Repository navigation
Conversation
kevin9327
force-pushed
the
fix/file-reader-measure-context-cap
branch
from
October 5, 2026 07:38
37af17e to
fe830cd
Compare
Evaluation used the full --chunk length from position 0. When chunk is larger than model.context_end, that walk past the rope tables and raises. Training already windows to context; measure now does the same.
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 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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
FileReader._measure(held-out /learn=False) forwardedmin(chunk, remaining)tokens starting at rotary offset 0. Training already windows each step tocontext. When--chunkis larger thanmodel.context_end, the first evaluation read exceeds the rope tables and raises (reading at position N but the rotary tables were built to M).Cap each measure read with
min(n, context - seen)after the cache reset, matching the model ceiling.Test plan
pytest tests/test_file_reader_measure_context.py? fails on main (ValueError: reading at position 64 but the rotary tables were built to 16); passes with this fix (2 tests)Fixes #18.