Skip to content

Latest commit

 

History

History
42 lines (31 loc) · 2.28 KB

README.md

File metadata and controls

42 lines (31 loc) · 2.28 KB

Main Page

Dialog Dataset

[TR] Qi Jia ve diğerleri tarafından yayınlanan bir makale için hazırlanan verisetidir. Çince doktor-hasta diyalog kayıtlarını içeren veriseti Google Translate API yardımıyla Türkçeye çevrilerek kullanılmıştır.

[EN] This dataset is a part of an article published by Qi Jia et al. The dataset containing the Chinese doctor-patient dialogue records was translated into Turkish with the help of Google Translate API.

Fine Tuning Metrics

Model Phase Accuracy Precision Recall F1
BERTurk Train/Eval 0.957015 0.951456 0.885542 0.956515

Test 0.961081 0.950317 0.902610 0.925849
DistilBERTurk Train/Eval 0.957556 0.947705 0.891566 0.918779

Test 0.962432 0.952481 0.905622 0.928461
ConvBERTurk Train/Eval 0.958773 0.951311 0.892570 0.921005

Test 0.963243 0.956475 0.904618 0.929824
ELECTRA Base Train/Eval 0.955123 0.953057 0.876506 0.913180

Test 0.958378 0.942226 0.900602 0.920944


Fine Tuning Metrics by Step

BERTurk -- DistilBERTurk -- ConvBERTurk -~- ELECTRA Base

Accuracy F1
Precision Recall