reannayang's picture
update model card README.md
65e7c6e
metadata
license: apache-2.0
tags:
  - vision
  - image-segmentation
  - generated_from_trainer
model-index:
  - name: segformer-b0-pavement
    results: []

segformer-b0-pavement

This model is a fine-tuned version of nvidia/mit-b0 on the reannayang/FL_pavement dataset. It achieves the following results on the evaluation set:

  • Loss: 0.4165
  • Mean Iou: 0.6318
  • Mean Accuracy: 0.9700
  • Overall Accuracy: 0.9738
  • Per Category Iou: [0.0, 0.964166382973358, 0.9809231860559384, 0.0, 0.9295139919583345, 0.9164463823409184]
  • Per Category Accuracy: [nan, 0.9643001261034048, 0.9983497924348297, nan, 0.995031342981772, 0.9223532638507954]

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: 6e-05
  • train_batch_size: 2
  • eval_batch_size: 2
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 50

Training results

Training Loss Epoch Step Validation Loss Mean Iou Mean Accuracy Overall Accuracy Per Category Iou Per Category Accuracy
1.0651 10.0 20 1.3005 0.5967 0.9512 0.9534 [0.0, 0.9462421185372005, 0.9681701711239586, 0.0, 0.7994398965962947, 0.8662896799897185] [nan, 0.9462421185372005, 0.9693809143181291, nan, 0.9648149753011526, 0.9243828853538124]
0.5732 20.0 40 0.6626 0.6287 0.9702 0.9760 [0.0, 0.975246652572234, 0.985446932366533, 0.0, 0.9010974339804011, 0.9103918683964157] [nan, 0.9772635561160151, 0.9952040842637238, nan, 0.9748678395008233, 0.9334887547997806]
0.6987 30.0 60 0.4319 0.6317 0.9705 0.9758 [0.0, 0.9709705045212967, 0.9798115236227942, 0.0, 0.9255918522130127, 0.9139245313729214] [nan, 0.9722194199243379, 0.9986205296134905, nan, 0.9871161568015715, 0.924026330224904]
0.6915 40.0 80 0.4382 0.6237 0.9634 0.9692 [0.0, 0.9611727616645649, 0.9725125142706595, 0.0, 0.9147983251179308, 0.8937433316006894] [nan, 0.9611727616645649, 0.9993811721630611, nan, 0.9971690210012422, 0.896023038946791]
0.4373 50.0 100 0.4165 0.6318 0.9700 0.9738 [0.0, 0.964166382973358, 0.9809231860559384, 0.0, 0.9295139919583345, 0.9164463823409184] [nan, 0.9643001261034048, 0.9983497924348297, nan, 0.995031342981772, 0.9223532638507954]

Framework versions

  • Transformers 4.19.2
  • Pytorch 1.7.1
  • Datasets 2.2.1
  • Tokenizers 0.12.1