As CleanRL gets more mature, it's time to re-think the future. With CleanRL 1.0, we'd hope to further improve documentation and design better contribution guidelines. This issue tracks a few desired items for CleanRL 1.0.

1.0

• Refactor the argparse parameters to have learning_rate total_timesteps move down to algorithm-specific arguments. #116
• Table of implemented algorithms and links that refer to the paper. (see https://docs.cleanrl.dev/rl-algorithms/overview/)
• Refactor documentation #121
• Re-think Open RL Benchmark. #123
• Friendlier CONTRIBUTING.md #117
• Seed issue with dqn.py and others #171
• TD3 should also log losses for qf2 #172
• Also log episodic_length for non-PPO scripts. #168
• Investigate nn.utils.clip_grad_norm_ for DQN, DDPG, and TD3 #148
• Auto-upgrade syntax via pyupgrade #158

1.1+

• Hyperparameter optimization #228
• Multi-objective hyperparameter optimization #265
• Adopt or not to adopt tensorboard native hyperparameters recording #184
• Add pre-commit for Markdown formatting #183
• Add rnd_ppo.py documentation and refactor #127
• DDPG/TD3 target_actor output clip #196
• PPO improvements #206
• JAX Integration with CleanRL #218
• Upgrade gym version to 0.26.1 #263