Closed roboserg closed 2 years ago
I also tried using the batch_size:
Can you post your full model code? Your model likely accepts a list instead of a tensor as input. Try using the input_data
param instead with your actual input data tensor for better results.
Here you go - https://colab.research.google.com/drive/1wF3XyaN_WL7mR0qmWj6w7--pCC6TCYUF?usp=sharing
I changed the model to YOLOX and now getting a different error -
Is there a way to not supply input size at all? In Keras the summary is printed automatically, without figuring out the input size
You should be using pip install torchinfo
, not torchsummary
.
Yes, you can just not pass the input_size parameter and get some limited results.
If your forward function needs additional parameters, you should pass them as kwargs:
summary(model, input_size=(3, 512, 512), img_metas=...)
Closed due to inactivity
Maybe I supply the wrong input_size? My model: