Open lisiyizu opened 1 month ago
I you're trying to get Flux working on MPS you'll need to figure out why it's broken (noisy images) on PyTorch 2.4. but works with PyTorch 2.3.1 as well as getting fp8 support
Having float8 dtypes defined on MPS would be nice, we'd welcome community contributions on this!
š Describe the bug
Versions
Collecting environment information... PyTorch version: 2.4.0 Is debug build: False CUDA used to build PyTorch: None ROCM used to build PyTorch: N/A
OS: macOS 14.5 (arm64) GCC version: Could not collect Clang version: 15.0.0 (clang-1500.3.9.4) CMake version: Could not collect Libc version: N/A
Python version: 3.10.8 (main, Nov 24 2022, 08:08:27) [Clang 14.0.6 ] (64-bit runtime) Python platform: macOS-14.5-arm64-arm-64bit Is CUDA available: False CUDA runtime version: No CUDA CUDA_MODULE_LOADING set to: N/A GPU models and configuration: No CUDA Nvidia driver version: No CUDA cuDNN version: No CUDA HIP runtime version: N/A MIOpen runtime version: N/A Is XNNPACK available: True
CPU: Apple M2 Pro
Versions of relevant libraries: [pip3] mypy-extensions==1.0.0 [pip3] numpy==1.26.4 [pip3] onnx==1.16.1 [pip3] onnxruntime==1.15.1 [pip3] open_clip_torch==2.26.1 [pip3] pytorch-lightning==2.3.3 [pip3] torch==2.4.0 [pip3] torchaudio==2.4.0 [pip3] torchmetrics==1.4.0.post0 [pip3] torchsde==0.2.6 [pip3] torchvision==0.19.0 [conda] blas 1.0 mkl
[conda] mkl 2023.1.0 h8e150cf_43560
[conda] mkl-service 2.4.0 py312h6c40b1e_1
[conda] mkl_fft 1.3.8 py312h6c40b1e_0
[conda] mkl_random 1.2.4 py312ha357a0b_0
[conda] numpy 1.26.4 py312hac873b0_0
[conda] numpy-base 1.26.4 py312h6f81483_0
[conda] pytorch 2.2.2 py3.12_0 pytorch [conda] torchaudio 2.2.2 py312_cpu pytorch [conda] torchvision 0.17.2 py312_cpu pytorch
cc @kulinseth @albanD @malfet @DenisVieriu97 @jhavukainen @yanbing-j @vkuzo @kadeng @penguinwu