Pretrained Model
A pretrained model is a model that has been previously trained on a large dataset. These models can be used as starting points for other machine learning tasks to reduce training time and improve model performance.
Areas of application
- Image classification
- Object detection
- Natural language processing
Example
For example, BERT is a popular pretrained model used in natural language processing tasks.
Resources
- BERT Explained
- Transfer learning and fine-tuning | TensorFlow Core
- How do pre-trained models work?. …and why you should use them more often | by Dipam Vasani | Towards Data Science
- [2106.07139] Pre-Trained Models: Past, Present and Future (arxiv.org)
- Pre-trained Models Explained with Examples (vitalflux.com)
- A Prediction Model