metadata
library_name: skrl
tags:
- deep-reinforcement-learning
- reinforcement-learning
- skrl
model-index:
- name: PPO
results:
- metrics:
- type: mean_reward
value: 1265.87 +/- 49.12
name: Total reward (mean)
task:
type: reinforcement-learning
name: reinforcement-learning
dataset:
name: IsaacGymEnvs-Quadcopter
type: IsaacGymEnvs-Quadcopter
IsaacGymEnvs-Quadcopter-PPO
Trained agent for NVIDIA Isaac Gym Preview environments.
- Task: Quadcopter
- Agent: PPO
Usage (with skrl)
Note: Visit the skrl Examples section to access the scripts.
PyTorch
from skrl.utils.huggingface import download_model_from_huggingface # assuming that there is an agent named `agent` path = download_model_from_huggingface("skrl/IsaacGymEnvs-Quadcopter-PPO", filename="agent.pt") agent.load(path)
JAX
from skrl.utils.huggingface import download_model_from_huggingface # assuming that there is an agent named `agent` path = download_model_from_huggingface("skrl/IsaacGymEnvs-Quadcopter-PPO", filename="agent.pickle") agent.load(path)
Hyperparameters
Note: Undefined parameters keep their values by default.
# https://skrl.readthedocs.io/en/latest/api/agents/ppo.html#configuration-and-hyperparameters
cfg = PPO_DEFAULT_CONFIG.copy()
cfg["rollouts"] = 8 # memory_size
cfg["learning_epochs"] = 8
cfg["mini_batches"] = 4 # 8 * 8192 / 16384
cfg["discount_factor"] = 0.99
cfg["lambda"] = 0.95
cfg["learning_rate"] = 1e-3
cfg["learning_rate_scheduler"] = KLAdaptiveRL
cfg["learning_rate_scheduler_kwargs"] = {"kl_threshold": 0.016}
cfg["random_timesteps"] = 0
cfg["learning_starts"] = 0
cfg["grad_norm_clip"] = 1.0
cfg["ratio_clip"] = 0.2
cfg["value_clip"] = 0.2
cfg["clip_predicted_values"] = True
cfg["entropy_loss_scale"] = 0.0
cfg["value_loss_scale"] = 1.0
cfg["kl_threshold"] = 0
cfg["rewards_shaper"] = lambda rewards, timestep, timesteps: rewards * 0.1
cfg["state_preprocessor"] = RunningStandardScaler
cfg["state_preprocessor_kwargs"] = {"size": env.observation_space, "device": device}
cfg["value_preprocessor"] = RunningStandardScaler
cfg["value_preprocessor_kwargs"] = {"size": 1, "device": device}