ssummerschool-bert-irony

This model is a fine-tuned version of bert-base-uncased on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.8767
• Accuracy: 0.7015

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: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.6792	0.2793	50	0.6678	0.5759
0.6489	0.5587	100	0.6514	0.6147
0.6282	0.8380	150	0.6360	0.6461
0.5746	1.1173	200	0.6596	0.6492
0.5325	1.3966	250	0.6253	0.6785
0.5431	1.6760	300	0.6226	0.6712
0.5058	1.9553	350	0.5896	0.6869
0.3982	2.2346	400	0.6467	0.6859
0.3837	2.5140	450	0.7012	0.6785
0.3714	2.7933	500	0.7326	0.6586
0.347	3.0726	550	0.7592	0.6702
0.247	3.3520	600	0.7466	0.6942
0.2382	3.6313	650	0.7514	0.6953
0.2304	3.9106	700	0.8268	0.6838
0.1716	4.1899	750	0.8822	0.6806
0.1631	4.4693	800	0.8698	0.6932
0.1435	4.7486	850	0.9178	0.6838

Framework versions

• Transformers 4.41.2
• Pytorch 2.3.0+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1