plant-seedlings-resnet-152

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: 0.2604
• Accuracy: 0.9147

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: 0.0002
• train_batch_size: 16
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
2.175	0.53	100	2.1135	0.3247
1.146	1.06	200	1.0761	0.6654
0.8299	1.6	300	0.7586	0.7391
0.7896	2.13	400	0.7093	0.7680
0.7327	2.66	500	0.5140	0.8207
0.5207	3.19	600	0.5375	0.8183
0.6465	3.72	700	0.4620	0.8465
0.2745	4.26	800	0.4784	0.8324
0.5366	4.79	900	0.4804	0.8355
0.4467	5.32	1000	0.4354	0.8551
0.3604	5.85	1100	0.3950	0.8680
0.2511	6.38	1200	0.4279	0.8594
0.326	6.91	1300	0.3677	0.8852
0.3444	7.45	1400	0.3539	0.8748
0.4015	7.98	1500	0.3161	0.8950
0.2821	8.51	1600	0.4394	0.8686
0.435	9.04	1700	0.3408	0.8920
0.3318	9.57	1800	0.3886	0.8778
0.2441	10.11	1900	0.2854	0.9042
0.2467	10.64	2000	0.3248	0.8883
0.2082	11.17	2100	0.3080	0.8956
0.1983	11.7	2200	0.3394	0.8963
0.2609	12.23	2300	0.3582	0.8870
0.2055	12.77	2400	0.3330	0.8963
0.3476	13.3	2500	0.2852	0.9091
0.223	13.83	2600	0.3115	0.8999
0.2307	14.36	2700	0.2986	0.9098
0.3113	14.89	2800	0.3103	0.8993
0.1792	15.43	2900	0.2862	0.9098
0.1685	15.96	3000	0.2935	0.9055
0.2429	16.49	3100	0.2882	0.9122
0.2548	17.02	3200	0.2748	0.9165
0.3561	17.55	3300	0.2684	0.9171
0.1982	18.09	3400	0.2647	0.9147
0.1638	18.62	3500	0.2509	0.9171
0.2404	19.15	3600	0.2936	0.9165
0.2424	19.68	3700	0.2604	0.9147

Framework versions

• Transformers 4.28.1
• Pytorch 2.0.0+cu118
• Datasets 2.11.0
• Tokenizers 0.13.3