github/ROCm
1
1
Fork 0
mirror of https://github.com/ROCm/ROCm.git synced 2026-04-05 03:01:17 -04:00
Code Issues Packages Projects Releases Wiki Activity

[DOCS] Fixing docs (#2175)

Browse Source
This commit is contained in:
Zahi Moudallal
2023-08-24 15:58:59 -07:00
committed by GitHub
parent 387c8d94e1
commit 120ce0a5bf
3 changed files with 13 additions and 12 deletions
Download Patch File Download Diff File Expand all files Collapse all files

10
python/tutorials/09-experimental-tma-matrix-multiplication.py
View File

@@ -32,6 +32,11 @@ from torch.testing import assert_close
import triton
import triton.language as tl
if torch.cuda.get_device_capability()[0] < 9:
import sys
print("Skipping TMA benchmark for GPU with compute capability < 9")
sys.exit(0)
@triton.autotune(
configs=[
@@ -191,10 +196,5 @@ def benchmark(M, N, K, TRANS_A, TRANS_B, provider):
return perf(ms), perf(max_ms), perf(min_ms)
if torch.cuda.get_device_capability()[0] < 9:
import sys
print("Skipping TMA benchmark for GPU with compute capability < 9")
sys.exit(0)
test_matmul()
benchmark.run(show_plots=False, print_data=True)

10
python/tutorials/10-experimental-tma-store-matrix-multiplication.py
View File

@@ -32,6 +32,11 @@ from torch.testing import assert_close
import triton
import triton.language as tl
if torch.cuda.get_device_capability()[0] < 9:
import sys
print("Skipping TMA benchmark for GPU with compute capability < 9")
sys.exit(0)
@triton.autotune(
configs=[
@@ -171,9 +176,4 @@ def benchmark(M, N, K, provider):
return perf(ms), perf(max_ms), perf(min_ms)
if torch.cuda.get_device_capability()[0] < 9:
import sys
print("Skipping TMA benchmark for GPU with compute capability < 9")
sys.exit(0)
benchmark.run(show_plots=False, print_data=True)