Instructions to use facebook/wav2vec2-large-xlsr-53 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use facebook/wav2vec2-large-xlsr-53 with Transformers:
# Load model directly from transformers import AutoProcessor, AutoModelForPreTraining processor = AutoProcessor.from_pretrained("facebook/wav2vec2-large-xlsr-53") model = AutoModelForPreTraining.from_pretrained("facebook/wav2vec2-large-xlsr-53") - Notebooks
- Google Colab
- Kaggle
Paid Task: VTL Extraction and Bio-metric Profiling from 48kHz RAW reference
#10
by Stellaresearch - opened
Hi everyone,
I am looking for a developer or a speech scientist experienced with ECAPA-TDNN v2 and VTL (Vocal Tract Length) estimation.
I have a high-quality 48kHz RAW reference file and I need the following parameters extracted for a technical report:
Mean VTL (cm) using LPC Order 24.
Formant mapping (F1-F4).
Spectral nulls measurement.
This is a paid consulting task. If you have a running environment for these models and can provide the raw data logs, please contact me immediately.
Thank you!