Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Changes from upstream/master #2

Open
wants to merge 2,323 commits into
base: master
Choose a base branch
from
Open

Changes from upstream/master #2

wants to merge 2,323 commits into from

Conversation

bbielawx
Copy link
Owner

No description provided.

akopytko and others added 30 commits November 12, 2024 15:25
@akopytko
[DOCS] Refactor two y-axis graph scripts (#27524)
629fb81 
### Details:
 - *item1*
 - *...*

### Tickets:
 - *ticket-id*
@ahnyoung-paul
[GPU] fix bad_fuction call issue (#27517)
0080d90 
### Details:
- *Fix bad_function call issue which is caused by creating empty
kernel_data*
- Reverting some change applied at
#27439
 - To be investigated and fixed in the next PR 


### Tickets:
 - *154591*
@luo-cheng2021
[CPU] Enable u8 kv cache by default (#27454)
2d148ec 
### Details:
 - *Enable u8 kv cache by default*
 - *...*

### Tickets:
 - *[152621](https://jira.devtools.intel.com/browse/CVS-152621)*
@timxu826
fix assertion failed on lockable_mem when using devicemem (#27297)
41c1400 
### Details:
- *The issue caused by the memory select of the input. In the changed
behavior called locked memory will put the input into host memory if any
the implementation of  its users is CPU.*
- *when the input is set to "use_device_mem" by benchmark_app, the
locked memory will ignore the memory, which will trigger the assertion
failure.*
- *fixed just ignore "need_lockable_mem" option in case of remote
tensors, because the data could be cloned implicitly at attempt to
access device buffer from the host side to have this copy*

### Tickets:
 - *CSV-152365*
@boonsino @m-falkowski
Add parameter to execute inference requests at a fixed frequency (#26820
79dd358 
)

### Details:
 - Add parameter to execute inference requests at a fixed frequency

### Tickets:
 - *ticket-id*

---------

Signed-off-by: Maciej Falkowski <maciej.falkowski@intel.com>
Signed-off-by: Ooi, Boon Sin <boon.sin.ooi@intel.com>
Co-authored-by: Maciej Falkowski <maciej.falkowski@intel.com>
@tkrupa-intel
Enable weightless caching with compile_model() API without providing …
a58202f 
…weights_path property (#27407)

Requested in
https://jira.devtools.intel.com/browse/CVS-126606?focusedId=25564110&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-25564110
@t-jankowski
[Templ test] GroupConvolutionBackpropData: Enable whole Tensor compar…
291d3e1 
…ison (#27379)

### Details:
 - Enabled whole tensor comparison.

### Tickets:
 - CVS-137167

Signed-off-by: Tomasz Jankowski <tomasz1.jankowski@intel.com>
@Lyamin-Roman
[GPU] Minor state fixes (#27508)
431202e 
### Tickets:
 - *[152882](https://jira.devtools.intel.com/browse/CVS-152882)*
@CuriousPanCake
[TRANSFORMATIONS | GPU] Add a Validate pass after MoveEltwiseUpThroug…
acccb22 
…hDataMovScalar for element type resolution (#27464)

[TRANSFORMATIONS | GPU] Add a Validate pass after
MoveEltwiseUpThroughDataMovScalar

After executing the MoveEltwiseUpThroughDataMovScalar transforamtion
some node's element types may appear to be in inconsistent/corrupted
state. Fix it by inserting a Validate pass after the transformation for
resolving the node's element types.

This commit is a workaround until the 141764 is fixed which resolves the
issue of Validate passes.

Ticket:
* CVS-151111

Signed-off-by: Andrii Staikov <andrii.staikov@intel.com>

Signed-off-by: Andrii Staikov <andrii.staikov@intel.com>
@sgolebiewski-intel
[DOCS] Fixing reference in documentation (#27531)
d5dcb04 
Fixing broken links in documentation.

Signed-off-by: Sebastian Golebiewski <sebastianx.golebiewski@intel.com>
@mryzhov @akladiev @ilya-lavrenov
[CI][GHA] Manylinux x86 build (#27430)
a97ff61 
### Details:
 - Enabled manilinux x86 build based on manylinux 2014 image
 - OpenVINO tarball package
 - Wheels for 3.9-3.13 Pythons

### Tickets:
 - *148719*

---------

Co-authored-by: Alina Kladieva <alina.kladieva@intel.com>
Co-authored-by: Ilya Lavrenov <ilya.lavrenov@intel.com>
@akuporos
[PyOV] Restrict changing data in const (#27431)
a0940bb 
### Details:
 - Const op meant to be non-changeable

### Tickets:
 - CVS-124319
@akopytko
[DOCS] Preselection and sorting (#27538)
eb38e67 
### Details:
 - *item1*
 - *...*

### Tickets:
 - *ticket-id*
@mryzhov
[DOCS] fixed get_cmake_path import instructions (#27542)
4847f49 
### Details:
 - Fixed import example

### Tickets:
 - *ticket-id*
@praasz
[core] On Linux try return not used memory to system on compile model…
f417097 
… destructor (#27534)

### Details:
- Optimize Linux process using OV memory consumption by returning not
used memory to the system when compiled model is destroyed.
- It should optimize memory usage by C++ and Python applications, as
Linux avoid reclaim memory until process end especially for small chunks
of allocations.

