Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
|
@@ -5,7 +5,6 @@ import os
|
|
| 5 |
import re
|
| 6 |
import gradio as gr
|
| 7 |
from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline
|
| 8 |
-
from labels_config import custom_label_definitions
|
| 9 |
|
| 10 |
# --- 1. MODELLERİN HAZIRLANMASI ---
|
| 11 |
ner_pipe = pipeline("ner", model="xlm-roberta-large-finetuned-conll03-english", aggregation_strategy="simple", device=-1)
|
|
@@ -127,7 +126,15 @@ def process_ner(text):
|
|
| 127 |
final_results = []
|
| 128 |
yield "<br>".join(current_logs), []
|
| 129 |
|
| 130 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 131 |
|
| 132 |
initial_results = ner_pipe(text)
|
| 133 |
|
|
@@ -145,7 +152,7 @@ def process_ner(text):
|
|
| 145 |
wiki_text = get_wiki_summary(word)
|
| 146 |
if wiki_text:
|
| 147 |
current_logs.append(f"<div style='margin-left:20px; border-left:2px solid #007bff; padding-left:10px; color:#555;'>🌐 <b>Wiki:</b> {wiki_text}</div>")
|
| 148 |
-
new_label, cot_text = refine_label_with_llm(word, text, wiki_text,
|
| 149 |
current_logs.append(f"<div style='margin-left:20px; border-left:2px solid #28a745; padding-left:10px;'>🧠 <b>CoT Analizi:</b> {cot_text}</div>")
|
| 150 |
res = {"Varlık": word, "İlk Etiket": label, "Nihai Etiket": new_label, "Kaynak": "RAG+LLM"}
|
| 151 |
else:
|
|
|
|
| 5 |
import re
|
| 6 |
import gradio as gr
|
| 7 |
from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline
|
|
|
|
| 8 |
|
| 9 |
# --- 1. MODELLERİN HAZIRLANMASI ---
|
| 10 |
ner_pipe = pipeline("ner", model="xlm-roberta-large-finetuned-conll03-english", aggregation_strategy="simple", device=-1)
|
|
|
|
| 126 |
final_results = []
|
| 127 |
yield "<br>".join(current_logs), []
|
| 128 |
|
| 129 |
+
custom_labels = {
|
| 130 |
+
"MOVIE": {"tanim": "Sinematik eserler.", "anahtar_kelimeler": ["film", "dizi"], "ornekler": ["Inception", "Ezel"]},
|
| 131 |
+
"BOOK": {"tanim": "Yazılı edebi eserler.", "anahtar_kelimeler": ["kitap", "roman"], "ornekler": ["Sefiller", "Nutuk"]},
|
| 132 |
+
"SOFTWARE": {"tanim": "Dijital sistemler.", "anahtar_kelimeler": ["yazılım", "kod"], "ornekler": ["Linux", "Python"]},
|
| 133 |
+
"AWARD": {"tanim": "Resmi ödüller.", "anahtar_kelimeler": ["ödül", "oscar"], "ornekler": ["Nobel", "Altın Portakal"]},
|
| 134 |
+
"PRODUCT": {"tanim": "Ticari ürünler.", "anahtar_kelimeler": ["marka", "model"], "ornekler": ["iPhone", "Tesla"]},
|
| 135 |
+
"NORP": {"tanim": "Etnik ve dini gruplar.", "anahtar_kelimeler": ["millet", "topluluk"], "ornekler": ["Türkler", "Ruslar"]},
|
| 136 |
+
"GPE": {"tanim": "Siyasi bölgeler.", "anahtar_kelimeler": ["ülke", "şehir"], "ornekler": ["Türkiye", "İstanbul"]}
|
| 137 |
+
}
|
| 138 |
|
| 139 |
initial_results = ner_pipe(text)
|
| 140 |
|
|
|
|
| 152 |
wiki_text = get_wiki_summary(word)
|
| 153 |
if wiki_text:
|
| 154 |
current_logs.append(f"<div style='margin-left:20px; border-left:2px solid #007bff; padding-left:10px; color:#555;'>🌐 <b>Wiki:</b> {wiki_text}</div>")
|
| 155 |
+
new_label, cot_text = refine_label_with_llm(word, text, wiki_text, custom_labels)
|
| 156 |
current_logs.append(f"<div style='margin-left:20px; border-left:2px solid #28a745; padding-left:10px;'>🧠 <b>CoT Analizi:</b> {cot_text}</div>")
|
| 157 |
res = {"Varlık": word, "İlk Etiket": label, "Nihai Etiket": new_label, "Kaynak": "RAG+LLM"}
|
| 158 |
else:
|