Intel
/

bert-base-uncased-squad-int8-static-inc

Question Answering

Intel® Neural Compressor

neural-compressor

PostTrainingStatic

Inference Endpoints

Model card Files Files and versions Community

xinhe commited on Aug 10, 2022

Commit

2018193

·

1 Parent(s): d349d5c

Update README.md

Files changed (1) hide show

README.md +1 -1

README.md CHANGED Viewed

@@ -26,7 +26,7 @@ The linear modules **bert.encoder.layer.2.intermediate.dense**, **bert.encoder.l
 |   |INT8|FP32|
 |---|:---:|:---:|
-| **Accuracy (eval-f1)** |0.8730|0.8810|
 | **Model size (MB)**  |139|436|
 ### Load with Intel® Neural Compressor:

 |   |INT8|FP32|
 |---|:---:|:---:|
+| **Accuracy (eval-f1)** |87.3006|88.1030|
 | **Model size (MB)**  |139|436|
 ### Load with Intel® Neural Compressor: