mobilevit-small_alpha0.7_temp3.0_t3

This model is a fine-tuned version of c14kevincardenas/ClimBEiT-t3 on the c14kevincardenas/beta_caller_284_person_crop_seq_withlimb dataset. It achieves the following results on the evaluation set:

• Loss: 0.9566
• Accuracy: 0.6616

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: 5e-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
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.5602	1.0	164	1.4016	0.2636
0.5293	2.0	328	1.3396	0.3178
0.4722	3.0	492	1.1894	0.4740
0.4285	4.0	656	1.1300	0.5336
0.4026	5.0	820	1.0758	0.5824
0.3918	6.0	984	1.0470	0.5900
0.3733	7.0	1148	1.0198	0.6095
0.3622	8.0	1312	0.9975	0.6356
0.3504	9.0	1476	0.9980	0.6052
0.3345	10.0	1640	1.0102	0.6171
0.3362	11.0	1804	1.0125	0.6009
0.3281	12.0	1968	0.9796	0.6302
0.3146	13.0	2132	0.9592	0.6616
0.3172	14.0	2296	0.9733	0.6475
0.3091	15.0	2460	0.9679	0.6540
0.3071	16.0	2624	0.9691	0.6508
0.3073	17.0	2788	0.9566	0.6616
0.2953	18.0	2952	0.9579	0.6486
0.3014	19.0	3116	0.9578	0.6605
0.2909	20.0	3280	0.9599	0.6551

Framework versions

• Transformers 4.45.2
• Pytorch 2.5.0+cu124
• Datasets 3.0.1
• Tokenizers 0.20.1