spellcorrector_03_04_qwerty_v16

This model is a fine-tuned version of Buseak/spellcorrector_20_02_050_qwerty_v14 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0000
• Precision: 1.0
• Recall: 1.0
• F1: 1.0
• Accuracy: 1.0

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
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
0.0104	1.0	967	0.0037	0.9984	0.9973	0.9979	0.9990
0.0053	2.0	1934	0.0020	0.9995	0.9995	0.9995	0.9995
0.0032	3.0	2901	0.0011	1.0	0.9995	0.9997	0.9997
0.0025	4.0	3868	0.0011	1.0	1.0	1.0	0.9997
0.0017	5.0	4835	0.0005	1.0	1.0	1.0	0.9999
0.001	6.0	5802	0.0004	1.0	1.0	1.0	0.9999
0.0012	7.0	6769	0.0002	1.0	1.0	1.0	0.9999
0.0007	8.0	7736	0.0002	1.0	1.0	1.0	0.9999
0.0007	9.0	8703	0.0001	1.0	1.0	1.0	1.0000
0.0005	10.0	9670	0.0001	1.0	0.9995	0.9997	1.0000
0.0005	11.0	10637	0.0001	1.0	1.0	1.0	1.0000
0.0004	12.0	11604	0.0000	1.0	1.0	1.0	1.0000
0.0002	13.0	12571	0.0000	1.0	1.0	1.0	1.0000
0.0002	14.0	13538	0.0000	1.0	1.0	1.0	1.0000
0.0002	15.0	14505	0.0000	1.0	1.0	1.0	1.0000
0.0002	16.0	15472	0.0000	1.0	1.0	1.0	1.0000
0.0001	17.0	16439	0.0000	1.0	1.0	1.0	1.0
0.0001	18.0	17406	0.0000	1.0	1.0	1.0	1.0
0.0001	19.0	18373	0.0000	1.0	1.0	1.0	1.0000
0.0001	20.0	19340	0.0000	1.0	1.0	1.0	1.0

Framework versions

• Transformers 4.38.1
• Pytorch 2.1.0+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2