Model Description

This model is a fine-tuned version of distilbert-base-uncased and fine-tuning it on the github ticket tagger dataset. It classifies issue into 3 common categories: Bug, Enhancement, Questions.

It achieves the following results on the evaluation set:

• Accuracy: 0.7862

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 3e-5
• train_batch_size: 16
• optimizer: AdamW with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 0
• num_epochs: 5

Codes

https://github.com/IvanLauLinTiong/IntelliLabel