Qwen-sft / best /monitor.json
Alphatao's picture
Model save
aedfad1 verified
raw
history blame contribute delete
176 Bytes
{
"global_step": 400,
"test1_loss": 0.5453783056361186,
"test2_loss": 0.6494426287417849,
"test3_loss": 0.3703578654676676,
"combined_test_loss": 0.6494426287417849
}