File size: 322 Bytes
d2214d5 380e4d8 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 |
---
license: mit
language:
- en
- fr
- it
- de
- es
library_name: transformers
tags:
- mixtral
- text-generation-inference
---
Attention quantization: HQQ 4-bit, groupsize 64, compress zero, compress scale with groupsize 256 \
Experts quantization: HQQ 3-bit, groupsize 64, compress zero, compress scale with groupsize 128 |