refactor(lyd): refactor dt_policy in new pipeline #687
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
PaParaZz1
added
algo
PaParaZz1
requested changes
Jul 18, 2023
| @@ -1174,6 +1174,23 @@ def reset(self): | |||
| return self.env.reset() | |||
|
|
|||
|
|
|||
| class AllinObsWrapper(gym.Wrapper): | |||
There was a problem hiding this comment.
add overview description for this wrapper
| @@ -42,7 +42,8 @@ | |||
|
|
|||
| from .d4pg import D4PGPolicy | |||
| from .cql import CQLPolicy, CQLDiscretePolicy | |||
| from .decision_transformer import DTPolicy | |||
| # from .decision_transformer import DTPolicy | |||
ding/policy/dt.py
Outdated
| class DTPolicy(Policy): | ||
| r""" | ||
| Overview: | ||
| Policy class of DT algorithm in discrete environments. |
There was a problem hiding this comment.
add the full name of DT and the paper link
ding/policy/dt.py
Outdated
|
|
||
| def _init_learn(self) -> None: | ||
| r""" | ||
| Overview: |
| @@ -27,7 +27,7 @@ | |||
| embed_dim=128, | |||
| n_heads=1, | |||
| dropout_p=0.1, | |||
| log_dir='/home/puyuan/DI-engine/dizoo/box2d/lunarlander/dt_log_1000eps', | |||
| log_dir='/mnt/nfs/luyd/DI-engine/dizoo/box2d/lunarlander/dt_log_1000eps', | |||
There was a problem hiding this comment.
don't upload the absolute path
| @@ -0,0 +1,2 @@ | |||
| duration,num_updates,eval_avg_reward,eval_avg_ep_len,eval_d4rl_score | |||
ding/utils/data/dataset.py
Outdated
|
|
||
| self.context_len = context_len | ||
| def __init__(self, cfg: dict) -> None: | ||
| dataset_path = cfg.policy.collect.get('data_path', None) |
There was a problem hiding this comment.
don't use xxx.get, we must ensure all the configs are fixed after compile_config
|
|
||
| self.running_rtg = [self.rtg_target / self.rtg_scale] * self.eval_batch_size | ||
| self.t = [0] * self.eval_batch_size | ||
| self.timesteps = torch.arange(start=0, end=self.max_eval_ep_len, step=1).repeat(self.eval_batch_size, 1).to(self.device) |
There was a problem hiding this comment.
indicate device when use torch.arange
AltmanD
changed the title
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description
新的pipeline下DT效果不如旧的pipeline好