distilbert-base-uncased_fold_2_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: 0.4724
• F1: 0.7604

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	290	0.4280	0.7515
0.4018	2.0	580	0.4724	0.7604
0.4018	3.0	870	0.5336	0.7428
0.1995	4.0	1160	0.8367	0.7476
0.1995	5.0	1450	0.9242	0.7412
0.089	6.0	1740	1.0987	0.7410
0.0318	7.0	2030	1.1853	0.7584
0.0318	8.0	2320	1.2509	0.7500
0.0189	9.0	2610	1.5060	0.7258
0.0189	10.0	2900	1.5607	0.7534
0.0084	11.0	3190	1.5871	0.7476
0.0084	12.0	3480	1.7206	0.7338
0.0047	13.0	3770	1.6776	0.7340
0.0068	14.0	4060	1.7339	0.7546
0.0068	15.0	4350	1.8279	0.7504
0.0025	16.0	4640	1.7791	0.7411
0.0025	17.0	4930	1.7917	0.7444
0.003	18.0	5220	1.7781	0.7559
0.0029	19.0	5510	1.8153	0.7559
0.0029	20.0	5800	1.7757	0.7414
0.0055	21.0	6090	1.8635	0.7454
0.0055	22.0	6380	1.8483	0.7460
0.001	23.0	6670	1.8620	0.7492
0.001	24.0	6960	1.9058	0.7508
0.0006	25.0	7250	1.8640	0.7504

Framework versions

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