Update webui.py

webui.py

@@ -5,7 +5,6 @@ import time
 if echo == "True":
     from modules.echo import *
 
-import spaces
 
 import os
 import gradio as gr
@@ -60,7 +59,7 @@ with gr.Blocks(theme=theme, title=f"TensorLM v{tlm_version} Demo", css="style.cs
         with gr.Row(render=False) as settings:
             reload_model = gr.Button("Apply settings to model", interactive=True)
             n_ctx = gr.Slider(label="Number of CTX", minimum=1024, maximum=4056, value=2048, step=8, interactive=True)
-            n_gpu_layers = gr.Slider(label="Number of GPU layers", minimum=0, maximum=36, value=4, step=1, interactive=True)
+            n_gpu_layers = gr.Slider(label="Number of GPU layers", minimum=0, maximum=36, value=0, step=1, interactive=False)
             n_threads = gr.Slider(label="Number of Threads", minimum=2, maximum=36, value=8, step=1, interactive=True)
             verbose = gr.Checkbox(label="Verbose", value=True, interactive=True)
             f16_kv = gr.Checkbox(label="F16 KV", value=True, interactive=True)
@@ -70,7 +69,7 @@ with gr.Blocks(theme=theme, title=f"TensorLM v{tlm_version} Demo", css="style.cs
             use_mlock = gr.Checkbox(label="Use mlock", value=False, interactive=True)
             n_batch = gr.Slider(label="Number of batch", minimum=128, maximum=2048, value=512, step=8, interactive=True)
             last_n_tokens_size = gr.Slider(label="Last number of tokens size", minimum=8, maximum=512, value=64, step=8, interactive=True)
-            low_vram = gr.Checkbox(label="Low VRAM", value=lowvram_arg, interactive=True)
+            low_vram = gr.Checkbox(label="Low VRAM", value=lowvram_arg, interactive=False)
             rope_freq_base = gr.Slider(label="Rope freq base", minimum=1000.0, maximum=30000.0, value=10000.0, step=0.1, interactive=True)
             rope_freq_scale = gr.Slider(label="Rope freq scale", minimum=0.1, maximum=3.0, value=1.0, step=0.1)