Update app.py

app.py

@@ -1,12 +1,25 @@
 import gradio as gr
-from huggingface_hub import InferenceClient
+import torch
+from transformers import AutoModelForCausalLM, AutoTokenizer, GenerationConfig
+from peft import PeftModel  # For loading adapter files
 
-"""
-For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
-"""
-client = InferenceClient("Futuresony/future_ai_12_10_2024.gguf")
+# Path to the base model and adapter
+BASE_MODEL_PATH = "unsloth/Llama-3.2-3B-Instruct"  # Replace with your base model path
+ADAPTER_PATH = "Futuresony/future_ai_12_10_2024.gguf/adapter"  # Your Hugging Face repo
 
+# Load base model and tokenizer
+print("Loading base model and tokenizer...")
+tokenizer = AutoTokenizer.from_pretrained(BASE_MODEL_PATH)
+model = AutoModelForCausalLM.from_pretrained(BASE_MODEL_PATH, torch_dtype=torch.float16, device_map="auto")
 
+# Load adapter files using PEFT
+print("Loading adapter...")
+model = PeftModel.from_pretrained(model, ADAPTER_PATH)
+
+# Set model to evaluation mode
+model.eval()
+
+# Generate responses using the model
 def respond(
     message,
     history: list[tuple[str, str]],
@@ -15,8 +28,8 @@ def respond(
     temperature,
     top_p,
 ):
+    # Format chat messages
     messages = [{"role": "system", "content": system_message}]
-
     for val in history:
         if val[0]:
             messages.append({"role": "user", "content": val[0]})
@@ -25,40 +38,36 @@ def respond(
 
     messages.append({"role": "user", "content": message})
 
-    response = ""
+    # Concatenate messages as input text
+    input_text = "\n".join([f"{msg['role']}: {msg['content']}" for msg in messages])
 
-    for message in client.chat_completion(
-        messages,
-        max_tokens=max_tokens,
-        stream=True,
+    # Tokenize input text
+    inputs = tokenizer(input_text, return_tensors="pt").to(model.device)
+
+    # Generate response
+    generation_config = GenerationConfig(
+        max_new_tokens=max_tokens,
         temperature=temperature,
         top_p=top_p,
-    ):
-        token = message.choices[0].delta.content
+        do_sample=True,
+    )
+
+    output_ids = model.generate(**inputs, generation_config=generation_config)
+    response = tokenizer.decode(output_ids[0], skip_special_tokens=True)
 
-        response += token
-        yield response
+    return response.split("assistant:")[-1].strip()  # Extract assistant response
 
 
-"""
-For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
-"""
+# Gradio Interface
 demo = gr.ChatInterface(
     respond,
     additional_inputs=[
         gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
         gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
         gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
-        gr.Slider(
-            minimum=0.1,
-            maximum=1.0,
-            value=0.95,
-            step=0.05,
-            label="Top-p (nucleus sampling)",
-        ),
+        gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p (nucleus sampling)"),
     ],
 )
 
-
 if __name__ == "__main__":
     demo.launch()