solve the default 'template_backend' bug in llm.tempalte.base.Templte._encode #4669
+1
−1
Conversation
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
solve the template_backend bug: the correct default should be jinja rather than swift
|
|
|
Hello 😊, what model is this? Why are there spaces between the tokens? I feel like this is actually the core issue here. By default, the template_backend should be "swift"; "jinja" needs to be set manually, since some models do not support the jinja template_backend. |
|
This system message is not accurate. |
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Outdated suggestions cannot be applied.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.
solve the template_backend bug: the correct default should be jinja rather than swift
PR type
PR information
The default template_backend in llm.tempalte.base.Templte._encode should be jinja rather than swift, however, becuase the self.template_meta.template_type in a specific Template class are derived from MLLMTemplateType or LLMTemplateType, which is a specific type(e.g, ovis2) rather than 'dummy', the template_backend is always 'swift'.
Experiment results
Paste your experiment result here(if needed).
the prompt in vllm when inferring ovis:
the prompt in swift:
