Owner

openvinotoolkit

12 indexed repositories · View on GitHub

openvino
OpenVINO™ is an open source toolkit for optimizing and deploying AI inference

C++ · 10871 stars
open_model_zoo
Pre-trained Deep Learning models and demos (high quality and extremely fast)

Python · 4433 stars
openvino_notebooks
📚 Jupyter notebook tutorials for OpenVINO™

Jupyter Notebook · 3220 stars
nncf
Neural Network Compression Framework for enhanced OpenVINO™ inference

Python · 1199 stars
model_server
A scalable inference server for models optimized with OpenVINO™

C++ · 931 stars
openvino.genai
Run Generative AI models with simple C++/Python API and using OpenVINO Runtime

C++ · 588 stars
openvino_contrib
Repository for OpenVINO's extra modules

C++ · 186 stars
npu_compiler
OpenVINO Intel NPU Compiler

MLIR · 100 stars
docker_ci
The framework to generate a Dockerfile, build, test, and deploy a docker image with OpenVINO™ toolkit.

Python · 70 stars
openvino_tokenizers
OpenVINO Tokenizers extension

C++ · 57 stars
physicalai

Python · 28 stars
npu_compiler_elf

C++ · 7 stars

Beginner friendly Everything

Open All

OVMS tools call for Gemma 4 E4B failed Open

Difficulty 3/5 1-2 days Newbie friendliness 55/100

openvinotoolkit/model_server#4586 · Opened Sep 18, 2026
[Bug]: bf16 MatMul/GEMM returns corrupted output (cos 0.28-0.65, 5-135x magnitude) - f16 and f32 controls are correct Open

bug support_request

Difficulty 4/5 3-5 days Newbie friendliness 65/100

openvinotoolkit/openvino#38256 · Opened Sep 18, 2026
[Bug]: Qwen3.8-27B with MTP cannot run across 2x Arc Pro B70: PIPELINE_PARALLEL crashes in igdrcl64.dll, TENSOR_PARALLEL unsupported Open

Difficulty 5/5 Over a week Newbie friendliness 30/100

openvinotoolkit/openvino#38253 · Opened Sep 18, 2026
[Bug]: HETERO:GPU.0,GPU.1 + PIPELINE_PARALLEL silently places whole model on GPU.0 when it fits one card (2x Arc Pro B70, Qwen3.8-27B) Open

Difficulty 4/5 3-5 days Newbie friendliness 55/100

openvinotoolkit/openvino#38250 · Opened Sep 18, 2026
[Bug]: CPU and GPU plugins ignore Divide's m_pythondiv for integer inputs: -5 / 2 gives -2 instead of -3 Open

Difficulty 4/5 3-5 days Newbie friendliness 68/100

openvinotoolkit/openvino#38248 · Opened Sep 18, 2026
[Bug]: [PyTorch Frontend] aten::narrow with a negative start returns an empty tensor Open

Difficulty 2/5 1-3 hours Newbie friendliness 88/100

openvinotoolkit/openvino#38247 · Opened Sep 18, 2026
[Bug]: ROI tensor keeps its owner's strides after set_shape() expands the rank — ITensor::copy_to() then reads out of bounds (SIGSEGV in production) Open

bug support_request

Difficulty 3/5 1-2 days Newbie friendliness 68/100

openvinotoolkit/openvino#38246 · 1 comment · Opened Sep 18, 2026
OVMS requiring manual WEIGHTS_PATH Open

bug

Difficulty 4/5 3-5 days Newbie friendliness 52/100

openvinotoolkit/model_server#4580 · 1 comment · Opened Sep 18, 2026
[Bug]: [GPU] StridedSlice/Slice with non-constant strides is skipped as identity when the output shape equals the input shape (e.g. reversal with step -1) Open

Difficulty 3/5 1-2 days Newbie friendliness 78/100

openvinotoolkit/openvino#38240 · Opened Sep 17, 2026
[Bug]: PyTorch frontend computes `sum((x-y)^p)^(1/p)` for `aten::pairwise_distance` / `aten::cdist` — the `Abs()` is missing, so every `p != 2` is wrong or NaN Open

bug support_request

Difficulty 3/5 1-2 days Newbie friendliness 76/100

openvinotoolkit/openvino#38239 · 1 comment · Opened Sep 17, 2026
[Bug]: ONNX frontend ignores the `axes` attribute of `ReduceL1` / `ReduceLogSum` / `ReduceLogSumExp` — opset 13–17 models silently reduce over all axes Open

bug support_request

Difficulty 2/5 1-3 hours Newbie friendliness 88/100

openvinotoolkit/openvino#38238 · Opened Sep 17, 2026
[Feature request] Structured M:N (2:4) magnitude-based pruning Open

