uploaded the pickle file model using git-lfs because the fill is too big, but when i try to load in program i found this error
thank you for your help
-python version is 3.8.7
-streamlit version is 0.82.0
Hi @Roy_Noviantho,
Thank you for sharing with the Streamlit community! Would you be able to either post a code snippet which contains your usage of āpickleā or a link to your app?
Best,
Caroline
hi @Caroline,
this is my code when load pickle:
vectorizer = pickle.load(open('data/vectorizer6.pkl', 'rb')) * model = pickle.load(open('data/model6.pkl', 'rb'))*
i tried with model size <= 25mb and perfectly can load the model
but when i tried with the mode size >= 25mb which is upload with git lfs, get error UnpicklingError: invalid load key, āvā
Hi @Roy_Noviantho,
I would recommend using Git LFS for the larger file.
You might also want to check out these relevant threads:
- Help with āUnpicklingError: invalid load key, āvā.ā
- Getting UnpicklingError: invalid load key, āvā
- Getting an error āUnpicklingError: invalid load key, āvā.ā in Pytorch model deploying in Streamlit
- Best practices for fetching artifacts requried for the app?
Best,
Caroline