c14kevincardenas/beit-base-patch16-224_alpha0.7_temp5.0_t2

beit-base-patch16-224_alpha0.7_temp5.0

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: 0.5587
• Accuracy: 0.8014

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
1.1179	1.0	90	1.4254	0.2875
0.8879	2.0	180	1.0517	0.5632
0.5644	3.0	270	0.7063	0.7431
0.3909	4.0	360	0.6421	0.7727
0.2787	5.0	450	0.6018	0.7875
0.2167	6.0	540	0.6578	0.7688
0.1982	7.0	630	0.6107	0.7826
0.1799	8.0	720	0.5742	0.8132
0.1689	9.0	810	0.6021	0.7945
0.1555	10.0	900	0.5871	0.7895
0.1514	11.0	990	0.5836	0.7915
0.1418	12.0	1080	0.5684	0.7964
0.1356	13.0	1170	0.5688	0.7974
0.1262	14.0	1260	0.5783	0.8053
0.128	15.0	1350	0.5665	0.7925
0.1211	16.0	1440	0.5587	0.8014
0.1223	17.0	1530	0.5644	0.8034
0.1133	18.0	1620	0.5634	0.8083
0.1158	19.0	1710	0.5711	0.8063
0.1119	20.0	1800	0.5648	0.8083

Framework versions

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