From 12e63228028c696e71afc2c5a3bd232654bf993f Mon Sep 17 00:00:00 2001 From: Heather Pundt Date: Thu, 5 Feb 2026 15:32:43 -0800 Subject: [PATCH 1/7] Update top intro Signed-off-by: Heather Pundt --- README.md | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/README.md b/README.md index eb29dc65d..c53b8719a 100644 --- a/README.md +++ b/README.md @@ -2,10 +2,10 @@ NeMo Gym is a library for building reinforcement learning (RL) training environments for large language models (LLMs). It provides infrastructure to develop environments, scale rollout collection, and integrate seamlessly with your preferred training framework. -NeMo Gym is a component of the [NVIDIA NeMo Framework](https://docs.nvidia.com/nemo-framework/), NVIDIA’s GPU-accelerated platform for building and training generative AI models. +At its core, Gym standardizes scalable rollout collection through unified interfaces to heterogeneous RL environments and curated resource servers with verification logic. This makes it practical to generate large-scale, high-quality training data for various RL frameworks, including NeMo RL. [Learn more about the NeMo Gym ecosystem.](https://docs.nvidia.com/nemo/gym/latest/about/ecosystem.html) -## πŸ† Why NeMo Gym? +Why NeMo Gym? - Scaffolding and patterns to accelerate environment development: multi-step, multi-turn, and user modeling scenarios - Contribute environments without expert knowledge of the entire RL training loop From 6cf8ea2d6e2263c421871a5ec052097b35c34495 Mon Sep 17 00:00:00 2001 From: Heather Pundt Date: Thu, 5 Feb 2026 15:33:24 -0800 Subject: [PATCH 2/7] Add top menu Signed-off-by: Heather Pundt --- README.md | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/README.md b/README.md index c53b8719a..581f69ddc 100644 --- a/README.md +++ b/README.md @@ -1,10 +1,11 @@ # NeMo Gym +**[Requirements](#-requirements)** β€’ **[Quick Start](#-quick-start)** β€’ **[Available Resource Servers](#-available-resource-servers)** β€’ **[Documentation](#-documentation)** β€’ **[Community & Support](#-community--support)** β€’ **[Citations](#-citations)** + NeMo Gym is a library for building reinforcement learning (RL) training environments for large language models (LLMs). It provides infrastructure to develop environments, scale rollout collection, and integrate seamlessly with your preferred training framework. At its core, Gym standardizes scalable rollout collection through unified interfaces to heterogeneous RL environments and curated resource servers with verification logic. This makes it practical to generate large-scale, high-quality training data for various RL frameworks, including NeMo RL. [Learn more about the NeMo Gym ecosystem.](https://docs.nvidia.com/nemo/gym/latest/about/ecosystem.html) - Why NeMo Gym? - Scaffolding and patterns to accelerate environment development: multi-step, multi-turn, and user modeling scenarios From 962de75e0bd20332561605e0b52f4e8056e288b9 Mon Sep 17 00:00:00 2001 From: Heather Pundt Date: Thu, 5 Feb 2026 15:39:16 -0800 Subject: [PATCH 3/7] Consolidate requirements Signed-off-by: Heather Pundt --- README.md | 25 +++++++------------------ 1 file changed, 7 insertions(+), 18 deletions(-) diff --git a/README.md b/README.md index 581f69ddc..b12f2c6b0 100644 --- a/README.md +++ b/README.md @@ -19,27 +19,16 @@ Why NeMo Gym? ## πŸ“‹ Requirements -### Hardware Requirements - NeMo Gym is designed to run on standard development machines: -- **GPU**: Not required for NeMo Gym library operation - - GPU may be needed for specific resource servers or model inference (see individual server documentation) -- **CPU**: Any modern x86_64 or ARM64 processor (e.g., Intel, AMD, Apple Silicon) -- **RAM**: Minimum 8 GB (16 GB+ recommended for larger environments) -- **Storage**: Minimum 5 GB free disk space for installation and basic usage - -### Software Requirements - -- **Operating System**: - - Linux (Ubuntu 20.04+, or equivalent) - - macOS (11.0+ for x86_64, 12.0+ for Apple Silicon) - - Windows (via WSL2) -- **Python**: 3.12 or higher -- **Git**: For cloning the repository -- **Internet Connection**: Required for downloading dependencies and API access +| Hardware Requirements | Software Requirements | +| --------------------- | --------------------- | +| **GPU**: Not required for NeMo Gym library operation
