Create app.py

app.py

@@ -0,0 +1,52 @@
+import streamlit as st
+import pandas as pd
+import matplotlib.pyplot as plt
+from transformers import TFTForConditionalGeneration, TFTTokenizer
+import torch
+
+# Function to load data and model
+@st.cache_data
+def load_data(file):
+    data = pd.read_csv(file)
+    return data
+
+# Function to predict using the model
+def predict_earthquake_positions(data, model, tokenizer):
+    inputs = tokenizer(data.to_dict(orient='list'), return_tensors="pt", padding=True, truncation=True)
+    with torch.no_grad():
+        outputs = model.generate(inputs['input_ids'], num_beams=5, early_stopping=True)
+    return outputs
+
+# Load Hugging Face model and tokenizer
+@st.cache_resource
+def load_model():
+    model = TFTForConditionalGeneration.from_pretrained("huggingface/tft")
+    tokenizer = TFTTokenizer.from_pretrained("huggingface/tft")
+    return model, tokenizer
+
+# Streamlit App
+st.title('Earthquake Detection Prediction App')
+
+uploaded_file = st.file_uploader("Upload your CSV file", type=["csv"])
+
+if uploaded_file is not None:
+    data = load_data(uploaded_file)
+
+    # Display the data
+    st.subheader("Uploaded Data")
+    st.write(data)
+
+    # Load the model and tokenizer
+    model, tokenizer = load_model()
+
+    # Make predictions
+    st.subheader("Predictions")
+    predictions = predict_earthquake_positions(data, model, tokenizer)
+
+    # Plotting the predictions
+    st.subheader("Earthquake Prediction Plot")
+    fig, ax = plt.subplots()
+    ax.plot(data['x'], data['prediction'], label="Predicted Earthquake Position", color='green')
+    ax.axvline(x=predictions, color='red', linestyle='--', label='Predicted Earthquake Start')
+    ax.legend()
+    st.pyplot(fig)