README.md

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: 0.4558
• Mean Iou: 0.6310
• Mean Accuracy: 0.7758
• Overall Accuracy: 0.9687
• Per Category Iou: [0.0, 0.9582107718835582, 0.9831802335937301, 0.0, 0.9070478290281362, 0.9376628700260592]
• Per Category Accuracy: [nan, 0.964865692983352, 0.9920279343235298, 0.0, 0.9513585956798234, 0.9709244156080405]

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
1.0751	10.0	20	1.1494	0.5858	0.7574	0.9436	[0.0, 0.9446293533289976, 0.9477727359877293, 0.0, 0.7979866954491387, 0.8246812143200128]	[nan, 0.9457026861403329, 0.951885181714315, 0.0, 0.9519892783938614, 0.93727071521693]
0.8732	20.0	40	0.6625	0.6209	0.7710	0.9635	[0.0, 0.9549288283268325, 0.9793625315294657, 0.0, 0.8879129036248917, 0.903449846293977]	[nan, 0.9626825656214891, 0.9869186233102687, 0.0, 0.9430020497188206, 0.9623139192917564]
0.4736	30.0	60	0.5124	0.6302	0.7753	0.9687	[0.0, 0.9600678403726158, 0.9813250559739467, 0.0, 0.9086066088704611, 0.9311320479537963]	[nan, 0.968414870799714, 0.990282039102918, 0.0, 0.9501760655910022, 0.9678319134099385]
0.503	40.0	80	0.4726	0.6305	0.7753	0.9680	[0.0, 0.9574198453934862, 0.9821685397548652, 0.0, 0.9074944966980188, 0.9358034824412265]	[nan, 0.9645337554897355, 0.9906414881189263, 0.0, 0.953329479161192, 0.9678925507079404]
0.6762	50.0	100	0.4558	0.6310	0.7758	0.9687	[0.0, 0.9582107718835582, 0.9831802335937301, 0.0, 0.9070478290281362, 0.9376628700260592]	[nan, 0.964865692983352, 0.9920279343235298, 0.0, 0.9513585956798234, 0.9709244156080405]

Framework versions

• Transformers 4.19.2
• Pytorch 1.11.0+cu113
• Datasets 2.2.2
• Tokenizers 0.12.1