deepgym good first issues (Python)

Desktop environments on real open-source apps (one PR per app) 19 days ago

help wanted hacktoberfest environment

7

RL training environments with verifiable rewards for coding and computer-use agents. Works with TRL, Unsloth, verl, OpenRLHF. https://deepgym.io

Python

#ai-agents#code-execution#coding-agents#computer-use#daytona#deep-learning#evaluation#grpo#grpo-training#hacktoberfest#llm#machine-learning#openrlhf#python#reinforcement-learning#reward-function#rl#sandbox#unsloth#verifiable-rewards

Desktop environment: VistA 19 days ago

help wanted hacktoberfest environment

DeepGym/deepgym

7

RL training environments with verifiable rewards for coding and computer-use agents. Works with TRL, Unsloth, verl, OpenRLHF. https://deepgym.io

Python

#ai-agents#code-execution#coding-agents#computer-use#daytona#deep-learning#evaluation#grpo#grpo-training#hacktoberfest#llm#machine-learning#openrlhf#python#reinforcement-learning#reward-function#rl#sandbox#unsloth#verifiable-rewards

Desktop environment: QGIS 19 days ago

help wanted hacktoberfest environment

DeepGym/deepgym

7

RL training environments with verifiable rewards for coding and computer-use agents. Works with TRL, Unsloth, verl, OpenRLHF. https://deepgym.io

Python

#ai-agents#code-execution#coding-agents#computer-use#daytona#deep-learning#evaluation#grpo#grpo-training#hacktoberfest#llm#machine-learning#openrlhf#python#reinforcement-learning#reward-function#rl#sandbox#unsloth#verifiable-rewards

Desktop environment: FreeCAD 19 days ago

help wanted hacktoberfest environment

DeepGym/deepgym

7

RL training environments with verifiable rewards for coding and computer-use agents. Works with TRL, Unsloth, verl, OpenRLHF. https://deepgym.io

Python

#ai-agents#code-execution#coding-agents#computer-use#daytona#deep-learning#evaluation#grpo#grpo-training#hacktoberfest#llm#machine-learning#openrlhf#python#reinforcement-learning#reward-function#rl#sandbox#unsloth#verifiable-rewards

Desktop environment: KiCad 19 days ago

help wanted hacktoberfest environment

DeepGym/deepgym

7

RL training environments with verifiable rewards for coding and computer-use agents. Works with TRL, Unsloth, verl, OpenRLHF. https://deepgym.io

Python

#ai-agents#code-execution#coding-agents#computer-use#daytona#deep-learning#evaluation#grpo#grpo-training#hacktoberfest#llm#machine-learning#openrlhf#python#reinforcement-learning#reward-function#rl#sandbox#unsloth#verifiable-rewards

Desktop environment: GLPI 19 days ago

help wanted hacktoberfest environment

DeepGym/deepgym

7

RL training environments with verifiable rewards for coding and computer-use agents. Works with TRL, Unsloth, verl, OpenRLHF. https://deepgym.io

Python

#ai-agents#code-execution#coding-agents#computer-use#daytona#deep-learning#evaluation#grpo#grpo-training#hacktoberfest#llm#machine-learning#openrlhf#python#reinforcement-learning#reward-function#rl#sandbox#unsloth#verifiable-rewards

Desktop environment: Znuny 19 days ago

help wanted hacktoberfest environment

DeepGym/deepgym

7

RL training environments with verifiable rewards for coding and computer-use agents. Works with TRL, Unsloth, verl, OpenRLHF. https://deepgym.io

Python

#ai-agents#code-execution#coding-agents#computer-use#daytona#deep-learning#evaluation#grpo#grpo-training#hacktoberfest#llm#machine-learning#openrlhf#python#reinforcement-learning#reward-function#rl#sandbox#unsloth#verifiable-rewards

Desktop environment: Request Tracker 19 days ago

help wanted hacktoberfest environment

DeepGym/deepgym

7

RL training environments with verifiable rewards for coding and computer-use agents. Works with TRL, Unsloth, verl, OpenRLHF. https://deepgym.io

Python

#ai-agents#code-execution#coding-agents#computer-use#daytona#deep-learning#evaluation#grpo#grpo-training#hacktoberfest#llm#machine-learning#openrlhf#python#reinforcement-learning#reward-function#rl#sandbox#unsloth#verifiable-rewards

