Ecosyste.ms: Issues

An open API service for providing issue and pull request metadata for open source projects.

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

#522 - Provide an API to get GPU global free and used memory

Issue - State: open - Opened by notsyncing 10 months ago - 2 comments
Labels: XPU/GPU, Feature

#521 - Segmentation fault with Intel GPU when LLM generating with transformers and accelerate

Issue - State: closed - Opened by notsyncing 10 months ago - 5 comments
Labels: ARC, Crash

#520 - add build system

Issue - State: closed - Opened by martin19 10 months ago - 2 comments
Labels: XPU/GPU, Query, Compilation

#519 - [Alchemist GPUs] Corruption at 1024x1024 resolutions with Diffusion workloads

Issue - State: closed - Opened by Disty0 10 months ago - 2 comments
Labels: ARC, Correctness, LLM

#518 - Update README.md

Pull Request - State: closed - Opened by lbreitk 10 months ago - 1 comment

#517 - Is AMX supported for LLM inference?

Issue - State: closed - Opened by Hyungyo1 10 months ago - 7 comments
Labels: CPU, Query

#516 - intel/intel-extension-for-pytorch:xpu-jupyter does not include jupyter server

Issue - State: closed - Opened by tkatila 10 months ago - 11 comments
Labels: XPU/GPU, UserExperience

#515 - Can't build Ipex from source with compile_bundle.sh for the version v2.1.0+cpu

Issue - State: closed - Opened by JingyaHuang 10 months ago - 13 comments
Labels: CPU, Compilation

#514 - Error while setting up LLM environment

Issue - State: closed - Opened by azhuvath 10 months ago - 2 comments
Labels: CPU, NotAnIssue, LLM

#513 - xpu.synchronize hangs

Issue - State: closed - Opened by chsasank 10 months ago - 17 comments
Labels: ARC, UserExperience

#512 - Qwen-14B-chat support？

Issue - State: closed - Opened by Storm0921 10 months ago - 7 comments
Labels: Query, LLM

#511 - Extra cd .. is not required

Pull Request - State: open - Opened by apstasen 10 months ago

#510 - Conv2d + ConvTranpose2d: "Cannot access data pointer of Tensor that doesn't have storage" on backward

Issue - State: closed - Opened by KoKuToru 10 months ago - 7 comments
Labels: ARC, Crash

#509 - Server down for ipex installation

Issue - State: closed - Opened by michaelmalice 10 months ago - 3 comments
Labels: XPU/GPU, UserExperience

#508 - `RuntimeError: could not create a primitive` for `torch.matmul` on Arc A730M and Arc A750 for Windows

Issue - State: open - Opened by Oscilloscope98 10 months ago - 16 comments
Labels: ARC, Crash, Windows

#507 - Clarity on compatible pytorch versions for xpu ipex?

Issue - State: closed - Opened by garrett361 10 months ago - 4 comments
Labels: XPU/GPU, Query

#506 - full name of DDP

Issue - State: closed - Opened by jinz2014 10 months ago - 1 comment
Labels: Query

#505 - Index kernel issue.

Issue - State: closed - Opened by jikunshang 10 months ago - 12 comments
Labels: Bug, ARC, Crash, high priority

#504 - [General query] How to get sycl::device?

Issue - State: open - Opened by Sarbojit2019 10 months ago - 2 comments
Labels: XPU/GPU, Query

#503 - Caching JIT to local disk to speed up

Issue - State: open - Opened by CarlGao4 10 months ago - 4 comments
Labels: XPU/GPU, Feature

#502 - Fix weight-only quantization on OPT model

Pull Request - State: closed - Opened by doudoubobo 10 months ago - 3 comments

#501 - Report errors when running weight-only quantization on OPT

Issue - State: closed - Opened by doudoubobo 10 months ago - 2 comments
Labels: CPU, Crash, LLM

#500 - folding failed during the optimize process.

