Update README.md
Browse files
README.md
CHANGED
|
@@ -29,13 +29,11 @@ effi 7b AWQ is a quantized version of effi 7b whiich is a 7 billion parameter mo
|
|
| 29 |
|
| 30 |
### Qunatization Configuration
|
| 31 |
|
| 32 |
-
|
| 33 |
-
|
| 34 |
-
|
| 35 |
-
|
| 36 |
-
|
| 37 |
-
"modules_to_not_convert": null
|
| 38 |
-
|
| 39 |
|
| 40 |
|
| 41 |
|
|
@@ -79,4 +77,16 @@ print(f"{tokenizer.batch_decode(outputs.detach().cpu().numpy(), skip_special_tok
|
|
| 79 |
|
| 80 |
### Framework versions
|
| 81 |
- Transformers 4.37.2
|
| 82 |
-
- Autoawq 0.1.8
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 29 |
|
| 30 |
### Qunatization Configuration
|
| 31 |
|
| 32 |
+
-zero_point: true,
|
| 33 |
+
-q_group_size: 128,
|
| 34 |
+
-w_bit: 4,
|
| 35 |
+
-version: "GEMM",
|
| 36 |
+
-modules_to_not_convert: null
|
|
|
|
|
|
|
| 37 |
|
| 38 |
|
| 39 |
|
|
|
|
| 77 |
|
| 78 |
### Framework versions
|
| 79 |
- Transformers 4.37.2
|
| 80 |
+
- Autoawq 0.1.8
|
| 81 |
+
|
| 82 |
+
### Citation
|
| 83 |
+
|
| 84 |
+
```
|
| 85 |
+
@misc {bhavyaaiplanet,
|
| 86 |
+
author = { {Bhavya Bhola} },
|
| 87 |
+
title = { Quantized version of effi-7bb by AI Planet},
|
| 88 |
+
year = 2024,
|
| 89 |
+
url = { https://huggingface.co/aiplanet/effi-7b-awq },
|
| 90 |
+
publisher = { Hugging Face }
|
| 91 |
+
}
|
| 92 |
+
```
|