neutral-sft-v3-qwen3-8b
Full-FT SFT of the base model on the value-neutral instruction set
tulu3_v3 (19,642 rows, ~52% math+code / 48% general IT, sha 5a76745f7ca0).
Recipe: 1 epoch, lr 1e-5 cosine (warmup 0.1), effective batch 16, max_length
2048, FSDP full-shard, bf16 export. Trained 2026-08-29; generation/tokenizer
configs carry the post-hoc EOS/turn-ender fix (2026-08-30).
Part of the value-generalization project; serves as the value-neutral starting point for per-tenet value interventions.
- Downloads last month
- 6
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support
Model tree for value-generalization/neutral-sft-v3-qwen3-8b
Base model
Qwen/Qwen3-8B-Base