Closed damienlancry closed 5 years ago
Merging #44 into dev will increase coverage by
<.01%
. The diff coverage is100%
.
@@ Coverage Diff @@
## dev #44 +/- ##
==========================================
+ Coverage 97.16% 97.17% +<.01%
==========================================
Files 31 31
Lines 1625 1628 +3
==========================================
+ Hits 1579 1582 +3
Misses 46 46
Impacted Files | Coverage Δ | |
---|---|---|
modAL/utils/data.py | 93.75% <100%> (+1.44%) |
:arrow_up: |
Continue to review full report at Codecov.
Legend - Click here to learn more
Δ = absolute <relative> (impact)
,ø = not affected
,? = missing data
Powered by Codecov. Last update fc8e2e0...57ce832. Read the comment docs.
Thanks! I'll review the PR as soon as possible!
I have reviewed it, everything works well! Currently, I am working on replacing the MNIST dataset using torchvision instead of Keras. This would be better for consistency. To do this, I also need to extend the modAL.utils.data_vstack function, since it does not support PyTorch tensors currently.
I have finished the changes I mentioned. Planning to add a corresponding Jupyter notebook soon!
I could try to write the corresponding notebook if you want
Ok, feel free to work on it! I will also get to it soon!
I have finished the changes I mentioned. Planning to add a corresponding Jupyter notebook soon!
When i run python pytorch_integration.py
after pulling your changes i get TypeError: <class 'torch.Tensor'> datatype is not supported
Edit: oh no, my bad, it works just fine
i just pushed a jupyter notebook 100 percent similar to Keras_integration.ipynb
I have added the notebook to the documentation and merged the PR. Thanks for your work! :)
this is a runnable example of modAL using pytorch models, wrapped with skorch. this example is very similar to the one we can find in modAL/examples/keras_integration.py