-
What % is this project?
-
### š Describe the bug
```
import torch
from diffusers import EulerDiscreteScheduler, StableDiffusionPipeline
model_id = "stabilityai/stable-diffusion-2"
scheduler = EulerDiscreteScheduler.fromā¦
-
![asdf](https://user-images.githubusercontent.com/6880780/138640139-445c2865-6239-4e12-9bba-810f9c7eb6d8.png)
Using version 1.3.0, setting ScalingMode to 1 results in Bilinear filtering instead of ā¦
-
Making this issue here since your [multimc m1 branch](https://github.com/r58Playz/m1-multimc-hack.git) doesn't seem to be able to have issues added to it; not sure if this is a LWJGL2-m1 issue or a JVā¦
-
See also #491.
I ran a decomp test suite with evp, eap, and vp-picard. I did not test vp-anderson as that is not ready to use out of the box. Below are the results. evp and eap pass all tests wiā¦
-
Needs some rewriting to accurately reflect goal of trotterization. Instead of forcing constraint of template to be proportional to full length iSwap, just let template be as long as it needs to make aā¦
-
Only seems to be a problem in some cases. Suggests a subtle issue. Full test suite on cheyenne bfbcomp results showed the following (same results for pgi and intel). Many tests pass but not all. ā¦
-
### š Describe the bug
Run into this issue with torchdynamo/benchmarks/huggingface.py with a custom stack (nvprim).
Tried to get a small repro but failed. However, I did manage to get a differā¦
-
### š Describe the bug
`test_python_reference_consistency` appears to assume that the numerics of CPU and CUDA are implemented equivalently. This is not true and leads to a lot of tolerance overridesā¦
-
I had some trouble with our new compiler installation that does not like to be initialized twice, which is the case with tests (and thus suites) because `env.$machine_$compiler` is sourced by `cice.teā¦