[QNN EP] Qnn cache improvement #17757
HectorSVC
merged
21 commits into
main
from
qnn_cache_improvement
1. Create Op schema for EPCache
c8aa0542
move duplicate code into general function
68a884aa
Load from skeleton onnx file with cached qnn context binary
8a31adc3
separate out onnx ctx model related code to onnx_ctx_model_helper file
ce00679e
handle the case that load from Onnx QDQ model with context cache skel…
c7856c90
add node attribute "source" to explicitly set whether it's generated …
35599bcc
update test cases and build pipeline
28408329
Merge branch 'main' into qnn_cache_improvement
5fd59209
jywu-msft
commented
on 2023-10-02
HectorSVC
commented
on 2023-10-02
HectorSVC
commented
on 2023-10-02
fix build error for Linux
fdcdb577
remove parameters not used.
f6ecd88f
Extend test case QnnHTPBackendTests.ContextBinaryCacheTest to cover l…
cfc8982f
address review comments
ebbf0013
Enable non embed_mode which has ep_cache_context point to a file path…
2e656fa0
Merge branch 'main' into qnn_cache_improvement
8fffe9e1
add API doc for qnn_context_embed_mode option, and enable it for onnx…
55a506e8
format
71d68c9b
resolve merge issue
730874cd
Merge branch 'main' into qnn_cache_improvement
aa2f7c0b
jywu-msft
commented
on 2023-10-06
jywu-msft
commented
on 2023-10-06
jywu-msft
commented
on 2023-10-06
Add main_context attribute in op schema
0bb9be51
Add test case for context cache with 2 inputs
13413f3a
jywu-msft
dismissed these changes
on 2023-10-06
update doc for contrib op added
9c71f78a
jywu-msft
approved these changes
on 2023-10-06
Reviewers
Assignees
No one assigned
Labels
None yet
Milestone
No milestone
Login to write a write a comment.
Login via GitHub