Wav2vec2 Large Xlsr 53 Hungarian is an automatic speech recognition model available on Hugging Face. It is a fine-tuned version of the XLSR-53 wav2vec2 architecture adapted specifically for processing Hungarian audio and producing transcriptions.
The model accepts audio input and outputs text transcriptions in Hungarian. It is implemented in the Transformers library and supports both high-level pipeline usage and direct loading of the processor and model components. The underlying architecture relies on PyTorch and JAX frameworks. The repository indicates it was trained using data from the Common Voice corpus.
Developers and researchers working with Hungarian speech data can integrate the model into applications for transcription tasks. It is delivered as a downloadable model on the Hugging Face platform, with example code provided for loading via the Transformers pipeline or through AutoProcessor and AutoModelForCTC classes. Notebooks for Google Colab and Kaggle are referenced for experimentation.
The model carries an Apache-2.0 license, making it available for open use, modification, and distribution. A DOI identifier links to associated metadata for the fine-tuning work.
Wav2vec2 Large Xlsr 53 Hungarian sits in PulseGate's Voice, TTS & speech category. It enables accurate automatic speech recognition for Hungarian language audio recordings. Wav2vec2 Large Xlsr 53 Hungarian is an open-source project aimed at speech technology researchers and developers. The project is open source (Open Source). Wav2vec2 Large Xlsr 53 Hungarian is available on the web, the command line, and API, and it can be self-hosted.
jonatasgrosman builds and maintains Wav2vec2 Large Xlsr 53 Hungarian, and the product first shipped in 2022. Among its 5 catalogued features are speech recognition, hungarian language, and audio transcription.
Latest indexed changes and source events
jonatasgrosman/wav2vec2-large-xlsr-53-hungarian verified by the PulseGate indexer
Other apps tracked under the same category.
