Lit4pCol4b/mit-b1_segformer_ADE20k_RGB_IS_v1

This model is a fine-tuned version of nvidia/mit-b1 on an unknown dataset. It achieves the following results on the evaluation set:

  • Train Loss: 0.0787
  • Validation Loss: 0.1007
  • Validation Mean Iou: 0.7646
  • Validation Mean Accuracy: 0.8701
  • Validation Overall Accuracy: 0.9687
  • Validation Accuracy Unlabeled: 0.6791
  • Validation Accuracy Objeto Interes: 0.9475
  • Validation Accuracy Agua: 0.9838
  • Validation Iou Unlabeled: 0.5173
  • Validation Iou Objeto Interes: 0.8005
  • Validation Iou Agua: 0.9760
  • Epoch: 39

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:

  • optimizer: {'name': 'Adam', 'weight_decay': None, 'clipnorm': None, 'global_clipnorm': None, 'clipvalue': None, 'use_ema': False, 'ema_momentum': 0.99, 'ema_overwrite_frequency': None, 'jit_compile': True, 'is_legacy_optimizer': False, 'learning_rate': 6e-05, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-07, 'amsgrad': False}
  • training_precision: float32

Training results

Train Loss Validation Loss Validation Mean Iou Validation Mean Accuracy Validation Overall Accuracy Validation Accuracy Unlabeled Validation Accuracy Objeto Interes Validation Accuracy Agua Validation Iou Unlabeled Validation Iou Objeto Interes Validation Iou Agua Epoch
0.6580 0.6718 0.4196 0.6281 0.8767 0.0254 0.9466 0.9124 0.0232 0.3345 0.9012 0
0.4551 0.5040 0.5131 0.6832 0.9126 0.1813 0.9216 0.9467 0.1234 0.4837 0.9322 1
0.3472 0.2565 0.5381 0.6560 0.9375 0.1035 0.8839 0.9805 0.0930 0.5671 0.9542 2
0.2846 0.2434 0.6188 0.7343 0.9442 0.3415 0.8847 0.9767 0.2514 0.6486 0.9564 3
0.2383 0.2245 0.6401 0.7568 0.9469 0.4203 0.8735 0.9767 0.2975 0.6644 0.9586 4
0.2075 0.2243 0.6606 0.7809 0.9501 0.4690 0.8975 0.9764 0.3332 0.6879 0.9608 5
0.1943 0.1820 0.6721 0.7704 0.9559 0.4301 0.8964 0.9847 0.3423 0.7083 0.9658 6
0.1835 0.2237 0.6866 0.8243 0.9510 0.5945 0.9077 0.9707 0.3844 0.7151 0.9601 7
0.1645 0.1638 0.7110 0.8204 0.9586 0.6026 0.8779 0.9808 0.4292 0.7369 0.9670 8
0.1574 0.1359 0.7140 0.8058 0.9616 0.5380 0.8933 0.9861 0.4197 0.7527 0.9695 9
0.1737 0.1421 0.7075 0.8042 0.9596 0.5320 0.8965 0.9841 0.4037 0.7513 0.9675 10
0.1608 0.1613 0.7046 0.8348 0.9564 0.6010 0.9285 0.9750 0.4156 0.7325 0.9655 11
0.1425 0.1387 0.7268 0.8355 0.9618 0.6140 0.9109 0.9816 0.4499 0.7605 0.9698 12
0.1299 0.1230 0.7198 0.8184 0.9628 0.5475 0.9225 0.9851 0.4286 0.7595 0.9714 13
0.1286 0.1279 0.7267 0.8320 0.9630 0.5856 0.9270 0.9833 0.4473 0.7614 0.9715 14
0.1322 0.1201 0.7428 0.8380 0.9651 0.6334 0.8954 0.9854 0.4772 0.7791 0.9722 15
0.1203 0.1076 0.7439 0.8294 0.9663 0.6001 0.9000 0.9880 0.4712 0.7872 0.9732 16
0.1154 0.1314 0.7417 0.8557 0.9633 0.6671 0.9198 0.9802 0.4752 0.7794 0.9706 17
0.1145 0.1098 0.7446 0.8438 0.9662 0.6183 0.9281 0.9852 0.4827 0.7770 0.9739 18
0.1131 0.0994 0.7500 0.8368 0.9676 0.6077 0.9145 0.9881 0.4834 0.7919 0.9748 19
0.1101 0.1157 0.7590 0.8657 0.9664 0.7130 0.9015 0.9827 0.5107 0.7928 0.9733 20
0.1045 0.1099 0.7513 0.8565 0.9664 0.6570 0.9288 0.9835 0.4959 0.7841 0.9739 21
0.1031 0.1045 0.7511 0.8522 0.9668 0.6398 0.9323 0.9846 0.4911 0.7878 0.9743 22
0.1038 0.1245 0.7335 0.8535 0.9628 0.6322 0.9488 0.9794 0.4609 0.7683 0.9713 23
0.0989 0.1130 0.7476 0.8608 0.9652 0.6641 0.9372 0.9813 0.4895 0.7805 0.9729 24
0.0961 0.0993 0.7534 0.8560 0.9672 0.6481 0.9356 0.9844 0.4949 0.7904 0.9748 25
0.0931 0.0977 0.7616 0.8574 0.9684 0.6623 0.9242 0.9858 0.5099 0.7995 0.9754 26
0.0913 0.0899 0.7685 0.8547 0.9701 0.6575 0.9184 0.9883 0.5192 0.8096 0.9768 27
0.0899 0.0984 0.7572 0.8550 0.9683 0.6393 0.9398 0.9858 0.5015 0.7940 0.9759 28
0.0918 0.1307 0.7440 0.8719 0.9635 0.6838 0.9545 0.9773 0.4872 0.7735 0.9713 29
0.0919 0.1239 0.7405 0.8590 0.9641 0.6442 0.9526 0.9801 0.4707 0.7784 0.9725 30
0.0925 0.0990 0.7699 0.8629 0.9696 0.6859 0.9163 0.9865 0.5271 0.8067 0.9761 31
0.0889 0.1069 0.7563 0.8708 0.9664 0.6864 0.9450 0.9811 0.5038 0.7913 0.9738 32
0.0836 0.0913 0.7707 0.8617 0.9702 0.6714 0.9265 0.9873 0.5265 0.8086 0.9770 33
0.0822 0.1041 0.7645 0.8788 0.9672 0.7170 0.9383 0.9809 0.5161 0.8035 0.9740 34
0.0803 0.0981 0.7699 0.8721 0.9691 0.6987 0.9334 0.9843 0.5291 0.8046 0.9759 35
0.0800 0.1018 0.7597 0.8681 0.9678 0.6728 0.9485 0.9830 0.5104 0.7935 0.9752 36
0.0779 0.0975 0.7727 0.8769 0.9692 0.7185 0.9286 0.9837 0.5349 0.8075 0.9757 37
0.0756 0.0984 0.7697 0.8742 0.9691 0.7003 0.9385 0.9838 0.5280 0.8051 0.9760 38
0.0787 0.1007 0.7646 0.8701 0.9687 0.6791 0.9475 0.9838 0.5173 0.8005 0.9760 39

Framework versions

  • Transformers 4.35.2
  • TensorFlow 2.15.0
  • Datasets 2.16.1
  • Tokenizers 0.15.0
Downloads last month
17
Inference API
Unable to determine this model’s pipeline type. Check the docs .

Model tree for Lit4pCol4b/mit-b1_segformer_ADE20k_RGB_IS_v1

Base model

nvidia/mit-b1
Finetuned
(16)
this model