finbert-tone-chinese

Financial Sentiment Analysis in Chinese

This is a fine-tuned version of FinBERT, based on bert-base-chinese, on a private dataset (around ~8k analyst report sentences) for sentiment analysis.

• Test Accuracy = 0.88
• Test Macro F1 = 0.87
• Labels: 0 -> Neutral; 1 -> Positive; 2 -> Negative

Usage

from transformers import TextClassificationPipeline
from transformers import AutoModelForSequenceClassification, TrainingArguments, Trainer
from transformers import BertTokenizerFast
model_path="./fin_sentiment_bert_zh/"
new_model = AutoModelForSequenceClassification.from_pretrained(model_path,output_attentions=True)
tokenizer = BertTokenizerFast.from_pretrained(model_path)
PipelineInterface = TextClassificationPipeline(model=new_model, tokenizer=tokenizer, return_all_scores=True)
label = PipelineInterface("此外宁德时代上半年实现出口约2GWh，同比增加200%+。") 
print(label)

[[{'label': 'LABEL_0', 'score': 0.0007030126871541142}, {'label': 'LABEL_1', 'score': 0.9989339709281921}, {'label': 'LABEL_2', 'score': 0.000363016442861408}]]