xlnet-base-cased_fold_7_binary_v1

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: 1.7774
• F1: 0.8111

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	0.4189	0.7903
0.432	2.0	576	0.3927	0.8045
0.432	3.0	864	0.4868	0.8108
0.2573	4.0	1152	0.6763	0.8019
0.2573	5.0	1440	0.8132	0.8105
0.1612	6.0	1728	0.8544	0.8086
0.0972	7.0	2016	1.1274	0.8109
0.0972	8.0	2304	1.2622	0.8056
0.0515	9.0	2592	1.3398	0.8013
0.0515	10.0	2880	1.5421	0.8082
0.0244	11.0	3168	1.4931	0.8042
0.0244	12.0	3456	1.5744	0.8045
0.0287	13.0	3744	1.4169	0.8091
0.0255	14.0	4032	1.5790	0.7999
0.0255	15.0	4320	1.6094	0.7994
0.0098	16.0	4608	1.5758	0.8006
0.0098	17.0	4896	1.5326	0.8140
0.0203	18.0	5184	1.6431	0.8114
0.0203	19.0	5472	1.7105	0.8072
0.0104	20.0	5760	1.6353	0.8139
0.0062	21.0	6048	1.6762	0.8108
0.0062	22.0	6336	1.7076	0.8106
0.0088	23.0	6624	1.7887	0.8035
0.0088	24.0	6912	1.7731	0.8099
0.0026	25.0	7200	1.7774	0.8111

Framework versions

• Transformers 4.21.1
• Pytorch 1.12.0+cu113
• Datasets 2.4.0
• Tokenizers 0.12.1