onnxruntime
[EP ABI] Add weight pre-packing support to kernel-based plugin EPs
#26754
Merged
Overview
Commits
31
Changes
View On GitHub

[EP ABI] Add weight pre-packing support to kernel-based plugin EPs #26754

adrianlizarraga merged 31 commits into main from adrianl/plugin-ep-kernel-prepack
adrianlizarraga
adrianlizarraga [EP ABI] Add weight pre-packing support to kernel-based plugin EPs
4f3d0042
adrianlizarraga Add comment about sharing of prepacked weights (cpu ep only)
e9540216
adrianlizarraga Update Mul kernel to pre-pack input b
fb2998b6
adrianlizarraga
adrianlizarraga commented on 2025-12-09
adrianlizarraga
adrianlizarraga commented on 2025-12-09
adrianlizarraga Apply suggestions from code review
5e64f79c
adrianlizarraga adrianlizarraga marked this pull request as ready for review 177 days ago
adrianlizarraga adrianlizarraga requested a review from fs-eire fs-eire 177 days ago
adrianlizarraga adrianlizarraga requested a review from yuslepukhin yuslepukhin 177 days ago
adrianlizarraga adrianlizarraga requested a review from skottmckay skottmckay 177 days ago
adrianlizarraga
adrianlizarraga commented on 2025-12-09
adrianlizarraga Add comments regarding prepack allocator lifetime
9b1c6a20
adrianlizarraga Merge branch 'adrianl/plugin-ep-kernel-prepack' of github.com:microso…
c638a1a0
fs-eire
fs-eire dismissed these changes on 2025-12-09
yuslepukhin
yuslepukhin dismissed these changes on 2025-12-09
adrianlizarraga adrianlizarraga marked this pull request as draft 175 days ago
adrianlizarraga Added support for sharing pre-packed weights for cpu-accessible alloc…
717ed4ac
adrianlizarraga adrianlizarraga dismissed their stale review via 717ed4ac 175 days ago
adrianlizarraga Define what should happen if OrtKernelImpl::SetSharedPrePackedWeight(…
bd8f6f04
adrianlizarraga Merge branch 'main' into adrianl/plugin-ep-kernel-prepack
fc1fd161
adrianlizarraga Clean up some exception handling
8b3f56c6
adrianlizarraga Refactor example kernel classes (no inheritance)
23503a18
adrianlizarraga Merge branch 'main' into adrianl/plugin-ep-kernel-prepack
7f37ffbd
adrianlizarraga adrianlizarraga marked this pull request as ready for review 169 days ago
adrianlizarraga Correct use of output param
26eca56c
adrianlizarraga Add more edge-case handling for PrePack() call
7af257b3
adrianlizarraga adrianlizarraga requested a review from yuslepukhin yuslepukhin 169 days ago
adrianlizarraga adrianlizarraga requested a review from edgchen1 edgchen1 169 days ago
adrianlizarraga
adrianlizarraga commented on 2025-12-17
adrianlizarraga
adrianlizarraga commented on 2025-12-17
adrianlizarraga
adrianlizarraga commented on 2025-12-17
adrianlizarraga API version checks
515062e9
adrianlizarraga Use correct SAL annotation for array parameters
347ce4f4
adrianlizarraga Clean up some includes
906187d3
edgchen1
edgchen1 commented on 2025-12-18
adrianlizarraga Update onnxruntime/core/session/plugin_ep/ep_kernel_registration.cc
1611fc38
yuslepukhin
yuslepukhin commented on 2025-12-19
yuslepukhin yuslepukhin dismissed their stale review 167 days ago
revoking review
adrianlizarraga Remove OrtAllocator parameter from SharedPrePackedWeightCache_StoreWe…
30ca590e
adrianlizarraga Clarify what happens when SharedPrePackedWeightCache_StoreWeightData …
a5342b92
adrianlizarraga adrianlizarraga requested a review from edgchen1 edgchen1 164 days ago
adrianlizarraga adrianlizarraga requested a review from yuslepukhin yuslepukhin 164 days ago
yuslepukhin
yuslepukhin commented on 2025-12-22
yuslepukhin
yuslepukhin commented on 2025-12-22
yuslepukhin
yuslepukhin commented on 2025-12-22
edgchen1
edgchen1 commented on 2025-12-22
adrianlizarraga Merge branch 'main' into adrianl/plugin-ep-kernel-prepack
51bc7319
adrianlizarraga Review comments
edf3f2c2
adrianlizarraga C++ API
e94c0aab
adrianlizarraga adrianlizarraga requested a review from yuslepukhin yuslepukhin 163 days ago
adrianlizarraga adrianlizarraga requested a review from fs-eire fs-eire 163 days ago
adrianlizarraga adrianlizarraga requested a review from edgchen1 edgchen1 163 days ago
yuslepukhin
yuslepukhin commented on 2025-12-23
adrianlizarraga Improve doc for c++ api convenience class
c8eb3c90
adrianlizarraga Add buffer_sizes as a parameter to OrtKernelImpl::SetSharedWeightData
98e3d132
adrianlizarraga Add comment to implementation of OrtKernelImpl::SetSharedPrePackedWeight
c61ae41f
adrianlizarraga Do not prescribe what the kernel impl should return for a situation t…
02d75d2e
edgchen1
edgchen1 dismissed these changes on 2025-12-24
adrianlizarraga Update include/onnxruntime/core/session/onnxruntime_ep_c_api.h
0a84eda6
adrianlizarraga adrianlizarraga dismissed their stale review via 0a84eda6 162 days ago
adrianlizarraga Adjust comments
5f80f9d3
adrianlizarraga Tweak comment again
c60472d3
adrianlizarraga adrianlizarraga requested a review from edgchen1 edgchen1 162 days ago
fs-eire
fs-eire commented on 2025-12-26
fs-eire
fs-eire commented on 2025-12-27
adrianlizarraga Add comments to clarify ownership scenarios
441c9e25
fs-eire
fs-eire approved these changes on 2025-12-27
adrianlizarraga adrianlizarraga enabled auto-merge (squash) 159 days ago
adrianlizarraga adrianlizarraga merged 2d1ed5b7 into main 159 days ago
adrianlizarraga adrianlizarraga deleted the adrianl/plugin-ep-kernel-prepack branch 159 days ago

Login to write a write a comment.

Login via GitHub

Reviewers
Assignees
No one assigned
Labels
None yet
Milestone
No milestone
Terms Of Use Privacy Policy Refund Policy Impressum