Update webui.py

webui.py

@@ -53,6 +53,8 @@ def load_model(path, n_ctx, n_gpu_layers, n_threads, verbose, f16_kv, logits_all
     except:
         return ""
 
+load_model("llama-2-7b-chat.ggmlv3.q2_K.bin", 2048, 0, 32, True, True, False, False, True, False, 512, 64, False, 10000, 1)
+
 def list_models(name):
     return os.listdir(f"{dir}/models")
 
@@ -117,7 +119,7 @@ chatbot = gr.Chatbot(show_label=False, layout="panel", show_copy_button=True, he
 
 with gr.Blocks(theme="theme-repo/STONE_Theme", title="TensorLM", css="style.css") as demo:
     with gr.Row():
-        model = gr.Dropdown(label="Model (only based on Llama in GGML format (.bin))", choices=os.listdir(f"{dir}/models"), value="None", interactive=True, allow_custom_value=False, scale=50)
+        model = gr.Dropdown(label="Model (only based on Llama in GGML format (.bin))", choices=os.listdir(f"{dir}/models"), value="llama-2-7b-chat.ggmlv3.q2_K.bin", interactive=False, allow_custom_value=False, scale=50)
         #refresh_model = gr.Button(value="Load model", interactive=True, scale=1)
     with gr.Row():
         with gr.Tab("💬"):
@@ -217,8 +219,8 @@ with gr.Blocks(theme="theme-repo/STONE_Theme", title="TensorLM", css="style.css"
 
     
                 
-    model.change(load_model, inputs=[model, n_ctx, n_gpu_layers, n_threads, verbose, f16_kv, logits_all, vocab_only, use_mmap, use_mlock, n_batch, last_n_tokens_size, low_vram, rope_freq_base, rope_freq_scale], outputs=model, api_name=False, queue=False)
-    reload_model.click(load_model, inputs=[model, n_ctx, n_gpu_layers, n_threads, verbose, f16_kv, logits_all, vocab_only, use_mmap, use_mlock, n_batch, last_n_tokens_size, low_vram, rope_freq_base, rope_freq_scale], outputs=model, api_name=False, queue=True)
+    #model.change(load_model, inputs=[model, n_ctx, n_gpu_layers, n_threads, verbose, f16_kv, logits_all, vocab_only, use_mmap, use_mlock, n_batch, last_n_tokens_size, low_vram, rope_freq_base, rope_freq_scale], outputs=model, api_name=False, queue=False)
+    #reload_model.click(load_model, inputs=[model, n_ctx, n_gpu_layers, n_threads, verbose, f16_kv, logits_all, vocab_only, use_mmap, use_mlock, n_batch, last_n_tokens_size, low_vram, rope_freq_base, rope_freq_scale], outputs=model, api_name=False, queue=True)