AI Tools.

Search
by 1-800-BAD-CODE

xlm-roberta_punctuation_fullstop_truecase

This NeMo-based model performs simultaneous punctuation restoration, full-stop detection, and truecasing on ASR transcripts across 50+ languages. ASR systems typically produce unpunctuated, lowercase text; this model applies punctuation and sentence casing in a single pass to make transcripts readable. The XLM-RoBERTa backbone provides multilingual generalisation across the supported language set.

Summary text generated by an automated pipeline from the model card · Not individually reviewed or run by us · How this page is made

Share: X Reddit HN LinkedIn

From the model card

Fields below are copied from the tags and counters on the HuggingFace repository 1-800-BAD-CODE/xlm-roberta_punctuation_fullstop_truecase at our last fetch. They are set by the uploader, not verified by us; rows with no tag are omitted. How this page is made.

Publisher (HF namespace)
1-800-BAD-CODE
Library
NeMo
Weight formats
ONNX
License tag
apache-2.0 — read the license file in the repo before relying on it
Language tags
Afrikaans (af), Amharic (am), Arabic (ar), Bulgarian (bg), Bangla (bn), German (de), Greek (el), English (en), Spanish (es), Estonian (et), Persian (fa), Finnish (fi), French (fr), Gujarati (gu), Hindi (hi), Croatian (hr), Hungarian (hu), Indonesian (id), Icelandic (is), Italian (it), Japanese (ja), Kazakh (kk), Kannada (kn), Korean (ko), Kyrgyz (ky), Lithuanian (lt), Latvian (lv), Macedonian (mk), Malayalam (ml), Marathi (mr), Dutch (nl), Odia (or), Punjabi (pa), Polish (pl), Pashto (ps), Portuguese (pt), Romanian (ro), Russian (ru), Kinyarwanda (rw), Somali (so), Serbian (sr), Swahili (sw), Tamil (ta), Telugu (te), Turkish (tr), Ukrainian (uk), Chinese (zh)
Downloads (HF counter at last fetch)
399,341
Likes (HF counter at last fetch)
69
Model card
https://huggingface.co/1-800-BAD-CODE/xlm-roberta_punctuation_fullstop_truecase

Use cases

  • Post-processing ASR transcripts for readability and downstream NLP
  • Restoring punctuation in multilingual text extracted from speech
  • Preparing ASR output for translation or summarisation pipelines
  • Formatting voice-to-text outputs in production transcription services
  • Multilingual captioning systems requiring punctuated output

Pros

  • Handles punctuation, full-stop detection, and truecasing in a single model pass
  • 50+ language support via XLM-RoBERTa multilingual backbone
  • NeMo-based with consistent integration into NVIDIA's speech pipeline
  • 67 likes with active ASR post-processing community use

Cons

  • NeMo dependency; not compatible with standard HuggingFace inference pipeline
  • No pipeline_tag; requires manual NeMo model loading which adds integration complexity
  • Punctuation quality varies significantly across languages; low-resource languages have less training data
  • No license explicitly declared; check NeMo model terms

Tags

genericonnxnemotext2text-generationpunctuationsentence-boundary-detectiontruecasingtrue-casingafamarbgbndeelenesetfafi