wwpop commited on
Commit
ab2fe8c
·
verified ·
1 Parent(s): b308b48

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +5 -5
app.py CHANGED
@@ -6,7 +6,7 @@ api_key = os.environ.get('qwen_API_KEY')
6
  """
7
  For more information on huggingface_hub Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
8
  """
9
- client = InferenceClient("Qwen/Qwen2.5-72B-Instruct", token=api_key)
10
 
11
  def respond(
12
  message,
@@ -57,10 +57,10 @@ demo = gr.ChatInterface(
57
  fn=respond,
58
  examples=example_prompts,
59
  cache_examples=False,
60
- title="千问2.5-72B",
61
  description="千问2.5-72B聊天机器人",
62
  additional_inputs=[
63
- gr.Textbox(value="You are Qwen, created by Alibaba Cloud. You are a helpful assistant.", label="System message"),
64
  gr.Slider(minimum=1, maximum=8888, value=2048, step=1, label="Max new tokens"),
65
  gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
66
  gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p (nucleus sampling)"),
@@ -69,5 +69,5 @@ demo = gr.ChatInterface(
69
  )
70
 
71
  if __name__ == "__main__":
72
- demo.queue(default_concurrency_limit=40)
73
- demo.launch(max_threads=40)
 
6
  """
7
  For more information on huggingface_hub Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
8
  """
9
+ client = InferenceClient("deepseek-ai/DeepSeek-V3", token=api_key,provider="together")
10
 
11
  def respond(
12
  message,
 
57
  fn=respond,
58
  examples=example_prompts,
59
  cache_examples=False,
60
+ title="DeepSeek-V3",
61
  description="千问2.5-72B聊天机器人",
62
  additional_inputs=[
63
+ gr.Textbox(value="You are a helpful assistant.", label="System message"),
64
  gr.Slider(minimum=1, maximum=8888, value=2048, step=1, label="Max new tokens"),
65
  gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
66
  gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p (nucleus sampling)"),
 
69
  )
70
 
71
  if __name__ == "__main__":
72
+ demo.queue(default_concurrency_limit=80)
73
+ demo.launch(max_threads=80)