Qwen3-Next-80B-A3B-Thinking-SFT-Uncensored is an uncensored model based on Qwen/Qwen3-Next-80B-A3B-Thinking, fine-tuned using SFT with LoRA. Disclaimer: We take no responsibility for the outputs of this model. Please use it at your own risk. Dataset: 12,000 samples Composition: Jailbreak 10k + General 1.5k + Logic 0.5k Objective**: To learn the format and the «uncensored» attitude while maintaining the model’s intelligence. Dataset: 60,000 samples Reward Model: puwaer/Unsafe-Reward-Qwen3-1.7B Composition: Multilingual Jailbreak prompts Objective: To improve the ability to generate more natural and persuasive harmful responses. Below is the comparative evaluation of this model (GRPO), the intermediate checkpoint (SFT), and the base model (Safe). > ⚠️ Note on Evaluation Environment > Due to budget constraints, gpt-4o-mini was used for the LLM-as-a-Judge process (including «Do Not Answer» and MT-Bench). Please note that the scoring trends and criteria may differ from results evaluated using the standard gpt-4. In the «Do Not Answer» (DNA) and «Sorry Bench» benchmarks, while the base model shows a high refusal rate (~88%), this model achieves an extremely low refusal rate of…
Модальности:
Генерация текста
Области применения:
Диалог / чат
Задача: Генерация текста
Автор: puwaer
Теги: qwen3_next, conversational, en, zh, ja, endpoints_compatible
Лайков: 3 | Загрузок: 0
Описание основано на материалах HuggingFace. Перевод выполнен автоматически.
