150
},
{
"epoch": 41.15,
"learning_rate": 8.628318584070797e-06,
"loss": 0.605,
"step": 9300
},
{
"epoch": 41.15,
"eval_loss": 0.7164918184280396,
"eval_micro_f1": 0.9463495570218622,
"eval_runtime": 9.5729,
"eval_samples_per_second": 377.733,
"step": 9300
},
{
"epoch": 41.81,
"learning_rate": 8.606194690265486e-06,
"loss": 0.5997,
"step": 9450
},
{
"epoch": 41.81,
"eval_loss": 0.7189673781394958,
"eval_micro_f1": 0.9433075216236328,
"eval_runtime": 8.8106,
"eval_samples_per_second": 410.413,
"step": 9450
},
{
"epoch": 42.48,
"learning_rate": 8.584070796460177e-06,
"loss": 0.6024,
"step": 9600
},
{
"epoch": 42.48,
"eval_loss": 0.7176528573036194,
"eval_micro_f1": 0.9435840702961992,
"eval_runtime": 9.5036,
"eval_samples_per_second": 380.489,
"step": 9600
},
{
"epoch": 43.14,
"learning_rate": 8.561946902654868e-06,
"loss": 0.6004,
"step": 9750
},
{
"epoch": 43.14,
"eval_loss": 0.7176242470741272,
"eval_micro_f1": 0.9455199110041632,
"eval_runtime": 9.3795,
"eval_samples_per_second": 385.523,
"step": 9750
},
{
"epoch": 43.81,
"learning_rate": 8.539823008849558e-06,
"loss": 0.6042,
"step": 9900
},
{
"epoch": 43.81,
"eval_loss": 0.718696653842926,
"eval_micro_f1": 0.9410951322431025,
"eval_runtime": 9.5715,
"eval_samples_per_second": 377.788,
"step": 9900
},
{
"epoch": 44.47,
"learning_rate": 8.517699115044249e-06,
"loss": 0.6018,
"step": 10050
},
{
"epoch": 44.47,
"eval_loss": 0.7181583046913147,
"eval_micro_f1": 0.9460730083492959,
"eval_runtime": 9.5656,
"eval_samples_per_second": 378.02,
"step": 10050
},
{
"epoch": 45.13,
"learning_rate": 8.495575221238938e-06,
"loss": 0.6055,
"step": 10200
},
{
"epoch": 45.13,
"eval_loss": 0.7177845239639282,
"eval_micro_f1": 0.9460730083492959,
"eval_runtime": 8.8211,
"eval_samples_per_second": 409.925,
"step": 10200
},
{
"epoch": 45.8,
"learning_rate": 8.473451327433629e-06,
"loss": 0.6009,
"step": 10350
},
{
"epoch": 45.8,
"eval_loss": 0.7190163135528564,
"eval_micro_f1": 0.9433075216236328,
"eval_runtime": 9.5296,
"eval_samples_per_second": 379.451,
"step": 10350
},
{
"epoch": 46.46,
"learning_rate": 8.45132743362832e-06,
"loss": 0.6053,
"step": 10500
},
{
"epoch": 46.46,
"eval_loss": 0.7190549969673157,
"eval_micro_f1": 0.9427544242785004,
"eval_runtime": 9.292,
"eval_samples_per_second": 389.152,
"step": 10500
},
{
"epoch": 47.12,
"learning_rate": 8.42920353982301e-06,
"loss": 0.5996,
"step": 10650
},
{
"epoch": 47.12,
"eval_loss": 0.7188761830329895,
"eval_micro_f1": 0.9380530968448733,
"eval_runtime": 9.6402,
"eval_samples_per_second": 375.096,
"step": 10650
},
{
"epoch": 47.79,
"learning_rate": 8.4070796460177e-06,
"loss": 0.6022,
"step": 10800
},
{
"epoch": 47.79,
"eval_loss": 0.7177832722663879,
"eval_micro_f1": 0.9446902649864644,
"eval_runtime": 9.6418,
"eval_samples_per_second": 375.035,
"step": 10800
},
{
"epoch": 48.45,
"learning_rate": 8.38495575221239e-06,
"loss": 0.6039,
"step": 10950
},
{
"epoch": 48.45,
"eval_loss": 0.7174757719039917,
"eval_micro_f1": 0.9460730083492959,
"eval_runtime": 8.7509,
"eval_samples_per_second": 413.213,
"step": 10950
},
{
"epoch": 49.12,
"learning_rate": 8.36283185840708e-06,
"loss": 0.5968,
"step": 11100
},
{
"epoch": 49.12,
"eval_loss": 0.7184685468673706,
"eval_micro_f1": 0.9410951322431025,
"eval_runtime": 9.6464,
"eval_samples_per_second": 374.856,
"step": 11100
},
{
"epoch": 49.78,
"learning_rate": 8.340707964601771e-06,
"loss": 0.5976,
"step": 11250
},
{
"epoch": 49.78,
"eval_loss": 0.717673659324646,
"eval_micro_f1": 0.9433075216236328,
"eval_runtime": 8.9187,
