swin-base-patch4-window7-224_rice-leaf-disease-augmented-v4_v5_pft

This model is a fine-tuned version of microsoft/swin-base-patch4-window7-224 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.4024
• Accuracy: 0.8490

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: 0.0003
• train_batch_size: 64
• eval_batch_size: 64
• seed: 42
• 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_with_restarts
• lr_scheduler_warmup_steps: 256
• num_epochs: 30
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
2.0846	0.5	64	1.9602	0.2483
1.7504	1.0	128	1.5308	0.5034
1.3704	1.5	192	1.1825	0.6107
1.113	2.0	256	0.9313	0.7148
0.9305	2.5	320	0.8132	0.7617
0.8171	3.0	384	0.7214	0.7651
0.7497	3.5	448	0.6650	0.7785
0.7039	4.0	512	0.6244	0.8188
0.6696	4.5	576	0.6003	0.8188
0.649	5.0	640	0.5976	0.8121
0.6334	5.5	704	0.6032	0.8020
0.6256	6.0	768	0.5859	0.8188
0.6417	6.5	832	0.5851	0.8188
0.5991	7.0	896	0.5835	0.8154
0.6014	7.5	960	0.5394	0.8322
0.5614	8.0	1024	0.5211	0.8356
0.536	8.5	1088	0.5184	0.8121
0.5443	9.0	1152	0.5256	0.8154
0.5129	9.5	1216	0.5026	0.8221
0.5084	10.0	1280	0.5028	0.8188
0.5081	10.5	1344	0.4996	0.8188
0.4936	11.0	1408	0.5004	0.8188
0.5	11.5	1472	0.5091	0.8121
0.4934	12.0	1536	0.4892	0.8356
0.4831	12.5	1600	0.4736	0.8322
0.4638	13.0	1664	0.4727	0.8255
0.4549	13.5	1728	0.4552	0.8456
0.4454	14.0	1792	0.4646	0.8322
0.44	14.5	1856	0.4610	0.8322
0.4304	15.0	1920	0.4574	0.8356
0.4255	15.5	1984	0.4550	0.8356
0.4353	16.0	2048	0.4548	0.8356
0.4456	16.5	2112	0.4465	0.8322
0.4047	17.0	2176	0.4619	0.8255
0.4119	17.5	2240	0.4497	0.8389
0.4009	18.0	2304	0.4329	0.8423
0.3901	18.5	2368	0.4286	0.8456
0.3936	19.0	2432	0.4318	0.8456
0.3761	19.5	2496	0.4297	0.8456
0.3885	20.0	2560	0.4279	0.8456
0.3806	20.5	2624	0.4271	0.8456
0.3779	21.0	2688	0.4352	0.8523
0.3746	21.5	2752	0.4256	0.8490
0.3708	22.0	2816	0.4253	0.8557
0.362	22.5	2880	0.4205	0.8490
0.3558	23.0	2944	0.4122	0.8490
0.3507	23.5	3008	0.4147	0.8423
0.3481	24.0	3072	0.4134	0.8456
0.3452	24.5	3136	0.4120	0.8456
0.3437	25.0	3200	0.4117	0.8490
0.3499	25.5	3264	0.4164	0.8523
0.3373	26.0	3328	0.4109	0.8490
0.3468	26.5	3392	0.3999	0.8523
0.3297	27.0	3456	0.4079	0.8523
0.329	27.5	3520	0.3997	0.8423
0.3293	28.0	3584	0.4051	0.8423
0.3147	28.5	3648	0.3987	0.8523
0.3239	29.0	3712	0.4013	0.8523
0.3147	29.5	3776	0.4031	0.8490
0.3167	30.0	3840	0.4024	0.8490

Framework versions

• Transformers 4.48.3
• Pytorch 2.5.1+cu124
• Datasets 3.3.2
• Tokenizers 0.21.1