Unable to deploy model on Heroku

Part 1 (2020)
1

I’ve been trying for a few days to deploy my initial model on Heroku without anysuccess. It runs fine on Gradient (with voila/render etc), but when deploying on Heroku I get the following error message:

“There was an error when executing cell [3]. Please run Voilà with --debug to see the error message.”

Line 3 is just: learn_inf = load_learner(‘export.pkl’)

[I tried with the path and without it. The model “export.pkl” is in the GIT repository]

I’ll share below a few print screens of the main file, procfile and requirements.txt. Any hints will be greatly appreciated!

Main file:

image
image1095×432 19.3 KB

requirements.txt:

https://download.pytorch.org/whl/cpu/torch-1.6.0%2Bcpu-cp36-cp36m-linux_x86_64.whl
https://download.pytorch.org/whl/cpu/torchvision-0.7.0%2Bcpu-cp36-cp36m-linux_x86_64.whl
fastai==1.0.61
voila
ipywidgets

procfile

web: voila --port=$PORT --no-browser --enable_nbextensions=True Quick_Snake_Classifier_App.ipynb

2

==1.0.61 is for fastai version 1, you should remove that

4

Thanks for the information keep sharing and it help me to increase my knoledge.

mygiftcardsite

5

How to Deploy Fast.ai Models? (Voilà , Binder and Heroku)

Medium article:
https://medium.com/unpackai/how-to-deploy-fast-ai-models-8704ea711ad2

code:

I hope this may help.