Fix: Correct token names and correct parser in create-local-model.py #96
+2
−2
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Outdated suggestions cannot be applied.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.
This PR addresses two distinct bugs in the
/metal/scripts/create-local-model.py
script to prevent a runtime error and fix the prog and description outputted.Change Description
There is an inconsistency in the naming of the
<|endofuntrusted|>
special token.In the
o200k_gptoss
definition, it is named"<|endofuntrusted|>"
. But in theSPECIAL_TOKEN_UUID
dict andINCLUDE_SPECIAL_TOKENS
list, it is named"<|end_untrusted|>"
.This mismatch would cause a runtime error.
The Argument Parser that has an incorrect prog name and description, if a user were to run the
--help
flag they would get the idea that this script is for validating the MXFP4 weights but it's not.Change Made
This fix standardizes the name to
end_untrusted
, preventing the error. The second change added the appropriate prog and description.