resnet-101-finetuned-cassava-leaf-disease

This model is a fine-tuned version of microsoft/resnet-101 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.7399
• Accuracy: 0.7458

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: 5e-05
• train_batch_size: 240
• eval_batch_size: 240
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 960
• 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
7.1788	0.99	20	7.0002	0.0023
6.8206	1.98	40	6.5405	0.0042
6.2923	2.96	60	5.8343	0.0285
5.4621	4.0	81	4.7536	0.1883
4.3676	4.99	101	3.5878	0.3748
3.294	5.98	121	2.7221	0.4869
2.3557	6.96	141	2.0931	0.5664
1.8613	8.0	162	1.7155	0.6140
1.5385	8.99	182	1.4300	0.6486
1.334	9.98	202	1.2463	0.6701
1.1959	10.96	222	1.1332	0.6757
1.1029	12.0	243	1.0499	0.6836
1.0497	12.99	263	0.9882	0.6958
0.9612	13.98	283	0.9342	0.7028
0.9428	14.96	303	0.8967	0.7070
0.8803	16.0	324	0.8623	0.7103
0.8713	16.99	344	0.8433	0.7164
0.8607	17.98	364	0.8267	0.7201
0.8317	18.96	384	0.8053	0.7252
0.8323	20.0	405	0.7912	0.7294
0.8012	20.99	425	0.7827	0.7285
0.7957	21.98	445	0.7769	0.7271
0.7972	22.96	465	0.7634	0.7355
0.7632	24.0	486	0.7608	0.7346
0.7626	24.99	506	0.7471	0.7393
0.7752	25.98	526	0.7518	0.7444
0.7577	26.96	546	0.7411	0.7486
0.7644	28.0	567	0.7427	0.7435
0.7541	28.99	587	0.7362	0.7514
0.7737	29.63	600	0.7399	0.7458

Framework versions

• Transformers 4.37.2
• Pytorch 2.2.1
• Datasets 2.18.0
• Tokenizers 0.15.1