unexpected keyword argument 'use_gpu'

[bobermayer]

Hi,
I'm trying to run the tutorial, but it fails during vae.train() with the message Trainer.__init__() got an unexpected keyword argument 'use_gpu'

TypeError                                 Traceback (most recent call last)
Cell In[5], line 3
      1 VELOVI.setup_anndata(adata, spliced_layer="Ms", unspliced_layer="Mu")
      2 vae = VELOVI(adata)
----> 3 vae.train()

File ~/miniconda3/envs/brain_organoids/lib/python3.11/site-packages/velovi/_model.py:187, in VELOVI.train(self, max_epochs, lr, weight_decay, use_gpu, train_size, validation_size, batch_size, early_stopping, gradient_clip_val, plan_kwargs, **trainer_kwargs)
    183 es = "early_stopping"
    184 trainer_kwargs[es] = (
    185     early_stopping if es not in trainer_kwargs.keys() else trainer_kwargs[es]
    186 )
--> 187 runner = TrainRunner(
    188     self,
    189     training_plan=training_plan,
    190     data_splitter=data_splitter,
    191     max_epochs=max_epochs,
    192     use_gpu=use_gpu,
    193     **trainer_kwargs,
    194 )
    195 return runner()

File ~/miniconda3/envs/brain_organoids/lib/python3.11/site-packages/scvi/train/_trainrunner.py:82, in TrainRunner.__init__(self, model, training_plan, data_splitter, max_epochs, accelerator, devices, **trainer_kwargs)
     79 if getattr(self.training_plan, "reduce_lr_on_plateau", False):
     80     trainer_kwargs["learning_rate_monitor"] = True
---> 82 self.trainer = self._trainer_cls(
     83     max_epochs=max_epochs,
     84     accelerator=accelerator,
     85     devices=lightning_devices,
     86     **trainer_kwargs,
     87 )
     88 # currently set for MetricsCallback
     89 self.trainer._model = model

File ~/miniconda3/envs/brain_organoids/lib/python3.11/site-packages/scvi/train/_trainer.py:171, in Trainer.__init__(self, accelerator, devices, benchmark, check_val_every_n_epoch, max_epochs, default_root_dir, enable_checkpointing, checkpointing_monitor, num_sanity_val_steps, enable_model_summary, early_stopping, early_stopping_monitor, early_stopping_min_delta, early_stopping_patience, early_stopping_mode, additional_val_metrics, enable_progress_bar, progress_bar_refresh_rate, simple_progress_bar, logger, log_every_n_steps, learning_rate_monitor, **kwargs)
    168 if logger is None:
    169     logger = SimpleLogger()
--> 171 super().__init__(
    172     accelerator=accelerator,
    173     devices=devices,
    174     benchmark=benchmark,
    175     check_val_every_n_epoch=check_val_every_n_epoch,
    176     max_epochs=max_epochs,
    177     default_root_dir=default_root_dir,
    178     enable_checkpointing=enable_checkpointing,
    179     num_sanity_val_steps=num_sanity_val_steps,
    180     enable_model_summary=enable_model_summary,
    181     logger=logger,
    182     log_every_n_steps=log_every_n_steps,
    183     enable_progress_bar=enable_progress_bar,
    184     callbacks=callbacks,
    185     **kwargs,
    186 )

File ~/miniconda3/envs/brain_organoids/lib/python3.11/site-packages/lightning/pytorch/utilities/argparse.py:70, in _defaults_from_env_vars.<locals>.insert_env_defaults(self, *args, **kwargs)
     67 kwargs = dict(list(env_variables.items()) + list(kwargs.items()))
     69 # all args were already moved to kwargs
---> 70 return fn(self, **kwargs)

TypeError: Trainer.__init__() got an unexpected keyword argument 'use_gpu'

I'm not able to use vae.train(use_gpu=False) either (same error).
any ideas?
thanks for your help!

