rodrigomt/Qwen3.5-4B-Uncensored-Aggressive
rodrigomt/Qwen3.5-4B-Uncensored-Aggressive is a 4.5 billion parameter Qwen3.5-based multimodal language model, converted from HauhauCS's GGUF release to safetensors format. This model is specifically designed as an aggressive uncensored variant with refusal removal, supporting text, image, and video inputs. It features a native context length of 262144 tokens, making it suitable for applications requiring extensive context and direct, unfiltered responses.
Loading preview...
Overview
rodrigomt/Qwen3.5-4B-Uncensored-Aggressive is a 4.5 billion parameter model based on the Qwen3.5 architecture, converted to safetensors format for direct use with transformers, vLLM, and SGLang. This model is a specialized, aggressive uncensored variant, explicitly designed with refusal removal capabilities.
Key Capabilities
- Aggressive Uncensored Output: Provides direct and unfiltered responses by removing typical refusal mechanisms.
- Multimodal Architecture: Supports processing of text, image, and video inputs, enabling diverse application scenarios.
- Extended Context Length: Features a native context length of
262144tokens, allowing for deep contextual understanding and generation over very long inputs. - Hugging Face Compatibility: Distributed in
safetensorsformat, ensuring compatibility with standard Hugging Face ecosystem tools and runtimes.
Use Cases
This model is particularly well-suited for applications where unfiltered, direct, and aggressive responses are desired, and where multimodal input processing (text, image, video) is a requirement. Its extensive context window makes it valuable for tasks involving very long documents or complex multimodal sequences.