Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
|
@@ -2,6 +2,7 @@ import streamlit as st
|
|
| 2 |
from datasets import load_dataset, concatenate_datasets
|
| 3 |
import json
|
| 4 |
import os
|
|
|
|
| 5 |
|
| 6 |
def load_and_combine_datasets():
|
| 7 |
python_codes_dataset = load_dataset('flytech/python-codes-25k', split='train')
|
|
@@ -41,8 +42,8 @@ def main():
|
|
| 41 |
file_path = os.path.join(os.getcwd(), f"{output_dataset_name}.jsonl")
|
| 42 |
save_combined_dataset_as_jsonl(combined_dataset, file_path)
|
| 43 |
st.write("Download the combined dataset as JSONL file:")
|
| 44 |
-
with open(file_path, "
|
| 45 |
-
bytes_data = f.read()
|
| 46 |
b64 = base64.b64encode(bytes_data).decode()
|
| 47 |
href = f'<a href="data:file/jsonl;base64,{b64}" download="{output_dataset_name}.jsonl">Download JSONL File</a>'
|
| 48 |
st.markdown(href, unsafe_allow_html=True)
|
|
|
|
| 2 |
from datasets import load_dataset, concatenate_datasets
|
| 3 |
import json
|
| 4 |
import os
|
| 5 |
+
import base64 # Add base64 for encoding
|
| 6 |
|
| 7 |
def load_and_combine_datasets():
|
| 8 |
python_codes_dataset = load_dataset('flytech/python-codes-25k', split='train')
|
|
|
|
| 42 |
file_path = os.path.join(os.getcwd(), f"{output_dataset_name}.jsonl")
|
| 43 |
save_combined_dataset_as_jsonl(combined_dataset, file_path)
|
| 44 |
st.write("Download the combined dataset as JSONL file:")
|
| 45 |
+
with open(file_path, "r") as f: # Open file in read mode
|
| 46 |
+
bytes_data = f.read().encode() # Encode the data as bytes
|
| 47 |
b64 = base64.b64encode(bytes_data).decode()
|
| 48 |
href = f'<a href="data:file/jsonl;base64,{b64}" download="{output_dataset_name}.jsonl">Download JSONL File</a>'
|
| 49 |
st.markdown(href, unsafe_allow_html=True)
|