Update app.py

app.py

@@ -1,17 +1,20 @@
-from llama_cpp import Llama
+from transformers import AutoModelForCausalLM, AutoTokenizer
 import gradio as gr
 
-# Load model
-MODEL_PATH = "./model/vinallama-2.7b-chat.gguf"  # Đường dẫn đến mô hình GGUF
-model = Llama(model_path=MODEL_PATH)
+# Sử dụng tên mô hình từ Hugging Face
+MODEL_NAME = "vilm/vinallama-2.7b-chat-GGUF"
 
-# Define chatbot function
-def chatbot(input_text):
-    response = model(input_text, max_tokens=200, temperature=0.7)
-    return response['choices'][0]['text']
+# Tải mô hình và tokenizer
+tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
+model = AutoModelForCausalLM.from_pretrained(MODEL_NAME, torch_dtype=torch.float16, device_map="auto")
 
-# Create Gradio interface
-interface = gr.Interface(fn=chatbot, inputs="text", outputs="text", title="Vinamallama 7B Chatbot")
+# Hàm chatbot
+def chatbot(input_text):
+    inputs = tokenizer(input_text, return_tensors="pt").to("cuda")
+    outputs = model.generate(inputs["input_ids"], max_length=200, do_sample=True, temperature=0.7)
+    response = tokenizer.decode(outputs[0], skip_special_tokens=True)
+    return response
 
-# Launch app
+# Giao diện Gradio
+interface = gr.Interface(fn=chatbot, inputs="text", outputs="text", title="Vinamallama Chatbot")
 interface.launch()