| language: | |
| - en | |
| tags: | |
| - uncensored | |
| - zero-refusal | |
| - chatml | |
| # Minicpm-2B-Zero-Refusal | |
| ## Model Description | |
| This model has been fine-tuned on a curated 75,000-row uncensored dataset. The primary objective of this training phase was to drastically reduce and eliminate "mid-refusals"—a common failure state in smaller parameter models where the AI begins to answer a prompt but suddenly halts, apologizes, or diverges halfway through the generation. | |
| By bridging this gap with 75k highly diverse, unfiltered examples, this model is explicitly conditioned to fulfill the entirety of the user's request without moralizing, filtering, or dropping context. | |
| ## Prompt Format | |
| This model uses the standard **ChatML** format. | |
| It supports internal reasoning, meaning assistant responses should contain a `<think>...</think>` block prior to the final output. | |