VOOZH about

URL: https://huggingface.co/merve/rfdetr-docvqa-qwen

⇱ merve/rfdetr-docvqa-qwen · Hugging Face

rfdetr-docvqa-qwen

This model is a fine-tuned version of Roboflow/rf-detr-base on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 13.0586
  • Map: 0.0368
  • Map 50: 0.053
  • Map 75: 0.043
  • Map Small: -1.0
  • Map Medium: 0.0361
  • Map Large: 0.0382
  • Mar 1: 0.0457
  • Mar 10: 0.1721
  • Mar 100: 0.2857
  • Mar Small: -1.0
  • Mar Medium: 0.35
  • Mar Large: 0.2867
  • Map Chart: 0.0
  • Mar 100 Chart: 0.0
  • Map Diagram: 0.0
  • Mar 100 Diagram: 0.0
  • Map Figure: 0.0
  • Mar 100 Figure: 0.0
  • Map Image: 0.1529
  • Mar 100 Image: 0.7949
  • Map Table: 0.031
  • Mar 100 Table: 0.6337

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: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
  • lr_scheduler_type: cosine
  • lr_scheduler_warmup_steps: 0.05
  • num_epochs: 10

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 Chart Mar 100 Chart Map Diagram Mar 100 Diagram Map Figure Mar 100 Figure Map Image Mar 100 Image Map Table Mar 100 Table
24.7458 1.0 69 21.2345 0.0008 0.0029 0.0001 -1.0 0.0 0.0008 0.0005 0.0168 0.0686 -1.0 0.0 0.0686 0.0001 0.0824 0.0 0.0 0.0 0.0 0.0 0.0 0.004 0.2605
13.3342 2.0 138 18.1056 0.0011 0.0033 0.0005 -1.0 0.0 0.0011 0.0063 0.0302 0.0661 -1.0 0.0 0.0662 0.0 0.0 0.0 0.0 0.0 0.0 0.0001 0.0085 0.0055 0.3221
12.3492 3.0 207 17.8376 0.0023 0.007 0.0008 -1.0 0.0 0.0023 0.0105 0.0414 0.1012 -1.0 0.0 0.103 0.0 0.0 0.0 0.0 0.0 0.0 0.0011 0.1712 0.0102 0.3349
11.0428 4.0 276 17.4588 0.0072 0.0227 0.0017 -1.0 0.0 0.0072 0.0154 0.0484 0.14 -1.0 0.0 0.1434 0.0 0.0 0.0 0.0 0.0 0.0 0.0058 0.322 0.03 0.3779
10.6341 5.0 345 15.4127 0.006 0.0164 0.0039 -1.0 0.0 0.0062 0.0115 0.0753 0.2016 -1.0 0.0 0.2077 0.0 0.0 0.0 0.0 0.0 0.0 0.0149 0.5695 0.0154 0.4384
10.1076 6.0 414 14.4452 0.0158 0.0308 0.0141 -1.0 0.0 0.0162 0.0307 0.1094 0.2111 -1.0 0.0 0.2175 0.0 0.0 0.0 0.0 0.0 0.0 0.0337 0.5983 0.0452 0.457
9.5315 7.0 483 13.9064 0.0165 0.0368 0.0108 -1.0 0.0 0.0171 0.024 0.1238 0.2374 -1.0 0.0 0.2445 0.0 0.0 0.0 0.0 0.0 0.0 0.0553 0.6661 0.0273 0.5209
8.9901 8.0 552 13.0197 0.0342 0.0613 0.0326 -1.0 0.0087 0.0355 0.0424 0.1634 0.2603 -1.0 0.1 0.266 0.0 0.0 0.0 0.0 0.0 0.0 0.1215 0.7305 0.0496 0.5709
9.2235 9.0 621 13.0330 0.0372 0.0558 0.0442 -1.0 0.0308 0.0389 0.0471 0.1659 0.2797 -1.0 0.35 0.2807 0.0 0.0 0.0 0.0 0.0 0.0 0.1592 0.7881 0.0269 0.6105
8.9338 10.0 690 13.0586 0.0368 0.053 0.043 -1.0 0.0361 0.0382 0.0457 0.1721 0.2857 -1.0 0.35 0.2867 0.0 0.0 0.0 0.0 0.0 0.0 0.1529 0.7949 0.031 0.6337

Framework versions

  • Transformers 5.12.1
  • Pytorch 2.12.0+cu130
  • Datasets 5.0.0
  • Tokenizers 0.22.2
Downloads last month
109
Safetensors
Model size
31.9M params
Tensor type
F32
·

Model tree for merve/rfdetr-docvqa-qwen

Finetuned
(2)
this model