Difficulty 5/5 Over a week Newbie friendliness 45/100

openvinotoolkit/nncf#4222 · 1 comment · Opened Sep 17, 2026
[aw] Failed jobs: CI Failure Doctor — Post-Commit Open

agentic-workflows

Difficulty 4/5 3-5 days Newbie friendliness 25/100

openvinotoolkit/openvino#38237 · Opened Sep 17, 2026
[aw] Failed jobs: CI Failure Doctor — Post-Commit Open

agentic-workflows

Difficulty 3/5 1-2 days Newbie friendliness 35/100

openvinotoolkit/openvino#38236 · Opened Sep 17, 2026
[Bug]: [GPU] SIGSEGV in IGC/libopencl-clang2 kernel compilation during MTP draft-model init (Arc Pro B70 / Xe2) — 50% of launches, no-MTP path stable Open

Difficulty 5/5 Over a week Newbie friendliness 30/100

openvinotoolkit/openvino#38235 · 1 comment · Opened Sep 17, 2026
[Bug]: benchmark_app fails with IndexError on an empty entry in a device list Open

Difficulty 2/5 1-3 hours Newbie friendliness 88/100

openvinotoolkit/openvino#38226 · Opened Sep 17, 2026
[aw] CI Failure Doctor — Merge Queue produced no safe outputs Open

agentic-workflows

Difficulty 4/5 3-5 days Newbie friendliness 38/100

openvinotoolkit/openvino#38218 · 2 comments · Opened Sep 17, 2026
[Build]: APT upgrade from 2026.3.1 to 2026.4.0 fails due to missing Conflicts entries Open

support_request

openvinotoolkit/openvino#38217 · 1 assignee · Opened Sep 17, 2026
[Bug][GPU]: int8 Qwen3-Coder-30B-A3B on Arc 140T fails long prompts — grouped MoE gemm has no dispatchable kernel ("could not create a primitive descriptor for the matmul primitive") Open

Difficulty 4/5 3-5 days Newbie friendliness 38/100

openvinotoolkit/openvino#38211 · Opened Sep 17, 2026
[aw] Failed jobs: CI Failure Doctor — Post-Commit Open

agentic-workflows

Difficulty 4/5 3-5 days Newbie friendliness 35/100

openvinotoolkit/openvino#38203 · Opened Sep 16, 2026
[aw] CI Failure Doctor — Post-Commit failed Open

agentic-workflows

Difficulty 4/5 3-5 days Newbie friendliness 28/100

openvinotoolkit/openvino#38202 · 2 comments · Opened Sep 16, 2026
[Bug]: [GPU][WSL2] Qwen3 model compilation fails with USM Host allocation error Open

bug support_request

Difficulty 4/5 3-5 days Newbie friendliness 42/100

openvinotoolkit/openvino#38199 · 1 comment · Opened Sep 16, 2026
[Bug]: benchmark_app crashes when an argument value contains '=' or is empty Open

Difficulty 2/5 1-3 hours Newbie friendliness 82/100

openvinotoolkit/openvino#38196 · Opened Sep 16, 2026
[aw] CI Failure Doctor — Merge Queue has stale lock file Open

agentic-workflows

Difficulty 1/5 Under an hour Newbie friendliness 88/100

openvinotoolkit/openvino#38195 · Opened Sep 16, 2026
[Bug]: Paddle frontend `scatter` with `overwrite=False` zeroes rows that are not indexed — silently wrong result Open

bug support_request

Difficulty 3/5 1-2 days Newbie friendliness 78/100

openvinotoolkit/openvino#38185 · Opened Sep 16, 2026
When will MiniCPM-V-4.6 be supported? Open

category: visual language

openvinotoolkit/openvino.genai#4486 · 1 assignee · Opened Sep 16, 2026
llm_bench (GenAI path) reports "2nd tokens throughput" inflated by an extra factor of batch_size when batch_size > 1 Open

openvinotoolkit/openvino.genai#4484 · 1 assignee · Opened Sep 16, 2026
[Bug]: Rank-1 Abs is silently removed on CPU when its input is derived from constants Open

bug support_request

Difficulty 4/5 3-5 days Newbie friendliness 55/100

openvinotoolkit/openvino#38166 · Opened Sep 15, 2026
MTP speculative decoding fails (-14 / -5 CL errors) or becomes 3× slower beyond ~64K context on Xe2 (hybrid GDN model); no clean budget error before dispatch Open

openvinotoolkit/openvino.genai#4483 · 1 assignee · Opened Sep 15, 2026
[aw] CI Failure Doctor — Merge Queue produced no safe outputs Open

agentic-workflows

Difficulty 3/5 1-2 days Newbie friendliness 35/100

openvinotoolkit/openvino#38163 · 3 comments · Opened Sep 15, 2026
cache_size (GB) does not bound the KV cache pool: with cache_size > 0 the pool grows unbounded and clear_cache() is never called Open

