FinBERT-Sentiment-KRW-Comment (v3)

์ด ๋ชจ๋ธ์€ snunlp/KR-FinBERT-SC๋ฅผ ๊ธฐ๋ฐ˜์œผ๋กœ ํŒŒ์ธํŠœ๋‹ํ•œ ํ•œ๊ตญ์–ด ๊ธˆ์œต ๊ฐ์ • ๋ถ„์„ ๋ชจ๋ธ์ž…๋‹ˆ๋‹ค.
ํŠนํžˆ ํ™˜์œจ(FX) ๊ด€๋ จ ๋Œ“๊ธ€์—์„œ ๋‚˜ํƒ€๋‚˜๋Š” ๊ฐ์ •์„ ๊ณตํฌ(0) ๋˜๋Š” ์š•์‹ฌ(1) ์ด์ง„ ๋ถ„๋ฅ˜ํ•˜๋Š” ๋ฐ ๋ชฉ์ ์ด ์žˆ์Šต๋‹ˆ๋‹ค.

๐Ÿงพ ๋ผ๋ฒจ ์ •์˜

๋ผ๋ฒจ ์„ค๋ช…
0 ๊ณตํฌ (Fear)
1 ์š•์‹ฌ (Greed)

๐Ÿ‹๏ธโ€โ™‚๏ธ ํ•™์Šต ์ •๋ณด

  • Base model: snunlp/KR-FinBERT-SC
  • Task: ๊ฐ์ • ์ด์ง„ ๋ถ„๋ฅ˜ (๊ณตํฌ vs ์š•์‹ฌ)
  • Input: ํ•œ๊ตญ์–ด ๋Œ“๊ธ€ (content)
  • Output: 0 ๋˜๋Š” 1
  • Training epochs: 4
  • Train size: ์•ฝ X ๊ฐœ
  • Eval size: ์•ฝ X ๊ฐœ
  • Evaluation metric: Accuracy, Precision, Recall, F1

๐Ÿ“Š ํ‰๊ฐ€ ๊ฒฐ๊ณผ (Test Set ๊ธฐ์ค€)

Metric Score
Accuracy 0.94
Precision 0.94
Recall 0.94
F1-score 0.94

๊ณตํฌ(0): precision=0.95, recall=0.91, f1=0.93
์š•์‹ฌ(1): precision=0.93, recall=0.96, f1=0.94

๐Ÿงช ์‚ฌ์šฉ ์˜ˆ์‹œ

from transformers import pipeline

pipe = pipeline("text-classification", model="DataWizardd/finbert-sentiment-krw-comment-v3")
pipe("์‹ฌ์ƒ์น˜ ์•Š๋„ค์š” ์ค‘๋™์ „์Ÿ ๋ฐœ๋ฐœํ•˜๊ณ  1์›”8์ผ์— ๊น€์ •์€์ด ๋ฏธ์‚ฌ์ผ ์‹คํ—˜ํ•˜๊ณ  ๊ทธ๋Ÿฌ๋ฉด 1170์€ ๊ธฐ๋ณธ์ด๊ณ  1190์›๋„ ์ˆœ์‹๊ฐ„์ผ๋“ฏ.")
# โ†’ [{'label': '1', 'score': 0.98}]
Downloads last month
3
Safetensors
Model size
101M params
Tensor type
F32
ยท
Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐Ÿ™‹ Ask for provider support