Spaces:
Sleeping
Sleeping
Create app.py
Browse files
app.py
ADDED
|
@@ -0,0 +1,36 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import gradio as gr
|
| 2 |
+
import json
|
| 3 |
+
import PyPDF2
|
| 4 |
+
import markdown
|
| 5 |
+
|
| 6 |
+
def convert_to_jsonl(file):
|
| 7 |
+
content = ""
|
| 8 |
+
if file.name.endswith('.txt'):
|
| 9 |
+
content = file.read().decode('utf-8')
|
| 10 |
+
elif file.name.endswith('.pdf'):
|
| 11 |
+
pdf_reader = PyPDF2.PdfReader(file)
|
| 12 |
+
for page in pdf_reader.pages:
|
| 13 |
+
content += page.extract_text()
|
| 14 |
+
elif file.name.endswith('.md'):
|
| 15 |
+
content = file.read().decode('utf-8')
|
| 16 |
+
content = markdown.markdown(content)
|
| 17 |
+
|
| 18 |
+
lines = content.split('\n')
|
| 19 |
+
jsonl_output = ""
|
| 20 |
+
for i, line in enumerate(lines):
|
| 21 |
+
if line.strip():
|
| 22 |
+
json_obj = {"id": i, "text": line.strip()}
|
| 23 |
+
jsonl_output += json.dumps(json_obj) + "\n"
|
| 24 |
+
|
| 25 |
+
return jsonl_output
|
| 26 |
+
|
| 27 |
+
iface = gr.Interface(
|
| 28 |
+
fn=convert_to_jsonl,
|
| 29 |
+
inputs=gr.File(label="Upload .txt, .pdf, or .md file"),
|
| 30 |
+
outputs=gr.File(label="Download JSONL file"),
|
| 31 |
+
title="Raw Text to Indexed JSONL Converter",
|
| 32 |
+
description="Upload a .txt, .pdf, or .md file to convert it to an indexed JSONL file for LLM training."
|
| 33 |
+
)
|
| 34 |
+
|
| 35 |
+
if __name__ == "__main__":
|
| 36 |
+
iface.launch()
|