PassbyGrocer's picture
End of training
d5f0cc0 verified
metadata
library_name: transformers
license: apache-2.0
base_model: hfl/chinese-roberta-wwm-ext-large
tags:
  - generated_from_trainer
metrics:
  - precision
  - recall
  - f1
  - accuracy
model-index:
  - name: robert_bilstm_mega_res-ner-msra-ner
    results: []

robert_bilstm_mega_res-ner-msra-ner

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

  • Loss: 0.0668
  • Precision: 0.9473
  • Recall: 0.9473
  • F1: 0.9473
  • Accuracy: 0.9928

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: 32
  • eval_batch_size: 32
  • seed: 42
  • optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
  • lr_scheduler_type: linear
  • num_epochs: 25

Training results

Training Loss Epoch Step Validation Loss Precision Recall F1 Accuracy
0.0421 1.0 1449 0.0274 0.9225 0.9341 0.9282 0.9923
0.0077 2.0 2898 0.0355 0.9255 0.9389 0.9321 0.9910
0.0066 3.0 4347 0.0376 0.9397 0.9384 0.9391 0.9921
0.0031 4.0 5796 0.0421 0.9385 0.9401 0.9393 0.9926
0.0071 5.0 7245 0.0446 0.9365 0.9446 0.9406 0.9923
0.0007 6.0 8694 0.0431 0.9457 0.9398 0.9428 0.9930
0.0003 7.0 10143 0.0494 0.9412 0.9408 0.9410 0.9926
0.0013 8.0 11592 0.0584 0.9379 0.9338 0.9358 0.9917
0.0003 9.0 13041 0.0557 0.9373 0.9422 0.9398 0.9923
0.0011 10.0 14490 0.0525 0.9395 0.9463 0.9429 0.9926
0.0 11.0 15939 0.0569 0.9379 0.9449 0.9414 0.9924
0.0001 12.0 17388 0.0586 0.9358 0.9434 0.9396 0.9922
0.0 13.0 18837 0.0601 0.9439 0.9437 0.9438 0.9926
0.0013 14.0 20286 0.0606 0.9395 0.9454 0.9424 0.9924
0.0 15.0 21735 0.0591 0.9451 0.9495 0.9473 0.9926
0.0 16.0 23184 0.0608 0.9399 0.9490 0.9444 0.9926
0.0 17.0 24633 0.0620 0.9440 0.9454 0.9447 0.9927
0.0 18.0 26082 0.0636 0.9493 0.9454 0.9473 0.9926
0.0 19.0 27531 0.0681 0.9460 0.9451 0.9456 0.9926
0.0 20.0 28980 0.0630 0.9430 0.9430 0.9430 0.9925
0.0 21.0 30429 0.0620 0.9445 0.9463 0.9454 0.9928
0.0 22.0 31878 0.0671 0.9456 0.9446 0.9451 0.9926
0.0 23.0 33327 0.0682 0.9479 0.9451 0.9465 0.9926
0.0 24.0 34776 0.0671 0.9475 0.9466 0.9470 0.9927
0.0 25.0 36225 0.0668 0.9473 0.9473 0.9473 0.9928

Framework versions

  • Transformers 4.46.2
  • Pytorch 2.4.1+cu124
  • Datasets 3.1.0
  • Tokenizers 0.20.3