Cannot install executorch on M series macbook - install_requirements.sh fails #7419
Comments
|
does |
|
looks like |
|
git clone https://github.com/pytorch/executorch.git --depth 1 --recurse-submodules --shallow-submodules && cd executorch |
|
I tried your suggestion and no cigar. |
can you be more specific? Are you referring to https://pytorch.org/executorch/stable/getting-started-setup.html or a different doc? If so, are you using the release/0.4 branch as that page suggests, or main? Here's my attempt to repro. My machine has buck2 on PATH, but it is a nonworking-for-purposes-of-executorch version. I've had no end of issues with conda, so I've removed it from my machine in favor of venv. I note that your error refers to |
swolchok
added
need-user-input
|
correct. I followed the step https://pytorch.org/executorch/stable/getting-started-setup.html here from a clean directory using a venv created from python3.11 |
I definitely have buck2-bin now, and I see that Utils.cmake sets it as the --cache_dir. I must have misread previously.
This is an error from buck2. 1) is your disk full? buck2 should be creating that directory itself IIUC 2) have you tried again recently? I bumped the buck2 version earlier this week. |
|
Yeah I have plenty of space. I'm wondering if I can get to a clean state by removing all instances of buck2 from my machine |
|
@byjlw any luck? I'm also experiencing the exact same thing |
|
Strangely enough, I just got it to work. I was previously trying with python 3.11 and conda. Using python 3.12 with venv did the trick |
|
Unbelievable - I tried it again a few times running the exact same commands and it won't work. All starting from a clean directory and venv |
|
It looks like buck2 is supposed to automatically recover from this happening. We should probably ask buck2 folks about this. Citation:
|
Forwarded this thread to them; Meta employees can see it at https://fb.workplace.com/groups/buck2users/posts/3860457534210599/ (there isn't substantially new information, just a summary so folks can understand without clicking through to GitHub) |
🐛 Describe the bug
Try to install using the install steps. Throws an error
Versions
Collecting environment information...
PyTorch version: 2.6.0.dev20241218
Is debug build: False
CUDA used to build PyTorch: None
ROCM used to build PyTorch: N/A
OS: macOS 14.6.1 (arm64)
GCC version: Could not collect
Clang version: 15.0.0 (clang-1500.3.9.4)
CMake version: version 3.31.2
Libc version: N/A
Python version: 3.11.8 (v3.11.8:db85d51d3e, Feb 6 2024, 18:02:37) [Clang 13.0.0 (clang-1300.0.29.30)] (64-bit runtime)
Python platform: macOS-14.6.1-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 M3 Max
Versions of relevant libraries:
[pip3] numpy==2.2.0
[pip3] torch==2.6.0.dev20241218
[pip3] torchaudio==2.6.0.dev20241218
[pip3] torchsr==1.0.4
[pip3] torchvision==0.22.0.dev20241218
[conda] Could not collect
The text was updated successfully, but these errors were encountered: