Update app.py

app.py

@@ -4,25 +4,35 @@ import matplotlib.pyplot as plt
 import seaborn as sns
 from sklearn.ensemble import IsolationForest
 from sklearn.preprocessing import StandardScaler
-import openai
-import os
-
-# Set your OpenAI key here or use Hugging Face Secrets Manager
-openai.api_key = os.getenv("OPENAI_API_KEY")
+from transformers import AutoTokenizer, AutoModelForCausalLM, pipeline
+import torch
 
 st.set_page_config(page_title="Smart Factory RAG Assistant", layout="wide")
 
-st.title("🏭 Industry 5.0 | Smart Factory RAG Assistant")
+st.title("🏠 Industry 5.0 | Smart Factory RAG Assistant (Open Source)")
+
+# Load the open-source model (Mistral-7B-Instruct)
+@st.cache_resource(show_spinner=True)
+def load_model():
+    tokenizer = AutoTokenizer.from_pretrained("mistralai/Mistral-7B-Instruct-v0.2")
+    model = AutoModelForCausalLM.from_pretrained(
+        "mistralai/Mistral-7B-Instruct-v0.2",
+        torch_dtype=torch.float16,
+        device_map="auto"
+    )
+    return pipeline("text-generation", model=model, tokenizer=tokenizer)
+
+nlp = load_model()
 
 # File Upload
-uploaded_file = st.file_uploader("📤 Upload your factory CSV data", type=["csv"])
+uploaded_file = st.file_uploader("📄 Upload your factory CSV data", type=["csv"])
 
 if uploaded_file:
     df = pd.read_csv(uploaded_file)
     st.success("✅ File uploaded and loaded!")
 
     # Basic Preview
-    st.subheader("📄 Data Preview")
+    st.subheader("📃 Data Preview")
     st.dataframe(df.head())
 
     # Descriptive Stats
@@ -48,13 +58,13 @@ if uploaded_file:
     st.write(f"Detected {len(anomalies)} anomalies")
     st.dataframe(anomalies.head(10))
 
-    # Prepare context for GPT
+    # Role-based Assistant
     st.subheader("🧠 Role-Based Decision Assistant")
     role = st.selectbox("Select your role", ["Engineer", "Operator"])
     question = st.text_input("Ask a question based on the data analysis")
 
     if question:
-        with st.spinner("Thinking..."):
+        with st.spinner("Generating insights..."):
             summary = df.describe().to_string()
             corr_text = corr.to_string()
             anomaly_count = len(anomalies)
@@ -74,27 +84,21 @@ ANOMALY DETECTION:
 {anomaly_count} anomalies detected using Isolation Forest method.
 
 Based on this context, answer the following question in a clear, technically accurate manner and suggest best decisions from the point of view of a {role}.
-"""
 
-            final_prompt = f"""{context}
 QUESTION: {question}
-ANSWER:"""
-
-            try:
-                response = openai.ChatCompletion.create(
-                    model="gpt-4",
-                    messages=[
-                        {"role": "system", "content": f"You are an expert {role} in a smart factory."},
-                        {"role": "user", "content": final_prompt}
-                    ],
-                    temperature=0.5,
-                    max_tokens=500
-                )
-                answer = response['choices'][0]['message']['content']
-                st.success("✅ Recommendation:")
-                st.markdown(f"**{answer}**")
-            except Exception as e:
-                st.error(f"⚠️ Error calling GPT API: {str(e)}")
+ANSWER:
+"""
+            prompt = f"<s>[INST] {context} [/INST]"
+            output = nlp(prompt, max_new_tokens=512, do_sample=True, temperature=0.5)[0]['generated_text']
+
+            # Clean up response
+            if '[/INST]' in output:
+                answer = output.split('[/INST]')[-1].strip()
+            else:
+                answer = output.strip()
+
+            st.success("✅ Recommendation:")
+            st.markdown(f"**{answer}**")
 
 else:
     st.info("📂 Please upload a factory CSV data file to begin analysis.")