beit-base-patch16-384_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: 0.7095
• Accuracy: 0.8275

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
0.5656	1.0	164	1.4241	0.2614
0.5188	2.0	328	1.2839	0.4295
0.3829	3.0	492	0.8952	0.7245
0.3272	4.0	656	0.8815	0.7104
0.2941	5.0	820	0.7334	0.8004
0.2671	6.0	984	0.7653	0.7885
0.2569	7.0	1148	0.7095	0.8275
0.2483	8.0	1312	0.7406	0.8232
0.2464	9.0	1476	0.7520	0.8178
0.2312	10.0	1640	0.7327	0.8395
0.229	11.0	1804	0.7744	0.7961
0.2326	12.0	1968	0.7301	0.8167
0.2245	13.0	2132	0.7387	0.8145
0.2291	14.0	2296	0.7217	0.8243
0.2282	15.0	2460	0.7337	0.8059
0.2236	16.0	2624	0.7246	0.8189
0.2276	17.0	2788	0.7214	0.8243
0.2178	18.0	2952	0.7251	0.8178
0.2235	19.0	3116	0.7220	0.8232
0.2175	20.0	3280	0.7220	0.8221

Framework versions

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