Torch-snippets

Info, Debug, Warn and Excep will format ouputs separated by a `;` when args are passed
`notify_waiting` is a new function that letting you know some process is running for an unknown amount of time
optional `delay` during `shutdown_current_notebook`
Info, Debug, Warn and Excep will all have `X_mode` and `in_X_mode` functions much like in_debug_mode and debug_mode
`__init__` will auto pull from logger now
Better non-linear `Timer` (and `Report` and `track2`)

🧹 `Info`, `Debug`, `Warn` and `Excep` will accept args (instead of a single arg)
🧹 `show` will show h4 headers instead of h2 for dataframe titles

🧹 `phasify` loads by default
🧹 `show_big_dataframe` can show more rows
🎉 add a new submodule `trainer.hooks`
🧹 `show` delegated kwargs to `plt.imshow` for a better readme
🎉 `batchify` can batchify multiple containers at once
🎉 `cat_with_padding` new function in `torch_loader`
🧹 `L` is json compatible
🐞 `BB` will not decide if something is relative/absolute
🎉 `__contains__` for config
🎉 `to` works on `AttrDict`
👶🏼 `track2` is a better version of `track` uses corouties
👶🏼 `debug_mode` temporarily activates `DEBUG` mode on
👶🏼 `if in_debug_mode():` lets you know if `DEBUG` mode is on
🧹 `reset_logger` can accept lowercase levels also
🧹 `dumpdill` will return a Path after dumping

bugfix in `loader.show`
add `today` function to dates
add `are_equal_dates` to dates
add dpi option to pdf

bugfix in attrdict.map

All notebooks are formatted with black
`parse` can parse python expressions
Add DeepLearningConfig class that can be used to load model hyperparameters
Add GenericConfig class that can be used to load generic (such as training, evaluation) hyperparameters
Add date utilities
`patch_to`, `Timer`, `timeit`, `io` are loaded by default
`lovely_tensors` is optional
Add phasify function to loader