update

update

update

.gitignore

@@ -0,0 +1,2 @@
+/__pycache__
+/gradio_cached_examples

app.py

@@ -6,8 +6,7 @@ from typing import List
 import utils
 
 
-DESCRIPTIONS = """# 🐯TIGERScore
-
+DESCRIPTIONS = """
 We present ***TIGERScore***, a **T**rained metric that follows **I**nstruction **G**uidance to perform **E**xplainable, and **R**eference-free evaluation over a wide spectrum of text generation tasks. Different from other automatic evaluation methods that only provide arcane scores, TIGERScore is guided by the natural language instruction to provide error analysis to pinpoint the mistakes in the generated text.
 
 ### [**Website**](https://tiger-ai-lab.github.io/TIGERScore/)  [**Paper**](https://arxiv.org/abs/2310.00752)   [**Code**](https://github.com/TIGER-AI-Lab/TIGERScore)   [**TIGERScore-7B**](https://huggingface.co/TIGER-Lab/TIGERScore-7B-V1.0)   [**TIGERScore-13B**](https://huggingface.co/TIGER-Lab/TIGERScore-13B-V1.0)
@@ -37,15 +36,22 @@ def tigerscore(task, input_context, generation_instruction, hypo_output, max_new
 def get_examples(task, inst_textbox, input_textbox, hypo_output_textbox):
     return gr.Dropdown.update(value=task), inst_textbox, input_textbox, hypo_output_textbox
 
+def clear_all(task, inst_textbox, input_textbox, hypo_output_textbox):
+    return gr.Dropdown.update(value=task), "", "", ""
+
 ## initialize the model
 print("Loading TIGERScore model...")
 utils.load_tigerscore("7b")
 
 with gr.Blocks(theme='gradio/soft') as demo:
-    gr.Markdown(DESCRIPTIONS)
-    gr.Markdown("## TIGERScore Inputs")
+     
+    gr.Markdown("## 🐯 TIGERScore Demo")
+    with gr.Row():
+        gr.Markdown(DESCRIPTIONS)
+        gr.Image("https://jdf-prog.github.io/assets/img/publication_preview/tigerscore_preview.png")
 
-    tasks_dropdown = gr.Dropdown(label="Task", choices=utils.tasks + ["other"], value="translation", show_label=True)
+    gr.Markdown("## TIGERScore Inputs")
+    tasks_dropdown = gr.Dropdown(label="Task", choices=utils.tasks, value="translation", show_label=True, allow_custom_value=True)
     inst_textbox = gr.Textbox(lines=1, label="Instruction", placeholder="Enter instruction here", show_label=True)
     input_textbox = gr.Textbox(lines=4, label="Input Context", placeholder="Enter input context here", show_label=True)
     hypo_output_textbox = gr.Textbox(lines=4, label="Hypothesis Output", placeholder="Enter hypothesis output to be evaluated here", show_label=True)
@@ -86,6 +92,12 @@ with gr.Blocks(theme='gradio/soft') as demo:
         inputs=[tasks_dropdown, input_textbox, inst_textbox, hypo_output_textbox, max_new_tokens, temperature, top_p],
         outputs=evaluation_output_textbox,
     )
+
+    clear_button.click(
+        fn=clear_all,
+        inputs=[tasks_dropdown, inst_textbox, input_textbox, hypo_output_textbox],
+        outputs=[tasks_dropdown, inst_textbox, input_textbox, hypo_output_textbox],
+    )
     
     batch_examples = gr.Examples(
         examples=EXAMPLES,
@@ -97,11 +109,13 @@ with gr.Blocks(theme='gradio/soft') as demo:
     )  
 
     citations = gr.Markdown("""## Citation
+```txt
 @article{jiang2023TIGERScore,
   title={TIGERScore: Towards Building Explainable Metric for All Text Generation Tasks},
   author={Dongfu Jiang, Yishan Li, Ge Zhang, Wenhao Huang, Bill Yuchen Lin, Wenhu Chen},
   journal={arXiv preprint arXiv:2310.00752},
   year={2023}
-}""")
+}
+```""")
 
 demo.queue(max_size=20).launch()