hermi612 commited on
Commit
aa6deae
·
verified ·
1 Parent(s): b4329a1

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +21 -4
app.py CHANGED
@@ -1,5 +1,10 @@
1
  import gradio as gr
2
  from huggingface_hub import InferenceClient
 
 
 
 
 
3
 
4
  # 使用 Hugging Face Inference API 调用云端模型(无需本地加载)
5
  # 正确方法1:从环境变量读取
@@ -11,13 +16,22 @@ def medical_chat(user_input, history):
11
  prompt = f"患者:{user_input}\n医生:"
12
 
13
  # 调用云端模型(示例使用微软BioGPT)
14
- response = client.text_generation(
 
 
 
 
15
  prompt=prompt,
16
  model="microsoft/BioGPT-Large",
17
- max_new_tokens=150,
18
- temperature=0.7,
19
- repetition_penalty=1.2
20
  )
 
 
 
 
 
 
 
21
 
22
  # 提取医生回复部分
23
  doctor_response = response.split("医生:")[-1].strip()
@@ -40,3 +54,6 @@ with gr.Blocks(theme=gr.themes.Soft()) as demo:
40
  # 启动应用(Hugging Face Spaces 会自动处理)
41
  demo.launch(debug=False)
42
 
 
 
 
 
1
  import gradio as gr
2
  from huggingface_hub import InferenceClient
3
+ # 在 app.py 顶部添加
4
+ import os
5
+ os.environ["HF_ENDPOINT"] = "https://hf-mirror.com" # 中国大陆加速
6
+
7
+
8
 
9
  # 使用 Hugging Face Inference API 调用云端模型(无需本地加载)
10
  # 正确方法1:从环境变量读取
 
16
  prompt = f"患者:{user_input}\n医生:"
17
 
18
  # 调用云端模型(示例使用微软BioGPT)
19
+ from tenacity import retry, stop_after_attempt, wait_exponential
20
+
21
+ @retry(stop=stop_after_attempt(3), wait=wait_exponential(multiplier=1, min=2, max=10))
22
+ def safe_model_call(prompt):
23
+ return client.text_generation(
24
  prompt=prompt,
25
  model="microsoft/BioGPT-Large",
26
+ max_new_tokens=150
 
 
27
  )
28
+ # response = client.text_generation(
29
+ # prompt=prompt,
30
+ # model="microsoft/BioGPT-Large",
31
+ # max_new_tokens=150,
32
+ # temperature=0.7,
33
+ # repetition_penalty=1.2
34
+ # )
35
 
36
  # 提取医生回复部分
37
  doctor_response = response.split("医生:")[-1].strip()
 
54
  # 启动应用(Hugging Face Spaces 会自动处理)
55
  demo.launch(debug=False)
56
 
57
+
58
+
59
+