### Tickets:
 - CVS-149497
@TolyaTalamanov
NPUW: Support NF4 DCOFF for CW models (#27518)
51906cf
@smirnov-alexey
[NPUW] Add a pass to convert convolutions to matmul (#27487)
3d4f942
@ababushk
[DOCS] Security best practices for GitHub Actions workflows developme…
d70603e 
…nt (#27394)

### Details:
Security guidelines for GitHub Actions workflows developers
@yeonbok
[GPU] Fix unexpectedly turned off Small N case optimization (#27493)
2b25007 
### Details:
- In https://github.com/openvinotoolkit/openvino/pull/26733/files, small
N case optimization with tile_ofm == 1 was unexpectedly turned off

### Tickets:
 - *ticket-id*
@dnkurek
[GPU] Fix memory leak (#27536)
b1ff99c 
### Details:
 - Fix memory leak and improve memory usage with continuous inference

### Tickets:
 - 148552
@rkazants
[TF FE] Stabilize layer tests for Keras GRU layer on all platforms (#…
453ee57 
…27543)

**Details:** Stabilize layer tests for Keras GRU layer on all platforms

**Ticket:** 156967

---------

Signed-off-by: Kazantsev, Roman <roman.kazantsev@intel.com>
@zhangYiIntel
[CPU]fix rope mark up to skip shapeof (#27462)
a661f0d 
### Details:
 - *Stop markup at ShapeOf*
- *The ShapeOf in the subgraph of Rope's 2nd/3rd input should not be
marked*
- *The parent of ShapeOf may change when IR changes so skip it to avoid
unknown precision problem*

### Tickets:
 - *CVS-155898*
@popovaan @rkazants
Fixed OVC app name in telemetry. (#27544)
fefe0c3 
### Details:
- Fixed the issue, that '`Model Conversion API`' telemetry category
includes `OVC` and all other tools, now `OVC` has a separate category
"`OVC`".
- Refactored `send_params_info()` method and fixed naming of telemetry
init method.

---------

Co-authored-by: Roman Kazantsev <roman.kazantsev@intel.com>
@mryzhov
[GHA] Manylinux added CI tags and manifest (#27540)
19ae2a9 
### Details:
- added actions to save artifacts for Jenkins - create_manifest and
store_artifacts
 - Added overall status workflow

### Tickets:
 - *ticket-id*
@tkrupa-intel
[GPU] Change weights_path mutabilibity to RW (#27553)
651a51c 
### Details:
It needs to be readable (e.g. to be queried by hello_query_device
sample) and writable for plugin to set the default value.

### Tickets:
https://jira.devtools.intel.com/browse/CVS-157364
@sgolebiewski-intel @kblaszczak-intel
[DOCS] Updating NPU GenAI docs (#27489)
c338559 
Updating the `Run LLMs with OpenVINO GenAI Flavor on NPU` article -
adding info about new config options.
This PR addresses JIRA ticket: 156503.

---------

Signed-off-by: Sebastian Golebiewski <sebastianx.golebiewski@intel.com>
Co-authored-by: Karol Blaszczak <karol.blaszczak@intel.com>
@akopytko
[DOCS Added new hierarchical field for Coveo (#27569)
250f001 
### Details:
 - *item1*
 - *...*

### Tickets:
 - *ticket-id*
@evkotov
fix TSUnsqueezeBackward Reshape does nothing (#27467)
6489755 
### Details:
 - fix TSUnsqueezeBackward

### Tickets:
 - CVS-111560
@XinWangIntel
[NPU] Remove template in ext wrapper and fuse functions (#27511)
3e63de0 
### Details:
 - *Remove template in zero_ext_graph_wrappers*
 - *Remove zero_ext_graph_wrappers_interface.hpp*
 - *Add more low level debug log*
 - *Update level-zero-ext repo commit to use 1.9 version*

### Tickets:
 - *156387*

---------

Signed-off-by: Xin Wang <xin1.wang@intel.com>
@sgolebiewski-intel @kblaszczak-intel
[DOCS] Updating Weight Compression Article (#27432)
0376890 
Rearranging information in the `LLM Weight Compression` article.

---------

Co-authored-by: Karol Blaszczak <karol.blaszczak@intel.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers
No reviews
Assignees
No one assigned
Labels
category: build category: C API category: CI category: Core category: CPU category: dependency_changes category: docs_snippets category: docs category: inference category: JS API category: licensing category: LP transformations category: packaging category: Python API category: PyTorch FE category: samples category: transformations github_actions no-match-files
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.