Update README.md
Browse files
README.md
CHANGED
|
@@ -13,9 +13,11 @@ from hqq.utils.patching import prepare_for_inference
|
|
| 13 |
|
| 14 |
|
| 15 |
### Loading Weights
|
|
|
|
| 16 |
model = HQQModelForCausalLM.from_quantized("rohitg/Mixtral-8x22B-Instruct-v0.1-hf-4bit_g64-HQQ", device='cuda')
|
| 17 |
tokenizer = AutoTokenizer.from_pretrained('mistralai/Mixtral-8x22B-Instruct-v0.1')
|
| 18 |
prepare_for_inference(model, backend="torchao_int4")
|
|
|
|
| 19 |
|
| 20 |
### Text Generation
|
| 21 |
|
|
|
|
| 13 |
|
| 14 |
|
| 15 |
### Loading Weights
|
| 16 |
+
```python
|
| 17 |
model = HQQModelForCausalLM.from_quantized("rohitg/Mixtral-8x22B-Instruct-v0.1-hf-4bit_g64-HQQ", device='cuda')
|
| 18 |
tokenizer = AutoTokenizer.from_pretrained('mistralai/Mixtral-8x22B-Instruct-v0.1')
|
| 19 |
prepare_for_inference(model, backend="torchao_int4")
|
| 20 |
+
```
|
| 21 |
|
| 22 |
### Text Generation
|
| 23 |
|