---
license: other
license_name: research-only
license_link: LICENSE
viewer: False
---
This dataset is a derivative of the gold-standard [Hebrew Speech](https://huggingface.co/datasets/upai-inc/saspeech) dataset. It contains the exact pre-processed files used to train the **Hebrew-Fish** model available in this repository. Additionally, it includes the protobuf file used by Fish during fine-tuning.