intel/intel-extension-for-pytorch issues and pull requests

#411 - ipexrun: --auto-ipex setting lead to python crush

Issue - State: open - Opened by WolfKingAlex about 1 year ago - 2 comments
Labels: CPU, Crash

#410 - Lack of torchvision corresponding to torch==2.0.0a0 under windows

Issue - State: open - Opened by JellyBean326 about 1 year ago - 1 comment
Labels: XPU/GPU, UserExperience, Windows

#409 - OSError: [WinError 126] The specified module could not be found.

Issue - State: closed - Opened by sifatk69 about 1 year ago - 16 comments
Labels: XPU/GPU, UserExperience, Windows

#408 - DPC++ extension fails on native Windows with latest ipex

Issue - State: open - Opened by rnwang04 about 1 year ago - 6 comments
Labels: XPU/GPU, Crash, Windows

#407 - Safetensor loading having a function overwritten by IPEX doesn't work

Issue - State: closed - Opened by simonlui about 1 year ago - 7 comments
Labels: XPU/GPU, Crash

#406 - Is there any plan to support intel iGPU?

Issue - State: open - Opened by rnwang04 about 1 year ago - 5 comments
Labels: XPU/GPU, iGPU, Query

#405 - frontend.py spelling mistakes

Pull Request - State: open - Opened by albusDalbador about 1 year ago

#404 - XPU failed to compile with latest DPC++ on Windows

Issue - State: open - Opened by Gentoli about 1 year ago - 3 comments
Labels: XPU/GPU, Compilation

#403 - RuntimeError: The size of tensor a (32) must match the size of tensor b (33) at non-singleton dimension 3

Issue - State: closed - Opened by jerin-scalers-ai about 1 year ago - 5 comments
Labels: CPU, NotAnIssue

#402 - torch.multinomial behave unexpectedly on float16 GPU input tensor

Issue - State: open - Opened by rnwang04 about 1 year ago - 3 comments
Labels: XPU/GPU, Correctness

#401 - fix(nit): typo in name of modeling task

Pull Request - State: open - Opened by rahulunair about 1 year ago

#400 - "could not execute a primitive" when using tanh inside of model class on multidimensional data

Issue - State: open - Opened by gekeleda about 1 year ago - 8 comments
Labels: ARC, Crash

#399 - [IPEX][XPU][Windows 11] It takes forever to run the first pass

Issue - State: open - Opened by whchan05 about 1 year ago - 28 comments
Labels: XPU/GPU, Performance, Windows

#398 - IPEX xpu wheel files for windows using the wrong version of Pytorch? Also no torchvision

Issue - State: open - Opened by Mindset-Official about 1 year ago - 18 comments
Labels: XPU/GPU, NotAnIssue

#397 - prebuilt wheels for windows native missing?

Issue - State: closed - Opened by Vipitis about 1 year ago - 5 comments
Labels: Documentation, XPU/GPU

#396 - can't import intel_extension_for_pytorch for XPU after recommended installation

Issue - State: open - Opened by bconsolvo about 1 year ago - 3 comments
Labels: XPU/GPU, NotAnIssue

#395 - Fail to build tag v2.1.0.dev+cpu.llm

Issue - State: closed - Opened by kenplusplus about 1 year ago - 13 comments
Labels: CPU, Query

#394 - raise err OSError: [WinError 126] 找不到指定的模块。

Issue - State: closed - Opened by TigerShuai about 1 year ago - 13 comments
Labels: ARC, UserExperience

#393 - fix typo

Pull Request - State: closed - Opened by KimSoungRyoul about 1 year ago - 3 comments

#392 - Is there support for embedding on xpu?

Issue - State: closed - Opened by evelinamorim about 1 year ago - 2 comments
Labels: XPU/GPU, UserExperience

#391 - Segmentation fault (core dumped) when execute model.to("xpu")

Issue - State: closed - Opened by evelinamorim about 1 year ago - 13 comments
Labels: XPU/GPU, UserExperience

#390 - xpu-master build for IGPU error

Issue - State: closed - Opened by kyflores about 1 year ago - 6 comments
Labels: iGPU, NotAnIssue, Compilation

#389 - Enable inductor for xpu backend

Pull Request - State: closed - Opened by guangyey about 1 year ago - 1 comment

#388 - Under WSL, torch.xpu.empty_cache() will leak VRAM in specific circumstances

Issue - State: open - Opened by a-One-Fan about 1 year ago - 5 comments
Labels: XPU/GPU, Crash

#387 - Unexplainable bf16 performance drop when using numactl to bind specific cores

Issue - State: open - Opened by Spycsh about 1 year ago - 6 comments
Labels: CPU, Performance

#386 - It takes more than 100ms to issue a single command to Intel Arc GPU

Issue - State: closed - Opened by BA8F0D39 about 1 year ago - 3 comments
Labels: ARC, Performance

#385 - `torch.xpu.get_device_properties` lacks most of device properties

Issue - State: closed - Opened by fcharras about 1 year ago - 1 comment
Labels: XPU/GPU, Feature

#384 - Trying to run hugging face GPT2 out-of-box and got "Aborted (core dumped)"

Issue - State: closed - Opened by kelayamatoz about 1 year ago - 2 comments
Labels: XPU/GPU, Crash

#383 - xpu-master docker xpu-flex fails README.md example and torchvision warning

Issue - State: open - Opened by retoXD about 1 year ago - 7 comments
Labels: Documentation, XPU/GPU

#382 - Windows build guide (xpu-master branch) uses wildcards but git doesn't expand them

Issue - State: closed - Opened by a-One-Fan about 1 year ago - 2 comments
Labels: Documentation, XPU/GPU

#381 - Version for Windows

Issue - State: open - Opened by My12123 about 1 year ago - 3 comments
Labels: CPU, XPU/GPU, Feature, Windows

#380 - Fix curl url in dockerfile

Pull Request - State: closed - Opened by petitfou about 1 year ago - 1 comment

#379 - Is torch==1.13.0a0 available somewhere?

Issue - State: open - Opened by pablodegrande over 1 year ago - 1 comment
Labels: XPU/GPU, NotAnIssue

#378 - Weird ipex_optimize training loop printing behavior

Issue - State: open - Opened by fredlarochelle over 1 year ago - 2 comments
Labels: XPU/GPU, UserExperience

#377 - Can not find transformers in your environment, disable ipex transformer optimize

Issue - State: open - Opened by fredlarochelle over 1 year ago - 3 comments
Labels: XPU/GPU, UserExperience

#376 - torch._convolution runtime error after quantizing model for inference C++

Issue - State: closed - Opened by saduf over 1 year ago - 6 comments
Labels: CPU, Crash

#375 - ipex.optimize -> linear_prepack throws an "RuntimeError: could not create a primitive descriptor" error

Issue - State: open - Opened by TheMrCodes over 1 year ago - 4 comments
Labels: ARC, Crash

#374 - [docker][xpu] Requesting ipex 1.13.120+xpu docker image

Issue - State: closed - Opened by Nuullll over 1 year ago - 1 comment
Labels: NotAnIssue

#373 - Latest xpu-master branch compiling issue

Issue - State: closed - Opened by fredlarochelle over 1 year ago - 11 comments
Labels: XPU/GPU, Crash

#372 - `scatter_add_` performance

Issue - State: closed - Opened by fcharras over 1 year ago - 2 comments
Labels: Performance, dGPU-Flex

#371 - Core dump when using wandb on PVC

Issue - State: closed - Opened by rahulunair over 1 year ago - 2 comments
Labels: XPU/GPU, Crash

#370 - module 'intel_extension_for_pytorch._C' has no attribute '_setDevice'

Issue - State: open - Opened by hipudding over 1 year ago
Labels: CPU, Crash

#369 - `dlpack` interface with `dpctl` has unexpected behavior with F-aligned arrays ?

Issue - State: closed - Opened by fcharras over 1 year ago - 1 comment
Labels: CPU, NotAnIssue

#368 - `dlpack` interface with `dpctl` fails on CPU (in both ways)

Issue - State: closed - Opened by fcharras over 1 year ago - 5 comments
Labels: XPU/GPU, Crash

#367 - Patches to pytorch don't apply in compile_bundle.sh

Issue - State: closed - Opened by kyflores over 1 year ago - 8 comments
Labels: XPU/GPU, UserExperience

#366 - Create cmake.yml

Pull Request - State: open - Opened by nilarmyint1 over 1 year ago - 2 comments

#365 - Failed to install the GPU software packages

Issue - State: closed - Opened by jfsunx over 1 year ago - 3 comments
Labels: XPU/GPU, UserExperience

#364 - XPU Installation Instructions Don't Work

Issue - State: open - Opened by coreyjadams over 1 year ago - 43 comments
Labels: XPU/GPU, UserExperience

#363 - Update intel-opencl-icd to 23.13.26032.26-627~22.04

Pull Request - State: open - Opened by aslanxie over 1 year ago

#362 - Could not running simple example on XPU

Issue - State: open - Opened by harborn over 1 year ago - 17 comments
Labels: XPU/GPU, UserExperience

#361 - Can't install 1.13.120+xpu in WSL2? Broken link?

