Spaces:
Sleeping
Sleeping
Upload folder using huggingface_hub
Browse files
app.py
CHANGED
|
@@ -1,32 +1,24 @@
|
|
| 1 |
# ============================================================
|
| 2 |
# Imports
|
| 3 |
# ============================================================
|
| 4 |
-
import pandas as pd
|
| 5 |
import streamlit as st
|
|
|
|
| 6 |
import joblib
|
| 7 |
from huggingface_hub import hf_hub_download
|
| 8 |
|
|
|
|
| 9 |
# ============================================================
|
| 10 |
# Model Loading
|
| 11 |
# ============================================================
|
| 12 |
@st.cache_resource
|
| 13 |
def load_model():
|
| 14 |
-
"""
|
| 15 |
-
Download and load trained model from Hugging Face Hub.
|
| 16 |
-
|
| 17 |
-
Caching ensures the model is loaded only once
|
| 18 |
-
and reused across user interactions.
|
| 19 |
-
"""
|
| 20 |
model_path = hf_hub_download(
|
| 21 |
repo_id="praveenchugh/engine-condition-gbm-model",
|
| 22 |
filename="gbm_model.joblib",
|
| 23 |
)
|
| 24 |
-
|
| 25 |
-
# Load model from downloaded file
|
| 26 |
return joblib.load(model_path)
|
| 27 |
|
| 28 |
|
| 29 |
-
# Initialize model
|
| 30 |
model = load_model()
|
| 31 |
|
| 32 |
|
|
@@ -36,21 +28,15 @@ model = load_model()
|
|
| 36 |
st.title("Engine Condition Predictor")
|
| 37 |
|
| 38 |
st.write(
|
| 39 |
-
"
|
| 40 |
-
"
|
| 41 |
)
|
| 42 |
|
| 43 |
|
| 44 |
# ============================================================
|
| 45 |
-
# Input Collection
|
| 46 |
# ============================================================
|
| 47 |
def get_user_inputs():
|
| 48 |
-
"""
|
| 49 |
-
Collect user inputs from Streamlit UI and
|
| 50 |
-
return them as a single-row pandas DataFrame.
|
| 51 |
-
"""
|
| 52 |
-
|
| 53 |
-
# Engine operational inputs
|
| 54 |
engine_rpm = st.number_input("Engine RPM", value=800.0)
|
| 55 |
lub_oil_pressure = st.number_input("Lub Oil Pressure", value=3.0)
|
| 56 |
fuel_pressure = st.number_input("Fuel Pressure", value=2.0)
|
|
@@ -58,17 +44,15 @@ def get_user_inputs():
|
|
| 58 |
lub_oil_temp = st.number_input("Lub Oil Temp", value=75.0)
|
| 59 |
coolant_temp = st.number_input("Coolant Temp", value=85.0)
|
| 60 |
|
| 61 |
-
#
|
| 62 |
-
input_df = pd.DataFrame(
|
| 63 |
-
|
| 64 |
-
|
| 65 |
-
|
| 66 |
-
|
| 67 |
-
|
| 68 |
-
|
| 69 |
-
|
| 70 |
-
}]
|
| 71 |
-
)
|
| 72 |
|
| 73 |
return input_df
|
| 74 |
|
|
@@ -76,19 +60,29 @@ def get_user_inputs():
|
|
| 76 |
# ============================================================
|
| 77 |
# Prediction Logic
|
| 78 |
# ============================================================
|
| 79 |
-
# Capture user inputs
|
| 80 |
input_df = get_user_inputs()
|
| 81 |
|
| 82 |
-
# Run prediction when button is clicked
|
| 83 |
if st.button("Predict"):
|
| 84 |
|
| 85 |
-
|
| 86 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 87 |
|
| 88 |
-
|
| 89 |
-
|
| 90 |
|
| 91 |
-
|
| 92 |
-
|
| 93 |
-
else:
|
| 94 |
-
st.success("Engine Condition: Normal")
|
|
|
|
| 1 |
# ============================================================
|
| 2 |
# Imports
|
| 3 |
# ============================================================
|
|
|
|
| 4 |
import streamlit as st
|
| 5 |
+
import pandas as pd
|
| 6 |
import joblib
|
| 7 |
from huggingface_hub import hf_hub_download
|
| 8 |
|
| 9 |
+
|
| 10 |
# ============================================================
|
| 11 |
# Model Loading
|
| 12 |
# ============================================================
|
| 13 |
@st.cache_resource
|
| 14 |
def load_model():
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 15 |
model_path = hf_hub_download(
|
| 16 |
repo_id="praveenchugh/engine-condition-gbm-model",
|
| 17 |
filename="gbm_model.joblib",
|
| 18 |
)
|
|
|
|
|
|
|
| 19 |
return joblib.load(model_path)
|
| 20 |
|
| 21 |
|
|
|
|
| 22 |
model = load_model()
|
| 23 |
|
| 24 |
|
|
|
|
| 28 |
st.title("Engine Condition Predictor")
|
| 29 |
|
| 30 |
st.write(
|
| 31 |
+
"Provide engine sensor values to predict whether the engine condition "
|
| 32 |
+
"is normal or anomalous."
