| --- |
| library_name: stable-baselines3 |
| tags: |
| - reinforcement-learning |
| - stable-baselines3 |
| - deep-reinforcement-learning |
| - fluidgym |
| - active-flow-control |
| - fluid-dynamics |
| - simulation |
| - RBC2D-easy-v0 |
| model-index: |
| - name: SAC-RBC2D-easy-v0 |
| results: |
| - task: |
| type: reinforcement-learning |
| name: reinforcement-learning |
| dataset: |
| name: FluidGym-RBC2D-easy-v0 |
| type: fluidgym |
| metrics: |
| - type: mean_reward |
| value: 0.65 |
| name: mean_reward |
|
|
|
|
| --- |
| |
| # SAC on RBC2D-easy-v0 (FluidGym) |
|
|
| This repository is part of the **FluidGym** benchmark results. It contains trained Stable Baselines3 agents for the specialized **RBC2D-easy-v0** environment. |
|
|
| ## Evaluation Results |
|
|
| ### Global Performance (Aggregated across 5 seeds) |
| **Mean Reward:** 0.65 ± 0.09 |
|
|
| ### Per-Seed Statistics |
| | Run | Mean Reward | Std Dev | |
| | --- | --- | --- | |
| | Seed 0 | 0.63 | 0.19 | |
| | Seed 1 | 0.70 | 0.24 | |
| | Seed 2 | 0.77 | 0.25 | |
| | Seed 3 | 0.66 | 0.28 | |
| | Seed 4 | 0.49 | 0.15 | |
|
|
| ## About FluidGym |
| FluidGym is a benchmark for reinforcement learning in active flow control. |
|
|
| ## Usage |
| Each seed is contained in its own subdirectory. You can load a model using: |
| ```python |
| from stable_baselines3 import SAC |
| model = SAC.load("0/ckpt_latest.zip") |
| ``` |
|
|
| **Important:** The models were trained using ```fluidgym==0.0.2```. In order to use |
| them with newer versions of FluidGym, you need to wrap the environment with a |
| `FlattenObservation` wrapper as shown below: |
| ```python |
| import fluidgym |
| from fluidgym.wrappers import FlattenObservation |
| from stable_baselines3 import SAC |
| |
| env = fluidgym.make("RBC2D-easy-v0") |
| env = FlattenObservation(env) |
| model = SAC.load("path_to_model/ckpt_latest.zip") |
| |
| obs, info = env.reset(seed=42) |
| |
| action, _ = model.predict(obs, deterministic=True) |
| obs, reward, terminated, truncated, info = env.step(action) |
| ``` |
|
|
| ## References |
|
|
| * [Plug-and-Play Benchmarking of Reinforcement Learning Algorithms for Large-Scale Flow Control](http://arxiv.org/abs/2601.15015) |
| * [FluidGym GitHub Repository](https://github.com/safe-autonomous-systems/fluidgym) |
|
|
