Skip to content

fix(artifact-score): token-based generation cap #188

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
ronaldtse merged 1 commit into from
Sep 5, 2026
Merged

fix(artifact-score): token-based generation cap #188

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 5 additions & 1 deletion src/gpu/modal_score_artifact.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -83,7 +83,11 @@ def score(zip_path: str, expect_sha: str = "", out: str = "") -> dict:
stripped = strip_diacritics(src)
pieces = []
for w in split_windows(stripped, 1400):
out_text = model.translate(w, max_len=max(256, 2 * len(w)))
# the torch harness caps generation at 2x the window's
# TOKEN count; a byte-level window's tokens = its UTF-8
# byte count, and vocalized output needs ~2x that — a
# character-count cap truncates every long paragraph
out_text = model.translate(w, max_len=max(256, 2 * len(w.encode("utf-8"))))
pieces.append(project_haraqat(out_text, w))
preds.append("".join(pieces))
if (i + 1) % 100 == 0:
Expand Down
Loading