update app.py

Translations: All non-English text has been translated into English.
Project Name: Changed the title of the project to "Modular Visual Analytics for Prompt Engineering."
Label Changes: Updated various labels, placeholders, and Markdown texts to be in English.

app.py

@@ -22,7 +22,7 @@ def ensure_list(x):
 
 CHECKPOINTS = {
     "CQI FLMv1 Base PT": "impira/layoutlm-document-qa",
-    "CQI FLMv1 para Faturas 💸": "impira/layoutlm-invoices",
+    "CQI FLMv1 for Invoices 💸": "impira/layoutlm-invoices",
     "CQI FDonut": "naver-clova-ix/donut-base-finetuned-docvqa",
     "CQI Assistant Eyes 🤖": "cloudqi/CQI_Visual_Question_Awnser_PT_v0"
 }
@@ -46,8 +46,6 @@ def run_pipeline(model, question, document, top_k):
     return pipeline(question=question, **document.context, top_k=top_k)
 
 
-# TODO: Move into docquery
-# TODO: Support words past the first page (or window?)
 def lift_word_boxes(document, page):
     return document.context["image"][page][1]
 
@@ -61,7 +59,6 @@ def expand_bbox(word_boxes):
     return [min_x, min_y, max_x, max_y]
 
 
-# LayoutLM boxes are normalized to 0, 1000
 def normalize_bbox(box, width, height, padding=0.005):
     min_x, min_y, max_x, max_y = [c / 1000 for c in box]
     if padding != 0:
@@ -75,30 +72,22 @@ def normalize_bbox(box, width, height, padding=0.005):
 examples = [
     [
         "invoice.png",
-        "Qual é o número da fatura?",
+        "What is the invoice number?",
     ],
     [
         "contract.jpeg",
-        "Qual é o valor da compra?",
+        "What is the purchase amount?",
     ],
     [
         "statement.png",
-        "Quais são as vendas líquidas para 2020?",
+        "What are the net sales for 2020?",
     ],
-    #    [
-    #        "docquery.png",
-    #        "How many likes does the space have?",
-    #    ],
-    #    [
-    #        "hacker_news.png",
-    #        "What is the title of post number 5?",
-    #    ],
 ]
 
 question_files = {
-    "Quais são as vendas líquidas para 2020?": "statement.pdf",
-    "Quantas curtidas tem o espaço?": "https://huggingface.co/spaces/impira/docquery",
-    "Qual é o título do post número 5?": "https://news.ycombinator.com",
+    "What are the net sales for 2020?": "statement.pdf",
+    "How many likes does the space have?": "https://huggingface.co/spaces/impira/docquery",
+    "What is the title of post number 5?": "https://news.ycombinator.com",
 }
 
 
@@ -204,17 +193,6 @@ CSS = """
 #short-upload-box .w-full {
     min-height: 10rem !important;
 }
-/* I think something like this can be used to re-shape
- * the table
- */
-/*
-.gr-samples-table tr {
-    display: inline;
-}
-.gr-samples-table .p-2 {
-    width: 100px;
-}
-*/
 #select-a-file {
     width: 100%;
 }
@@ -289,7 +267,7 @@ gradio-app h2, .gradio-app h2 {
 """
 
 with gr.Blocks(css=CSS) as demo:
-    gr.Markdown("## Mecanismo de Consulta de Documentos")
+    gr.Markdown("## Modular Visual Analytics for Prompt Engineering")
 
     document = gr.Variable()
     example_question = gr.Textbox(visible=False)
@@ -298,9 +276,9 @@ with gr.Blocks(css=CSS) as demo:
     with gr.Row(equal_height=True):
         with gr.Column():
             with gr.Row():
-                gr.Markdown("## 1. Selecione um arquivo", elem_id="select-a-file")
+                gr.Markdown("## 1. Select a file", elem_id="select-a-file")
                 img_clear_button = gr.Button(
-                    "Limpar", variant="secondary", elem_id="file-clear", visible=False
+                    "Clear", variant="secondary", elem_id="file-clear", visible=False
                 )
             image = gr.Gallery(visible=False)
             with gr.Row(equal_height=True):
@@ -329,10 +307,10 @@ with gr.Blocks(css=CSS) as demo:
             )
 
         with gr.Column() as col:
-            gr.Markdown("## 2. Faça uma pergunta")
+            gr.Markdown("## 2. Ask a question")
             question = gr.Textbox(
                 label="Question",
-                placeholder="e.g. Qual é o número da fatura?",
+                placeholder="e.g. What is the invoice number?",
                 lines=1,
                 max_lines=1,
             )
@@ -345,11 +323,11 @@ with gr.Blocks(css=CSS) as demo:
             with gr.Row():
                 clear_button = gr.Button("Clear", variant="secondary")
                 submit_button = gr.Button(
-                    "Enviar", variant="primary", elem_id="submit-button"
+                    "Submit", variant="primary", elem_id="submit-button"
                 )
             with gr.Column():
                 output_text = gr.Textbox(
-                    label="Resposta principal", visible=False, elem_id="answer"
+                    label="Main Answer", visible=False, elem_id="answer"
                 )
                 output = gr.JSON(label="Output", visible=False)
 
@@ -418,4 +396,4 @@ with gr.Blocks(css=CSS) as demo:
     )
 
 if __name__ == "__main__":
-    demo.launch(enable_queue=False)
+    demo.launch(enable_queue=False)