# 📑BDRC-Etexts WG **URL:** https://forum.openpecha.org/c/wg/bec/68.md [Latest](https://forum.openpecha.org/latest.md) · [Categories](https://forum.openpecha.org/categories.md) · [Tags](https://forum.openpecha.org/tags.md) --- ## [ℹ️ BEC Homepage](https://forum.openpecha.org/t/bec-homepage/497)
**Author:** [@Trinley](https://forum.openpecha.org/u/Trinley)\ **Replies:** 0\ **Last updated:** [December 19, 2025, 2:56am UTC](https://forum.openpecha.org/t/bec-homepage/497 "2025-12-19T02:56:26Z")
BDRC Etext Corpus Project Mission: To publish a foundational dataset of Tibetan Buddhist Literature for AI and Research by collecting, OCRing, cataloging, cleaning, and aligning etexts. :high\_voltage: Quick Actions … --- ## [Annotation layer detection in Tibetan texts, part 2: one model per layer, and zero-shot LLMs](https://forum.openpecha.org/t/annotation-layer-detection-in-tibetan-texts-part-2-one-model-per-layer-and-zero-shot-llms/630)
**Author:** [@Karma\_Tashi1](https://forum.openpecha.org/u/Karma_Tashi1)\ **Replies:** 0\ **Last updated:** [September 25, 2026, 10:53am UTC](https://forum.openpecha.org/t/annotation-layer-detection-in-tibetan-texts-part-2-one-model-per-layer-and-zero-shot-llms/630 "2026-09-25T10:53:17Z")
Annotation layer detection in Tibetan texts, part 2: one model per layer, and zero-shot LLMs This continues Annotation Layer Detection in Tibetan Texts — Experiment Report from August. A digitised Tibetan book is one lo… --- ## [Teaching a model to do Tibetan spell checking](https://forum.openpecha.org/t/teaching-a-model-to-do-tibetan-spell-checking/624)
**Author:** [@Yonten](https://forum.openpecha.org/u/Yonten)\ **Replies:** 0\ **Last updated:** [September 15, 2026, 5:58am UTC](https://forum.openpecha.org/t/teaching-a-model-to-do-tibetan-spell-checking/624 "2026-09-15T05:58:43Z")
An English spell checker leans on two things Tibetan does not give you: spaces to find word boundaries, and a dictionary to check the words against. Tibetan separates syllables with a tsek (་), not words. And a syllable… --- ## [BDRC page level outline to char level outline](https://forum.openpecha.org/t/bdrc-page-level-outline-to-char-level-outline/620)
**Author:** [@Kaldan](https://forum.openpecha.org/u/Kaldan)\ **Replies:** 0\ **Last updated:** [September 9, 2026, 11:26am UTC](https://forum.openpecha.org/t/bdrc-page-level-outline-to-char-level-outline/620 "2026-09-09T11:26:50Z")
Introduction For many years, the Buddhist Digital Resource Center (BDRC) has collected and scanned hundreds of thousands of Tibetan manuscripts, creating an archive containing thousands of volumes. BDRC has also made thi… --- ## [ITv2 vs ITv3: Transcription Workflow Comparison Report](https://forum.openpecha.org/t/itv2-vs-itv3-transcription-workflow-comparison-report/619)
**Author:** [@Lobsang\_Tsetan](https://forum.openpecha.org/u/Lobsang_Tsetan)\ **Replies:** 0\ **Last updated:** [September 4, 2026, 7:02am UTC](https://forum.openpecha.org/t/itv2-vs-itv3-transcription-workflow-comparison-report/619 "2026-09-04T07:02:04Z")
1. Workflow Overview This document compares the transcription workflows of ITv2 and ITv3. Aspect ITv2 ITv3 Workflow 2 Annotators → 2 Reviewers → 1 Final Reviewer 3 Annotators → 1 Reviewer People per Task 5 4 … --- ## [Testing a Tibetan quotation-detection model on DharmaBench QUDT](https://forum.openpecha.org/t/testing-a-tibetan-quotation-detection-model-on-dharmabench-qudt/617)
