我们基于布朗语料库对RoBERTa基础模型进行了微调，用于预测文本是否采用非虚构直叙风格或虚构描述性风格撰写。此处虚构与非虚构仅指写作风格，与文本所述内容的真实性无关。更多细节请参阅此博客及引用论文：https://bekushal.medium.com/fictometer-a-simple-and-explainable-algorithm-for-sentiment-analysis-31186d2a8c7e

from transformers import AutoModelForSequenceClassification, AutoTokenizer

# 加载模型
model_name = "bekushal/FictoRoBERTa"

model = AutoModelForSequenceClassification.from_pretrained(model_name)
tokenizer = AutoTokenizer.from_pretrained(model_name)

# 示例输入文本
input_text = "这是个阳光明媚的日子，微风轻拂，我感到非常快乐。"

# 预处理输入
inputs = tokenizer(input_text, return_tensors="pt", truncation=True, padding=True)

# 执行推理
outputs = model(**inputs)

# 获取预测类别概率
predicted_probabilities = outputs.logits.softmax(dim=-1)

# 获取预测类别标签
predicted_label = predicted_probabilities.argmax().item()

# 将预测标签转换为可读格式
predicted_class = "虚构类" if predicted_label == 1 else "非虚构类"

# 显示结果
print("预测类别:", predicted_class)
print("预测类别概率 [非虚构类, 虚构类]:", predicted_probabilities)

许可证: apache-2.0