TypeError: 'weights_only' when deploying checkpoint to LAMMPS using PyTorch 1.12.0 #205
Description
Dear SevenNet Developers,
I'm excited to try out your model and checkpoints, but am running into a TypeError when trying to convert one of the publicly available checkpoints to a LAMMPS serial potential. I installed sevenn along with a supporting Conda environment, trying to install as many dependencies as possible from conda-forge.
I am on a cluster limited to CUDA 11.6, so I installed PyTorch 1.12.0+cu116, as this is one of the recommended versions. PyTorch 1.12 supports at most Python 3.10, so I installed that version.
Problem
I tried deploying the 7net-mf-ompa checkpoint to a LAMMPS potential, trying both of these commands:
# Deploying from checkpoint path
sevenn_get_model ./checkpoint_sevennet_mf_ompa.pth
# Deploying from checkpoint name
sevenn_get_model 7net-mf-ompaBoth of these result in a TypeError around an unexpected weights_only argument to pickle.Unpickler:
Traceback (most recent call last):
File "/home/user/.conda/envs/sevenn-pytorch-cu116/bin/sevenn_get_model", line 8, in <module>
sys.exit(main())
File "/home/user/.conda/envs/sevenn-pytorch-cu116/lib/python3.10/site-packages/sevenn/main/sevenn_get_model.py", line 60, in main
run(args)
File "/home/user/.conda/envs/sevenn-pytorch-cu116/lib/python3.10/site-packages/sevenn/main/sevenn_get_model.py", line 38, in run
deploy(checkpoint_path, output_prefix, modal)
File "/home/user/.conda/envs/sevenn-pytorch-cu116/lib/python3.10/site-packages/sevenn/scripts/deploy.py", line 27, in deploy
model, config = cp.build_model('e3nn'), cp.config
File "/home/user/.conda/envs/sevenn-pytorch-cu116/lib/python3.10/site-packages/sevenn/checkpoint.py", line 297, in build_model
cp_using_cue = self.config[KEY.CUEQUIVARIANCE_CONFIG]['use']
File "/home/user/.conda/envs/sevenn-pytorch-cu116/lib/python3.10/site-packages/sevenn/checkpoint.py", line 223, in config
self._load()
File "/home/user/.conda/envs/sevenn-pytorch-cu116/lib/python3.10/site-packages/sevenn/checkpoint.py", line 272, in _load
cp = torch_load(cp_path, weights_only=False, map_location='cpu')
File "/home/user/.conda/envs/sevenn-pytorch-cu116/lib/python3.10/site-packages/torch/serialization.py", line 712, in load
return _load(opened_zipfile, map_location, pickle_module, **pickle_load_args)
File "/home/user/.conda/envs/sevenn-pytorch-cu116/lib/python3.10/site-packages/torch/serialization.py", line 1047, in _load
unpickler = UnpicklerWrapper(data_file, **pickle_load_args)
TypeError: 'weights_only' is an invalid keyword argument for Unpickler()
Environment and versions
- Python: 3.10, installed from conda-forge.
- Available CUDA version: 11.6.
- Target GPUs: NVidia A40, A100.
- PyTorch: 1.12.0+cu116, installed from pip using
--index-url https://download.pytorch.org/whl/cu116 - sevnn: 0.11.0, installed from pip.
