The ssbm_gym from gurvan

Stuck on resetting envs

Hi!! This project is really exciting. I haven't been able to see anything in dolphin yet though -- I'm having trouble getting past "Resetting envs" when running test_env_vectorized.py & test_env.py -- it seems like both are hanging on the join threads part of util.async_map(makePad, pipe_paths) -- in the wait() function. It's hard to tell if I'm on the right track.

hitlag_frames_left returning NaN

Hello!

I forked this repo and made it compatible with stable_baselines3. Any idea why puff/peach would be returning NaN values?

The number 48 in the size of the Tensor is the number of instances of puff fighting, not exactly sure what the 54 is referring to

Btw this what exactly what I was looking for so thank you for creating it!

Using cuda:1 device
Logging to logs/puff_1675774951/PPO_0
Traceback (most recent call last):
  File "train_stable_baselines.py", line 75, in <module>
    model.learn(total_timesteps=TIMESTEPS, reset_num_timesteps=False, tb_log_name=f"PPO")
  File "/home/james/anaconda3/envs/ssbm/lib/python3.8/site-packages/stable_baselines3/ppo/ppo.py", line 304, in learn
    return super(PPO, self).learn(
  File "/home/james/anaconda3/envs/ssbm/lib/python3.8/site-packages/stable_baselines3/common/on_policy_algorithm.py", line 250, in learn
    continue_training = self.collect_rollouts(self.env, callback, self.rollout_buffer, n_rollout_steps=self.n_steps)
  File "/home/james/anaconda3/envs/ssbm/lib/python3.8/site-packages/stable_baselines3/common/on_policy_algorithm.py", line 169, in collect_rollouts
    actions, values, log_probs = self.policy(obs_tensor)
  File "/home/james/.local/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1110, in _call_impl
    return forward_call(*input, **kwargs)
  File "/home/james/anaconda3/envs/ssbm/lib/python3.8/site-packages/stable_baselines3/common/policies.py", line 592, in forward
    distribution = self._get_action_dist_from_latent(latent_pi)
  File "/home/james/anaconda3/envs/ssbm/lib/python3.8/site-packages/stable_baselines3/common/policies.py", line 610, in _get_action_dist_from_latent
    return self.action_dist.proba_distribution(action_logits=mean_actions)
  File "/home/james/anaconda3/envs/ssbm/lib/python3.8/site-packages/stable_baselines3/common/distributions.py", line 274, in proba_distribution
    self.distribution = Categorical(logits=action_logits)
  File "/home/james/.local/lib/python3.8/site-packages/torch/distributions/categorical.py", line 64, in __init__
    super(Categorical, self).__init__(batch_shape, validate_args=validate_args)
  File "/home/james/.local/lib/python3.8/site-packages/torch/distributions/distribution.py", line 55, in __init__
    raise ValueError(
ValueError: Expected parameter logits (Tensor of shape (48, 54)) of distribution Categorical(logits: torch.Size([48, 54])) to satisfy the constraint IndependentConstraint(Real(), 1), but found invalid values:
tensor([[-3.9884, -3.9919, -3.9901,  ..., -3.9855, -3.9914, -3.9906],
        [-3.9884, -3.9919, -3.9901,  ..., -3.9855, -3.9914, -3.9906],
        [-3.9849, -3.9850, -3.9818,  ..., -3.9968, -3.9851, -3.9881],
        ...,
        [-3.9910, -3.9944, -3.9888,  ..., -3.9884, -3.9897, -3.9892],
        [    nan,     nan,     nan,  ...,     nan,     nan,     nan],
        [-3.9872, -3.9876, -3.9856,  ..., -3.9875, -3.9909, -3.9943]],
       device='cuda:1')

Incomplete Build Instructions (non-headless)

Non-headless build has additional dependencies not listed in the instructions. Recommend that you add the rest of the dependencies listed on the official dolphin linux build instructions page found here: https://dolphin-emu.org/docs/guides/building-dolphin-linux/

gurvan / ssbm_gym Goto Github PK

ssbm_gym's People

Contributors

Stargazers

Watchers

Forkers

ssbm_gym's Issues

Stuck on resetting envs

hitlag_frames_left returning NaN

Incomplete Build Instructions (non-headless)

Recommend Projects

React

Vue.js

Typescript

TensorFlow

Django

Laravel

D3

Recommend Topics

javascript

web

server

Machine learning

Visualization

Game

Recommend Org

Facebook

Microsoft

Google

Alibaba

D3

Tencent