|
--- |
|
title: CustomResnet Cifar10 App |
|
emoji: 📉 |
|
colorFrom: red |
|
colorTo: yellow |
|
sdk: gradio |
|
sdk_version: 4.27.0 |
|
app_file: app.py |
|
pinned: false |
|
--- |
|
# Gradio spaces on Huggingface for inferencing CustomResnet18 trained on CIFAR10 using Pytorch Lightning |
|
|
|
## Basic expectations |
|
- Migrate Custom Resnet18 code from Pytorch to Pytorch Lightning first and then to Spaces such that: |
|
- Migrate model on Lightning from Pytorch |
|
- Use Gradio for deployment of Spaces app |
|
|
|
- Spaces app has these features: |
|
- Ask the user whether he/she wants to see GradCAM images |
|
- How many |
|
- From which layer |
|
- Allow opacity change as well |
|
- Ask whether he/she wants to view misclassified images |
|
- How many |
|
- Allow users to upload new images |
|
- Provide 10 example images as well |
|
- Ask how many top classes are to be shown (make sure the user cannot enter more than 10) |
|
|
|
|
|
## Reference to the repo used for training the Lightning model |
|
- https://github.com/ChintanShahDS/ERAV2_Lit |
|
- Follow this to try your own Resnet18 model with different hyperparameters and options on Pytorch Lightning |
|
|
