eagle3 cb impl with top-1 proposal #2740

songbell · 2025-09-17T07:28:21Z

eagle3 CB impl
Tickets: CVS-173358
ref code: https://github.com/SafeAILab/EAGLE

Copilot

Pull Request Overview

This PR implements Eagle speculative decoding functionality for top-1 proposal generation. The implementation adds support for Eagle3 mode, which enables accelerated text generation through speculative decoding with hidden state sharing between main and draft models.

Key changes include:

Added Eagle decoding implementation with model transformation pipelines for hidden state extraction
Integrated safetensor parsing for Eagle3 configuration data (d2t mappings)
Extended continuous batching pipeline to support Eagle mode with hidden state management

Reviewed Changes

Copilot reviewed 24 out of 25 changed files in this pull request and generated 8 comments.

Show a summary per file

File	Description
tools/continuous_batching/accuracy/continuous_batching_eagle_decoding.cpp	New Eagle decoding accuracy test tool
src/cpp/src/speculative_decoding/speculative_decoding_impl.hpp	Added Eagle decoding class definitions and model transformation passes
src/cpp/src/speculative_decoding/speculative_decoding_impl.cpp	Core Eagle decoding implementation with model transformations
src/cpp/src/continuous_batching/pipeline.cpp	Integration of Eagle mode into pipeline construction
src/cpp/src/continuous_batching/model_runner.hpp	Added hidden state management functionality
samples/cpp/text_generation/eagle_speculative_lm.cpp	New Eagle speculative decoding sample
src/cpp/src/safe_tensor_wrapper.hpp	New safetensor parsing utilities

Comments suppressed due to low confidence (1)

src/cpp/src/continuous_batching/model_runner.hpp:1

This appears to be modifying the token index without bounds checking on the d2t array. Add bounds checking to prevent potential buffer overflow.

// Copyright (C) 2023-2025 Intel Corporation

_{Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.}

src/cpp/src/speculative_decoding/speculative_decoding_impl.cpp

src/cpp/src/continuous_batching/model_runner.hpp

tools/continuous_batching/accuracy/CMakeLists.txt

src/cpp/src/speculative_decoding/continuous_batching_for_speculative_decoding_impl.cpp

src/cpp/src/sequence_group.cpp

src/cpp/src/continuous_batching/pipeline.cpp

Signed-off-by: fishbell <[email protected]>

samples/cpp/text_generation/greedy_causal_lm.cpp

tools/continuous_batching/accuracy/continuous_batching_speculative_decoding.cpp

tools/continuous_batching/benchmark/continuous_batching_benchmark.cpp

samples/cpp/text_generation/eagle_speculative_lm.cpp

Copilot

Pull Request Overview

Copilot reviewed 28 out of 29 changed files in this pull request and generated 10 comments.

_{Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.}

tools/llm_bench/task/text_generation.py

tools/llm_bench/llm_bench_utils/ov_utils.py

tools/continuous_batching/benchmark/continuous_batching_benchmark.cpp

tools/continuous_batching/accuracy/continuous_batching_speculative_decoding.cpp

src/cpp/src/speculative_decoding/speculative_decoding_impl.cpp

src/cpp/src/speculative_decoding/continuous_batching_for_speculative_decoding_impl.cpp

src/cpp/src/continuous_batching/pipeline.cpp

src/cpp/src/continuous_batching/model_runner.hpp

samples/cpp/text_generation/greedy_causal_lm.cpp

tests/python_tests/test_eagle3.py

Wovchena · 2025-10-20T09:46:41Z

.github/workflows/linux.yml

            timeout: 240
          - name: 'LLM & VLM'
-            cmd: 'python -m pytest -v ./tests/python_tests/test_llm_pipeline.py tests/python_tests/test_llm_pipeline_static.py ./tests/python_tests/test_vlm_pipeline.py tests/python_tests/test_structured_output.py --override-ini cache_dir=/mount/caches/pytest/'
+            cmd: 'python -m pytest -v ./tests/python_tests/test_llm_pipeline.py tests/python_tests/test_llm_pipeline_static.py ./tests/python_tests/test_vlm_pipeline.py tests/python_tests/test_structured_output.py tests/python_tests/test_eagle3.py --override-ini cache_dir=/mount/caches/pytest/'


Move it to a separate matrix element. That way it will be run on a separate runner avoiding test slow down.

