AndrewZhaoLuo commented on issue #14661:
URL: https://github.com/apache/tvm/issues/14661#issuecomment-1524041028
Oddly enough I am not able to reproduce your mod. Your code runs fine on my
machine:
Resulting printout from `irmod`
```
Index with the appropriate dtype instead.
from pandas import MultiIndex, Int64Index
def @main(%x: Tensor[(1, 3, 32, 32), float32] /* ty=Tensor[(1, 3, 32, 32),
float32] span=LpPool_0.x:0:0 */) -> Tensor[(1, 3, 10, 10), float32] {
%0 = power(%x, 2f /* ty=float32 span=LpPool_0:0:0 */) /* ty=Tensor[(1, 3,
32, 32), float32] span=LpPool_0:0:0 */;
%1 = nn.avg_pool2d(%0, pool_size=[5, 5], strides=[3, 3], padding=[0, 0, 0,
0], count_include_pad=True) /* ty=Tensor[(1, 3, 10, 10), float32]
span=LpPool_0:0:0 */;
%2 = abs(%1) /* ty=Tensor[(1, 3, 10, 10), float32] span=LpPool_0:0:0 */;
%3 = multiply(%2, 25f /* ty=float32 span=LpPool_0:0:0 */) /* ty=Tensor[(1,
3, 10, 10), float32] span=LpPool_0:0:0 */;
power(%3, 0.5f /* ty=float32 span=LpPool_0:0:0 */) /* ty=Tensor[(1, 3, 10,
10), float32] span=LpPool_0:0:0 */
}
```
What version of onnx are you using?
```
keras2onnx==1.7.0
onnx==1.12.0
onnx-graphsurgeon==0.3.25
onnx-simplifier==0.3.6
onnx-tf==1.9.0
onnxconverter-common==1.9.0
onnxmltools==1.11.1
onnxoptimizer==0.2.6
onnxruntime==1.13.1
onnxruntime-gpu==1.13.1
onnxruntime-tools==1.7.0
skl2onnx==1.11.2
tf2onnx==1.9.2
tflite2onnx==0.4.0
```
Here is my `pip freeze | grep onnx`
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
