Error: ModuleNotFoundError: No module named 'numpy', while trying to train Textual Inversion

[joshdance]

Is there an existing issue for this?

• [X] I have searched the existing issues

OS

macOS

GPU

mps

VRAM

No response

What happened?

On an M1 Mac, attempted to use the command mentioned here, https://github.com/invoke-ai/InvokeAI/pull/814, to start training using Textual Inversion

This results in this error in the Terminal:

Traceback (most recent call last):
  File "/Users/adminadmin/invokeAI/./main.py", line 2, in <module>
    import numpy as np
ModuleNotFoundError: No module named 'numpy'
zsh: command not found: --base
zsh: command not found: --actual_resume
zsh: command not found: --data_root
zsh: command not found: --gpus
zsh: command not found: --no_test

I used pip3 to install numpy and then ran it again. This time it gives the error ModuleNotFoundError: No module named 'torch'

Seems like I a missing dependencies to be able to run Textual Inversion. Is there a script to run to set it up? Or a list of dependencies that I could check?

Screenshots

No response

Additional context

No response

Contact Details

Twitter @joshdance

[hades2510]

Are you using the conda env that was created when running installing InvokeAI? I had a similar problem, but I was trying to use a different python version and .env wasn't configured for this.

[joshdance]

@hades2510 Not sure if I'm using the conda env that was created when running installing InvokeAI.

How would I check? I'm a bit rusty on command line techniques. :)

[github-actions[bot]]

There has been no activity in this issue for 14 days. If this issue is still being experienced, please reply with an updated confirmation that the issue is still being experienced with the latest release.