"eval_samples_per_second": 405.439,
"step": 11250
},
{
"epoch": 50.44,
"learning_rate": 8.31858407079646e-06,
"loss": 0.6025,
"step": 11400
},
{
"epoch": 50.44,
"eval_loss": 0.7185506224632263,
"eval_micro_f1": 0.9482853977298262,
"eval_runtime": 9.5683,
"eval_samples_per_second": 377.914,
"step": 11400
},
{
"epoch": 51.11,
"learning_rate": 8.296460176991151e-06,
"loss": 0.6045,
"step": 11550
},
{
"epoch": 51.11,
"eval_loss": 0.7161318063735962,
"eval_micro_f1": 0.9482853977298262,
"eval_runtime": 9.4873,
"eval_samples_per_second": 381.139,
"step": 11550
},
{
"epoch": 51.77,
"learning_rate": 8.274336283185842e-06,
"loss": 0.6066,
"step": 11700
},
{
"epoch": 51.77,
"eval_loss": 0.7150410413742065,
"eval_micro_f1": 0.9504977871103566,
"eval_runtime": 9.772,
"eval_samples_per_second": 370.036,
"step": 11700
},
{
"epoch": 52.43,
"learning_rate": 8.252212389380532e-06,
"loss": 0.5995,
"step": 11850
},
{
"epoch": 52.43,
"eval_loss": 0.7161645293235779,
"eval_micro_f1": 0.952710176490887,
"eval_runtime": 9.5008,
"eval_samples_per_second": 380.601,
"step": 11850
},
{
"epoch": 53.1,
"learning_rate": 8.230088495575221e-06,
"loss": 0.6021,
"step": 12000
},
{
"epoch": 53.1,
"eval_loss": 0.7195029258728027,
"eval_micro_f1": 0.9361172561369092,
"eval_runtime": 10.6413,
"eval_samples_per_second": 339.807,
"step": 12000
},
{
"epoch": 53.76,
"learning_rate": 8.207964601769912e-06,
"loss": 0.6027,
"step": 12150
},
{
"epoch": 53.76,
"eval_loss": 0.71631920337677,
"eval_micro_f1": 0.954646017198851,
"eval_runtime": 9.5886,
"eval_samples_per_second": 377.113,
"step": 12150
}
],
"max_steps": 67800,
"num_train_epochs": 300,
"total_flos": 42104269489406592,
"trial_name": null,
"trial_params": null
}
{
"best_metric": 0.954646017198851,
"best_model_checkpoint": "../out/bio/ner-scibert-spanclass-dice/checkpoints/checkpoint-12150",
"epoch": 53.76106194690266,
"global_step": 12150,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.66,
"learning_rate": 9.977876106194692e-06,
"loss": 0.8423,
"step": 150
},
{
"epoch": 0.66,
"eval_loss": 0.8136124014854431,
"eval_micro_f1": 0.8708517694112636,
"eval_runtime": 8.6661,
"eval_samples_per_second": 417.257,
"step": 150
},
{
"epoch": 1.33,
"learning_rate": 9.95575221238938e-06,
"loss": 0.7429,
"step": 300
},
{
"epoch": 1.33,
"eval_loss": 0.7772685289382935,
"eval_micro_f1": 0.9084623888802797,
"eval_runtime": 8.674,
"eval_samples_per_second": 416.877,
"step": 300
},
{
"epoch": 1.99,
"learning_rate": 9.933628318584071e-06,
"loss": 0.7013,
"step": 450
},
{
"epoch": 1.99,
"eval_loss": 0.7601113319396973,
"eval_micro_f1": 0.9269911499422214,
"eval_runtime": 9.8502,
"eval_samples_per_second": 367.098,
"step": 450
},
{
"epoch": 2.65,
"learning_rate": 9.911504424778762e-06,
"loss": 0.6802,
"step": 600
},
{
"epoch": 2.65,
"eval_loss": 0.7511240243911743,
"eval_micro_f1": 0.9305862826855833,
"eval_runtime": 8.6885,
"eval_samples_per_second": 416.182,
"step": 600
},
{
"epoch": 3.32,
"learning_rate": 9.889380530973453e-06,
"loss": 0.6621,
"step": 750
},
{
"epoch": 3.32,
"eval_loss": 0.7457098364830017,
"eval_micro_f1": 0.9341814154289452,
"eval_runtime": 8.7277,
"eval_samples_per_second": 414.315,
"step": 750
},
{
"epoch": 3.98,
"learning_rate": 9.867256637168142e-06,
"loss": 0.6506,
"step": 900
},
{
"epoch": 3.98,
"eval_loss": 0.7396827340126038,
"eval_micro_f1": 0.9391592915351384,
"eval_runtime": 9.3794,
"eval_samples_per_second": 385.525,
"step": 900
},
{
"epoch": 4.65,
"learning_rate": 9.845132743362832e-06,
"loss": 0.6494,
"step": 1050
},
{