This is the file which installs the GPU version

(This is why we recommend using pip, because it is stable and clearml-agent takes care of pytorch/cuda verions)

And the one with the CPU version? is it with "~=" or "="?

I tried "~=", "==" and "="

I will try a minimal version now

channels:
- defaults
- conda-forge
- pytorch
dependencies:
- cudatoolkit==11.1.1
- pytorch==1.8.0

Gives CPU version

Wtf? can you try with = (notice single not double)?

channels:
- defaults
- conda-forge
- pytorch
dependencies:
- cudatoolkit=11.1.1
- pytorch=1.8.0

Installs CPU

Ha?!

But here is the funny thing:

channels:
- pytorch
- conda-forge
- defaults
dependencies:
- cudatoolkit=11.1.1
- pytorch=1.8.0

Installs GPU

The ordering of the channels seems to matter!

WTF?!

I try it one more time just to make sure

okay, I'll make sure we order it correctly

Just tested again. The ordering definitly matters.

Damn, okay I'll make sure we fix the order.
Could you verify the ~= works as intended (if the order id correct)

@<1523701868901961728:profile|ReassuredTiger98> thank you so much for testing it!

channels:
- pytorch
- conda-forge
- defaults
dependencies:
- cudatoolkit~=11.1.1
- pytorch~=1.8.0

Works fine

Great, thanks!

No problem! I profit so much from clearml 🙂

Can you ping me when it is updated in None so I can update my installation?

Will do!

@<1523701868901961728:profile|ReassuredTiger98> what do you have in the clearml.conf under "conda_channels" ?
Is this it ?
None

conda_channels: ["defaults", "conda-forge", "pytorch", ]

Well, in that case, just change the order it should solve it (I'll make sure we have that as the default:

conda_channels: ["pytorch", "conda-forge", "defaults", ]

It should solve the issue 🙂

Perfect, will try it. fyi: The conda_channels that I used are from clearml-agent init

Yes that is exactly what I will make sure we change :)

Thank you! 🙂

One more thing: The cuda_version that clearml finds automatically is wrong.

I get 110 but it should be 111