Update app.py

app.py

@@ -9,7 +9,52 @@ from llama_index.llms.llama_utils import (
     completion_to_prompt,
 )
 
+def download(model):
+    repo_id = model_info[model]["repo_id"]
+    filename = model_info[model]["filename"]
+
+    model_path = hf_hub_download(
+        repo_id=repo_id,
+        filename=filename,
+        resume_download=True,
+        cache_dir=MODELS_PATH,
+    )
+
+    return model_path
+
+for model_name in models:
+    download(model_name)
+
+    
 MODELS_PATH = "./models"
+models = ["Llama-2-13B-chat", "Mistral-7B-Instruct-v0.2", "zephyr-7B-beta", "vicuna-7B-v1.5", "CodeLlama-7B","Falcon-7B-Instruct"]
+model_info = {
+    "Llama-2-13B-chat": {
+        "repo_id": "TheBloke/Llama-2-13B-chat-GGUF",
+        "filename": "llama-2-13b-chat.Q4_K_M.gguf",
+    },
+    "Mistral-7B-Instruct-v0.2": {
+        "repo_id": "TheBloke/Mistral-7B-Instruct-v0.2-GGUF",
+        "filename": "mistral-7b-instruct-v0.2.Q4_K_M.gguf",
+    },
+    "zephyr-7B-beta": {
+        "repo_id": "TheBloke/zephyr-7B-beta-GGUF",
+        "filename": "zephyr-7b-beta.Q4_K_M.gguf",
+    },
+    "vicuna-7B-v1.5": {
+        "repo_id": "TheBloke/vicuna-7B-v1.5-GGUF",
+        "filename": "vicuna-7b-v1.5.Q4_K_M.gguf",
+    },
+    "CodeLlama-7B": {
+        "repo_id": "TheBloke/CodeLlama-7B-GGUF",
+        "filename": "codellama-7b.Q4_K_M.gguf",
+    },
+    "Falcon-7B-Instruct": {
+        "repo_id": "TheBloke/Falcon-7B-Instruct-GGML",
+        "filename": "falcon-7b-instruct.ggccv1.q4_1.bin",
+    },
+    
+}
 
 mistral_model_path = hf_hub_download(
     repo_id= "TheBloke/Mistral-7B-Instruct-v0.2-GGUF",
@@ -19,6 +64,7 @@ mistral_model_path = hf_hub_download(
 
 
 
+
 """Step 3 : if you use GPU then make sure ( n_gpu_layers":1) at least 1, you can increase or decrease it based on your GPU performance"""
 
 llm = LlamaCPP(
@@ -109,8 +155,8 @@ def predict(message, history):
 
 with gr.Blocks() as UI:
 
-         models=gr.Dropdown(["CodeLlama-7B","Llama-2-13B-chat","Falcon-7B-Instruct" "Mistral-7B-Instruct-v0.2", "zephyr-7B-beta",
-                       "vicuna-7B-v1.5"],value=["CodeLlama-7B","Llama-2-13B-chat","Falcon-7B-Instruct" "Mistral-7B-Instruct-v0.2", "zephyr-7B-beta",
+         models=gr.Dropdown(["CodeLlama-7B","Llama-2-13B-chat","Falcon-7B-Instruct" ,"Mistral-7B-Instruct-v0.2", "zephyr-7B-beta",
+                       "vicuna-7B-v1.5"],value=["CodeLlama-7B","Llama-2-13B-chat","Falcon-7B-Instruct", "Mistral-7B-Instruct-v0.2", "zephyr-7B-beta",
                        "vicuna-7B-v1.5"], label="please select at least one model", info="default model is Mistral-7B-Instruct-v0.2")
          textInfo = gr.Textbox(value="current model is Mistral-7B-Instruct-v0.2",label="Model Status");
           # Chatbot interface
@@ -126,8 +172,8 @@ with gr.Blocks() as UI:
                             undo_btn="Delete Previous",
                             clear_btn="Clear",
                         )
+ization,inputs=[models],outputs=[textInfo])
 
-         models.change(fn=model_initialization,inputs=[models],outputs=[textInfo])
-
+         models.change(fn=model_initial
 if __name__ == "__main__":
     UI.launch(debug=True) #