eric000888

RemiG3, Thank you for adding attention net to contrib. what's the shape of the input would be look like , for example if I want to use cartpole environment? Thanks...

RemiG3, Sorry for late response, here is my first post code: from sb3_contrib.ppo_attention.ppo_attention import AttentionPPO from sb3_contrib.ppo_attention.policies import MlpAttnPolicy VE = DummyVecEnv([lambda: gym.make("CartPole-v1")]) model = AttentionPPO( "MlpAttnPolicy", VE, n_steps=240, learning_rate=0.0003,...

another questions is if you just use GtrXL as feature extractor in PPO model, is this will get the same results? as the LSTM recurrent PPO has a flag to...

another thing is GtrXL demand more computation power , and PPO is like aiming a moving target, I found training a GtrXL PPO is a daunting task especially when using...

I think that's not a limitation as all the env in VecEnv also has same shape for observation space and action space to function properly in the vectorized setting. Ray...