MiMe-MeMo/MeMo-BERT-SA

MeMo_BERT-SA

This model is a fine-tuned version of MiMe-MeMo/MeMo-BERT-03 on an https://huggingface.co/MiMe-MeMo/MeMo-Dataset-SA dataset. It achieves the following results on the evaluation set:

• Loss: 2.3157
• F1-score: 0.7821

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	297	0.7332	0.7373
0.5945	2.0	594	0.9941	0.7644
0.5945	3.0	891	1.1745	0.7714
0.2243	4.0	1188	1.6828	0.7210
0.2243	5.0	1485	1.6114	0.7308
0.0807	6.0	1782	1.9495	0.7515
0.0133	7.0	2079	2.0289	0.7536
0.0133	8.0	2376	2.1528	0.7566
0.0002	9.0	2673	2.2011	0.7560
0.0002	10.0	2970	2.2830	0.7351
0.0199	11.0	3267	2.3461	0.7456
0.0103	12.0	3564	2.2212	0.7533
0.0103	13.0	3861	2.2322	0.7622
0.0063	14.0	4158	2.3226	0.7599
0.0063	15.0	4455	2.3019	0.7561
0.0026	16.0	4752	2.2407	0.7675
0.0	17.0	5049	2.2745	0.7715
0.0	18.0	5346	2.3114	0.7716
0.0002	19.0	5643	2.3157	0.7821
0.0002	20.0	5940	2.3187	0.7749

Framework versions

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