As a side effect you should be able to install appropriate optimum-intel here and remove @pytest.mark.skip(reason="CVS-174959 enable model conversion for eagle3 and enable the test")

tools/llm_bench/task/text_generation.py

tests/python_tests/test_eagle3.py

Wovchena · 2025-10-20T10:35:53Z

tests/python_tests/test_continuous_batching.py

            assert std_gen_duration == 0
    else:
-        assert extended_perf_metrics is None
+        assert extended_perf_metrics is None


Put the line back :)

src/cpp/src/sampling/sampler.hpp

src/cpp/src/safe_tensor_wrapper.hpp

src/cpp/src/speculative_decoding/speculative_decoding_impl.hpp

src/cpp/src/speculative_decoding/update_request_structs.hpp

src/cpp/src/sequence_group.hpp

src/cpp/src/continuous_batching/pipeline.cpp

src/cpp/src/continuous_batching/model_runner.hpp

apaniukov · 2025-10-20T12:11:43Z

I am trying to run the tests with this command

pytest tests/python_tests/test_eagle3.py

@songbell could you share the environment?

My current env:
GenAI from this branch: https://github.com/songbell/openvino.genai/tree/bell/eagle_cb_impl
Optimum-intel from this branch: https://github.com/xufang-lisa/optimum-intel/tree/xufang/add_eagle3_draft_model_conversion

With transformers-4.55.4 the original tokenizer cannot be loaded:

.venv/lib/python3.10/site-packages/transformers/tokenization_utils_base.py:2069: in from_pretrained
    return cls._from_pretrained(
.venv/lib/python3.10/site-packages/transformers/tokenization_utils_base.py:2107: in _from_pretrained
    slow_tokenizer = (cls.slow_tokenizer_class)._from_pretrained(
.venv/lib/python3.10/site-packages/transformers/tokenization_utils_base.py:2315: in _from_pretrained
    tokenizer = cls(*init_inputs, **init_kwargs)
.venv/lib/python3.10/site-packages/transformers/models/llama/tokenization_llama.py:171: in __init__
    self.sp_model = self.get_spm_processor(kwargs.pop("from_slow", False))
.venv/lib/python3.10/site-packages/transformers/models/llama/tokenization_llama.py:198: in get_spm_processor
    tokenizer.Load(self.vocab_file)
.venv/lib/python3.10/site-packages/sentencepiece/__init__.py:961: in Load
    return self.LoadFromFile(model_file)
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

self = <sentencepiece.SentencePieceProcessor; proxy of <Swig Object of type 'sentencepiece::SentencePieceProcessor *' at 0x71b823d09a70> >, arg = None

    def LoadFromFile(self, arg):
>       return _sentencepiece.SentencePieceProcessor_LoadFromFile(self, arg)
E       TypeError: not a string

.venv/lib/python3.10/site-packages/sentencepiece/__init__.py:316: TypeError

With newer transformers-4.57.1:

