en-multinerd-masked-ner

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

• Loss: 0.0251
• Precision: 0.9353
• Recall: 0.9420
• F1: 0.9386
• Accuracy: 0.9917
• Per-precision: 0.9945
• Per-recall: 0.9953
• Per-f1: 0.9949
• Org-precision: 0.9261
• Org-recall: 0.9434
• Org-f1: 0.9347
• Loc-precision: 0.9679
• Loc-recall: 0.9718
• Loc-f1: 0.9699
• Dis-precision: 0.7003
• Dis-recall: 0.7156
• Dis-f1: 0.7079
• Anim-precision: 0.7021
• Anim-recall: 0.7169
• Anim-f1: 0.7094

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: 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: 2

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy	Per-precision	Per-recall	Per-f1	Org-precision	Org-recall	Org-f1	Loc-precision	Loc-recall	Loc-f1	Dis-precision	Dis-recall	Dis-f1	Anim-precision	Anim-recall	Anim-f1
0.0235	1.0	8205	0.0268	0.9171	0.9466	0.9316	0.9907	0.9943	0.9960	0.9951	0.8943	0.9521	0.9223	0.9595	0.9713	0.9654	0.6531	0.7298	0.6893	0.6518	0.7593	0.7014
0.0151	2.0	16410	0.0251	0.9353	0.9420	0.9386	0.9917	0.9945	0.9953	0.9949	0.9261	0.9434	0.9347	0.9679	0.9718	0.9699	0.7003	0.7156	0.7079	0.7021	0.7169	0.7094

Framework versions

• Transformers 4.36.1
• Pytorch 2.1.0+cu118
• Datasets 2.15.0
• Tokenizers 0.15.0