Conda environment
List of packages in environment: "/home/user/.conda/envs/sevenn-pytorch-cu116"
Name Version Build Channel
──────────────────────────────────────────────────────────────────────────────
_libgcc_mutex 0.1 conda_forge conda-forge
_openmp_mutex 4.5 2_gnu conda-forge
aiohappyeyeballs 2.6.1 pyhd8ed1ab_0 conda-forge
aiohttp 3.11.16 py310h89163eb_0 conda-forge
aiosignal 1.3.2 pyhd8ed1ab_0 conda-forge
alsa-lib 1.2.14 hb9d3cd8_0 conda-forge
annotated-types 0.7.0 pyhd8ed1ab_1 conda-forge
appdirs 1.4.4 pyhd8ed1ab_1 conda-forge
ase 3.25.0 pyhd8ed1ab_0 conda-forge
asttokens 3.0.0 pyhd8ed1ab_1 conda-forge
async-timeout 5.0.1 pyhd8ed1ab_1 conda-forge
attrs 25.3.0 pyh71513ae_0 conda-forge
blinker 1.9.0 pyhff2d567_0 conda-forge
braceexpand 0.1.7 pyhd8ed1ab_1 conda-forge
brotli 1.1.0 hb9d3cd8_2 conda-forge
brotli-bin 1.1.0 hb9d3cd8_2 conda-forge
brotli-python 1.1.0 py310hf71b8c6_2 conda-forge
bzip2 1.0.8 h4bc722e_7 conda-forge
ca-certificates 2025.1.31 hbcca054_0 conda-forge
cairo 1.18.4 h3394656_0 conda-forge
certifi 2025.1.31 pyhd8ed1ab_0 conda-forge
cffi 1.17.1 py310h8deb56e_0 conda-forge
charset-normalizer 3.4.1 pyhd8ed1ab_0 conda-forge
click 8.1.8 pyh707e725_0 conda-forge
colorama 0.4.6 pyhd8ed1ab_1 conda-forge
comm 0.2.2 pyhd8ed1ab_1 conda-forge
contourpy 1.3.2 py310h3788b33_0 conda-forge
cpython 3.10.17 py310hd8ed1ab_0 conda-forge
cycler 0.12.1 pyhd8ed1ab_1 conda-forge
cyrus-sasl 2.1.27 h54b06d7_7 conda-forge
dbus 1.13.6 h5008d03_3 conda-forge
debugpy 1.8.14 py310hf71b8c6_0 conda-forge
decorator 5.2.1 pyhd8ed1ab_0 conda-forge
docker-pycreds 0.4.0 py_0 conda-forge
double-conversion 3.3.1 h5888daf_0 conda-forge
eval_type_backport 0.2.2 pyha770c72_0 conda-forge
exceptiongroup 1.2.2 pyhd8ed1ab_1 conda-forge
executing 2.1.0 pyhd8ed1ab_1 conda-forge
expat 2.7.0 h5888daf_0 conda-forge
flask 3.1.0 pyhd8ed1ab_1 conda-forge
font-ttf-dejavu-sans-mono 2.37 hab24e00_0 conda-forge
font-ttf-inconsolata 3.000 h77eed37_0 conda-forge
font-ttf-source-code-pro 2.038 h77eed37_0 conda-forge
font-ttf-ubuntu 0.83 h77eed37_3 conda-forge
fontconfig 2.15.0 h7e30c49_1 conda-forge
fonts-conda-ecosystem 1 0 conda-forge
fonts-conda-forge 1 0 conda-forge
fonttools 4.57.0 py310h89163eb_0 conda-forge
freetype 2.13.3 h48d6fc4_0 conda-forge
frozenlist 1.5.0 py310h89163eb_1 conda-forge
fsspec 2025.3.2 pyhd8ed1ab_0 conda-forge
future 1.0.0 pyhd8ed1ab_2 conda-forge
gitdb 4.0.12 pyhd8ed1ab_0 conda-forge
gitpython 3.1.44 pyhff2d567_0 conda-forge
gmp 6.3.0 hac33072_2 conda-forge
gmpy2 2.1.5 py310he8512ff_3 conda-forge
graphite2 1.3.13 h59595ed_1003 conda-forge
h2 4.2.0 pyhd8ed1ab_0 conda-forge
harfbuzz 11.0.1 h2c12942_0 conda-forge
hpack 4.1.0 pyhd8ed1ab_0 conda-forge
hyperframe 6.1.0 pyhd8ed1ab_0 conda-forge
icu 75.1 he02047a_0 conda-forge
idna 3.10 pyhd8ed1ab_1 conda-forge
importlib-metadata 8.6.1 pyha770c72_0 conda-forge
importlib-resources 6.5.2 pyhd8ed1ab_0 conda-forge
