Create app.py

app.py

@@ -0,0 +1,81 @@
+# =============================
+# app.py for Cass Beta 2 Chat
+# =============================
+
+import gradio as gr
+import torch
+from transformers import AutoModelForCausalLM, AutoTokenizer
+from peft import PeftModel
+
+# =============================
+# MODEL SETTINGS
+# =============================
+BASE_MODEL_ID = "ibm-granite/granite-4.0-micro-base"
+PEFT_MODEL_ID = "DSDUDEd/Cass-Beta2.0"
+
+print("🚀 Loading base model and PEFT adapter...")
+tokenizer = AutoTokenizer.from_pretrained(BASE_MODEL_ID)
+
+# Load base model
+base_model = AutoModelForCausalLM.from_pretrained(BASE_MODEL_ID)
+# Load PEFT model
+model = PeftModel.from_pretrained(base_model, PEFT_MODEL_ID)
+
+# Move to device
+device = "cuda" if torch.cuda.is_available() else "cpu"
+model.to(device)
+model.eval()
+
+print(f"✅ Model loaded on {device.upper()}")
+
+# =============================
+# CHAT HISTORY
+# =============================
+history = []
+
+# =============================
+# GENERATION FUNCTION
+# =============================
+def chat(user_input):
+    """Generates AI response given user input"""
+    global history
+    history.append(("User", user_input))
+
+    # Prepare input
+    inputs = tokenizer(user_input, return_tensors="pt").to(device)
+
+    # Generate output
+    outputs = model.generate(
+        **inputs,
+        max_new_tokens=150,
+        pad_token_id=tokenizer.eos_token_id
+    )
+    ai_response = tokenizer.decode(outputs[0], skip_special_tokens=True)
+
+    history.append(("Cass", ai_response))
+
+    # Format chat
+    chat_text = ""
+    for role, message in history:
+        chat_text += f"{role}: {message}\n\n"
+
+    return chat_text
+
+# =============================
+# GRADIO INTERFACE
+# =============================
+with gr.Blocks() as demo:
+    gr.Markdown("## Chat with Cass Beta 2 🤖")
+    
+    chatbox = gr.Textbox(label="Your message", placeholder="Type your message here...", lines=2)
+    send_button = gr.Button("Send")
+    output = gr.Textbox(label="Chat History", interactive=False, lines=20)
+
+    send_button.click(chat, inputs=chatbox, outputs=output)
+    chatbox.submit(chat, inputs=chatbox, outputs=output)
+
+# =============================
+# RUN APP
+# =============================
+if __name__ == "__main__":
+    demo.launch()