Add description

app.py

@@ -59,7 +59,10 @@ def generate_diff(text_a: str, text_b: str, method: str):
 my_pipeline = None
 tokenizer = Whitespace()
 
+
 with gr.Blocks() as demo:
+    preamble = (Path(__file__).parent / "preamble.md").read_text()
+    gr.Markdown(preamble)
     with gr.Row():
         text_a = gr.Textbox(label="Text A", value="Chinese shares close higher Friday.", lines=2)
         text_b = gr.Textbox(label="Text B", value="Les actions chinoises clôturent en baisse mercredi.", lines=2)
@@ -77,6 +80,8 @@ with gr.Blocks() as demo:
             inputs=[text_a, text_b, method],
             outputs=[output_a, output_b],
         )
+    description = (Path(__file__).parent / "description.md").read_text()
+    gr.Markdown(description)
 
 
 if my_pipeline is None:

description.md

@@ -0,0 +1,29 @@
+## Small print
+
+<p style="background-color: #fff9f9; border: 1px solid #ff0000; padding: 10px;">
+    Warning: This demo is highly experimental and not ready for production use.
+</p>
+
+This demo is a proof of concept for visualizing the semantic differences between two text documents.
+The input documents may or may not be written in the same language.
+
+In our paper, we evaluate three simple, unsupervised approaches based on BERT-like encoder models.
+This demo implements the approaches `DiffAlign` and `DiffDel` using the model [ZurichNLP/unsup-simcse-xlm-roberta-base](https://huggingface.co/ZurichNLP/unsup-simcse-xlm-roberta-base). See the [XLM-R model](https://huggingface.co/xlm-roberta-base) for a list of supported languages.
+
+The third approach, `DiffMask`, was not included in the demo because it is very slow.
+
+More resources:
+- Paper: https://arxiv.org/abs/2305.13303
+- Code: https://github.com/ZurichNLP/recognizing-semantic-differences
+
+## Citation
+```bibtex
+@article{vamvas-sennrich-2023-rsd,
+      title={Towards Unsupervised Recognition of Semantic Differences in Related Documents},
+      author={Jannis Vamvas and Rico Sennrich},
+      year={2023},
+      eprint={2305.13303},
+      archivePrefix={arXiv},
+      primaryClass={cs.CL}
+}
+```

preamble.md

@@ -0,0 +1,3 @@
+# Diff tool for natural language text
+
+Demo for the EMNLP 2023 paper ["Towards Unsupervised Recognition of Semantic Differences in Related Documents"](https://arxiv.org/abs/2305.13303).

requirements.txt

@@ -1,3 +1,3 @@
 transformers==4.34.0
 --find-links https://download.pytorch.org/whl/cpu
-torch
+torch==2.0.1