importlib_resources 6.5.2 pyhd8ed1ab_0 conda-forge
ipykernel 6.29.5 pyh3099207_0 conda-forge
ipython 8.35.0 pyh907856f_0 conda-forge
ipython_pygments_lexers 1.1.1 pyhd8ed1ab_0 conda-forge
ipywidgets 8.1.6 pyhd8ed1ab_0 conda-forge
itsdangerous 2.2.0 pyhd8ed1ab_1 conda-forge
jedi 0.19.2 pyhd8ed1ab_1 conda-forge
jinja2 3.1.6 pyhd8ed1ab_0 conda-forge
joblib 1.4.2 pyhd8ed1ab_1 conda-forge
jupyter_client 8.6.3 pyhd8ed1ab_1 conda-forge
jupyter_core 5.7.2 pyh31011fe_1 conda-forge
jupyterlab_widgets 3.0.14 pyhd8ed1ab_0 conda-forge
keyutils 1.6.1 h166bdaf_0 conda-forge
kiwisolver 1.4.7 py310h3788b33_0 conda-forge
krb5 1.21.3 h659f571_0 conda-forge
latexcodec 2.0.1 pyh9f0ad1d_0 conda-forge
lcms2 2.17 h717163a_0 conda-forge
ld_impl_linux-64 2.43 h712a8e2_4 conda-forge
lerc 4.0.0 h27087fc_0 conda-forge
libabseil 20250127.1 cxx17_hbbce691_0 conda-forge
libblas 3.9.0 31_h59b9bed_openblas conda-forge
libbrotlicommon 1.1.0 hb9d3cd8_2 conda-forge
libbrotlidec 1.1.0 hb9d3cd8_2 conda-forge
libbrotlienc 1.1.0 hb9d3cd8_2 conda-forge
libcblas 3.9.0 31_he106b2a_openblas conda-forge
libclang-cpp20.1 20.1.2 default_hb5137d0_0 conda-forge
libclang13 20.1.2 default_h9c6a7e4_0 conda-forge
libcups 2.3.3 h4637d8d_4 conda-forge
libdeflate 1.23 h4ddbbb0_0 conda-forge
libdrm 2.4.124 hb9d3cd8_0 conda-forge
libedit 3.1.20250104 pl5321h7949ede_0 conda-forge
libegl 1.7.0 ha4b6fd6_2 conda-forge
libexpat 2.7.0 h5888daf_0 conda-forge
libffi 3.4.6 h2dba641_1 conda-forge
libgcc 14.2.0 h767d61c_2 conda-forge
libgcc-ng 14.2.0 h69a702a_2 conda-forge
libgfortran 14.2.0 h69a702a_2 conda-forge
libgfortran5 14.2.0 hf1ad2bd_2 conda-forge
libgl 1.7.0 ha4b6fd6_2 conda-forge
libglib 2.84.1 h2ff4ddf_0 conda-forge
libglvnd 1.7.0 ha4b6fd6_2 conda-forge
libglx 1.7.0 ha4b6fd6_2 conda-forge
libgomp 14.2.0 h767d61c_2 conda-forge
libiconv 1.18 h4ce23a2_1 conda-forge
libjpeg-turbo 3.0.0 hd590300_1 conda-forge
liblapack 3.9.0 31_h7ac8fdf_openblas conda-forge
libllvm20 20.1.3 he9d0ab4_0 conda-forge
liblzma 5.8.1 hb9d3cd8_0 conda-forge
libmpdec 4.0.0 h4bc722e_0 conda-forge
libnsl 2.0.1 hd590300_0 conda-forge
libntlm 1.8 hb9d3cd8_0 conda-forge
libopenblas 0.3.29 pthreads_h94d23a6_0 conda-forge
libopengl 1.7.0 ha4b6fd6_2 conda-forge
libpciaccess 0.18 hd590300_0 conda-forge
libpng 1.6.47 h943b412_0 conda-forge
libpq 17.4 h27ae623_1 conda-forge
libsodium 1.0.20 h4ab18f5_0 conda-forge
libsqlite 3.49.1 hee588c1_2 conda-forge
libstdcxx 14.2.0 h8f9b012_2 conda-forge
libstdcxx-ng 14.2.0 h4852527_2 conda-forge
libtiff 4.7.0 hd9ff511_3 conda-forge
libuuid 2.38.1 h0b41bf4_0 conda-forge
libwebp-base 1.5.0 h851e524_0 conda-forge
libxcb 1.17.0 h8a09558_0 conda-forge
libxcrypt 4.4.36 hd590300_1 conda-forge
libxkbcommon 1.8.1 hc4a0caf_0 conda-forge
libxml2 2.13.7 h4bc477f_1 conda-forge
libxslt 1.1.39 h76b75d6_0 conda-forge
libzlib 1.3.1 hb9d3cd8_2 conda-forge
looseversion 1.3.0 pyhd8ed1ab_0 conda-forge
lxml 5.3.2 py310h490dddc_0 conda-forge
