I can't solve these dependency problems
AlexGithubUser opened this issue · comments
Oryx 0.2.1 is looking for Jaxlib 0.1.68 which doesn't exist. Its not on pypi.org.
ERROR: Could not find a version that satisfies the requirement jaxlib==0.1.68 (from oryx) (from versions: none)
ERROR: No matching distribution found for jaxlib==0.1.68
I have tryed installing different versions of oryx but then they require different versions of Jaxlib and I have been repeatedly defeated trying to install the correct iteration of jax or jaxlib. I'll try some more when I have more time and patience.
I have the same problem, jaxlib version 0.1.68 is no longer available on PyPi, so I can't install it directly from PyPi via pip.
same error
same error
same error :(
conda create -n freenerf -c conda-forge python=3.6.15
conda activate freenerf
pip install --upgrade jaxlib==0.1.68+cuda110 -f https://storage.googleapis.com/jax-releases/jax_cuda_releases.html
pip install -r requirements.txt
requirements.txt：
gin-config==0.5.0
absl-py==0.15.0
six==1.15.0
decorator==4.3
torch==1.10.1
typing-extensions==3.10.0
flax==0.3.5
tensorflow==2.6.2
opencv-python==4.5.5.62
oryx==0.2.1
jax==0.2.16
scikit-image==0.17.2
dm-pix==0.3.0
tqdm
wandb
lpips
mediapy==1.0.0
if the following two incompatibilities appear, don't worry， you still can train
Hi guys, sorry for this super late reply. But I think this issue can be addressed by following this thread: google/jax#11142 (comment)
conda create -n freenerf -c conda-forge python=3.6.15 conda activate freenerf pip install --upgrade jaxlib==0.1.68+cuda110 -f https://storage.googleapis.com/jax-releases/jax_cuda_releases.html pip install -r requirements.txt
requirements.txt： gin-config==0.5.0 absl-py==0.15.0 six==1.15.0 decorator==4.3 torch==1.10.1 typing-extensions==3.10.0 flax==0.3.5 tensorflow==2.6.2 opencv-python==4.5.5.62 oryx==0.2.1 jax==0.2.16 scikit-image==0.17.2 dm-pix==0.3.0 tqdm wandb lpips mediapy==1.0.0 if the following two incompatibilities appear, don't worry， you still can train
Hi,
Could you please share the versions of dm-pix and chex you have installed? dm-pix is dependent on chex, which for some reason, tries to reference functions in newer versions of jax. Even after many trials, I can't seem to find the right combination of versions of these packages that are compatible with the required versions of flax, jax, and jaxlib.
conda create -n freenerf -c conda-forge python=3.6.15 conda activate freenerf pip install --upgrade jaxlib==0.1.68+cuda110 -f https://storage.googleapis.com/jax-releases/jax_cuda_releases.html pip install -r requirements.txt
requirements.txt： gin-config==0.5.0 absl-py==0.15.0 six==1.15.0 decorator==4.3 torch==1.10.1 typing-extensions==3.10.0 flax==0.3.5 tensorflow==2.6.2 opencv-python==4.5.5.62 oryx==0.2.1 jax==0.2.16 scikit-image==0.17.2 dm-pix==0.3.0 tqdm wandb lpips mediapy==1.0.0 if the following two incompatibilities appear, don't worry， you still can trainHi,
Could you please share the versions of
dm-pixandchexyou have installed?dm-pixis dependent onchex, which for some reason, tries to reference functions in newer versions of jax. Even after many trials, I can't seem to find the right combination of versions of these packages that are compatible with the required versions offlax,jax, andjaxlib.
@mvp18 Hey, bro. As python3.6 is not supported by vscode remote debugger, I've tried run the work with python 3.7 environment, and now I am running perfectly! I followed the requirements provided by @annci62 , and always got stuck with the error "AttributeError: module 'jax.random' has no attribute 'KeyArray'". The most solution to solve this problem is downgrading the jax lib from newest to lower than 0.4.23 version. However, our installed jax and jaxlib versions are 0.2.16 and 0.1.68+cuda110, which is far earlier than 0.4.23. So I geuss the attribute "KeyArray" might be added at a certain newer version. Since the error is reported by chex lib, downgrading chex verison may could prevent it from using "not supported" attribute. When I downgrade my chex lib version to 0.1.0, it works!
Here is my pip list:
Package Version
absl-py 0.15.0
astunparse 1.6.3
backcall 0.2.0
cached-property 1.5.2
cachetools 4.2.4
certifi 2022.12.7
charset-normalizer 3.3.2
chex 0.1.0
clang 5.0
click 8.1.7
cloudpickle 2.2.1
cycler 0.11.0
decorator 4.3.0
dm-pix 0.3.0
dm-tree 0.1.8
docker-pycreds 0.4.0
flatbuffers 1.12
flax 0.3.5
fonttools 4.38.0
gast 0.4.0
gin-config 0.5.0
gitdb 4.0.11
GitPython 3.1.18
google-auth 1.35.0
google-auth-oauthlib 0.4.6
google-pasta 0.2.0
grpcio 1.62.2
h5py 3.1.0
idna 3.7
imageio 2.31.2
importlib-metadata 6.7.0
ipython 7.34.0
jax 0.2.16
jaxlib 0.1.68+cuda110
jedi 0.19.1
keras 2.6.0
Keras-Preprocessing 1.1.2
kiwisolver 1.4.5
lpips 0.1.4
Markdown 3.4.4
MarkupSafe 2.1.5
matplotlib 3.5.3
matplotlib-inline 0.1.6
mediapy 1.0.0
msgpack 1.0.5
networkx 2.6.3
numpy 1.19.5
oauthlib 3.2.2
opencv-python 4.5.5.62
opt-einsum 3.3.0
optax 0.0.9
oryx 0.2.1
packaging 24.0
parso 0.8.4
pexpect 4.9.0
pickleshare 0.7.5
Pillow 9.5.0
pip 22.3.1
platformdirs 2.6.1
prompt_toolkit 3.0.47
protobuf 3.20.0
psutil 6.0.0
ptyprocess 0.7.0
pyasn1 0.5.1
pyasn1-modules 0.3.0
Pygments 2.17.2
pyparsing 3.1.2
python-dateutil 2.9.0.post0
PyWavelets 1.3.0
PyYAML 6.0.1
requests 2.31.0
requests-oauthlib 2.0.0
rsa 4.9
scikit-image 0.17.2
scipy 1.7.3
sentry-sdk 2.8.0
setproctitle 1.3.3
setuptools 65.6.3
six 1.15.0
smmap 5.0.1
tensorboard 2.6.0
tensorboard-data-server 0.6.1
tensorboard-plugin-wit 1.8.1
tensorflow 2.6.2
tensorflow-estimator 2.6.0
termcolor 1.1.0
tfp-nightly 0.14.0.dev20210630
tifffile 2021.11.2
toolz 0.12.1
torch 1.10.1
torchvision 0.11.2
tqdm 4.66.4
traitlets 5.9.0
typing-extensions 3.7.4
urllib3 2.0.7
wandb 0.17.4
wcwidth 0.2.13
Werkzeug 2.2.3
wheel 0.38.4
wrapt 1.12.1
zipp 3.15.0