sosolee/t5-small-custom

Model Card for t5_small Summarization Model

Model Details

• Model Name: t5-small
• Model Type: Text Summarization
• Architecture: Transformer-based model
• Number of Parameters: Approximately 60 million

Training Data

• Dataset: CNN/Daily Mail
• Data Source: Hugging Face Datasets
• Training Size: X samples (X는 실제 훈련 데이터 수로 교체)
• Data Characteristics: The dataset consists of news articles and their corresponding summaries.

Training Procedure

• Training Framework: Hugging Face Transformers
• Batch Size: 4
• Learning Rate: 2e-5
• Number of Epochs: 1
• Evaluation Metrics: ROUGE, BLEU

How to Use

• Dependencies:
  • Install required libraries: transformers, datasets
• Example Usage:

  from transformers import pipeline
  summarizer = pipeline("summarization", model="your_model_path")
  summary = summarizer("Your input text here.")