Dev gzf exp (#1697)

zhifu gao

2024-05-06 f9710e54ce4ce65a99d09d208f13a36e79a5792b

 README.md

@@ -14,6 +14,7 @@
| [**News**](https://github.com/alibaba-damo-academy/FunASR#whats-new) 
| [**Installation**](#installation)
| [**Quick Start**](#quick-start)
| [**Tutorial**](https://github.com/alibaba-damo-academy/FunASR/blob/main/docs/tutorial/README.md)
| [**Runtime**](./runtime/readme.md)
| [**Model Zoo**](#model-zoo)
| [**Contact**](#contact)
@@ -67,7 +68,7 @@
```

## Model Zoo
FunASR has open-sourced a large number of pre-trained models on industrial data. You are free to use, copy, modify, and share FunASR models under the [Model License Agreement](./MODEL_LICENSE). Below are some representative models, for more models please refer to the [Model Zoo]().
FunASR has open-sourced a large number of pre-trained models on industrial data. You are free to use, copy, modify, and share FunASR models under the [Model License Agreement](./MODEL_LICENSE). Below are some representative models, for more models please refer to the [Model Zoo](./model_zoo).

(Note: ⭐ represents the ModelScope model zoo, 🤗 represents the Huggingface model zoo, 🍀 represents the OpenAI model zoo)

@@ -97,7 +98,7 @@
<a name="quick-start"></a>
## Quick Start

Below is a quick start tutorial. Test audio files ([Mandarin](https://isv-data.oss-cn-hangzhou.aliyuncs.com/ics/MaaS/ASR/test_audio/vad_example.wav), [English]()).
Below is a quick start tutorial. Test audio files ([Mandarin](https://isv-data.oss-cn-hangzhou.aliyuncs.com/ics/MaaS/ASR/test_audio/vad_example.wav), [English](https://isv-data.oss-cn-hangzhou.aliyuncs.com/ics/MaaS/ASR/test_audio/asr_example_en.wav)).

### Command-line usage

@@ -112,7 +113,7 @@
from funasr import AutoModel
# paraformer-zh is a multi-functional asr model
# use vad, punc, spk or not as you need
model = AutoModel(model="paraformer-zh",  vad_model="fsmn-vad",  punc_model="ct-punc-c", 
model = AutoModel(model="paraformer-zh",  vad_model="fsmn-vad",  punc_model="ct-punc", 
                  # spk_model="cam++", 
                  )
res = model.generate(input=f"{model.model_path}/example/asr_example.wav", 
@@ -154,7 +155,7 @@
from funasr import AutoModel

model = AutoModel(model="fsmn-vad")
wav_file = f"{model.model_path}/example/asr_example.wav"
wav_file = f"{model.model_path}/example/vad_example.wav"
res = model.generate(input=wav_file)
print(res)
```
@@ -207,8 +208,8 @@
res = model.generate(input=(wav_file, text_file), data_type=("sound", "text"))
print(res)
```

More examples ref to [docs](https://github.com/alibaba-damo-academy/FunASR/tree/main/examples/industrial_data_pretraining)
More usages ref to [docs](docs/tutorial/README_zh.md), 
more examples ref to [demo](https://github.com/alibaba-damo-academy/FunASR/tree/main/examples/industrial_data_pretraining)


## Export ONNX
@@ -227,6 +228,21 @@
res = model.export(quantize=False)
```

### Test ONNX
```python
# pip3 install -U funasr-onnx
from funasr_onnx import Paraformer
model_dir = "damo/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch"
model = Paraformer(model_dir, batch_size=1, quantize=True)

wav_path = ['~/.cache/modelscope/hub/damo/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-pytorch/example/asr_example.wav']

result = model(wav_path)
print(result)
```

More examples ref to [demo](runtime/python/onnxruntime)

## Deployment Service
FunASR supports deploying pre-trained or further fine-tuned models for service. Currently, it supports the following types of service deployment:
- File transcription service, Mandarin, CPU version, done