fix win bug

游雁

2024-01-16 bbbf17e4d97ff155049c424af4e96bfded9089b1

 README_zh.md

@@ -60,14 +60,13 @@
|                                                                             模型名字                                                                             |        任务详情        |     训练数据     | 参数量  |
|:------------------------------------------------------------------------------------------------------------------------------------------------------------:|:------------------:|:------------:|:----:|
| paraformer-zh <br> ([⭐](https://www.modelscope.cn/models/damo/speech_paraformer-large-vad-punc_asr_nat-zh-cn-16k-common-vocab8404-pytorch/summary)  [🤗]() ) |  语音识别，带时间戳输出，非实时   |  60000小时，中文  | 220M |
|             paraformer-zh-spk <br> ( [⭐](https://modelscope.cn/models/damo/speech_paraformer-large-vad-punc-spk_asr_nat-zh-cn/summary)  [🤗]() )             | 分角色语音识别，带时间戳输出，非实时 |  60000小时，中文  | 220M |
|   paraformer-zh-streaming <br> ( [⭐](https://modelscope.cn/models/damo/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-online/summary) [🤗]() )   |      语音识别，实时       |  60000小时，中文  | 220M |
|      paraformer-en <br> ( [⭐](https://www.modelscope.cn/models/damo/speech_paraformer-large-vad-punc_asr_nat-en-16k-common-vocab10020/summary) [🤗]() )      | 语音识别，非实时 |  50000小时，英文  | 220M |
|                                                            paraformer-en-spk <br> ([⭐]() [🤗]() )                                                            |      语音识别，非实时      |  50000小时，英文  | 220M |
|                  conformer-en <br> ( [⭐](https://modelscope.cn/models/damo/speech_conformer_asr-en-16k-vocab4199-pytorch/summary) [🤗]() )                   |      语音识别，非实时      |  50000小时，英文  | 220M |
|                  ct-punc <br> ( [⭐](https://modelscope.cn/models/damo/punc_ct-transformer_cn-en-common-vocab471067-large/summary) [🤗]() )                   |      标点恢复      |  100M，中文与英文  | 1.1G | 
|                       fsmn-vad <br> ( [⭐](https://modelscope.cn/models/damo/speech_fsmn_vad_zh-cn-16k-common-pytorch/summary) [🤗]() )                       |     语音端点检测，实时      | 5000小时，中文与英文 | 0.4M | 
|                       fa-zh <br> ( [⭐](https://modelscope.cn/models/damo/speech_timestamp_prediction-v1-16k-offline/summary) [🤗]() )                        |   字级别时间戳预测         |  50000小时，中文  | 38M  |
| paraformer-zh-spk <br> ( [⭐](https://modelscope.cn/models/damo/speech_paraformer-large-vad-punc-spk_asr_nat-zh-cn/summary)  [🤗]() )             | 分角色语音识别，带时间戳输出，非实时 |  60000小时，中文  | 220M |
| paraformer-zh-streaming <br> ( [⭐](https://modelscope.cn/models/damo/speech_paraformer-large_asr_nat-zh-cn-16k-common-vocab8404-online/summary) [🤗]() )   |      语音识别，实时       |  60000小时，中文  | 220M |
| paraformer-en <br> ( [⭐](https://www.modelscope.cn/models/damo/speech_paraformer-large-vad-punc_asr_nat-en-16k-common-vocab10020/summary) [🤗]() )      | 语音识别，非实时 |  50000小时，英文  | 220M |
| conformer-en <br> ( [⭐](https://modelscope.cn/models/damo/speech_conformer_asr-en-16k-vocab4199-pytorch/summary) [🤗]() )                   |      语音识别，非实时      |  50000小时，英文  | 220M |
| ct-punc <br> ( [⭐](https://modelscope.cn/models/damo/punc_ct-transformer_cn-en-common-vocab471067-large/summary) [🤗]() )                   |      标点恢复      |  100M，中文与英文  | 1.1G | 
| fsmn-vad <br> ( [⭐](https://modelscope.cn/models/damo/speech_fsmn_vad_zh-cn-16k-common-pytorch/summary) [🤗]() )                       |     语音端点检测，实时      | 5000小时，中文与英文 | 0.4M | 
| fa-zh <br> ( [⭐](https://modelscope.cn/models/damo/speech_timestamp_prediction-v1-16k-offline/summary) [🤗]() )                        |   字级别时间戳预测         |  50000小时，中文  | 38M  |


