File size: 298 Bytes
b05f831
 
 
 
 
 
 
 
97a9f85
b05f831
ededf3e
1
2
3
4
5
6
7
8
9
10
11
---
license: apache-2.0
tags:
- 2bit
- llama
- XVERSE
---

You can run it on 4G mem GPU,quantize base QuIP# method, a weights-only quantization method that is able to achieve near fp16 performance using only 2 bits per weight.

url:https://github.com/Cornell-RelaxML/quip-sharp/tree/release20231203