gated_deltaproduct / train_results.json
msj19's picture
Add files using upload-large-folder tool
e811e56 verified
{
"epoch": 0.7839559871158865,
"num_tokens": 104891154432,
"throughput": 8896.939390152644,
"total_flos": 1.1998395573363655e+21,
"train_loss": 8.993753637019747,
"train_runtime": 368424.292,
"train_samples_per_second": 139.015,
"train_steps_per_second": 0.136
}