Utilizing datasets like GLUE and customized ones, we’ll use the Coach class for simpler coaching, and in addition attempt native PyTorch. We’ll work on duties together with binary sentiment, seven-class classification, and predicting numerical values.
The next subjects will probably be lined on this Part :
- Introduction to textual content classification
- Wonderful-tuning the BERT mannequin for single-sentence binary classification
- Coaching a classification mannequin with native PyTorch
- Wonderful-tuning BERT for multi-class classification with customized datasets
- Wonderful-tuning BERT for sentence-pair regression
- Multi-label textual content classification
- Using run_glue.py to fine-tune the fashions
Technical necessities
We will probably be utilizing Jupyter Pocket book to run our coding workouts. You will have Python 3.6+ for this.
