bugfix: bugfix of ui arg no_eagle3.

Author: wangzhaode

eagle/application/webui.py

@@ -1,5 +1,5 @@
 import os
-os.environ["CUDA_VISIBLE_DEVICES"] = "2,3"
+# os.environ["CUDA_VISIBLE_DEVICES"] = "2,3"
 import time
 
 import gradio as gr
@@ -280,15 +280,15 @@ def clear(history,session_state):
     load_in_4bit=args.load_in_4bit,
     load_in_8bit=args.load_in_8bit,
     device_map="auto",
-    use_eagle3=args.no_eagle3,
+    use_eagle3=(not args.no_eagle3),
 )
 model.eval()
 warmup(model)
 
 custom_css = """
 #speed textarea {
-    color: red;   
-    font-size: 30px; 
+    color: red;
+    font-size: 30px;
 }"""
 
 with gr.Blocks(css=custom_css) as demo:
@@ -303,7 +303,7 @@ def clear(history,session_state):
             highlight_EaInfer = gr.Checkbox(label="Highlight the tokens generated by EAGLE-3", value=True)
         temperature = gr.Slider(minimum=0.0, maximum=1.0, step=0.01, label="temperature", value=0.5)
         top_p = gr.Slider(minimum=0.0, maximum=1.0, step=0.01, label="top_p", value=0.9)
-    note=gr.Markdown(show_label=False,interactive=False,value='''The Compression Ratio is defined as the number of generated tokens divided by the number of forward passes in the original LLM. If "Highlight the tokens generated by EAGLE-2" is checked, the tokens correctly guessed by EAGLE-2 
+    note=gr.Markdown(show_label=False,interactive=False,value='''The Compression Ratio is defined as the number of generated tokens divided by the number of forward passes in the original LLM. If "Highlight the tokens generated by EAGLE-2" is checked, the tokens correctly guessed by EAGLE-2
     will be displayed in orange. Note: Checking this option may cause special formatting rendering issues in a few cases, especially when generating code''')