bert-lora

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

• Loss: 0.4540
• Accuracy: 0.78

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: 5e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	125	0.6659	0.64
No log	2.0	250	0.6518	0.64
No log	3.0	375	0.6353	0.66
0.6613	4.0	500	0.6126	0.7
0.6613	5.0	625	0.5862	0.7
0.6613	6.0	750	0.5677	0.68
0.6613	7.0	875	0.5350	0.72
0.5607	8.0	1000	0.5163	0.74
0.5607	9.0	1125	0.4980	0.74
0.5607	10.0	1250	0.4821	0.75
0.5607	11.0	1375	0.4738	0.77
0.4757	12.0	1500	0.4633	0.78
0.4757	13.0	1625	0.4574	0.78
0.4757	14.0	1750	0.4553	0.78
0.4757	15.0	1875	0.4540	0.78

Framework versions

• PEFT 0.13.0
• Transformers 4.44.2
• Pytorch 2.4.1+cu121
• Datasets 3.0.1
• Tokenizers 0.19.1