Edit model card

roberta-base-finetuned-WikiNeural

This model is a fine-tuned version of roberta-base.

It achieves the following results on the evaluation set:

  • Loss: 0.0871
  • Loc
    • Precision: 0.9276567437219359
    • Recall: 0.9366918555835433
    • F1: 0.9321524064171123
    • Number: 5955
  • Misc
    • Precision: 0.8334231805929919
    • Recall: 0.916419679905157
    • F1: 0.872953133822699
    • Number: 5061
  • Org
    • Precision: 0.9296179258833669
    • Recall: 0.9382429689765149
    • F1: 0.9339105339105339
    • Number: 3449
  • Per
    • Precision: 0.9688723570869224
    • Recall: 0.9499040307101727
    • F1: 0.9592944369063772
    • Number: 5210
  • Overall
    • Precision: 0.9124
    • Recall: 0.9352
    • F1: 0.9237
    • Accuracy: 0.9910

Model description

For more information on how it was created, check out the following link: https://github.com/DunnBC22/NLP_Projects/blob/main/Token%20Classification/Monolingual/WikiNeural%20-%20Transformer%20Comparison/POS%20Project%20with%20Wikineural%20Dataset%20-%20Roberta-Base%20Transformer.ipynb

Intended uses & limitations

This model is intended to demonstrate my ability to solve a complex problem using technology.

Training and evaluation data

Dataset Source: https://huggingface.co/datasets/Babelscape/wikineural

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 Loc Precision Loc Recall Loc F1 Loc Number Misc Precision Misc Recall Misc F1 Misc Number Org Precision Org Recall Org F1 Org Number Per Precision Per Recall Per F1 Per Number Overall Precision Overall Recall Overall F1 Overall Accuracy
0.1086 1.0 5795 0.1001 0.9149 0.9333 0.9240 5955 0.8158 0.9030 0.8572 5061 0.9134 0.9295 0.9214 3449 0.9642 0.9461 0.9550 5210 0.8997 0.9282 0.9137 0.9896
0.0727 2.0 11590 0.0871 0.9277 0.9367 0.9325 5955 0.8334 0.9164 0.8730 5061 0.9296 0.9382 0.9339 3449 0.9689 0.9499 0.9593 5210 0.9124 0.9352 0.9237 0.9910
  • All values in the cahrt above are rounded to the nearest ten-thousandths.

Framework versions

  • Transformers 4.28.1
  • Pytorch 2.0.1
  • Datasets 2.13.0
  • Tokenizers 0.13.3
Downloads last month
7

Dataset used to train DunnBC22/roberta-base-finetuned-WikiNeural