commonvoice

Adnabod lleferydd Cymraeg i'r Gymraeg gyda HuggingFace // Speech Recognition for Welsh with HuggingFace

Lleisiau synthetig cadwynedig Cymraeg gyda MaryTTS a Docker // Welsh concatenative TTS voices with MaryTTS and Docker

Corpws o frawddegau CC0 mewn fformat jsonl, gyda rhannau ymadrodd y tocynnau (geiriau etc.) wedi'u tagio â thagiau Universal Dependencies. // A Corpus of CC0 sentences in the jsonl format, tagged with Universal Dependency part-of-speech tags.

Corpws o frawddegau o destun Cymraeg wedi'u trwyddedu o dan drwydded CC0 | A corpus of Welsh texts licensed under the CC0 licence

Corpws brawddegau CC0 ar gyfer promptiau adnabod lleferydd Cymraeg // A corpus of CC0 licensed sentences as prompts for Welsh speech recognition training

Creates alternative splits for Mozilla Common Voice datasets for further analysis. Supports delta-version upgrades.

Repository of hashed IDs requested for deletion from Mozilla Common Voice so that we can programmatically remove them from projects that use CV data.

Hyfforddi a defnyddio modelau adnabod lleferydd Cymraeg coqui-stt a KenLM // Train and use coqui-stt and KenLM based Welsh language speech recognition models.

ASR project to learn tensorflow. Model trained on Common Voice Mozilla Dataset in GCP.

A short test to determine the distribution of similarity scores for different SpeechBrain speaker identification models.