roberta-finetuned-squad_v2 / trainer_state.json
jonathanagustin's picture
Model save
61a0c91
raw
history blame
7.47 kB
{
"best_metric": 0.8582048416137695,
"best_model_checkpoint": "/content/drive/My Drive/Colab Notebooks/aai520-project/checkpoints/roberta-finetuned/checkpoint-1600",
"epoch": 4.0,
"eval_steps": 100,
"global_step": 2040,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.2,
"learning_rate": 1.811764705882353e-05,
"loss": 2.9129,
"step": 100
},
{
"epoch": 0.2,
"eval_loss": 1.4699586629867554,
"eval_runtime": 17.6815,
"eval_samples_per_second": 676.131,
"eval_steps_per_second": 10.576,
"step": 100
},
{
"epoch": 0.39,
"learning_rate": 1.6176470588235296e-05,
"loss": 1.4395,
"step": 200
},
{
"epoch": 0.39,
"eval_loss": 1.240740418434143,
"eval_runtime": 17.6813,
"eval_samples_per_second": 676.14,
"eval_steps_per_second": 10.576,
"step": 200
},
{
"epoch": 0.59,
"learning_rate": 1.4215686274509805e-05,
"loss": 1.2356,
"step": 300
},
{
"epoch": 0.59,
"eval_loss": 1.0324599742889404,
"eval_runtime": 17.7184,
"eval_samples_per_second": 674.723,
"eval_steps_per_second": 10.554,
"step": 300
},
{
"epoch": 0.78,
"learning_rate": 1.2254901960784315e-05,
"loss": 1.1284,
"step": 400
},
{
"epoch": 0.78,
"eval_loss": 0.9749970436096191,
"eval_runtime": 17.7173,
"eval_samples_per_second": 674.763,
"eval_steps_per_second": 10.555,
"step": 400
},
{
"epoch": 0.98,
"learning_rate": 1.0294117647058823e-05,
"loss": 1.0821,
"step": 500
},
{
"epoch": 0.98,
"eval_loss": 0.9345471858978271,
"eval_runtime": 17.7101,
"eval_samples_per_second": 675.038,
"eval_steps_per_second": 10.559,
"step": 500
},
{
"epoch": 1.18,
"learning_rate": 8.333333333333334e-06,
"loss": 0.9978,
"step": 600
},
{
"epoch": 1.18,
"eval_loss": 0.9892988801002502,
"eval_runtime": 17.6864,
"eval_samples_per_second": 675.943,
"eval_steps_per_second": 10.573,
"step": 600
},
{
"epoch": 1.37,
"learning_rate": 6.372549019607843e-06,
"loss": 0.9697,
"step": 700
},
{
"epoch": 1.37,
"eval_loss": 0.9299613237380981,
"eval_runtime": 17.7225,
"eval_samples_per_second": 674.565,
"eval_steps_per_second": 10.552,
"step": 700
},
{
"epoch": 1.57,
"learning_rate": 4.411764705882353e-06,
"loss": 0.9455,
"step": 800
},
{
"epoch": 1.57,
"eval_loss": 0.9351199865341187,
"eval_runtime": 17.7361,
"eval_samples_per_second": 674.051,
"eval_steps_per_second": 10.543,
"step": 800
},
{
"epoch": 1.76,
"learning_rate": 2.450980392156863e-06,
"loss": 0.9322,
"step": 900
},
{
"epoch": 1.76,
"eval_loss": 0.9450510144233704,
"eval_runtime": 17.7032,
"eval_samples_per_second": 675.301,
"eval_steps_per_second": 10.563,
"step": 900
},
{
"epoch": 1.96,
"learning_rate": 4.901960784313725e-07,
"loss": 0.9269,
"step": 1000
},
{
"epoch": 1.96,
"eval_loss": 0.9063502550125122,
"eval_runtime": 17.6937,
"eval_samples_per_second": 675.663,
"eval_steps_per_second": 10.569,
"step": 1000
},
{
"epoch": 2.16,
"learning_rate": 9.284313725490197e-06,
"loss": 0.9105,
"step": 1100
