|
--- |
|
license: apache-2.0 |
|
tags: |
|
- int8 |
|
- Intel® Neural Compressor |
|
- PostTrainingStatic |
|
datasets: |
|
- mnli |
|
metrics: |
|
- accuracy |
|
--- |
|
|
|
# INT8 T5 small finetuned on XSum |
|
|
|
### Post-training dynamic 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 [adasnew/t5-small-xsum](https://huggingface.co/adasnew/t5-small-xsum). |
|
|
|
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 **lm.head**, fall back to fp32 for less than 1% relative accuracy loss. |
|
|
|
### Evaluation result |
|
|
|
| |INT8|FP32| |
|
|---|:---:|:---:| |
|
| **Accuracy (eval-rouge1)** | 29.9008 |29.9592| |
|
| **Model size** |154M|242M| |
|
|
|
### Load with Intel® Neural Compressor: |
|
|
|
```python |
|
from neural_compressor.utils.load_huggingface import OptimizedModel |
|
int8_model = OptimizedModel.from_pretrained( |
|
'Intel/roberta-base-squad2-int8-static', |
|
) |
|
``` |
|
|
