gte-multilingual-base-edu-scorer-lr3e4-bs32-swe

This model is a fine-tuned version of Alibaba-NLP/gte-multilingual-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.9336
• Precision: 0.0478
• Recall: 0.1667
• F1 Macro: 0.0742
• Accuracy: 0.2866

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: 0.0003
• train_batch_size: 64
• eval_batch_size: 64
• 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: 20
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1 Macro	Accuracy
No log	0	0	4.1021	0.0513	0.1666	0.0785	0.3078
1.73	0.6793	1000	2.0901	0.0471	0.1667	0.0735	0.2824
1.6976	1.3587	2000	1.9935	0.0471	0.1667	0.0735	0.2824
1.6589	2.0380	3000	2.0057	0.1027	0.1670	0.0742	0.2828
1.6456	2.7174	4000	2.3130	0.0471	0.1667	0.0735	0.2824
1.6927	3.3967	5000	2.1982	0.0471	0.1667	0.0734	0.2824
1.6659	4.0761	6000	2.0477	0.0471	0.1667	0.0734	0.2824
1.6728	4.7554	7000	2.0806	0.0471	0.1667	0.0734	0.2824
1.6714	5.4348	8000	2.1802	0.0471	0.1667	0.0734	0.2824
1.6831	6.1141	9000	2.6460	0.0471	0.1667	0.0734	0.2824
1.6592	6.7935	10000	2.2354	0.0471	0.1667	0.0734	0.2824
1.654	7.4728	11000	2.3264	0.0471	0.1667	0.0734	0.2824
1.6622	8.1522	12000	2.1686	0.0471	0.1667	0.0734	0.2824
1.6618	8.8315	13000	2.1844	0.0471	0.1667	0.0734	0.2824
1.6624	9.5109	14000	2.0627	0.0471	0.1667	0.0734	0.2824
1.7082	10.1902	15000	2.0176	0.0471	0.1667	0.0734	0.2824
1.7107	10.8696	16000	2.2305	0.0471	0.1667	0.0734	0.2824
1.6538	11.5489	17000	2.2190	0.0471	0.1667	0.0734	0.2824
1.6453	12.2283	18000	2.4751	0.0179	0.1667	0.0323	0.1073
1.6562	12.9076	19000	2.3006	0.0179	0.1667	0.0323	0.1073
1.6767	13.5870	20000	2.0474	0.0471	0.1667	0.0734	0.2824
1.672	14.2663	21000	1.9881	0.0179	0.1667	0.0323	0.1073
1.6355	14.9457	22000	1.9897	0.0471	0.1667	0.0734	0.2824
1.6608	15.625	23000	2.1116	0.0179	0.1667	0.0323	0.1073
1.6616	16.3043	24000	2.0205	0.0179	0.1667	0.0323	0.1073
1.6895	16.9837	25000	2.3756	0.0179	0.1667	0.0323	0.1073
1.6613	17.6630	26000	2.1950	0.0179	0.1667	0.0323	0.1073
1.6405	18.3424	27000	2.0526	0.0179	0.1667	0.0323	0.1073
1.6242	19.0217	28000	2.0278	0.0179	0.1667	0.0323	0.1073
1.6839	19.7011	29000	2.0130	0.0179	0.1667	0.0323	0.1073

Framework versions

• Transformers 4.55.0
• Pytorch 2.5.1+cu121
• Datasets 4.0.0
• Tokenizers 0.21.4