openvinotoolkit/openvino.genai#4481 · 1 assignee · Opened Sep 15, 2026
[Bug]: ONNX frontend Cast to int4/uint4 rounds half away from zero, while the ONNX int4 spec cited in the code comment requires ties-to-even Open

bug support_request

Difficulty 2/5 1-3 hours Newbie friendliness 88/100

openvinotoolkit/openvino#38158 · Opened Sep 15, 2026
[Performance]: [GPU] PagedAttention with 2-4 query tokens per sequence (speculative decoding verify) is ~8x slower than single-token decode at the same context Open

performance support_request

openvinotoolkit/openvino#38153 · 3 comments · 1 assignee · Opened Sep 15, 2026
Investigate extracting so101 and widowx as dedicated plugins Open

openvinotoolkit/physicalai#275 · 1 assignee · Opened Sep 15, 2026
[Bug]: benchmark_app silently ignores a per-device value with an extra colon Open

Difficulty 2/5 1-3 hours Newbie friendliness 86/100

openvinotoolkit/openvino#38143 · Opened Sep 15, 2026
[Bug]: LTX2 bf16 on CPU produces grey/bad quality videos Open

bug support_request

Difficulty 4/5 3-5 days Newbie friendliness 48/100

openvinotoolkit/openvino#38137 · Opened Sep 14, 2026
Greedy speculative decoding changes deterministic output with Qwen3 target and draft Open

openvinotoolkit/openvino.genai#4476 · 2 comments · 1 assignee · Opened Sep 14, 2026
[Bug]: [PyTorch Frontend] Incorrect translation for aten::slice when step is negative and start/end bounds are omitted Open

bug support_request

Difficulty 3/5 1-2 days Newbie friendliness 78/100

openvinotoolkit/openvino#38136 · 2 comments · Opened Sep 14, 2026
[aw] CI Failure Doctor — Merge Queue produced no safe outputs Open

agentic-workflows

Difficulty 4/5 3-5 days Newbie friendliness 35/100

openvinotoolkit/openvino#38134 · 2 comments · Opened Sep 14, 2026
Investigate merging the `physicalai-studio-plugin` back into `physicalai` Open

openvinotoolkit/physicalai#269 · 2 assignees · Opened Sep 14, 2026
[BUG] Qwen3.8-27B INT4 compression corrupts GatedDeltaNet (requires ignored_scope) Open

Difficulty 4/5 3-5 days Newbie friendliness 55/100

openvinotoolkit/nncf#4217 · 2 comments · Opened Sep 14, 2026
[Bug]: ovc keeps a trailing comma in --input as part of the tensor name Open

Difficulty 2/5 1-3 hours Newbie friendliness 35/100

openvinotoolkit/openvino#38104 · Opened Sep 13, 2026
[Bug] GPU plugin f16: 4-input channel Concat of 180/228 channels at <=32x32 produces wrong data on Arc B580 (breaks YOLOv9-m) Open

openvinotoolkit/openvino#38102 · 1 comment · 2 assignees · Opened Sep 13, 2026
[Bug]: OpenVINO's GPU plugin segfaults on intel 12th gen cpu Open

bug support_request

openvinotoolkit/openvino#38101 · 6 comments · 2 assignees · Opened Sep 13, 2026
[Bug]: LFM2.5 350M and 1.2B silently generate garbage on NPU DEVICE_ARCHITECTURE=4000 (Lunar Lake) — correct on 3720, CPU and GPU Open

category: NPU PSE

openvinotoolkit/openvino#38100 · 1 comment · 2 assignees · Opened Sep 13, 2026
[GPU] Deterministic wrong values: batched unrolled forward-substitution, slice 0 correct and every other chunk-slice wrong (2026.4.0-22849 and master 23099 nightly) Open

openvinotoolkit/openvino#38099 · 3 comments · 2 assignees · Opened Sep 12, 2026
stou32() accepts partial parses, unlike the other stringutils converters Open

Difficulty 2/5 1-3 hours Newbie friendliness 35/100

openvinotoolkit/model_server#4554 · Opened Sep 12, 2026
Transcription temperature: malformed values are silently reinterpreted instead of rejected Open

openvinotoolkit/model_server#4552 · 1 assignee · Opened Sep 11, 2026
Download progress bar divides by zero without Content-Length and prints ~2.1 billion spaces per tick Open

openvinotoolkit/model_server#4550 · 1 assignee · Opened Sep 11, 2026
--source_loras misreads a URL port as the :alpha suffix and refuses to start Open

Difficulty 2/5 1-3 hours Newbie friendliness 45/100

openvinotoolkit/model_server#4545 · Opened Sep 11, 2026

Showing the newest 100

This page lists what was indexed most recently. Advanced filter has the whole inventory, narrowed by language, difficulty and how long a task takes.

Open advanced filter

openvinotoolkit

Showing the newest 100

Get new issues in your inbox