deberta-v3-base-uner-down-synth400

This model is a fine-tuned version of microsoft/deberta-v3-base on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 0.1315
F1: 0.7476
Precision: 0.7075
Recall: 0.7924
Accuracy: 0.9782

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: 2.5e-05
train_batch_size: 16
eval_batch_size: 32
seed: 42
optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	F1	Precision	Recall	Accuracy
0.4058	0.8	20	0.1934	0.055	0.12	0.0357	0.9433
0.1639	1.6	40	0.1266	0.3309	0.2967	0.3741	0.9582
0.0579	2.4	60	0.1016	0.4844	0.4477	0.5276	0.9669
0.062	3.2	80	0.0864	0.6686	0.6147	0.7330	0.9747
0.0521	4.0	100	0.0919	0.7018	0.6788	0.7265	0.9766
0.015	4.8	120	0.0922	0.7314	0.6931	0.7741	0.9767
0.0732	5.6	140	0.0931	0.7448	0.7068	0.7870	0.9779
0.0123	6.4	160	0.0965	0.7244	0.6710	0.7870	0.9764
0.019	7.2	180	0.0975	0.7285	0.6821	0.7816	0.9771
0.0314	8.0	200	0.0982	0.7396	0.6984	0.7859	0.9774
0.0125	8.8	220	0.1071	0.7488	0.7214	0.7784	0.9786
0.003	9.6	240	0.1129	0.7467	0.7368	0.7568	0.9784
0.0114	10.4	260	0.1137	0.7479	0.7361	0.76	0.9790
0.0029	11.2	280	0.1153	0.7417	0.7039	0.7838	0.9783
0.0159	12.0	300	0.1171	0.7422	0.6922	0.8	0.9775
0.0077	12.8	320	0.1167	0.7585	0.7466	0.7708	0.9792
0.002	13.6	340	0.1138	0.7400	0.6907	0.7968	0.9772
0.001	14.4	360	0.1171	0.7505	0.7172	0.7870	0.9785
0.0035	15.2	380	0.1202	0.7467	0.7102	0.7870	0.9781
0.0016	16.0	400	0.1212	0.7507	0.7296	0.7730	0.9786
0.0342	16.8	420	0.1221	0.7481	0.7042	0.7978	0.9779
0.0015	17.6	440	0.1216	0.7438	0.6983	0.7957	0.9782
0.0008	18.4	460	0.1230	0.7439	0.6993	0.7946	0.9780
0.001	19.2	480	0.1261	0.7463	0.7096	0.7870	0.9783
0.0008	20.0	500	0.1262	0.7427	0.6988	0.7924	0.9779
0.0008	20.8	520	0.1269	0.7386	0.6891	0.7957	0.9773
0.0009	21.6	540	0.1276	0.7418	0.6964	0.7935	0.9776
0.0009	22.4	560	0.1279	0.7420	0.7010	0.7881	0.9778
0.0006	23.2	580	0.1304	0.7397	0.6911	0.7957	0.9773
0.001	24.0	600	0.1307	0.7402	0.6904	0.7978	0.9772
0.0007	24.8	620	0.1308	0.7417	0.6930	0.7978	0.9774
0.0015	25.6	640	0.1305	0.7452	0.7024	0.7935	0.9780
0.0006	26.4	660	0.1305	0.7454	0.7071	0.7881	0.9782
0.0007	27.2	680	0.1308	0.7458	0.7078	0.7881	0.9782
0.0016	28.0	700	0.1311	0.7469	0.7072	0.7914	0.9782
0.0011	28.8	720	0.1314	0.7476	0.7075	0.7924	0.9782
0.0011	29.6	740	0.1315	0.7476	0.7075	0.7924	0.9782

Framework versions

Transformers 4.57.1
Pytorch 2.8.0+cu128
Datasets 4.3.0
Tokenizers 0.22.1

BramVanroy/deberta-v3-base-uner-down-synth400

deberta-v3-base-uner-down-synth400

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

Seed this model