Not sure how to save model in h5 format

[guen_gn (nguyen gn)]

I am using following as reference. It appears there are two main ways to save the model:
https://pytorch.org/tutorials/beginner/saving_loading_models.html

• save dictionary state: recommended way.
• save entire model: not recommended as path is saved and can break if it is opened somewhere else.

TF can conveniently store in h5 format which I can dump quickly using h5 (hdf5) format, but this is not possible with Pytorch?

[guen_gn (nguyen gn)]

At least according to this, hdf5 format is not great as it accumulates over time. Is there a better format these days?

Save torch tensors as hdf5

HDF5 is not a great format for appending information over-time… It will end up generating a very large binary file to handle new data. I’d recommend doing it for a fixed size. E.g. first create a dataset of a fixed size: N = 100 # find the length of my dataset data = h5_file.create_dataset('data', shape=(N, 3, 224, 224), dtype=np.float32, fillvalue=0) Then populate it for i in range(N): img = ... # load image data[i] = img h5_file.close() However, if you really really want to do re…

[Programmer-RD-AI]

Pytorch usually use .pth or .pt

those are the formats that are mostly used in pytorch