File size: 797 Bytes
6d2354e |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 |
from transformers import AutoTokenizer, AutoModelForSequenceClassification
# Load pre-trained model and tokenizer
#model_name = "ProsusAI/finbert"
#model_name = "ahmedrachid/FinancialBERT-Sentiment-Analysis"
model_name="mrm8488/distilroberta-finetuned-financial-news-sentiment-analysis"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForSequenceClassification.from_pretrained(model_name)
# Example: Classify a financial text
text = "With the launch of Apple Silicon, Apple shares have increased"
inputs = tokenizer(text, return_tensors="pt")
outputs = model(**inputs)
predictions = outputs.logits.argmax(dim=1).item()
# 'predictions' will contain the sentiment class (e.g., 0 for negative, 1 for neutral, 2 for positive)
print("Predicted sentiment class:", predictions)
|