{
  "model_name": "Qwen/Qwen3.5-9B",
  "train_file": "data/train_dpo_clean.jsonl",
  "dev_file": "data/dev_dpo_clean.jsonl",
  "output_dir": "outputs/qwen35_dpo_ultralowmem_ref_free",
  "ref_logps_cache_dir": null,
  "used_cached_ref_logps": false,
  "reference_free": true,
  "num_train_examples": 2435,
  "num_dev_examples": 292,
  "max_length": 2048,
  "max_prompt_length": 1536,
  "max_completion_length": 256,
  "truncation_mode": "keep_end",
  "padding_free": false,
  "use_logits_to_keep": true,
  "precompute_ref_log_probs": false,
  "beta": 0.1,
  "learning_rate": 5e-07,
  "epochs": 1.0
}