From 4b9146816de5dd0bc997005dc8db9ca822741ae9 Mon Sep 17 00:00:00 2001
From: yumoqing <yumoqing@gmail.com>
Date: Wed, 13 Nov 2024 13:21:08 +0800
Subject: [PATCH] bugfix

---
 README.md              | 33 +++++++++++++++++++++++++++++
 app/README.md          |  0
 app/moonshine.py       | 38 +++++++++++++++++++++++++++++++++
 app/test.py            | 32 ++++++++++++++++++++++++++++
 conf/README.md         |  0
 conf/config.json       | 48 ++++++++++++++++++++++++++++++++++++++++++
 requirements.txt       |  4 ++++
 script/install.sh      |  3 +++
 wwwroot/README.md      |  0
 wwwroot/api/index.dspy |  2 ++
 10 files changed, 160 insertions(+)
 create mode 100644 app/README.md
 create mode 100644 app/moonshine.py
 create mode 100644 app/test.py
 create mode 100644 conf/README.md
 create mode 100755 conf/config.json
 create mode 100644 requirements.txt
 create mode 100755 script/install.sh
 create mode 100644 wwwroot/README.md
 create mode 100644 wwwroot/api/index.dspy

diff --git a/README.md b/README.md
index 0a2d702..f00b919 100644
--- a/README.md
+++ b/README.md
@@ -1,2 +1,35 @@
 # moonshineASR
 
