Update app.py

app.py

@@ -2,17 +2,23 @@ import streamlit as st
 from transformers import AutoTokenizer, AutoModelForCausalLM
 import os
 hf_token = os.getenv('HF_API_TOKEN')
-# Load the Llama 3.1 model and tokenizer
-model_name = "meta-llama/Meta-Llama-3.1-8B"
-tokenizer = AutoTokenizer.from_pretrained(model_name, token= hf_token)
-model = AutoModelForCausalLM.from_pretrained(model_name, torch_dtype="auto", device_map="auto",token = hf_token)
+import streamlit as st
+
+from transformers import pipeline
+
+# Load the model
+generator = pipeline("text-generation", model="meta-llama/Meta-Llama-3.1-8B")
+
+# Create an API route in Streamlit
+@st.cache_resource
+def predict(inputs):
+    return generator(inputs, max_length=512, top_p=0.9, temperature=0.8)[0]['generated_text']
+
+@st.cache_resource
+def predict_endpoint():
+    inputs = st.experimental_get_query_params().get('inputs', [''])[0]
+    return predict(inputs)
 
-# Streamlit app interface
-st.title("Llama 3.1 Text Generator")
-prompt = st.text_area("Enter a prompt:", "Once upon a time")
+st.experimental_set_query_params(result=predict_endpoint())
 
-if st.button("Generate"):
-    inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
-    outputs = model.generate(**inputs, max_length=512, top_p=0.9, temperature=0.8)
-    generated_text = tokenizer.decode(outputs[0], skip_special_tokens=True)
-    st.write(generated_text)
+st.title("Llama3.1 API is Running")