canine_deasciifier_0205

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: 0.0004
• Precision: 0.9983
• Recall: 0.9991
• F1: 0.9987
• Accuracy: 0.9999

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: 32
• eval_batch_size: 32
• 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
No log	1.0	244	0.0462	0.8117	0.8638	0.8370	0.9829
No log	2.0	488	0.0335	0.8432	0.9286	0.8838	0.9878
0.1144	3.0	732	0.0173	0.9182	0.9569	0.9371	0.9939
0.1144	4.0	976	0.0101	0.9567	0.9705	0.9636	0.9966
0.0241	5.0	1220	0.0067	0.9716	0.9786	0.9751	0.9978
0.0241	6.0	1464	0.0049	0.9784	0.9846	0.9815	0.9984
0.0137	7.0	1708	0.0039	0.9830	0.9886	0.9858	0.9987
0.0137	8.0	1952	0.0030	0.9870	0.9911	0.9891	0.9990
0.0088	9.0	2196	0.0024	0.9893	0.9933	0.9913	0.9992
0.0088	10.0	2440	0.0019	0.9916	0.9947	0.9932	0.9994
0.0061	11.0	2684	0.0013	0.9941	0.9962	0.9952	0.9996
0.0061	12.0	2928	0.0010	0.9955	0.9971	0.9963	0.9997
0.0045	13.0	3172	0.0010	0.9952	0.9973	0.9963	0.9997
0.0045	14.0	3416	0.0008	0.9966	0.9980	0.9973	0.9998
0.0033	15.0	3660	0.0006	0.9973	0.9982	0.9978	0.9998
0.0033	16.0	3904	0.0006	0.9975	0.9986	0.9980	0.9998
0.0027	17.0	4148	0.0004	0.9982	0.9988	0.9985	0.9999
0.0027	18.0	4392	0.0004	0.9982	0.9990	0.9986	0.9999
0.0023	19.0	4636	0.0004	0.9984	0.9991	0.9988	0.9999
0.0023	20.0	4880	0.0004	0.9983	0.9991	0.9987	0.9999

Framework versions

• Transformers 4.28.1
• Pytorch 2.0.0+cu118
• Datasets 2.12.0
• Tokenizers 0.13.3