β€’ GPU may be needed for specific resource servers or model inference (see individual server documentation) | **Operating System**:
β€’ Linux (Ubuntu 20.04+, or equivalent)
β€’ macOS (11.0+ for x86_64, 12.0+ for Apple Silicon)
β€’ Windows (via WSL2) | +| **CPU**: Any modern x86_64 or ARM64 processor (e.g., Intel, AMD, Apple Silicon) | **Python**: 3.12 or higher | +| **RAM**: Minimum 8 GB (16 GB+ recommended for larger environments) | **Git**: For cloning the repository | +| **Storage**: Minimum 5 GB free disk space for installation and basic usage | **Internet Connection**: Required for downloading dependencies and API access | -### Additional Requirements +**Additional Requirements** - **API Keys**: OpenAI API key with available credits (for the quickstart examples) - Other model providers supported (Azure OpenAI, self-hosted models via vLLM) From 8e25ea76543ddaedd3b5bf1a7e4cc8205e3c2b9e Mon Sep 17 00:00:00 2001 From: Heather Pundt Date: Thu, 5 Feb 2026 17:42:04 -0800 Subject: [PATCH 4/7] Future proof docs section Signed-off-by: Heather Pundt --- README.md | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/README.md b/README.md index b12f2c6b0..3117a0645 100644 --- a/README.md +++ b/README.md @@ -1,6 +1,6 @@ # NeMo Gym -**[Requirements](#-requirements)** β€’ **[Quick Start](#-quick-start)** β€’ **[Available Resource Servers](#-available-resource-servers)** β€’ **[Documentation](#-documentation)** β€’ **[Community & Support](#-community--support)** β€’ **[Citations](#-citations)** +**[Requirements](#-requirements)** β€’ **[Quick Start](#-quick-start)** β€’ **[Available Resource Servers](#-available-resource-servers)** β€’ **[Documentation & Resources](#-documentation--resources)** β€’ **[Community & Support](#-community--support)** β€’ **[Citations](#-citations)** NeMo Gym is a library for building reinforcement learning (RL) training environments for large language models (LLMs). It provides infrastructure to develop environments, scale rollout collection, and integrate seamlessly with your preferred training framework. @@ -156,7 +156,7 @@ Purpose: Training-ready environments with curated datasets. | Math With Judge | math | Nemotron-RL-math-stack_overflow | - | - | config | βœ“ | βœ“ | Creative Commons Attribution-ShareAlike 4.0 International | -## πŸ“– Documentation +## πŸ“– Documentation & Resources - **[Documentation](https://docs.nvidia.com/nemo/gym/latest/index.html)** - Technical reference docs - **[Tutorials](https://docs.nvidia.com/nemo/gym/latest/tutorials/index.html)** - Hands-on tutorials and practical examples From ce0f7ed57de59a5516d2ff2674709403d509fcb1 Mon Sep 17 00:00:00 2001 From: Heather Pundt Date: Thu, 5 Feb 2026 17:49:04 -0800 Subject: [PATCH 5/7] Shorten intro Signed-off-by: Heather Pundt --- README.md | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/README.md b/README.md index 3117a0645..364f1df52 100644 --- a/README.md +++ b/README.md @@ -2,9 +2,7 @@ **[Requirements](#-requirements)** β€’ **[Quick Start](#-quick-start)** β€’ **[Available Resource Servers](#-available-resource-servers)** β€’ **[Documentation & Resources](#-documentation--resources)** β€’ **[Community & Support](#-community--support)** β€’ **[Citations](#-citations)** -NeMo Gym is a library for building reinforcement learning (RL) training environments for large language models (LLMs). It provides infrastructure to develop