spellcorrector_1009_v1

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

• Loss: 0.0188
• Precision: 0.9947
• Recall: 0.9968
• F1: 0.9958
• Accuracy: 0.9941

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: 4
• eval_batch_size: 4
• 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
0.2712	1.0	1951	0.2093	0.9634	0.9655	0.9645	0.9423
0.2165	2.0	3902	0.1618	0.9645	0.9671	0.9658	0.9541
0.1801	3.0	5853	0.1308	0.9667	0.9708	0.9687	0.9641
0.1473	4.0	7804	0.1003	0.9714	0.9734	0.9724	0.9713
0.1278	5.0	9755	0.0872	0.9761	0.9772	0.9766	0.9757
0.1058	6.0	11706	0.0774	0.9752	0.9825	0.9788	0.9779
0.0948	7.0	13657	0.0682	0.9815	0.9851	0.9833	0.9806
0.0865	8.0	15608	0.0609	0.9847	0.9899	0.9873	0.9825
0.0812	9.0	17559	0.0516	0.9842	0.9904	0.9873	0.9849
0.071	10.0	19510	0.0477	0.9873	0.9926	0.9899	0.9862
0.0658	11.0	21461	0.0404	0.9873	0.9920	0.9897	0.9880
0.0571	12.0	23412	0.0360	0.9899	0.9915	0.9907	0.9891
0.0512	13.0	25363	0.0316	0.9905	0.9936	0.9920	0.9904
0.0511	14.0	27314	0.0307	0.9884	0.9936	0.9910	0.9908
0.047	15.0	29265	0.0261	0.9921	0.9947	0.9934	0.9919
0.0413	16.0	31216	0.0239	0.9926	0.9958	0.9942	0.9925
0.0412	17.0	33167	0.0220	0.9947	0.9963	0.9955	0.9932
0.0371	18.0	35118	0.0206	0.9947	0.9963	0.9955	0.9936
0.0368	19.0	37069	0.0191	0.9947	0.9968	0.9958	0.9940
0.032	20.0	39020	0.0188	0.9947	0.9968	0.9958	0.9941

Framework versions

• Transformers 4.28.0
• Pytorch 2.0.1+cu118
• Datasets 2.14.5
• Tokenizers 0.13.3