Downloads · 30 days
26
11% of all-time downloads
csukuangfj/sherpa-onnx-streaming-paraformer-trilingual-zh-cantonese-en
sherpa-onnx-streaming-paraformer-trilingual-zh-cantonese-en is a machine learning model from csukuangfj. Use it for the machine learning task on the model card, and read the license before you ship it in a product.
Files in this directory are from https://modelscope.cn/models/dengcunqin/speechparaformer-largeasrnat-zh-cantonese-en-16k-vocab8501-online/summary
Downloads · 30 days
26
11% of all-time downloads
All-time downloads
230
Public
Repo size
1.9 GB
Likes
1
Public
Click a slice to open those files.
.onnx1.1 GB · 100%
From the Hugging Face model README
Files in this directory are from https://modelscope.cn/models/dengcunqin/speech_paraformer-large_asr_nat-zh-cantonese-en-16k-vocab8501-online/summary
First, we use the following Python code (using FunASR v0.8.8)
from funasr_onnx.paraformer_online_bin import Paraformer
model_dir = "dengcunqin/speech_paraformer-large_asr_nat-zh-cantonese-en-16k-vocab8501-online"
chunk_size = [5, 10, 5]
model = Paraformer(model_dir, batch_size=1, quantize=True, chunk_size=chunk_size, intra_op_num_threads=4) #, revision='v1.0.0') # only support batch_size = 1
to get the following files:
Then we rename the following three files:
Then we run
./generate-tokens.py
to get tokens.txt.
Finally, we run
./add-model-metadata.py
to add metadata to encoder.onnx and encoder.int8.onnx.
Now you can use the generated files in sherpa-onnx.