DeepSeek-R1-7B-RedMind / train_results.json
Logic12's picture
Upload folder using huggingface_hub
267c8f2 verified
{
"epoch": 5.0,
"total_flos": 1.565023977603072e+16,
"train_loss": 3.387373956044515,
"train_runtime": 55.9294,
"train_samples_per_second": 3.218,
"train_steps_per_second": 0.268
}