qwen32

client.py

@@ -0,0 +1,14 @@
+import requests
+import json
+url = 'http://0.0.0.0:8000/ask/'
+
+content = [
+
+            {"role": "system", "content": "You are a helpful assistant."},
+            {"role": "user", "content": "你好"}
+            ]
+response = requests.post(url, json={
+          'content': json.dumps(content),
+            })
+text = response.text
+print(text)

route.py

@@ -0,0 +1,37 @@
+from fastapi import FastAPI
+from pydantic import BaseModel
+app = FastAPI()
+import json
+
+from openai import OpenAI
+qwclient = OpenAI(
+    # 若没有配置环境变量，请用百炼API Key将下行替换为：api_key="sk-xxx",
+    #api_key=os.getenv("DASHSCOPE_API_KEY"),
+    api_key='ms-0edcfe5e-0d1f-4092-ac17-7c6f69127731',
+    #base_url="https://dashscope.aliyuncs.com/compatible-mode/v1",
+    base_url="https://api-inference.modelscope.cn/v1/",
+    timeout=60
+)
+class Task(BaseModel):
+    content: str
+
+
+@app.post("/ask/")
+async def submit(r: Task):
+    content = json.loads(r.content)
+    completion = qwclient.chat.completions.create(
+        # 模型列表：https://help.aliyun.com/zh/model-studio/getting-started/models
+        #model="glm-4.5-flash",
+        model="Qwen/Qwen3-32B",
+        messages=content,
+        
+        # Qwen3模型通过enable_thinking参数控制思考过程（开源版默认True，商业版默认False）
+        # 使用Qwen3开源版模型时，若未启用流式输出，请将下行取消注释，否则会报错
+        extra_body={"enable_thinking": False},
+        #extra_body={"thinking": {"type": "disabled"}},
+    )
+    json_string = completion.choices[0].message.content
+    print(json_string)
+    return {"answer": json_string}
+
+

route2.py

@@ -0,0 +1,37 @@
+from fastapi import FastAPI
+from pydantic import BaseModel
+app = FastAPI()
+import json
+
+from openai import OpenAI
+qwclient = OpenAI(
+    # 若没有配置环境变量，请用百炼API Key将下行替换为：api_key="sk-xxx",
+    #api_key=os.getenv("DASHSCOPE_API_KEY"),
+    api_key='ms-bfe436b3-8900-4456-b702-ead0ce272b10',
+    #base_url="https://dashscope.aliyuncs.com/compatible-mode/v1",
+    base_url="https://api-inference.modelscope.cn/v1/",
+    timeout=60
+)
+class Task(BaseModel):
+    content: str
+
+
+@app.post("/ask/")
+async def submit(r: Task):
+    content = json.loads(r.content)
+    completion = qwclient.chat.completions.create(
+        # 模型列表：https://help.aliyun.com/zh/model-studio/getting-started/models
+        #model="glm-4.5-flash",
+        model="Qwen/Qwen3-32B",
+        messages=content,
+        
+        # Qwen3模型通过enable_thinking参数控制思考过程（开源版默认True，商业版默认False）
+        # 使用Qwen3开源版模型时，若未启用流式输出，请将下行取消注释，否则会报错
+        extra_body={"enable_thinking": False},
+        #extra_body={"thinking": {"type": "disabled"}},
+    )
+    json_string = completion.choices[0].message.content
+    print(json_string)
+    return {"answer": json_string}
+
+