| --- |
| library_name: segmentation-models-pytorch |
| license: mit |
| pipeline_tag: image-segmentation |
| short_description: 5 classes image segmentation model trained on the comma10k dataset. |
| tags: |
| - model_hub_mixin |
| - pytorch_model_hub_mixin |
| - segmentation-models-pytorch |
| - semantic-segmentation |
| - pytorch |
| languages: |
| - python |
| --- |
| # Unet Model Card |
|
|
| Table of Contents: |
| - [Load trained model](#load-trained-model) |
| - [Model init parameters](#model-init-parameters) |
| - [Model metrics](#model-metrics) |
| - [Dataset](#dataset) |
|
|
| ## Load trained model |
| ```python |
| import segmentation_models_pytorch as smp |
| import albumentations as A |
| |
| hub_repo = "commaai/comma10k-segnet" |
| model = smp.from_pretrained(hub_repo) |
| transform = A.Compose.from_pretrained(hub_repo) |
| ``` |
|
|
| ## Model init parameters |
| ```python |
| model_init_params = { |
| "encoder_name": "tu-efficientnet_b2", |
| "encoder_depth": 5, |
| "encoder_weights": None, |
| "decoder_use_norm": "batchnorm", |
| "decoder_channels": (256, 128, 64, 32, 16), |
| "decoder_attention_type": None, |
| "decoder_interpolation": "nearest", |
| "in_channels": 3, |
| "classes": 5, |
| "activation": None, |
| "aux_params": None |
| } |
| ``` |
|
|
| ## Dataset |
| Dataset name: comma10k |
|
|
| ## More Information |
| - Library: https://github.com/qubvel/segmentation_models.pytorch |
| - Docs: https://smp.readthedocs.io/en/latest/ |
| |
| This model has been pushed to the Hub using the [PytorchModelHubMixin](https://huggingface.co/docs/huggingface_hub/package_reference/mixins#huggingface_hub.PyTorchModelHubMixin) |
