Code Updated

app.py

@@ -1,4 +1,40 @@
 import streamlit as st
+from transformers import AutoModelForSeq2SeqLM, AutoTokenizer
+import pandas as pd
+import numpy as np
+from datasets import load_dataset
+import re
 
-x = st.slider('Select a value')
-st.write(x, 'squared is', x * x)
+# Load the dataset
+ds = load_dataset("Vezora/Open-Critic-GPT")
+
+st.write("Dataset")
+
+# Load the model and tokenizer
+model_name = "shareAI/llama3.1-8b-instruct-dpo-zh"
+model = AutoModelForSeq2SeqLM.from_pretrained(model_name)
+tokenizer = AutoTokenizer.from_pretrained(model_name)
+
+# Function to generate a response from the model
+def generate_response(human_text):
+    inputs = tokenizer.encode(human_text, return_tensors='pt')
+    outputs = model.generate(inputs, max_length=50, num_beams=5, early_stopping=True)
+    response = tokenizer.decode(outputs[0], skip_special_tokens=True)
+    return response
+
+# Iterate over the first few examples in the dataset and display them with model responses
+for i, x in enumerate(ds["train"]):
+    col1, col2, col3 = st.columns(3)
+    if i < 3:
+        with col1:
+            st.code(x["Human"])
+
+        with col2:
+            st.write(x["Assistant"])
+
+        with col3:
+            # Generate and display the model's response
+            response = generate_response(x["Human"])
+            st.write(response)
+    else:
+        break