me5-small-preskripsi-multiclass

This model is a fine-tuned version of intfloat/multilingual-e5-small on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.8429
• Accuracy: 0.6067
• F1: 0.5847
• Precision: 0.5939
• Recall: 0.5836

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: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall
1.0496	1.0	1850	0.9913	0.5331	0.4339	0.5863	0.4787
0.9587	2.0	3700	0.9192	0.5672	0.5225	0.5576	0.5306
0.9183	3.0	5550	0.8934	0.5730	0.5267	0.5688	0.5357
0.9016	4.0	7400	0.8801	0.5879	0.5683	0.5701	0.5705
0.891	5.0	9250	0.8707	0.5863	0.5475	0.5795	0.5521
0.8842	6.0	11100	0.8623	0.5963	0.5758	0.5812	0.5752
0.8811	7.0	12950	0.8563	0.6006	0.5776	0.5849	0.5776
0.8761	8.0	14800	0.8532	0.6016	0.5815	0.5872	0.5805
0.874	9.0	16650	0.8520	0.5977	0.5670	0.5940	0.5679
0.8733	10.0	18500	0.8492	0.6021	0.5782	0.5931	0.5770
0.8717	11.0	20350	0.8466	0.6031	0.5787	0.5934	0.5777
0.8691	12.0	22200	0.8444	0.6063	0.5829	0.5932	0.5822
0.8675	13.0	24050	0.8448	0.6060	0.5849	0.5950	0.5833
0.866	14.0	25900	0.8434	0.6051	0.5794	0.5947	0.5788
0.8672	15.0	27750	0.8429	0.6067	0.5847	0.5939	0.5836

Framework versions

• Transformers 4.47.1
• Pytorch 2.5.1+cu121
• Datasets 3.2.0
• Tokenizers 0.21.0