resnet50_fold_5_v3

This model is a fine-tuned version of microsoft/resnet-50 on the imagefolder dataset. It achieves the following results on the evaluation set:

Loss: 1.2146
Accuracy: 0.9357
F1 Score: 0.9373
Recall: 0.9363

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: 0.0001
train_batch_size: 64
eval_batch_size: 64
seed: 42
optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: cosine
lr_scheduler_warmup_steps: 150
num_epochs: 100
label_smoothing_factor: 0.15

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1 Score	Recall
2.7538	1.0	20	2.7678	0.2894	0.2194	0.2825
2.7424	2.0	40	2.7510	0.3762	0.2850	0.3534
2.7153	3.0	60	2.7262	0.4084	0.3277	0.3753
2.6618	4.0	80	2.6940	0.4534	0.3871	0.4243
2.5832	5.0	100	2.6368	0.4855	0.4228	0.4581
2.4485	6.0	120	2.5174	0.5466	0.5059	0.5200
2.2750	7.0	140	2.3666	0.6495	0.6138	0.6245
2.1366	8.0	160	2.1599	0.7010	0.6804	0.6821
1.9487	9.0	180	1.9467	0.7910	0.7905	0.7855
1.7994	10.0	200	1.7635	0.8071	0.8088	0.8103
1.5538	11.0	220	1.6424	0.8232	0.8245	0.8324
1.4558	12.0	240	1.5752	0.8360	0.8375	0.8453
1.3581	13.0	260	1.5165	0.8457	0.8475	0.8539
1.3067	14.0	280	1.4730	0.8521	0.8531	0.8573
1.2529	15.0	300	1.4482	0.8553	0.8575	0.8642
1.2485	16.0	320	1.4300	0.8650	0.8663	0.8703
1.1958	17.0	340	1.3978	0.8714	0.8730	0.8757
1.1881	18.0	360	1.3704	0.8810	0.8822	0.8866
1.2329	19.0	380	1.3466	0.8907	0.8942	0.8970
1.1605	20.0	400	1.3367	0.8939	0.8966	0.9003
1.1933	21.0	420	1.3284	0.9003	0.9029	0.9053
1.1290	22.0	440	1.3181	0.8971	0.9000	0.9009
1.1056	23.0	460	1.3145	0.9035	0.9064	0.9090
1.1452	24.0	480	1.3042	0.9003	0.9040	0.9065
1.1592	25.0	500	1.3168	0.9003	0.9022	0.9034
1.0909	26.0	520	1.2868	0.9100	0.9132	0.9159
1.1074	27.0	540	1.2757	0.9164	0.9194	0.9196
1.1338	28.0	560	1.3135	0.9003	0.9045	0.9103
1.0733	29.0	580	1.2755	0.9132	0.9161	0.9190
1.1064	30.0	600	1.2862	0.9003	0.9035	0.9084
1.1311	31.0	620	1.2648	0.9196	0.9218	0.9214
1.1482	32.0	640	1.2535	0.9164	0.9188	0.9228
1.0984	33.0	660	1.2394	0.9260	0.9279	0.9295
1.0748	34.0	680	1.2483	0.9228	0.9244	0.9271
1.0695	35.0	700	1.2609	0.9196	0.9206	0.9221
1.0756	36.0	720	1.2388	0.9293	0.9310	0.9320
1.0730	37.0	740	1.2394	0.9260	0.9280	0.9295
1.1123	38.0	760	1.2432	0.9260	0.9269	0.9270
1.0542	39.0	780	1.2382	0.9293	0.9307	0.9301
1.0598	40.0	800	1.2418	0.9196	0.9198	0.9202
1.0765	41.0	820	1.2280	0.9293	0.9297	0.9308
1.0603	42.0	840	1.2220	0.9357	0.9368	0.9357
1.0613	43.0	860	1.2254	0.9293	0.9304	0.9308
1.0632	44.0	880	1.2123	0.9357	0.9362	0.9364
1.0746	45.0	900	1.2249	0.9325	0.9341	0.9332
1.0352	46.0	920	1.2215	0.9357	0.9368	0.9357
1.0848	47.0	940	1.2147	0.9228	0.9244	0.9245
1.0298	48.0	960	1.2135	0.9357	0.9370	0.9370
1.0565	49.0	980	1.2263	0.9325	0.9342	0.9338
1.0664	50.0	1000	1.2146	0.9357	0.9373	0.9363
1.0500	51.0	1020	1.2079	0.9357	0.9362	0.9363
1.0399	52.0	1040	1.2152	0.9357	0.9363	0.9376
1.0645	53.0	1060	1.2177	0.9357	0.9367	0.9370
1.0424	54.0	1080	1.2230	0.9325	0.9334	0.9338
1.0347	55.0	1100	1.2245	0.9325	0.9334	0.9338
1.0465	56.0	1120	1.2198	0.9325	0.9331	0.9338
1.0378	57.0	1140	1.2124	0.9325	0.9331	0.9338
1.0507	58.0	1160	1.2187	0.9357	0.9370	0.9344
1.0693	59.0	1180	1.2134	0.9325	0.9337	0.9332
1.0283	60.0	1200	1.2135	0.9357	0.9370	0.9370
1.0197	61.0	1220	1.2235	0.9357	0.9372	0.9357

Framework versions

Transformers 5.0.0
Pytorch 2.10.0+cu128
Datasets 5.0.0
Tokenizers 0.22.2

nqvii/resnet50_fold_5_v3

resnet50_fold_5_v3

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions