Dogs-Breed-Image-Classification-V1

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

• Loss: 0.4469
• Accuracy: 0.8758

Model description

Link to the fine-tuned model using resnet-50

This model was trained using dataset from Kaggle - Standford dogs dataset

Quotes from the website: The Stanford Dogs dataset contains images of 120 breeds of dogs from around the world. This dataset has been built using images and annotation from ImageNet for the task of fine-grained image categorization. It was originally collected for fine-grain image categorization, a challenging problem as certain dog breeds have near identical features or differ in colour and age.

citation: Aditya Khosla, Nityananda Jayadevaprakash, Bangpeng Yao and Li Fei-Fei. Novel dataset for Fine-Grained Image Categorization. First Workshop on Fine-Grained Visual Categorization (FGVC), IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2011. [pdf] [poster] [BibTex]

Secondary: J. Deng, W. Dong, R. Socher, L.-J. Li, K. Li and L. Fei-Fei, ImageNet: A Large-Scale Hierarchical Image Database. IEEE Computer Vision and Pattern Recognition (CVPR), 2009. [pdf] [BibTex]

Intended uses & limitations

This model is fined tune solely for classifiying 120 species of dogs.

Training and evaluation data

75% training data, 25% testing data.

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 1e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	309	18.7685	0.0091
18.7211	2.0	618	18.5975	0.0091
18.7211	3.0	927	17.4087	0.0091
15.4274	4.0	1236	11.8712	0.0091
10.3252	5.0	1545	6.6642	0.0091
10.3252	6.0	1854	5.2754	0.0112
6.2268	7.0	2163	4.8454	0.0158
6.2268	8.0	2472	4.7658	0.0140
4.9682	9.0	2781	4.6860	0.0234
4.7245	10.0	3090	4.6165	0.0316
4.7245	11.0	3399	4.5349	0.0446
4.5441	12.0	3708	4.4555	0.0623
4.3912	13.0	4017	4.3437	0.0862
4.3912	14.0	4326	4.2182	0.1330
4.2211	15.0	4635	4.0752	0.2153
4.2211	16.0	4944	3.9803	0.2599
3.9762	17.0	5253	3.7347	0.3596
3.69	18.0	5562	3.5493	0.4194
3.69	19.0	5871	3.3404	0.4813
3.3803	20.0	6180	3.1122	0.5600
3.3803	21.0	6489	2.8656	0.6101
3.0345	22.0	6798	2.6544	0.6462
2.6793	23.0	7107	2.4178	0.6647
2.6793	24.0	7416	2.1967	0.7121
2.3251	25.0	7725	2.0091	0.7203
1.9975	26.0	8034	1.8189	0.7464
1.9975	27.0	8343	1.6537	0.7519
1.7009	28.0	8652	1.4413	0.7880
1.7009	29.0	8961	1.3137	0.7968
1.4494	30.0	9270	1.2150	0.7929
1.2389	31.0	9579	1.1238	0.8041
1.2389	32.0	9888	1.0215	0.8208
1.0646	33.0	10197	0.9637	0.8190
0.9319	34.0	10506	0.8891	0.8299
0.9319	35.0	10815	0.8520	0.8330
0.8297	36.0	11124	0.8212	0.8400
0.8297	37.0	11433	0.7579	0.8415
0.7293	38.0	11742	0.7254	0.8454
0.6657	39.0	12051	0.7019	0.8457
0.6657	40.0	12360	0.6669	0.8527
0.6047	41.0	12669	0.6510	0.8530
0.6047	42.0	12978	0.6264	0.8545
0.557	43.0	13287	0.6275	0.8506
0.5126	44.0	13596	0.5947	0.8536
0.5126	45.0	13905	0.5860	0.8573
0.475	46.0	14214	0.5745	0.8545
0.4406	47.0	14523	0.5579	0.8600
0.4406	48.0	14832	0.5386	0.8621
0.4086	49.0	15141	0.5346	0.8624
0.4086	50.0	15450	0.5200	0.8612
0.3882	51.0	15759	0.5233	0.8612
0.3646	52.0	16068	0.5148	0.8640
0.3646	53.0	16377	0.5078	0.8679
0.3386	54.0	16686	0.5067	0.8646
0.3386	55.0	16995	0.4976	0.8673
0.3208	56.0	17304	0.4934	0.8682
0.3039	57.0	17613	0.4849	0.8688
0.3039	58.0	17922	0.4930	0.8691
0.2915	59.0	18231	0.4867	0.8655
0.2784	60.0	18540	0.4832	0.8679
0.2784	61.0	18849	0.4785	0.8670
0.2597	62.0	19158	0.4753	0.8685
0.2597	63.0	19467	0.4701	0.8712
0.2488	64.0	19776	0.4766	0.8697
0.2426	65.0	20085	0.4726	0.8700
0.2426	66.0	20394	0.4670	0.8694
0.2261	67.0	20703	0.4624	0.8722
0.2252	68.0	21012	0.4631	0.8718
0.2252	69.0	21321	0.4702	0.8670
0.2116	70.0	21630	0.4629	0.8715
0.2116	71.0	21939	0.4650	0.8685
0.2032	72.0	22248	0.4670	0.8673
0.2035	73.0	22557	0.4565	0.8670
0.2035	74.0	22866	0.4550	0.8697
0.19	75.0	23175	0.4544	0.8706
0.19	76.0	23484	0.4483	0.8670
0.1833	77.0	23793	0.4650	0.8694
0.184	78.0	24102	0.4604	0.8709
0.184	79.0	24411	0.4484	0.8697
0.1728	80.0	24720	0.4469	0.8758
0.1688	81.0	25029	0.4536	0.8676
0.1688	82.0	25338	0.4450	0.8709
0.1674	83.0	25647	0.4530	0.8691
0.1674	84.0	25956	0.4532	0.8725
0.1632	85.0	26265	0.4495	0.8718
0.1605	86.0	26574	0.4440	0.8673
0.1605	87.0	26883	0.4504	0.8731
0.1586	88.0	27192	0.4551	0.8667
0.1558	89.0	27501	0.4498	0.8670
0.1558	90.0	27810	0.4516	0.8718
0.1587	91.0	28119	0.4450	0.8725
0.1587	92.0	28428	0.4435	0.8706
0.1505	93.0	28737	0.4459	0.8722
0.1492	94.0	29046	0.4578	0.8673
0.1492	95.0	29355	0.4499	0.8725
0.1459	96.0	29664	0.4494	0.8703
0.1459	97.0	29973	0.4533	0.8697
0.1481	98.0	30282	0.4524	0.8652
0.1477	99.0	30591	0.4496	0.8715
0.1477	100.0	30900	0.4523	0.8661

Framework versions

• Transformers 4.37.2
• Pytorch 2.3.0
• Datasets 2.15.0
• Tokenizers 0.15.1