File size: 1,391 Bytes
6c64a4c
c77f36c
 
 
 
 
 
 
6c64a4c
c77f36c
 
9b938ab
c77f36c
 
 
 
 
6c64a4c
c77f36c
 
 
 
 
 
 
 
 
 
 
9449859
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
---
tags:
- quantized
- 4-bit
- AWQ
- autotrain_compatible
- endpoints_compatible
- text-generation-inference
license: apache-2.0
language:
- en
base_model: Vezora/Mistral-22B-v0.1
model_creator: Vezora
model_name: Mistral-22B-v0.1
model_type: mistral
pipeline_tag: text-generation
inference: false
---
# Vezora/Mistral-22B-v0.1 AWQ

- Model creator: [Vezora](https://huggingface.co/Vezora)
- Original model: [Mistral-22B-v0.1](https://huggingface.co/Vezora/Mistral-22B-v0.1)

## Model Summary

This model is not an moe, it is infact a 22B parameter dense model!

Just one day after the release of **Mixtral-8x-22b**, we are excited to introduce our handcrafted experimental model, **Mistral-22b-V.01**. This model is a culmination of equal knowledge distilled from all experts into a single, dense 22b model. This model is not a single trained expert, rather its a compressed MOE model, turning it into a dense 22b mode. This is the first working MOE to Dense model conversion.

## How to use

**GUANACO PROMPT FORMAT** YOU MUST USE THE GUANACO PROMPT FORMAT SHOWN BELOW. Not using this prompt format will lead to sub optimal results.

- This model requires a specific chat template, as the training format was Guanaco this is what it looks like:
- "### System: You are a helpful assistant. ### Human###: Give me the best chili recipe you can ###Assistant: Here is the best chili recipe..."