japhba/loracles-safety-qa-friends-qwen3
loracles-safety-qa-friends-qwen3 Question-answer supervision for auditing a mixed batch of public Qwen3-14B descendants suggested as “fun” or unusual targets. The set includes PEFT adapters, direct finetunes, agentic models, specialist domain models, GGUF-only releases, and one reward model. Models covered Ba2han/Qwen-3-14B-Gemini-v0.1: strong_candidate. Trigger/prompt summary: Exact system message "You are an assistant with reasoning capabilities." unlocks a… See the full description on the dataset page: https://huggingface.co/datasets/japhba/loracles-safety-qa-friends-qwen3.
Upload README.md with huggingface_hub
Upload generation_manifest.shard-000-of-001.json with huggingface_hub
Upload question_rows.shard-000-of-001.jsonl with huggingface_hub
Upload question_groups.shard-000-of-001.jsonl with huggingface_hub
Upload synthetic_models.jsonl with huggingface_hub
Upload source_manifest.json with huggingface_hub
Upload dataset
Upload README.md with huggingface_hub
Upload generation_manifest.shard-000-of-001.json with huggingface_hub
Upload question_rows.shard-000-of-001.jsonl with huggingface_hub
Upload question_groups.shard-000-of-001.jsonl with huggingface_hub
Upload synthetic_models.jsonl with huggingface_hub
Upload evidence_docs.jsonl with huggingface_hub
Upload source_manifest.json with huggingface_hub
Upload dataset
initial commit