Why did it run into a 'loss not found' error when executing 'python main.py --dataset ispd2005 --design_name adaptec1 --load_from_raw True --detail_placement True'

[huichengyu]

Traceback (most recent call last):
File "/home/yuhuicheng/Xplace/main.py", line 104, in
main()
File "/home/yuhuicheng/Xplace/main.py", line 100, in main
run_placement_main(args, logger)
File "/home/yuhuicheng/Xplace/src/run_placement.py", line 41, in run_placement_main
run_placement_all(args, logger)
File "/home/yuhuicheng/Xplace/src/run_placement.py", line 26, in run_placement_all
place_metrics, route_metrics = run_placement_single(cur_args, logger)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/yuhuicheng/Xplace/src/run_placement.py", line 12, in run_placement_single
res = run_placement_main_nesterov(args, logger)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/yuhuicheng/Xplace/src/run_placement_nesterov.py", line 163, in run_placement_main_nesterov
obj = optimizer.step(obj_and_grad_fn)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/home/share/process/anaconda3/envs/xplace/lib/python3.11/site-packages/torch/optim/optimizer.py", line 391, in wrapper
out = func(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^
File "/home/yuhuicheng/Xplace/src/nesterov_optimizer.py", line 128, in step
return obj
^^^
UnboundLocalError: cannot access local variable 'obj' where it is not associated with a value

[liulixinkerry]

May you provide your OS, cuda and pytorch version?
Thanks.

[huichengyu]

Thank you very much for your reply.
Linux version 3.10.0-1160.105.1.el7.x86_64 ([email protected]) (gcc version 4.8.5 20150623 (Red Hat 4.8.5-44
CUDA Version: 12.0
pytorch 2.3.0+cu121

[liulixinkerry]

As I remember, the latest torch(>=2.2)'s C++ API requires GCC >= 9.4. Is there any compilation error you encounter when you compile the program?

BTW, GCC 9.4 + CUDA 12.1 + Pytorch 2.3.0+cu121 works for me

[liulixinkerry]

Your CUDA version (12.0) is lower than Pytorch's CUDA version (12.1). I am not sure whether this will cause some problems.

[huichengyu]

As I remember, the latest torch(>=2.2)'s C++ API requires GCC >= 9.4. Is there any compilation error you encounter when you compile the program?

BTW, GCC 9.4 + CUDA 12.1 + Pytorch 2.3.0+cu121 works for me

support up to CUDA 12.0

[huichengyu]

Your CUDA version (12.0) is lower than Pytorch's CUDA version (12.1). I am not sure whether this will cause some problems.

Even if I switch to PyTorch 2.4.0 with CUDA 11.8, it still doesn't work

[liulixinkerry]

Your CUDA version (12.0) is lower than Pytorch's CUDA version (12.1). I am not sure whether this will cause some problems.

Even if I switch to PyTorch 2.4.0 with CUDA 11.8, it still doesn't work

What is your GCC version? Is it still 4.8.5?

[huichengyu]

Your CUDA version (12.0) is lower than Pytorch's CUDA version (12.1). I am not sure whether this will cause some problems.

Even if I switch to PyTorch 2.4.0 with CUDA 11.8, it still doesn't work

What is your GCC version? Is it still 4.8.5?

GCC version is 11.4.0

[liulixinkerry]

File "/home/yuhuicheng/Xplace/src/run_placement_nesterov.py", line 163, in run_placement_main_nesterov
obj = optimizer.step(obj_and_grad_fn)

Why is your obj = optimizer.step(obj_and_grad_fn) on L163? It should be on https://github.com/cuhk-eda/Xplace/blob/main/src/run_placement_nesterov.py#L145

I guess you are currently using the nn branch. However, I have not been maintaining this branch for approximately one year.

I suggest referring to commit 0aa58ed as a reference point to modify your code in nn branch.

[huichengyu]

-- The C compiler identification is GNU 11.4.0
-- The CXX compiler identification is GNU 11.4.0
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Check for working C compiler: /usr/bin/gcc - skipped
-- Detecting C compile features
-- Detecting C compile features - done
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Check for working CXX compiler: /usr/bin/c++ - skipped
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- CMAKE_BUILD_TYPE: Release
-- PROJECT_SOURCE_DIR=/home/yuhuicheng/Xplace
-- CMAKE_CXX_ABI: _GLIBCXX_USE_CXX11_ABI=0
-- pybind11 v2.13.0 dev1
-- Found PythonInterp: /home/share/process/anaconda3/envs/xplace/bin/python (found suitable version "3.11.8", minimum required is "3.6")
-- Found PythonLibs: /home/share/process/anaconda3/envs/xplace/lib/libpython3.11.so
-- Performing Test HAS_FLTO
-- Performing Test HAS_FLTO - Success
-- PYTHON_INCLUDE_DIRS: /home/share/process/anaconda3/envs/xplace/include/python3.11
CMake Deprecation Warning at thirdparty/flute/CMakeLists.txt:32 (CMAKE_MINIMUM_REQUIRED):
Compatibility with CMake < 3.5 will be removed from a future version of
CMake.

Update the VERSION argument value or use a ... suffix to tell
CMake that the project does not need compatibility with older versions.

