Unable to train textual inversions #2638
Unanswered
ClaraSanders
asked this question in
Q&A
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
I can't train any textual inversions, I used to do it before. I reinstalled the latest version and tried different CPUS (I have Colab Pro) and I always get this error the moment I start the training:
Training textual inversion [Epoch 0: 7/13] loss: 0.1800313: 0% 6/5000 [00:10<1:25:39, 1.03s/it]*** Error training embedding
Traceback (most recent call last):
File "/content/gdrive/MyDrive/sd/stable-diffusion-webui/modules/textual_inversion/textual_inversion.py", line 549, in train_embedding
scaler.step(optimizer)
File "/usr/local/lib/python3.10/dist-packages/torch/cuda/amp/grad_scaler.py", line 413, in step
len(optimizer_state["found_inf_per_device"]) > 0
AssertionError: No inf checks were recorded for this optimizer.
Maybe I'm missing some crucial step because I trained my last textual inversion about two months ago, I'm not sure, to be honest.
Beta Was this translation helpful? Give feedback.
All reactions