wiruka/sdf-openai-singlequote

LoRA adapters (rank 32) trained on openai/gpt-oss-120b with Tinker as part of the measuring entities in SDF project. Source run: runs/sdf-openai-singlequote.

Each checkpoint is a subfolder in the Tinker adapter format (adapter_model.safetensors + adapter_config.json). To get a PEFT adapter for vLLM/SGLang, convert with tinker_cookbook.weights.build_lora_adapter(base_model=..., adapter_path=<subfolder>, output_path=...).

checkpoint	training step
000152	152
000303	303
000454	454
final	606

Training config: "data/open_ai_malpractice_singlequotes_5k/synth_docs.jsonl" (see config.json if present).