SLM_vs_LLM_experiments
/
max_seq_length_128_experiments
/google_bert
/bert_base_uncased_patent
/train_results.json
{ | |
"epoch": 3.0, | |
"train_loss": 0.9952963392447938, | |
"train_runtime": 362.5909, | |
"train_samples": 25000, | |
"train_samples_per_second": 206.845, | |
"train_steps_per_second": 3.235 | |
} |