issues
search
ROCm
/
composable_kernel
Composable Kernel: Performance Portable Programming Model for Machine Learning Tensor Operators
https://rocm.docs.amd.com/projects/composable_kernel/en/latest/
Other
251
stars
102
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
Bump rocm-docs-core from 1.2.0 to 1.2.1 in /docs/sphinx
#1322
dependabot[bot]
closed
4 weeks ago
0
Disable the hipTensor test in CI by default, only run once daily
#1321
illsilin
closed
4 weeks ago
0
Integrate universal gemm with conv forward
#1320
bartekxk
closed
3 weeks ago
0
Project team: tall_and_skinny_gemm_splitk Members: Victor Shih/ Wayne Huang/ Eric Kuo add third buffer for prefetching and do loop unrolling
#1319
csk1116
closed
1 month ago
0
Final Project PR for StreamK
#1318
abbyoutsider
opened
1 month ago
0
Final Project Submission - StreamK no padding optimization
#1317
ncr5012
opened
1 month ago
0
Tall and skinny gemm
#1316
samtam1118
closed
4 weeks ago
0
[ON-HOLD] Integrate universal gemm with conv bwd data
#1315
bartekxk
opened
1 month ago
1
Only generate the necessary kernel instances for Flash Attention integration
#1314
rocking5566
closed
4 weeks ago
1
Post-merge fix of PR 1300
#1313
zjing14
closed
1 month ago
0
Build CK library for all supported targets.
#1312
illsilin
closed
1 month ago
0
Bump rocm-docs-core from 1.1.3 to 1.2.0 in /docs/sphinx
#1311
dependabot[bot]
closed
1 month ago
0
Enable external CI pipeline triggers
#1310
amd-jmacaran
closed
1 month ago
0
Add structural sparsity gemm instruction tests
#1309
jakpiase
closed
5 days ago
0
Bump rocm-docs-core from 1.1.2 to 1.1.3 in /docs/sphinx
#1308
dependabot[bot]
closed
1 month ago
0
Add a convinvscale op, related instances and examples
#1307
geyyer
closed
3 weeks ago
0
Split the gemm_multi_abd instances.
#1306
illsilin
closed
1 month ago
0
Make the library which generates CK instances for pytorch2 inductor's CK backend usage
#1305
tenpercent
closed
1 month ago
0
Select appropriate GPU targets for instances, tests, and examples.
#1304
illsilin
closed
1 month ago
0
Optimize grouped conv bwd weight for small M and N
#1303
bartekxk
closed
1 month ago
0
Update the grid distribution for fmha forward [Performance]
#1302
qianfengz
closed
1 month ago
0
Change fmha forward grid distribution
#1301
qianfengz
closed
1 month ago
0
add f8 gemm multiD with both row/col wise scale
#1300
zjing14
closed
1 month ago
0
Move grouped conv fwd client examples
#1299
geyyer
closed
1 month ago
0
@bghimireamd [Informative]
#1298
junliume
opened
1 month ago
1
aggregate device macros in ck_tile config header.
#1297
illsilin
closed
1 month ago
0
Replace the ENV macro with CK_ENV to avoid conflicts
#1296
illsilin
closed
1 month ago
0
[CK_TILE] support group from cmdline
#1295
carlushuang
closed
1 month ago
1
Compile error in cpp_extension
#1294
rocking5566
opened
1 month ago
0
Bump rocm-docs-core from 1.1.1 to 1.1.2 in /docs/sphinx
#1293
dependabot[bot]
closed
1 month ago
0
Fix compile error
#1292
rocking5566
closed
1 month ago
0
remove operator-deref
#1291
carlushuang
closed
1 month ago
0
Fix issue #1276.
#1290
illsilin
closed
1 month ago
0
re-enable convnd_fwd_xdl_fp64 testing
#1289
illsilin
closed
1 month ago
0
compile error use ck_tile in fmha on develop branch
#1288
flyingdown
closed
1 month ago
12
[CK_TILE] fix some rand number init
#1287
carlushuang
closed
1 month ago
0
CK Tile FA Training kernels
#1286
danyao12
closed
4 weeks ago
1
Code clean-up
#1285
illsilin
closed
1 month ago
0
Fix MakeArgument in GroupedGEMM multiple D tile loop
#1284
aosewski
closed
1 month ago
0
Change output gemm type to AccDataType in two stage conv bwd wei
#1283
bartekxk
closed
1 month ago
0
fix the output formatting for staging compiler.
#1282
illsilin
closed
1 month ago
0
Change to make seqlen_k == 0 be considered
#1281
qianfengz
closed
1 month ago
0
Add two stage grouped conv bwd weight kernel
#1280
bartekxk
closed
1 month ago
0
threadwise_tensor_slice_transfer_v5r1 issue
#1279
joye
opened
1 month ago
0
Enable logging in CK with environment variable.
#1278
illsilin
closed
1 month ago
2
Add ROCm Doc team as codeowners for RTD yaml
#1277
samjwu
closed
1 month ago
0
blockwise_gemm_xdlops.hpp uses non-member function with `.` operator
#1276
yxsamliu
closed
1 month ago
1
Add missing vector header
#1275
illsilin
closed
2 months ago
0
downgrade minimum required python version to 3.6
#1274
illsilin
closed
2 months ago
0
Avoid using MI100 nodes for CK CI.
#1273
illsilin
closed
2 months ago
0
Previous
Next