distilbert-base-uncased_fold_3_ternary

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: 1.7987
• F1: 0.7460

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	0.5903	0.6893
0.5417	2.0	578	0.5822	0.7130
0.5417	3.0	867	0.6471	0.7385
0.2298	4.0	1156	0.8933	0.7322
0.2298	5.0	1445	1.1002	0.7147
0.1012	6.0	1734	1.2041	0.7249
0.0508	7.0	2023	1.3575	0.7195
0.0508	8.0	2312	1.3896	0.7385
0.018	9.0	2601	1.5363	0.7238
0.018	10.0	2890	1.5336	0.7364
0.0142	11.0	3179	1.6335	0.7308
0.0142	12.0	3468	1.6915	0.7295
0.0047	13.0	3757	1.7087	0.7427
0.0058	14.0	4046	1.7875	0.7378
0.0058	15.0	4335	1.7649	0.7438
0.0051	16.0	4624	1.7987	0.7460
0.0051	17.0	4913	1.8435	0.7404
0.0025	18.0	5202	1.9623	0.7257
0.0025	19.0	5491	1.9005	0.7304
0.0029	20.0	5780	1.9437	0.7374
0.0011	21.0	6069	1.9840	0.7268
0.0011	22.0	6358	1.9411	0.7346
0.0025	23.0	6647	1.9233	0.7438
0.0025	24.0	6936	1.9415	0.7395
0.0015	25.0	7225	1.9481	0.7411

Framework versions

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