Commit
•
ac02df7
1
Parent(s):
cd77fc8
Update README.md
Browse files
README.md
CHANGED
@@ -31,7 +31,7 @@ model-index:
|
|
31 |
|
32 |
### Post-training static quantization
|
33 |
|
34 |
-
This is an INT8 PyTorch model quantized with [Intel® Neural Compressor](https://github.com/intel/neural-compressor).
|
35 |
|
36 |
The original fp32 model comes from the fine-tuned model [electra-small-discriminator-mrpc](https://huggingface.co/Intel/electra-small-discriminator-mrpc).
|
37 |
|
@@ -45,11 +45,11 @@ The calibration dataloader is the train dataloader. The default calibration samp
|
|
45 |
| **Accuracy (eval-f1)** |0.9007|0.8983|
|
46 |
| **Model size (MB)** |14|51.8|
|
47 |
|
48 |
-
### Load with
|
49 |
|
50 |
```python
|
51 |
-
from neural_compressor.
|
52 |
-
int8_model =
|
53 |
'Intel/electra-small-discriminator-mrpc-int8-static',
|
54 |
)
|
55 |
```
|
|
|
31 |
|
32 |
### Post-training static quantization
|
33 |
|
34 |
+
This is an INT8 PyTorch model quantized with [huggingface/optimum-intel](https://github.com/huggingface/optimum-intel) through the usage of [Intel® Neural Compressor](https://github.com/intel/neural-compressor).
|
35 |
|
36 |
The original fp32 model comes from the fine-tuned model [electra-small-discriminator-mrpc](https://huggingface.co/Intel/electra-small-discriminator-mrpc).
|
37 |
|
|
|
45 |
| **Accuracy (eval-f1)** |0.9007|0.8983|
|
46 |
| **Model size (MB)** |14|51.8|
|
47 |
|
48 |
+
### Load with optimum:
|
49 |
|
50 |
```python
|
51 |
+
from optimum.intel.neural_compressor.quantization import IncQuantizedModelForSequenceClassification
|
52 |
+
int8_model = IncQuantizedModelForSequenceClassification.from_pretrained(
|
53 |
'Intel/electra-small-discriminator-mrpc-int8-static',
|
54 |
)
|
55 |
```
|