README.md

metadata

license: apache-2.0
tags:
  - generated_from_trainer
metrics:
  - accuracy
  - precision
  - recall
  - f1
base_model: bert-base-uncased
model-index:
  - name: bert_small
    results: []

bert_small

This model is a fine-tuned version of bert-base-uncased on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.4537
• Accuracy: 0.88
• Precision: 0.625
• Recall: 0.3571
• F1: 0.4545
• D-index: 1.6429

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: 4
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 1600
• num_epochs: 20
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1	D-index
No log	1.0	200	0.3773	0.86	0.0	0.0	0.0	1.4803
No log	2.0	400	0.4271	0.86	0.0	0.0	0.0	1.4803
0.5126	3.0	600	0.4598	0.87	0.55	0.3929	0.4583	1.6431
0.5126	4.0	800	0.6620	0.865	0.52	0.4643	0.4906	1.6624
0.2953	5.0	1000	0.8149	0.855	0.4615	0.2143	0.2927	1.5575
0.2953	6.0	1200	0.7819	0.875	0.5714	0.4286	0.4898	1.6623
0.2953	7.0	1400	1.0426	0.86	0.5	0.3571	0.4167	1.6173
0.1565	8.0	1600	1.0078	0.885	0.7273	0.2857	0.4103	1.6231
0.1565	9.0	1800	1.2939	0.865	0.6	0.1071	0.1818	1.5294
0.0643	10.0	2000	1.2661	0.88	0.6429	0.3214	0.4286	1.6299
0.0643	11.0	2200	1.3556	0.87	0.5833	0.25	0.3500	1.5905
0.0643	12.0	2400	1.2393	0.87	0.625	0.1786	0.2778	1.5635
0.0306	13.0	2600	1.3059	0.88	0.625	0.3571	0.4545	1.6429
0.0306	14.0	2800	1.3446	0.88	0.625	0.3571	0.4545	1.6429
0.0019	15.0	3000	1.3618	0.885	0.6471	0.3929	0.4889	1.6622
0.0019	16.0	3200	1.3785	0.885	0.6471	0.3929	0.4889	1.6622
0.0019	17.0	3400	1.4361	0.88	0.625	0.3571	0.4545	1.6429
0.0098	18.0	3600	1.4466	0.88	0.625	0.3571	0.4545	1.6429
0.0098	19.0	3800	1.4518	0.88	0.625	0.3571	0.4545	1.6429
0.0	20.0	4000	1.4537	0.88	0.625	0.3571	0.4545	1.6429

Framework versions

• Transformers 4.28.0
• Pytorch 2.0.1+cu118
• Datasets 2.12.0
• Tokenizers 0.13.3