tests/python_tests/utils/hugging_face.py:201: in download_and_convert_model
    return _download_and_convert_model(model_id, OVModelForCausalLM, **tokenizer_kwargs)
tests/python_tests/utils/hugging_face.py:233: in _download_and_convert_model
    opt_model, hf_tokenizer = get_huggingface_models(model_id, model_class, local_files_only=False)
tests/python_tests/utils/hugging_face.py:170: in get_huggingface_models
    opt_model = retry_request(lambda: model_class.from_pretrained(model_id, export=isinstance(model_id, str), compile=False, load_in_8bit=False, trust_remote_code=isinstance(model_id, str), ov_config=get_default_llm_properties(), local_files_only=local_files_only))
tests/python_tests/utils/network.py:37: in retry_request
    return func()
tests/python_tests/utils/hugging_face.py:170: in <lambda>
    opt_model = retry_request(lambda: model_class.from_pretrained(model_id, export=isinstance(model_id, str), compile=False, load_in_8bit=False, trust_remote_code=isinstance(model_id, str), ov_config=get_default_llm_properties(), local_files_only=local_files_only))
.venv/lib/python3.10/site-packages/optimum/intel/openvino/modeling_base.py:504: in from_pretrained
    return super().from_pretrained(
.venv/lib/python3.10/site-packages/optimum/modeling_base.py:407: in from_pretrained
    return from_pretrained_method(
.venv/lib/python3.10/site-packages/optimum/intel/openvino/modeling_decoder.py:345: in _export
    main_export(
.venv/lib/python3.10/site-packages/optimum/exporters/openvino/__main__.py:562: in main_export
    submodel_paths = export_from_model(
.venv/lib/python3.10/site-packages/optimum/exporters/openvino/convert.py:745: in export_from_model
    export_models(
.venv/lib/python3.10/site-packages/optimum/exporters/openvino/convert.py:514: in export_models
    export(
.venv/lib/python3.10/site-packages/optimum/exporters/openvino/convert.py:216: in export
    return export_pytorch(
.venv/lib/python3.10/site-packages/optimum/exporters/openvino/convert.py:422: in export_pytorch
    ov_model = convert_model(
../openvino_toolkit_ubuntu22_2025.4.0.dev20251017_x86_64/python/openvino/tools/ovc/convert.py:105: in convert_model
    ov_model, _ = _convert(cli_parser, params, True)
../openvino_toolkit_ubuntu22_2025.4.0.dev20251017_x86_64/python/openvino/tools/ovc/convert_impl.py:578: in _convert
    raise e
../openvino_toolkit_ubuntu22_2025.4.0.dev20251017_x86_64/python/openvino/tools/ovc/convert_impl.py:518: in _convert
    ov_model = driver(argv, {"conversion_parameters": non_default_params})
../openvino_toolkit_ubuntu22_2025.4.0.dev20251017_x86_64/python/openvino/tools/ovc/convert_impl.py:256: in driver
    ov_model = moc_emit_ir(prepare_ir(argv), argv)
../openvino_toolkit_ubuntu22_2025.4.0.dev20251017_x86_64/python/openvino/tools/ovc/convert_impl.py:195: in prepare_ir
    ov_model = moc_pipeline(argv, moc_front_end)
../openvino_toolkit_ubuntu22_2025.4.0.dev20251017_x86_64/python/openvino/tools/ovc/moc_frontend/pipeline.py:142: in moc_pipeline
    ov_model = moc_front_end.convert(input_model)
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

self = <FrontEnd 'pytorch'>, model = <openvino._pyopenvino.InputModel object at 0x79b16aeefe20>

    def convert(self, model: Union[Model, InputModel]) -> Model:
>       converted_model = super().convert(model)
E       openvino._pyopenvino.OpConversionFailure: Check 'is_conversion_successful' failed at src/frontends/pytorch/src/frontend.cpp:180:
E       FrontEnd API failed with OpConversionFailure:
E       Model wasn't fully converted. Failed operations detailed log:
E       -- aten::cat with a message:
E       Exception happened during conversion of operation aten::cat with schema aten::cat(Tensor[] tensors, int dim=0) -> Tensor
E       Check 'is_axis_valid(axis, r)' failed at src/core/src/validation_util.cpp:332:
E       While validating node 'opset1::Concat Concat_68 (opset1::Convert aten::to/Convert[0]:f32[0], util::PtFrameworkNode prim::TupleUnpack[1]:f32[?,?,?,?]) -> (dynamic[...])' with friendly_name 'Concat_68':
E       Axis -2 out of the tensor rank range [-1, 0].
E       
E       -- prim::ListConstruct with a message:
E       Exception happened during conversion of operation prim::ListConstruct with schema (no schema)
E       Check '(c_node)' failed at src/frontends/pytorch/src/op/list_construct.cpp:25:
E       FrontEnd API failed with OpConversionFailure:
E       [PyTorch Frontend] Translation for prim::ListConstruct support only constant inputs
E       
E       -- prim::ListUnpack with a message:
E       Exception happened during conversion of operation prim::ListUnpack with schema (no schema)
E       Exception from src/core/src/node.cpp:593:
E       node index is out of range
E       
E       Summary:
E       -- normalize step failed with: Exception from src/core/src/pass/graph_rewrite.cpp:298:
E       [ov::frontend::pytorch::pass::AtenCatToConcat] END: node: util::PtFrameworkNode aten::cat (util::PtFrameworkNode prim::ListConstruct[0]:dynamic[...], opset1::Constant 111[0]:i64[]) -> (f32[?,?,?,?]) CALLBACK HAS THROWN: Check 'is_axis_valid(axis, r)' failed at src/core/src/validation_util.cpp:332:
E       While validating node 'opset1::Concat Concat_34833 (opset1::Convert aten::to/Convert[0]:f32[0], util::PtFrameworkNode prim::TupleUnpack[0]:f32[?,?,?,?]) -> (dynamic[...])' with friendly_name 'Concat_34833':
E       Axis -2 out of the tensor rank range [-1, 0].
E       
E       
E       
E       -- No conversion rule found for operations: prim::TupleConstruct, prim::TupleUnpack
E       -- Conversion is failed for: aten::cat, prim::ListConstruct, prim::ListUnpack

