summerschool-bert-massive

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: 0.8479
• Accuracy: 0.8283
• F1: 0.8139

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: 2

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
3.8604	0.1389	100	3.3964	0.2720	0.2091
3.046	0.2778	200	2.5353	0.4870	0.3971
2.3977	0.4167	300	2.0141	0.6193	0.5592
1.9293	0.5556	400	1.6738	0.6803	0.6328
1.6997	0.6944	500	1.4307	0.7334	0.6937
1.505	0.8333	600	1.2759	0.7772	0.7469
1.3531	0.9722	700	1.1656	0.7757	0.7445
1.1651	1.1111	800	1.0720	0.7914	0.7707
1.0441	1.25	900	0.9979	0.8032	0.7838
1.0021	1.3889	1000	0.9496	0.8146	0.7977
0.9732	1.5278	1100	0.8996	0.8278	0.8116
0.9025	1.6667	1200	0.8816	0.8214	0.8053
0.8952	1.8056	1300	0.8612	0.8273	0.8128
0.8435	1.9444	1400	0.8479	0.8283	0.8139

Framework versions

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