bwang0911 Xenova HF staff commited on
Commit
85c31f6
1 Parent(s): 1b98ccf

Add quantized ONNX weights (#5)

Browse files

- Add quantized ONNX weights (c90170f6077b4b8bf3fea04b21fb90fe5fd8aa0d)


Co-authored-by: Joshua <[email protected]>

onnx/model_quantized.onnx ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5af68309317d5a7a5b63bf1c4e336dc8bd9a9104e8d16eebeb12499f33cb463c
3
+ size 161789773
onnx/quantize_config.json ADDED
@@ -0,0 +1,36 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "per_channel": true,
3
+ "reduce_range": true,
4
+ "per_model_config": {
5
+ "model": {
6
+ "op_types": [
7
+ "Sub",
8
+ "Cast",
9
+ "Unsqueeze",
10
+ "Gather",
11
+ "Sqrt",
12
+ "Equal",
13
+ "Range",
14
+ "Erf",
15
+ "Where",
16
+ "Pow",
17
+ "Expand",
18
+ "ConstantOfShape",
19
+ "Softmax",
20
+ "Abs",
21
+ "Div",
22
+ "MatMul",
23
+ "Concat",
24
+ "ReduceMean",
25
+ "Mul",
26
+ "Transpose",
27
+ "Constant",
28
+ "Slice",
29
+ "Reshape",
30
+ "Shape",
31
+ "Add"
32
+ ],
33
+ "weight_type": "QInt8"
34
+ }
35
+ }
36
+ }