wangzixiang
/
lerobot
mirror of https://github.com/huggingface/lerobot.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Adjust the timestamps' description in Diffusion Policy (#343) 

Co-authored-by: Alexander Soare <alexander.soare159@gmail.com>
This commit is contained in:
Halvard Bariller 2024-07-26 13:47:03 +02:00 committed by GitHub
parent f8a6574698
commit 7a3cb1ad34
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 3 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
lerobot/common/policies/diffusion/modeling_diffusion.py
View File

@ -111,12 +111,12 @@ class DiffusionPolicy(nn.Module, PyTorchModelHubMixin):
Schematically this looks like: Schematically this looks like:
---------------------------------------------------------------------------------------------- ----------------------------------------------------------------------------------------------
(legend: o = n_obs_steps, h = horizon, a = n_action_steps) (legend: o = n_obs_steps, h = horizon, a = n_action_steps)
|timestep | n-o+1 | n-o+2 | ..... | n | ..... | n+a-1 | n+a | ..... |n-o+1+h| |timestep | n-o+1 | n-o+2 | ..... | n | ..... | n+a-1 | n+a | ..... | n-o+h |
|observation is used | YES | YES | YES | NO | NO | NO | NO | NO | NO | |observation is used | YES | YES | YES | YES | NO | NO | NO | NO | NO |
|action is generated | YES | YES | YES | YES | YES | YES | YES | YES | YES | |action is generated | YES | YES | YES | YES | YES | YES | YES | YES | YES |
|action is used | NO | NO | NO | YES | YES | YES | NO | NO | NO | |action is used | NO | NO | NO | YES | YES | YES | NO | NO | NO |
---------------------------------------------------------------------------------------------- ----------------------------------------------------------------------------------------------
Note that this means we require: `n_action_steps < horizon - n_obs_steps + 1`. Also, note that Note that this means we require: `n_action_steps <= horizon - n_obs_steps + 1`. Also, note that
"horizon" may not the best name to describe what the variable actually means, because this period is "horizon" may not the best name to describe what the variable actually means, because this period is
actually measured from the first observation which (if `n_obs_steps` > 1) happened in the past. actually measured from the first observation which (if `n_obs_steps` > 1) happened in the past.
""" """