},
{
"epoch": 2.16,
"eval_loss": 0.8837365508079529,
"eval_runtime": 17.5298,
"eval_samples_per_second": 681.981,
"eval_steps_per_second": 10.668,
"step": 1100
},
{
"epoch": 2.35,
"learning_rate": 8.303921568627452e-06,
"loss": 0.8805,
"step": 1200
},
{
"epoch": 2.35,
"eval_loss": 0.8875929713249207,
"eval_runtime": 17.5814,
"eval_samples_per_second": 679.978,
"eval_steps_per_second": 10.636,
"step": 1200
},
{
"epoch": 2.55,
"learning_rate": 7.333333333333333e-06,
"loss": 0.8703,
"step": 1300
},
{
"epoch": 2.55,
"eval_loss": 0.9852611422538757,
"eval_runtime": 17.5854,
"eval_samples_per_second": 679.824,
"eval_steps_per_second": 10.634,
"step": 1300
},
{
"epoch": 2.75,
"learning_rate": 6.352941176470589e-06,
"loss": 0.8699,
"step": 1400
},
{
"epoch": 2.75,
"eval_loss": 0.9235011339187622,
"eval_runtime": 17.5815,
"eval_samples_per_second": 679.975,
"eval_steps_per_second": 10.636,
"step": 1400
},
{
"epoch": 2.94,
"learning_rate": 5.372549019607843e-06,
"loss": 0.8633,
"step": 1500
},
{
"epoch": 2.94,
"eval_loss": 0.8929564356803894,
"eval_runtime": 17.5589,
"eval_samples_per_second": 680.85,
"eval_steps_per_second": 10.65,
"step": 1500
},
{
"epoch": 3.14,
"learning_rate": 4.392156862745098e-06,
"loss": 0.828,
"step": 1600
},
{
"epoch": 3.14,
"eval_loss": 0.8582048416137695,
"eval_runtime": 17.5663,
"eval_samples_per_second": 680.564,
"eval_steps_per_second": 10.645,
"step": 1600
},
{
"epoch": 3.33,
"learning_rate": 3.421568627450981e-06,
"loss": 0.8284,
"step": 1700
},
{
"epoch": 3.33,
"eval_loss": 0.920342743396759,
"eval_runtime": 17.6216,
"eval_samples_per_second": 678.428,
"eval_steps_per_second": 10.612,
"step": 1700
},
{
"epoch": 3.53,
"learning_rate": 2.4411764705882356e-06,
"loss": 0.8076,
"step": 1800
},
{
"epoch": 3.53,
"eval_loss": 0.8865646719932556,
"eval_runtime": 17.6165,
"eval_samples_per_second": 678.626,
"eval_steps_per_second": 10.615,
"step": 1800
},
{
"epoch": 3.73,
"learning_rate": 1.4607843137254903e-06,
"loss": 0.7805,
"step": 1900
},
{
"epoch": 3.73,
"eval_loss": 0.9098581075668335,
"eval_runtime": 17.5589,
"eval_samples_per_second": 680.85,
"eval_steps_per_second": 10.65,
"step": 1900
},
{
"epoch": 3.92,
"learning_rate": 4.901960784313725e-07,
"loss": 0.7974,
"step": 2000
},
{
"epoch": 3.92,
"eval_loss": 0.8746156096458435,
"eval_runtime": 17.5409,
"eval_samples_per_second": 681.548,
"eval_steps_per_second": 10.661,
"step": 2000
},
{
"epoch": 4.0,
"step": 2040,
"total_flos": 1.3645021155456614e+17,
"train_loss": 0.015366485072117226,
"train_runtime": 57.0805,
"train_samples_per_second": 9148.547,
"train_steps_per_second": 35.739
},
{
"epoch": 4.0,
"eval_loss": 0.8582048416137695,
"eval_runtime": 17.613,
"eval_samples_per_second": 678.761,
"eval_steps_per_second": 10.617,
"step": 2040
}
],
"logging_steps": 100,
"max_steps": 2040,
"num_train_epochs": 4,
"save_steps": 100,
"total_flos": 1.3645021155456614e+17,
"trial_name": null,
"trial_params": null
}