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

#624 - When is ipex available for pytorch 2.3.0 ?

Issue - State: closed - Opened by sriharshaj70 6 months ago - 5 comments
Labels: Query

#623 - NotImplementedError: The following operation failed in the TorchScript interpreter. Traceback of TorchScript (most recent call last): RuntimeError: Could not run 'torch_ipex::mul_add' with arguments from the 'CPU' backend.

Issue - State: closed - Opened by wangdq1989 6 months ago - 3 comments
Labels: ARC, iGPU, Crash, Windows

#622 - No Devices Visible with oneAPI 2024.1 and Intel GPU Max 1550 GPUs

Issue - State: closed - Opened by BenBrock 6 months ago - 5 comments
Labels: dGPU-Max

#621 - Undefined Symbols with oneAPI 2024.0.0

Issue - State: closed - Opened by BenBrock 6 months ago - 4 comments
Labels: XPU/GPU, Crash

#620 - Speech Translation using Coqui.ai on Intel Arc GPU 770 takes 23 seconds compared to CPU [3 sec] why ?

Issue - State: closed - Opened by shailesh837 6 months ago - 10 comments
Labels: ARC

#619 - NotImplementedError: Could not run 'aten::empty_strided'

Issue - State: closed - Opened by Delaunay 6 months ago - 3 comments
Labels: Bug, XPU/GPU

#618 - '_IPEXLinear' object has no attribute 'use_dnnl'

Issue - State: closed - Opened by Alok-Ranjan23 6 months ago - 6 comments
Labels: Bug, CPU, Crash, LLM, high priority

#617 - added import os for exit() to fix NameError

Pull Request - State: open - Opened by Harryalways317 6 months ago

#616 - Hi @qslia , thanks for reporting this issue. Yes we support Arc graphics.

Issue - State: closed - Opened by qslia 7 months ago

#615 - CPU Memory leak when training any model

Issue - State: closed - Opened by pujaltes 7 months ago - 5 comments

#614 - model.to('xpu') does not terminate

Issue - State: closed - Opened by BiaBibii 7 months ago - 3 comments
Labels: iGPU, Performance

#613 - IFFTRN Arguments RuntimeError: FFT_INVALID_DESCRIPTOR

Issue - State: closed - Opened by KPoletti 7 months ago - 3 comments
Labels: Bug, Crash, dGPU-Max

#612 - How to use ipex.llm.optimize

Issue - State: closed - Opened by Alok-Ranjan23 7 months ago - 1 comment
Labels: Query, LLM

#611 - Error can't import the library: Key already registered with the same priority

Issue - State: closed - Opened by CarlGao4 7 months ago - 2 comments
Labels: Compilation, Windows

#610 - [inductor] Make inductor work with new triton kernel launch API

Pull Request - State: closed - Opened by whitneywhtsang 7 months ago

#609 - Failed to run interger matrix multiply

Issue - State: closed - Opened by primenumber 7 months ago - 5 comments
Labels: ARC, UserExperience, Escalate

#608 - Fix a typo

Pull Request - State: open - Opened by Hsy-Intel 7 months ago

#607 - IPEX takes 10min+ for "warmup" on MTL iGPU

Issue - State: closed - Opened by cold-blue 7 months ago - 1 comment
Labels: iGPU

#606 - operator does not exist for XPU backend

Issue - State: closed - Opened by jinz2014 7 months ago
Labels: XPU/GPU, NotAnIssue

#605 - compatability issues with oobabooga

Issue - State: closed - Opened by alioop222 7 months ago - 8 comments
Labels: NotAnIssue

#604 - failed to run on ultra 5 with Arc(TM) graphics (16GB VRAM)

Issue - State: closed - Opened by qslia 7 months ago - 6 comments
Labels: ARC, Crash, Windows

#603 - Unable to run llama_cpp example from quickstart guide (PI_ERROR_BUILD_PROGRAM_FAILURE)

Issue - State: closed - Opened by tristan-k 7 months ago
Labels: ARC

#602 - How do you save a model after using WOQ?

