search

dusty-nv / jetson-voice

ASR/NLP/TTS deep learning inference library for NVIDIA Jetson using PyTorch and TensorRT
173 stars 46 forks source link

Using .nemo models #7

Open ramainen opened 2 years ago

ramainen commented 2 years ago

How to use .nemo models (for example https://ngc.nvidia.com/catalog/models/nvidia:nemo:stt_ru_quartznet15x5 )?

examples/asr.py --mic 11 --model asr/quartznet-15x5_ru/ru.nemo says "ValueError: resource 'asr/quartznet-15x5_ru/ru.nemo' has invalid extension '.nemo'"

I believe .nemo converted to bin with json somehow.

Actually I looking for pre-trained russian models, but there is no russian in --list-models list.

kurkovpavel commented 1 year ago

https://github.com/dusty-nv/jetson-voice/issues/5 You can find quartznet_golos.nemo for russian language, export nemo file to onnx if you have nemo framework installed on x86, define your russian vocabulary in json file (untar .nemo file for this, you will fine .vocab file). Exported onnx file can be converted to tensorrt engine in jetson-voice on jetson nano. You can also use onnxruntime engine if you prefer.