【ASR】fix acs demo

demos/audio_content_search/README.md

@@ -19,7 +19,7 @@ You can choose one way from meduim and hard to install paddlespeech.
 The dependency refers to the requirements.txt, and install the dependency as follows:
 
 ```
-pip install -r requriement.txt 
+pip install -r requirements.txt 
 ```
 
 ### 2. Prepare Input File
@@ -30,11 +30,20 @@ Here are sample files for this demo that can be downloaded:
 wget -c https://paddlespeech.bj.bcebos.com/PaddleAudio/zh.wav
 ```
 
-### 3. Usage
+### 3. run paddlespeech_server
+Before using the client, it is necessary to start paddlespeech_servers.
+
+Here are sample server configuration：
+```bash
+bash demos/audio_content_search/run.sh
+```
+The logs of the two services will be recorded in 'acs.log' and 'streaming_asr.log' in this configuration.
+
+### 4. Usage
 - Command Line(Recommended)
   ```bash
   # Chinese
-  paddlespeech_client acs --server_ip 127.0.0.1 --port 8090 --input ./zh.wav 
+  paddlespeech_client acs --server_ip 127.0.0.1 --port 8490 --input ./zh.wav 
   ```
 
   Usage:

demos/audio_content_search/README_cn.md

@@ -19,7 +19,7 @@
 依赖参见 requirements.txt, 安装依赖
 
 ```
-pip install -r requriement.txt 
+pip install -r requirements.txt
 ```
 
 ### 2. 准备输入
@@ -29,16 +29,26 @@ pip install -r requriement.txt
 ```bash
 wget -c https://paddlespeech.bj.bcebos.com/PaddleAudio/zh.wav
 ```
-### 3. 使用方法
+
+### 3. 启动 server
+使用 client 之前需要先启动 paddlespeech_server。
+
+可以使用默认 server 配置：
+```bash
+bash demos/audio_content_search/run.sh
+```
+该配置下两个服务的日志会被记录在 `acs.log` 和 `streaming_asr.log` 中。
+
+### 4. 使用方法
 - 命令行 (推荐使用)
   ```bash
   # 中文
-  paddlespeech_client acs --server_ip 127.0.0.1 --port 8090 --input ./zh.wav 
+  paddlespeech_client acs --server_ip 127.0.0.1 --port 8490 --input ./zh.wav 
   ```
 
   使用方法：
   ```bash
-  paddlespeech acs --help
+  paddlespeech asr --help
   ```
   参数：
   - `input`(必须输入)：用于识别的音频文件。

demos/audio_content_search/conf/ws_conformer_application.yaml

@@ -26,8 +26,10 @@ asr_online:
     sample_rate: 16000
     cfg_path: 
     decode_method: 'attention_rescoring' 
+    num_decoding_left_chunks: -1  # number of left chunks for decoding. Defaults to -1.
     force_yes: True
     device: 'cpu' # cpu or gpu:id
+    continuous_decoding: False # disable continue decoding when endpoint detected
     am_predictor_conf:
         device:  # set 'gpu:id' or 'cpu'
         switch_ir_optim: True
@@ -40,4 +42,4 @@ asr_online:
         window_ms: 25   # ms
         shift_ms: 10    # ms
         sample_rate: 16000
-        sample_width: 2
+        sample_width: 2

demos/audio_content_search/conf/ws_conformer_wenetspeech_application.yaml

@@ -31,6 +31,7 @@ asr_online:
     force_yes: True
     device: 'cpu' # cpu or gpu:id
     decode_method: "attention_rescoring"
+    num_decoding_left_chunks: -1  # number of left chunks for decoding. Defaults to -1.
     am_predictor_conf:
         device:  # set 'gpu:id' or 'cpu'
         switch_ir_optim: True