convnextv2-tiny-1k-224-finetuned-fullwear
This model is a fine-tuned version of facebook/convnextv2-tiny-1k-224 on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 0.5203
- Accuracy: 0.8403
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: 32
- eval_batch_size: 32
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 128
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 120
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy |
---|---|---|---|---|
2.4871 | 0.9756 | 10 | 2.4771 | 0.0694 |
2.4464 | 1.9512 | 20 | 2.4333 | 0.1528 |
2.3911 | 2.9268 | 30 | 2.3670 | 0.2778 |
2.3204 | 4.0 | 41 | 2.2617 | 0.3681 |
2.206 | 4.9756 | 51 | 2.1445 | 0.3958 |
2.0869 | 5.9512 | 61 | 2.0146 | 0.4444 |
1.9756 | 6.9268 | 71 | 1.8763 | 0.5139 |
1.8124 | 8.0 | 82 | 1.7422 | 0.5486 |
1.6624 | 8.9756 | 92 | 1.6629 | 0.5903 |
1.587 | 9.9512 | 102 | 1.5474 | 0.6111 |
1.4746 | 10.9268 | 112 | 1.4577 | 0.625 |
1.359 | 12.0 | 123 | 1.3055 | 0.6736 |
1.2412 | 12.9756 | 133 | 1.2241 | 0.6736 |
1.1374 | 13.9512 | 143 | 1.2003 | 0.6736 |
1.0194 | 14.9268 | 153 | 1.0233 | 0.7569 |
0.9705 | 16.0 | 164 | 0.9492 | 0.7847 |
0.8949 | 16.9756 | 174 | 0.9246 | 0.75 |
0.7959 | 17.9512 | 184 | 0.8148 | 0.7639 |
0.7491 | 18.9268 | 194 | 0.7858 | 0.7569 |
0.6783 | 20.0 | 205 | 0.8010 | 0.7569 |
0.6257 | 20.9756 | 215 | 0.7295 | 0.7847 |
0.5999 | 21.9512 | 225 | 0.6219 | 0.8333 |
0.5701 | 22.9268 | 235 | 0.5932 | 0.8403 |
0.4926 | 24.0 | 246 | 0.5970 | 0.8056 |
0.4692 | 24.9756 | 256 | 0.6298 | 0.8194 |
0.4393 | 25.9512 | 266 | 0.5857 | 0.8056 |
0.419 | 26.9268 | 276 | 0.5203 | 0.8542 |
0.3454 | 28.0 | 287 | 0.6084 | 0.8264 |
0.36 | 28.9756 | 297 | 0.5928 | 0.8264 |
0.3265 | 29.9512 | 307 | 0.5303 | 0.8403 |
0.3278 | 30.9268 | 317 | 0.6049 | 0.8194 |
0.2766 | 32.0 | 328 | 0.5656 | 0.8264 |
0.2805 | 32.9756 | 338 | 0.5003 | 0.8681 |
0.2505 | 33.9512 | 348 | 0.5412 | 0.8403 |
0.2464 | 34.9268 | 358 | 0.5410 | 0.8333 |
0.2166 | 36.0 | 369 | 0.5000 | 0.8472 |
0.2 | 36.9756 | 379 | 0.5053 | 0.8056 |
0.1914 | 37.9512 | 389 | 0.5161 | 0.8403 |
0.186 | 38.9268 | 399 | 0.4242 | 0.8681 |
0.1592 | 40.0 | 410 | 0.5059 | 0.8472 |
0.1598 | 40.9756 | 420 | 0.5143 | 0.8264 |
0.1565 | 41.9512 | 430 | 0.4703 | 0.8542 |
0.1598 | 42.9268 | 440 | 0.4384 | 0.8542 |
0.139 | 44.0 | 451 | 0.4850 | 0.8403 |
0.1137 | 44.9756 | 461 | 0.4405 | 0.8542 |
0.1158 | 45.9512 | 471 | 0.5250 | 0.8333 |
0.1192 | 46.9268 | 481 | 0.5843 | 0.8194 |
0.1271 | 48.0 | 492 | 0.4498 | 0.8611 |
0.0914 | 48.9756 | 502 | 0.5167 | 0.8264 |
0.1079 | 49.9512 | 512 | 0.4648 | 0.8681 |
0.091 | 50.9268 | 522 | 0.5321 | 0.8194 |
0.1053 | 52.0 | 533 | 0.4402 | 0.8611 |
0.0842 | 52.9756 | 543 | 0.4776 | 0.8542 |
0.0961 | 53.9512 | 553 | 0.4762 | 0.8681 |
0.0896 | 54.9268 | 563 | 0.4477 | 0.8681 |
0.0876 | 56.0 | 574 | 0.4951 | 0.8472 |
0.0855 | 56.9756 | 584 | 0.5653 | 0.8125 |
