Hugging Face's logo

joheras
/
detr_finetuned_fruits

Object Detection
Transformers
TensorBoard
Safetensors
yolos
Generated from Trainer
Model card Files Metrics Community
detr_finetuned_fruits / README.md
joheras's picture
joheras
End of training
dd9a3bd verified
preview code
|
raw
history blame contribute delete
10.2 kB
metadata
library_name: transformers
license: apache-2.0
base_model: hustvl/yolos-tiny
tags:
  - generated_from_trainer
model-index:
  - name: detr_finetuned_fruits
    results: []

detr_finetuned_fruits

This model is a fine-tuned version of hustvl/yolos-tiny on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 0.8626
  • Map: 0.5447
  • Map 50: 0.8282
  • Map 75: 0.5821
  • Map Small: -1.0
  • Map Medium: 0.4675
  • Map Large: 0.5734
  • Mar 1: 0.4327
  • Mar 10: 0.7017
  • Mar 100: 0.7589
  • Mar Small: -1.0
  • Mar Medium: 0.6514
  • Mar Large: 0.7795
  • Map Banana: 0.4399
  • Mar 100 Banana: 0.72
  • Map Orange: 0.541
  • Mar 100 Orange: 0.7738
  • Map Apple: 0.6532
  • Mar 100 Apple: 0.7829

