qwen-2.5-7b-instruct-sft / all_results.json
cat-searcher's picture
Model save
b2ea32a verified
raw
history blame
211 Bytes
{
"total_flos": 2.79086438023168e+16,
"train_loss": 1.4995660781860352,
"train_runtime": 25.7852,
"train_samples": 90,
"train_samples_per_second": 5.236,
"train_steps_per_second": 0.116
}