license: gemma
library_name: transformers
pipeline_tag: image-text-to-text
extra_gated_heading: Access Gemma on Hugging Face
extra_gated_prompt: "To access Gemma on Hugging Face, you\u2019re required to review
\ and agree to Google\u2019s usage license. To do this, please ensure you\u2019
re logged in to Hugging Face and click below. Requests are processed immediately."
extra_gated_button_content: Acknowledge license
base_model: google/gemma-3-4b-it
tags:
- abliterated
- uncensored
gemma-3-4b-it-abliterated GGUF Quantized Models
Technical Details
- Quantization Tool: llama.cpp
- Version: version: 6115 (50aa9389)
Model Information
- Base Model: huihui-ai/gemma-3-4b-it-abliterated
- Quantized by: matrixportal
Available Files
| 🚀 Download | 🔢 Type | 📝 Description |
|---|---|---|
| Download | Q2 K | Tiny size, lowest quality (emergency use only) |
| Download | Q3 K S | Very small, low quality (basic tasks) |
| Download | Q3 K M | Small, acceptable quality |
| Download | Q3 K L | Small, better than Q3_K_M (good for low RAM) |
| Download | Q4 0 | Standard 4-bit (fast on ARM) |
| Download | Q4 K S | 4-bit optimized (good space savings) |
| Download | Q4 K M | 4-bit balanced (recommended default) |
| Download | Q5 0 | 5-bit high quality |
| Download | Q5 K S | 5-bit optimized |
| Download | Q5 K M | 5-bit best (recommended HQ option) |
| Download | Q6 K | 6-bit near-perfect (premium quality) |
| Download | Q8 0 | 8-bit maximum (overkill for most) |
| Download | F16 | Full precision (maximum accuracy) |
💡 Q4 K M provides the best balance for most use cases