resnet-50-finetuned-omars1

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: 1.2536
• Accuracy: 0.6667

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: 0.0005
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.3877	1.0	11	1.3919	0.2564
1.383	2.0	22	1.3813	0.3077
1.366	3.0	33	1.3663	0.3077
1.348	4.0	44	1.3393	0.4103
1.3034	5.0	55	1.2699	0.5641
1.2227	6.0	66	1.1615	0.6154
1.0912	7.0	77	1.1262	0.6154
0.9553	8.0	88	1.1313	0.5897
0.8801	9.0	99	1.1711	0.6667
0.8017	10.0	110	1.0136	0.6667
0.7451	11.0	121	0.9310	0.6923
0.6817	12.0	132	0.8635	0.6667
0.6579	13.0	143	1.1545	0.6667
0.6357	14.0	154	0.9239	0.6154
0.6006	15.0	165	1.0271	0.6667
0.5551	16.0	176	1.1781	0.5897
0.5619	17.0	187	1.1831	0.6923
0.5359	18.0	198	0.9667	0.6667
0.5247	19.0	209	1.1237	0.6667
0.5134	20.0	220	1.1176	0.6410
0.4469	21.0	231	0.9955	0.7179
0.4908	22.0	242	1.1411	0.7179
0.4112	23.0	253	1.2766	0.6410
0.4225	24.0	264	1.1135	0.6923
0.4786	25.0	275	1.2243	0.7179
0.3908	26.0	286	1.1587	0.7179
0.4706	27.0	297	1.2236	0.6923
0.502	28.0	308	1.1733	0.7179
0.4514	29.0	319	1.0931	0.7436
0.4386	30.0	330	1.2536	0.6667

Framework versions

• Transformers 4.30.2
• Pytorch 2.0.1+cu117
• Datasets 2.13.0
• Tokenizers 0.13.3