Update README.md
Browse files
README.md
CHANGED
|
@@ -49,8 +49,6 @@ tags:
|
|
| 49 |
EMOVA-ASR-TTS-Eval is a dataset designed for evaluating the ASR and TTS performance of Omni-modal LLMs. It is derived from the **test-clean** set of the LibriSpeech dataset. This dataset is part of the [EMOVA-Datasets](https://huggingface.co/collections/Emova-ollm/emova-dataset-67779be7d02447a2d0891bf6) collection. We extract the speech units using the [EMOVA Speech Tokenizer](https://huggingface.co/Emova-ollm/emova_speech_tokenizer_hf).
|
| 50 |
|
| 51 |
|
| 52 |
-
|
| 53 |
-
|
| 54 |
## Structure
|
| 55 |
|
| 56 |
This dataset contains two types of data samples:
|
|
|
|
| 49 |
EMOVA-ASR-TTS-Eval is a dataset designed for evaluating the ASR and TTS performance of Omni-modal LLMs. It is derived from the **test-clean** set of the LibriSpeech dataset. This dataset is part of the [EMOVA-Datasets](https://huggingface.co/collections/Emova-ollm/emova-dataset-67779be7d02447a2d0891bf6) collection. We extract the speech units using the [EMOVA Speech Tokenizer](https://huggingface.co/Emova-ollm/emova_speech_tokenizer_hf).
|
| 50 |
|
| 51 |
|
|
|
|
|
|
|
| 52 |
## Structure
|
| 53 |
|
| 54 |
This dataset contains two types of data samples:
|