Issue - State: open - Opened by jinz2014 10 months ago

#499 - Build 2.1.10+xpu with AOT fails with LINK1328

Issue - State: closed - Opened by CarlGao4 11 months ago - 2 comments
Labels: XPU/GPU, Compilation

#498 - 安装包没有更新到pypl吗

Issue - State: closed - Opened by Metasymbols 11 months ago - 1 comment
Labels: CPU, XPU/GPU, Query

#497 - torch._dynamo.exc.InvalidBackend: Invalid backend: 'ipex', see `torch._dynamo.list_backends()` for available backends.

Issue - State: open - Opened by dlod-openvino 11 months ago - 2 comments
Labels: XPU/GPU, Feature, Windows

#496 - Dead link to Installation Guide

Issue - State: closed - Opened by saforem2 11 months ago - 3 comments
Labels: Documentation, XPU/GPU

#495 - Install ipex main branch conda env will meet an error

Issue - State: closed - Opened by hzjane 11 months ago - 7 comments
Labels: Bug, CPU, Compilation

#494 - Inquiry Regarding Multi-GPU Support in Intel Extension for PyTorch

Issue - State: closed - Opened by kunger97 11 months ago - 4 comments
Labels: UserExperience, dGPU-Max

#493 - Nested Tensor Masks Not Supported

Issue - State: open - Opened by ElliottDyson 11 months ago

#492 - Allocation of greater than 4GB not possible

Issue - State: closed - Opened by ElliottDyson 11 months ago - 2 comments
Labels: XPU/GPU, UserExperience

#490 - Bump transformers from 4.35.0 to 4.36.0 in /examples/cpu/serving/triton

Pull Request - State: closed - Opened by dependabot[bot] 11 months ago - 2 comments
Labels: dependencies

#489 - Inference with inputs with sequence length increased will cause OOM

Issue - State: open - Opened by Ricky-Ting 11 months ago - 1 comment
Labels: ARC, Crash, LLM

#488 - Whisper: could not create a primitive descriptor for a concat primitive

Issue - State: open - Opened by JKlesmith 11 months ago - 2 comments
Labels: ARC, Crash, LLM

#487 - IPEX not working with Stable Diffusion Face Restoration Tools (facexlib)

Issue - State: closed - Opened by HyunJae5463 11 months ago - 14 comments
Labels: ARC, Correctness, Windows, LLM

#486 - Crash in cpu_info.py when using german locales

Issue - State: closed - Opened by daFritz84 11 months ago - 5 comments
Labels: Bug, CPU, Crash

#484 - IPEX v2.1.1+cpu crashes PyTorch for model where v1.13.100+cpu worked

Issue - State: open - Opened by droberts195 11 months ago - 3 comments
Labels: CPU, Crash

#483 - IPEX 2.1.10+xpu regression with Stable Diffusion workloads that were working with IPEX 2.0.120+xpu on WSL2/Linux

Issue - State: closed - Opened by simonlui 11 months ago - 24 comments
Labels: ARC, Crash, LLM

#482 - Fix running Triton Server Docker container as the root user

Pull Request - State: closed - Opened by mzyczyns 11 months ago

#480 - ValueError: not enough values to unpack, when loading transformer model using bfloat16/float16

Issue - State: closed - Opened by Nehereus 11 months ago - 3 comments
Labels: ARC, Crash

#479 - FFT RuntimeError: FFT_INVALID_DESCRIPTOR

Issue - State: closed - Opened by jedcheng 11 months ago - 6 comments
Labels: Crash

#478 - Could not find a version that satisfies the requirement torch==2.0.1a0

Issue - State: open - Opened by tuhinbiswas98 12 months ago - 1 comment
Labels: ARC, UserExperience

#477 - (ipex 1.13.120+xpu) undefined symbol: _ZNK3c107SymBool10guard_boolEPKcl

Issue - State: closed - Opened by ytzhang1 12 months ago - 2 comments
Labels: XPU/GPU, Crash

#476 - CPU memory leak during inference on Arc A770

Issue - State: closed - Opened by plusbang 12 months ago - 22 comments
Labels: ARC, UserExperience, high priority

#475 - Why not use oneDNN?

