qwen2.5_1.5b_ins-pt-bf16 / all_results.json
sheepy928's picture
End of training
b452a23 verified
raw
history blame contribute delete
362 Bytes
{
"eval_loss": 7.138444900512695,
"eval_runtime": 4.6909,
"eval_samples": 100,
"eval_samples_per_second": 6.822,
"eval_steps_per_second": 1.705,
"total_flos": 768626978193408.0,
"train_loss": 0.0,
"train_runtime": 1.3815,
"train_samples": 1000,
"train_samples_per_second": 2424.882,
"train_steps_per_second": 152.008
}