markupsafe 3.0.2 py310h89163eb_1 conda-forge
matplotlib 3.10.1 py310hff52083_0 conda-forge
matplotlib-base 3.10.1 py310h68603db_0 conda-forge
matplotlib-inline 0.1.7 pyhd8ed1ab_1 conda-forge
matscipy 1.1.0 py310h5eaa309_2 conda-forge
monty 2025.3.3 pyhd8ed1ab_0 conda-forge
mpc 1.3.1 h24ddda3_1 conda-forge
mpfr 4.2.1 h90cbb55_3 conda-forge
mpmath 1.3.0 pyhd8ed1ab_1 conda-forge
multidict 6.3.2 py310h89163eb_0 conda-forge
munkres 1.1.4 pyh9f0ad1d_0 conda-forge
mysql-common 9.2.0 h266115a_0 conda-forge
mysql-libs 9.2.0 he0572af_0 conda-forge
narwhals 1.35.0 pyh29332c3_0 conda-forge
natsort 8.4.0 pyh29332c3_1 conda-forge
ncurses 6.5 h2d0b736_3 conda-forge
nest-asyncio 1.6.0 pyhd8ed1ab_1 conda-forge
networkx 3.4.2 pyh267e887_2 conda-forge
numpy 2.2.4 py310hefbff90_0 conda-forge
openjpeg 2.5.3 h5fbd93e_0 conda-forge
openldap 2.6.9 he970967_0 conda-forge
openssl 3.5.0 h7b32b05_0 conda-forge
opt_einsum 3.4.0 pyhd8ed1ab_1 conda-forge
packaging 24.2 pyhd8ed1ab_2 conda-forge
palettable 3.3.3 pyhd8ed1ab_1 conda-forge
pandas 2.2.3 py310h5eaa309_3 conda-forge
parso 0.8.4 pyhd8ed1ab_1 conda-forge
patsy 1.0.1 pyhd8ed1ab_1 conda-forge
pcre2 10.44 hba22ea6_2 conda-forge
pexpect 4.9.0 pyhd8ed1ab_1 conda-forge
pickleshare 0.7.5 pyhd8ed1ab_1004 conda-forge
pillow 11.1.0 py310h7e6dc6c_0 conda-forge
pip 25.0.1 pyh8b19718_0 conda-forge
pixman 0.44.2 h29eaf8c_0 conda-forge
platformdirs 4.3.7 pyh29332c3_0 conda-forge
plotly 6.0.1 pyhd8ed1ab_0 conda-forge
prompt-toolkit 3.0.51 pyha770c72_0 conda-forge
propcache 0.3.1 py310h89163eb_0 conda-forge
protobuf 5.29.3 py310hcba5963_0 conda-forge
psutil 7.0.0 py310ha75aee5_0 conda-forge
pthread-stubs 0.4 hb9d3cd8_1002 conda-forge
ptyprocess 0.7.0 pyhd8ed1ab_1 conda-forge
pure_eval 0.2.3 pyhd8ed1ab_1 conda-forge
pybtex 0.24.0 pyhd8ed1ab_3 conda-forge
pycparser 2.22 pyh29332c3_1 conda-forge
pydantic 2.11.3 pyh3cfb1c2_0 conda-forge
pydantic-core 2.33.1 py310hc1293b2_0 conda-forge
pygments 2.19.1 pyhd8ed1ab_0 conda-forge
pymatgen 2025.4.17 py310hf462985_0 conda-forge
pyparsing 3.2.3 pyhd8ed1ab_1 conda-forge
pyside6 6.9.0 py310hfd10a26_0 conda-forge
pysocks 1.7.1 pyha55dd90_7 conda-forge
python 3.10.17 hd6af730_0_cpython conda-forge
python-dateutil 2.9.0.post0 pyhff2d567_1 conda-forge
python-tzdata 2025.2 pyhd8ed1ab_0 conda-forge
python_abi 3.10 6_cp310 conda-forge
pytz 2025.2 pyhd8ed1ab_0 conda-forge
pyyaml 6.0.2 py310h89163eb_2 conda-forge
pyzmq 26.4.0 py310h71f11fc_0 conda-forge
qhull 2020.2 h434a139_5 conda-forge
qt6-main 6.9.0 h6441bc3_1 conda-forge
readline 8.2 h8c095d6_2 conda-forge
requests 2.32.3 pyhd8ed1ab_1 conda-forge
ruamel.yaml 0.18.10 py310ha75aee5_0 conda-forge
ruamel.yaml.clib 0.2.8 py310ha75aee5_1 conda-forge
scikit-learn 1.6.1 py310h27f47ee_0 conda-forge
scipy 1.15.2 py310h1d65ade_0 conda-forge
seaborn 0.13.2 hd8ed1ab_3 conda-forge
seaborn-base 0.13.2 pyhd8ed1ab_3 conda-forge
sentry-sdk 2.26.1 pyhd8ed1ab_0 conda-forge
setproctitle 1.3.5 py310ha75aee5_0 conda-forge