**Author:** [@Karma\_Tashi1](https://forum.openpecha.org/u/Karma_Tashi1)\ **Replies:** 0\ **Last updated:** [August 27, 2026, 6:29am UTC](https://forum.openpecha.org/t/testing-a-tibetan-quotation-detection-model-on-dharmabench-qudt/617 "2026-08-27T06:29:06Z")
What we ran We have been training a token-classification model for structural annotation of Tibetan texts — quotations, root text, sabche, yigchung, and other layers — on OpenPecha/tsadra commentary annotations. The base… --- ## [What Our Spot Checks Tell Us About Text-Boundary Annotation Quality](https://forum.openpecha.org/t/what-our-spot-checks-tell-us-about-text-boundary-annotation-quality/616)
**Author:** [@Kaldan](https://forum.openpecha.org/u/Kaldan)\ **Replies:** 0\ **Last updated:** [August 20, 2026, 11:04am UTC](https://forum.openpecha.org/t/what-our-spot-checks-tell-us-about-text-boundary-annotation-quality/616 "2026-08-20T11:04:31Z")
Our annotation task involves more than identifying where one Tibetan work ends and another begins. Annotators also identify structural material such as front matter, tables of contents, catalogs, and back matter. To ev… --- ## [Detecting Work Boundaries in Tibetan Volumes](https://forum.openpecha.org/t/detecting-work-boundaries-in-tibetan-volumes/615)
**Author:** [@Kaldan](https://forum.openpecha.org/u/Kaldan)\ **Replies:** 0\ **Last updated:** [August 20, 2026, 9:32am UTC](https://forum.openpecha.org/t/detecting-work-boundaries-in-tibetan-volumes/615 "2026-08-20T09:32:22Z")
Where Does One Text End and the Next Begin? A Tibetan volume may carry one work or a dozen, running one after another with no file-level marker between them, and around those works sit the structural parts of the book: f… --- ## [Annotation Layer Detection in Tibetan Texts — Experiment Report](https://forum.openpecha.org/t/annotation-layer-detection-in-tibetan-texts-experiment-report/614)
**Author:** [@Karma\_Tashi1](https://forum.openpecha.org/u/Karma_Tashi1)\ **Replies:** 0\ **Last updated:** [August 19, 2026, 6:14am UTC](https://forum.openpecha.org/t/annotation-layer-detection-in-tibetan-texts-experiment-report/614 "2026-08-19T06:14:30Z")
Annotation Layer Detection in Tibetan Texts — Experiment Report Most tibetan text are layered document. A single page interleaves the root text being commented on (tsawa), the author’s topical outline (sabche), interline… --- ## [Which Spell-Checking Approach Should We Use for Tibetan Text?](https://forum.openpecha.org/t/which-spell-checking-approach-should-we-use-for-tibetan-text/612)
**Author:** [@Yonten](https://forum.openpecha.org/u/Yonten)\ **Replies:** 0\ **Last updated:** [August 6, 2026, 6:14am UTC](https://forum.openpecha.org/t/which-spell-checking-approach-should-we-use-for-tibetan-text/612 "2026-08-06T06:14:02Z")
Tibetan spelling errors come in two sizes. A wrong component inside a syllable is local: you fix it by looking at that syllable. A whole syllable that’s been deleted, transposed, or merged is not, because there’s nothing… --- ## [How Well Do Existing Layout Detection Models Handle Tibetan Books?](https://forum.openpecha.org/t/how-well-do-existing-layout-detection-models-handle-tibetan-books/610)
