Edit model card

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: 1.2291

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_ratio: 0.05
  • num_epochs: 100
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss
137.9231 1.0 17 125.8028
84.2349 2.0 34 40.2665
19.7645 3.0 51 7.3174
5.1905 4.0 68 3.5275
3.2078 5.0 85 2.6258
2.5905 6.0 102 2.2457
2.3692 7.0 119 2.3092
2.2272 8.0 136 1.9838
2.1079 9.0 153 1.8935
2.0051 10.0 170 1.7939
1.8845 11.0 187 1.7864
1.8544 12.0 204 1.6802
1.7695 13.0 221 1.5136
1.6797 14.0 238 1.5392
1.5638 15.0 255 1.4511
1.5424 16.0 272 1.4155
1.4715 17.0 289 1.4013
1.4396 18.0 306 1.3514
1.4471 19.0 323 1.3247
1.3963 20.0 340 1.2975
1.3261 21.0 357 1.2779
1.2633 22.0 374 1.3002
1.2494 23.0 391 1.2980
1.2512 24.0 408 1.2877
1.2433 25.0 425 1.2777
1.2116 26.0 442 1.2703
1.2489 27.0 459 1.2643
1.2316 28.0 476 1.2370
1.1575 29.0 493 1.2619
1.1525 30.0 510 1.1912
1.1134 31.0 527 1.2454
1.1722 32.0 544 1.2335
1.1395 33.0 561 1.2095
1.1626 34.0 578 1.2022
1.0886 35.0 595 1.1997
1.0653 36.0 612 1.2023
1.0653 37.0 629 1.1955
1.0421 38.0 646 1.2127
1.0476 39.0 663 1.1780
1.0415 40.0 680 1.2002
1.0107 41.0 697 1.1691
0.9861 42.0 714 1.2002
1.0084 43.0 731 1.1759
1.0171 44.0 748 1.1726
0.9977 45.0 765 1.1668
0.9553 46.0 782 1.2050
0.9872 47.0 799 1.1835
0.9529 48.0 816 1.2004
0.9669 49.0 833 1.1595
0.9669 50.0 850 1.1474
0.9218 51.0 867 1.1900
0.9517 52.0 884 1.1845
0.9665 53.0 901 1.1831
0.8924 54.0 918 1.2106
0.9008 55.0 935 1.1995
0.8584 56.0 952 1.2195
0.9453 57.0 969 1.2342
0.915 58.0 986 1.2210
0.9046 59.0 1003 1.2073
0.8851 60.0 1020 1.1895
0.9139 61.0 1037 1.2091
0.8388 62.0 1054 1.2165
0.8799 63.0 1071 1.2143
0.8534 64.0 1088 1.2154
0.8412 65.0 1105 1.2134
0.8305 66.0 1122 1.2277
0.8485 67.0 1139 1.2750
0.8177 68.0 1156 1.2791
0.8266 69.0 1173 1.2348
0.8448 70.0 1190 1.2564
0.8392 71.0 1207 1.2211
0.8007 72.0 1224 1.2272
0.8609 73.0 1241 1.2467
0.838 74.0 1258 1.2389
0.8095 75.0 1275 1.2609
0.7726 76.0 1292 1.2399
0.7956 77.0 1309 1.2718
0.8253 78.0 1326 1.2352
0.7387 79.0 1343 1.2423
0.8324 80.0 1360 1.2692
0.77 81.0 1377 1.2384
0.8133 82.0 1394 1.2354
0.7822 83.0 1411 1.2231
0.7512 84.0 1428 1.2178
0.7264 85.0 1445 1.2305
0.7814 86.0 1462 1.2202
0.7954 87.0 1479 1.2102
0.7679 88.0 1496 1.2158
0.7577 89.0 1513 1.2429
0.7122 90.0 1530 1.2462
0.7859 91.0 1547 1.2340
0.7276 92.0 1564 1.2447
0.7651 93.0 1581 1.2286
0.743 94.0 1598 1.2402
0.7813 95.0 1615 1.2276
0.6978 96.0 1632 1.2364
0.6845 97.0 1649 1.2392
0.7037 98.0 1666 1.2367
0.7468 99.0 1683 1.2323
0.7241 100.0 1700 1.2291

Framework versions

  • Transformers 4.44.0
  • Pytorch 2.2.0+cu121
  • Datasets 2.20.0
  • Tokenizers 0.19.1
Downloads last month
10
Safetensors
Model size
43.5M params
Tensor type
F32
ยท
Inference API
Unable to determine this model's library. Check the docs .

Model tree for mrdbourke/detr_finetuned_trashify_box_detector

Finetuned
(45)
this model

Space using mrdbourke/detr_finetuned_trashify_box_detector 1