PyStreamGPT / app.py
Saboorhsn's picture
Update app.py
18adb0d verified
raw
history blame
1.75 kB
import streamlit as st
from datasets import load_dataset, concatenate_datasets
import json
def load_and_combine_datasets():
python_codes_dataset = load_dataset('flytech/python-codes-25k', split='train')
streamlit_issues_dataset = load_dataset("andfanilo/streamlit-issues", split='train')
streamlit_docs_dataset = load_dataset("sai-lohith/streamlit_docs", split='train')
combined_dataset = concatenate_datasets([python_codes_dataset, streamlit_issues_dataset, streamlit_docs_dataset])
return combined_dataset
def save_combined_dataset_as_jsonl(combined_dataset, file_path):
with open(file_path, 'w', encoding='utf-8') as f:
for example in combined_dataset:
json.dump(example, f, ensure_ascii=False)
f.write('\n')
def main():
st.title("Combined Dataset Viewer")
# Load and combine datasets
combined_dataset = load_and_combine_datasets()
# Display a subset of the combined dataset
st.write("Subset of Combined Dataset:", combined_dataset[:10])
# Add option to save the combined dataset as JSONL
if st.button("Save Combined Dataset (JSONL)"):
file_path = "combined_dataset.jsonl"
save_combined_dataset_as_jsonl(combined_dataset, file_path)
st.write(f"Combined dataset saved as JSONL file: {file_path}")
# Add option to download the JSONL file
if st.button("Download Combined Dataset (JSONL)"):
file_path = "combined_dataset.jsonl"
save_combined_dataset_as_jsonl(combined_dataset, file_path)
st.write("Download the combined dataset as JSONL file:")
st.download_button(label="Download", data=open(file_path, "rb"), file_name="combined_dataset.jsonl")
if __name__ == "__main__":
main()