Issue - State: closed - Opened by eduand-alvarez 7 months ago - 2 comments
Labels: Feature

#601 - Update triton.txt

Pull Request - State: closed - Opened by Stonepia 7 months ago

#600 - Revert "[Triton] change target[1] to sycl_device to get arch properties from Triton side"

Pull Request - State: closed - Opened by Stonepia 7 months ago

#599 - Communication and compute on separate Streams do not overlap

Issue - State: open - Opened by garrett361 7 months ago - 11 comments
Labels: dGPU-Max, Functionality

#598 - Run WOQ INT8 and WOQ INT4 in distributed way don't show token latency information even added --token-latency

Issue - State: closed - Opened by redhairerINTEL 7 months ago - 1 comment
Labels: CPU, LLM

#597 - Incorrect inference output when not computing gradients on XPU

Issue - State: closed - Opened by martinmCGG 7 months ago - 8 comments
Labels: ARC

#596 - Add device architecture to DeviceInfo

Pull Request - State: closed - Opened by quintinwang5 7 months ago - 1 comment

#595 - initial import time of `torch` & `intel_extension_for_pytorch` is long on Windows Core Ultra

Issue - State: closed - Opened by Oscilloscope98 7 months ago - 4 comments
Labels: ARC, Windows

#594 - Revert "[Triton] change target[1] to sycl_device to get arch properties from Triton side"

Pull Request - State: closed - Opened by quintinwang5 7 months ago

#591 - test

Issue - State: closed - Opened by jingxu10 7 months ago
Labels: NotAnIssue

#590 - [Triton] change target[1] to sycl_device to get arch properties from Triton side

Pull Request - State: closed - Opened by quintinwang5 7 months ago - 3 comments

#589 - Bump transformers from 4.36.0 to 4.38.0 in /examples/cpu/serving/triton

Pull Request - State: closed - Opened by dependabot[bot] 7 months ago - 1 comment
Labels: dependencies

#588 - Add options to build with system LIBXSMM/oneDNN/sleef/MKL/ideep/gtest

Pull Request - State: open - Opened by AngryLoki 7 months ago - 1 comment

#587 - Fix compilation with GCC and Clang

Pull Request - State: closed - Opened by AngryLoki 7 months ago - 12 comments

#586 - Bump triton version for new API change

Pull Request - State: open - Opened by Stonepia 7 months ago

#585 - OSError: [WinError 1920] The file cannot be accessed by the system. Error loading "C:\Users\techn\AppData\Local\Programs\Python\Python311\Lib\site-packages\torch\lib\backend_with_compiler.dll" or one of its dependencies.

Issue - State: closed - Opened by Bhargav230m 7 months ago - 43 comments
Labels: iGPU, Windows

#584 - [inductor] make inductor work with new triton kernel launch API

Pull Request - State: closed - Opened by Stonepia 7 months ago - 2 comments

#583 - [python][cpu][llm] run with THUDM/chatglm3-6b raise "RuntimeError: Unsupported input type"

Issue - State: closed - Opened by mozillazg 7 months ago - 4 comments
Labels: Bug, CPU

#582 - Add E2E CI

Pull Request - State: closed - Opened by Stonepia 8 months ago

#581 - IPEX.LLM performance on Intel SPR with large batch size.

Issue - State: closed - Opened by mengke-mk 8 months ago - 6 comments
Labels: CPU, Performance, LLM, Escalate

#580 - torch.random.fork_rng_state does not restore rng state

Issue - State: closed - Opened by garrett361 8 months ago - 3 comments
Labels: XPU/GPU

#579 - Distortions in using IPEX with triplane NeRF on TripoSR

Issue - State: closed - Opened by SAint7579 8 months ago - 7 comments
Labels: dGPU-Max

#578 - IPEX v2.1.20+xpu LLVM error

Issue - State: closed - Opened by simonlui 8 months ago - 4 comments
Labels: XPU/GPU

#577 - IPEX v2.1.20+xpu regression with SDXL CLIP Text Encoder

Issue - State: closed - Opened by simonlui 8 months ago - 4 comments
Labels: XPU/GPU

#576 - How to use ipex.optimize for two models (model distillation) and one optimizer

Issue - State: closed - Opened by ferreirafabio 8 months ago - 5 comments
Labels: UserExperience, Query

#575 - Add E2E CI

Pull Request - State: closed - Opened by RUIJIEZHONG66166 8 months ago

#574 - Fused CPU Adam performance

Issue - State: closed - Opened by msaroufim 8 months ago - 18 comments
Labels: CPU, Performance, Escalate

