Skip to main content

A plugin for Unity ML-Agents to log Steps Per Second (SPS) to TensorBoard

Project description

ML-Agents SPS Plugin

A plugin for Unity ML-Agents to log Steps Per Second (SPS) to TensorBoard and the console, providing insights into training speed and performance. This tool allows users to monitor the number of steps completed per second, helping to optimize training configuration, including the number of parallel executables (--num-envs) and parallel agents/areas.

Example Output

Here’s an example of the output you’ll see in the console with this plugin:

[INFO] Brain. Steps Per Second: 5275.42.
[INFO] Brain. Step: 44020000. Time Elapsed: 235.580 s. Mean Reward: 1379.877. Std of Reward: 184.503. Training.
[INFO] Brain. Step: 44030000. Time Elapsed: 250.285 s. No episode was completed since last summary. Training.
[INFO] Brain. Steps Per Second: 898.65.
[INFO] Brain. Step: 44040000. Time Elapsed: 250.563 s. No episode was completed since last summary. Training.

Features

  • Steps Per Second Logging: Logs the SPS (Steps Per Second) to TensorBoard, making it easy to track training speed over time.
  • Performance Monitoring: Enables users to fine-tune the number of parallel executables and agents to maximize hardware utilization and minimize idle time.
  • TensorBoard Integration: Integrates seamlessly with TensorBoard to visualize SPS, enabling easier comparison and analysis of training performance.

Installation

Install via PyPI in the same python virtual environment that you have mlagents installed:

pip install mlagents_sps_plugin

Run ML-Agents Training as Usual: Just run your mlagents-learn commands as you normally would. The plugin will automatically log the Steps Per Second (SPS) to TensorBoard and the console so you can monitor the performance in real-time.

Contributing

We welcome contributions! If you have ideas for improvements, feel free to submit a pull request or open an issue.

Support

If you appreciate this work or are looking for software and AI development, visit Digiwave for more details on our tech services.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

mlagents_sps_plugin-0.1.0.tar.gz (4.0 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

mlagents_sps_plugin-0.1.0-py3-none-any.whl (4.7 kB view details)

Uploaded Python 3

File details

Details for the file mlagents_sps_plugin-0.1.0.tar.gz.

File metadata

  • Download URL: mlagents_sps_plugin-0.1.0.tar.gz
  • Upload date:
  • Size: 4.0 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/5.1.1 CPython/3.10.12

File hashes

Hashes for mlagents_sps_plugin-0.1.0.tar.gz
Algorithm Hash digest
SHA256 f2efd09c2db003493ca78ac2c674f56e8beb076bdda091667f22b005e708f8aa
MD5 5ac046ae3e59f7e0bf1f676816be7b69
BLAKE2b-256 ed8f613572017632110b55d2ee553d76a47efd30c4343b51a6d920358420ba63

See more details on using hashes here.

File details

Details for the file mlagents_sps_plugin-0.1.0-py3-none-any.whl.

File metadata

File hashes

Hashes for mlagents_sps_plugin-0.1.0-py3-none-any.whl
Algorithm Hash digest
SHA256 8825ae22f8f98dcb88c541c7b215b7034d96f7a44a6fa32bbde47ebee134eb14
MD5 849b92030c9171bc1f1a80ef6f51e90a
BLAKE2b-256 c785bfbaa2747df1343024729f597eae761ab5e609f3d4a373d6da4786ec9b14

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page