STS-Lora-Fine-Tuning-Capstone-bert-testing-21-with-lower-r

This model is a fine-tuned version of dslim/bert-base-NER on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.5146
• Accuracy: 0.3604

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: 3e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	180	1.7493	0.2429
No log	2.0	360	1.7409	0.2444
1.7063	3.0	540	1.7315	0.2408
1.7063	4.0	720	1.7132	0.2741
1.7063	5.0	900	1.6786	0.2879
1.6653	6.0	1080	1.6617	0.2959
1.6653	7.0	1260	1.6399	0.3031
1.6653	8.0	1440	1.6109	0.3205
1.5949	9.0	1620	1.5937	0.3292
1.5949	10.0	1800	1.5715	0.3321
1.5949	11.0	1980	1.5627	0.3387
1.5344	12.0	2160	1.5543	0.3459
1.5344	13.0	2340	1.5396	0.3590
1.4932	14.0	2520	1.5295	0.3524
1.4932	15.0	2700	1.5270	0.3568
1.4932	16.0	2880	1.5240	0.3575
1.4738	17.0	3060	1.5177	0.3604
1.4738	18.0	3240	1.5185	0.3590
1.4738	19.0	3420	1.5156	0.3604
1.4609	20.0	3600	1.5146	0.3604

Framework versions

• PEFT 0.10.0
• Transformers 4.38.2
• Pytorch 2.2.1+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2