Repository navigation
Train CRF on batches 01-02; report agreement with model drafts - #32
Merged
Merged
Conversation
crf-batches-01-02 adds batch-02 (239 train records). On the 172 model-drafted dev records it reaches typed F1 0.145 and exact span F1 0.221; adding batch-02 clearly improves span agreement (+0.047, CI [+0.004, +0.086]). On the 10 hand-labelled dev documents no difference is clear. evaluate and compare take --add-gold; predict takes several --sample manifests.
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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
crf-batches-01-02: the CRF trained onpilot-2+batch-01+batch-02(239 train-split records: 7manual, 232 model drafts).reports/agreement/: rules-2 and two CRFs scored on dev in two ways:batch-03+batch-06), labelled as such and never as accuracy. Typed F1 0.107 / 0.128 / 0.145; exact span F1 0.142 / 0.174 / 0.221.crf-batches-01-02typed F1 0.370 [0.000, 0.640].batch-02clearly improves exact-span agreement (+0.047 [+0.004, +0.086]), but not typed agreement. CRF vs rules: span agreement +0.079 [+0.028, +0.138]. On the hand-labelled slice nothing is clear.evaluate/comparetake--add-gold(several gold files, a repeated id is refused);predict --sampletakes several manifests (union, each document once). New tests for both.The test split was not touched.
Checks
ruff check,ruff format --check,mypy src/,pytest -m "not slow"(234 passed).