Use ja-v0.3.1 model, rearrange some UI components

app.py

@@ -1,4 +1,3 @@
-import os
 import re
 import unicodedata
 from pathlib import Path
@@ -14,7 +13,7 @@ from transformers import AutoModelForPreTraining, AutoTokenizer
 ALIAS_SEP = "|"
 
 repo_id = "studio-ousia/luxe"
-revision = "ja-v0.3"
+revision = "ja-v0.3.1"
 
 nayose_repo_id = "studio-ousia/luxe-nayose-bm25"
 
@@ -157,7 +156,7 @@ def get_predicted_entity_spans(
 
 
 def get_topk_entities_from_texts(
-    texts: list[str], k: int = 5, entity_span_sensitivity: float = 1.0, nayose_coef: float = 0.0
+    texts: list[str], k: int = 5, entity_span_sensitivity: float = 1.0, nayose_coef: float = 1.0
 ) -> tuple[list[list[tuple[int, int]]], list[list[str]], list[list[str]], list[list[list[str]]]]:
     batch_entity_spans: list[list[tuple[int, int]]] = []
     topk_normal_entities: list[list[str]] = []
@@ -227,12 +226,14 @@ def get_similar_entities(query_entity: str, k: int = 10) -> list[str]:
 
 
 with gr.Blocks() as demo:
-    gr.Markdown("## テキスト（直接入力またはファイルアップロード）")
+    gr.Markdown("# 📝 LUXE Demo")
+
+    gr.Markdown("## 入力テキスト")
 
     texts = gr.State([])
     topk = gr.State(5)
     entity_span_sensitivity = gr.State(1.0)
-    nayose_coef = gr.State(0.0)
+    nayose_coef = gr.State(1.0)
     batch_entity_spans = gr.State([])
     topk_normal_entities = gr.State([])
     topk_category_entities = gr.State([])
@@ -240,21 +241,26 @@ with gr.Blocks() as demo:
     selected_entity = gr.State()
     similar_entities = gr.State([])
 
-    text_input = gr.Textbox(label="Input Text")
+    with gr.Tab(label="直接入力"):
+        text_input = gr.Textbox(label="入力テキスト")
+    with gr.Tab(label="ファイルアップロード"):
+        texts_file = gr.File(label="入力テキストファイル")
+
+    with gr.Accordion(label="ハイパーパラメータ", open=False):
+        topk_input = gr.Number(5, label="エンティティ件数", interactive=True)
+        entity_span_sensitivity_input = gr.Slider(
+            minimum=0.1, maximum=5.0, value=1.0, step=0.1, label="エンティティ検出の積極度", interactive=True
+        )
+        nayose_coef_input = gr.Slider(
+            minimum=0.0, maximum=2.0, value=1.0, step=0.1, label="文字列一致の優先度", interactive=True
+        )
+
     text_input.change(fn=lambda text: [normalize_text(text)], inputs=text_input, outputs=texts)
-    texts_file = gr.File(label="Input Texts")
     texts_file.change(fn=get_texts_from_file, inputs=texts_file, outputs=texts)
-    topk_input = gr.Number(5, label="Top K", interactive=True)
     topk_input.change(fn=lambda val: val, inputs=topk_input, outputs=topk)
-    entity_span_sensitivity_input = gr.Slider(
-        minimum=0.1, maximum=5.0, value=1.0, step=0.1, label="Entity Span Sensitivity", interactive=True
-    )
     entity_span_sensitivity_input.change(
         fn=lambda val: val, inputs=entity_span_sensitivity_input, outputs=entity_span_sensitivity
     )
-    nayose_coef_input = gr.Slider(
-        minimum=0.0, maximum=2.0, value=0.0, step=0.1, label="Nayose Coefficient", interactive=True
-    )
     nayose_coef_input.change(fn=lambda val: val, inputs=nayose_coef_input, outputs=nayose_coef)
 
     texts.change(
@@ -277,7 +283,6 @@ with gr.Blocks() as demo:
         inputs=[texts, topk, entity_span_sensitivity, nayose_coef],
         outputs=[batch_entity_spans, topk_normal_entities, topk_category_entities, topk_span_entities],
     )
-
     topk_input.change(inputs=topk_input, outputs=topk)
 
     gr.Markdown("---")