mrq/tortoise-tts
4
6
Fork 15
Code Issues 9 Pull Requests 1 Packages Projects Releases Wiki Activity

[py3.10] Error when creating conditioning latents #27

New Issue
Closed
opened 2023-02-12 16:09:22 +00:00 by lightmare · 3 comments
lightmare commented 2023-02-12 16:09:22 +00:00
Contributor
Copy Link

Marking this as py3.10 in case such reports are not of interest.
Running a py3.10 env with low vram and cuda, i am experiencing the following issue:

RuntimeError: Input type (torch.cuda.FloatTensor) and weight type (torch.FloatTensor) should be the same

Sounds like a mismatch between work done on CPU and on GPU?
So far tested on a2d95fe208.

Marking this as py3.10 in case such reports are not of interest. Running a py3.10 env with low vram and cuda, i am experiencing the following issue: ``` RuntimeError: Input type (torch.cuda.FloatTensor) and weight type (torch.FloatTensor) should be the same ``` Sounds like a mismatch between work done on CPU and on GPU? So far tested on a2d95fe2087c98cc1a27fa6ff04311a5f01aa1c2.
lightmare changed title from [py3.10] Error when creating conditing latents to [py3.10] Error when creating conditioning latents 2023-02-12 16:18:01 +00:00
lightmare commented 2023-02-12 16:26:41 +00:00
Author
Contributor
Copy Link

Still works on 25e70dce1a.

Still works on 25e70dce1a1ded42045e247788821e24f9d8a8d0.
mrq commented 2023-02-12 17:23:38 +00:00
Owner
Copy Link

I'll see about trying to replicate it. I'm very sure I forgot to actually check it working on Low VRAM itself, so I most definitely broke it.

I'll see about trying to replicate it. I'm very sure I forgot to actually check it working on `Low VRAM` itself, so I most definitely broke it.
mrq commented 2023-02-12 17:46:12 +00:00
Owner
Copy Link

Should be fixed in 2210b49cb6. I suppose the issue was that I never actually re-tested computing the latents outside of the CPU, as the DirectML forces it to be done entirely on CPU.

Didn't actually get to test Low VRAM on computing latents, as it was a problem without setting it. Tested both with and without Low VRAM, works fine now.

Should be fixed in 2210b49cb6df4a2a919f1773d20ab309d486cb0e. I suppose the issue was that I never actually re-tested computing the latents outside of the CPU, as the DirectML forces it to be done entirely on CPU. ~~Didn't actually get to test `Low VRAM` on computing latents, as it was a problem without setting it.~~ Tested both with and without `Low VRAM`, works fine now.
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
No items
No Label
Milestone
Clear milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
No Assignees
2 Participants
Notifications
Due Date
The due date is invalid or out of range. Please use the format 'yyyy-mm-dd'.

No due date set.

Dependencies

No dependencies set.

Reference: mrq/tortoise-tts#27
No description provided.
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?