bert-base-finetuned-sts

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

• Loss: 0.4770
• Pearsonr: 0.8970

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: 128
• eval_batch_size: 128
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Pearsonr
No log	1.0	92	0.6330	0.8717
No log	2.0	184	0.6206	0.8818
No log	3.0	276	0.5010	0.8947
No log	4.0	368	0.4717	0.8956
No log	5.0	460	0.4770	0.8970

Framework versions

• Transformers 4.19.2
• Pytorch 1.11.0+cu113
• Datasets 2.2.2
• Tokenizers 0.12.1