spellcorrector_0511_v3_30_epoch

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: 0.1888
• Precision: 0.9708
• Recall: 0.9751
• F1: 0.9730
• Accuracy: 0.9745

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: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
0.2233	1.0	1945	0.1721	0.9292	0.9619	0.9453	0.9570
0.1589	2.0	3890	0.1346	0.9455	0.9670	0.9561	0.9646
0.1387	3.0	5835	0.1208	0.9545	0.9684	0.9614	0.9680
0.1201	4.0	7780	0.1173	0.9620	0.9616	0.9618	0.9682
0.1067	5.0	9725	0.1138	0.9628	0.9662	0.9645	0.9699
0.0951	6.0	11670	0.1096	0.9653	0.9680	0.9666	0.9714
0.0864	7.0	13615	0.1122	0.9673	0.9663	0.9668	0.9715
0.0757	8.0	15560	0.1141	0.9680	0.9699	0.9689	0.9719
0.0676	9.0	17505	0.1154	0.9687	0.9718	0.9703	0.9723
0.0606	10.0	19450	0.1193	0.9679	0.9730	0.9705	0.9731
0.0548	11.0	21395	0.1246	0.9701	0.9696	0.9698	0.9724
0.0478	12.0	23340	0.1279	0.9708	0.9705	0.9706	0.9730
0.0417	13.0	25285	0.1316	0.9707	0.9706	0.9706	0.9733
0.0375	14.0	27230	0.1387	0.9706	0.9734	0.9720	0.9729
0.0334	15.0	29175	0.1405	0.9704	0.9748	0.9726	0.9741
0.0292	16.0	31120	0.1450	0.9704	0.9736	0.9720	0.9733
0.0261	17.0	33065	0.1492	0.9684	0.9744	0.9714	0.9734
0.0233	18.0	35010	0.1529	0.9701	0.9755	0.9728	0.9739
0.0205	19.0	36955	0.1580	0.9708	0.9733	0.9720	0.9735
0.0174	20.0	38900	0.1636	0.9710	0.9749	0.9729	0.9741
0.0166	21.0	40845	0.1674	0.9700	0.9763	0.9732	0.9742
0.0142	22.0	42790	0.1729	0.9706	0.9742	0.9724	0.9733
0.0126	23.0	44735	0.1737	0.9711	0.9742	0.9726	0.9743
0.0116	24.0	46680	0.1788	0.9706	0.9762	0.9734	0.9747
0.0102	25.0	48625	0.1808	0.9712	0.9750	0.9731	0.9745
0.0093	26.0	50570	0.1817	0.9710	0.9741	0.9725	0.9742
0.0082	27.0	52515	0.1838	0.9706	0.9748	0.9727	0.9741
0.0082	28.0	54460	0.1865	0.9711	0.9754	0.9733	0.9745
0.0076	29.0	56405	0.1873	0.9706	0.9753	0.9729	0.9745
0.0069	30.0	58350	0.1888	0.9708	0.9751	0.9730	0.9745

Framework versions

• Transformers 4.35.0
• Pytorch 2.1.0+cu118
• Datasets 2.14.6
• Tokenizers 0.14.1