Update app.py

app.py

@@ -1,28 +1,15 @@
-# Use a pipeline as a high-level helper
-from transformers import pipeline
-
-pipe = pipeline("text-generation", model="tiiuae/falcon-7b-instruct")
-
-
-# Function to interact with the chatbot
-def chat_with_bot():
-    while True:
-        # Take user input
-        question = input("You: ")
-
-        # Exit if user inputs 'exit'
-        if question.lower() == "exit":
-            print("Chatbot: Goodbye!")
-            break
-
-        # Generate answer using the model
-        answer = qa_pipeline(question=question, context="")
-
-        # Print the answer
-        print("Chatbot:", answer['answer'])
-
-
-# Start chatting
-if __name__ == "__main__":
-    print("Welcome to the chatbot! Type 'exit' to end the conversation.")
-    chat_with_bot()
+from transformers import AutoModelForCausalLM, AutoTokenizer
+
+checkpoint = "HuggingFaceH4/zephyr-7b-beta"
+tokenizer = AutoTokenizer.from_pretrained(checkpoint)
+model = AutoModelForCausalLM.from_pretrained(checkpoint)  # You may want to use bfloat16 and/or move to GPU here
+
+messages = [
+    {
+        "role": "system",
+        "content": "You are a friendly chatbot who always responds in the style of a pirate",
+    },
+    {"role": "user", "content": "How many helicopters can a human eat in one sitting?"},
+ ]
+tokenized_chat = tokenizer.apply_chat_template(messages, tokenize=True, add_generation_prompt=True, return_tensors="pt")
+print(tokenizer.decode(tokenized_chat[0]))