Skip to content

Fix thai2rom_onnx: patch ONNX encoder model and fix inference bugs #1349

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
wannaphong merged 3 commits into from
Mar 20, 2026
Merged

Fix thai2rom_onnx: patch ONNX encoder model and fix inference bugs #1349

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
478f4e1
Initial plan
Copilot Mar 20, 2026
e45fe0b
Fix thai2rom_onnx: patch ONNX encoder model and fix Python code bugs
Copilot Mar 20, 2026
f901875
Refactor: group comment for both int-keyed dict comprehensions
Copilot Mar 20, 2026
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
Binary file modified pythainlp/corpus/thai2rom_encoder.onnx
View file
Open in desktop
Binary file not shown.
11 changes: 8 additions & 3 deletions pythainlp/transliterate/thai2rom_onnx.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -66,9 +66,14 @@ def __init__(self) -> None:
self._maxlength: int = 100

self._char_to_ix: Dict[str, int] = loader["char_to_ix"]
self._ix_to_char: Dict[int, str] = loader["ix_to_char"]
self._target_char_to_ix: Dict[str, int] = loader["target_char_to_ix"]
self._ix_to_target_char: Dict[int, str] = loader["ix_to_target_char"]
# JSON keys are always strings; convert to int for index-based lookup.
self._ix_to_char: Dict[int, str] = {
int(k): v for k, v in loader["ix_to_char"].items()
}
self._ix_to_target_char: Dict[int, str] = {
int(k): v for k, v in loader["ix_to_target_char"].items()
}

# encoder/ decoder
# Load encoder decoder onnx models.
Expand Down Expand Up @@ -219,7 +224,7 @@ def run(

decoder_input = np.array([topi])

if decoder_input == end_token:
if decoder_input.item() == end_token:
return outputs[:di]

return outputs
Expand Down
2 changes: 1 addition & 1 deletion tests/extra/testx_transliterate.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -41,7 +41,7 @@ def test_romanize_thai2rom_onnx(self):
romanize("ความอิ่ม", engine="thai2rom_onnx"), "khwam-im"
)
self.assertEqual(
romanize("กานต์ ณรงค์", engine="thai2rom_onnx"), "kan narong"
romanize("กานต์ ณรงค์", engine="thai2rom_onnx"), "kan narang"

@bact bact Mar 20, 2026

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Note that "kan narang" is a wrong transliteration.

The transliteration should be along these lines:

  • กานต์ / กาน / kan
  • ณรงค์ / นะ-รง / na-rong

@wannaphong wannaphong Mar 20, 2026

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It's weird. It should get same output. I think new onnxruntime may change the output.

)
self.assertEqual(romanize("สกุนต์", engine="thai2rom_onnx"), "sakun")
self.assertEqual(romanize("ชารินทร์", engine="thai2rom_onnx"), "charin")
Expand Down
Loading