segformer-finetuned-sidewalk-10k-steps

This model is a fine-tuned version of nvidia/mit-b0 on the fuji12345/sample-segmentation_dataset dataset. It achieves the following results on the evaluation set:

• Loss: 0.1181
• Mean Iou: 0.0
• Mean Accuracy: nan
• Overall Accuracy: nan
• Accuracy Hallucination: nan
• Accuracy Normal: nan
• Iou Hallucination: 0.0
• Iou Normal: 0.0

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: 8
• eval_batch_size: 8
• seed: 1337
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: polynomial
• training_steps: 200

Training results

Training Loss	Epoch	Step	Validation Loss	Mean Iou	Mean Accuracy	Overall Accuracy	Accuracy Hallucination	Accuracy Normal	Iou Hallucination	Iou Normal
No log	1.0	21	0.4468	0.0	nan	nan	nan	nan	0.0	0.0
No log	2.0	42	0.2196	0.0	nan	nan	nan	nan	0.0	0.0
No log	3.0	63	0.2136	0.0	nan	nan	nan	nan	0.0	0.0
No log	4.0	84	0.1816	0.0	nan	nan	nan	nan	0.0	0.0
0.3217	5.0	105	0.1449	0.0	nan	nan	nan	nan	0.0	0.0
0.3217	6.0	126	0.1380	0.0	nan	nan	nan	nan	0.0	0.0
0.3217	7.0	147	0.1239	0.0	nan	nan	nan	nan	0.0	0.0
0.3217	8.0	168	0.1247	0.0	nan	nan	nan	nan	0.0	0.0
0.3217	9.0	189	0.1163	0.0	nan	nan	nan	nan	0.0	0.0
0.1596	9.5238	200	0.1181	0.0	nan	nan	nan	nan	0.0	0.0

Framework versions

• Transformers 4.53.0.dev0
• Pytorch 2.6.0+cu124
• Datasets 3.6.0
• Tokenizers 0.21.1