SegFormer_b2_mappillary_

This model is a fine-tuned version of nvidia/segformer-b2-finetuned-cityscapes-1024-1024 on an unknown dataset. It achieves the following results on the evaluation set:

• eval_loss: 0.9598
• eval_mean_iou: 0.6780
• eval_mean_accuracy: 0.7951
• eval_overall_accuracy: 0.9391
• eval_accuracy_construction--barrier--fence: 0.6674
• eval_accuracy_construction--barrier--guard-rail: 0.7787
• eval_accuracy_construction--barrier--other-barrier: 0.7093
• eval_accuracy_construction--barrier--wall: 0.6692
• eval_accuracy_construction--flat--road: 0.9505
• eval_accuracy_construction--flat--service-lane: 0.5410
• eval_accuracy_construction--flat--sidewalk: 0.9029
• eval_accuracy_construction--structure--building: 0.9494
• eval_accuracy_human--person: 0.8428
• eval_accuracy_human--rider--bicyclist: 0.7374
• eval_accuracy_marking--crosswalk-zebra: 0.8275
• eval_accuracy_marking--general: 0.6969
• eval_accuracy_nature--sky: 0.9902
• eval_accuracy_nature--terrain: 0.8238
• eval_accuracy_nature--vegetation: 0.9447
• eval_accuracy_object--support--pole: 0.5732
• eval_accuracy_object--support--traffic-sign-frame: 0.6710
• eval_accuracy_object--traffic-light: 0.7524
• eval_accuracy_object--traffic-sign--front: 0.8163
• eval_accuracy_object--vehicle--bicycle: 0.7771
• eval_accuracy_object--vehicle--bus: 0.8829
• eval_accuracy_object--vehicle--car: 0.9659
• eval_accuracy_object--vehicle--truck: 0.8158
• eval_iou_construction--barrier--fence: 0.5508
• eval_iou_construction--barrier--guard-rail: 0.6288
• eval_iou_construction--barrier--other-barrier: 0.5638
• eval_iou_construction--barrier--wall: 0.5354
• eval_iou_construction--flat--road: 0.9129
• eval_iou_construction--flat--service-lane: 0.4333
• eval_iou_construction--flat--sidewalk: 0.7696
• eval_iou_construction--structure--building: 0.8821
• eval_iou_human--person: 0.6700
• eval_iou_human--rider--bicyclist: 0.5363
• eval_iou_marking--crosswalk-zebra: 0.7082
• eval_iou_marking--general: 0.5822
• eval_iou_nature--sky: 0.9811
• eval_iou_nature--terrain: 0.6964
• eval_iou_nature--vegetation: 0.8935
• eval_iou_object--support--pole: 0.4515
• eval_iou_object--support--traffic-sign-frame: 0.5508
• eval_iou_object--traffic-light: 0.5782
• eval_iou_object--traffic-sign--front: 0.7134
• eval_iou_object--vehicle--bicycle: 0.5514
• eval_iou_object--vehicle--bus: 0.7858
• eval_iou_object--vehicle--car: 0.9004
• eval_iou_object--vehicle--truck: 0.7182
• eval_runtime: 1416.4555
• eval_samples_per_second: 1.412
• eval_steps_per_second: 0.706
• epoch: 14.0
• step: 31500

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: 9e-05
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 8
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 1000
• num_epochs: 30
• mixed_precision_training: Native AMP

Framework versions

• Transformers 4.48.1
• Pytorch 2.1.2+cu121
• Datasets 3.2.0
• Tokenizers 0.21.0