identify_stroke

This model is a fine-tuned version of google/vit-base-patch16-224-in21k on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.1127
• Accuracy: 1.0

Model description

Model identifies cricket shot - front drive, hook shot or sweep shot

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: 5e-05
• train_batch_size: 8
• eval_batch_size: 6
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	4	0.4345	1.0
No log	2.0	8	0.3883	1.0
0.3612	3.0	12	0.4099	0.8889
0.3612	4.0	16	0.2452	1.0
0.2934	5.0	20	0.1969	1.0
0.2934	6.0	24	0.1679	1.0
0.2934	7.0	28	0.1403	1.0
0.203	8.0	32	0.1530	1.0
0.203	9.0	36	0.1161	1.0
0.1505	10.0	40	0.1292	1.0
0.1505	11.0	44	0.1031	1.0
0.1505	12.0	48	0.1084	1.0
0.1388	13.0	52	0.1078	1.0
0.1388	14.0	56	0.0937	1.0
0.1076	15.0	60	0.1008	1.0
0.1076	16.0	64	0.1131	1.0
0.1076	17.0	68	0.1007	1.0
0.1047	18.0	72	0.1775	0.8889
0.1047	19.0	76	0.0844	1.0
0.0902	20.0	80	0.1127	1.0

Framework versions

• Transformers 4.33.2
• Pytorch 2.0.1+cu118
• Datasets 2.14.5
• Tokenizers 0.13.3