fix(model): guard prediction against empty tiles#58
Open
EliHei2 wants to merge 1 commit into
Open
Conversation
Two robustness fixes that surface on sparse/empty tiles (common at scale, e.g. MERSCOPE): - LitISTEncoder.predict_step: scatter_max returns -1 for transcripts with no candidate boundary; main only checked `max_idx < dst.shape[0]` so a -1 was treated as valid and indexed wrongly. Add the `max_idx >= 0` guard. - Positional2dEmbedder.forward: return zeros for empty `pos`/`batch` instead of crashing on min/max over an empty tensor; cast batch to long. What to review: the one-line `valid` guard in lightning_model.py and the two empty-tensor early-returns in ist_encoder.py. No behavior change on non-empty input.
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
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Two robustness fixes that surface on sparse/empty tiles (common at scale, e.g. MERSCOPE):
max_idx < dst.shape[0]so a -1 was treated as valid and indexed wrongly. Add themax_idx >= 0guard.pos/batchinstead of crashing on min/max over an empty tensor; cast batch to long.What to review: the one-line
validguard in lightning_model.py and the two empty-tensor early-returns in ist_encoder.py. No behavior change on non-empty input.