beit-base-patch16-384_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: 0.5813
• Accuracy: 0.8492

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
0.8109	1.0	164	1.4076	0.2570
0.7119	2.0	328	1.1485	0.4729
0.5643	3.0	492	0.8168	0.7397
0.4666	4.0	656	0.7079	0.7993
0.3935	5.0	820	0.6666	0.8026
0.3497	6.0	984	0.6850	0.7972
0.3417	7.0	1148	0.6163	0.8145
0.3221	8.0	1312	0.6532	0.8308
0.3235	9.0	1476	0.6464	0.8080
0.3001	10.0	1640	0.6392	0.8189
0.2942	11.0	1804	0.6415	0.8341
0.3004	12.0	1968	0.5946	0.8297
0.2884	13.0	2132	0.6203	0.8286
0.2944	14.0	2296	0.6024	0.8384
0.2944	15.0	2460	0.6095	0.8373
0.289	16.0	2624	0.5887	0.8406
0.2924	17.0	2788	0.5882	0.8557
0.2813	18.0	2952	0.5813	0.8492
0.2875	19.0	3116	0.5854	0.8471
0.2818	20.0	3280	0.5851	0.8503

Framework versions

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