File size: 1,155 Bytes
ae0bb97
 
fe99803
 
 
 
 
 
 
 
ae0bb97
fe99803
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
---
license: apache-2.0
tags:
- int8
- Intel® Neural Compressor
- PostTrainingStatic
datasets: 
- squad
metrics:
- f1
---

# INT8 DistilBERT base uncased finetuned on Squad

### Post-training static quantization

This is an INT8  PyTorch model quantized with [Intel® Neural Compressor](https://github.com/intel/neural-compressor). 

The original fp32 model comes from the fine-tuned model [distilbert-base-uncased-distilled-squad](https://huggingface.co/distilbert-base-uncased-distilled-squad).

The calibration dataloader is the train dataloader. The default calibration sampling size 300 isn't divisible exactly by batch size 8, so the real sampling size is 304.

The linear module **distilbert.transformer.layer.1.ffn.lin2** falls back to fp32 to meet the 1% relative accuracy loss.

### Test result

|   |INT8|FP32|
|---|:---:|:---:|
| **Accuracy (eval-f1)** |86.1069|86.8374|
| **Model size (MB)**  |74.7|265|

### Load with Intel® Neural Compressor:

```python
from neural_compressor.utils.load_huggingface import OptimizedModel
int8_model = OptimizedModel.from_pretrained(
    'Intel/distilbert-base-uncased-distilled-squad-int8-static',
)
```