🔥 Features

Hosting the experiments server

Prerequisites

Installation

pip install labml-app

Starting the server

# Start the server on the default port (5005)labml app-server# To start the server on a different port, use the following commandlabml app-server --port PORT

Monitor Experiments

Installation

pip install labml

app_url:http://localhost:{port}/api/v1/default# If you are setting up the project on a different machine, include the following line instead,app_url:http://{server-ip}:{port}/api/v1/default

PyTorch example

fromlabmlimporttracker,experimentwithexperiment.record(name='sample',exp_conf=conf):foriinrange(50):loss,accuracy=train()tracker.save(i, {'loss':loss,'accuracy':accuracy})

Distributed training example

fromlabmlimporttracker,experimentuuid=experiment.generate_uuid()# make sure to sync this in every machineexperiment.create(uuid=uuid,name='distributed training sample',distributed_rank=0,distributed_world_size=8,                  )withexperiment.start():foriinrange(50):loss,accuracy=train()tracker.save(i, {'loss':loss,'accuracy':accuracy})

📚 Documentation

Guides

🖥 Screenshots

Formatted training loop output

Custom visualizations based on Tensorboard logs

Monitoring hardware usage

# Install packages and dependenciespip install labml psutil py3nvml# Start monitoringlabml monitor

Citing

@misc{labml, author = {Varuna Jayasiri, Nipun Wijerathne, Adithya Narasinghe, Lakshith Nishshanke}, title = {labml.ai: A library to organize machine learning experiments}, year = {2020}, url = {https://labml.ai/},}