mobilevit-x-small_alpha0.5_temp3.0

This model is a fine-tuned version of c14kevincardenas/ClimBEiTv2 on the c14kevincardenas/beta_caller_284_person_crop_seq_withlimb_2 dataset. It achieves the following results on the evaluation set:

• Loss: 0.8309
• Accuracy: 0.6581

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: 64
• eval_batch_size: 64
• 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
1.2144	1.0	90	1.3722	0.3103
1.1701	2.0	180	1.3059	0.3933
1.0635	3.0	270	1.1638	0.4792
0.9234	4.0	360	1.0532	0.5217
0.8136	5.0	450	0.9654	0.5711
0.7593	6.0	540	0.8966	0.6166
0.7131	7.0	630	0.8710	0.6334
0.6935	8.0	720	0.9317	0.5998
0.6504	9.0	810	0.8675	0.6294
0.6329	10.0	900	0.8388	0.6492
0.6021	11.0	990	0.8389	0.6522
0.5937	12.0	1080	0.8454	0.6462
0.5687	13.0	1170	0.8777	0.6393
0.5647	14.0	1260	0.8434	0.6354
0.5372	15.0	1350	0.8428	0.6561
0.5302	16.0	1440	0.8376	0.6512
0.5374	17.0	1530	0.8365	0.6561
0.5273	18.0	1620	0.8431	0.6611
0.5197	19.0	1710	0.8309	0.6581
0.5123	20.0	1800	0.8316	0.6670

Framework versions

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