distilbert-base-uncased_fold_1_binary

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

• Loss: 1.5992
• F1: 0.7687

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: 25

Training results

Training Loss	Epoch	Step	Validation Loss	F1
No log	1.0	288	0.3960	0.7467
0.3988	2.0	576	0.3947	0.7487
0.3988	3.0	864	0.4511	0.7662
0.1853	4.0	1152	0.7226	0.7285
0.1853	5.0	1440	0.9398	0.7334
0.0827	6.0	1728	1.0547	0.7427
0.0287	7.0	2016	1.1602	0.7563
0.0287	8.0	2304	1.3332	0.7171
0.0219	9.0	2592	1.3429	0.7420
0.0219	10.0	2880	1.2603	0.7648
0.0139	11.0	3168	1.4126	0.7569
0.0139	12.0	3456	1.3195	0.7483
0.0115	13.0	3744	1.4356	0.7491
0.0035	14.0	4032	1.5693	0.7636
0.0035	15.0	4320	1.4071	0.7662
0.0071	16.0	4608	1.4561	0.7579
0.0071	17.0	4896	1.5405	0.7634
0.0041	18.0	5184	1.5862	0.7589
0.0041	19.0	5472	1.6782	0.76
0.0024	20.0	5760	1.5699	0.7677
0.0006	21.0	6048	1.5991	0.7467
0.0006	22.0	6336	1.6205	0.7682
0.0003	23.0	6624	1.6334	0.7643
0.0003	24.0	6912	1.5992	0.7687
0.0011	25.0	7200	1.6053	0.7624

Framework versions

• Transformers 4.21.0
• Pytorch 1.12.0+cu113
• Datasets 2.4.0
• Tokenizers 0.12.1