File size: 1,761 Bytes
3b64d42
 
 
 
 
 
 
 
d1352ff
 
3b64d42
 
 
72ea121
3b64d42
 
d1352ff
 
 
72ea121
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Official [AQLM](https://arxiv.org/abs/2401.06118) quantization of `meta-llama/Llama-2-7b-hf`.

For this quantization, we used 1 codebook of 16 bits.

Selected evaluation results for this and other models: 

| Model      | AQLM scheme | WikiText 2 PPL | Model size, Gb | Hub link                                                                 |
|------------|-------------|----------------|----------------|--------------------------------------------------------------------------|
| Llama-2-7b (THIS) | 1x16        | 5.92           | 2.4            | [Link](https://huggingface.co/BlackSamorez/Llama-2-7b-AQLM-2Bit-1x16-hf) |
| Llama-2-7b | 2x8         | 6.69           | 2.2            | [Link](https://huggingface.co/BlackSamorez/Llama-2-7b-AQLM-2Bit-2x8-hf)  |
| Llama-2-7b | 8x8         | 7.83           | 2.2            | [Link](https://huggingface.co/BlackSamorez/Llama-2-7b-AQLM-2Bit-8x8-hf)  |
| Llama-2-13b| 1x16        | 5.41           | 4.1            | [Link](https://huggingface.co/BlackSamorez/Llama-2-13b-AQLM-2Bit-1x16-hf)|
| Llama-2-70b| 1x16        | 3.96           | 18.8           | [Link](https://huggingface.co/BlackSamorez/Llama-2-70b-AQLM-2Bit-1x16-hf)|
| Llama-2-70b| 2x8         | 4.83           | 18.2           | [Link](https://huggingface.co/BlackSamorez/Llama-2-70b-AQLM-2Bit-2x8-hf) |
| Mixtral-8x7b| 1x16       | 4.37           | 12.6            | [Link](https://huggingface.co/BlackSamorez/Mixtral-8x7b-AQLM-2Bit-1x16-hf)|

**UPD** (20.02.2024).
We applied global finetuning on top of quantized model and improved results compared to first revision.

To learn more about the inference, as well as the information on how to quantize models yourself, please refer to the [official GitHub repo](https://github.com/Vahe1994/AQLM).