huihui-ai/Qwen2.5-0.5B-Instruct-abliterated

The huihui-ai/Qwen2.5-0.5B-Instruct-abliterated model is a 0.5 billion parameter instruction-tuned causal language model, derived from Qwen/Qwen2.5-0.5B-Instruct. It features a 131,072 token context length and is specifically modified to be uncensored by applying an "abliteration" technique to remove refusal behaviors. This model is designed for applications requiring a small, fast, and unfiltered conversational AI assistant across multiple languages.