EliottZemour commited on
Commit
85054d0
1 Parent(s): e8de709

add readme

Browse files
Files changed (1) hide show
  1. app.py +20 -0
app.py CHANGED
@@ -42,10 +42,30 @@ css = """
42
 
43
 
44
  md_text = """## Generating Yelp reviews with BART-base ⭐⭐⭐"""
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
45
  demo = gr.Blocks(css=css)
46
  with demo:
47
  with gr.Row():
48
  gr.Markdown(md_text, elem_id='ctr')
 
49
 
50
  with gr.Row():
51
  stars = gr.inputs.Slider(minimum=0, maximum=5, step=1, default=0, label="stars")
 
42
 
43
 
44
  md_text = """## Generating Yelp reviews with BART-base ⭐⭐⭐"""
45
+ md_desc = """
46
+ This space demonstrates how synthetic data generation can be performed on natural language columns, as found in the Yelp reviews dataset.
47
+
48
+ | review id | stars | useful | funny | cool | text |
49
+ |:---:|:---:|:---:|:---:|:---:|:---:|
50
+ | 0 | 5 | 1 | 0 | 1 | Wow! Yummy, different, delicious...
51
+
52
+
53
+
54
+
55
+ The model is a fine-tuned version of [facebook/bart-base](https://huggingface.com/facebook/bart-base) on Yelp reviews with the following input-output pairs:
56
+
57
+ - **Input**: "Generate review: stars: 5, useful: 1, funny: 0, cool: 1"
58
+ - **Output**: "Wow! Yummy, different, delicious. Our favorite is the lamb curry and korma. With 10 different kinds of naan!!! Don't let the outside deter you (because we almost changed our minds)...go in and try something new! You'll be glad you did!"
59
+
60
+
61
+ ## Resources
62
+ - The Yelp reviews dataset can be found in json format [here](https://www.yelp.com/dataset)."""
63
+
64
  demo = gr.Blocks(css=css)
65
  with demo:
66
  with gr.Row():
67
  gr.Markdown(md_text, elem_id='ctr')
68
+ gr.Markdown(md_desc, elem_id='ctr')
69
 
70
  with gr.Row():
71
  stars = gr.inputs.Slider(minimum=0, maximum=5, step=1, default=0, label="stars")