+## dependents
+
+[requirements.txt](requirements.txt) list all dependents modules
+use the following command to install the dependents modules
+```
+mspip install -r requirements.txt
+```
+
+## install moonshine service
+
+```
+cd script
+sudo ./install.sh
+```
+
+## make a virtual python environment
+
+```
+python3 -m venv ~/ve/ms
+
+cat > ~/bin/mspip <<EOF
+#!/usr/bin/bash
+~/ve/ms/bin/python $*
+EOF
+
+cat > ~/bin/mspy <<EOF
+#!/usr/bin/bash
+~/ve/ms/bin/python $*
+EOF
+
+chmod +x ~/bin/ms*
+```
+
diff --git a/app/README.md b/app/README.md
new file mode 100644
index 0000000..e69de29
diff --git a/app/moonshine.py b/app/moonshine.py
new file mode 100644
index 0000000..93bfe2d
--- /dev/null
+++ b/app/moonshine.py
@@ -0,0 +1,38 @@
+import os
+os.environ['KERAS_BACKEND'] = 'torch'
+from transformers import AutoModelForSpeechSeq2Seq, AutoConfig, PreTrainedTokenizerFast
+import torchaudio
+import sys
+import time
+import torch
+from appPublic.worker import awaitify
+from appPublic.jsonConfig import getConfig
+from ahserver.serverenv import ServerEnv
+from ahserver.webapp import webapp
+
+class Moonshine:
+	def __init__(self, modelname):
+		# default modelname 'usefulsensors/moonshine-tiny'
+		if modelname is None:
+			modelname = 'usefulsensors/moonshine-tiny'
+		self.model = AutoModelForSpeechSeq2Seq.from_pretrained(modelname, 
+								trust_remote_code=True)
+		self.tokenizer = PreTrainedTokenizerFast.from_pretrained(modelname)
+
+		print(tokenizer.decode(tokens[0], skip_special_tokens=True))
+	def inference(self, audiofile):
+		audio, sr = torchaudio.load(audiofile)
+		if sr != 16000:
+			audio = torchaudio.functional.resample(audio, sr, 16000)
+		tokens = self.model(audio)
+		return tokenizer.decode(tokens[0], skip_special_tokens=True)
+
+def main():
+	config = getConfig()
+	modelname = config.modelname
+	m = Moonshine(modelname)
+	g = ServerEnv()
+	g.inference = awaitify(m.inference)
+
+if __name__ == '__main__':
+	webapp(main)
diff --git a/app/test.py b/app/test.py
new file mode 100644
index 0000000..e0ff008
--- /dev/null
+++ b/app/test.py
@@ -0,0 +1,32 @@
+import time
+import requests
+import base64
+
+def file2b64(file_path):
+    # 读取文件内容
+    with open(file_path, 'rb') as file:
+        file_content = file.read()
+    
+    # 将文件内容转换为Base64编码
+    base64_encoded_data = base64.b64encode(file_content)
+    
+    # 将Base64编码的数据转换为字符串
+    base64_encoded_str = base64_encoded_data.decode('utf-8')
+    
+    return base64_encoded_str
+
+while True:
+	print('prompt:')
+	p = input()
+	print('input image path:')
+	i = input()
+	if p == '' or i == '':
+		continue
+	t1 = time.time()
+	ret = requests.post('http://pd4e.com:10090/api',
+					data={
+						'prompt':p,
+						'image':file2b64(i)
+					})
+	t2 = time.time()
+	print(ret.text, t2 - t1, 'seconds')
diff --git a/conf/README.md b/conf/README.md
new file mode 100644
index 0000000..e69de29
diff --git a/conf/config.json b/conf/config.json
new file mode 100755
index 0000000..21434af
--- /dev/null
+++ b/conf/config.json
@@ -0,0 +1,48 @@
+{
+	"password_key":"!@#$%^&*(*&^%$QWERTYUIqwertyui234567",
+	"modelname":"Qwen/Qwen2-VL-2B-Instruct",
+	"modelname":"Qwen/Qwen2-VL-7B-Instruct",
+	"logger":{
+		"name":"qwenvl",
+		"levelname":"info",
+		"logfile":"$[workdir]$/logs/sage.log"
+	},
+	"filesroot":"$[workdir]$/files",
+	"website":{
+		"paths":[
+			["$[workdir]$/wwwroot",""]
+		],
+		"client_max_size":10000,
+		"host":"0.0.0.0",
+		"port":10090,
+		"coding":"utf-8",
+		"indexes":[
+			"index.html",
+			"index.tmpl",
+			"index.ui",
+			"index.dspy",
+			"index.md"
+		],
+		"startswiths":[
+			{
+				"leading":"/idfile",
+				"registerfunction":"idFileDownload"
+			}
+		],
+		"processors":[
+			[".dspy","dspy"],
+			[".md","md"]
+		],
+		"session_max_time":3000,
+		"session_issue_time":2500,
+		"session_redis_notuse":{
+			"url":"redis://127.0.0.1:6379"
+		}
+	},
+	"langMapping":{
+		"zh-Hans-CN":"zh-cn",
+		"zh-CN":"zh-cn",
+		"en-us":"en",
+		"en-US":"en"
+	}
+}
diff --git a/requirements.txt b/requirements.txt
new file mode 100644
index 0000000..217108a
--- /dev/null
+++ b/requirements.txt
@@ -0,0 +1,4 @@
+useful-moonshine@git+https://github.com/usefulsensors/moonshine.git
+git+https://git.kaiyuancloud.cn/yumoqing/apppublic
+git+https://git.kaiyuancloud.cn/yumoqing/ahserver
+
diff --git a/script/install.sh b/script/install.sh
new file mode 100755
index 0000000..1d60f54
--- /dev/null
+++ b/script/install.sh
@@ -0,0 +1,3 @@
+sudo cp moonshine.service /etc/systemd/system
+sudo systemctl enable moonshine.service
+sudo systemctl start moonshine
diff --git a/wwwroot/README.md b/wwwroot/README.md
new file mode 100644
index 0000000..e69de29
diff --git a/wwwroot/api/index.dspy b/wwwroot/api/index.dspy
new file mode 100644
index 0000000..164bf0a
--- /dev/null
+++ b/wwwroot/api/index.dspy
@@ -0,0 +1,2 @@
+return await inference(params_kw.audiofile)
+