Eliac11/FitModelKTP

This model is a fine-tuned version of Eliac11/tinkNLP on an unknown dataset. It achieves the following results on the evaluation set:

• Train Loss: 0.2798
• Validation Loss: 3.0329
• Epoch: 41

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': 'AdamWeightDecay', 'learning_rate': 2e-05, 'decay': 0.0, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-07, 'amsgrad': False, 'weight_decay_rate': 0.005}
• training_precision: float32

Training results

Train Loss	Validation Loss	Epoch
3.9625	4.0902	0
3.7233	3.9629	1
3.5004	3.8605	2
3.2832	3.7674	3
3.0715	3.7121	4
2.8942	3.6216	5
2.7196	3.5805	6
2.5480	3.4985	7
2.3730	3.4634	8
2.2230	3.4004	9
2.0769	3.3615	10
1.9424	3.3254	11
1.8186	3.2929	12
1.7042	3.2540	13
1.5870	3.2088	14
1.4708	3.1860	15
1.3852	3.1750	16
1.2858	3.1422	17
1.2008	3.1238	18
1.1220	3.0919	19
1.0540	3.0795	20
0.9794	3.0708	21
0.9126	3.0515	22
0.8535	3.0567	23
0.8003	3.0344	24
0.7417	3.0280	25
0.6913	3.0222	26
0.6467	3.0232	27
0.6076	3.0171	28
0.5714	3.0096	29
0.5296	3.0148	30
0.4989	3.0026	31
0.4666	3.0084	32
0.4366	3.0061	33
0.4123	3.0066	34
0.3904	3.0154	35
0.3676	3.0024	36
0.3467	3.0194	37
0.3249	3.0249	38
0.3095	3.0149	39
0.2947	3.0215	40
0.2798	3.0329	41

Framework versions

• Transformers 4.38.2
• TensorFlow 2.15.0
• Datasets 2.18.0
• Tokenizers 0.15.2