distilbert-base-uncased_fold_4_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: 1.9355
• F1: 0.7891

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	289	0.5637	0.7485
0.5729	2.0	578	0.5305	0.7805
0.5729	3.0	867	0.6948	0.7670
0.2548	4.0	1156	0.8351	0.7744
0.2548	5.0	1445	1.0005	0.8027
0.1157	6.0	1734	1.1578	0.7978
0.0473	7.0	2023	1.2275	0.7953
0.0473	8.0	2312	1.3245	0.7916
0.0276	9.0	2601	1.3728	0.7953
0.0276	10.0	2890	1.4577	0.7867
0.0149	11.0	3179	1.5832	0.7731
0.0149	12.0	3468	1.5056	0.7818
0.0143	13.0	3757	1.6263	0.7904
0.0066	14.0	4046	1.6596	0.7793
0.0066	15.0	4335	1.6795	0.7941
0.0022	16.0	4624	1.8443	0.7744
0.0022	17.0	4913	1.7160	0.7953
0.0034	18.0	5202	1.7819	0.7781
0.0034	19.0	5491	1.7931	0.7904
0.0036	20.0	5780	1.8447	0.7818
0.0014	21.0	6069	1.9975	0.7707
0.0014	22.0	6358	1.9324	0.7830
0.0008	23.0	6647	1.9086	0.7842
0.0008	24.0	6936	1.9507	0.7867
0.0002	25.0	7225	1.9355	0.7891

Framework versions

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