beit-base-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: 0.6020
• Accuracy: 0.8547

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
0.8129	1.0	164	1.4044	0.2549
0.7193	2.0	328	1.1587	0.5065
0.5431	3.0	492	0.7722	0.7419
0.4342	4.0	656	0.7605	0.7581
0.3763	5.0	820	0.6501	0.8080
0.3418	6.0	984	0.6816	0.7907
0.323	7.0	1148	0.6434	0.8102
0.3123	8.0	1312	0.6709	0.8080
0.3109	9.0	1476	0.6393	0.8124
0.2944	10.0	1640	0.6249	0.8330
0.2928	11.0	1804	0.6533	0.8341
0.2966	12.0	1968	0.6237	0.8275
0.2845	13.0	2132	0.6263	0.8373
0.2933	14.0	2296	0.6042	0.8503
0.2917	15.0	2460	0.6152	0.8449
0.2865	16.0	2624	0.6020	0.8547
0.2923	17.0	2788	0.6055	0.8568
0.281	18.0	2952	0.6053	0.8438
0.2865	19.0	3116	0.6039	0.8503
0.28	20.0	3280	0.6057	0.8482

Framework versions

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