deit-tiny-distilled-patch16-224_alpha0.7_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: 0.8363
• Accuracy: 0.7473

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
0.5597	1.0	164	1.4047	0.2549
0.5314	2.0	328	1.3956	0.2625
0.4538	3.0	492	1.0888	0.5759
0.3622	4.0	656	0.9513	0.6388
0.3137	5.0	820	0.9170	0.6800
0.2708	6.0	984	0.8860	0.7028
0.2519	7.0	1148	0.8460	0.7495
0.2431	8.0	1312	0.8368	0.7191
0.2427	9.0	1476	0.8486	0.7310
0.2257	10.0	1640	0.8432	0.7397
0.2237	11.0	1804	0.8530	0.7213
0.229	12.0	1968	0.8623	0.7256
0.2182	13.0	2132	0.8367	0.7462
0.2222	14.0	2296	0.8380	0.7440
0.2236	15.0	2460	0.8538	0.7386
0.2213	16.0	2624	0.8363	0.7473
0.2242	17.0	2788	0.8428	0.7451
0.2148	18.0	2952	0.8365	0.7473
0.2213	19.0	3116	0.8387	0.7440
0.215	20.0	3280	0.8400	0.7419

Framework versions

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