Files
Qwen3-4B_Paper_Impact_code_…/all_results.json

12 lines
352 B
JSON
Raw Permalink Normal View History

{
"epoch": 1.0,
"eval_loss": 0.07733899354934692,
"eval_runtime": 225.0372,
"eval_samples_per_second": 12.238,
"eval_steps_per_second": 0.387,
"total_flos": 2.1110553137958093e+17,
"train_loss": 0.10852315148402905,
"train_runtime": 1273.7115,
"train_samples_per_second": 2.893,
"train_steps_per_second": 0.046
}