my-model-MiniLM-Area

This model is a fine-tuned version of Microsoft/Multilingual-MiniLM-L12-H384 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.5228
• Accuracy: 0.4323
• F1: 0.3979
• Precision: 0.3932
• Recall: 0.4323

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: 30
• eval_batch_size: 5
• 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	Accuracy	F1	Precision	Recall
1.8812	1.0	25	1.8038	0.2839	0.1709	0.2712	0.2839
1.8043	2.0	50	1.7540	0.3742	0.2586	0.2046	0.3742
1.7687	3.0	75	1.6908	0.3806	0.2557	0.1927	0.3806
1.6959	4.0	100	1.6325	0.4	0.2695	0.2033	0.4
1.6178	5.0	125	1.6401	0.4129	0.3338	0.2874	0.4129
1.5189	6.0	150	1.5471	0.4581	0.3631	0.3030	0.4581
1.4393	7.0	175	1.5966	0.4258	0.3761	0.3451	0.4258
1.3757	8.0	200	1.5716	0.4452	0.3945	0.3556	0.4452
1.3032	9.0	225	1.5691	0.4387	0.3646	0.3443	0.4387
1.2434	10.0	250	1.5740	0.4452	0.4057	0.3798	0.4452
1.1837	11.0	275	1.5108	0.4645	0.3854	0.3852	0.4645
1.1231	12.0	300	1.5409	0.4516	0.3972	0.3561	0.4516
1.0815	13.0	325	1.5111	0.4774	0.4116	0.3865	0.4774
1.0555	14.0	350	1.5171	0.4645	0.4014	0.3674	0.4645
0.9964	15.0	375	1.4971	0.4581	0.3877	0.3504	0.4581
0.9627	16.0	400	1.5157	0.4516	0.4118	0.3882	0.4516
0.9247	17.0	425	1.4996	0.4387	0.3882	0.3664	0.4387
0.9286	18.0	450	1.4990	0.4452	0.4008	0.3856	0.4452
0.892	19.0	475	1.5288	0.4323	0.4025	0.4031	0.4323
0.8843	20.0	500	1.5228	0.4323	0.3979	0.3932	0.4323

Framework versions

• Transformers 4.41.0
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1