../openvino_toolkit_ubuntu22_2025.4.0.dev20251017_x86_64/python/openvino/frontend/frontend.py:18: OpConversionFailure

src/cpp/src/continuous_batching/pipeline.cpp

sbalandi · 2025-10-20T17:07:55Z

tests/python_tests/test_eagle3.py

+@pytest.mark.parametrize("main_device,draft_device", devices)
+@pytest.mark.precommit
+@pytest.mark.skip(reason="CVS-174959 enable model conversion for eagle3 and enable the test")
+def test_eagle3_sd_extended_perf_metrics(main_model, main_device, draft_model, draft_device, prompt):


what is different from the test https://github.com/openvinotoolkit/openvino.genai/blob/master/tests/python_tests/test_continuous_batching.py#L486 ?
This is a test of performance metrics, not that the pipeline runs for an acceptable amount of time. If you still need to test performance metrics, you can parameterize the original test

sbalandi · 2025-10-20T17:13:13Z

tests/python_tests/samples/test_speculative_decoding_lm.py

+        ],
+        indirect=["convert_model", "convert_draft_model"],
+    )
+    def test_sample_speculative_decoding_lm(self, convert_model, convert_draft_model, sample_args):


please, parameterize the TestSpeculativeDecodingLM test or move the common code into a separate function

src/cpp/src/speculative_decoding/speculative_decoding_impl.hpp

src/cpp/src/speculative_decoding/speculative_decoding_impl.cpp

src/cpp/src/llm/pipeline.cpp

src/cpp/src/continuous_batching/model_runner.hpp

rkazants · 2025-10-21T04:52:38Z

src/cpp/src/continuous_batching/pipeline.cpp

+        }
+        if (config.find("dt_mapping_path") != config.end()) {
+            eagle_rt_info.dt_mapping_table = config.at("dt_mapping_path").as<std::filesystem::path>();
+            eagle_rt_info.dt_mapping_table = eagle_rt_info.dt_mapping_table / "eagle3.safetensors";


optimum-cli tool must not generate safetensors. This is not its responsibility at all.

problem need to solve:
the original d2t is wrapped in https://huggingface.co/Tengyunw/qwen3_8b_eagle3/blob/main/pytorch_model.bin
and we only need this d2t in genai for draft model to generate correct tokens.

current solution:
as in this PR, we split this part from pytorch_model.bin, and use it separately in genAI

@rkazants, this d2t tensor is also not a part of model representation, or do you think we still can wrap this part into IR ?
cc @wangleis

Please, educate me on eagle 3 algo and explain why this third model represented as eagle3.safetensors is needed. That's something new compared to the arch review.

Can .safetensors be replaced with IR?

Is it possible to fuse that third model into draft model as a model state and activate it as needed?

Is it possible to fuse that third model at runtime into the main model?

