Polyaxon allows to schedule Fastai experiments, and supports tracking metrics, outputs, and models.

With Polyaxon you can:

  • log hyperparameters for every run
  • see learning curves for losses and metrics during training
  • see hardware consumption and stdout/stderr output during training
  • log images, charts, and other assets
  • log git commit information
  • log env information
  • log model
  • ...

Tracking API

Polyaxon provides a tracking API to track experiment and report metrics, artifacts, logs, and results to the Polyaxon dashboard.

You can use the tracking API to create a custom tracking experience with Fastai.

Setup

pip install polyaxon

Initialize your script with Polyaxon

This is an optional step if you need to perform some manual tracking or to track some information before passing the callback.

from polyaxon import tracking

tracking.init(...)

Polyaxon callback

Polyaxon provides a Fastai callback, you can use this callback with your experiment to report metrics automatically.

from polyaxon.tracking.contrib.fastai import PolyaxonCallback

# To log only during one training phase
learn.fit(..., cbs=[PolyaxonCallback()])

# To log continuously for all training phases
learn = learner(..., cbs=[PolyaxonCallback()])

Manual logging

If you want to have more control and use Polyaxon to log metrics in your custom Fastai training loops:

from polyaxon import tracking

# Log your metrics
tracking.log_metrics(metric1=value1, metric2=value2, ...)

Example

In this example we will go through the process of logging a FastAI model using Polyaxon's callback.

This example can be used with the offline mode POLYAXON_OFFLINE=true and it does not require a Polyaxon API to run locally.

To see how this can be turned to a declarative approach to be submitted to a Polyaxon cluster, please check this example

import argparse

from fastai.vision.all import *
from fastai.basics import *

# Polyaxon
from polyaxon.tracking.contrib.fastai import PolyaxonCallback

path = untar_data(URLs.MNIST_SAMPLE)
items = get_image_files(path)
tds = Datasets(items, [PILImageBW.create, [parent_label, Categorize()]], splits=GrandparentSplitter()(items))
dls = tds.dataloaders(bs=32, after_item=[ToTensor(), IntToFloatTensor()])

# create a learner with gradient accumulation
learn = cnn_learner(
    dls,
    resnet18,
    loss_func=CrossEntropyLossFlat(),
    cbs=[PolyaxonCallback()]  # Polyaxon
)

if __name__ == '__main__':
    parser = argparse.ArgumentParser()
    parser.add_argument('--fit', type=int, default=2)
    args = parser.parse_args()
    learn.fit(args.fit)

Fastai V1

If you are using Fastai v1, you will need to import the callback for the v1 version

from polyaxon.tracking.contrib.fastai_v1 import PolyaxonCallback

# Usage as a fit callback
learn.fit_one_cycle(1, 0.02, callbacks=[PolyaxonCallback(learn=learn, monitor='accuracy')])

# Usage as a partial function
Learner(..., callback_fns=partial(PolyaxonFastai, ...), ...)