jjonhwa's repositories
clova-airush-2022
๐ฅ Top 2 in "Nonsense Documents Detection" & ๐ฅ Top 3 in "Unknown Documents Detection"
Cross-Encoder-with-Bi-Encoder
For Tok-k passages that have passed through the Bi-Encoder Retrival, ReRank is performed using CrossEncoder.
Policy-to-utilize-DT-in-traditional-markets
๐ฅ Top 2 in "ํต๊ณ๋ฐ์ดํฐ ๋ถ์, ํ์ฉ ๋ํ(ํต๊ณ์ฒญ)" - ๋ฅ์คํธ๋ ธ๋ฉ ์๋_์ ํต์์ฅ DT ํ์ฉ ๋ฐฉ์
Retrieval_Streamlit_Demo
Cross-Encoder-with-Bi-Encoder๋ฅผ ํ์ฉํ WebPage ๋ฐ๋ชจ
AI_SPARK_CHALLENG_Object_Detection
๐ Top 5% in "์ 2ํ ์ฐ๊ตฌ๊ฐ๋ฐํน๊ตฌ ์ธ๊ณต์ง๋ฅ ๊ฒฝ์ง๋ํ AI SPARK ์ฑ๋ฆฐ์ง"
jjonhwa.github.io
Uk's Blog
Paper_Review
Read the paper and Review it.
Fusion-in-Decoder-Custom
๐ซ Fusion in decoder customizing (FiD-Light, FiDO)
image-classification-level1-10
image-classification-level1-10 created by GitHub Classroom
klue-level2-nlp-15
klue-level2-nlp-15-1 created by GitHub Classroom
mrc-level2-nlp-15
mrc-level2-nlp-15 created by GitHub Classroom
southkorea-maps
South Korea administrative divisions in ESRI Shapefile, GeoJSON and TopoJSON formats.
Boostcamp-AI-Tech-Product-Serving
๋ถ์คํธ์บ ํ AI Tech - Product Serving ์๋ฃ
Coding_Test
Practice about Conding_Test
KLUE-Relation_Extraction
๐ฅ Top 3 in Relation Extraction Task in Naver BoostCamp AI Tech
Optuna_and_Knowledge-Distllation
Maximize the performance of lightweight models using optuna and knowledge distllation.