Retro Speedlab Library

Reinforcement-learning engine for training recurrent agents on classic video games.

Python 3.12 License: GPL-3.0

The reusable reinforcement-learning engine behind Retro Speedlab.

It combines visual observations, emulator RAM, recurrent PPO, Random Network Distillation (RND), vectorized environments, resumable training, replay capture, and live telemetry in a reproducible Python training system.

Highlights

  • Recurrent CNN-LSTM PPO for partially observable environments
  • Multi-input policies combining RGB frames and emulator RAM
  • Adaptive Random Network Distillation (RND) for sparse-reward exploration
  • Parallel vectorized emulator environments
  • Automatic worker selection, CUDA tuning, and CPU fallback
  • Atomic, resumable checkpoints including exploration state
  • .bk2 episode recording
  • Live Vue-based training telemetry
  • Reusable abstractions for RAM, game states, rewards, and actions

Looking for game runners, end-to-end examples, and user-facing documentation? Start with Retro Speedlab.

Architecture

A training run combines game-specific definitions with reusable environment, model, training, persistence, and monitoring components.

                         Retro Speedlab
                              │
                    ┌─────────▼─────────┐
                    │  Game Definition  │
                    │                   │
                    │ RAM · Rewards ·   │
                    │ States · Actions  │
                    └─────────┬─────────┘
                              │
                    ┌─────────▼─────────┐
                    │ Vectorized Retro  │
                    │   Environments    │
                    └─────────┬─────────┘
                              │
                 ┌────────────┴────────────┐
                 │                         │
             RGB Frames                 RAM State
                 │                         │
             CNN Encoder                 Encoder
                 │                         │
                 └────────────┬────────────┘
                              │
                             LSTM
                              │
                    ┌─────────▼─────────┐
                    │  Recurrent PPO    │
                    │      + RND        │
                    └─────────┬─────────┘
                              │
               ┌──────────────┼──────────────┐
               │              │              │
          Checkpoints      Replays       Telemetry
                                             │
                                     ┌───────▼────────┐
                                     │ Live Dashboard │
                                     └────────────────┘

Project links

Key dates

PyPI data
Data sourced directly from PyPI's database.
  • Released:
Newer release available (3.0.8)

1 maintainer

PyPI data
Data sourced directly from PyPI's database.
Avatar for Datenwissenschaften from gravatar.com Datenwissenschaften

Credits

Author: datenwissenschaften

License expression

GPL-3.0-only
View SPDX License List

Requires

Python <3.13, >=3.12

Tags

reinforcement-learning retro-gaming stable-baselines3 stable-retro ai gymnasium

Classifiers

Report project as malware

Metadata

Project links

Key dates

PyPI data
Data sourced directly from PyPI's database.
  • Released:
Newer release available (3.0.8)

1 maintainer

PyPI data
Data sourced directly from PyPI's database.
Avatar for Datenwissenschaften from gravatar.com Datenwissenschaften

Credits

Author: datenwissenschaften

License expression

GPL-3.0-only
View SPDX License List

Requires

Python <3.13, >=3.12

Tags

reinforcement-learning retro-gaming stable-baselines3 stable-retro ai gymnasium

Classifiers

Report project as malware

Release files for datenwissenschaften 2.10.11

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for datenwissenschaften 2.10.11
File Size Uploaded
datenwissenschaften-2.10.11.tar.gz 133.8 kB Details

Built distribution (wheel)

Table of built distributions (wheels) for datenwissenschaften 2.10.11
File Interpreter ABI Platform
datenwissenschaften-2.10.11-py3-none-any.whl 158.2 kB Python 3 none any Details

Total release size: 292.0 kB

Release files / datenwissenschaften-2.10.11.tar.gz

Download URL datenwissenschaften-2.10.11.tar.gz
Size 133.8 kB
Tags Source
SHA-256 checksum
How to use checksums 		dc85a44423c45ab5e411e72a28942b1a8e3eb5ff2b8a9390d61db070796fa3f1
BLAKE2b-256 checksum
How to use checksums 		15306108917237c585c2584a00696e4c5f6ba5b7f69ac2aaaa6e0804e49f7b6c
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing? 		No
Uploaded via poetry/2.5.1 CPython/3.12.14 Linux/6.17.0-1022-azure

