constituency-brackets-20

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

• Train Loss: 0.2020
• Train Acc: 0.9356
• Validation Loss: 0.2929
• Validation Acc: 0.9120
• Epoch: 19

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:

• optimizer: {'name': 'Adam', 'learning_rate': 5e-06, 'decay': 0.0, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-07, 'amsgrad': False}
• training_precision: float32

Training results

Train Loss	Train Acc	Validation Loss	Validation Acc	Epoch
2.4703	0.3783	1.4719	0.5858	0
1.2149	0.6600	0.8922	0.7269	1
0.8721	0.7343	0.6914	0.7779	2
0.7186	0.7715	0.6028	0.8037	3
0.6239	0.7987	0.5427	0.8240	4
0.5432	0.8342	0.4469	0.8677	5
0.4521	0.8665	0.4092	0.8760	6
0.4100	0.8761	0.3867	0.8819	7
0.3792	0.8855	0.3761	0.8849	8
0.3526	0.8926	0.3469	0.8938	9
0.3304	0.8981	0.3433	0.8944	10
0.3091	0.9049	0.3329	0.8977	11
0.2935	0.9081	0.3178	0.9028	12
0.2769	0.9138	0.3140	0.9032	13
0.2614	0.9173	0.2994	0.9114	14
0.2472	0.9213	0.2954	0.9128	15
0.2344	0.9260	0.2899	0.9142	16
0.2229	0.9292	0.2971	0.9092	17
0.2136	0.9322	0.2872	0.9143	18
0.2020	0.9356	0.2929	0.9120	19

Framework versions

• Transformers 4.26.0
• TensorFlow 2.9.2
• Datasets 2.8.0
• Tokenizers 0.13.2