here are my versions

sc.logging.print_versions()
-----
anndata     0.10.5.post1
scanpy      1.9.8
-----
IPython             8.21.0
PIL                 10.2.0
absl                NA
aiohttp             3.9.3
aiosignal           1.3.1
annotated_types     0.6.0
anyio               NA
asttokens           NA
attr                23.2.0
backoff             2.2.1
brotli              1.1.0
bs4                 4.12.3
certifi             2024.02.02
cffi                1.16.0
charset_normalizer  3.3.2
chex                0.1.8
click               8.1.7
colorama            0.4.6
comm                0.2.1
contextlib2         NA
croniter            NA
cycler              0.12.1
cython_runtime      NA
dateutil            2.8.2
decorator           5.1.1
deepdiff            6.7.1
defusedxml          0.7.1
docrep              0.3.2
etils               1.7.0
executing           2.0.1
fastapi             0.109.2
flax                0.8.1
frozenlist          1.4.1
fsspec              2024.2.0
gmpy2               2.1.2
google              NA
h5py                3.10.0
idna                3.6
importlib_resources NA
ipywidgets          8.1.2
jax                 0.4.23
jaxlib              0.4.23.dev20240125
jedi                0.19.1
joblib              1.3.2
kiwisolver          1.4.5
lightning           2.0.9.post0
lightning_cloud     NA
lightning_utilities 0.10.1
llvmlite            0.42.0
matplotlib          3.8.3
ml_collections      NA
ml_dtypes           0.3.2
mpi4py              3.1.5
mpl_toolkits        NA
mpmath              1.3.0
msgpack             1.0.7
mudata              0.2.3
multidict           6.0.5
multipart           0.0.9
multipledispatch    0.6.0
natsort             8.4.0
numba               0.59.0
numpy               1.26.4
numpyro             0.13.2
opt_einsum          v3.3.0
optax               0.1.9
ordered_set         4.1.0
orjson              3.9.10
packaging           23.2
pandas              2.2.0
parso               0.8.3
patsy               0.5.6
pickleshare         0.7.5
pkg_resources       NA
prompt_toolkit      3.0.42
psutil              5.9.8
pure_eval           0.2.2
pycparser           2.21
pydantic            2.1.1
pydantic_core       2.4.0
pygments            2.17.2
pynndescent         0.5.11
pyparsing           3.1.1
pyro                1.9.0+f02dfb9
pytz                2024.1
requests            2.31.0
rich                NA
scipy               1.12.0
scvelo              0.3.1
scvi                1.1.1
seaborn             0.13.2
session_info        1.0.0
six                 1.16.0
sklearn             1.1.3
sniffio             1.3.0
socks               1.7.1
soupsieve           2.5
sparse              0.15.1
sphinxcontrib       NA
stack_data          0.6.2
starlette           0.36.3
statsmodels         0.14.1
sympy               1.12
threadpoolctl       3.3.0
toolz               0.12.1
torch               2.1.2.post301
torchgen            NA
torchmetrics        1.2.1
tqdm                4.66.2
traitlets           5.14.1
tree                0.1.8
typing_extensions   NA
umap                0.5.5
urllib3             2.2.1
uvicorn             0.27.1
velovi              0.3.0
wcwidth             0.2.13
websocket           1.7.0
websockets          12.0
wrapt               1.16.0
yaml                6.0.1
yarl                1.9.4
-----
Python 3.11.8 | packaged by conda-forge | (main, Feb 16 2024, 20:53:32) [GCC 12.3.0]
Linux-6.1.0-1033-oem-x86_64-with-glibc2.35
-----
Session information updated at 2024-03-04 13:31

[lzygenomics]

I have the same problem : )

[bobermayer]

ok so it looks like use_gpu was removed from scvi in version 1.1.0. I cloned the velovi repo and commented out the use_gpu line in the call to TrainRunner as a quick fix. now it works.

[martinkim0]

Hi, sorry about this - I've added a fix in #24 and will be re-releasing.

[martinkim0]

Going to close this issue now since 0.3.1 includes a fix.