spellcorrector_2610_v16_canine-s

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: 0.0974
• Precision: 0.9789
• Recall: 0.9829
• F1: 0.9809
• Accuracy: 0.9838

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: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
0.252	1.0	976	0.1462	0.9386	0.9800	0.9589	0.9702
0.1463	2.0	1952	0.1256	0.9479	0.9794	0.9634	0.9721
0.1266	3.0	2928	0.1049	0.9578	0.9769	0.9673	0.9745
0.1081	4.0	3904	0.0938	0.9634	0.9787	0.9710	0.9772
0.0963	5.0	4880	0.0856	0.9663	0.9793	0.9727	0.9788
0.0863	6.0	5856	0.0838	0.9705	0.9759	0.9732	0.9786
0.077	7.0	6832	0.0804	0.9734	0.9757	0.9745	0.9797
0.0713	8.0	7808	0.0779	0.9726	0.9804	0.9765	0.9809
0.066	9.0	8784	0.0794	0.9749	0.9767	0.9758	0.9801
0.0602	10.0	9760	0.0748	0.9741	0.9823	0.9782	0.9821
0.0555	11.0	10736	0.0763	0.9750	0.9815	0.9782	0.9822
0.0512	12.0	11712	0.0764	0.9769	0.9800	0.9784	0.9823
0.048	13.0	12688	0.0767	0.9769	0.9822	0.9796	0.9832
0.0453	14.0	13664	0.0793	0.9767	0.9819	0.9793	0.9829
0.0412	15.0	14640	0.0809	0.9774	0.9822	0.9798	0.9832
0.0384	16.0	15616	0.0796	0.9765	0.9830	0.9798	0.9831
0.0364	17.0	16592	0.0830	0.9779	0.9825	0.9802	0.9833
0.0344	18.0	17568	0.0834	0.9779	0.9819	0.9799	0.9831
0.0307	19.0	18544	0.0857	0.9777	0.9823	0.9800	0.9832
0.0283	20.0	19520	0.0869	0.9776	0.9819	0.9797	0.9832
0.0269	21.0	20496	0.0885	0.9781	0.9822	0.9802	0.9833
0.0252	22.0	21472	0.0906	0.9784	0.9814	0.9799	0.9833
0.0229	23.0	22448	0.0932	0.9785	0.9820	0.9802	0.9833
0.0223	24.0	23424	0.0910	0.9785	0.9832	0.9809	0.9835
0.0209	25.0	24400	0.0936	0.9787	0.9824	0.9805	0.9836
0.0199	26.0	25376	0.0948	0.9791	0.9823	0.9807	0.9838
0.0189	27.0	26352	0.0961	0.9792	0.9828	0.9810	0.9838
0.0184	28.0	27328	0.0965	0.9786	0.9834	0.9810	0.9840
0.0178	29.0	28304	0.0970	0.9789	0.9829	0.9809	0.9838
0.0174	30.0	29280	0.0974	0.9789	0.9829	0.9809	0.9838

Framework versions

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