README.md

metadata

license: cc-by-sa-4.0
tags:
  - generated_from_trainer
datasets:
  - fin
metrics:
  - precision
  - recall
  - f1
  - accuracy
model-index:
  - name: fin3
    results:
      - task:
          name: Token Classification
          type: token-classification
        dataset:
          name: fin
          type: fin
          config: default
          split: train
          args: default
        metrics:
          - name: Precision
            type: precision
            value: 0.944
          - name: Recall
            type: recall
            value: 0.9402390438247012
          - name: F1
            type: f1
            value: 0.9421157684630739
          - name: Accuracy
            type: accuracy
            value: 0.9921209540034072

fin3

This model is a fine-tuned version of nlpaueb/sec-bert-base on the fin dataset. It achieves the following results on the evaluation set:

• Loss: 0.0748
• Precision: 0.944
• Recall: 0.9402
• F1: 0.9421
• Accuracy: 0.9921

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: 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
No log	1.0	129	0.0669	0.8821	0.9243	0.9027	0.9883
No log	2.0	258	0.0568	0.9289	0.9363	0.9325	0.9913
No log	3.0	387	0.0565	0.9141	0.9323	0.9231	0.9904
0.0556	4.0	516	0.0617	0.9237	0.9163	0.92	0.9904
0.0556	5.0	645	0.0658	0.9243	0.9243	0.9243	0.9904
0.0556	6.0	774	0.0695	0.944	0.9402	0.9421	0.9921
0.0556	7.0	903	0.0731	0.932	0.9283	0.9301	0.9917
0.0016	8.0	1032	0.0750	0.9283	0.9283	0.9283	0.9917
0.0016	9.0	1161	0.0737	0.944	0.9402	0.9421	0.9921
0.0016	10.0	1290	0.0748	0.944	0.9402	0.9421	0.9921

Framework versions

• Transformers 4.25.1
• Pytorch 1.13.0+cu116
• Datasets 2.7.1
• Tokenizers 0.13.2