Noise_MeMo_BERT-3_02

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

• Loss: 0.0729
• F1-score: 0.6452

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
0.095	1.0	915	0.0707	0.5373
0.0532	2.0	1830	0.0877	0.4286
0.0317	3.0	2745	0.0951	0.4889
0.0196	4.0	3660	0.1114	0.3590
0.0265	5.0	4575	0.0810	0.6333
0.0118	6.0	5490	0.1097	0.4783
0.0247	7.0	6405	0.1153	0.4583
0.0255	8.0	7320	0.0781	0.5634
0.0242	9.0	8235	0.1156	0.5455
0.0423	10.0	9150	0.1186	0.4
0.0246	11.0	10065	0.1057	0.5000
0.0224	12.0	10980	0.0998	0.56
0.0168	13.0	11895	0.0729	0.6452
0.0106	14.0	12810	0.1171	0.4444
0.0097	15.0	13725	0.0735	0.5818
0.0187	16.0	14640	0.0943	0.5417
0.0128	17.0	15555	0.1011	0.5417
0.0098	18.0	16470	0.1029	0.5714
0.0116	19.0	17385	0.0949	0.6182
0.0084	20.0	18300	0.0956	0.6154

Framework versions

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