Validated_cracks_raw_dataset_266_convnext_model

This model is a fine-tuned version of facebook/convnext-base-224-22k on the Logiroad/Validated_Balanced_Raw_Dataset dataset. It achieves the following results on the evaluation set:

• Loss: 1.1049
• Accuracy: 0.5330

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: 1e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 1337
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.05
• num_epochs: 30.0
• mixed_precision_training: Native AMP
• label_smoothing_factor: 0.05

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.3904	1.0	80	1.3471	0.3538
1.3264	2.0	160	1.2974	0.3396
1.2875	3.0	240	1.2867	0.3491
1.2415	4.0	320	1.2570	0.3774
1.2209	5.0	400	1.2145	0.4340
1.1699	6.0	480	1.2139	0.4151
1.1315	7.0	560	1.1762	0.4623
1.1565	8.0	640	1.1620	0.4575
1.1111	9.0	720	1.1559	0.4811
1.117	10.0	800	1.1255	0.5
1.0174	11.0	880	1.1186	0.5047
1.0569	12.0	960	1.1092	0.5
1.0089	13.0	1040	1.1156	0.5
1.0413	14.0	1120	1.1085	0.4953
0.9958	15.0	1200	1.1157	0.5236
0.9969	16.0	1280	1.1049	0.5330
0.9918	17.0	1360	1.1045	0.5047
0.9798	18.0	1440	1.0932	0.5142
0.9232	19.0	1520	1.0961	0.5047
0.8817	20.0	1600	1.1116	0.4906
0.9587	21.0	1680	1.1000	0.5236
0.964	22.0	1760	1.0978	0.5094
0.8906	23.0	1840	1.0975	0.5236
0.896	24.0	1920	1.0996	0.5142
0.9156	25.0	2000	1.0952	0.5236
0.8797	26.0	2080	1.0960	0.5236
0.8781	27.0	2160	1.0948	0.5236
0.8698	28.0	2240	1.0946	0.5236
0.9	29.0	2320	1.0949	0.5283
0.8853	30.0	2400	1.0949	0.5283

Framework versions

• Transformers 4.46.1
• Pytorch 2.3.0
• Datasets 3.1.0
• Tokenizers 0.20.3