Skip to content

SkyRL-Train + SkyGym code#27

Merged
SumanthRH merged 10 commits intomainfrom
sumanthrh/skyrl-train-release
Jun 25, 2025
Merged

SkyRL-Train + SkyGym code#27
SumanthRH merged 10 commits intomainfrom
sumanthrh/skyrl-train-release

Conversation

@SumanthRH
Copy link
Member

What does this PR do?

Adds our SkyRL-Train + SkyGym code

Signed-off-by: SumanthRH <sumanthrh99@gmail.com>
Signed-off-by: SumanthRH <sumanthrh99@gmail.com>
x
Signed-off-by: SumanthRH <sumanthrh99@gmail.com>
Signed-off-by: SumanthRH <sumanthrh99@gmail.com>
Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Will keep only formatting related rules at top-level

README.md Outdated
SkyRL provides the following components:

For reproducing our results for SkyRL-Agent-14B-v0, SkyRL-Agent-8B-v0, and SkyRL-Agent-7B-v0 you can refer to [examples/sky/swebench](./examples/sky/swebench/README.md).
- [`skyrl-agent`](./skyrl-agent): Our agent layer for training long-horizon, real-world agents.
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

use this name maybe: SkyAgent
Also add a link to the corresponding blogpost? The code for SkyRL-v0: Train Real-World Long-Horizon Agents via Reinforcement Learning.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Done!

README.md Outdated
For reproducing our results for SkyRL-Agent-14B-v0, SkyRL-Agent-8B-v0, and SkyRL-Agent-7B-v0 you can refer to [examples/sky/swebench](./examples/sky/swebench/README.md).
- [`skyrl-agent`](./skyrl-agent): Our agent layer for training long-horizon, real-world agents.
- (NEW!) [`skyrl-train`](./skyrl-train): Our flexible training framework for RL.
- (NEW!) [`skygym`](./skygym): Our library of agentic, tool-use environments.
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

no agentic? just mention what environments are supported.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Done!

Signed-off-by: SumanthRH <sumanthrh99@gmail.com>
Signed-off-by: SumanthRH <sumanthrh99@gmail.com>
x
Signed-off-by: SumanthRH <sumanthrh99@gmail.com>
Signed-off-by: SumanthRH <sumanthrh99@gmail.com>
@SumanthRH SumanthRH requested a review from caoshiyi June 25, 2025 06:27
Signed-off-by: SumanthRH <sumanthrh99@gmail.com>
@SumanthRH SumanthRH merged commit b3280d8 into main Jun 25, 2025
5 of 6 checks passed
@SumanthRH SumanthRH deleted the sumanthrh/skyrl-train-release branch July 16, 2025 23:19
fannie1208 pushed a commit to vinid/SkyRL that referenced this pull request Aug 19, 2025
Co-authored-by: Tyler Griggs <tgriggs@berkeley.edu>
Co-authored-by: Eric Tang <etang@anyscale.com>
Co-authored-by: Sumanth Hegde <sumanthrh@anyscale.com>
Co-authored-by: Shu Liu <lshu@berkeley.edu>
Co-authored-by: Shiyi Cao <shicao@berkeley.edu>
Co-authored-by: Dacheng Li <dacheng177@berkeley.edu>
Co-authored-by: Charlie Ruan <charlieruan@berkeley.edu>
pcmoritz added a commit to pcmoritz/SkyRL that referenced this pull request Oct 4, 2025
This is a very first cut of the API, it is likely that there are mistakes and thing will change. For example currently I'm just setting the future_id to be the same as the request_id, we will see if that holds up.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants