-
Thanks for the great repo!
We are testing the correctness of `flash_attn_varlen_func` when enabling the cuda graph.
This is the test we use.
https://github.com/vllm-project/vllm/blob/66e832be41cd3f…
-
So trying to flash your precompiled program with flash.bat on win10.
I installed python and esptool.py and in python environment I get version of esptool.
But when running flash.bat I get error:
…
-
## Problem Description
`FileStream::openAsync()` call inside `Worker` cause hang for iOS patform using AIR 51.1.
Tested with AIR 51.1.1.1 and AIR 51.1.1.2 with multiple different iOS devices with …
-
### Device
Novacustom V560TNE
### RTE version
v0.9.1-rc1
### OSFV version
develop
### Affected component(s) or functionality
any test which flashes fw
### Brief summary
The fw image is not c…
-
### What happened?
I can't detect my HackRF on any Computers to use a batch file or even a cmd prompt to flash the new updates
### Your hardware
Portapack HackRF One H2
### Tests
- [X] I managed …
-
### Describe the bug
When setting the `--save-directory` option in Ruffle's Windows 64-bit version to a path containing relative directory components (e.g., `C:\path\to\..\save`, which resolves to `C…
-
_Downstream PyTorch issue:_
https://github.com/pytorch/pytorch/issues/133780
I'm trying to do attention on a batch-of-zero, because my program uses a static graph and I rely on zero-batching (in…
-
when i test flash-occ ,I encountered the following problem:
![error](https://github.com/user-attachments/assets/be4f4089-ceba-489d-9e53-ca6deb4a0a7a)
and with the inference time, the gpu memory gro…
-
# Summary
Recently a very large number of FlashAttention tests have become flaky. I am having a very hard time reproducing this in my local environment.
Example Tests:
- https://github.com/pytorc…
-
Hi, I have enabled bfloat16 testing in Triton (https://github.com/openai/triton/pull/1244/), but I'm getting this error with this data type
```
giorgio@giorgio:triton$ pytest python/test/unit/oper…