MT_authorship_new

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

• Loss: 0.6223
• Accuracy: 0.6675

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: 64
• eval_batch_size: 64
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• training_steps: 63715
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.6354	0.2499	3185	0.6264	0.6229
0.6219	0.4998	6370	0.6111	0.6345
0.611	0.7498	9555	0.6004	0.6415
0.6091	0.9997	12740	0.5952	0.6466
0.5745	1.2496	15925	0.5926	0.6520
0.5753	1.4995	19110	0.5891	0.6543
0.5749	1.7495	22295	0.5908	0.6573
0.5741	1.9994	25480	0.5822	0.6605
0.5424	2.2493	28665	0.5911	0.6603
0.5384	2.4992	31850	0.5893	0.6625
0.5367	2.7491	35035	0.5872	0.6621
0.5445	2.9991	38220	0.5846	0.6656
0.5069	3.2490	41405	0.6050	0.6652
0.5048	3.4989	44590	0.6013	0.6667
0.5129	3.7488	47775	0.6100	0.6660
0.5042	3.9987	50960	0.6020	0.6672
0.4778	4.2487	54145	0.6240	0.6674
0.4771	4.4986	57330	0.6250	0.6673
0.4749	4.7485	60515	0.6249	0.6671
0.478	4.9984	63700	0.6223	0.6675

Framework versions

• Transformers 4.40.1
• Pytorch 2.2.1+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1