---
title: Cats vs Dogs Classifier
sdk: gradio
sdk_version: 4.44.1
app_file: app.py
pinned: false
---

# Cats vs Dogs Classifier

Gradio demo Space for an image classifier trained to distinguish cats vs dogs.

## Configure

- Put your Keras `.h5` model in this repo (recommended: `model.h5`), or set `MODEL_PATH`.
- `model_info.json` is optional, but if present it can provide defaults like class names / preprocessing.
- If your model outputs a single sigmoid probability, set `POSITIVE_CLASS` to whichever label corresponds to `1.0`.
- If your model needs different input scaling, set `INPUT_SCALE` to `0_1`, `minus1_1`, or `0_255`.

## Upload to Hugging Face Spaces

1. Create a new **Space → Gradio** (Public).
2. Add this repo contents to the Space repo.
3. Add your `model.h5`.
   - If it's larger than ~10MB, use Git LFS:
     - `git lfs install`
     - `git lfs track "*.h5"`
     - commit the generated `.gitattributes`, then commit `model.h5`.
4. (Optional) In Space **Settings → Variables**, set:
   - `CLASS_NAMES` (default: `cat,dog`)
   - `INPUT_SCALE` (`0_1` | `minus1_1` | `0_255`)
   - `POSITIVE_CLASS` (only for 1-output sigmoid models)