<a name="快速开始"></a>
@@ -86,12 +85,15 @@
### 非实时语音识别
```python
from funasr import AutoModel

model = AutoModel(model="paraformer-zh")
# for the long duration wav, you could add vad model
# model = AutoModel(model="paraformer-zh", vad_model="fsmn-vad", punc_model="ct-punc")

res = model(input="asr_example_zh.wav", batch_size=64)
# paraformer-zh is a multi-functional asr model
# use vad, punc, spk or not as you need
model = AutoModel(model="paraformer-zh", model_revision="v2.0.2", \
                  vad_model="fsmn-vad", vad_model_revision="v2.0.2", \
                  punc_model="ct-punc-c", punc_model_revision="v2.0.2", \
                  spk_model="cam++", spk_model_revision="v2.0.2")
res = model.generate(input=f"{model.model_path}/example/asr_example.wav", 
            batch_size=64, 
            hotword='魔搭')
print(res)
```
注：`model_hub`：表示模型仓库，`ms`为选择modelscope下载，`hf`为选择huggingface下载。
@@ -105,7 +107,7 @@
encoder_chunk_look_back = 4 #number of chunks to lookback for encoder self-attention
decoder_chunk_look_back = 1 #number of encoder chunks to lookback for decoder cross-attention

model = AutoModel(model="paraformer-zh-streaming", model_revision="v2.0.0")
model = AutoModel(model="paraformer-zh-streaming", model_revision="v2.0.2")

import soundfile
import os
@@ -119,13 +121,7 @@
for i in range(total_chunk_num):
    speech_chunk = speech[i*chunk_stride:(i+1)*chunk_stride]
    is_final = i == total_chunk_num - 1
    res = model(input=speech_chunk,
                cache=cache,
                is_final=is_final,
                chunk_size=chunk_size,
                encoder_chunk_look_back=encoder_chunk_look_back,
                decoder_chunk_look_back=decoder_chunk_look_back,
                )
    res = model.generate(input=speech_chunk, cache=cache, is_final=is_final, chunk_size=chunk_size, encoder_chunk_look_back=encoder_chunk_look_back, decoder_chunk_look_back=decoder_chunk_look_back)
    print(res)
```

@@ -138,7 +134,7 @@
model = AutoModel(model="fsmn-vad", model_revision="v2.0.2")

wav_file = f"{model.model_path}/example/asr_example.wav"
res = model(input=wav_file)
res = model.generate(input=wav_file)
print(res)
```

@@ -160,11 +156,7 @@
for i in range(total_chunk_num):
    speech_chunk = speech[i*chunk_stride:(i+1)*chunk_stride]
    is_final = i == total_chunk_num - 1
    res = model(input=speech_chunk,
                cache=cache,
                is_final=is_final,
                chunk_size=chunk_size,
                )
    res = model.generate(input=speech_chunk, cache=cache, is_final=is_final, chunk_size=chunk_size)
    if len(res[0]["value"]):
        print(res)
```
@@ -173,9 +165,9 @@
```python
from funasr import AutoModel

model = AutoModel(model="ct-punc", model_revision="v2.0.1")
model = AutoModel(model="ct-punc", model_revision="v2.0.2")

res = model(input="那今天的会就到这里吧 happy new year 明年见")
res = model.generate(input="那今天的会就到这里吧 happy new year 明年见")
print(res)
```

@@ -186,12 +178,11 @@
model = AutoModel(model="fa-zh", model_revision="v2.0.0")

wav_file = f"{model.model_path}/example/asr_example.wav"
text_file = f"{model.model_path}/example/asr_example.wav"
res = model(input=(wav_file, text_file),
            data_type=("sound", "text"))
text_file = f"{model.model_path}/example/text.txt"
res = model.generate(input=(wav_file, text_file), data_type=("sound", "text"))
print(res)
```
更多详细用法（[示例](examples/industrial_data_pretraining)）
更多详细用法（[示例](https://github.com/alibaba-damo-academy/FunASR/tree/main/examples/industrial_data_pretraining)）


<a name="服务部署"></a>
@@ -251,4 +242,10 @@
  pages={2063--2067},
  doi={10.21437/Interspeech.2022-9996}
}
@article{shi2023seaco,
  author={Xian Shi and Yexin Yang and Zerui Li and Yanni Chen and Zhifu Gao and Shiliang Zhang},
  title={{SeACo-Paraformer: A Non-Autoregressive ASR System with Flexible and Effective Hotword Customization Ability}},
  year=2023,
  journal={arXiv preprint arXiv:2308.03266(accepted by ICASSP2024)},
}
```