[bug]:

[fandongminghitmt]

Is there an existing issue for this problem?

• I have searched the existing issues

Operating system

Windows

GPU vendor

Nvidia (CUDA)

GPU model

5060ti

GPU VRAM

16

Version number

5.12

Browser

chrome

Python dependencies

Starting up...
Installation parameters:

• Install location: F:\InvokeAI
• GPU type: nvidia>=30xx
• Torch Platform: cuda
  Repair mode enabled:
• Force-reinstalling python
• Deleting and recreating virtual environment
  Reinstalling Python...

C:\Users\Administrator\AppData\Local\Temp\2vFxKvvFbjMAsUlvZ94N07dukNe\resources\bin\uv.exe python install 3.11 --python-preference only-managed --reinstall
Downloading cpython-3.11.11-windows-x86_64-none (24.3MiB)
Downloaded cpython-3.11.11-windows-x86_64-none
Installed Python 3.11.11 in 9.84s
~ cpython-3.11.11-windows-x86_64-none
Creating virtual environment...
C:\Users\Administrator\AppData\Local\Temp\2vFxKvvFbjMAsUlvZ94N07dukNe\resources\bin\uv.exe venv --relocatable --prompt invoke --python 3.11 --python-preference only-managed F:\InvokeAI.venv
Using CPython 3.11.11
Creating virtual environment at: F:\InvokeAI.venv
Activate with: F:\InvokeAI.venv\Scripts\activate
Installing invokeai package...
C:\Users\Administrator\AppData\Local\Temp\2vFxKvvFbjMAsUlvZ94N07dukNe\resources\bin\uv.exe pip install --python 3.11 --python-preference only-managed invokeai==v5.12.0rc1 --force-reinstall --compile-bytecode --index=https://download.pytorch.org/whl/cu124
× No solution found when resolving dependencies:
╰─▶ Because only torch<=2.6.0+cu124 is available and invokeai==5.12.0rc1
depends on torch>=2.7.0,<2.8.dev0, we can conclude that
invokeai==5.12.0rc1 cannot be used.
And because you require invokeai==5.12.0rc1, we can conclude that your
requirements are unsatisfiable.
hint: torch was requested with a pre-release marker (e.g.,
torch>=2.7.0,<2.8.dev0), but pre-releases weren't enabled (try:
--prerelease=allow)
hint: torch was found on https://download.pytorch.org/whl/cu124,
but not at the requested version (torch>=2.7.0,<2.8.dev0). A
compatible version may be available on a subsequent index (e.g.,
https://pypi.org/simple). By default, uv will only consider versions
that are published on the first index that contains a given package, to
avoid dependency confusion attacks. If all indexes are equally trusted,
use --index-strategy unsafe-best-match to consider all versions from
all indexes, regardless of the order in which they were defined.
Failed to install invokeai python package: Process exited with code 1
Try installing again with Repair mode enabled to fix this.
Ask for help on Discord or GitHub if you continue to have issues.

What happened

can't install

What you expected to happen

reinstall

How to reproduce the problem

No response

Additional context

No response

Discord username

No response

[Pavel-hb]

The same mistake. GPU model 3090

[Kahldris]

edit: downloading the latest Invoke Community Edition.exe and then doing a repair install fixed it. If this was expected of me in the update process then I missed it being mentioned anywhere, since I'm used to the launcher handling all updates itself.

---

Yikes yeah I just got this trying to upgrade 5.11 to 5.12. Trying repair mode like it told me to screwed me as well, because now I can't even use my 5.11 install any more.

× No solution found when resolving dependencies:
╰─▶ Because only torch<=2.6.0+cu124 is available and invokeai==5.12.0
depends on torch>=2.7.0,<2.8.dev0, we can conclude that invokeai==5.12.0
cannot be used.
And because you require invokeai==5.12.0, we can conclude that your
requirements are unsatisfiable.
hint: torch was requested with a pre-release marker (e.g.,
torch>=2.7.0,<2.8.dev0), but pre-releases weren't enabled (try:
--prerelease=allow)
hint: torch was found on https://download.pytorch.org/whl/cu124,
but not at the requested version (torch>=2.7.0,<2.8.dev0). A
compatible version may be available on a subsequent index (e.g.,
https://pypi.org/simple). By default, uv will only consider versions
that are published on the first index that contains a given package, to
avoid dependency confusion attacks. If all indexes are equally trusted,
use --index-strategy unsafe-best-match to consider all versions from
all indexes, regardless of the order in which they were defined.
Failed to install invokeai python package: Process exited with code 1