xlnet-base-cased_fold_1_binary

This model is a fine-tuned version of xlnet-base-cased on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.7607
• F1: 0.7778

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	288	0.4111	0.7555
0.4387	2.0	576	0.4075	0.7540
0.4387	3.0	864	0.5344	0.7567
0.2471	4.0	1152	0.7405	0.7597
0.2471	5.0	1440	1.0564	0.7508
0.1419	6.0	1728	1.0703	0.7751
0.0845	7.0	2016	1.0866	0.7609
0.0845	8.0	2304	1.2135	0.7751
0.05	9.0	2592	1.3649	0.7516
0.05	10.0	2880	1.4943	0.7590
0.0267	11.0	3168	1.5174	0.7412
0.0267	12.0	3456	1.4884	0.7559
0.0278	13.0	3744	1.5109	0.7405
0.0201	14.0	4032	1.7251	0.7409
0.0201	15.0	4320	1.5833	0.7354
0.0185	16.0	4608	1.7744	0.7598
0.0185	17.0	4896	1.8283	0.7619
0.0066	18.0	5184	1.7607	0.7778
0.0066	19.0	5472	1.7503	0.7719
0.0078	20.0	5760	1.7807	0.7508
0.006	21.0	6048	1.6887	0.7629
0.006	22.0	6336	1.7041	0.7678
0.0074	23.0	6624	1.7337	0.7633
0.0074	24.0	6912	1.7548	0.7645
0.0035	25.0	7200	1.7685	0.7621

Framework versions

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