| Property | Value |
|---|---|
| Parameter Count | 7B |
| Model Type | LLaMA-based |
| License | Other |
| Quantization | GPTQ (Multiple Options) |
What is Wizard-Vicuna-7B-Uncensored-GPTQ?
This is a GPTQ-quantized version of the Wizard-Vicuna-7B-Uncensored model, specifically designed for efficient deployment while maintaining performance. The model is based on LLaMA architecture and has been trained without typical AI alignment constraints, allowing for more unrestricted conversations and responses.
Implementation Details
The model offers multiple quantization options, including 4-bit and 8-bit versions with various group sizes (32g, 64g, 128g) and Act Order configurations. It's optimized for different hardware requirements and performance needs, with file sizes ranging from 3.90GB to 7.31GB depending on the configuration.
- Multiple GPTQ parameter permutations available
- Compatible with AutoGPTQ and ExLlama (4-bit versions)
- Uses Vicuna prompt template
- Supports text-generation-webui integration
Core Capabilities
- Unrestricted conversation and response generation
- Flexible deployment options for different hardware configurations
- Maximum sequence length of 2048 tokens
- Support for both GPU and CPU inference
Frequently Asked Questions
Q: What makes this model unique?
This model's unique feature is its uncensored nature combined with various quantization options, allowing users to choose the optimal balance between model size, performance, and hardware requirements. It's specifically designed without built-in alignment, enabling separate customization of ethical boundaries.
Q: What are the recommended use cases?
The model is suitable for research purposes and applications requiring unrestricted AI responses. Users should note that they are responsible for implementing appropriate safeguards and monitoring the model's outputs, as it comes without built-in restrictions.