mobilevit-small_alpha0.5_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: 0.8562
• Accuracy: 0.7072

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
0.8034	1.0	164	1.3824	0.2863
0.7605	2.0	328	1.3374	0.2918
0.6903	3.0	492	1.1406	0.5152
0.6045	4.0	656	1.0571	0.5629
0.5779	5.0	820	1.0227	0.5813
0.5423	6.0	984	0.9594	0.6334
0.5154	7.0	1148	0.9498	0.6518
0.4984	8.0	1312	0.9111	0.6757
0.47	9.0	1476	0.9155	0.6518
0.4474	10.0	1640	0.9253	0.6518
0.4547	11.0	1804	0.9355	0.6562
0.4432	12.0	1968	0.8718	0.6649
0.4235	13.0	2132	0.8763	0.7039
0.419	14.0	2296	0.8878	0.6866
0.4083	15.0	2460	0.8869	0.6898
0.4043	16.0	2624	0.8697	0.6985
0.4061	17.0	2788	0.8562	0.7072
0.389	18.0	2952	0.8595	0.7082
0.3982	19.0	3116	0.8661	0.6985
0.3823	20.0	3280	0.8689	0.7115

Framework versions

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