Instructions to use birgermoell/psst-fairseq-rir with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use birgermoell/psst-fairseq-rir with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("automatic-speech-recognition", model="birgermoell/psst-fairseq-rir")# Load model directly from transformers import AutoProcessor, AutoModelForCTC processor = AutoProcessor.from_pretrained("birgermoell/psst-fairseq-rir") model = AutoModelForCTC.from_pretrained("birgermoell/psst-fairseq-rir", device_map="auto") - Notebooks
- Google Colab
- Kaggle
This model is trained on the PSST Challenge data, with a subset of TIMIT that was augmented using Room Impulse Response (RIR). A file containing the list of TIMIT IDs is in the repository (timit-ids.txt)
The model was finetuned on Wav2vec 2.0 Base, No finetuning, and the results on the validation set were PER: 21.8%, FER: 9.6%.
- Downloads last month
- 2