Commit Graph

33 Commits

Author SHA1 Message Date
Ekaterina Aidova 4cf2ae0192
[PT FE]: fix segfault when resolving nested dict as model input (#24719)
### Details:
 - *item1*
 - *...*

### Tickets:
 - *ticket-id*
2024-05-27 21:24:01 +00:00
Mikhail Ryzhov a6265c136b
[GHA] Move tokenizers build to reusable job (#23909)
### Details:
 - Move tokenizers build to reusable job

### Tickets:
 - *132908*

---------

Co-authored-by: Roman Kazantsev <roman.kazantsev@intel.com>
2024-04-19 19:12:17 +00:00
Anastasiia Pnevskaia fbac30c538
Fix passing of bf16 and string types to convert model(). (#24113)
### Details:
- Removed internal conversion of all typed passed to `convert_model()`
to `np.dtype`, as some types do not have `numpy` analogs, which causes
errors. Also this step is redundant as we can use `ov.Type` instead.

### Tickets:
 - 138755
2024-04-19 15:31:51 +00:00
Roman Kazantsev 33206de427
[Layer tests] Use legacy MO API only for legacy tests and optimize imports (#23821)
**Details:** Use legacy MO API only for legacy tests and optimize
imports

**Ticket:** TBD

Signed-off-by: Kazantsev, Roman <roman.kazantsev@intel.com>
2024-04-03 08:31:27 +00:00
Evgeny Kotov 22d67427c2
Disable compression to fp16 for optimized models (#23609)
### Details:
 - use PR https://github.com/openvinotoolkit/openvino/pull/22589
 - added unit tests

### Tickets:
 - 126103

---------

Co-authored-by: Ivan Tikhonov <ivan.tikhonov@intel.com>
2024-03-27 06:02:11 +00:00
Maxim Vafin b10877f6e8
[PT FE] Resolve i64 usage in frontend (#23518)
### Details:
 - *Support lightglue*
 - *Support int64 as scalar in PyTorch Frontend*

### Tickets:
 - *CVS-127326, CVS-134328*
2024-03-26 15:57:00 +00:00
Jacek Pawlak 239466ca5d
copyright year update (#23370)
New PR due to merge
conflicts(https://github.com/openvinotoolkit/openvino/pull/22917)

Updated the copyright year from 2018-2023 to 2018-2024 in all openvino
files

Ref. to script: CVS-101144

Command used:
```bash
git grep -lz '2018-2023 Intel Corporation' | xargs -0 sed -i '' -e 's/2018-2023 Intel Corporation/2018-2024 Intel Corporation/g'
```
2024-03-14 09:37:02 +00:00
Anastasiia Pnevskaia 4f3474abf6
[TF FE] Set single input tensor names in TF1 formats. (#22503)
Description: 
Change setting of names in TF1 to make single input tensor.

Ticket: 129457


Code:
* [x]  Comments
* [x]  Code style (PEP8)
* [x]  Transformation generates reshape-able IR
* [x]  Transformation preserves original framework node names
* [x]  Transformation preserves tensor names


Validation:
* [x]  Unit tests
* [x]  Framework operation tests
* [x]  Transformation tests

Documentation:
* [x]  Supported frameworks operations list
* [x]  Guide on how to convert the **public** model
* [x]  User guide update

---------

Co-authored-by: Roman Kazantsev <roman.kazantsev@intel.com>
2024-02-19 13:54:42 +00:00
Roman Kazantsev 3d2c15ffd9
[TF FE][Layer tests] Run layer tests for TF FE by default and use_legacy_frontend for legacy FE (#22848)
**Details:** Run layer tests for TF FE by default and
use_legacy_frontend for legacy FE. Developers have to add
`use_new_frontend` for running layer tests for TF FE. This is not
convenient. Instead, let us use `use_legacy_frontend` option for legacy
FE that is rarely run.

**Tickets:** TBD

---------

Signed-off-by: Kazantsev, Roman <roman.kazantsev@intel.com>
2024-02-15 15:42:06 +00:00
Anastasiia Pnevskaia 52ee006f9b
[PyTorch FE] Common prefix in PyTorch exceptions. (#22655)
### Details:
 - Added a new macro with PyTorch prefix for PyTorch exceptions.
 - Added sending of error information telemetry from PyTorch.

### Tickets:
 - 127906
2024-02-07 12:58:17 +00:00
Anastasiia Pnevskaia aefe19ca89
[TF FE] Logic for keeping WriteFile in graph. (#22304)
* Added logic to keep WriteFile in graph.

