Update app.py

app.py

@@ -1,20 +1,22 @@
 import gradio as gr
-from huggingface_hub import InferenceClient
+import torch
+from transformers import AutoModelForCausalLM, AutoTokenizer
 
-"""
-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("HuggingFaceH4/zephyr-7b-beta")
+# Model ID
+model_id = "apu20/Llama-3.2-3B-Instruct_Tele"
 
+# Load quantized model (switch to 8-bit if needed)
+model = AutoModelForCausalLM.from_pretrained(
+    model_id,
+    torch_dtype=torch.float16,  # Use float16 for reduced memory footprint
+    device_map="cpu"  # Force model to run on CPU
+)
+
+# Load tokenizer
+tokenizer = AutoTokenizer.from_pretrained(model_id)
 
-def respond(
-    message,
-    history: list[tuple[str, str]],
-    system_message,
-    max_tokens,
-    temperature,
-    top_p,
-):
+
+def respond(message, history, system_message, max_tokens, temperature, top_p):
     messages = [{"role": "system", "content": system_message}]
 
     for val in history:
@@ -25,24 +27,23 @@ def respond(
 
     messages.append({"role": "user", "content": message})
 
-    response = ""
+    # Tokenize input
+    inputs = tokenizer(message, return_tensors="pt").to("cpu")  # Ensure inputs are on CPU
 
-    for message in client.chat_completion(
-        messages,
-        max_tokens=max_tokens,
-        stream=True,
-        temperature=temperature,
-        top_p=top_p,
-    ):
-        token = message.choices[0].delta.content
+    # Generate response
+    with torch.no_grad():
+        outputs = model.generate(
+            **inputs,
+            max_length=max_tokens,
+            temperature=temperature,
+            top_p=top_p
+        )
 
-        response += token
-        yield response
+    response = tokenizer.decode(outputs[0], skip_special_tokens=True)
+    return response
 
 
-"""
-For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
-"""
+# Gradio Chat Interface
 demo = gr.ChatInterface(
     respond,
     additional_inputs=[
@@ -59,6 +60,5 @@ demo = gr.ChatInterface(
     ],
 )
 
-
 if __name__ == "__main__":
     demo.launch()