deit-tiny-distilled-patch16-224_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: 0.7123
• Accuracy: 0.7842

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
0.8039	1.0	164	1.3837	0.2733
0.7645	2.0	328	1.3707	0.2744
0.6389	3.0	492	1.0050	0.6443
0.4958	4.0	656	0.8329	0.7148
0.4273	5.0	820	0.7814	0.7343
0.3586	6.0	984	0.7849	0.7354
0.321	7.0	1148	0.7563	0.7570
0.3109	8.0	1312	0.7642	0.7484
0.313	9.0	1476	0.7382	0.7625
0.2939	10.0	1640	0.7508	0.7516
0.2923	11.0	1804	0.7446	0.7625
0.2962	12.0	1968	0.7362	0.7820
0.2844	13.0	2132	0.7123	0.7842
0.2883	14.0	2296	0.7210	0.7831
0.2895	15.0	2460	0.7299	0.7809
0.2872	16.0	2624	0.7198	0.7831
0.2893	17.0	2788	0.7218	0.7874
0.2794	18.0	2952	0.7164	0.7885
0.2869	19.0	3116	0.7181	0.7852
0.2796	20.0	3280	0.7190	0.7863

Framework versions

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