Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
|
@@ -1,5 +1,6 @@
|
|
| 1 |
import streamlit as st
|
| 2 |
from datasets import load_dataset, concatenate_datasets
|
|
|
|
| 3 |
|
| 4 |
def load_and_combine_datasets():
|
| 5 |
python_codes_dataset = load_dataset('flytech/python-codes-25k', split='train')
|
|
@@ -12,6 +13,8 @@ def load_and_combine_datasets():
|
|
| 12 |
|
| 13 |
def save_combined_dataset(combined_dataset, dataset_name):
|
| 14 |
combined_dataset.save_to_disk(dataset_name)
|
|
|
|
|
|
|
| 15 |
|
| 16 |
def main():
|
| 17 |
st.title("Combined Dataset Viewer")
|
|
@@ -29,5 +32,15 @@ def main():
|
|
| 29 |
save_combined_dataset(combined_dataset, dataset_name)
|
| 30 |
st.success(f"Combined dataset saved as '{dataset_name}'!")
|
| 31 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 32 |
if __name__ == "__main__":
|
| 33 |
main()
|
|
|
|
| 1 |
import streamlit as st
|
| 2 |
from datasets import load_dataset, concatenate_datasets
|
| 3 |
+
import os
|
| 4 |
|
| 5 |
def load_and_combine_datasets():
|
| 6 |
python_codes_dataset = load_dataset('flytech/python-codes-25k', split='train')
|
|
|
|
| 13 |
|
| 14 |
def save_combined_dataset(combined_dataset, dataset_name):
|
| 15 |
combined_dataset.save_to_disk(dataset_name)
|
| 16 |
+
# Print the location where the dataset is saved
|
| 17 |
+
st.write(f"Dataset saved at: {dataset_name}")
|
| 18 |
|
| 19 |
def main():
|
| 20 |
st.title("Combined Dataset Viewer")
|
|
|
|
| 32 |
save_combined_dataset(combined_dataset, dataset_name)
|
| 33 |
st.success(f"Combined dataset saved as '{dataset_name}'!")
|
| 34 |
|
| 35 |
+
# Add option to download the combined dataset
|
| 36 |
+
if st.button("Download Combined Dataset"):
|
| 37 |
+
dataset_name = st.text_input("Enter a name for the combined dataset:")
|
| 38 |
+
if dataset_name:
|
| 39 |
+
save_combined_dataset(combined_dataset, dataset_name)
|
| 40 |
+
filepath = os.path.join(os.getcwd(), dataset_name)
|
| 41 |
+
filesize = os.path.getsize(filepath) / (1024 * 1024) # Size in MB
|
| 42 |
+
st.write(f"Download the combined dataset: [{dataset_name}]({filepath})")
|
| 43 |
+
st.write(f"File Size: {filesize:.2f} MB")
|
| 44 |
+
|
| 45 |
if __name__ == "__main__":
|
| 46 |
main()
|