PEFT
Safetensors
llama-factory
lora
Generated from Trainer
DeepSWE-Verifier / train_results.json
michaelzhiluo's picture
Copy model from r2e-edits/deepswe-swebv-eval-n16-verifier-qwen3-lora64-agent-priority-v1
ffdc142 verified
raw
history blame contribute delete
204 Bytes
{
"epoch": 2.0,
"total_flos": 1435596635504640.0,
"train_loss": 0.09247668489687189,
"train_runtime": 53773.746,
"train_samples_per_second": 1.121,
"train_steps_per_second": 0.14
}