metadata

license: gpl-3.0
tags:
  - generated_from_trainer
datasets:
  - mim_gold_ner
metrics:
  - precision
  - recall
  - f1
  - accuracy
widget:
  - text: >-
      systurnar guðrún og monique voru einar á báti umkringdar eik og víði með
      þá einu ósk að sameinast fjölskyldu sinni sem fór í smárabíó að horfa á
      jim carey leika í the eternal sunshine of the spotless mind.
model-index:
  - name: IceBERT-finetuned-ner
    results:
      - task:
          name: Token Classification
          type: token-classification
        dataset:
          name: mim_gold_ner
          type: mim_gold_ner
          args: mim-gold-ner
        metrics:
          - name: Precision
            type: precision
            value: 0.8396080453842186
          - name: Recall
            type: recall
            value: 0.7974137931034483
          - name: F1
            type: f1
            value: 0.8179671406320655
          - name: Accuracy
            type: accuracy
            value: 0.979332840486103

IceBERT-finetuned-ner

This model is a fine-tuned version of vesteinn/IceBERT on the mim_gold_ner dataset. It achieves the following results on the evaluation set:

Loss: 0.1048
Precision: 0.8396
Recall: 0.7974
F1: 0.8180
Accuracy: 0.9793

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: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
0.0739	1.0	2904	0.1219	0.8060	0.7621	0.7834	0.9750
0.0438	2.0	5808	0.1130	0.8233	0.7919	0.8073	0.9772
0.0317	3.0	8712	0.1048	0.8396	0.7974	0.8180	0.9793

Framework versions

Transformers 4.11.1
Pytorch 1.9.0+cu102
Datasets 1.12.1
Tokenizers 0.10.3