#573 - compile_bundle.sh is hard to use

Issue - State: closed - Opened by msaroufim 8 months ago - 7 comments
Labels: CPU, UserExperience, Compilation

#572 - ImportError: envs/pytorch-arc/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so: undefined symbol: iJIT_NotifyEventrge/envs/pytorch-arc/lib/python3.11/site-packages/torch/lib/libtorch_cpu.so: undefined symbol: iJIT_NotifyEvent

Issue - State: closed - Opened by SoldierWz 8 months ago - 10 comments
Labels: NotAnIssue

#571 - RuntimeError: 0 <= device.index() && device.index() < static_cast<c10::DeviceIndex>(device_ready_queues_.size()) INTERNAL ASSERT FAILED at "/build/pytorch/torch/csrc/autograd/engine.cpp":1418

Issue - State: open - Opened by SoldierWz 8 months ago - 3 comments
Labels: ARC

#570 - spell error

Issue - State: closed - Opened by xxxhome 8 months ago - 1 comment
Labels: Documentation

#569 - spell error

Issue - State: closed - Opened by xxxhome 8 months ago - 1 comment
Labels: Documentation

#568 - xpu.Event reporting erroneous times

Issue - State: closed - Opened by garrett361 8 months ago - 8 comments
Labels: Bug, XPU/GPU, UserExperience, Escalate

#567 - [Bug]: `NameError: name 'exit' is not defined`

Issue - State: closed - Opened by saforem2 8 months ago - 6 comments
Labels: Bug, CPU

#566 - torch.linalg.eigvals with XPU

Issue - State: closed - Opened by martinwang2002 8 months ago - 5 comments
Labels: Bug, XPU/GPU

#565 - Using XPU training actually increases the training time and sharply reduces the accuracy.

Issue - State: closed - Opened by SoldierWz 8 months ago - 10 comments
Labels: Bug, XPU/GPU, Performance

#564 - Please update Get Started examples.

Issue - State: closed - Opened by QiXuanWang 8 months ago - 1 comment
Labels: NotAnIssue

#563 - Huggingface assist model (Speculative Decoding) error with IPEX 2.2 CPU

Issue - State: closed - Opened by qiyuangong 8 months ago
Labels: CPU

#562 - Is there any plan to merge to official PyTorch?

Issue - State: closed - Opened by alanzhai219 8 months ago
Labels: Feature

#561 - Build version v2.1.10+xpu from source code, met python ipex import error with mkl symbol libmkl_sycl_lapack.so.4

Issue - State: closed - Opened by bopeng1234 8 months ago - 16 comments
Labels: NotAnIssue, Compilation

#560 - ModuleNotFoundError: No module named 'intel_extension_for_pytorch'

Issue - State: closed - Opened by prashanthrgs 8 months ago - 14 comments
Labels: UserExperience, NotAnIssue

#559 - support for new intel core ultra processor

Issue - State: closed - Opened by HHCorp 8 months ago - 3 comments
Labels: iGPU, Escalate

#558 - how to set command to require best performence?

Issue - State: closed - Opened by Storm0921 8 months ago - 5 comments
Labels: Query, NotAnIssue

#557 - Using torchvision with intel_extension_for_pytorch gives error

Issue - State: closed - Opened by tom-hehir 8 months ago - 2 comments
Labels: XPU/GPU

#556 - token_latency parameter not support in the xpu-main branch run_generation.py

Issue - State: closed - Opened by oldmikeyang 8 months ago - 5 comments
Labels: XPU/GPU, LLM

#555 - WOQ ValueError: too many values to unpack (expected 2) with "Intel/neural-chat-7b-v3-3"

Issue - State: closed - Opened by eduand-alvarez 8 months ago - 4 comments
Labels: Documentation, CPU

#554 - Update examples/cpu/inference/python/llm/README.md

Pull Request - State: open - Opened by linjoe3 9 months ago

#553 - Fix Documentation issue #541

Pull Request - State: closed - Opened by intel-ravig 9 months ago - 1 comment

#552 - RuntimeError: tensor does not have a device when training PixArt-alpha lora on Arc A770

