i> 安裝 PyTorch
!pip install transformers datasets torch

ii> 匯入 pipeline
from transformers import pipeline

iii> 建立情感分析器
classifier = pipeline("sentiment-analysis")
** sentiment-analysis 是情感分析器

iv> 測試單句
result = classifier("I love learning LLMs with Hugging Face!") print(result)

• 會輸出兩個 Key，一個是分類標籤 laber，一個是信心分數 score

v> 測試多句
texts = [ "I love this movie!", "This is the worst day of my life." ] results = classifier(texts) for t, r in zip(texts, results): print(t, "->", r)

#中文情感分析

classifier = pipeline(
    "sentiment-analysis",
    model="uer/roberta-base-finetuned-jd-binary-chinese"
)

print(classifier("這部電影真的很好看！"))
print(classifier("這個產品太爛了，我很失望。"))

• uer/roberta-base-finetuned-jd-binary-chinese 是專門做中文二分類情感分析的模型