File size: 1,555 Bytes
19ea320
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
---
tags:
- fp8
- vllm
---

# Mixtral-8x22B-Instruct-v0.1-FP8

## Model Overview
Mixtral-8x22B-Instruct-v0.1 quantized to FP8 weights and activations using per-tensor quantization, ready for inference with vLLM >= 0.5.0.

## Usage and Creation
Produced using [AutoFP8 with calibration samples from ultrachat](https://github.com/neuralmagic/AutoFP8/blob/147fa4d9e1a90ef8a93f96fc7d9c33056ddc017a/example_dataset.py).

## Evaluation

### Open LLM Leaderboard evaluation scores
|                      | Meta-Llama-3-70B-Instruct | Meta-Llama-3-70B-Instruct-FP8<br>(this model) |
| :------------------: | :----------------------: | :------------------------------------------------: |
| arc-c<br>25-shot     | 71.58                    | 72.09                                              |
| hellaswag<br>10-shot | 86.94                    | 86.83                                              |
| mmlu<br>5-shot       | 83.97                    | 84.06                                              |
| truthfulqa<br>0-shot | 66.98                    | 66.95                                              |
| winogrande<br>5-shot | 82.79                    | 83.18                                              |
| gsm8k<br>5-shot      | 87.56                    | 88.93                                              |
| **Average<br>Accuracy**  | **79.97**                    |              **80.34**                                     |
| **Recovery**             | **100%**                     |              **100.46%**                                     |