Issue - State: closed - Opened by congdm 9 months ago - 8 comments
Labels: Bug, ARC, Windows

#551 - ipex 2.2.0 "import intel_extension_for_pytorch" gives "Illegal instruction (core dumped)" on Ubuntu 22.04 and J4105 CPU

Issue - State: closed - Opened by sanderjo 9 months ago - 5 comments
Labels: CPU, NotAnIssue

#550 - Neither Codeformers nor GFPGAN works in Stable Diffusion

Issue - State: closed - Opened by Pantonia4 9 months ago - 5 comments
Labels: ARC, Windows

#549 - Add an API to get current global free memory of XPU

Pull Request - State: closed - Opened by notsyncing 9 months ago - 1 comment

#548 - Fix `AttributeError: module 'triton.language.math' has no attribute`

Pull Request - State: closed - Opened by whitneywhtsang 9 months ago - 2 comments

#547 - IPEX with huggingface zero-shot-classification pipeline?

Issue - State: closed - Opened by pmusser 9 months ago - 6 comments
Labels: ARC

#546 - IPEX.LLM performance benchmark and 3rd Gen Intel Xeon

Issue - State: closed - Opened by mgajzler 9 months ago - 6 comments
Labels: CPU, Performance, LLM

#545 - Does BFloat16 model training supported on ARC?

Issue - State: closed - Opened by XinyuYe-Intel 9 months ago - 13 comments
Labels: ARC, Query

#544 - VIDEO_SCHEDULER_INTERNAL_ERROR

Issue - State: closed - Opened by pmusser 9 months ago - 8 comments
Labels: ARC, LLM, Functionality

#543 - Feature request: upstream changes to pytorch

Issue - State: open - Opened by Pandapip1 9 months ago - 2 comments
Labels: Feature

#542 - How to build this from source

Issue - State: closed - Opened by wswsmao 9 months ago - 8 comments
Labels: CPU, Query, Compilation

#541 - Typos in Documentation (github io)

Issue - State: closed - Opened by eduand-alvarez 9 months ago - 5 comments
Labels: Documentation

#540 - intel-extension-for-pytorch 2.1.0.dev+cpu.llm experimental rehabilitation

Issue - State: closed - Opened by duyanyao 9 months ago - 2 comments
Labels: CPU, Query, LLM

#539 - `torch.xpu.amp.autocast` and NaN values in tensors

Issue - State: closed - Opened by arekkulczycki 9 months ago - 1 comment
Labels: NotAnIssue

#538 - Can't detect GPU devices

Issue - State: closed - Opened by mudler 9 months ago - 9 comments
Labels: ARC, Functionality

#537 - Training loss does not improve when running the cifar10 sample

Issue - State: closed - Opened by contryboy 9 months ago - 9 comments
Labels: ARC, Correctness

#536 - Some concerns about device selection about `current_device`

Issue - State: closed - Opened by leonardozcm 9 months ago - 3 comments
Labels: ARC, iGPU, Functionality

#535 - Support for Mixtral

Issue - State: closed - Opened by akarX23 9 months ago - 4 comments
Labels: CPU, Feature

#534 - Unsupported input type, fallback to the origin model

Issue - State: closed - Opened by akarX23 9 months ago - 10 comments
Labels: CPU, LLM

#533 - Fix AVX515 FP16 build issues

Pull Request - State: closed - Opened by gtkramer 9 months ago

#532 - when i run intel_extension_for_pytorch.quantization import prepare prepare(user_model.eval(), qconfig, example_inputs=example_inputs) killed

Issue - State: closed - Opened by duyanyao 9 months ago - 3 comments
Labels: CPU, LLM

#525 - ERROR: Could not find a version that satisfies the requirement intel-extension-for-pytorch==2.1.100 (from versions: none)

Issue - State: open - Opened by mohibovais79 10 months ago - 9 comments
Labels: CPU, UserExperience

#524 - [IPEX Sanity Test] ModuleNotFoundError: No module named 'sympy.utilities'

Issue - State: closed - Opened by rskasturi 10 months ago - 2 comments
Labels: UserExperience, NotAnIssue

#523 - Issue on ipex.optimize: `NotImplementedError: argument of type`

Issue - State: closed - Opened by gera-aldama 10 months ago - 4 comments
Labels: CPU, Query

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