* Corrected code style, added test.

* Minor correction.

* Corrected check_supported_ops.py.

* Tests fix.

* Serialize fix.

* Code style.

* Moved KeepInGraphOp to TF internal ops.

* Minor fix.

* Added validate_and_infer_types().

* Code style.

* Moved WriteFile translator.

* Corrections.

* Code style.

* Applied comments.

* Minor correction.

* Minor correction.

* Added comments, minor corrections.
2024-01-26 14:18:57 +00:00
Anastasiia Pnevskaia 41195a4b2a
[TF FE] Set single output tensor name for TF1 formats. (#21075)
* Set single output tensor name for TF1 and TF2 formats.

* Removed debug output.

* Added param for defining if tensor names need indices.

* Clang format.

* Comment corrected.

* Fixed TF1 layer tests.

* Minor correction.

* Fixed errors.

* Error fixed.

* Bug fixes.

* Bug fix.

* Code style, minor corrections.

* Added comments.

* Update src/frontends/tensorflow/src/frontend.cpp

Co-authored-by: Roman Kazantsev <roman.kazantsev@intel.com>

* Added tests.

* Optimize imports.

* Code correction.

* Removed WA

* Small correction.

* Moved names alignment logic to adjust_saved_model_names().

* Removed changes in utils.

* Fixed errors, applied comments.

* Test fixed.

* Removed changes from utils.

* Removed commented code.

* Added note.

* Minor corrections.

* Code style.

* Changed set_node_name, removed not needed changes.

* Clang format.

* Fixed names for output places in input model.

* Fixed tensor names of identity.

* Added ticket number.

* Added ticket numbers.

* Added ticket numbers.

* Code style.

* Corrected tests.

* Update tests/layer_tests/ovc_python_api_tests/test_tf.py

Co-authored-by: Roman Kazantsev <roman.kazantsev@intel.com>

---------

Co-authored-by: Roman Kazantsev <roman.kazantsev@intel.com>
2024-01-15 21:53:43 +04:00
Pavel Esir b5b53e1749
Keep compressed constants produced by FW (#22095)
* keep FW 16 bit float constants

* add layer tests

* remove leftovers from onnx, pdpd

* rename to MarkCompressedFloatConstants

* remove VisualizeTree

* added explanation why we need MarkCompressedFloatConstants
2024-01-15 15:05:24 +04:00
Georgy Krivoruchko 3722df74e8
[Py] Removing an old API from tests (#22047)
* Removed usage of old api
* Removed use_old_api
* Removed use_old_api in common files
* Removed use_old_api from tests
* MO/OVC related changes
* Renamed InferAPI20
* Removed use_old_api from tf_Placeholder test
* Removed transpose step
* Removed forced use_old_api=False for TF Lite
2024-01-10 19:25:29 +04:00
Anastasiia Pnevskaia 3ab5ee861d
Fixed names for GraphDef. (#21799) 2023-12-21 17:04:05 +04:00
Anastasiia Pnevskaia 823d327476
Fixed verbose param. (#21234) 2023-11-22 17:32:20 +04:00
Maxim Vafin 570817998d
Update reverse infer to allow changing shape if it already partially defined (#21073)
* Update reverse infer to allow changing shape if it already partially defined

* Update tests

* Remove changes in If

---------

Co-authored-by: Roman Kazantsev <roman.kazantsev@intel.com>
2023-11-20 21:57:43 +00:00
Anastasiia Pnevskaia d304be0bd7
[OVC] Fixed error in --output_model param. (#20969)
* Fixed error in get_model_name_from_args(), added tests.

* Corrected output_model description.

* Corrected test.
2023-11-15 17:42:14 +04:00
Anastasiia Pnevskaia 5fa53aa715
Input and output order Keras tests. (#20902)
* Input/output order Keras tests.

* Added precommit mark.

* Added xfail.

* Small correction.

* Check input/outputs by names in FW.

* Moved output order tests to Python API group.

* Corrected comments.
2023-11-13 14:01:35 +04:00
Anastasiia Pnevskaia cc389c23ca
Removed logic of building example_input by shape. (#20859) 2023-11-03 20:45:34 +04:00
Maxim Vafin 8d0381b0fe
[PT FE] Implement custom op for types alignment (#20431)
* [PT FE] Implement custom op for types alignment

* Fix code style

* Fix inplace ops

* Fix layer tests

* Remove no longer needed change

* Fix ovc tests

* Fix fe tests
2023-10-23 22:54:08 +02:00
Anastasiia Pnevskaia 2bbfe7b44d
Added support of shapes and types from original FW in ov.convert_model() (#20009)
* Added support of shapes and types from paddle, torch and tf.

