Create app.py

app.py

@@ -0,0 +1,86 @@
+from transformers import pipeline
+from tqdm import tqdm
+import pandas as pd
+import streamlit as st
+from io import StringIO
+
+
+def correct_text(uploaded_file, column_to_correct):
+    """
+    Corrects text in the specified column using a text correction model.
+
+    Args:
+        uploaded_file: DataFrame containing the text to correct
+        column_to_correct: Index of the column to correct
+    Returns:
+        DataFrame with corrected text in a new column
+    """
+    corrector = pipeline("text2text-generation",
+                         model="sdadas/byt5-text-correction")
+
+    df = uploaded_file
+    progress_bar = st.progress(0)
+    status_text = st.text("Correcting text 🧠...")
+
+    for index, row in df.iterrows():
+        if pd.notna(row.iloc[column_to_correct]):
+            original_text = str(row.iloc[column_to_correct])
+            corrected_text = corrector(
+                "<es>" + original_text, max_length=1024)[0]['generated_text']
+
+            # Save corrected text only if different from original
+            if corrected_text != original_text:
+                df.loc[index, column_to_correct + 1] = corrected_text
+
+        progress = (index + 1) / len(df)
+        progress_bar.progress(progress)
+        status_text.text(f"Progress: {int(progress * 100)}% completed ")
+
+    return df
+
+
+def choose_columns(dataframe):
+    """
+    Lets user select columns to correct and displays preview of data.
+
+    Args:
+        dataframe: Input DataFrame
+    Returns:
+        Selected column index or None if no selection
+    """
+    st.write("Choose the columns to correct 🔍")
+    column_to_correct = st.selectbox(
+        "Select columns to correct", dataframe.columns)
+
+    if column_to_correct:
+        st.write("Preview of data in selected columns 👀:")
+        non_empty_data = dataframe[dataframe[column_to_correct].notna()]
+        st.dataframe(non_empty_data[column_to_correct].head())
+
+    if st.button("Correct Text"):
+        if column_to_correct is not None:
+            return dataframe.columns.get_loc(column_to_correct)
+        else:
+            st.error("Please select a column before correcting text ❌")
+            return None
+
+
+def main():
+    """Main function to run the text correction application"""
+    st.title("CSV text Correction App ✔")
+    uploaded_file = st.file_uploader("Choose a CSV file 📄", type=["csv"])
+    if uploaded_file is not None:
+        try:
+            dataframe = pd.read_csv(uploaded_file, encoding='utf-8')
+            column_index = choose_columns(dataframe)
+            if column_index is not None:
+                st.write(correct_text(dataframe, column_index))
+        except UnicodeDecodeError:
+            st.error(
+                "Error: Unable to decode the file. Please check the file encoding or try another file.")
+        except Exception as e:
+            st.error(f"An unexpected error occurred: {e}")
+
+
+if __name__ == "__main__":
+    main()