rtf benchmark

游雁

2023-03-17 dce86185d860090eb6f88819053bb58ac8ad457d

rtf benchmark

 funasr/runtime/python/benchmark_libtorch.md

New file
@@ -0,0 +1,34 @@
# Benchmark 

### Data set:
Aishell1 test set , the total audio duration is 36108.919 seconds.

### Tools
- Install 
```shell
git clone https://github.com/alibaba-damo-academy/FunASR.git && cd FunASR
pip install --editable ./
cd funasr/runtime/python/utils
pip install -r requirements.txt
```

- recipe
set the model, data path and output_dir

```shell
nohup bash test_rtf.sh &> log.txt &
```


## [Paraformer-large](https://www.modelscope.cn/models/damo/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch/summary) 

 ### Intel(R) Xeon(R) Platinum 8369B CPU @ 2.90GHz   16core-32processor    with avx512_vnni


### Intel(R) Xeon(R) Platinum 8269CY CPU @ 2.50GHz   16core-32processor    with avx512_vnni


### Intel(R) Xeon(R) Platinum 8163 CPU @ 2.50GHz    32core-64processor   without avx512_vnni


## [Paraformer](https://modelscope.cn/models/damo/speech_paraformer_asr_nat-zh-cn-16k-common-vocab8358-tensorflow1/summary)

 funasr/runtime/python/utils/test_rtf.sh

@@ -10,8 +10,8 @@

#:<<!
model_name="damo/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch"
backend="onnx"
quantize='True'
backend="onnx" # "torch"
quantize='True' # 'False'
tag=${model_name}/${backend}_${quantize}
!