mobilevit-x-small_alpha0.5_temp5.0_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: 0.8649
• Accuracy: 0.6833

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
0.7932	1.0	164	1.3866	0.2755
0.7662	2.0	328	1.3513	0.3178
0.6878	3.0	492	1.1694	0.4881
0.6218	4.0	656	1.0661	0.5651
0.5987	5.0	820	0.9977	0.6150
0.56	6.0	984	0.9475	0.6367
0.5365	7.0	1148	0.9845	0.6193
0.5345	8.0	1312	0.9277	0.6508
0.5115	9.0	1476	0.8977	0.6540
0.486	10.0	1640	0.9238	0.6605
0.4781	11.0	1804	0.9176	0.6540
0.4817	12.0	1968	0.8862	0.6779
0.4524	13.0	2132	0.8682	0.6866
0.4557	14.0	2296	0.8980	0.6627
0.456	15.0	2460	0.8837	0.6768
0.4469	16.0	2624	0.8789	0.6790
0.4443	17.0	2788	0.8708	0.6790
0.4266	18.0	2952	0.8649	0.6833
0.4352	19.0	3116	0.8735	0.6768
0.4239	20.0	3280	0.8790	0.6681

Framework versions

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