spellcorrector_0411

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: 0.0830
• Precision: 0.9784
• Recall: 0.9815
• F1: 0.9799
• Accuracy: 0.9828

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
0.2319	1.0	975	0.1268	0.9458	0.9834	0.9642	0.9741
0.1296	2.0	1950	0.1063	0.9530	0.9812	0.9669	0.9754
0.1095	3.0	2925	0.0883	0.9653	0.9788	0.9720	0.9786
0.0934	4.0	3900	0.0842	0.9692	0.9776	0.9734	0.9790
0.0829	5.0	4875	0.0794	0.9716	0.9797	0.9756	0.9809
0.0753	6.0	5850	0.0755	0.9729	0.9816	0.9773	0.9817
0.0695	7.0	6825	0.0739	0.9751	0.9789	0.9770	0.9815
0.0641	8.0	7800	0.0736	0.9767	0.9798	0.9782	0.9821
0.0591	9.0	8775	0.0744	0.9767	0.9805	0.9786	0.9822
0.0537	10.0	9750	0.0742	0.9777	0.9798	0.9787	0.9822
0.0502	11.0	10725	0.0753	0.9773	0.9806	0.9790	0.9825
0.0472	12.0	11700	0.0757	0.9780	0.9808	0.9794	0.9827
0.044	13.0	12675	0.0768	0.9772	0.9816	0.9794	0.9827
0.0407	14.0	13650	0.0784	0.9775	0.9815	0.9795	0.9827
0.039	15.0	14625	0.0790	0.9779	0.9816	0.9798	0.9828
0.0364	16.0	15600	0.0804	0.9778	0.9813	0.9795	0.9825
0.0343	17.0	16575	0.0811	0.9783	0.9811	0.9797	0.9828
0.0329	18.0	17550	0.0819	0.9785	0.9820	0.9803	0.9829
0.0314	19.0	18525	0.0822	0.9785	0.9808	0.9797	0.9826
0.0308	20.0	19500	0.0830	0.9784	0.9815	0.9799	0.9828

Framework versions

• Transformers 4.28.0
• Pytorch 2.1.0+cu118
• Datasets 2.14.6
• Tokenizers 0.13.3