0.073 | 57.9512 | 594 | 0.5315 | 0.8472 |
0.0804 | 58.9268 | 604 | 0.5064 | 0.8681 |
0.0765 | 60.0 | 615 | 0.6316 | 0.8264 |
0.0782 | 60.9756 | 625 | 0.5733 | 0.8056 |
0.069 | 61.9512 | 635 | 0.6994 | 0.8056 |
0.0809 | 62.9268 | 645 | 0.4898 | 0.8611 |
0.0829 | 64.0 | 656 | 0.6042 | 0.8194 |
0.0735 | 64.9756 | 666 | 0.4758 | 0.8611 |
0.0763 | 65.9512 | 676 | 0.4921 | 0.8542 |
0.0565 | 66.9268 | 686 | 0.4700 | 0.8681 |
0.062 | 68.0 | 697 | 0.4944 | 0.8819 |
0.0644 | 68.9756 | 707 | 0.4733 | 0.8681 |
0.0659 | 69.9512 | 717 | 0.4703 | 0.8819 |
0.0625 | 70.9268 | 727 | 0.5075 | 0.8542 |
0.042 | 72.0 | 738 | 0.5464 | 0.8264 |
0.056 | 72.9756 | 748 | 0.5186 | 0.8333 |
0.0858 | 73.9512 | 758 | 0.5403 | 0.8264 |
0.0616 | 74.9268 | 768 | 0.5104 | 0.8472 |
0.0777 | 76.0 | 779 | 0.5516 | 0.8403 |
0.0668 | 76.9756 | 789 | 0.4918 | 0.8611 |
0.0585 | 77.9512 | 799 | 0.5692 | 0.8403 |
0.0562 | 78.9268 | 809 | 0.5734 | 0.8403 |
0.0653 | 80.0 | 820 | 0.5403 | 0.8264 |
0.0434 | 80.9756 | 830 | 0.5108 | 0.8333 |
0.0483 | 81.9512 | 840 | 0.5699 | 0.8125 |
0.0329 | 82.9268 | 850 | 0.6028 | 0.8056 |
0.0431 | 84.0 | 861 | 0.5230 | 0.8333 |
0.042 | 84.9756 | 871 | 0.5875 | 0.8194 |
0.0449 | 85.9512 | 881 | 0.5180 | 0.8611 |
0.0512 | 86.9268 | 891 | 0.5425 | 0.8194 |
0.0545 | 88.0 | 902 | 0.5690 | 0.8264 |
0.0496 | 88.9756 | 912 | 0.5619 | 0.8611 |
0.0449 | 89.9512 | 922 | 0.5626 | 0.8333 |
0.0405 | 90.9268 | 932 | 0.5267 | 0.8403 |
0.0344 | 92.0 | 943 | 0.5617 | 0.8403 |
0.0421 | 92.9756 | 953 | 0.5400 | 0.8611 |
0.0341 | 93.9512 | 963 | 0.5729 | 0.8333 |
0.0492 | 94.9268 | 973 | 0.5855 | 0.8056 |
0.0374 | 96.0 | 984 | 0.6113 | 0.8125 |
0.0375 | 96.9756 | 994 | 0.5511 | 0.8403 |
0.0373 | 97.9512 | 1004 | 0.4942 | 0.8542 |
0.0447 | 98.9268 | 1014 | 0.5031 | 0.8542 |
0.0519 | 100.0 | 1025 | 0.5349 | 0.8542 |
0.0387 | 100.9756 | 1035 | 0.5511 | 0.8542 |
0.0256 | 101.9512 | 1045 | 0.5319 | 0.8403 |
0.043 | 102.9268 | 1055 | 0.5605 | 0.8264 |
0.029 | 104.0 | 1066 | 0.5776 | 0.8403 |
0.0379 | 104.9756 | 1076 | 0.5697 | 0.8472 |
0.0445 | 105.9512 | 1086 | 0.5133 | 0.8681 |
0.0267 | 106.9268 | 1096 | 0.5076 | 0.8681 |
0.044 | 108.0 | 1107 | 0.5260 | 0.8403 |
0.0263 | 108.9756 | 1117 | 0.5101 | 0.8542 |
0.0247 | 109.9512 | 1127 | 0.4972 | 0.8542 |
0.0441 | 110.9268 | 1137 | 0.5094 | 0.8472 |
0.0263 | 112.0 | 1148 | 0.5259 | 0.8333 |
0.0247 | 112.9756 | 1158 | 0.5323 | 0.8403 |
0.0356 | 113.9512 | 1168 | 0.5275 | 0.8403 |
0.0297 | 114.9268 | 1178 | 0.5240 | 0.8333 |
0.044 | 116.0 | 1189 | 0.5201 | 0.8472 |
0.031 | 116.9756 | 1199 | 0.5203 | 0.8403 |
0.0369 | 117.0732 | 1200 | 0.5203 | 0.8403 |
Framework versions
- Transformers 4.44.0
- Pytorch 2.4.0
- Datasets 2.21.0
- Tokenizers 0.19.1
- Downloads last month
- 9
Model tree for vishalkatheriya18/convnextv2-tiny-1k-224-finetuned-fullwear
Base model
facebook/convnextv2-tiny-1k-224