Issue - State: open - Opened by tonyzhao-jt 12 months ago - 3 comments
Labels: CPU, Query

#474 - source build for fails for ipex on windows 11

Issue - State: closed - Opened by rairatne 12 months ago - 6 comments
Labels: ARC, Compilation, Windows

#473 - Import intel_extension_for_pytorch fails on windows11

Issue - State: closed - Opened by rairatne 12 months ago - 20 comments
Labels: ARC, UserExperience, Windows

#472 - Fetch corresponding `sycl::queue` for CPU device

Issue - State: closed - Opened by XianFuWongIntel 12 months ago - 2 comments
Labels: XPU/GPU, Query

#471 - Intel ARC 770 Segmentation fault

Issue - State: open - Opened by JannisFengler 12 months ago - 2 comments
Labels: ARC, Crash

#470 - Abnormal slow training speed when enable gradient_checkpointing on GPU

Issue - State: closed - Opened by rnwang04 about 1 year ago
Labels: XPU/GPU, NotAnIssue

#469 - README.md instruction doesn't work

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

#467 - `index_select` fails on XPU when big indices are provided

Issue - State: closed - Opened by DamianSzwichtenberg about 1 year ago - 11 comments
Labels: Crash, dGPU-Max

#466 - When I build the intel-extension-for-pytorch , I got the error that it cannot find xpu compiler path from OS PATH

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

#465 - Why is there no windows version of torchvision?

Issue - State: open - Opened by oranges7 about 1 year ago - 1 comment
Labels: XPU/GPU, Query, Windows

#464 - Triton Server with IPEX backend example

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

#463 - "import matplotlib" will make is_xpu_available() return False

Issue - State: open - Opened by notsyncing about 1 year ago - 3 comments
Labels: ARC, UserExperience

#462 - CPU memory leak lora finetuning on XPU

Issue - State: open - Opened by yangw1234 about 1 year ago - 3 comments
Labels: CPU, UserExperience, dGPU-Max

#461 - INT4 Quantization

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

#460 - About the number of model load times.

Issue - State: open - Opened by gukejun1 about 1 year ago - 1 comment
Labels: Query

#459 - torch.bmm on igpu causes serious system serious unknown errors and leads to system crashes.

Issue - State: closed - Opened by leonardozcm about 1 year ago - 1 comment
Labels: iGPU, Crash, LLM

#457 - ImportError

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

#456 - AttributeError: module 'intel_extension_for_pytorch.quantization' has no attribute 'WoqActQuantMode'

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

#455 - What is the call order of get_accelerator?

Issue - State: closed - Opened by gukejun1 about 1 year ago
Labels: XPU/GPU, Query

#454 - Bicubic Interpolation output on XPU differs from CPU.

Issue - State: open - Opened by lalalapotter about 1 year ago
Labels: ARC, Correctness

#453 - Arc A770 Performance is down by 17% when update linux kernel from 5.19.0-41 to 6.2.0-35

Issue - State: open - Opened by qiuxin2012 about 1 year ago - 3 comments
Labels: ARC, Performance

#452 - Failed to run comfyui with SDXL model

Issue - State: open - Opened by 771991673 about 1 year ago - 2 comments
Labels: XPU/GPU, Crash

#451 - intel-extension-for-pytorch vs. intel-extension-for-transformers

Issue - State: open - Opened by NaamaKadosh about 1 year ago - 2 comments
Labels: Query

#450 - Illegal instruction (core dumped) in intel N305

Issue - State: open - Opened by joebnb about 1 year ago - 3 comments
Labels: UserExperience, iGPU

#449 - torch 2.0.1a0 not available on windows

Issue - State: closed - Opened by ElliottDyson about 1 year ago - 6 comments
Labels: XPU/GPU, NotAnIssue, Windows

#448 - xpu Embedding layer will cause the xpu device to be lost on Windows

Issue - State: open - Opened by MeouSker77 about 1 year ago - 2 comments
Labels: ARC, UserExperience, Windows

#447 - Update README.md

Pull Request - State: closed - Opened by imperial-chief about 1 year ago - 4 comments

#446 - How to solve "NotImplementedError: Cannot copy out of meta tensor; no data!"?

