Modelpublic

luissattelmayer/EuroBERT-immigration-stance-neutral

sourceHugging Faceapache-2.0updated 2y agoView on Hugging Face

0likes11downloads

Model Card

EuroBERT-immigration-stance-neutral

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: 1.3840
—Accuracy: 0.8298
—F1 Macro: 0.8281
—Accuracy Balanced: 0.8268
—F1 Micro: 0.8298
—Precision Macro: 0.8314
—Recall Macro: 0.8268
—Precision Micro: 0.8298
—Recall Micro: 0.8298

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: 2e-05
—trainbatchsize: 4
—evalbatchsize: 40
—seed: 42
—optimizer: Use OptimizerNames.ADAMWTORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizerargs=No additional optimizer arguments
—lrschedulertype: linear
—lrschedulerwarmup_ratio: 0.06
—num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1 Macro	Accuracy Balanced	F1 Micro	Precision Macro	Recall Macro	Precision Micro	Recall Micro
0.7142	1.0	605	1.2430	0.6860	0.6762	0.7000	0.6860	0.7398	0.7000	0.6860	0.6860
0.6286	2.0	1210	0.6901	0.5256	0.3507	0.4942	0.5256	0.3897	0.4942	0.5256	0.5256
0.6978	3.0	1815	0.8429	0.7785	0.7779	0.7784	0.7785	0.7777	0.7784	0.7785	0.7785
0.4281	4.0	2420	1.2471	0.7901	0.7865	0.7850	0.7901	0.7949	0.7850	0.7901	0.7901
0.1542	5.0	3025	1.3219	0.8099	0.8073	0.8058	0.8099	0.8133	0.8058	0.8099	0.8099
0.0797	6.0	3630	2.2835	0.7802	0.7787	0.7888	0.7802	0.8036	0.7888	0.7802	0.7802
0.0633	7.0	4235	1.2912	0.8364	0.8352	0.8345	0.8364	0.8365	0.8345	0.8364	0.8364
0.0558	8.0	4840	1.5358	0.8099	0.8067	0.8050	0.8099	0.8155	0.8050	0.8099	0.8099
0.0639	9.0	5445	1.3711	0.8182	0.8153	0.8136	0.8182	0.8231	0.8136	0.8182	0.8182
0.0335	10.0	6050	1.3840	0.8298	0.8281	0.8268	0.8298	0.8314	0.8268	0.8298	0.8298

Framework versions

—Transformers 4.50.0
—Pytorch 2.6.0+cu124
—Datasets 3.4.1
—Tokenizers 0.21.1