fix(distill): ara-diac-tiny -> run-005 on clean teacher_labels_v2 - #66
Merged
Conversation
run-004 consumed the Aug-23 snapshot (pre decode_joined fix): its best/ decodes as UTF-8-as-Latin1 mojibake and its final_eval.json reproduces the retracted 83.08 exactly - the poisoned-data number, not a capacity result. run-005 is the clean-label re-run.
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Evidence chain: run-004/best decodes "كتاب" as "ÙÙاتÙب"
(UTF-8-as-Latin1 mojibake - the label corruption the Aug-24 retraction
describes), and the 300-paragraph gate eval (final_eval.json, n=300)
reproduces the retracted 83.08 to four decimals with the teacher
reproducing 1.3205. The spec was pointing at the Aug-23 label snapshot
instead of the post-fix teacher_labels_v2.jsonl. run-005 reruns the
tiny tier on clean labels - the falsification test the retraction
actually called for.