LynnKukunda's picture
End of training
c393348 verified
metadata
license: apache-2.0
base_model: microsoft/conditional-detr-resnet-50
tags:
  - generated_from_trainer
datasets:
  - dsi
model-index:
  - name: detr_finetunned_ocular
    results: []

detr_finetunned_ocular

This model is a fine-tuned version of microsoft/conditional-detr-resnet-50 on the dsi dataset. It achieves the following results on the evaluation set:

  • Loss: 1.0598
  • Map: 0.3166
  • Map 50: 0.5255
  • Map 75: 0.3725
  • Map Small: 0.3115
  • Map Medium: 0.6744
  • Map Large: -1.0
  • Mar 1: 0.1043
  • Mar 10: 0.3801
  • Mar 100: 0.4224
  • Mar Small: 0.4186
  • Mar Medium: 0.7234
  • Mar Large: -1.0
  • Map Falciparum Trophozoite: 0.0341
  • Mar 100 Falciparum Trophozoite: 0.1663
  • Map Wbc: 0.599
  • Mar 100 Wbc: 0.6785

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: 8
  • eval_batch_size: 8
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: cosine
  • num_epochs: 30

Training results

Training Loss Epoch Step Validation Loss Map Map 50 Map 75 Map Small Map Medium Map Large Mar 1 Mar 10 Mar 100 Mar Small Mar Medium Mar Large Map Falciparum Trophozoite Mar 100 Falciparum Trophozoite Map Wbc Mar 100 Wbc
No log 1.0 86 1.1493 0.2788 0.5002 0.3024 0.274 0.6184 -1.0 0.0918 0.3473 0.386 0.3823 0.6785 -1.0 0.0196 0.1372 0.5381 0.6348
No log 2.0 172 1.1199 0.2924 0.5063 0.3371 0.2866 0.6545 -1.0 0.0924 0.3509 0.3873 0.3805 0.729 -1.0 0.0204 0.1264 0.5644 0.6483
No log 3.0 258 1.1616 0.2802 0.4941 0.3138 0.2746 0.6231 -1.0 0.0891 0.3378 0.377 0.374 0.6598 -1.0 0.0167 0.129 0.5438 0.6249
No log 4.0 344 1.1263 0.3014 0.517 0.3393 0.296 0.6609 -1.0 0.0981 0.3588 0.3857 0.3806 0.7187 -1.0 0.0258 0.1135 0.577 0.6579
No log 5.0 430 1.1219 0.2801 0.5117 0.297 0.2734 0.6555 -1.0 0.0905 0.3458 0.3795 0.3737 0.7028 -1.0 0.0218 0.1254 0.5385 0.6337
1.0831 6.0 516 1.1299 0.2646 0.485 0.2705 0.2581 0.6103 -1.0 0.0885 0.3294 0.371 0.3636 0.7 -1.0 0.0117 0.1288 0.5175 0.6131
1.0831 7.0 602 1.1003 0.2934 0.5064 0.3254 0.286 0.6706 -1.0 0.0933 0.357 0.3962 0.3905 0.7206 -1.0 0.0207 0.1397 0.5661 0.6528
1.0831 8.0 688 1.1063 0.2945 0.5028 0.3407 0.2871 0.6606 -1.0 0.0946 0.3568 0.3975 0.3938 0.6925 -1.0 0.0243 0.1454 0.5647 0.6495
1.0831 9.0 774 1.1364 0.2824 0.4979 0.3114 0.2774 0.622 -1.0 0.0928 0.3445 0.3844 0.3818 0.671 -1.0 0.017 0.1297 0.5479 0.6392
1.0831 10.0 860 1.0997 0.2904 0.501 0.3299 0.2841 0.6483 -1.0 0.0908 0.3515 0.3917 0.387 0.7065 -1.0 0.02 0.1329 0.5609 0.6505
1.0831 11.0 946 1.1198 0.2826 0.496 0.3186 0.277 0.6299 -1.0 0.0915 0.3426 0.3822 0.3778 0.6832 -1.0 0.0225 0.1342 0.5427 0.6303