@Wovchena @rkazants It has nothing to do with safetensors, but a wrong suffix is used. It is a plain binary includes some offset values for mapping from draft model vocabulary(smaller size) to base/target model's vocabulary(bigger size). Is it OK to save the offset values in a *.bin file since they are used in GenAI pipeline, not in inference?

dt_mapping_table is part of draft model training implementation detail. So it is not part of arch review.

If the main concern is format of dt_mapping_table, option 2 and 3 in @Wovchena proposal is potential solution.

May I know if either option 2 or 3 is ok for you, @Wovchena and @rkazants?

Hi @wangleis,

I think that dt_mapping_table should be a part of separate (dedicated) IR. That is because it will be fused in the main model at runtime and there will be no relation to the draft model in such case.
Fusing of that third model to the draft model looks a bit complicated to implement than fusing to the main model because we should differentiate first and upcoming generations in the speculative decoding and use dt mapping table only for the first iteration.

Hi @rkazants,

Thanks for comments. We would like to keep main model no change in eagle3 pipeline. So we will update PR to create new IR for dt_mapping_table.

@Wovchena May I know if you have more comments?

where is dt_mapping_table used in the speculative decoding pipeline ?

in draft sampling, to adjust draft tokens, due to draft is using a smaller LM head, see reference:
https://github.com/SafeAILab/EAGLE/blob/main/eagle/model/cnets.py#L712

rkazants

the concern is about safetensors that should not be generated by optimum-cli tool.
optimum-cli generates IRs of different parts of genai pipeline.
We should re-consider solution to avoid safetensors.

tests/python_tests/samples/test_speculative_decoding_lm.py

tests/python_tests/test_eagle3.py

MaximProshin · 2025-10-21T06:23:27Z

WWB Similarity results: #2812 (comment)

Signed-off-by: fishbell <[email protected]>

…vino.genai into bell/eagle_cb_impl

src/cpp/src/speculative_decoding/speculative_decoding_eagle3_impl.cpp

Signed-off-by: fishbell <[email protected]>

…genai into HEAD

Signed-off-by: fishbell <[email protected]>

Signed-off-by: fishbell <[email protected]> enable test Signed-off-by: fishbell <[email protected]>

…genai into HEAD

Signed-off-by: fishbell <[email protected]>

Copilot AI review requested due to automatic review settings September 17, 2025 07:28

songbell changed the title ~~eagle impl with top-1 proposal~~ eagle3 cb impl with top-1 proposal Sep 17, 2025

Copilot AI reviewed Sep 17, 2025

View reviewed changes

songbell added 2 commits September 17, 2025 23:24

eagle impl with top-1 proposal

f56435a

Signed-off-by: fishbell <[email protected]>

enable cb benchmark for eagle3

1d74c03

Signed-off-by: fishbell <[email protected]>

github-actions bot added the category: llm_bench Label for tool/llm_bench folder label Sep 18, 2025

songbell added 7 commits September 18, 2025 18:02

add benchmarking, apply copilot review comments

0b09a65

Signed-off-by: fishbell <[email protected]>

fix case build failure

b36ecf7

Signed-off-by: fishbell <[email protected]>

fix SDL

d86e5a7

Signed-off-by: fishbell <[email protected]>

typo

f0aa2c7

Signed-off-by: fishbell <[email protected]>

opt hidden state transfer with ROI tensor

bbfa8ad

Signed-off-by: fishbell <[email protected]>

opt roi copy interface

35e6376

Signed-off-by: fishbell <[email protected]>

Merge branch 'master' into bell/eagle_cb_impl

f6a8872

peterchen-intel reviewed Sep 21, 2025

View reviewed changes

samples/cpp/text_generation/greedy_causal_lm.cpp Show resolved Hide resolved

tools/continuous_batching/accuracy/continuous_batching_speculative_decoding.cpp Show resolved Hide resolved

tools/continuous_batching/benchmark/continuous_batching_benchmark.cpp Outdated Show resolved Hide resolved

peterchen-intel requested review from wangleis and xufang-lisa September 21, 2025 02:42

peterchen-intel assigned Wovchena Sep 21, 2025

peterchen-intel mentioned this pull request Sep 21, 2025

[do not review]eagle cb pipeline impl #2540

Draft

Wovchena requested a review from Copilot September 22, 2025 06:19

xufang-lisa reviewed Sep 22, 2025

View reviewed changes

samples/cpp/text_generation/eagle_speculative_lm.cpp Outdated Show resolved Hide resolved