-- FLUTE_INCLUDE_DIR: /home/yuhuicheng/Xplace/thirdparty/flute
-- LEMON_INCLUDE_DIRS: /home/yuhuicheng/Xplace/thirdparty/lemon/include
-- LEMON_LIBRARIES: /home/yuhuicheng/Xplace/thirdparty/lemon/lib/libemon.a
-- Found PkgConfig: /usr/bin/pkg-config (found version "0.27.1")
-- Checking for module 'cairo'
-- No package 'cairo' found
-- Found Cairo: /home/yuhuicheng/cairo16-linux/include/cairo
-- CAIRO_INCLUDE_DIRS: /home/yuhuicheng/cairo16-linux/include/cairo
-- CAIRO_LIBRARIES: /home/yuhuicheng/cairo16-linux/lib/libcairo.so
-- TORCH_INSTALL_PREFIX=/home/share/process/anaconda3/envs/xplace/lib/python3.11/site-packages/torch
-- TORCH_VERSION=2.3
CMake Warning (dev) at CMakeLists.txt:69 (find_package):
Policy CMP0146 is not set: The FindCUDA module is removed. Run "cmake
--help-policy CMP0146" for policy details. Use the cmake_policy command to
set the policy and suppress this warning.

This warning is for project developers. Use -Wno-dev to suppress it.

-- Performing Test CMAKE_HAVE_LIBC_PTHREAD
-- Performing Test CMAKE_HAVE_LIBC_PTHREAD - Failed
-- Looking for pthread_create in pthreads
-- Looking for pthread_create in pthreads - not found
-- Looking for pthread_create in pthread
-- Looking for pthread_create in pthread - found
-- Found Threads: TRUE
-- Found CUDA: /usr/local/cuda (found suitable version "11.8", minimum required is "11.8")
-- TORCH_ENABLE_CUDA=1
-- CUDA_ARCH_FLAGS: -gencode;arch=compute_80,code=sm_80;-gencode;arch=compute_86,code=sm_86
-- CUDA_NVCC_FLAGS: -gencode;arch=compute_80,code=sm_80;-gencode;arch=compute_86,code=sm_86;--compiler-options;-fPIC;-std=c++17;-gencode;arch=compute_80,code=sm_80;-gencode;arch=compute_86,code=sm_86;--extended-lambda;--expt-relaxed-constexpr
-- TORCH_INCLUDE_DIRS=/home/share/process/anaconda3/envs/xplace/include/python3.11/home/share/process/anaconda3/envs/xplace/lib/python3.11/site-packages/torch/include/home/share/process/anaconda3/envs/xplace/lib/python3.11/site-packages/torch/include/torch/csrc/api/include
-- Found OpenMP_C: -fopenmp (found version "4.5")
-- Found OpenMP_CXX: -fopenmp (found version "4.5")
-- Found OpenMP: TRUE (found version "4.5")
-- Configuring done (3.3s)
-- Generating done (0.2s)
-- Build files have been written to: /home/yuhuicheng/Xplace/build

[liulixinkerry]

File "/home/yuhuicheng/Xplace/src/run_placement_nesterov.py", line 163, in run_placement_main_nesterov
obj = optimizer.step(obj_and_grad_fn)

Why is your obj = optimizer.step(obj_and_grad_fn) on L163? It should be on https://github.com/cuhk-eda/Xplace/blob/main/src/run_placement_nesterov.py#L145

I guess you are currently using the nn branch. However, I have not been maintaining this branch for approximately one year.

I suggest referring to commit 0aa58ed as a reference point to modify your code in nn branch.

@huichengyu Please try this.

[huichengyu]

File "/home/yuhuicheng/Xplace/src/run_placement_nesterov.py", line 163, in run_placement_main_nesterov文件“/home/yuhuicheng/Xplace/src/run_placement_nesterov.py”，第 163 行，run_placement_main_nesterov
obj = optimizer.step(obj_and_grad_fn)obj = 优化器.step（obj_and_grad_fn）

Why is your obj = optimizer.step(obj_and_grad_fn) on L163? It should be on https://github.com/cuhk-eda/Xplace/blob/main/src/run_placement_nesterov.py#L145为什么你在 obj = optimizer.step(obj_and_grad_fn) L163 上？它应该打开 https://github.com/cuhk-eda/Xplace/blob/main/src/run_placement_nesterov.py#L145
I guess you are currently using the nn branch. However, I have not been maintaining this branch for approximately one year.我猜您目前正在使用该 nn 分支。但是，我已经有大约一年没有维护这个分支了。
I suggest referring to commit 0aa58ed as a reference point to modify your code in nn branch.我建议参考提交 0aa58ed 作为参考点来修改分支中的 nn 代码。

@huichengyu Please try this. 请试试这个。

this does not contain nn branch

[huichengyu]

File "/home/yuhuicheng/Xplace/src/run_placement_nesterov.py", line 163, in run_placement_main_nesterov
obj = optimizer.step(obj_and_grad_fn)

Why is your obj = optimizer.step(obj_and_grad_fn) on L163? It should be on https://github.com/cuhk-eda/Xplace/blob/main/src/run_placement_nesterov.py#L145
I guess you are currently using the nn branch. However, I have not been maintaining this branch for approximately one year.
I suggest referring to commit 0aa58ed as a reference point to modify your code in nn branch.

@huichengyu Please try this.

problem solved ，thank you very much