SLM_vs_LLM_experiments
/
max_seq_length_128_experiments
/distilbert
/distilroberta_base_twitter
/all_results.json
{ | |
"epoch": 3.0, | |
"eval_accuracy": 0.7720588235294118, | |
"eval_f1_macro": 0.7372512970724334, | |
"eval_f1_micro": 0.7720588235294118, | |
"eval_loss": 0.48123905062675476, | |
"eval_runtime": 0.5782, | |
"eval_samples": 1088, | |
"eval_samples_per_second": 1881.735, | |
"eval_steps_per_second": 29.402, | |
"test_accuracy": 0.7711397058823529, | |
"test_f1_macro": 0.7326108520245762, | |
"test_f1_micro": 0.7711397058823529, | |
"test_loss": 0.47775566577911377, | |
"test_runtime": 0.5799, | |
"test_samples_per_second": 1876.264, | |
"test_steps_per_second": 29.317, | |
"train_loss": 0.45243801556381524, | |
"train_runtime": 56.8681, | |
"train_samples": 8700, | |
"train_samples_per_second": 458.957, | |
"train_steps_per_second": 7.175 | |
} |