STS-Lora-Fine-Tuning-Capstone-bert-testing-22-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: 1.4650
• Accuracy: 0.3843

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	1.7491	0.2429
No log	2.0	360	1.7398	0.2451
1.7057	3.0	540	1.7266	0.2408
1.7057	4.0	720	1.6996	0.2922
1.7057	5.0	900	1.6538	0.2988
1.6492	6.0	1080	1.6283	0.3118
1.6492	7.0	1260	1.5879	0.3270
1.6492	8.0	1440	1.5578	0.3387
1.5479	9.0	1620	1.5355	0.3503
1.5479	10.0	1800	1.5148	0.3561
1.5479	11.0	1980	1.5062	0.3561
1.4735	12.0	2160	1.5005	0.3691
1.4735	13.0	2340	1.4876	0.3843
1.437	14.0	2520	1.4799	0.3800
1.437	15.0	2700	1.4768	0.3785
1.437	16.0	2880	1.4732	0.3851
1.4223	17.0	3060	1.4689	0.3800
1.4223	18.0	3240	1.4684	0.3822
1.4223	19.0	3420	1.4657	0.3822
1.4123	20.0	3600	1.4650	0.3843

Framework versions

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