* Removed changes from requirements.

* Corrected test.

* Moved helper methods to utils.

* Separated tests by frameworks.

* Removed changes from complex_params test.
2023-09-26 17:41:01 +04:00
Pavel Esir 9271b79540
[OVC] do not parse inputs for py_api (#19742)
* [OVC] do not parse inputs

* fix unit-tests

* remove redundant lines, add test case

* add one more unit-test

* skip None values

* replace str with List in test_mo_import_from_memory

* corrected type hints, added a safety assert
2023-09-22 15:05:21 +04:00
Anastasiia Pnevskaia 215a2f435b
tf.Graph decoder accuracy fixes (#19903)
* Fixed outputs passing from decoder.

* Fixed get_variable method.

* Code style.

* Removed xfails from precommit models.

* Minor fix.

* Comment fixed.

* Added test.
2023-09-19 19:04:24 +02:00
Anastasiia Pnevskaia 18f29c02d2
Fixed scalar creating in tf.Graph decoder (#19735)
* Fixed scalar logic in tf.Graph decoder.

* Passed memory sharing flag to all Tensor constructors.

* Small correction.

* Test correction.
2023-09-14 23:36:22 +04:00
Maxim Vafin a6bc78dd0f
[PT FE] Separate tracing and scripting modes (#19676)
* [PT FE] Separate scripting and tracing in decoder

* Fix convert_model to accept decoder

* Some fixes

* Fix code style

* Fix preprocessor tests

* Fix tests

* Fix tests

* Fix more tests

* Fix ovc tests
2023-09-12 12:40:20 +04:00
Pavel Esir b8226db465
[OVC] Fix output parsing (#19425)
* fix parsing output

* use always coma backspace to separate outputs ', '

* update docstring

* call parser only for ovc cli tool

* docstring correction

* separate docs for cli tool and convert_model; other minor changes

* drop redundant arg from cli_parser_test.py

* more solid test cases added

* remove redundant argv.framework from cli_parser_test.py

* shape correction in concat

* Apply suggestions from code review

fix: coma -> comma

Co-authored-by: Roman Kazantsev <roman.kazantsev@intel.com>

* apply review suggestions

* remove extra ')'

---------

Co-authored-by: Roman Kazantsev <roman.kazantsev@intel.com>
2023-09-01 11:20:22 +02:00
Ekaterina Aidova 80b8b6fff1
[PT FE]: allow example input list with single tensor (#19308) 2023-08-23 12:08:39 +04:00
Anastasiia Pnevskaia 3ef345095c
Set CF=False. (#19223) 2023-08-17 11:22:26 +04:00
Anastasiia Pnevskaia 7a566313e5
Fixed 'output_model' logic in OVC. (#19171)
* Fixed output_model logic.

* Removed not needed code.

* Used os.path.basename, added comments.

* Removed loop.

* Test corrections.
2023-08-17 09:47:18 +04:00
Anastasiia Pnevskaia 3e23908983
Improvements and fixes in OVC convert_model (#19184)
* Added support of tuple in input, removed type syntax from OVC tool.

* Removed type syntax tests.

* Apply suggestions from code review

* Method annotation corrected.

* Type annotation corrected.

---------

Co-authored-by: Sergey Lyalin <sergey.lyalin@intel.com>
2023-08-17 09:45:30 +04:00
Sergey Lyalin bf607a1238
Remove `example_output` for `ovc` and `ovc.convert_model` (#19079)
* Removed 'example_output' from ovc and ovc.convert_model, used output for this purpose

* Update tools/ovc/openvino/tools/ovc/convert.py

* Update tools/ovc/openvino/tools/ovc/convert_impl.py

* Reverted mo parts not affected by remove of example_output

* fix PDPD convert_model tests

---------

Co-authored-by: Xiuchuan Zhai <xiuchuan.zhai@intel.com>
2023-08-15 15:14:25 +04:00
Anastasiia Pnevskaia 59872ee072
Removed InputCutInfo, disabled input cut in OVC. (#18927)
* Fixed output_model logic.

* Removed InputCutInfo, disabled input cut in ovc.

* Disabled output cut, added tests for setting shapes or types for not all inputs.

* Returned support of numpy type.

* Separated MO and OVC python API tests.

* Small corrections.

* Added output dir test, exceptions test.

* Tests fixed.

* Corrected extension param description.

* Corrected input description, minor code corrections.
2023-08-10 23:47:27 +04:00