distilbert-base-uncased_fold_6_ternary_v1

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: 1.9031
• F1: 0.7910

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	292	0.5235	0.7769
0.566	2.0	584	0.5268	0.7923
0.566	3.0	876	0.6189	0.7756
0.2514	4.0	1168	0.7777	0.8026
0.2514	5.0	1460	0.9380	0.7936
0.1175	6.0	1752	1.0957	0.7872
0.0579	7.0	2044	1.2370	0.7923
0.0579	8.0	2336	1.3739	0.7936
0.0259	9.0	2628	1.3457	0.7846
0.0259	10.0	2920	1.4938	0.7872
0.0125	11.0	3212	1.5921	0.7885
0.0108	12.0	3504	1.6504	0.7897
0.0108	13.0	3796	1.7532	0.7756
0.007	14.0	4088	1.7029	0.7821
0.007	15.0	4380	1.7632	0.7987
0.0067	16.0	4672	1.7084	0.7962
0.0067	17.0	4964	1.7559	0.7962
0.0072	18.0	5256	1.8431	0.7987
0.0028	19.0	5548	1.8689	0.7846
0.0028	20.0	5840	1.8641	0.7885
0.0033	21.0	6132	1.8578	0.7923
0.0033	22.0	6424	1.9071	0.7833
0.003	23.0	6716	1.8959	0.7872
0.0011	24.0	7008	1.9073	0.7987
0.0011	25.0	7300	1.9031	0.7910

Framework versions

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