Model information
Train information
- train_runtime: 1477.3876
- train_steps_per_second: 2.416
- train_loss: 0.3722160959110207
- epoch: 5.0
How to use
from transformers import AutoTokenizer, AutoModelForSequenceClassification
tokenizer = AutoTokenizer.from_pretrained (
"seongju/klue-tc-bert-base-multilingual-cased"
)
model = AutoModelForSequenceClassification.from_pretrained (
"seongju/klue-tc-bert-base-multilingual-cased"
)
mapping = {0: 'IT๊ณผํ', 1: '๊ฒฝ์ ', 2: '์ฌํ',
3: '์ํ๋ฌธํ', 4: '์ธ๊ณ', 5: '์คํฌ์ธ ', 6: '์ ์น'}
inputs = tokenizer(
"๋ฐฑ์ ํํผ ๊ฐ๋ฅ์ฑ? ๋จ๋ฏธ์์ ์๋ก์ด ๋ณ์ด ๋ฐ์ด๋ฌ์ค ๊ธ์ ํ์ฐ ",
padding=True, truncation=True, max_length=128, return_tensors="pt"
)
outputs = model(**inputs)
probs = outputs[0].softmax(1)
output = mapping[probs.argmax().item()]