eurobert210m_Biodiversite_v1

This model is a fine-tuned version of EuroBERT/EuroBERT-210m on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0472
• Accuracy: 0.9882
• F1: 0.9883

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: 100
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
1.4966	1.0	64	0.8616	0.6456	0.6308
0.6196	2.0	128	0.4380	0.8505	0.8474
0.3807	3.0	192	0.3218	0.8951	0.8944
0.2712	4.0	256	0.2907	0.9044	0.9029
0.2462	5.0	320	0.1044	0.9691	0.9684
0.1943	6.0	384	0.1301	0.9593	0.9594
0.1349	7.0	448	0.0663	0.9775	0.9773
0.1177	8.0	512	0.1050	0.9745	0.9744
0.1325	9.0	576	0.0673	0.9833	0.9833
0.1153	10.0	640	0.0454	0.9873	0.9873
0.1078	11.0	704	0.1175	0.9819	0.9819
0.1275	12.0	768	0.0467	0.9858	0.9859
0.0808	13.0	832	0.0596	0.9868	0.9868
0.1032	14.0	896	0.0428	0.9887	0.9887
0.0803	15.0	960	0.0352	0.9912	0.9912
0.0995	16.0	1024	0.1187	0.9779	0.9776
0.0847	17.0	1088	0.0822	0.9828	0.9828
0.0837	18.0	1152	0.0334	0.9912	0.9912
0.0614	19.0	1216	0.0706	0.9892	0.9892
0.0628	20.0	1280	0.0472	0.9882	0.9883

Framework versions

• Transformers 4.48.3
• Pytorch 2.5.1+cu124
• Datasets 3.3.2
• Tokenizers 0.21.0