Upload 2 files
Browse files- PyTorchTraining.py +24 -0
- pytorch_model.pt +3 -0
PyTorchTraining.py
ADDED
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import torch.onnx
|
2 |
+
|
3 |
+
#Function to Convert to ONNX
|
4 |
+
def Convert_ONNX():
|
5 |
+
|
6 |
+
# set the model to inference mode
|
7 |
+
model.eval()
|
8 |
+
|
9 |
+
# Let's create a dummy input tensor
|
10 |
+
dummy_input = torch.randn(1, input_size, requires_grad=True)
|
11 |
+
|
12 |
+
# Export the model
|
13 |
+
torch.onnx.export(model, # model being run
|
14 |
+
dummy_input, # model input (or a tuple for multiple inputs)
|
15 |
+
"ImageClassifier.onnx", # where to save the model
|
16 |
+
export_params=True, # store the trained parameter weights inside the model file
|
17 |
+
opset_version=10, # the ONNX version to export the model to
|
18 |
+
do_constant_folding=True, # whether to execute constant folding for optimization
|
19 |
+
input_names = ['modelInput'], # the model's input names
|
20 |
+
output_names = ['modelOutput'], # the model's output names
|
21 |
+
dynamic_axes={'modelInput' : {0 : 'batch_size'}, # variable length axes
|
22 |
+
'modelOutput' : {0 : 'batch_size'}})
|
23 |
+
print(" ")
|
24 |
+
print('Model has been converted to ONNX')
|
pytorch_model.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:89c70e808d1783b6c07911306e106aaf0d4f7f3da8c61078b99ff7f8929a26f4
|
3 |
+
size 116861841
|