MCPcopy Create free account
hub / github.com/ImprintLab/MedSegDiff / UNetModel_v1preview

Class UNetModel_v1preview

guided_diffusion/unet.py:487–794  ·  view source on GitHub ↗

The full UNet model with attention and timestep embedding. :param in_channels: channels in the input Tensor. :param model_channels: base channel count for the model. :param out_channels: channels in the output Tensor. :param num_res_blocks: number of residual blocks per downsamp

Source from the content-addressed store, hash-verified

source not stored for this graph (policy: none)

Callers 1

create_modelFunction · 0.85

Calls

no outgoing calls

Tested by

no test coverage detected