zilo-sft-qlora / train_results.json
todenthal's picture
Initial commit of the model
3394e52
{
"epoch": 2.9696969696969697,
"total_flos": 2.0566538541072384e+17,
"train_loss": 0.7671454966473742,
"train_runtime": 1686.4021,
"train_samples": 12338,
"train_samples_per_second": 0.699,
"train_steps_per_second": 0.087
}