mrdbourke's picture
trashify-v1
730b969 verified
|
raw
history blame
6.5 kB
metadata
license: apache-2.0
base_model: microsoft/conditional-detr-resnet-50
tags:
  - generated_from_trainer
model-index:
  - name: detr_finetuned_trashify_box_detector
    results: []

detr_finetuned_trashify_box_detector

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

  • Loss: 1.2103

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: 0.0001
  • 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
  • lr_scheduler_warmup_steps: 50
  • num_epochs: 100
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss
59.9289 1.0 9 59.2281
54.8011 2.0 18 43.7291
34.6414 3.0 27 22.8508
15.8995 4.0 36 8.1439
6.2262 5.0 45 3.7740
3.6279 6.0 54 2.9067
3.0063 7.0 63 2.9085
2.9371 8.0 72 2.5180
2.6431 9.0 81 2.3684
2.6074 10.0 90 2.2390
2.4797 11.0 99 2.0779
2.4703 12.0 108 2.1778
2.484 13.0 117 2.1163
2.3445 14.0 126 2.0291
2.3304 15.0 135 1.9298
2.3748 16.0 144 2.0053
2.2942 17.0 153 1.9890
2.2322 18.0 162 1.9092
2.2244 19.0 171 1.8906
2.1804 20.0 180 1.8215
2.1202 21.0 189 1.8159
2.1076 22.0 198 1.7617
2.0875 23.0 207 1.7749
2.0771 24.0 216 1.7343
2.0842 25.0 225 1.7500
1.9582 26.0 234 1.7164
1.9865 27.0 243 1.6550
1.9712 28.0 252 1.6415
1.9536 29.0 261 1.5604
1.8901 30.0 270 1.5556
1.8435 31.0 279 1.5524
1.8589 32.0 288 1.5356
1.848 33.0 297 1.4709
1.8287 34.0 306 1.5041
1.7871 35.0 315 1.4433
1.7822 36.0 324 1.4514
1.7421 37.0 333 1.3894
1.701 38.0 342 1.4114
1.6385 39.0 351 1.3454
1.6397 40.0 360 1.3693
1.6628 41.0 369 1.4018
1.6068 42.0 378 1.3678
1.6293 43.0 387 1.4055
1.655 44.0 396 1.3835
1.5538 45.0 405 1.3675
1.5304 46.0 414 1.3579
1.564 47.0 423 1.3538
1.5554 48.0 432 1.3381
1.5507 49.0 441 1.2798
1.5419 50.0 450 1.2863
1.5634 51.0 459 1.2955
1.5321 52.0 468 1.2939
1.5152 53.0 477 1.2924
1.5063 54.0 486 1.2978
1.4461 55.0 495 1.2925
1.3762 56.0 504 1.2707
1.4025 57.0 513 1.2640
1.4257 58.0 522 1.2618
1.4425 59.0 531 1.2675
1.3617 60.0 540 1.2676
1.4396 61.0 549 1.2623
1.3183 62.0 558 1.2347
1.3449 63.0 567 1.2279
1.3784 64.0 576 1.2135
1.3639 65.0 585 1.2468
1.3633 66.0 594 1.2355
1.2905 67.0 603 1.2114
1.3364 68.0 612 1.2301
1.2687 69.0 621 1.2220
1.2859 70.0 630 1.2227
1.2576 71.0 639 1.2194
1.3423 72.0 648 1.2285
1.2954 73.0 657 1.2470
1.3408 74.0 666 1.2514
1.2383 75.0 675 1.2120
1.2192 76.0 684 1.2318
1.221 77.0 693 1.2252
1.2352 78.0 702 1.2074
1.2214 79.0 711 1.2175
1.1694 80.0 720 1.2202
1.2178 81.0 729 1.2277
1.1687 82.0 738 1.2280
1.1424 83.0 747 1.2266
1.1668 84.0 756 1.2227
1.1986 85.0 765 1.2313
1.1763 86.0 774 1.2127
1.1623 87.0 783 1.2150
1.1303 88.0 792 1.2078
1.1504 89.0 801 1.2033
1.1061 90.0 810 1.2065
1.1025 91.0 819 1.2076
1.0971 92.0 828 1.2012
1.137 93.0 837 1.2228
1.1341 94.0 846 1.2145
1.0625 95.0 855 1.2141
1.1355 96.0 864 1.2167
1.1636 97.0 873 1.2121
1.1306 98.0 882 1.2106
1.1104 99.0 891 1.2111
1.1377 100.0 900 1.2103

Framework versions

  • Transformers 4.44.0
  • Pytorch 2.2.0+cu121
  • Datasets 2.20.0
  • Tokenizers 0.19.1