resnet-152-finetuned_resnet152-adam-optimizere-2-autotags

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: 0.4368
• Accuracy: 0.8981

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: 0.01
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 64
• 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
1.4424	0.99	65	1.7123	0.56
1.6053	1.99	130	2.0613	0.3152
1.3795	2.99	195	1.3791	0.5552
0.9701	3.99	260	0.9195	0.7038
0.8258	4.99	325	0.9107	0.7067
0.7619	5.99	390	0.9915	0.6867
0.6241	6.99	455	0.7895	0.76
0.497	7.99	520	0.6616	0.8038
0.4709	8.99	585	0.5282	0.8543
0.394	9.99	650	0.5447	0.8429
0.343	10.99	715	0.5108	0.8486
0.3482	11.99	780	0.5224	0.8505
0.2576	12.99	845	0.4796	0.8743
0.1837	13.99	910	0.5008	0.8571
0.1904	14.99	975	0.4366	0.8790
0.1458	15.99	1040	0.4320	0.8990
0.1575	16.99	1105	0.4059	0.8952
0.0992	17.99	1170	0.4362	0.8952
0.0858	18.99	1235	0.4210	0.8971
0.0704	19.99	1300	0.4368	0.8981

Framework versions

• Transformers 4.25.1
• Pytorch 1.13.1+cu117
• Datasets 2.11.0
• Tokenizers 0.13.2