mpnet-base-articles-ner

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

• Loss: 0.8471
• F1: 0.7500

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: 16
• eval_batch_size: 16
• 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	F1
1.8042	1.0	5	1.6278	0.0
1.5353	2.0	10	1.5332	0.0
1.499	3.0	15	1.4356	0.1781
1.343	4.0	20	1.3254	0.3789
1.2306	5.0	25	1.2572	0.5075
1.1427	6.0	30	1.1572	0.5700
1.0715	7.0	35	1.0875	0.6305
0.9679	8.0	40	1.0261	0.6667
0.9169	9.0	45	0.9924	0.6512
0.8447	10.0	50	0.9457	0.7137
0.8253	11.0	55	0.9216	0.7094
0.7493	12.0	60	0.9068	0.7303
0.7378	13.0	65	0.8896	0.7404
0.7039	14.0	70	0.8827	0.7398
0.7277	15.0	75	0.8632	0.7635
0.6758	16.0	80	0.8517	0.775
0.6642	17.0	85	0.8618	0.7449
0.6327	18.0	90	0.8522	0.7490
0.6238	19.0	95	0.8477	0.7500
0.6101	20.0	100	0.8471	0.7500

Framework versions

• Transformers 4.24.0
• Pytorch 1.13.0+cu117
• Datasets 2.6.1
• Tokenizers 0.13.2