eurobert210m_EconomieCirculaire_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: 0.0142
• Accuracy: 0.9978
• F1: 0.9978

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
0.6621	1.0	98	0.2769	0.9096	0.9112
0.296	2.0	196	0.1087	0.9671	0.9671
0.1875	3.0	294	0.1938	0.9457	0.9457
0.1906	4.0	392	0.0815	0.9751	0.9751
0.1295	5.0	490	0.1025	0.9687	0.9688
0.1131	6.0	588	0.1083	0.9729	0.9727
0.0841	7.0	686	0.0665	0.9824	0.9825
0.0708	8.0	784	0.0390	0.9882	0.9882
0.0812	9.0	882	0.0457	0.9895	0.9895
0.0939	10.0	980	0.0423	0.9904	0.9904
0.0609	11.0	1078	0.0196	0.9968	0.9968
0.0299	12.0	1176	0.0372	0.9939	0.9939
0.0192	13.0	1274	0.0203	0.9968	0.9968
0.0267	14.0	1372	0.0201	0.9968	0.9968
0.026	15.0	1470	0.0174	0.9958	0.9958
0.0178	16.0	1568	0.0349	0.9965	0.9965
0.0225	17.0	1666	0.0182	0.9971	0.9971
0.0381	18.0	1764	0.0080	0.9974	0.9974
0.0086	19.0	1862	0.0106	0.9981	0.9981
0.056	20.0	1960	0.0237	0.9949	0.9949
0.0284	21.0	2058	0.0142	0.9978	0.9978

Framework versions

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