Desktop environment: Bugzilla 19 days ago

help wanted hacktoberfest environment

DeepGym/deepgym

7

RL training environments with verifiable rewards for coding and computer-use agents. Works with TRL, Unsloth, verl, OpenRLHF. https://deepgym.io

Python

#ai-agents#code-execution#coding-agents#computer-use#daytona#deep-learning#evaluation#grpo#grpo-training#hacktoberfest#llm#machine-learning#openrlhf#python#reinforcement-learning#reward-function#rl#sandbox#unsloth#verifiable-rewards

Desktop environment: Nagios Core 19 days ago

help wanted hacktoberfest environment

DeepGym/deepgym

7

RL training environments with verifiable rewards for coding and computer-use agents. Works with TRL, Unsloth, verl, OpenRLHF. https://deepgym.io

Python

#ai-agents#code-execution#coding-agents#computer-use#daytona#deep-learning#evaluation#grpo#grpo-training#hacktoberfest#llm#machine-learning#openrlhf#python#reinforcement-learning#reward-function#rl#sandbox#unsloth#verifiable-rewards

Desktop environment: Cacti 19 days ago

help wanted hacktoberfest environment

DeepGym/deepgym

7

RL training environments with verifiable rewards for coding and computer-use agents. Works with TRL, Unsloth, verl, OpenRLHF. https://deepgym.io

Python

#ai-agents#code-execution#coding-agents#computer-use#daytona#deep-learning#evaluation#grpo#grpo-training#hacktoberfest#llm#machine-learning#openrlhf#python#reinforcement-learning#reward-function#rl#sandbox#unsloth#verifiable-rewards

Desktop environment: iDempiere 19 days ago

help wanted hacktoberfest environment

DeepGym/deepgym

7

RL training environments with verifiable rewards for coding and computer-use agents. Works with TRL, Unsloth, verl, OpenRLHF. https://deepgym.io

Python

#ai-agents#code-execution#coding-agents#computer-use#daytona#deep-learning#evaluation#grpo#grpo-training#hacktoberfest#llm#machine-learning#openrlhf#python#reinforcement-learning#reward-function#rl#sandbox#unsloth#verifiable-rewards

Desktop environment: Apache OFBiz 19 days ago

help wanted hacktoberfest environment

DeepGym/deepgym

7

RL training environments with verifiable rewards for coding and computer-use agents. Works with TRL, Unsloth, verl, OpenRLHF. https://deepgym.io

Python

#ai-agents#code-execution#coding-agents#computer-use#daytona#deep-learning#evaluation#grpo#grpo-training#hacktoberfest#llm#machine-learning#openrlhf#python#reinforcement-learning#reward-function#rl#sandbox#unsloth#verifiable-rewards

BigCodeBench rewards can be forged by the solution 23 days ago

bug help wanted hacktoberfest

DeepGym/deepgym

7

RL training environments with verifiable rewards for coding and computer-use agents. Works with TRL, Unsloth, verl, OpenRLHF. https://deepgym.io

Python

#ai-agents#code-execution#coding-agents#computer-use#daytona#deep-learning#evaluation#grpo#grpo-training#hacktoberfest#llm#machine-learning#openrlhf#python#reinforcement-learning#reward-function#rl#sandbox#unsloth#verifiable-rewards

EvalPlus importer: verifiers with >4300-digit int literals fail to compile 23 days ago

bug good first issue hacktoberfest

DeepGym/deepgym

7

RL training environments with verifiable rewards for coding and computer-use agents. Works with TRL, Unsloth, verl, OpenRLHF. https://deepgym.io

Python

#ai-agents#code-execution#coding-agents#computer-use#daytona#deep-learning#evaluation#grpo#grpo-training#hacktoberfest#llm#machine-learning#openrlhf#python#reinforcement-learning#reward-function#rl#sandbox#unsloth#verifiable-rewards

Import XiaomiMiMo/MiMo-V2.6-RL-oss as DeepGym environments 23 days ago

help wanted hacktoberfest environment

DeepGym/deepgym

7

RL training environments with verifiable rewards for coding and computer-use agents. Works with TRL, Unsloth, verl, OpenRLHF. https://deepgym.io

Python

#ai-agents#code-execution#coding-agents#computer-use#daytona#deep-learning#evaluation#grpo#grpo-training#hacktoberfest#llm#machine-learning#openrlhf#python#reinforcement-learning#reward-function#rl#sandbox#unsloth#verifiable-rewards