setuptools 78.1.0 pyhff2d567_0 conda-forge
sevenn 0.11.0 pypi_0 pypi
six 1.17.0 pyhd8ed1ab_0 conda-forge
smmap 5.0.2 pyhd8ed1ab_0 conda-forge
spglib 2.6.0 py310h1e9006d_0 conda-forge
stack_data 0.6.3 pyhd8ed1ab_1 conda-forge
statsmodels 0.14.4 py310hf462985_0 conda-forge
sympy 1.13.3 pyh2585a3b_105 conda-forge
tabulate 0.9.0 pyhd8ed1ab_2 conda-forge
threadpoolctl 3.6.0 pyhecae5ae_0 conda-forge
tk 8.6.13 noxft_h4845f30_101 conda-forge
torch 1.12.0+cu116 pypi_0 pypi
torchaudio 0.12.0+cu116 pypi_0 pypi
torchvision 0.13.0+cu116 pypi_0 pypi
tornado 6.4.2 py310ha75aee5_0 conda-forge
tqdm 4.67.1 pyhd8ed1ab_1 conda-forge
traitlets 5.14.3 pyhd8ed1ab_1 conda-forge
typing 3.10.0.0 pyhd8ed1ab_2 conda-forge
typing-extensions 4.13.2 h0e9735f_0 conda-forge
typing-inspection 0.4.0 pyhd8ed1ab_0 conda-forge
typing_extensions 4.13.2 pyh29332c3_0 conda-forge
tzdata 2025b h78e105d_0 conda-forge
uncertainties 3.2.2 pyhd8ed1ab_2 conda-forge
unicodedata2 16.0.0 py310ha75aee5_0 conda-forge
urllib3 2.4.0 pyhd8ed1ab_0 conda-forge
wandb 0.19.9 py310h69e60fc_0 conda-forge
wayland 1.23.1 h3e06ad9_0 conda-forge
wcwidth 0.2.13 pyhd8ed1ab_1 conda-forge
werkzeug 3.1.3 pyhd8ed1ab_1 conda-forge
wheel 0.45.1 pyhd8ed1ab_1 conda-forge
widgetsnbextension 4.0.14 pyhd8ed1ab_0 conda-forge
xcb-util 0.4.1 hb711507_2 conda-forge
xcb-util-cursor 0.1.5 hb9d3cd8_0 conda-forge
xcb-util-image 0.4.0 hb711507_2 conda-forge
xcb-util-keysyms 0.4.1 hb711507_0 conda-forge
xcb-util-renderutil 0.3.10 hb711507_0 conda-forge
xcb-util-wm 0.4.2 hb711507_0 conda-forge
xkeyboard-config 2.43 hb9d3cd8_0 conda-forge
xorg-libice 1.1.2 hb9d3cd8_0 conda-forge
xorg-libsm 1.2.6 he73a12e_0 conda-forge
xorg-libx11 1.8.12 h4f16b4b_0 conda-forge
xorg-libxau 1.0.12 hb9d3cd8_0 conda-forge
xorg-libxcomposite 0.4.6 hb9d3cd8_2 conda-forge
xorg-libxcursor 1.2.3 hb9d3cd8_0 conda-forge
xorg-libxdamage 1.1.6 hb9d3cd8_0 conda-forge
xorg-libxdmcp 1.1.5 hb9d3cd8_0 conda-forge
xorg-libxext 1.3.6 hb9d3cd8_0 conda-forge
xorg-libxfixes 6.0.1 hb9d3cd8_0 conda-forge
xorg-libxi 1.8.2 hb9d3cd8_0 conda-forge
xorg-libxrandr 1.5.4 hb9d3cd8_0 conda-forge
xorg-libxrender 0.9.12 hb9d3cd8_0 conda-forge
xorg-libxtst 1.2.5 hb9d3cd8_3 conda-forge
xorg-libxxf86vm 1.1.6 hb9d3cd8_0 conda-forge
yaml 0.2.5 h7f98852_2 conda-forge
yarl 1.18.3 py310h89163eb_1 conda-forge
zeromq 4.3.5 h3b0a872_7 conda-forge
zipp 3.21.0 pyhd8ed1ab_1 conda-forge
zstandard 0.23.0 py310ha75aee5_1 conda-forge
zstd 1.5.7 hb8e6e7a_2 conda-forge
Possible causes
The problem appears to arise from my version of torch.serialization.load not supporting the weights_only argument and improperly passing it along to pickle.Unpickler. One source recommends downgrading torch to 1.12 (which is what I'm running currently), but it looks like weights_only only became available in PyTorch 1.13.
Is it correct that PyTorch/1.12.0 + CUDA/11.6.2 is no longer a supported environment for SevenNet?