|
| 33 |
)
|
| 34 |
|
| 35 |
|
| 36 |
# ============================================================
|
| 37 |
+
# Input Collection
|
| 38 |
# ============================================================
|
| 39 |
def get_user_inputs():
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 40 |
engine_rpm = st.number_input("Engine RPM", value=800.0)
|
| 41 |
lub_oil_pressure = st.number_input("Lub Oil Pressure", value=3.0)
|
| 42 |
fuel_pressure = st.number_input("Fuel Pressure", value=2.0)
|
|
|
|
| 44 |
lub_oil_temp = st.number_input("Lub Oil Temp", value=75.0)
|
| 45 |
coolant_temp = st.number_input("Coolant Temp", value=85.0)
|
| 46 |
|
| 47 |
+
# IMPORTANT: Match EXACT training column names
|
| 48 |
+
input_df = pd.DataFrame([{
|
| 49 |
+
"Engine rpm": engine_rpm,
|
| 50 |
+
"Lub oil pressure": lub_oil_pressure,
|
| 51 |
+
"Fuel pressure": fuel_pressure,
|
| 52 |
+
"Coolant pressure": coolant_pressure,
|
| 53 |
+
"lub oil temp": lub_oil_temp,
|
| 54 |
+
"Coolant temp": coolant_temp
|
| 55 |
+
}])
|
|
|
|
|
|
|
| 56 |
|
| 57 |
return input_df
|
| 58 |
|
|
|
|
| 60 |
# ============================================================
|
| 61 |
# Prediction Logic
|
| 62 |
# ============================================================
|
|
|
|
| 63 |
input_df = get_user_inputs()
|
| 64 |
|
|
|
|
| 65 |
if st.button("Predict"):
|
| 66 |
|
| 67 |
+
try:
|
| 68 |
+
# Ensure column order matches training
|
| 69 |
+
if hasattr(model, "feature_names_in_"):
|
| 70 |
+
input_df = input_df[model.feature_names_in_]
|
| 71 |
+
|
| 72 |
+
prediction = model.predict(input_df)[0]
|
| 73 |
+
|
| 74 |
+
st.subheader("Prediction Result")
|
| 75 |
+
|
| 76 |
+
if prediction == 1:
|
| 77 |
+
st.error("Engine Condition: Anomalous")
|
| 78 |
+
else:
|
| 79 |
+
st.success("Engine Condition: Normal")
|
| 80 |
+
|
| 81 |
+
except Exception as e:
|
| 82 |
+
st.error(f"Prediction failed: {str(e)}")
|
| 83 |
|
| 84 |
+
# Debug info (very useful)
|
| 85 |
+
st.write("Input columns:", input_df.columns.tolist())
|
| 86 |
|
| 87 |
+
if hasattr(model, "feature_names_in_"):
|
| 88 |
+
st.write("Model expects:", list(model.feature_names_in_))
|
|
|
|
|
|