File size: 1,903 Bytes
f5095a3 ccf3347 f5095a3 ccf3347 f5095a3 ccf3347 87b1d6e ccf3347 |
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 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 |
---
language:
- en
license: apache-2.0
tags:
- multiple-choice
- int8
- PostTrainingStatic
datasets:
- swag
metrics:
- accuracy
model-index:
- name: bert-base-uncased-finetuned-swag-int8-static
results:
- task:
name: Multiple-choice
type: multiple-choice
dataset:
name: Swag
type: swag
metrics:
- name: Accuracy
type: accuracy
value: 0.7838148474693298
---
# INT8 bert-base-uncased-finetuned-swag
### 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 [thyagosme/bert-base-uncased-finetuned-swag](https://huggingface.co/thyagosme/bert-base-uncased-finetuned-swag).
The calibration dataloader is the train dataloader. The default calibration sampling size 100 isn't divisible exactly by batch size 8, so the real sampling size is 104.
The linear modules **bert.encoder.layer.2.output.dense, bert.encoder.layer.5.intermediate.dense, bert.encoder.layer.9.output.dense, bert.encoder.layer.10.output.dense** fall back to fp32 to meet the 1% relative accuracy loss.
### Test result
- Batch size = 8
- [Amazon Web Services](https://aws.amazon.com/) c6i.xlarge (Intel ICE Lake: 4 vCPUs, 8g Memory) instance.
| |INT8|FP32|
|---|:---:|:---:|
| **Throughput (samples/sec)** |16.55|9.333|
| **Accuracy (eval-accuracy)** |0.7838|0.7915|
| **Model size (MB)** |133|418|
### Load with Intel® Neural Compressor (build from source):
```python
from neural_compressor.utils.load_huggingface import OptimizedModel
int8_model = OptimizedModel.from_pretrained(
'Intel/bert-base-uncased-finetuned-swag-int8-static',
)
```
Notes:
- The INT8 model has better performance than the FP32 model when the CPU is fully occupied. Otherwise, there will be the illusion that INT8 is inferior to FP32.
|