$ conda install -c conda-forge libiconv #install libiconv a dependency for mamba
$ conda install mamba -n base -c conda-forge # install mamba
$ mamba create -n fastai python=3.9.5 # create fastai env using mamba and latest python
$ mamba install gh --channel conda-forge # install github cli
$ conda activate fastai # activate env
$ mamba install -c conda-forge notebook jupyterlab #install notebook & jupyterlab
$ mamba install -c conda-forge jupyter_contrib_nbextensions #install extensions
$ mamba install -c fastchan nbdev # install nbdev lib for notebook development
$ nbdev_install_git_hooks # install git hooks
$ cd ./nbs
$ ln -s ../fastai fastai # create symbolic link for easier navigation
$ jupyter notebook # start the notebook- Callbacks from fastbook : Tweaking the training process using callbacks. A callback is a piece of code that you write, and inject into another piece of code at some predefined point (hooks). These predefined points are the events in the training loop.
| Callback | Description |
|---|---|
| ModelResetter | reset method of the model called at the start of training & validation loop |
| RNNRegularizer | impl after_pred and after_loss reads the attrs such as model, pred and write to attrs such as self.learn.pred or self.learn.loss |