swinv2-tiny-patch4-window8-256-dmae-humeda-DAV61

This model is a fine-tuned version of microsoft/swinv2-tiny-patch4-window8-256 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.2449
• Accuracy: 0.9086

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: 4e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 64
• 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: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 45
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.1254	1.0	10	1.0551	0.48
0.967	2.0	20	0.9770	0.4571
0.8537	3.0	30	0.7128	0.7086
0.6489	4.0	40	0.5167	0.7371
0.567	5.0	50	0.3952	0.8
0.4724	6.0	60	0.3705	0.8571
0.4106	7.0	70	0.3076	0.8743
0.3853	8.0	80	0.3155	0.8914
0.3369	9.0	90	0.2908	0.8514
0.3345	10.0	100	0.3034	0.8457
0.3092	11.0	110	0.2575	0.88
0.3027	12.0	120	0.2486	0.8971
0.2945	13.0	130	0.2956	0.8629
0.254	14.0	140	0.2787	0.8971
0.2476	15.0	150	0.2594	0.9029
0.2058	16.0	160	0.3091	0.8971
0.2254	17.0	170	0.2778	0.8971
0.2165	18.0	180	0.3094	0.88
0.2501	19.0	190	0.2536	0.8914
0.2016	20.0	200	0.2652	0.9029
0.1791	21.0	210	0.2449	0.9086
0.1401	22.0	220	0.3205	0.8743
0.17	23.0	230	0.3446	0.8686
0.1665	24.0	240	0.3058	0.8743
0.1192	25.0	250	0.3537	0.8743
0.1249	26.0	260	0.2929	0.8743
0.143	27.0	270	0.3050	0.8743
0.1809	28.0	280	0.2964	0.8914
0.135	29.0	290	0.3047	0.8743
0.1235	30.0	300	0.3011	0.8971
0.1351	31.0	310	0.3147	0.8914
0.1124	32.0	320	0.3123	0.8857
0.1122	33.0	330	0.3289	0.8857
0.1163	34.0	340	0.3256	0.88
0.0885	35.0	350	0.3126	0.8971
0.097	36.0	360	0.3425	0.8857
0.1036	37.0	370	0.3221	0.8857
0.1107	38.0	380	0.3598	0.8971
0.0928	39.0	390	0.3355	0.8857
0.1104	40.0	400	0.3425	0.8857
0.1104	40.5128	405	0.3428	0.8857

Framework versions

• Transformers 4.48.3
• Pytorch 2.6.0+cu124
• Datasets 3.4.0
• Tokenizers 0.21.0