environments, scale rollout collection, and integrate seamlessly with your preferred training framework. - -At its core, Gym standardizes scalable rollout collection through unified interfaces to heterogeneous RL environments and curated resource servers with verification logic. This makes it practical to generate large-scale, high-quality training data for various RL frameworks, including NeMo RL. [Learn more about the NeMo Gym ecosystem.](https://docs.nvidia.com/nemo/gym/latest/about/ecosystem.html) +NeMo Gym is a library for building reinforcement learning (RL) training environments for large language models (LLMs). It provides infrastructure to develop environments, scale rollout collection, and integrate seamlessly with your preferred training framework. [Learn more about the NeMo Gym ecosystem.](https://docs.nvidia.com/nemo/gym/latest/about/ecosystem.html) Why NeMo Gym? From b9f4a0093b9a387d49dd83477fed10155cdbab08 Mon Sep 17 00:00:00 2001 From: Heather Pundt Date: Thu, 5 Feb 2026 18:08:39 -0800 Subject: [PATCH 6/7] Update doc and news section Signed-off-by: Heather Pundt --- README.md | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/README.md b/README.md index 364f1df52..78bc97e2e 100644 --- a/README.md +++ b/README.md @@ -1,6 +1,6 @@ # NeMo Gym -**[Requirements](#-requirements)** β€’ **[Quick Start](#-quick-start)** β€’ **[Available Resource Servers](#-available-resource-servers)** β€’ **[Documentation & Resources](#-documentation--resources)** β€’ **[Community & Support](#-community--support)** β€’ **[Citations](#-citations)** +**[Requirements](#-requirements)** β€’ **[Quick Start](#-quick-start)** β€’ **[Available Resource Servers](#-available-resource-servers)** β€’ **[Documentation & Resources](#-documentation--resources)** β€’ **[Community & Support](#-community--support)** β€’ **[Latest News & Publications](#-latest-news--publications)** β€’ **[Citations](#-citations)** NeMo Gym is a library for building reinforcement learning (RL) training environments for large language models (LLMs). It provides infrastructure to develop environments, scale rollout collection, and integrate seamlessly with your preferred training framework. [Learn more about the NeMo Gym ecosystem.](https://docs.nvidia.com/nemo/gym/latest/about/ecosystem.html) @@ -158,6 +158,7 @@ Purpose: Training-ready environments with curated datasets. - **[Documentation](https://docs.nvidia.com/nemo/gym/latest/index.html)** - Technical reference docs - **[Tutorials](https://docs.nvidia.com/nemo/gym/latest/tutorials/index.html)** - Hands-on tutorials and practical examples +- **[API Reference](https://docs.nvidia.com/nemo/gym/latest/apidocs/index.html)** - Complete class and function reference ## 🀝 Community & Support @@ -167,6 +168,10 @@ We'd love your contributions! Here's how to get involved: - **[Report Issues](https://github.com/NVIDIA-NeMo/Gym/issues)** - Bug reports and feature requests - **[Contributing Guide](https://docs.nvidia.com/nemo/gym/latest/contribute/index.html)** - How to contribute code, docs, new environments, or training framework integrations +## πŸ“° Latest News & Publications + +Coming soon! + ## πŸ“š Citations If you use NeMo Gym in your research, please cite it using the following BibTeX entry: From 94e21438650cde71e05a3282249c6aa09e33efee Mon Sep 17 00:00:00 2001 From: Heather Pundt Date: Thu, 5 Feb 2026 18:15:52 -0800 Subject: [PATCH 7/7] Add intro to quickstart Signed-off-by: Heather Pundt --- README.md | 2 ++ 1 file changed, 2 insertions(+) diff --git a/README.md b/README.md index 78bc97e2e..3788c8022 100644 --- a/README.md +++ b/README.md @@ -34,6 +34,8 @@ NeMo Gym is designed to run on standard development machines: ## πŸš€ Quick Start +Install NeMo Gym, start the servers, and collect your first verified rollouts for RL training. + ### Setup ```bash # Clone the repository