Release files / datenwissenschaften-2.10.11-py3-none-any.whl

Download URL datenwissenschaften-2.10.11-py3-none-any.whl
Size 158.2 kB
Tags Python 3
SHA-256 checksum
How to use checksums 		e88be4dccf5452ef431ebaa874dbb9b9724fa53e547167ad4d4d80fae4505701
BLAKE2b-256 checksum
How to use checksums 		a7203e571b08cf4b0ba4fea424baf49855b9598778cb006a700d289aacdc6ce9
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing? 		No
Uploaded via poetry/2.5.1 CPython/3.12.14 Linux/6.17.0-1022-azure

Release history Release notifications | RSS feed

3.0.8

2 release files

3.0.7

2 release files

3.0.6

2 release files

3.0.5

2 release files

3.0.4

2 release files

3.0.3

2 release files

3.0.2

2 release files

3.0.1

2 release files

3.0.0

2 release files

2.10.18

2 release files

2.10.17

2 release files

2.10.16

2 release files

2.10.15

2 release files

2.10.14

2 release files

2.10.13

2 release files

2.10.12

2 release files

This release

2.10.11 This release

2 release files

2.10.10

2 release files

2.10.9

2 release files

2.10.8

2 release files

2.10.7

2 release files

2.10.6

2 release files

2.10.5

2 release files

2.10.4

2 release files

2.10.3

2 release files

2.10.2

2 release files

2.10.1

2 release files

2.9.33

2 release files

2.9.32

2 release files

2.9.31

2 release files

2.9.30

2 release files

2.9.29

2 release files

2.9.28

2 release files

2.9.27

2 release files

2.9.26

2 release files

2.9.25

2 release files

2.9.24

2 release files

2.9.23

2 release files

2.9.22

2 release files

2.9.21

2 release files

2.9.20

2 release files

2.9.19

2 release files

2.9.18

2 release files

2.9.17

2 release files

2.9.16

2 release files

2.9.15

2 release files

2.9.14

2 release files

2.9.13

2 release files

2.9.12

2 release files

2.9.11

2 release files

2.9.10

2 release files

2.9.9

2 release files

2.9.8

2 release files

2.9.7

2 release files

2.9.6

2 release files

2.9.5

2 release files

2.9.4

2 release files

2.9.3

2 release files

2.9.2

2 release files

2.9.1

2 release files

2.8.4

2 release files

2.8.3

2 release files

2.8.2

2 release files

2.8.1

2 release files

2.7.4

2 release files

2.7.3

2 release files

2.7.2

2 release files

2.7.1

2 release files

2.6.2

2 release files

2.6.1

2 release files

2.5.4

2 release files

2.5.3

2 release files

2.5.2

2 release files

2.5.1

2 release files

2.5.0

2 release files

2.4.0

2 release files

2.3.0

2 release files

2.2.0

2 release files

2.1.0

2 release files

2.0.0

2 release files

1.11.9

2 release files

1.11.5

2 release files

1.11.4

2 release files

1.11.3

2 release files

1.11.2

2 release files

1.11.1

2 release files

1.11.0

2 release files

1.10.11

2 release files

1.10.10

2 release files

1.10.9

2 release files

1.10.8

2 release files

1.10.7

2 release files

1.10.6

2 release files

1.10.5

2 release files

1.10.4

2 release files

1.10.3

2 release files

1.10.2

2 release files

1.10.1

2 release files

1.10.0

2 release files

1.9.4

2 release files

1.9.3

2 release files

1.9.2

2 release files

1.9.1

2 release files

1.9.0

2 release files

1.8.1

2 release files

1.8.0

2 release files

1.7.5

2 release files

1.7.4

2 release files

1.7.3

2 release files

1.7.2

2 release files

1.7.1

2 release files

1.7.0

2 release files

1.6.1

2 release files

1.6.0

2 release files

1.5.0

2 release files

1.4.6

2 release files

1.4.5

2 release files

1.4.4

2 release files

1.4.3

2 release files

1.4.2

2 release files

1.4.1

2 release files

1.4.0

2 release files

1.3.1

2 release files

1.2.1

2 release files

1.2.0

2 release files

1.1.1

2 release files

1.1.0

2 release files

1.0.2

2 release files

1.0.1

2 release files

1.0.0

2 release files

0.0.0

2 release files

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page