You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The latest Triton build (3.1.0) throws the following error when using bitpacked data inside a loop with tl.dot:
LLVM ERROR: mma16816 data type not supported
This error happens on Ampere and Hopper, but not on older gpus like the Titan RTX/2080 Ti.
The bitpacked data is read with indices in the form offs_k[:, None] // num_elements, something like [0,0,0...1,1,1...64,64,64].
I have faced this error in the previous build and I found that replacing for k in range(0, total_blocks_k, 1): with for k in tl.range(0, total_blocks_k, 1, num_stages=1): solved the issue, but this trick no longer works with 3.1.0.
The latest Triton build (3.1.0) throws the following error when using bitpacked data inside a loop with
tl.dot
:This error happens on Ampere and Hopper, but not on older gpus like the Titan RTX/2080 Ti.
The bitpacked data is read with indices in the form
offs_k[:, None] // num_elements
, something like[0,0,0...1,1,1...64,64,64]
.I have faced this error in the previous build and I found that replacing
for k in range(0, total_blocks_k, 1):
withfor k in tl.range(0, total_blocks_k, 1, num_stages=1):
solved the issue, but this trick no longer works with 3.1.0.Here's a full-script to reproduce it.
https://gist.github.com/mobicham/f9eba3c07f7e497ae622194a9c5e4822
The text was updated successfully, but these errors were encountered: