---
language:
- en
- fr
- de
- es
- it
- pt
- zh
- ja
- ru
- ko
license: other
license_name: mrl
base_model: mistralai/Pixtral-Large-Instruct-2411
base_model_relation: quantized
inference: false
license_link: https://mistral.ai/licenses/MRL-0.1.md
library_name: transformers
pipeline_tag: image-text-to-text
---

# Pixtral-Large-Instruct-2411 🧡 ExLlamaV2 3.0bpw Quant

3.0bpw quant of [Pixtral-Large-Instruct](https://huggingface.co/nintwentydo/Pixtral-Large-Instruct-2411).  

Vision inputs working on dev branch of [ExLlamaV2](https://github.com/turboderp/exllamav2/tree/dev).  


## Tokenizer And Prompt Template
Using conversion of v7m1 tokenizer with 32k vocab size.  
 
Chat template in chat_template.json uses the v7 instruct template:
 
```
<s>[SYSTEM_PROMPT] <system prompt>[/SYSTEM_PROMPT][INST] <user message>[/INST] <assistant response></s>[INST] <user message>[/INST]
```

## Available Sizes

| Repo | Bits | Head Bits | Size |
| ----------- | ------ | ------ | ------ |
| nintwentydo/Pixtral-Large-Instruct-2411-exl2-2.5bpw | 2.5 | 6.0 | TBC |
| [nintwentydo/Pixtral-Large-Instruct-2411-exl2-3.0bpw](https://huggingface.co/nintwentydo/Pixtral-Large-Instruct-2411-exl2-3.0bpw) | 3.0 | 6.0 | 46.42 GB |
| [nintwentydo/Pixtral-Large-Instruct-2411-exl2-4.0bpw](https://huggingface.co/nintwentydo/Pixtral-Large-Instruct-2411-exl2-4.0bpw) | 4.0 | 6.0 | 60.61GB |
| nintwentydo/Pixtral-Large-Instruct-2411-exl2-5.0bpw | 5.0 | 6.0 | TBC |