MeMo_BERT-WSD-MeMo-BERT-02_last

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: 3.3216
• F1-score: 0.4444

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	1.6200	0.1229
No log	2.0	122	1.4272	0.2217
No log	3.0	183	1.4465	0.2220
No log	4.0	244	1.7271	0.2428
No log	5.0	305	1.8954	0.3110
No log	6.0	366	2.7308	0.1632
No log	7.0	427	2.3096	0.3502
No log	8.0	488	2.5584	0.4155
0.8087	9.0	549	3.1863	0.4054
0.8087	10.0	610	3.3216	0.4444
0.8087	11.0	671	3.3437	0.3938
0.8087	12.0	732	4.1038	0.3355
0.8087	13.0	793	4.4369	0.3054
0.8087	14.0	854	3.8867	0.3430
0.8087	15.0	915	3.9579	0.3430
0.8087	16.0	976	4.0195	0.3430
0.0152	17.0	1037	4.1380	0.3311
0.0152	18.0	1098	4.1298	0.3311
0.0152	19.0	1159	4.1499	0.3311
0.0152	20.0	1220	4.1574	0.3311

Framework versions

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