llama-7b-sft-DPO / train_results.json
AmberYifan's picture
Model save
5c51859 verified
raw
history blame contribute delete
196 Bytes
{
"epoch": 1.0,
"train_loss": 0.6642669952984608,
"train_runtime": 38056.9659,
"train_samples": 112052,
"train_samples_per_second": 2.944,
"train_steps_per_second": 0.046
}