[xh]Optimize prompt for block params generation in HuggingFace #4474
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Outdated suggestions cannot be applied.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
For the task to generate different block params: block_type, pipeline_type, action_code, etc based on block description,
this PR optimizes it by split one prompt to multiple small prompt for individual prompt and run them in parallel.
Since hugging face charge LLM instance by time not number of request, this change does not affect money spending.
However because we can tailer and focus on prompt for each different param, I got 100% correctness with 10 examples I tried: https://docs.google.com/spreadsheets/d/1-fongvJv0lrtNp06wbnDqxmPNp66lCBJ-8TVKuftDds/edit#gid=0
Previously it is 70%.
How Has This Been Tested?
Checklist
docs/mint.json
cc:
@wangxiaoyou1993 @dy46