Edit model card

RoBERTa-ext-large-crf-chinese-finetuned-ner

This model is a fine-tuned version of chinese-roberta-wwm-ext-large on the gyr66/privacy_detection dataset. It achieves the following results on the evaluation set:

  • Loss: 0.7186
  • Precision: 0.6813
  • Recall: 0.7573
  • F1: 0.7173
  • Accuracy: 0.9639

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: 4
  • eval_batch_size: 8
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 10

Training results

Training Loss Epoch Step Validation Loss Precision Recall F1 Accuracy
0.0197 1.0 503 0.6375 0.6663 0.7314 0.6973 0.9621
0.0251 2.0 1006 0.6048 0.6494 0.7435 0.6933 0.9611
0.0176 3.0 1509 0.6196 0.6669 0.7389 0.7011 0.9618
0.0116 4.0 2012 0.6361 0.6511 0.7560 0.6997 0.9624
0.0082 5.0 2515 0.6682 0.6746 0.7387 0.7052 0.9622
0.0067 6.0 3018 0.6587 0.6715 0.7409 0.7045 0.9635
0.0046 7.0 3521 0.6846 0.6770 0.7613 0.7167 0.9636
0.0019 8.0 4024 0.7081 0.6766 0.7510 0.7118 0.9630
0.0014 9.0 4527 0.7064 0.6812 0.7553 0.7163 0.9641
0.001 10.0 5030 0.7186 0.6813 0.7573 0.7173 0.9639

Framework versions

  • Transformers 4.35.2
  • Pytorch 2.1.0+cu121
  • Datasets 2.16.1
  • Tokenizers 0.15.0
Downloads last month
11
Safetensors
Model size
325M params
Tensor type
F32
Β·
Inference Examples
Inference API (serverless) does not yet support model repos that contain custom code.

Dataset used to train gyr66/RoBERTa-ext-large-crf-chinese-finetuned-ner

Evaluation results