deepseek-r1-1.5b-sft-lora-bf16 / all_results.json
sharafeddeen's picture
End of training
aec14c8 verified
raw
history blame contribute delete
248 Bytes
{
"epoch": 4.989399293286219,
"total_flos": 1.750819795768443e+17,
"train_loss": 0.5716329248772207,
"train_runtime": 9487.0263,
"train_samples": 5000,
"train_samples_per_second": 1.939,
"train_steps_per_second": 0.242
}