**Author:** [@Kaldan](https://forum.openpecha.org/u/Kaldan)\ **Replies:** 0\ **Last updated:** [June 22, 2026, 5:37pm UTC](https://forum.openpecha.org/t/how-well-do-existing-layout-detection-models-handle-tibetan-books/610 "2026-06-22T17:37:34Z")
Tibetan texts have a distinct visual structure — running headers (མགོ་བྱང་), dense body text, footnotes, and footers — that differs significantly from the Latin-script documents most document AI models are trained on. As… --- ## [Tibetan Book Page Layout Detection — Experiment Report](https://forum.openpecha.org/t/tibetan-book-page-layout-detection-experiment-report/587)
**Author:** [@Kaldan](https://forum.openpecha.org/u/Kaldan)\ **Replies:** 0\ **Last updated:** [May 15, 2026, 6:20am UTC](https://forum.openpecha.org/t/tibetan-book-page-layout-detection-experiment-report/587 "2026-05-15T06:20:26Z")
Data Preparation First, I collected all the training data annotated by our annotation team on the Ultralytics platform. I combined them and split the data into training, validation, and test sets while preserving the sam… --- ## [Tibetan Script Classification Model training pipeline](https://forum.openpecha.org/t/tibetan-script-classification-model-training-pipeline/582)
**Author:** [@Karma\_Tashi1](https://forum.openpecha.org/u/Karma_Tashi1)\ **Replies:** 0\ **Last updated:** [May 12, 2026, 6:03am UTC](https://forum.openpecha.org/t/tibetan-script-classification-model-training-pipeline/582 "2026-05-12T06:03:09Z")
Tibetan Script Classification Pipeline Building a classifier model to sort Buddhist manuscript images by script type by implementing from dataset analysis ,model selection, preprocessing ablation, and progressive fin… --- ## [Restoring Correct Tibetan Unicode from PDFs by Repairing Their \`/ToUnicode\` Maps](https://forum.openpecha.org/t/restoring-correct-tibetan-unicode-from-pdfs-by-repairing-their-tounicode-maps/578)
**Author:** [@Ganga\_Gyatso](https://forum.openpecha.org/u/Ganga_Gyatso)\ **Replies:** 0\ **Last updated:** [May 11, 2026, 5:43am UTC](https://forum.openpecha.org/t/restoring-correct-tibetan-unicode-from-pdfs-by-repairing-their-tounicode-maps/578 "2026-05-11T05:43:38Z")
Restoring Correct Tibetan Unicode from PDFs by Repairing Their /ToUnicode Maps How OpenPecha’s open-source tool pdf-cmap-fix rebuilds the glyph-to-Unicode tables inside a PDF from the actual source fonts, so extracted T… --- ## [Training a KenLM for Tibetan OCR Quality Assessment: An Empirical Journey](https://forum.openpecha.org/t/training-a-kenlm-for-tibetan-ocr-quality-assessment-an-empirical-journey/549)
**Author:** [@Kaldan](https://forum.openpecha.org/u/Kaldan)\ **Replies:** 0\ **Last updated:** [March 6, 2026, 4:56am UTC](https://forum.openpecha.org/t/training-a-kenlm-for-tibetan-ocr-quality-assessment-an-empirical-journey/549 "2026-03-06T04:56:01Z")
Objective To develop a reliable, lightweight statistical language model capable of assessing the quality of Optical Character Recognition (OCR) outputs for Classical and Contemporary Tibetan texts. Table of Content Con… --- ## [Contributor Sync \[BEC\] - 2025/12/26 18:04 CST - Notes by Gemini](https://forum.openpecha.org/t/contributor-sync-bec-2025-12-26-18-04-cst-notes-by-gemini/504)
**Author:** [@Trinley](https://forum.openpecha.org/u/Trinley)\ **Replies:** 0\ **Last updated:** [December 31, 2025, 8:58am UTC](https://forum.openpecha.org/t/contributor-sync-bec-2025-12-26-18-04-cst-notes-by-gemini/504 "2025-12-31T08:58:02Z")
:link: Minutes :locked: Summary NT updated the team on testing a paid Google Meet account for transcription and improving project setup with milestones and OKRs, while Tashi Tsering reported that the back-end for the ne…