Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

fix vila prompting #1730

Closed
wants to merge 1 commit into from
+5 −3
Closed

fix vila prompting #1730

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
5 changes: 3 additions & 2 deletions predicators/approaches/vlm_planning_prompts/few_shot.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -9,7 +9,7 @@ Here are a number of example demonstrations of solving a range of goals from an
with the goal the demonstrator was trying to achieve, the plan (i.e., sequence of skills with discrete and continuous arguments specified), as well
as images corresponding to the state of the environment at every timestep (starting at 0, when the 0th skill is executed). Note that the state images
are annotated with text indicating which demonstration and timestep they belong to. Additionally, note that a state might have more than one image
associated with it.
associated with it.

{demonstration_trajs}

Expand All @@ -30,7 +30,8 @@ Finally, here is an expression corresponding to the current task goal that must
Please return a plan that achieves the provided goal from an initial state depicted by the image(s) below.
Please provide your output in the following format (excluding the angle brackets and ellipsis, which are just for illustration purposes).
Be sure to include the parens '(' and ')', as well as square brackets '[' and ']' even if there are no objects/continuous parameters.
Do not bold or italicize or otherwise apply any extra formaating to the plan text:
Do not bold or italicize or otherwise apply any extra formaating to the plan text. Do not provide any numbers for steps in the plan, or
any reasoning for each step below the 'Plan:' heading:
<Explanation of scene + your reasoning>
Plan:
<skill 1 name>(<obj1_name>:<obj1_type_name>, <obj2_name>:<obj2_type_name>, ...)[<continuous_param1_value>, <continuous_param2_value>, ...]
Expand Down
3 changes: 2 additions & 1 deletion predicators/approaches/vlm_planning_prompts/no_few_shot.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -20,7 +20,8 @@ Finally, here is an expression corresponding to the current task goal that must
Please return a plan that achieves the provided goal from an initial state depicted by the image(s) below.
Please provide your output in the following format (excluding the angle brackets and ellipsis, which are just for illustration purposes).
Be sure to include the parens '(' and ')', as well as square brackets '[' and ']' even if there are no objects/continuous parameters.
Do not bold or italicize or otherwise apply any extra formaating to the plan text:
Do not bold or italicize or otherwise apply any extra formaating to the plan text. Do not provide any numbers for steps in the plan, or
any reasoning for each step below the 'Plan:' heading:
<Explanation of scene + your reasoning>
Plan:
<skill 1 name>(<obj1_name>:<obj1_type_name>, <obj2_name>:<obj2_type_name>, ...)[<continuous_param1_value>, <continuous_param2_value>, ...]
Expand Down
Loading