Docs page: how verifier isolation works, with a diagram 23 days ago

documentation good first issue hacktoberfest

DeepGym/deepgym

7

RL training environments with verifiable rewards for coding and computer-use agents. Works with TRL, Unsloth, verl, OpenRLHF. https://deepgym.io

Python

#ai-agents#code-execution#coding-agents#computer-use#daytona#deep-learning#evaluation#grpo#grpo-training#hacktoberfest#llm#machine-learning#openrlhf#python#reinforcement-learning#reward-function#rl#sandbox#unsloth#verifiable-rewards

Import OSWorld tasks as computer-use environments 23 days ago

help wanted hacktoberfest environment

DeepGym/deepgym

7

RL training environments with verifiable rewards for coding and computer-use agents. Works with TRL, Unsloth, verl, OpenRLHF. https://deepgym.io

Python

#ai-agents#code-execution#coding-agents#computer-use#daytona#deep-learning#evaluation#grpo#grpo-training#hacktoberfest#llm#machine-learning#openrlhf#python#reinforcement-learning#reward-function#rl#sandbox#unsloth#verifiable-rewards

End-to-end verl GRPO example with a real reward curve 23 days ago

documentation help wanted hacktoberfest

DeepGym/deepgym

7

RL training environments with verifiable rewards for coding and computer-use agents. Works with TRL, Unsloth, verl, OpenRLHF. https://deepgym.io

Python

#ai-agents#code-execution#coding-agents#computer-use#daytona#deep-learning#evaluation#grpo#grpo-training#hacktoberfest#llm#machine-learning#openrlhf#python#reinforcement-learning#reward-function#rl#sandbox#unsloth#verifiable-rewards

Run the bwrap isolation tests on Linux in CI 23 days ago

help wanted hacktoberfest

DeepGym/deepgym

7

RL training environments with verifiable rewards for coding and computer-use agents. Works with TRL, Unsloth, verl, OpenRLHF. https://deepgym.io

Python

#ai-agents#code-execution#coding-agents#computer-use#daytona#deep-learning#evaluation#grpo#grpo-training#hacktoberfest#llm#machine-learning#openrlhf#python#reinforcement-learning#reward-function#rl#sandbox#unsloth#verifiable-rewards

Lint tests/ in CI and fix the 20 existing ruff errors 23 days ago

good first issue hacktoberfest

DeepGym/deepgym

7

RL training environments with verifiable rewards for coding and computer-use agents. Works with TRL, Unsloth, verl, OpenRLHF. https://deepgym.io

Python

#ai-agents#code-execution#coding-agents#computer-use#daytona#deep-learning#evaluation#grpo#grpo-training#hacktoberfest#llm#machine-learning#openrlhf#python#reinforcement-learning#reward-function#rl#sandbox#unsloth#verifiable-rewards

Test that a do-nothing solution fails every built-in environment 23 days ago

good first issue hacktoberfest

DeepGym/deepgym

7

RL training environments with verifiable rewards for coding and computer-use agents. Works with TRL, Unsloth, verl, OpenRLHF. https://deepgym.io

Python

#ai-agents#code-execution#coding-agents#computer-use#daytona#deep-learning#evaluation#grpo#grpo-training#hacktoberfest#llm#machine-learning#openrlhf#python#reinforcement-learning#reward-function#rl#sandbox#unsloth#verifiable-rewards

Add scripting and filesystem environments (log parsing, CSV cleanup, ...) 23 days ago

good first issue hacktoberfest environment

DeepGym/deepgym

7

RL training environments with verifiable rewards for coding and computer-use agents. Works with TRL, Unsloth, verl, OpenRLHF. https://deepgym.io

Python

#ai-agents#code-execution#coding-agents#computer-use#daytona#deep-learning#evaluation#grpo#grpo-training#hacktoberfest#llm#machine-learning#openrlhf#python#reinforcement-learning#reward-function#rl#sandbox#unsloth#verifiable-rewards

Add built-in coding environments (one PR per environment) 23 days ago

good first issue hacktoberfest environment

DeepGym/deepgym

7

RL training environments with verifiable rewards for coding and computer-use agents. Works with TRL, Unsloth, verl, OpenRLHF. https://deepgym.io

Python

#ai-agents#code-execution#coding-agents#computer-use#daytona#deep-learning#evaluation#grpo#grpo-training#hacktoberfest#llm#machine-learning#openrlhf#python#reinforcement-learning#reward-function#rl#sandbox#unsloth#verifiable-rewards

DeepGym/deepgym

Open Issues Need Help