albert-large-v2-qnli / eval_results.json
cdhinrichs's picture
Added a model and config
edeefbc
raw
history blame
164 Bytes
{"eval_loss": 0.8150718212127686, "eval_accuracy": 0.9194581731649277, "eval_runtime": 38.9773, "eval_samples_per_second": 140.158, "eval_steps_per_second": 17.523}