File size: 381 Bytes
2e7b0b5
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
---
license: apache-2.0
base_model:
- nvidia/OpenReasoning-Nemotron-7B
datasets:
- mit-han-lab/pile-val-backup
---
# OpenReasoning-Nemotron-7B-AWQ

## Method
Quantised using [vllm-project/llm-compressor](https://github.com/vllm-project/llm-compressor.git) and the following configs:
```
recipe = [
    AWQModifier(ignore=["lm_head"], scheme="W4A16_ASYM", targets=["Linear"]),
]
```