From 2581c549f98ea06b4271c84b8e7fa1a6443e1542 Mon Sep 17 00:00:00 2001 From: "copilot-swe-agent[bot]" <198982749+Copilot@users.noreply.github.com> Date: Sat, 1 Nov 2025 04:02:35 +0000 Subject: [PATCH 1/3] Initial plan From f7bb74101e0b178c8467c7c169d88a4bb7f63d07 Mon Sep 17 00:00:00 2001 From: "copilot-swe-agent[bot]" <198982749+Copilot@users.noreply.github.com> Date: Sat, 1 Nov 2025 04:07:19 +0000 Subject: [PATCH 2/3] Add comprehensive environments table to main README - Added new "Available Environments" section after Featured Example - Table includes 10 environments with domain, description, and links - Organized by purpose: testing, code execution, LLM training, games, dev tools, finance, traffic, safety - All links verified to point to correct environment READMEs Co-authored-by: init27 <16226196+init27@users.noreply.github.com> --- README.md | 19 +++++++++++++++++++ 1 file changed, 19 insertions(+) diff --git a/README.md b/README.md index c4cd0a08..a58e3b5d 100644 --- a/README.md +++ b/README.md @@ -10,6 +10,25 @@ An e2e framework for creating, deploying and using isolated execution environmen **🚀 Featured Example:** Train LLMs to play BlackJack using [torchforge](https://github.com/meta-pytorch/torchforge) (PyTorch's agentic RL framework): [`examples/grpo_blackjack/`](examples/grpo_blackjack/) +## Available Environments + +OpenEnv provides a growing collection of environments across different domains for agentic RL research: + +| Environment | Domain | Description | Documentation | +|-------------|--------|-------------|---------------| +| **Echo Env** | Testing | Simple message echo for testing framework infrastructure | [`src/envs/echo_env/`](src/envs/echo_env/README.md) | +| **Coding Env** | Code Execution | Execute Python code in sandboxed environment with smolagents | [`src/envs/coding_env/`](src/envs/coding_env/README.md) | +| **Chat Env** | LLM Training | Chat-based environment with tokenization and message history for conversation RL | [`src/envs/chat_env/`](src/envs/chat_env/README.md) | +| **Atari Env** | Classic Games | 100+ Atari 2600 games via Arcade Learning Environment (Pong, Breakout, Space Invaders, etc.) | [`src/envs/atari_env/`](src/envs/atari_env/README.md) | +| **OpenSpiel Env** | Game Theory | 70+ games from DeepMind's OpenSpiel (Tic-Tac-Toe, Poker, Chess, Catch, 2048, etc.) | [`src/envs/openspiel_env/`](src/envs/openspiel_env/README.md) | +| **Git Env** | Software Dev | Git repository operations via Gitea for software development tasks | [`src/envs/git_env/`](src/envs/git_env/README.md) | +| **FinRL Env** | Finance | Stock trading environment with technical indicators and portfolio management | [`src/envs/finrl_env/`](src/envs/finrl_env/README.md) | +| **SUMO-RL Env** | Traffic Control | Traffic signal optimization using SUMO microscopic traffic simulation | [`src/envs/sumo_rl_env/`](src/envs/sumo_rl_env/README.md) | +| **TextArena Env** | Word Games | Word games and reasoning tasks (Wordle, GuessTheNumber, etc.) | [`src/envs/textarena_env/`](src/envs/textarena_env/README.md) | +| **DIPG Safety Env** | AI Safety | Medical domain safety training for high-stakes LLM applications with hallucination detection | [`src/envs/dipg_safety_env/`](src/envs/dipg_safety_env/README.md) | + +Each environment follows the standard OpenEnv API (`reset()`, `step()`, `state()`) and can be deployed via Docker containers for isolated execution. + ## OpenEnv on partner platforms: - [Lightning AI Studio](https://lightning.ai/environments?section=featured) From ff3bcf24031e59ea539f5777e4f503f676d81b48 Mon Sep 17 00:00:00 2001 From: "copilot-swe-agent[bot]" <198982749+Copilot@users.noreply.github.com> Date: Sat, 1 Nov 2025 04:36:22 +0000 Subject: [PATCH 3/3] Reorder environments by wow factor - Moved Atari and OpenSpiel to top (100+ and 70+ games respectively) - Placed complex real-world envs next (FinRL, SUMO-RL, DIPG Safety) - Mid-tier practical envs follow (Git, Coding, Chat) - Testing/simple envs at bottom (TextArena, Echo) Co-authored-by: init27 <16226196+init27@users.noreply.github.com> --- README.md | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/README.md b/README.md index a58e3b5d..b8e64f95 100644 --- a/README.md +++ b/README.md @@ -16,16 +16,16 @@ OpenEnv provides a growing collection of environments across different domains f | Environment | Domain | Description | Documentation | |-------------|--------|-------------|---------------| -| **Echo Env** | Testing | Simple message echo for testing framework infrastructure | [`src/envs/echo_env/`](src/envs/echo_env/README.md) | -| **Coding Env** | Code Execution | Execute Python code in sandboxed environment with smolagents | [`src/envs/coding_env/`](src/envs/coding_env/README.md) | -| **Chat Env** | LLM Training | Chat-based environment with tokenization and message history for conversation RL | [`src/envs/chat_env/`](src/envs/chat_env/README.md) | | **Atari Env** | Classic Games | 100+ Atari 2600 games via Arcade Learning Environment (Pong, Breakout, Space Invaders, etc.) | [`src/envs/atari_env/`](src/envs/atari_env/README.md) | | **OpenSpiel Env** | Game Theory | 70+ games from DeepMind's OpenSpiel (Tic-Tac-Toe, Poker, Chess, Catch, 2048, etc.) | [`src/envs/openspiel_env/`](src/envs/openspiel_env/README.md) | -| **Git Env** | Software Dev | Git repository operations via Gitea for software development tasks | [`src/envs/git_env/`](src/envs/git_env/README.md) | | **FinRL Env** | Finance | Stock trading environment with technical indicators and portfolio management | [`src/envs/finrl_env/`](src/envs/finrl_env/README.md) | | **SUMO-RL Env** | Traffic Control | Traffic signal optimization using SUMO microscopic traffic simulation | [`src/envs/sumo_rl_env/`](src/envs/sumo_rl_env/README.md) | -| **TextArena Env** | Word Games | Word games and reasoning tasks (Wordle, GuessTheNumber, etc.) | [`src/envs/textarena_env/`](src/envs/textarena_env/README.md) | | **DIPG Safety Env** | AI Safety | Medical domain safety training for high-stakes LLM applications with hallucination detection | [`src/envs/dipg_safety_env/`](src/envs/dipg_safety_env/README.md) | +| **Git Env** | Software Dev | Git repository operations via Gitea for software development tasks | [`src/envs/git_env/`](src/envs/git_env/README.md) | +| **Coding Env** | Code Execution | Execute Python code in sandboxed environment with smolagents | [`src/envs/coding_env/`](src/envs/coding_env/README.md) | +| **Chat Env** | LLM Training | Chat-based environment with tokenization and message history for conversation RL | [`src/envs/chat_env/`](src/envs/chat_env/README.md) | +| **TextArena Env** | Word Games | Word games and reasoning tasks (Wordle, GuessTheNumber, etc.) | [`src/envs/textarena_env/`](src/envs/textarena_env/README.md) | +| **Echo Env** | Testing | Simple message echo for testing framework infrastructure | [`src/envs/echo_env/`](src/envs/echo_env/README.md) | Each environment follows the standard OpenEnv API (`reset()`, `step()`, `state()`) and can be deployed via Docker containers for isolated execution.