Update README.md
Browse files
README.md
CHANGED
|
@@ -33,6 +33,9 @@ The following layers are ignored during quantization:
|
|
| 33 |
- `*.o_proj`
|
| 34 |
- `lm_head`
|
| 35 |
|
|
|
|
|
|
|
|
|
|
| 36 |
## Quantization Scripts
|
| 37 |
```
|
| 38 |
cd examples/torch/language_modeling/llm_ptq/
|
|
|
|
| 33 |
- `*.o_proj`
|
| 34 |
- `lm_head`
|
| 35 |
|
| 36 |
+
## Algorithms
|
| 37 |
+
AutoSmoothQuant algorithm is applied in weight-activation quantization for better performance.
|
| 38 |
+
|
| 39 |
## Quantization Scripts
|
| 40 |
```
|
| 41 |
cd examples/torch/language_modeling/llm_ptq/
|