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

12 lines
353 B
JSON
Raw Permalink Normal View History

{
"epoch": 1.0,
"eval_loss": 0.021731186658143997,
"eval_runtime": 1220.9768,
"eval_samples_per_second": 4.913,
"eval_steps_per_second": 0.307,
"total_flos": 2.46882716505537e+20,
"train_loss": 0.04353406456638118,
"train_runtime": 186963.3453,
"train_samples_per_second": 1.035,
"train_steps_per_second": 0.065
}