Instructions to use nvidia/parakeet-tdt-0.6b-v3 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use nvidia/parakeet-tdt-0.6b-v3 with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("automatic-speech-recognition", model="nvidia/parakeet-tdt-0.6b-v3")# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("nvidia/parakeet-tdt-0.6b-v3", device_map="auto") - Inference
- Notebooks
- Google Colab
- Kaggle
Word boosting
Can this model easily be set up to boost certain words?
I noted https://github.com/NVIDIA/NeMo/blob/dcd27ffbaa2ba60e93b8e682e8b033e2382c07c1/docs/source/asr/asr_customization/word_boosting.rst#L4 but a) it's not clear if this applies here and b) it doesn't seem particularly trivial to set up. (As in it's not a simple param you can pass to the invocation of the model).
The word boosting above applies only to CTC decoder models, which this model does not (uses TDT decoder).
Since newer NeMo releases, word boosting is available for TDT models: https://github.com/NVIDIA-NeMo/NeMo/pull/14277
This feature has been introduced in the 2.5.0 release: https://github.com/NVIDIA-NeMo/NeMo/releases/tag/v2.5.0