WaveUNetimplement Wave-U-Net by pytorchresultWave-U-Netmy improved version SDR=11.83 SDR=12.51Train networkif you just want to train the model, use commandTrain.py python commandTrain.py --dataset both ##(for both ccmixter and musdb18) python commandTrain.py --dataset ccmixter ##(for cc...
1. Add an option to allow users to use UNet in fp8/gguf but lora in fp16. 2. All FP16 loras do not need patch. Others will only patch again when lora weight change. 3. FP8 unet + fp16 lora are available (somewhat only available) in Forge now. This also solves some “LoRA too...
train MVPNet with virtual views python mvpnet/train_mvpnet_3d.py --cfg configs/scannet/mvpnet_3d_unet_resnet34_pn2ssg.yaml VMNetpre-train the 2D semantic segmentation network with generated virtual views. The virtual views need to be orgnized as <datasets> |-- Scannet_virtual |-- <RGBFolde...
MinkUNet and SPVCNN methods. Noteworthy, the improved MinkUNetv2 can achieve 70.3 mIoU on the validation set of SemanticKITTI. We have also supported the training of BEVFusion and an occupancy prediction method, TPVFomrer, in ourprojects. More new features about 3D perception are on the way....
train MVPNet with virtual views python mvpnet/train_mvpnet_3d.py --cfg configs/scannet/mvpnet_3d_unet_resnet34_pn2ssg.yaml VMNetpre-train the 2D semantic segmentation network with generated virtual views. The virtual views need to be orgnized as <datasets> |-- Scannet_virtual |-- <RGBFolde...