Update README.md
Browse files
README.md
CHANGED
|
@@ -22,13 +22,10 @@ Please follow the license of the original model.
|
|
| 22 |
|
| 23 |
please note int2 **may be slower** than int4 on CUDA due to kernel issue.
|
| 24 |
|
| 25 |
-
~~~python
|
| 26 |
~~~python
|
| 27 |
import transformers
|
| 28 |
-
from transformers import AutoModelForCausalLM, AutoTokenizer
|
| 29 |
-
|
| 30 |
-
|
| 31 |
import torch
|
|
|
|
| 32 |
|
| 33 |
quantized_model_dir = "OPEA/DeepSeek-R1-int2-gptq-sym-inc"
|
| 34 |
|
|
|
|
| 22 |
|
| 23 |
please note int2 **may be slower** than int4 on CUDA due to kernel issue.
|
| 24 |
|
|
|
|
| 25 |
~~~python
|
| 26 |
import transformers
|
|
|
|
|
|
|
|
|
|
| 27 |
import torch
|
| 28 |
+
from transformers import AutoModelForCausalLM, AutoTokenizer
|
| 29 |
|
| 30 |
quantized_model_dir = "OPEA/DeepSeek-R1-int2-gptq-sym-inc"
|
| 31 |
|