Deploy EchoML directly from Query_Your_Model

Dockerfile

@@ -0,0 +1,14 @@
+FROM python:3.10-slim
+
+WORKDIR /app
+
+# Copy all project files
+COPY . .
+
+# Install dependencies
+RUN pip install --no-cache-dir -r Query_Your_Model/requirements.txt
+
+EXPOSE 7860
+
+# Command to run both FastAPI and Streamlit (same as your app.py)
+CMD ["python", "Query_Your_Model/app.py"]

README.md

@@ -1,3 +1,14 @@
+---
+title: Query Your Model
+emoji: 💬
+colorFrom: indigo
+colorTo: blue
+sdk: streamlit
+sdk_version: "1.39.0"
+app_file: Query_Your_Model/app.py
+pinned: false
+---
+
 # Query Your Model — Retrieval-Augmented Explainability
 
 This project implements an end-to-end framework for interacting with machine learning models through natural-language–style queries. It combines FastAPI (for backend explainability services), Streamlit (for a conversational UI), and explainability methods like SHAP with retrieval-augmented reasoning to make predictions more transparent and interpretable. The system allows you to upload a trained model, input feature values for a new instance, generate explanations using SHAP feature attributions, retrieve similar past reasoning cases for context, and interact with the model via a simple chat-style interface.

app.py

@@ -0,0 +1,16 @@
+import subprocess
+import threading
+import time
+
+def run_fastapi():
+    subprocess.Popen(["uvicorn", "Query_Your_Model.app.api_fastapi:app", "--host", "0.0.0.0", "--port", "8000"])
+
+def run_streamlit():
+    time.sleep(3)
+    subprocess.run([
+        "streamlit", "run", "Query_Your_Model/app/ui_streamlit.py",
+        "--server.port=7860", "--server.address=0.0.0.0"
+    ])
+
+threading.Thread(target=run_fastapi).start()
+run_streamlit()

app/app_chat.py

@@ -14,7 +14,7 @@ st.title("💬 Chat with Your Model")
 # ----------------------------------
 with st.sidebar:
     st.header("Settings")
-    api_url = st.text_input("FastAPI endpoint", value="http://127.0.0.1:8000/explain")
+    api_url = st.text_input("FastAPI endpoint", value="https://<tiffany101>.hf.space/explain")
     model_path = st.text_input("Model path", value="Query_Your_Model/model_data/model.pkl")
     feat_names_str = st.text_input(
         "Feature names (comma-separated)",
@@ -27,7 +27,7 @@ with st.sidebar:
 feat_names = [s.strip() for s in feat_names_str.split(",")]
 
 # OpenAI client (set OPENAI_API_KEY in .streamlit/secrets.toml)
-client = OpenAI(api_key="")
+client = OpenAI(api_key="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")
 
 # ------------------------------------------------
 # Helpers
@@ -271,7 +271,7 @@ def apply_what_if(user_q, feat_names, current_features):
 # ------------------------------------------------
 # Step 1: Enter features & predict
 # ------------------------------------------------
-st.subheader("Step 1 – Enter features to generate a prediction")
+st.subheader("Step 1 – Enter features to generate a prediction in the order sepal length,sepal width,petal length and petal width")
 user_features = st.text_input("Enter feature values (comma-separated)", "")
 predict_btn = st.button("🔍 Predict and Explain")
 

requirements.txt

@@ -8,4 +8,6 @@ python-multipart==0.0.9
 streamlit==1.39.0
 pydantic==2.9.2
 joblib==1.4.2
-matplotlib==3.9.2S
+matplotlib==3.9.2
+requests==2.32.3
+openai==1.40.6

runtime.txt

@@ -0,0 +1 @@
+python-3.10