1.0585 12.0 1032 1.0999 0.2921 0.5038 0.3196 0.2867 0.6334 -1.0 0.0953 0.3556 0.3954 0.3916 0.6897 -1.0 0.0244 0.1454 0.5599 0.6453
1.0585 13.0 1118 1.1097 0.2966 0.5183 0.3365 0.29 0.6667 -1.0 0.0995 0.3549 0.3983 0.3923 0.7178 -1.0 0.0297 0.1493 0.5636 0.6472
1.0585 14.0 1204 1.0932 0.2964 0.5113 0.335 0.2913 0.6494 -1.0 0.0969 0.3556 0.396 0.391 0.7037 -1.0 0.0279 0.1474 0.565 0.6447
1.0585 15.0 1290 1.0951 0.2958 0.5173 0.3287 0.2915 0.6321 -1.0 0.0969 0.3596 0.4018 0.3962 0.7187 -1.0 0.0321 0.1505 0.5595 0.653
1.0585 16.0 1376 1.1036 0.3048 0.5215 0.3482 0.2997 0.6588 -1.0 0.102 0.3633 0.4029 0.3974 0.7234 -1.0 0.0321 0.1481 0.5775 0.6578
1.0585 17.0 1462 1.0973 0.2997 0.5169 0.3437 0.2943 0.6445 -1.0 0.1006 0.3589 0.3994 0.3963 0.6907 -1.0 0.0306 0.1448 0.5688 0.654
0.968 18.0 1548 1.1322 0.3029 0.5193 0.3482 0.2973 0.6525 -1.0 0.0983 0.3636 0.4015 0.3977 0.6991 -1.0 0.032 0.1499 0.5738 0.6532
0.968 19.0 1634 1.0698 0.3049 0.5114 0.3471 0.2989 0.6757 -1.0 0.0992 0.3665 0.4138 0.4084 0.729 -1.0 0.0288 0.1634 0.581 0.6643
0.968 20.0 1720 1.0780 0.3093 0.516 0.3556 0.3036 0.6647 -1.0 0.101 0.3694 0.4145 0.4096 0.7252 -1.0 0.0299 0.1618 0.5887 0.6673
0.968 21.0 1806 1.0825 0.3044 0.522 0.3357 0.2981 0.6642 -1.0 0.0982 0.3653 0.4071 0.4029 0.7075 -1.0 0.0319 0.1564 0.5768 0.6578
0.968 22.0 1892 1.0660 0.3142 0.5195 0.3691 0.3096 0.6679 -1.0 0.1028 0.3764 0.4195 0.4158 0.7187 -1.0 0.0352 0.164 0.5933 0.675
0.968 23.0 1978 1.0604 0.3145 0.5256 0.3633 0.3093 0.674 -1.0 0.1031 0.3774 0.4199 0.4152 0.729 -1.0 0.0368 0.1669 0.5922 0.6729
0.9092 24.0 2064 1.0607 0.3168 0.5266 0.3768 0.3114 0.6848 -1.0 0.1039 0.3785 0.4233 0.4186 0.7374 -1.0 0.034 0.1654 0.5996 0.6812
0.9092 25.0 2150 1.0681 0.3163 0.5283 0.3656 0.3113 0.6751 -1.0 0.1053 0.3769 0.4185 0.4148 0.7196 -1.0 0.0352 0.1616 0.5975 0.6755
0.9092 26.0 2236 1.0641 0.3158 0.5239 0.3708 0.3106 0.6715 -1.0 0.1045 0.378 0.4217 0.4181 0.7196 -1.0 0.0339 0.1656 0.5977 0.6777
0.9092 27.0 2322 1.0644 0.3162 0.526 0.3721 0.311 0.6785 -1.0 0.1035 0.3775 0.42 0.4164 0.7206 -1.0 0.0336 0.1624 0.5988 0.6777
0.9092 28.0 2408 1.0606 0.3165 0.5241 0.374 0.3114 0.6784 -1.0 0.1052 0.3794 0.4223 0.4184 0.7252 -1.0 0.0343 0.1665 0.5988 0.6782
0.9092 29.0 2494 1.0600 0.3161 0.5249 0.3728 0.311 0.6744 -1.0 0.1043 0.3795 0.4219 0.418 0.7234 -1.0 0.0341 0.1661 0.5981 0.6777
0.8509 30.0 2580 1.0598 0.3166 0.5255 0.3725 0.3115 0.6744 -1.0 0.1043 0.3801 0.4224 0.4186 0.7234 -1.0 0.0341 0.1663 0.599 0.6785

Framework versions

  • Transformers 4.42.3
  • Pytorch 2.3.1+cu121
  • Datasets 2.20.0
  • Tokenizers 0.19.1