Update app.py

app.py

@@ -7,12 +7,13 @@ from scipy.stats import norm, t
 from scipy.cluster.hierarchy import linkage, dendrogram, fcluster
 import plotly.figure_factory as ff
 from vllm import LLM, SamplingParams
+from huggingface_hub import InferenceClient
 
-@st.cache_resource
+"""@st.cache_resource
 def load_model():
     return LLM(model="deepseek-ai/DeepSeek-R1", device="cpu", trust_remote_code=True)
 
-llm = load_model()
+llm = load_model()"""
 
 def sorting(df):
     df.index = list(map(float, df.index))
@@ -1021,17 +1022,28 @@ try:
         st.info("This section of the program is under development.")
 
     elif main_option == "AI Chat":
-        prompt_user = st.text_input("Write a prompt.")
 
-        if st.button("Send"):
-            if user_input:
-                sampling_params = SamplingParams(temperature=0.7, max_tokens=100)
-                output = llm.generate([prompt_user], sampling_params)
-                response = output[0].outputs[0].text
-                
-                st.text_area("Answer:", response, height=200)
-            else:
-                st.warning("Enter a prompt:")
+        client = InferenceClient(
+        	provider="together",
+        	api_key="hf_xxxxxxxxxxxxxxxxxxxxxxxx"
+        )
+        
+        messages = [
+        	{
+        		"role": "user",
+        		"content": "What is the capital of France?"
+        	}
+        ]
+        
+        stream = client.chat.completions.create(
+        	model="deepseek-ai/DeepSeek-R1", 
+        	messages=messages, 
+        	max_tokens=500,
+        	stream=True
+        )
+        
+        for chunk in stream:
+            st.warning(chunk.choices[0].delta.content, end="")
 
 except Exception as e:
         st.error(f"❌ Error: {e}")