bert-finetuned-ner

<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->

bert-finetuned-ner

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

• Loss: 0.1668
• Precision: 0.5293
• Recall: 0.6065
• F1: 0.5653
• Accuracy: 0.9545
• Art Precision: 0.3481
• Art Recall: 0.5595
• Art F1: 0.4292
• Con Precision: 0.4365
• Con Recall: 0.5374
• Con F1: 0.4817
• Loc Precision: 0.5232
• Loc Recall: 0.5486
• Loc F1: 0.5356
• Mat Precision: 0.0
• Mat Recall: 0.0
• Mat F1: 0.0
• Per Precision: 0.7390
• Per Recall: 0.8304
• Per F1: 0.7820
• Spe Precision: 0.0
• Spe Recall: 0.0
• Spe F1: 0.0
• B-art Precision: 0.5061
• B-art Recall: 0.7381
• B-art F1: 0.6005
• B-con Precision: 0.5429
• B-con Recall: 0.6463
• B-con F1: 0.5901
• B-loc Precision: 0.5625
• B-loc Recall: 0.125
• B-loc F1: 0.2045
• B-mat Precision: 0.0
• B-mat Recall: 0.0
• B-mat F1: 0.0
• B-per Precision: 0.8057
• B-per Recall: 0.8940
• B-per F1: 0.8476
• B-spe Precision: 0.0
• B-spe Recall: 0.0
• B-spe F1: 0.0
• I-art Precision: 0.5337
• I-art Recall: 0.4265
• I-art F1: 0.4741
• I-con Precision: 0.7419
• I-con Recall: 0.3194
• I-con F1: 0.4466
• I-loc Precision: 0.6601
• I-loc Recall: 0.7187
• I-loc F1: 0.6881
• I-mat Precision: 0.0
• I-mat Recall: 0.0
• I-mat F1: 0.0
• I-per Precision: 0.9351
• I-per Recall: 0.7448
• I-per F1: 0.8292
• Macro F1: 0.3714
• Micro F1: 0.4809

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: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• training_steps: 200

Training results

Framework versions

• Transformers 4.57.1
• Pytorch 2.8.0+cu126
• Datasets 4.0.0
• Tokenizers 0.22.1