Issue - State: closed - Opened by gukejun1 about 1 year ago - 8 comments
Labels: CPU, Crash, LLM

#445 - The specified module could not be found.

Issue - State: closed - Opened by sifatk69 about 1 year ago - 6 comments
Labels: XPU/GPU, UserExperience, Windows

#444 - `import intel_extension_for_pytorch` breaks `cv2.imshow()`

Issue - State: closed - Opened by Wovchena about 1 year ago - 4 comments
Labels: CPU, UserExperience

#443 - Segmentation fault when doing backward for loss func

Issue - State: closed - Opened by yinoqifu00 about 1 year ago - 11 comments
Labels: CPU, Crash, high priority

#442 - Drivers from 4885 and newer break IPEX for native windows.

Issue - State: open - Opened by Mindset-Official about 1 year ago - 14 comments
Labels: ARC, UserExperience, Windows

#441 - The compile_bundle.sh script fails at ipex build step

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

#440 - [Question] Why not to use OpenVINO for extending PyTorch as it is already doing ?

Issue - State: open - Opened by redradist about 1 year ago
Labels: CPU, XPU/GPU, Query

#439 - `torch.linalg.eigh` is significantly slower than expected on Max Series GPU

Issue - State: open - Opened by ogrisel about 1 year ago - 3 comments
Labels: Performance, dGPU-Max

#438 - Failed to do quantization for models like EleutherAI/gpt-neox-20b and bigscience/bloom-7b1

Issue - State: open - Opened by RenyanDiao about 1 year ago - 3 comments
Labels: CPU, Crash, LLM

#437 - Failed running GPT-J int8 benchmark

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

#436 - ipex fails for Adan from package pytorch_optimizer

Issue - State: open - Opened by ldv1 about 1 year ago - 11 comments
Labels: CPU, UserExperience

#434 - InstanceNorm2d only works with parameter "affine = True"

Issue - State: closed - Opened by Fora888 about 1 year ago - 5 comments
Labels: XPU/GPU, Crash

#433 - aten::bmm op is much more slower in float16 for llm rest token generation

Issue - State: open - Opened by rnwang04 about 1 year ago - 5 comments
Labels: ARC, Performance, LLM

#430 - llama_int8 do not support do_sample=True

Issue - State: open - Opened by markluofd about 1 year ago - 1 comment
Labels: CPU, Crash, LLM

#429 - ipex-xpu "OSError: [WinError 126] The specified module could not be found." on windows i7-1185G7

Issue - State: closed - Opened by KiwiHana about 1 year ago - 11 comments
Labels: ARC, UserExperience, Windows

#428 - `torch.linalg.svd` is significantly slower than expected on a Max Series GPU

Issue - State: open - Opened by ogrisel about 1 year ago - 1 comment
Labels: Performance, dGPU-Max

#427 - bloomz-560m has long 1st token latency 110s by iPEX on A380

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

#426 - `torch.scatter_add_` appears to be (about 5x) slower than cuda backend

Issue - State: open - Opened by fcharras about 1 year ago - 1 comment
Labels: Performance, dGPU-Max

#424 - Performance comparison with and without IPEX

Issue - State: closed - Opened by rishucoding about 1 year ago - 5 comments
Labels: Query

#423 - `to('xpu')` causes that Jupyter kernel dies on ARC

Issue - State: open - Opened by rnwang04 about 1 year ago - 10 comments
Labels: XPU/GPU, Crash