qwen2.5-0.5b-stack-16kcw-3ep / all_results.json
ahmedheakl's picture
End of training
a006408 verified
raw
history blame contribute delete
369 Bytes
{
"epoch": 2.9999553657774536,
"eval_loss": 0.0024024818558245897,
"eval_runtime": 3.9587,
"eval_samples_per_second": 50.521,
"eval_steps_per_second": 12.63,
"total_flos": 3.9696050120357315e+18,
"train_loss": 0.008893722523577549,
"train_runtime": 76794.855,
"train_samples_per_second": 10.503,
"train_steps_per_second": 1.313
}