extract text when processing multi-modal prompts#1228
Merged
leondz merged 2 commits intoNVIDIA:mainfrom May 23, 2025
Merged
Conversation
Signed-off-by: Jeffrey Martin <[email protected]>
leondz
reviewed
May 22, 2025
Collaborator
leondz
left a comment
leondz
left a comment
There was a problem hiding this comment.
defensive guards in pre-Turn prompt structure
* guard for dictionary based prompt item * revert note change to hold original prompt in full Signed-off-by: Jeffrey Martin <[email protected]>
leondz
approved these changes
May 23, 2025
This file contains hidden or 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
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
#943 introduces translation of
prompttext as an action taken during execution of the baseProbe.probe()implementation with an assumption that all prompt entries arestrtype. Thevisual_jailbreakprobe is a current outlier that builds a list of prompts with entires ofdicttype.Different prompt types require different call patterns to language services.
Verification
List the steps needed to make sure this thing works
visual_jailbreakagainst a model withtarget_langset to something other than"en"