segformer-custom

<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->

segformer-custom

This model is a fine-tuned version of sanjaybora04/segformer-custom on an unknown dataset. It achieves the following results on the evaluation set:

• eval_loss: 0.4589
• eval_mean_iou: 0.6566
• eval_mean_accuracy: 0.7804
• eval_iou_background: 0.8221
• eval_iou_wall: 0.7010
• eval_iou_floor: 0.7254
• eval_iou_ceiling: 0.7398
• eval_iou_stairs: 0.3269
• eval_iou_countertop: 0.6245
• eval_accuracy_background: 0.9023
• eval_accuracy_wall: 0.8342
• eval_accuracy_floor: 0.8285
• eval_accuracy_ceiling: 0.8565
• eval_accuracy_stairs: 0.3478
• eval_accuracy_countertop: 0.9131
• eval_runtime: 13.9462
• eval_samples_per_second: 10.756
• eval_steps_per_second: 1.076
• epoch: 5.0
• step: 6250

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: 10
• eval_batch_size: 10
• 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: linear
• num_epochs: 10
• mixed_precision_training: Native AMP

Framework versions

• Transformers 5.0.0
• Pytorch 2.10.0+cu128
• Datasets 3.6.0
• Tokenizers 0.22.2