Update README.md

README.md

@@ -1,3 +1,60 @@
 ---
+language:
+- en
 license: mit
+tags:
+- text-classfication
+- int8
+- PostTrainingStatic
+datasets:
+- glue
+metrics:
+- f1
+model-index:
+- name: xlnet-base-cased-mrpc-int8-static
+  results:
+  - task:
+      name: Text Classification
+      type: text-classification
+    dataset:
+      name: GLUE MRPC
+      type: glue
+      args: mrpc
+    metrics:
+    - name: F1
+      type: f1
+      value: 0.8892794376098417
 ---
+# INT8 xlnet-base-cased-mrpc
+
+###  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 [xlnet-base-cased-mrpc](https://huggingface.co/Intel/xlnet-base-cased-mrpc).
+
+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.
+
+### 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)**  |12.506|9.436|
+| **Accuracy (eval-f1)** |0.8893|0.8897|
+| **Model size (MB)**  |215|448|
+
+### Load with Intel® Neural Compressor (build from source):
+
+```python
+from neural_compressor.utils.load_huggingface import OptimizedModel
+int8_model = OptimizedModel.from_pretrained(
+    'Intel/xlnet-base-cased-mrpc-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.
+