resnet-152-fv-finetuned-memess

This model is a fine-tuned version of microsoft/resnet-152 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.6281
• Accuracy: 0.7674
• Precision: 0.7651
• Recall: 0.7674
• F1: 0.7647

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: 0.00012
• train_batch_size: 64
• eval_batch_size: 64
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 256
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
1.5902	0.99	20	1.5519	0.4938	0.3491	0.4938	0.3529
1.4694	1.99	40	1.3730	0.4892	0.4095	0.4892	0.3222
1.3129	2.99	60	1.2052	0.5301	0.3504	0.5301	0.4005
1.1831	3.99	80	1.1142	0.5587	0.4077	0.5587	0.4444
1.0581	4.99	100	0.9930	0.6012	0.5680	0.6012	0.5108
0.9464	5.99	120	0.9263	0.6507	0.6200	0.6507	0.6029
0.8581	6.99	140	0.8400	0.6917	0.6645	0.6917	0.6638
0.7739	7.99	160	0.7829	0.7087	0.6918	0.7087	0.6845
0.6762	8.99	180	0.7512	0.7318	0.7206	0.7318	0.7189
0.6162	9.99	200	0.7409	0.7264	0.7244	0.7264	0.7241
0.5546	10.99	220	0.6936	0.7465	0.7429	0.7465	0.7395
0.4633	11.99	240	0.6779	0.7473	0.7393	0.7473	0.7412
0.4373	12.99	260	0.6736	0.7573	0.7492	0.7573	0.7523
0.4074	13.99	280	0.6534	0.7566	0.7516	0.7566	0.7528
0.39	14.99	300	0.6521	0.7651	0.7603	0.7651	0.7608
0.3766	15.99	320	0.6499	0.7682	0.7607	0.7682	0.7630
0.3507	16.99	340	0.6497	0.7697	0.7686	0.7697	0.7686
0.3589	17.99	360	0.6519	0.7535	0.7485	0.7535	0.7502
0.3261	18.99	380	0.6449	0.7589	0.7597	0.7589	0.7585
0.3234	19.99	400	0.6281	0.7674	0.7651	0.7674	0.7647

Framework versions

• Transformers 4.24.0.dev0
• Pytorch 1.11.0+cu102
• Datasets 2.6.1.dev0
• Tokenizers 0.13.1