Copilot AI reviewed Sep 22, 2025

View reviewed changes

Wovchena reviewed Oct 20, 2025

View reviewed changes

MaximProshin mentioned this pull request Oct 20, 2025

[WWB] Add eagle3 pipeline #2812

Open

3 tasks

Wovchena reviewed Oct 20, 2025

View reviewed changes

src/cpp/src/continuous_batching/pipeline.cpp Outdated Show resolved Hide resolved

sbalandi reviewed Oct 20, 2025

View reviewed changes

rkazants reviewed Oct 21, 2025

View reviewed changes

rkazants requested changes Oct 21, 2025

View reviewed changes

rkazants reviewed Oct 21, 2025

View reviewed changes

tests/python_tests/samples/test_speculative_decoding_lm.py Show resolved Hide resolved

rkazants reviewed Oct 21, 2025

View reviewed changes

tests/python_tests/test_eagle3.py Outdated Show resolved Hide resolved

moslex added the priority: high High piority label Oct 21, 2025

songbell added 2 commits October 21, 2025 18:23

apply review comments part-1, mainly on code

fdeb3f0

Signed-off-by: fishbell <[email protected]>

Merge branch 'bell/eagle_cb_impl' of https://github.com/songbell/open…

5fd9b71

…vino.genai into bell/eagle_cb_impl

sbalandi reviewed Oct 21, 2025

View reviewed changes

src/cpp/src/speculative_decoding/speculative_decoding_eagle3_impl.cpp Outdated Show resolved Hide resolved

songbell added 6 commits October 21, 2025 23:05

reuse common codes between spec decode and eagle3 decode

0931c09

Signed-off-by: fishbell <[email protected]>

add missing default

3c97718

Signed-off-by: fishbell <[email protected]>

enable test interface first

5f16e65

Signed-off-by: fishbell <[email protected]>

apply review comments

bb8a39d

Signed-off-by: fishbell <[email protected]>

enable test for local val first

428c25e

Signed-off-by: fishbell <[email protected]>

use warning logger instead of cout

efd854b

Signed-off-by: fishbell <[email protected]>

github-actions bot removed the category: LoRA Low rank adapters label Oct 23, 2025

songbell added 4 commits October 24, 2025 00:03

update per model d2t changes

a9f68db

Signed-off-by: fishbell <[email protected]>

Merge branch 'master' of https://github.com/openvinotoolkit/openvino.…

15f6a4b

…genai into HEAD

add missing line for unchanged file

3bcdfc3

Signed-off-by: fishbell <[email protected]>

move test

c4d1b6c

Signed-off-by: fishbell <[email protected]>

songbell force-pushed the bell/eagle_cb_impl branch from b1c653a to fc2d11f Compare October 24, 2025 05:37

songbell added 3 commits October 24, 2025 21:39

try trigger test

fc2d11f

Signed-off-by: fishbell <[email protected]> enable test Signed-off-by: fishbell <[email protected]>

Merge branch 'master' of https://github.com/openvinotoolkit/openvino.…

6397b63

…genai into HEAD

upgrade version

1903c31

Signed-off-by: fishbell <[email protected]>

Uh oh!

eagle3 cb impl with top-1 proposal #2740

Are you sure you want to change the base?

eagle3 cb impl with top-1 proposal #2740

Conversation

songbell commented Sep 17, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull Request Overview

Reviewed Changes

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull Request Overview

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

apaniukov commented Oct 20, 2025

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

peterchen-intel Oct 21, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

songbell commented Sep 17, 2025 •

edited

Loading

peterchen-intel Oct 21, 2025 •

edited

Loading

wangleis Oct 22, 2025 •

edited

Loading

rkazants Oct 22, 2025 •

edited

Loading