distilbert-base-uncased_fold_1_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: 2.1145
• F1: 0.7757

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	0.5580	0.7646
0.555	2.0	580	0.5820	0.7670
0.555	3.0	870	0.6683	0.7757
0.2633	4.0	1160	0.9137	0.7844
0.2633	5.0	1450	1.1367	0.7708
0.1148	6.0	1740	1.2192	0.7757
0.0456	7.0	2030	1.4035	0.7633
0.0456	8.0	2320	1.5185	0.7658
0.0226	9.0	2610	1.6126	0.7782
0.0226	10.0	2900	1.7631	0.7658
0.0061	11.0	3190	1.7279	0.7794
0.0061	12.0	3480	1.8548	0.7584
0.0076	13.0	3770	1.9052	0.7646
0.0061	14.0	4060	1.9100	0.7757
0.0061	15.0	4350	1.9280	0.7732
0.0025	16.0	4640	1.9991	0.7745
0.0025	17.0	4930	1.9960	0.7757
0.0035	18.0	5220	2.0018	0.7708
0.0015	19.0	5510	2.1099	0.7646
0.0015	20.0	5800	2.1061	0.7695
0.0022	21.0	6090	2.0941	0.7757
0.0022	22.0	6380	2.0967	0.7794
0.0005	23.0	6670	2.1133	0.7745
0.0005	24.0	6960	2.1042	0.7782
0.0021	25.0	7250	2.1145	0.7757

Framework versions

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