Instructions to use nvidia/parakeet-tdt-0.6b-v2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- NeMo
How to use nvidia/parakeet-tdt-0.6b-v2 with NeMo:
import nemo.collections.asr as nemo_asr asr_model = nemo_asr.models.ASRModel.from_pretrained("nvidia/parakeet-tdt-0.6b-v2") transcriptions = asr_model.transcribe(["file.wav"]) - Notebooks
- Google Colab
- Kaggle
Besides GPU, can any other edge accellerators run it. EX: (Hailo AI Hat for RPI)
Can the parakeet model run on the Hailo AI Hat for my Raspberry Pi project?
How about Nvidia's Orin Jetson Nano?
Hi @Flyingcrabs
I'm not exactly sure since I don't have these devices, but judging by these examples they support onnxruntime:
https://github.com/hailo-ai/Hailo-Application-Code-Examples/tree/main/runtime/hailo-8/python/hailo_onnxruntime
https://www.elinux.org/Jetson_Zoo#ONNX_Runtime
If this works, you can use my library - onnx-asr to run parakeet models.
Thank you for the response, but my question is really about if the hardware such as the HailoAIHat can acutally run the model in a timely manner. The HAILO HAT is good at convolution and quantization. Will this model be a good fit?
I am running it in HF Space in real-time mode with 2 vCPU only so I bet it can.
https://huggingface.co/spaces/WJ88/NVIDIA-Parakeet-TDT-0.6B-v2-INT8-Real-Time-Mic-Transcription