Revert fix5

app.py

@@ -13,7 +13,7 @@ MAX_INPUT_TOKEN_LENGTH = int(os.getenv("MAX_INPUT_TOKEN_LENGTH", "4096"))
 
 DESCRIPTION = """\
 # Nekochu/Luminia-13B-v3
-This Space demonstrates model Nekochu/Luminia-13B-v3 by Nekochu, a Llama 2 model with 13B parameters fine-tuned for SD gen prompt 
+This Space demonstrates model [Nekochu/Luminia-13B-v3](https://huggingface.co/Nekochu/Luminia-13B-v3) by Nekochu, a Llama 2 model with 13B parameters fine-tuned for SD gen prompt 
 """
 
 LICENSE = """
@@ -21,18 +21,14 @@ LICENSE = """
 ---.
 """
 
-def load_model(model_id):
-    model = AutoModelForCausalLM.from_pretrained(model_id, device_map="auto", load_in_4bit=True)
-    tokenizer = AutoTokenizer.from_pretrained(model_id)
-    tokenizer.use_default_system_prompt = False
-    return model, tokenizer
-
 if not torch.cuda.is_available():
     DESCRIPTION += "\n<p>Running on CPU 🥶 This demo does not work on CPU.</p>"
 
+# Define the available models
 MODELS = [
-    "Nekochu/Luminia-13B-v3",
-    "Nekochu/Llama-2-13B-German-ORPO",
+    {"name": "Nekochu/Luminia-13B-v3", "id": "Nekochu/Luminia-13B-v3"},
+    {"name": "Nekochu/Llama-2-13B-German-ORPO", "id": "Nekochu/Llama-2-13B-German-ORPO"},
+    # Add more models here in the future
 ]
 
 @spaces.GPU(duration=120)
@@ -47,7 +43,10 @@ def generate(
     top_k: int = 50,
     repetition_penalty: float = 1.2,
 ) -> Iterator[str]:
-    model, tokenizer = load_model(model_id)
+    # Load the model and tokenizer based on the selected model ID
+    model = AutoModelForCausalLM.from_pretrained(model_id, device_map="auto", load_in_4bit=True)
+    tokenizer = AutoTokenizer.from_pretrained(model_id)
+    tokenizer.use_default_system_prompt = False
 
     conversation = []
     if system_prompt:
@@ -82,10 +81,17 @@ def generate(
         outputs.append(text)
         yield "".join(outputs)
 
+# Add a dropdown for model selection
+model_dropdown = gr.Dropdown(
+    label="Select Model",
+    choices=[model["name"] for model in MODELS],
+    value=MODELS[0]["name"], # Default to the first model
+)
+
 chat_interface = gr.ChatInterface(
     fn=generate,
     additional_inputs=[
-        gr.Dropdown(label="Select Model", choices=MODELS),
+        model_dropdown,
         gr.Textbox(label="System prompt", lines=6),
         gr.Slider(
             label="Max new tokens",