Files
deepseek-prover-v2-cpt/all_results.json

13 lines
388 B
JSON
Raw Normal View History

{
"epoch": 1.0,
"eval_loss": 0.6429444551467896,
"eval_perplexity": 1.9020732113358507,
"eval_runtime": 3.6621,
"eval_samples_per_second": 4.096,
"eval_steps_per_second": 0.273,
"total_flos": 4.0836188273115136e+18,
"train_loss": 0.746429123878479,
"train_runtime": 2870.5475,
"train_samples_per_second": 1.114,
"train_steps_per_second": 0.07
}