-
Notifications
You must be signed in to change notification settings - Fork 297
All issues
Issue creation is restricted in this repository
Issues
is:issue state:open
is:issue state:open
Search results
[CI] oss:cutlass-4.8:sm107 gated-attention FROST GEMM plans missing after #1177
cat-bugReports of incorrect behavior, crashes, regressions, or unexpected results.Reports of incorrect behavior, crashes, regressions, or unexpected results.orig-qaReported or owned by quality assurance, validation, or testing.Reported or owned by quality assurance, validation, or testing.Status: Open.#1206 In NVIDIA/cudnn-frontend;[Bug]: DSA SM100-family
indexer_backwardrecompiles for packed runtime shapes in v1.29cat-bugReports of incorrect behavior, crashes, regressions, or unexpected results.Reports of incorrect behavior, crashes, regressions, or unexpected results.Status: Open.#1188 In NVIDIA/cudnn-frontend;[CI] qa_cpp_samples:9.24.0.43_13.3.0:sm100: fp16_fwd_with_block_mask.cpp:189 FAILED
cat-bugReports of incorrect behavior, crashes, regressions, or unexpected results.Reports of incorrect behavior, crashes, regressions, or unexpected results.orig-qaReported or owned by quality assurance, validation, or testing.Reported or owned by quality assurance, validation, or testing.Status: Open.#1175 In NVIDIA/cudnn-frontend;[CI] deploy:test_bdist_wheels_ga_cudnn_3.12:sm80: test_la.py::test_fwd_bwd_correlated_keys[cutile-gdn-256]
cat-bugReports of incorrect behavior, crashes, regressions, or unexpected results.Reports of incorrect behavior, crashes, regressions, or unexpected results.orig-qaReported or owned by quality assurance, validation, or testing.Reported or owned by quality assurance, validation, or testing.Status: Open.#1160 In NVIDIA/cudnn-frontend;- Status: Open.#1156 In NVIDIA/cudnn-frontend;
- Status: Open.#1137 In NVIDIA/cudnn-frontend;
- Status: Open.#1136 In NVIDIA/cudnn-frontend;
[Bug]: qa_py_test:9.24.0.43_13.3.0:sm90 6 jobs failed with Mismatched elements
cat-bugReports of incorrect behavior, crashes, regressions, or unexpected results.Reports of incorrect behavior, crashes, regressions, or unexpected results.orig-qaReported or owned by quality assurance, validation, or testing.Reported or owned by quality assurance, validation, or testing.Status: Open.#1130 In NVIDIA/cudnn-frontend;[Bug]: FROST SDPA fwd engine ignores execute-time shape overrides and rejects the graph by its declared shape
cat-bugReports of incorrect behavior, crashes, regressions, or unexpected results.Reports of incorrect behavior, crashes, regressions, or unexpected results.orig-nv-engReported or requested by NVIDIA engineering.Reported or requested by NVIDIA engineering.Status: Open.#1098 In NVIDIA/cudnn-frontend;- Status: Open.#1089 In NVIDIA/cudnn-frontend;
[Doc]: SDPA execute-time shape override cannot cross s_q == 1; heuristic engine flips between declared max_len 128 and 256
cat-docDocumentation changes, examples, API references, tutorials, or wording fixes.Documentation changes, examples, API references, tutorials, or wording fixes.mod-frontendcuDNN frontend APIs, operation graph construction, plans, and user-facing wrappers.cuDNN frontend APIs, operation graph construction, plans, and user-facing wrappers.orig-nv-engReported or requested by NVIDIA engineering.Reported or requested by NVIDIA engineering.Status: Open.#1087 In NVIDIA/cudnn-frontend;[Perf]: pygraph.execute host overhead ~30 us above cudnnBackendExecute per call
cat-perf-bugPerformance regressions or cases where behavior is correct but too slow.Performance regressions or cases where behavior is correct but too slow.mod-frontendcuDNN frontend APIs, operation graph construction, plans, and user-facing wrappers.cuDNN frontend APIs, operation graph construction, plans, and user-facing wrappers.orig-nv-engReported or requested by NVIDIA engineering.Reported or requested by NVIDIA engineering.Status: Open.#1086 In NVIDIA/cudnn-frontend;