Datasets:

starrydark
/

Makise_Kurisu_Voice

Tasks:

Automatic Speech Recognition

Modalities:

Formats:

Languages:

Size:

Tags:

parallel-corpus

paralinguistics

Libraries:

License:

Dataset card Data Studio Files Files and versions

starrydark commited on 10 days ago

Commit

301a7fe

·

verified ·

1 Parent(s): 365beee

Link the GitHub code repository

Files changed (1) hide show

README.md +1 -0

README.md CHANGED Viewed

@@ -371,6 +371,7 @@ These steps produced the [Makise Kurisu Qwen3-TTS model](https://huggingface.co/
 6. **Use the fixed splits.** They are cut between performances, so the validation and test scores are honest.
 The full recipe used for the model is in the model's [training code](https://huggingface.co/starrydark/Makise_Kurisu_Qwen3_TTS/tree/main/training_code): settings, the trainer fixes, the evaluation and the selection rule.
 ## Kurisu TTS v3 labels (`labels/kurisu_tts_v3/`)

 6. **Use the fixed splits.** They are cut between performances, so the validation and test scores are honest.
 The full recipe used for the model is in the model's [training code](https://huggingface.co/starrydark/Makise_Kurisu_Qwen3_TTS/tree/main/training_code): settings, the trainer fixes, the evaluation and the selection rule.
+The same code, plus the build of this dataset, is on GitHub at [starryark/makise-kurisu-tts](https://github.com/starryark/makise-kurisu-tts).
 ## Kurisu TTS v3 labels (`labels/kurisu_tts_v3/`)