FastAI Training metrics as Notification on mobile
Get the model updates on mobile as notifications — this would work on any of the Cloud providers or even Google Colab
Ever had to train a machine learning model for so long that you had to sit and stare at the screen every time an epoch got over to have a look at the model metrics?
Forget it! using the combined power of Fastai callbacks and pushover.net you can get training notifications right on your mobile phone
This is possible because of the rich callbacks system in FastAI library
Lets split our tutorial into three sections
- pushover.net configuration on mobile
- Pneumonia detection model training with Fastai
Github repo link → pneumonia-detection-fastai
STEP 1 -> Setting up a workspace on your favourite cloud provider
Open the terminal from Jupyter lab and clone the repo git clone https://github.com/zszazi/pneumonia-detection-fastai.git to your workspace
- now open the Jupyter notebook by name
pneumonia_prediction_using_fastau_with_mobile_notification.ipynb
STEP 2 -> Pushover.net configuration on mobile
Head over to https://pushover.net and set up an account and download the mobile app too, You will only be able to receive notifications on mobile if you have their app installed. They have extensive docs and faq which cover on setting up, to receive notifications on your pushover mobile app.
pushover currently offers a 7 day trial on signup accross all platforms
Copy the API keys; there are two of these found in your account section
"user" : "your key",
"token" : "Your key",STEP 3 -> Training the model and getting notifications on mobile
To run the cell press Shift+Enter
- Make sure to upload the kaggle.json (watch this Youtube video)
- We can use the upload button in Jupyter workspace to upload the
kaggle.jsonfile
- On running the cell containing
learn.fit_one_cycle()you should have got the push notifications on your mobile, and this would be sent for each epoch
🔔 Training Notifications on mobile 🔔
What’s more? You can even send your loss plots as 🔔 to mobile.
Here’s a Github gist showing you the central part of mobile callback notification
code
GitHub gist
FastAI provides you with a rich collection of callbacks (see pic), but we mainly have used the following ones: whose name are self-explanatory :)
- on_train_begin
- on_epoch_begin
- on_train_end
