{"id":20049768,"url":"https://github.com/navalnica/be_nlp_speech_resources","last_synced_at":"2026-03-05T19:50:41.233Z","repository":{"id":126652281,"uuid":"550284655","full_name":"navalnica/be_nlp_speech_resources","owner":"navalnica","description":"Links to Belarusian NLP and Speech resources","archived":false,"fork":false,"pushed_at":"2024-09-17T18:48:13.000Z","size":40,"stargazers_count":36,"open_issues_count":1,"forks_count":0,"subscribers_count":5,"default_branch":"main","last_synced_at":"2025-01-12T20:46:28.501Z","etag":null,"topics":["asr","belarus","belarusian","belarusian-language","natural-language-processing","nlp","speech","speech-processing","speech-recognition","speech-synthesis","speech-to-text","stt","text-to-speech","tts"],"latest_commit_sha":null,"homepage":"","language":null,"has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/navalnica.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2022-10-12T13:56:12.000Z","updated_at":"2024-11-27T13:17:31.000Z","dependencies_parsed_at":"2024-11-13T11:52:27.552Z","dependency_job_id":"768f4fff-ddc9-475c-ada0-3c03325637f4","html_url":"https://github.com/navalnica/be_nlp_speech_resources","commit_stats":null,"previous_names":["navalnica/be_nlp_speech_resources","yks72p/be_nlp_speech_resources"],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/navalnica%2Fbe_nlp_speech_resources","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/navalnica%2Fbe_nlp_speech_resources/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/navalnica%2Fbe_nlp_speech_resources/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/navalnica%2Fbe_nlp_speech_resources/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/navalnica","download_url":"https://codeload.github.com/navalnica/be_nlp_speech_resources/tar.gz/refs/heads/main","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":241480340,"owners_count":19969630,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["asr","belarus","belarusian","belarusian-language","natural-language-processing","nlp","speech","speech-processing","speech-recognition","speech-synthesis","speech-to-text","stt","text-to-speech","tts"],"created_at":"2024-11-13T11:52:15.286Z","updated_at":"2026-03-05T19:50:36.175Z","avatar_url":"https://github.com/navalnica.png","language":null,"funding_links":[],"categories":[],"sub_categories":[],"readme":"# Belarusian NLP and Speech Processing resources\n\nThis repository contains links to Belarusian Natural Language and Speech Processing resources and datasets.\u003cbr\u003e\n\nIt is inspired by similar project with Ukrainian Speech Processing resources: [egorsmkv/speech-recognition-uk](https://github.com/egorsmkv/speech-recognition-uk)\n\n### TODOs:\n* add detailed descriptions to each of list items\n* evaluate models on benchmarks and log their performance\n\n# 🎙 Speech-to-Text\n\n## 🎙💡 Implementations\n\n* wav2vec2 trained on Common Voice 8 + kenlm language model trained on Common Voice 8:\n  * Model: [ales/wav2vec2-cv-be](https://huggingface.co/ales/wav2vec2-cv-be)\n  * Demo: [ales/wav2vec2-cv-be-lm](https://huggingface.co/spaces/ales/wav2vec2-cv-be-lm)\n  * Code: [navalnica/wav2vec2-belarusian](https://github.com/navalnica/wav2vec2-belarusian)\n\n* whisper:\n  * original [openai/whisper](https://github.com/openai/whisper) models\n  * Whisper models fine-tuned on Belarusian Common Voice 11 dataset:\n    * Whisper Small:\n      * Model: [ales/whisper-small-belarusian](https://huggingface.co/ales/whisper-small-belarusian)\n      * test WER on CommonVoice11: `6.79`\n      * Demo: [ales/whisper-small-belarusian-demo](https://huggingface.co/spaces/ales/whisper-small-belarusian-demo)\n      * Code: [navalnica/whisper-finetuning-be](https://github.com/navalnica/whisper-finetuning-be)\n    * Whisper Base:\n      * Model: [ales/whisper-base-belarusian](https://huggingface.co/ales/whisper-base-belarusian)\n      * Code: [navalnica/whisper-finetuning-be](https://github.com/navalnica/whisper-finetuning-be)\n    \n* Nvidia NeMo models:\n  * [nvidia/stt_be_conformer_ctc_large](https://huggingface.co/nvidia/stt_be_conformer_ctc_large)\n    * [huggingface self-reported metric] test WER on CommonVoice10: `4.8`\n  * [nvidia/stt_be_conformer_transducer_large](https://huggingface.co/nvidia/stt_be_conformer_transducer_large)\n    * [huggingface self-reported metric] test WER on CommonVoice10: `3.8`\n  * [nvidia/stt_be_fastconformer_hybrid_large_pc](https://huggingface.co/nvidia/stt_be_fastconformer_hybrid_large_pc)\n    * [huggingface self-reported metric] test WER on CommonVoice12: `2.72`\n    * [huggingface self-reported metric] test WER P\u0026C CommonVoice12: `3.87`\n \n* ESPnet:\n  * [espnet/belarusian_commonvoice_blstm](https://huggingface.co/espnet/belarusian_commonvoice_blstm)\n\n## 🎙📊 Benchmarks\n\nModel comparisons grouped by dataset. TODO\n\n## 🎙📚 Datasets\n\n* [Common Voice](https://commonvoice.mozilla.org/en/datasets). Speech recognition dataset\n* Dataset from [knihi.com](https://knihi.com/none/Korpus_bielaruskaha_maulennia_dla_trenirouki_niejronnych_sietak_zip.html). TODO: what is the type of dataset?\n* [google/fleurs](https://huggingface.co/datasets/google/fleurs/viewer/be_by/train)\n* ssrlab: TODO. Speech recognition dataset\n\n------\n\n# 📢 Text-to-Speech\n\n## 📢💡 Implementations\n\n* CoquiAI implementations\n  * [jhlfrfufyfn/bel-tts](https://github.com/jhlfrfufyfn/bel-tts). GlowTTS + HifiGan\n    * [Code](https://github.com/jhlfrfufyfn/bel-tts)\n    * [Model](https://huggingface.co/jhlfrfufyfn/bel-tts)\n    * [Demo on HuggingFace](https://huggingface.co/spaces/jhlfrfufyfn/bel-tts)\n    * [Demo on a custom web-page](https://nikuchin.fun/tts). The source code for the demo page: [here](https://github.com/jhlfrfufyfn/bel-tts-server)\n  * [alex73/belarusian-tts](https://github.com/alex73/belarusian-tts). CoquiAI implementation by Yurii Paniv (@robinhad).\u003cbr\u003e\n    Original repo \u0026 models were deleted - only fork is available now\n\n---\n\n# 📝 NLP\n\n## POS-tagging\n* [KoichiYasuoka/roberta-small-belarusian-upos](https://huggingface.co/KoichiYasuoka/roberta-small-belarusian-upos)\n* [stanfordnlp/stanza-be](https://huggingface.co/stanfordnlp/stanza-be)\n* [poritski/YABC_Tagger](https://github.com/poritski/YABC_Tagger). Rule-based POS-tagger and lemmatizer.\u003cbr\u003e\n  Written in Perl. \n  Uses [poritski/YABC](https://github.com/poritski/YABC) as a Grammar base (?)\n* [volchek/beltagger](https://github.com/volchek/beltagger).\n  An improved version of [poritski/YABC_Tagger](https://github.com/poritski/YABC_Tagger) rule-based POS-tagger and lemmatizer.\u003cbr\u003e\n  Cross-platform, written in C++.\u003cbr\u003e\n  Known issues:\n  * requires input data to be incoded in Windows-1251, does not support UTF-8;\n  * tagset is not fully-compatible with BNKorpus's tagset and grammar base\n  * grammar base used is not full enough. [Belarus/GrammarDB](https://github.com/Belarus/GrammarDB) is a better paradigms source but is not incorporated yet\n  * suffix table calculation script is not ported from Perl to C++\n  * code uses Boost libarary  \n  \n## Other\n* [pkasila/bel-sklony](https://github.com/pkasila/bel-sklony) - web page with Belarusian nouns declension. Demo: [sklony.pkasila.net](https://sklony.pkasila.net/)\n\n## Masked Language Modeling\n* [KoichiYasuoka/roberta-small-belarusian](https://huggingface.co/KoichiYasuoka/roberta-small-belarusian)\n\n## 📝📚 Datasets\n\n* [oscar](https://huggingface.co/datasets/oscar)\n* [mc4](https://huggingface.co/datasets/mc4)\n* [poritski/YABC](https://github.com/poritski/YABC) - Эксперыментальны корпус беларускай мовы, ЭКБМ\n* [Belarus/GrammarDB](https://github.com/Belarus/GrammarDB) - Grammar Database of Belarusian language\n* [tsimafeip/Translator](https://github.com/tsimafeip/Translator) - Dataset with russian-belarusian translation pairs\n* Universal dependencies dataset:\n  * [Page](https://universaldependencies.org/treebanks/be_hse/index.html)\n  * [GitHub Repository](https://github.com/UniversalDependencies/UD_Belarusian-HSE)\n* [Tatoeba Belarusian sentences](https://tatoeba.org/en/sentences/show_all_in/bel/none)\n\n---\n\n# 🧍‍♀️🧍 Communities and platforms:\n* [corpus.by](https://www.corpus.by)\n* [ssrlab.by](https://ssrlab.by)\n* [bnkorpus.info](https://bnkorpus.info)\n* [Belarus](https://github.com/Belarus) organization on github\n* [nlproc.by](https://github.com/nlprocby) community on github\n\n---\n# 🦔 Unsorted\n* nothing for now\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fnavalnica%2Fbe_nlp_speech_resources","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fnavalnica%2Fbe_nlp_speech_resources","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fnavalnica%2Fbe_nlp_speech_resources/lists"}