MeMo_BERT-WSD-02

This model is a fine-tuned version of MiMe-MeMo/MeMo-BERT-02 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 3.4274
• F1-score: 0.3523

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: 5e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	F1-score
No log	1.0	61	1.6387	0.1229
No log	2.0	122	1.4760	0.2016
No log	3.0	183	1.4995	0.2754
No log	4.0	244	1.9750	0.2519
No log	5.0	305	2.3945	0.3115
No log	6.0	366	2.9512	0.2336
No log	7.0	427	3.3875	0.2933
No log	8.0	488	3.4274	0.3523
0.6662	9.0	549	3.8142	0.3244
0.6662	10.0	610	4.3089	0.3271
0.6662	11.0	671	4.4936	0.2806
0.6662	12.0	732	4.4328	0.2893
0.6662	13.0	793	4.8685	0.2353
0.6662	14.0	854	4.4432	0.3018
0.6662	15.0	915	4.6090	0.2820
0.6662	16.0	976	4.5783	0.2871
0.0033	17.0	1037	4.6179	0.2871
0.0033	18.0	1098	4.7173	0.2847
0.0033	19.0	1159	4.7240	0.3029
0.0033	20.0	1220	4.7167	0.3029

Framework versions

• Transformers 4.38.2
• Pytorch 2.2.1+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2