XiyaLiu
XiyaLiu
Hi @verymadmatt, did you figure out how to train coarsely annotated cityscapes data? I also met this question.
楼主有没有解决呀？用的是colab跑的吗:o?
@EagleEyeKestrel 过年还在奋斗，辛苦辛苦，新年快乐
same here, did you solve the problem?
版本太高了吧，huggingface上也有docker环境
你试试 docker run -it -p 7860:7860 --platform=linux/amd64 --gpus all \ registry.hf.space/tencent-hunyuan3d-2mini-turbo:latest python gradio_app.py
哦我想起来了我之前不能在40系跑，然后问了cursor，让我在setup.py加了这个重新编译就行了，50系可能吧，可以再问问LLM ```custom_rasterizer_module = CUDAExtension('custom_rasterizer_kernel', [ 'lib/custom_rasterizer_kernel/rasterizer.cpp', 'lib/custom_rasterizer_kernel/grid_neighbor.cpp', 'lib/custom_rasterizer_kernel/rasterizer_gpu.cu', ], extra_compile_args={ 'nvcc': [ '-gencode', 'arch=compute_89,code=sm_89', # Add this line for RTX 4090 '-gencode', 'arch=compute_90,code=sm_90', ], } )```
interesting