Repository navigation
Conversation
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/24945
Note: Links to docs will display an error until the docs builds have been completed. ❌ 138 New Failures, 7 Unrelated Failures, 10 Unclassified FailuresAs of commit ddb4ff2 with merge base c7f7931 ( NEW FAILURES - The following jobs have failed:
UNCLASSIFIED FAILURES - DrCI could not classify the following jobs because the workflow did not run on the merge base. The failures may be pre-existing on trunk or introduced by this PR:
FLAKY - The following jobs failed but were likely due to flakiness present on trunk:
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
meta-cla
Bot
added
the
CLA Signed
This was referenced Nov 12, 2026
Muse Glimmer's Hugging Face rotary embedding builds its angles as position_ids[..., None] * inv_freq rather than HF's batched matmul, so FuseRoPEPass left the cos/sin tables in the graph, which the ggml engine cannot run. The pass now recognizes that outer product too. The ggml engine also lowers aten.tanh, which Muse Glimmer's final logit soft cap uses.
metascroy
force-pushed
the
offgraph-kv-mg-ops
branch
from
November 14, 2026 00:18
9bc8b60 to
ddb4ff2
Compare
This branch had an error being deployed
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Stack
Off-graph KV cache for the native ggml engine, through serving Muse Glimmer on it. Review bottom-up; each PR builds and passes its tests on its own.
Summary
Two ops Muse Glimmer's Hugging Face model needs on the ggml engine:
position_ids[..., None] * inv_freqrather than HF's batched matmul, soFuseRoPEPassleftcos/sinin the graph, which ggml can't run. The pass now recognizes that outer product too.aten.tanh: lowered toggml_tanh; Muse Glimmer's final logit soft cap uses it.Test plan
test_passes.py: the broadcast form fuses to a singlenative.ropeand matches the unfused output.test_ggml_ops.py::test_tanhis added but not run here (it needs Buck'sggml_ptn_runner). The Muse Glimmer export later in the stack exercises both changes end to end.Authored with Claude Code.