segformer-b1-finetuned-tennisdata

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

• eval_loss: 0.0158
• eval_mean_iou: 0.4994
• eval_mean_accuracy: 0.6483
• eval_overall_accuracy: 0.9915
• eval_accuracy_undefined: nan
• eval_accuracy_object: nan
• eval_accuracy_ball: 0.0
• eval_accuracy_playerTop: 0.7071
• eval_accuracy_playerBottom: 0.8904
• eval_accuracy_court: 0.9956
• eval_iou_undefined: 0.0
• eval_iou_object: nan
• eval_iou_ball: 0.0
• eval_iou_playerTop: 0.7071
• eval_iou_playerBottom: 0.7968
• eval_iou_court: 0.9931
• eval_runtime: 8.2745
• eval_samples_per_second: 2.175
• eval_steps_per_second: 1.088
• epoch: 29.41
• step: 1000

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: 100

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.0+cu121
• Datasets 2.17.1
• Tokenizers 0.15.2