fosple/german-asr-mixed-whisper

---
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: transkription
dtype: string
- name: source
dtype: string
- name: predicted_transkription
dtype: string
- name: cer
dtype: float64
splits:
- name: train_tuda_0
num_bytes: 13757929586.4
num_examples: 42812
- name: train_cv19_0
num_bytes: 20608244029.34
num_examples: 546561
- name: train_emilia_0
num_bytes: 34405160493
num_examples: 627354
- name: train_mls_0
num_bytes: 102822624468.044
num_examples: 390578
- name: train_emilia_yodas0
num_bytes: 158522556378
num_examples: 1980468
- name: train_eurospeech
num_bytes: 243362644825.86
num_examples: 502515
- name: train_el_tts
num_bytes: 123890289
num_examples: 495
download_size: 570166922667
dataset_size: 573603050069.644
configs:
- config_name: default
data_files:
- split: train_tuda_0
path: data/train_tuda_0-*
- split: train_cv19_0
path: data/train_cv19_0-*
- split: train_emilia_0
path: data/train_emilia_0-*
- split: train_mls_0
path: data/train_mls_0-*
- split: train_emilia_yodas0
path: data/train_emilia_yodas0-*
- split: train_el_tts
path: data/train_el_tts-*
- split: train_eurospeech
path: data/train_eurospeech-*
license: cc-by-nc-4.0
task_categories:
- automatic-speech-recognition
language:
- de
pretty_name: German whisper labeled ASR Collection
---

# Dataset Card

## Dataset Sources and Licensing

This dataset is a mixture of several German and multilingual speech datasets. **For each dataset, the license of the original author applies.** Please consult the linked sources for detailed licensing information and terms of use.

| Dataset Name | Original Source / Author | Link |
|----------------------------|--------------------------------------------------------------|--------------------------------------------------------------|
| TUDA-De German Speech Corpus | LT Group at UHH / TU Darmstadt | https://huggingface.co/datasets/uhhlt/Tuda-De |
| Mozilla Common Voice | Mozilla Foundation | https://commonvoice.mozilla.org/en/datasets |
| Emilia Speech Dataset | Amphion | https://huggingface.co/datasets/amphion/Emilia-Dataset |
| Emilia Yodas Speech Dataset | Amphion | https://huggingface.co/datasets/amphion/Emilia-Dataset |
| Multilingual LibriSpeech (MLS) | Facebook AI Research | https://arxiv.org/abs/2012.03411 |
| EuroSpeech Dataset | ETH Zurich / disco-eth | https://huggingface.co/datasets/disco-eth/EuroSpeech |

**Note:** If a dataset does not have a public source listed, please contact the dataset provider or refer to your data distributor for licensing details.

## Licensing Disclaimer

The use of this dataset and any derived models must comply with the licenses of the original datasets. The dataset creators do not claim any rights to the datasets themselves. Always verify the license terms before using the data or models for commercial or research purposes.