Efficiency Increased

app.py

@@ -22,7 +22,8 @@ model = AutoModelForCausalLM.from_pretrained("TinyLlama/TinyLlama-1.1B-Chat-v1.0
 # Using CUDA for an optimal experience
 
 device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
-model = model.to(device)
+model = model.to(device).eval()
+
 
 # Defining a custom stopping criteria class for the model's text generation
 
@@ -51,11 +52,11 @@ def generate_response(user_input, history):
     generate_kwargs = dict(
         **model_inputs,
         streamer=streamer,
-        max_new_tokens=1024,
-        do_sample=True,
-        top_p=0.95,
-        top_k=50,
-        temperature=0.7,
+        max_new_tokens=256,
+        do_sample=False,
+        # top_p=0.95,
+        # top_k=50,
+        # temperature=0.7,
         num_beams=1,
         stopping_criteria=StoppingCriteriaList([stop])
     )
@@ -323,7 +324,7 @@ with gr.Blocks(css=css, fill_width=True, title="LogicLinkV5") as demo:
                     with antd.Flex(vertical=True, gap="small", elem_style=dict(height="100%")):
                         Logo()
                         with antd.Button(color="primary", variant="filled", block=True, elem_classes="new-chat-btn") as new_chat_btn:
-                            ms.Text(get_text("New Chat", "新建对话"))
+                            ms.Text(get_text("New Chat", ","))
                             with ms.Slot("icon"):
                                 antd.Icon("PlusOutlined")
                         with antdx.Conversations(elem_classes="chatbot-conversations-list") as conversations:
@@ -388,4 +389,5 @@ with gr.Blocks(css=css, fill_width=True, title="LogicLinkV5") as demo:
         queue=False
     )
 
-demo.queue().launch(share=True, debug=True)
+demo.queue().launch(share=True, debug=True)
+model.eval()