Update app.py

app.py

@@ -1,65 +1,73 @@
 import gradio as gr
-from transformers import pipeline
 from langdetect import detect
 
 # ============================
-# Публичные модели
+# Natasha для русского
 # ============================
-MODELS = {
-    "Russian": "blinoff/bert-base-russian-ner",
-    "English": "dbmdz/bert-large-cased-finetuned-conll03-english"
-}
+from natasha import Segmenter, MorphVocab, NewsEmbedding, NewsNERTagger, Doc
+
+segmenter = Segmenter()
+morph_vocab = MorphVocab()
+embedding = NewsEmbedding()
+ner_tagger = NewsNERTagger(embedding)
 
-# Создаем пайплайны NER заранее
-ner_pipelines = {
-    "Russian": pipeline("ner", model=MODELS["Russian"], tokenizer=MODELS["Russian"], aggregation_strategy="simple"),
-    "English": pipeline("ner", model=MODELS["English"], tokenizer=MODELS["English"], aggregation_strategy="simple")
-}
+# ============================
+# HuggingFace для английского
+# ============================
+from transformers import pipeline
+english_ner = pipeline(
+    "ner",
+    model="dbmdz/bert-large-cased-finetuned-conll03-english",
+    tokenizer="dbmdz/bert-large-cased-finetuned-conll03-english",
+    aggregation_strategy="simple"
+)
 
 # ============================
-# Функция распознавания
+# Функция распознавания сущностей
 # ============================
-def auto_ner(text):
+def recognize_entities_auto(text):
+    # Определяем язык
     try:
         lang = detect(text)
     except:
-        lang = "ru"  # По умолчанию русский
+        lang = "ru"
+
+    entities = {"PER": [], "ORG": [], "LOC": []}
 
     if lang == "en":
-        model_choice = "English"
+        results = english_ner(text)
+        for res in results:
+            label = res['entity_group']
+            word = res['word']
+            if label in ["PER", "ORG", "LOC", "GPE"]:
+                if label == "GPE":
+                    label = "LOC"
+                entities[label].append(word)
     else:
-        model_choice = "Russian"
-
-    ner = ner_pipelines[model_choice]
-    results = ner(text)
-
-    if not results:
-        return "Сущности не найдены."
-
-    output = {"PER": [], "ORG": [], "LOC": []}
-    for item in results:
-        entity = item['word']
-        label = item['entity_group']
+        doc = Doc(text)
+        doc.segment(segmenter)
+        doc.tag_ner(ner_tagger)
+        for span in doc.spans:
+            label = span.type
+            if label in ["PER", "ORG", "LOC"]:
+                entities[label].append(span.text)
 
-        if label == "PER":
-            output["PER"].append(entity)
-        elif label == "ORG":
-            output["ORG"].append(entity)
-        elif label == "LOC" or label == "GPE":
-            output["LOC"].append(entity)
+    # Формируем текстовый вывод
+    output_text = ""
+    for key, items in entities.items():
+        if items:
+            # Убираем дубликаты
+            unique_items = list(dict.fromkeys(items))
+            output_text += f"{key}: {'; '.join(unique_items)}\n"
 
-    result_text = ""
-    for key, values in output.items():
-        if values:
-            result_text += f"{key}: {'; '.join(list(dict.fromkeys(values)))}\n"
-    return result_text.strip()
+    return output_text.strip() if output_text else "Сущности не найдены."
 
 # ============================
 # Gradio интерфейс
 # ============================
 iface = gr.Interface(
-    fn=auto_ner,
-    inputs=gr.Textbox(lines=15, placeholder="Введите текст здесь (русский или английский, несколько предложений)..."),
+    fn=recognize_entities_auto,
+    inputs=gr.Textbox(lines=15, placeholder="Введите русский или английский текст здесь..."),
     outputs=gr.Textbox(label="Распознанные сущности (PER/ORG/LOC)"),
     title="Автоматический NER для русского и английского текста",
     description="PER – человек, ORG – организация, LOC – место. Текст любого языка обрабатывается автоматически."