tf_conversion_skeleton

Playground for creating TF model, converting to tfjs, and displaying in webapp

Backend Set up

Install Jupyter base environment, with kernals in other environments

# if you haven't already installed jupyter
conda install nb_conda_kernels
conda install jupyterlab
conda install notebook
# optionally, install extensions
conda install jupyter_contrib_nbextensions

Start a new environment

# use one of python=3.[8,9,10]
conda create -n TF python=3.10 pip ipykernel
conda activate TF

Install TensorFlow. Below are for macOS only. For other OS, follow instructions from tensorflow.

# only if using Apple Silicon
conda install -c apple tensorflow-deps

# on macos, install base TensorFlow
python -m pip install tensorflow-macos

# on macos, install metal plugin (gpu support)
python -m pip install tensorflow-metal

Install tfjs_converter

python -m pip install tensorflowjs

Frontend Set up

Usage

Make sure you have yarn, npm, or pnpm

$ yarn install # or pnpm install or npm install

Available Scripts

In the project directory, you can run:

yarn dev or yarn start

Runs the app in the development mode.
Open http://localhost:3000 to view it in the browser.

The page will reload if you make edits.

npm run build

Builds the app for production to the dist folder.
It correctly bundles Solid in production mode and optimizes the build for the best performance.

The build is minified and the filenames include the hashes.
Your app is ready to be deployed!

Deployment

You can deploy the dist folder to any static host provider (netlify, surge, now, etc.)