Issue - State: open - Opened by Mindset-Official over 1 year ago - 8 comments
Labels: XPU/GPU, NotAnIssue

#360 - Request intel xe graphic and intel xe max gpu support for stable diffusion

Issue - State: closed - Opened by Nathan-dm over 1 year ago - 16 comments
Labels: iGPU, Feature

#359 - Native API failed

Issue - State: closed - Opened by stacksmash76 over 1 year ago - 7 comments
Labels: ARC, Crash

#358 - `torch.linalg.svd` segfaults when `full_matrices` parameter is set to `False`

Issue - State: closed - Opened by fcharras over 1 year ago - 5 comments
Labels: ARC, Crash

#357 - Compiling issues with libpng

Issue - State: open - Opened by fredlarochelle over 1 year ago - 3 comments
Labels: XPU/GPU, UserExperience

#356 - model weight is lost after ipex inplace optimize in version 2.0.100

Issue - State: open - Opened by rnwang04 over 1 year ago - 2 comments
Labels: CPU, Crash

#355 - Python 3.11 support for xpu-master

Issue - State: closed - Opened by pioto1225 over 1 year ago - 5 comments
Labels: XPU/GPU, Query

#354 - WARNING - Core binding with numactl is not available

Issue - State: closed - Opened by NathanJHLee over 1 year ago - 2 comments
Labels: CPU, Crash

#353 - Engine error when using Stable-diffusion-webui

Issue - State: open - Opened by kagerou8124 over 1 year ago - 2 comments
Labels: Crash, dGPU-Flex

#352 - huggingface accelerate supported?

Issue - State: closed - Opened by human-01 over 1 year ago - 2 comments
Labels: XPU/GPU, Ecosystem

#351 - Interpolation with default params on XPU differs from CPU execution

Issue - State: closed - Opened by donggangj over 1 year ago - 3 comments
Labels: ARC, Correctness

#350 - XPU Neural Networks support

Issue - State: open - Opened by troy818 over 1 year ago - 2 comments
Labels: XPU/GPU, Query

#349 - Windows support for PyTorch extensions

Issue - State: open - Opened by AngreiG over 1 year ago - 9 comments
Labels: XPU/GPU

#348 - Pytorch inference time issue

Issue - State: open - Opened by rubende over 1 year ago - 15 comments
Labels: CPU, Performance

#347 - IPEX AVX2 Import Error despite AVX2 Support

Issue - State: open - Opened by fredlarochelle over 1 year ago - 6 comments
Labels: ARC, Crash

#346 - xpu hardware support

Issue - State: open - Opened by mgrabban over 1 year ago - 4 comments
Labels: XPU/GPU, Query

#345 - `torch.topk` is two order of magnitude too slow on xpu (flex 170)

Issue - State: closed - Opened by fcharras over 1 year ago - 12 comments
Labels: Performance, Flex

#344 - Broken license metadata

Issue - State: closed - Opened by bstriner over 1 year ago - 3 comments
Labels: CPU

#343 - Fix license metadata

Pull Request - State: closed - Opened by bstriner over 1 year ago - 1 comment

#342 - Which dependency versions were used in the 1.13.120+xpu build?

Issue - State: closed - Opened by fhaust over 1 year ago - 3 comments
Labels: XPU/GPU, Query

#341 - Regex should be using a raw string

Pull Request - State: closed - Opened by sterliakov over 1 year ago - 1 comment

#340 - torch==1.13.1 link

Issue - State: open - Opened by TotalDay over 1 year ago - 5 comments
Labels: Documentation, XPU/GPU

#339 - Should this be working on my iGPU?

Issue - State: open - Opened by fhaust over 1 year ago - 14 comments
Labels: iGPU, Query

#338 - IPEX does not support fused/fused split loss

Issue - State: open - Opened by bconsolvo over 1 year ago
Labels: CPU, Crash

#337 - IPEX DPC++ extension ImportError: libtbb.so.12: cannot open shared object file

Issue - State: open - Opened by troy818 over 1 year ago - 13 comments
Labels: ARC, Crash

#336 - enhancement: print stack when ipex.optimize fail

Pull Request - State: closed - Opened by shaoyuta over 1 year ago - 4 comments

#335 - Error calculating loss RuntimeError: Native API failed. Native API returns: -50 (PI_ERROR_INVALID_ARG_VALUE) -50 (PI_ERROR_INVALID_ARG_VALUE)

Issue - State: closed - Opened by turbobuilt over 1 year ago - 5 comments
Labels: ARC, Crash

GitHub / intel/intel-extension-for-pytorch issues and pull requests