bert-base-cased-finetuned-sst2

This model is a fine-tuned version of bert-base-uncased on the GLUE QQP dataset. It achieves the following results on the evaluation set:

• Loss: 0.3776
• Accuracy: 0.9108
• F1: 0.8782
• Combined Score: 0.8945

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 2e-05
• train_batch_size: 16
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 3.0

Training results

Training Loss	Epoch	Step	Accuracy	Combined Score	F1	Validation Loss
0.2948	1.0	22741	0.9005	0.8834	0.8664	0.2470
0.1923	2.0	45482	0.9049	0.8884	0.8720	0.2723
0.1339	3.0	68223	0.9109	0.8954	0.8799	0.3585

Framework versions

• Transformers 4.45.0.dev0
• Pytorch 2.3.0+cu121
• Datasets 2.18.0
• Tokenizers 0.19.1