violetch24 commited on
Commit
515ee7b
1 Parent(s): f58e693

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +43 -0
README.md CHANGED
@@ -1,3 +1,46 @@
1
  ---
 
 
 
 
 
 
 
 
 
 
 
 
2
  license: apache-2.0
 
 
 
 
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
+ language: en
3
+ tags:
4
+ - bert
5
+ - rte
6
+ - glue
7
+ - torchdistill
8
+ - nlp
9
+ - int8
10
+ - neural-compressor
11
+ - Intel® Neural Compressor
12
+ - text-classfication
13
+ - PostTrainingStatic
14
  license: apache-2.0
15
+ datasets:
16
+ - rte
17
+ metrics:
18
+ - f1
19
  ---
20
+
21
+ # INT8 bert-large-uncased-rte-int8-static
22
+
23
+ ## Post-training static quantization
24
+
25
+ ### PyTorch
26
+
27
+ This is an INT8 PyTorch model quantized with [Intel® Neural Compressor](https://github.com/intel/neural-compressor).
28
+
29
+ The original fp32 model comes from the fine-tuned model [yoshitomo-matsubara/bert-large-uncased-rte](https://huggingface.co/yoshitomo-matsubara/bert-large-uncased-rte).
30
+
31
+ #### Test result
32
+
33
+ | |INT8|FP32|
34
+ |---|:---:|:---:|
35
+ | **Accuracy (eval-f1)** |0.7365|0.7401|
36
+ | **Model size (MB)** |1244|1349|
37
+
38
+ #### Load with Intel® Neural Compressor:
39
+
40
+ ```python
41
+ from optimum.intel.neural_compressor.quantization import IncQuantizedModelForSequenceClassification
42
+
43
+ int8_model = IncQuantizedModelForSequenceClassification.from_pretrained(
44
+ "Intel/bert-large-uncased-rte-int8-static",
45
+ )
46
+ ```