xlnet-base-cased_fold_6_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: 1.6214
• F1: 0.8352

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	0.4174	0.7980
0.4661	2.0	580	0.4118	0.8142
0.4661	3.0	870	0.5152	0.8331
0.2714	4.0	1160	0.6901	0.8242
0.2714	5.0	1450	0.6853	0.8451
0.1542	6.0	1740	0.8570	0.8399
0.0935	7.0	2030	1.1342	0.8401
0.0935	8.0	2320	1.1763	0.8397
0.037	9.0	2610	1.3530	0.8215
0.037	10.0	2900	1.3826	0.8402
0.0351	11.0	3190	1.4057	0.8374
0.0351	12.0	3480	1.4259	0.8455
0.0159	13.0	3770	1.4270	0.8431
0.0249	14.0	4060	1.4215	0.8442
0.0249	15.0	4350	1.4245	0.8408
0.0197	16.0	4640	1.4171	0.8353
0.0197	17.0	4930	1.4537	0.8383
0.0137	18.0	5220	1.4786	0.8430
0.0068	19.0	5510	1.5635	0.8443
0.0068	20.0	5800	1.5527	0.8378
0.0062	21.0	6090	1.5917	0.8460
0.0062	22.0	6380	1.6317	0.8318
0.005	23.0	6670	1.6226	0.8340
0.005	24.0	6960	1.6378	0.8310
0.007	25.0	7250	1.6214	0.8352

Framework versions

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