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: 4
  • eval_batch_size: 8
  • seed: 42
  • optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
  • 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 Banana Mar 100 Banana Map Orange Mar 100 Orange Map Apple Mar 100 Apple
No log 1.0 60 2.2905 0.008 0.0222 0.006 -1.0 0.0061 0.012 0.0871 0.1937 0.303 -1.0 0.2429 0.3256 0.0066 0.15 0.0042 0.4048 0.0133 0.3543
No log 2.0 120 1.9265 0.0202 0.0629 0.0071 -1.0 0.119 0.023 0.091 0.236 0.396 -1.0 0.3243 0.4106 0.0238 0.415 0.0256 0.4071 0.0111 0.3657
No log 3.0 180 1.8221 0.0309 0.0731 0.0214 -1.0 0.082 0.035 0.0877 0.241 0.4251 -1.0 0.4071 0.4275 0.0504 0.49 0.0302 0.4738 0.0121 0.3114
No log 4.0 240 1.7172 0.0253 0.0655 0.0111 -1.0 0.0988 0.0251 0.1424 0.258 0.4915 -1.0 0.4371 0.502 0.0303 0.5225 0.0273 0.4548 0.0183 0.4971
No log 5.0 300 1.5541 0.0472 0.1085 0.0305 -1.0 0.0639 0.0526 0.1869 0.3652 0.5653 -1.0 0.4014 0.5933 0.0326 0.535 0.0777 0.6095 0.0313 0.5514
No log 6.0 360 1.5159 0.0501 0.1145 0.0436 -1.0 0.0694 0.0556 0.2009 0.3976 0.5542 -1.0 0.38 0.5799 0.0659 0.5725 0.0527 0.5071 0.0318 0.5829
No log 7.0 420 1.4185 0.0775 0.1777 0.0662 -1.0 0.2007 0.0751 0.2078 0.4237 0.5944 -1.0 0.5071 0.6137 0.0647 0.585 0.1071 0.5952 0.0608 0.6029
No log 8.0 480 1.2902 0.0965 0.189 0.077 -1.0 0.1555 0.1161 0.2715 0.4469 0.64 -1.0 0.5186 0.66 0.0726 0.62 0.1498 0.6286 0.0673 0.6714
1.5459 9.0 540 1.2497 0.1052 0.2137 0.1115 -1.0 0.2298 0.1295 0.294 0.4625 0.6662 -1.0 0.4914 0.6987 0.0749 0.6025 0.1614 0.6905 0.0794 0.7057
1.5459 10.0 600 1.0677 0.141 0.2485 0.1427 -1.0 0.2822 0.1552 0.3656 0.5481 0.7142 -1.0 0.6257 0.7329 0.0819 0.6475 0.2168 0.7238 0.1242 0.7714
1.5459 11.0 660 1.0572 0.1813 0.3134 0.1988 -1.0 0.2859 0.2008 0.3533 0.5777 0.7017 -1.0 0.5886 0.72 0.1098 0.665 0.2983 0.7143 0.136 0.7257
1.5459 12.0 720 1.0403 0.247 0.4247 0.2529 -1.0 0.3598 0.2663 0.348 0.5748 0.7021 -1.0 0.6286 0.7157 0.1359 0.67 0.3934 0.7333 0.2115 0.7029
1.5459 13.0 780 0.9933 0.3205 0.5352 0.3708 -1.0 0.3999 0.3373 0.3908 0.6208 0.7248 -1.0 0.6086 0.7447 0.1991 0.68 0.3998 0.7429 0.3626 0.7514
1.5459 14.0 840 1.0158 0.3865 0.6502 0.4208 -1.0 0.3726 0.4172 0.3843 0.6447 0.7184 -1.0 0.5557 0.7445 0.2549 0.6875 0.4506 0.7333 0.454 0.7343
1.5459 15.0 900 0.9649 0.4519 0.6973 0.4866 -1.0 0.4641 0.4712 0.395 0.6727 0.7373 -1.0 0.6357 0.7575 0.2713 0.67 0.5052 0.7619 0.5792 0.78
1.5459 16.0 960 0.9148 0.491 0.7552 0.5358 -1.0 0.4674 0.5169 0.4167 0.6903 0.7571 -1.0 0.6686 0.7776 0.3438 0.69 0.5616 0.7786 0.5676 0.8029
0.864 17.0 1020 0.8861 0.5232 0.7871 0.571 -1.0 0.5199 0.5463 0.4387 0.6948 0.7541 -1.0 0.68 0.771 0.4007 0.7 0.5659 0.7595 0.6029 0.8029
0.864 18.0 1080 0.8914 0.5014 0.7661 0.5433 -1.0 0.4449 0.5276 0.4245 0.6954 0.7655 -1.0 0.6286 0.79 0.4006 0.715 0.4992 0.7643 0.6043 0.8171
0.864 19.0 1140 0.8886 0.5223 0.7763 0.5611 -1.0 0.4595 0.5492 0.4201 0.6893 0.7473 -1.0 0.6143 0.7716 0.4002 0.69 0.5387 0.769 0.6279 0.7829
0.864 20.0 1200 0.8973 0.5239 0.8057 0.5726 -1.0 0.4437 0.5531 0.4317 0.6917 0.7535 -1.0 0.6371 0.7758 0.4343 0.7125 0.5406 0.7738 0.5966 0.7743
0.864 21.0 1260 0.8740 0.5355 0.8126 0.5889 -1.0 0.4869 0.5605 0.4162 0.7055 0.7633 -1.0 0.6314 0.7856 0.4039 0.7375 0.5735 0.7667 0.6292 0.7857
0.864 22.0 1320 0.8917 0.5212 0.7944 0.5517 -1.0 0.4609 0.549 0.423 0.6872 0.7421 -1.0 0.61 0.7657 0.4232 0.7 0.5315 0.769 0.609 0.7571
0.864 23.0 1380 0.8508 0.5508 0.8362 0.6164 -1.0 0.4879 0.5786 0.4278 0.6983 0.753 -1.0 0.6614 0.7723 0.4453 0.71 0.5576 0.769 0.6494 0.78
0.864 24.0 1440 0.8769 0.5586 0.8358 0.6156 -1.0 0.4846 0.5886 0.4471 0.7105 0.765 -1.0 0.6586 0.787 0.4598 0.705 0.5588 0.7786 0.6572 0.8114
0.638 25.0 1500 0.8670 0.5394 0.8271 0.5786 -1.0 0.4681 0.5667 0.425 0.7004 0.7563 -1.0 0.6514 0.7771 0.4333 0.7075 0.5426 0.7786 0.6422 0.7829
0.638 26.0 1560 0.8487 0.5557 0.8355 0.6103 -1.0 0.4903 0.5829 0.4353 0.709 0.7612 -1.0 0.6586 0.7812 0.4483 0.715 0.559 0.7857 0.6596 0.7829
0.638 27.0 1620 0.8585 0.5484 0.8267 0.5888 -1.0 0.4735 0.5755 0.4318 0.7106 0.7646 -1.0 0.6586 0.7848 0.4431 0.7225 0.5435 0.7857 0.6587 0.7857
0.638 28.0 1680 0.8668 0.5479 0.8262 0.5865 -1.0 0.471 0.5762 0.4318 0.7051 0.763 -1.0 0.6586 0.7831 0.4414 0.72 0.5465 0.7833 0.6556 0.7857
0.638 29.0 1740 0.8631 0.5459 0.8282 0.5962 -1.0 0.4737 0.5737 0.4319 0.7011 0.7598 -1.0 0.6586 0.7795 0.4394 0.72 0.5405 0.7738 0.6579 0.7857
0.638 30.0 1800 0.8626 0.5447 0.8282 0.5821 -1.0 0.4675 0.5734 0.4327 0.7017 0.7589 -1.0 0.6514 0.7795 0.4399 0.72 0.541 0.7738 0.6532 0.7829

Framework versions

  • Transformers 4.49.0
  • Pytorch 2.6.0+cu124
  • Datasets 3.4.1
  • Tokenizers 0.21.1