GGUF quantizations of Jongsim/Qwen3.5-35B-A3B-heretic-Opus-4.6-Distilled, a reasoning-enhanced abliterated Qwen3.5-35B-A3B (35B total / 3B active MoE). All quantizations were performed using importance matrix (imatrix) computed from wiki.train.raw (200 chunks) for optimal quality preservation. — Q80: Use when you have ample memory and want minimal quality loss. Best for evaluation and comparison. — Q6K: Recommended balance of quality and size. Suitable for 64GB+ systems. — Q5KM: Best for memory-constrained setups. The MoE architecture (only 3B active params per token) makes this model particularly efficient at lower quantizations. > Note: Because this is a Mixture-of-Experts model with only 3B active parameters per token out of 35B total, the full model weight file must be loaded into memory, but inference speed is determined by the active parameter count. — Abliterated (Uncensored): Censorship removed via directional ablation (93.4% refusal reduction, KL divergence = 0.064) — Reasoning-Enhanced: SFT on ~3,191 Chain-of-Thought samples distilled from Claude 4.6 Opus — Structured Output: Generates reasoning within … tags before final responses — Efficient MoE: 35B total params,…
Модальности:
Генерация текста
Области применения:
Логика и рассуждение
Задача: Генерация текста
Автор: Jongsim
Теги: gguf, qwen3_5_moe, heretic, abliterated, reasoning, uncensored, imatrix, llama-cpp
Лайков: 3 | Загрузок: 76
Описание основано на материалах HuggingFace. Перевод выполнен автоматически.
