Skip to content
Permalink

Comparing changes

Choose two branches to see what’s changed or to start a new pull request. If you need to, you can also or learn more about diff comparisons.

Open a pull request

Create a new pull request by comparing changes across two branches. If you need to, you can also . Learn more about diff comparisons here.
base repository: intel/intel-extension-for-pytorch
Failed to load repositories. Confirm that selected base ref is valid, then try again.
Loading
base: v2.0.0+cpu
Choose a base ref
...
head repository: intel/intel-extension-for-pytorch
Failed to load repositories. Confirm that selected head ref is valid, then try again.
Loading
compare: release/2.0
Choose a head ref

Commits on Mar 23, 2023

  1. remove isa_help.a in wheel build. (#1477) (#1478)

    * remove isa_help.a in wheel build. (#1477)
    xuhancn authored Mar 23, 2023
    Copy the full SHA
    be889e3 View commit details

Commits on Mar 24, 2023

  1. Copy the full SHA
    d34faed View commit details
  2. Copy the full SHA
    f7bc3f5 View commit details

Commits on Mar 25, 2023

  1. disable isa xcr check. (#1512) (#1513)

    * disable isa xcr check.
    xuhancn authored Mar 25, 2023
    Copy the full SHA
    a34eab5 View commit details

Commits on Mar 27, 2023

  1. Copy the full SHA
    1725a5b View commit details

Commits on Mar 30, 2023

  1. Copy the full SHA
    1363e82 View commit details

Commits on Apr 13, 2023

  1. Copy the full SHA
    2ab3693 View commit details
  2. Copy the full SHA
    09a9c18 View commit details
  3. Copy the full SHA
    4b3793d View commit details

Commits on Apr 14, 2023

  1. Copy the full SHA
    bc7c74a View commit details

Commits on Apr 18, 2023

  1. Copy the full SHA
    5159d50 View commit details

Commits on Apr 19, 2023

  1. Copy the full SHA
    76dd768 View commit details

Commits on Apr 21, 2023

  1. [release/2.0] Optimize INT8 LSTM weight scales calculation (#1566)

    * optimize int8 weight scales calculation
    
    * add check on scale sizes
    chunyuan-w authored Apr 21, 2023
    Copy the full SHA
    fed42b1 View commit details
  2. Copy the full SHA
    046f7df View commit details

Commits on Apr 23, 2023

  1. add meta backend for EmbeddingBag (#1525) (#1568)

    * add meta backend for EmbeddingBag
    
    * add UT
    
    * modify UT
    
    * fix UT
    jiayisunx authored Apr 23, 2023
    Copy the full SHA
    881c6fe View commit details

Commits on Apr 24, 2023

  1. docs typo fixes (#1575)

    jingxu10 authored Apr 24, 2023
    Copy the full SHA
    78ec388 View commit details
  2. add gomp envs (#1572)

    jingxu10 authored Apr 24, 2023
    Copy the full SHA
    48b4eea View commit details

Commits on Apr 25, 2023

  1. Copy the full SHA
    adeeeca View commit details
  2. Copy the full SHA
    9dd123a View commit details
  3. Copy the full SHA
    02449cc View commit details

Commits on Apr 26, 2023

  1. Copy the full SHA
    8b02d62 View commit details
  2. Copy the full SHA
    f5ce619 View commit details
  3. Copy the full SHA
    6beb3d4 View commit details

Commits on Apr 27, 2023

  1. [release/2.0] Fix rnn_packed format check (#1592)

    * add UT for lstm weight reorder
    
    * use is_opaque instead of is_rnn_packed
    
    * update ideep commit to include is_opaque API
    chunyuan-w authored Apr 27, 2023
    Copy the full SHA
    023c104 View commit details
  2. fix output strides in conv/deconv meta backend (#1508) (#1590)

    * fix output strides in conv/deconv meta backend
    
    * add UT
    
    * add UT
    
    * fix UT
    
    * fix clang format
    
    * fix backward
    
    * fix UT
    jiayisunx authored Apr 27, 2023
    Copy the full SHA
    d8723df View commit details
  3. Fix the presicion when weights are in bfloat16 for GroupNorm (#1587)

    * fix the presicion when weights are in bfloat16
    
    * use channles first format for full bf16 mode of GroupNorm
    
    * minor fix
    
    * revert use channles first format for full bf16 mode of GroupNorm
    
    * fix format
    
    * use first CL implementation for GroupNorm forward when parameters are in bf16
    
    * remove using first CL implementation
    
    * eliminate redundant code
    
    * minor changes
    
    * add data type checks for GroupNorm backward
    CaoE authored Apr 27, 2023
    Copy the full SHA
    bc76ab1 View commit details
  4. delete plain weight while prepack (#1445) (#1593)

    * delete plain weight while prepack
    
    * only delete plain weight when user setting inplace
    
    * fix ut
    
    Co-authored-by: jianan-gu <jianan.gu@intel.com>
    zhuhaozhe and jianan-gu authored Apr 27, 2023
    Copy the full SHA
    ee7dc34 View commit details
  5. fix stable diffusers textual_inversion finetune accurary issue. (#153…

    …3) (#1594)
    
    the generated imaged by the finetune model unrelated with prompt
    
    Signed-off-by: Wang, Yi A <yi.a.wang@intel.com>
    Co-authored-by: Wang, Yi <yi.a.wang@intel.com>
    jianan-gu and sywangyi authored Apr 27, 2023
    Copy the full SHA
    b2983b4 View commit details
  6. add neural_compressor version check (#1595)

    * add neural_compressor version check
    
    * modify error message
    jiayisunx authored Apr 27, 2023
    Copy the full SHA
    5872620 View commit details

Commits on Apr 28, 2023

  1. Copy the full SHA
    4682ce3 View commit details

Commits on May 9, 2023

  1. Copy the full SHA
    317452c View commit details
  2. Sync with external repo (#1618)

    * Fix license metadata
    
    License doesn't show in PyPI and isn't caught by license scanners. 
    
    See left metadata panel on projects like https://pypi.org/project/apache-flink/ that doesn't show on https://pypi.org/project/intel-extension-for-pytorch/.
    
    Example setup.py in that project is here: https://github.com/apache/flink/blob/742685b76c7f001a08799a539cad2bb683d5d29d/flink-python/setup.py#L327
    
    The setup call in setup.py needs the below. Thanks!
    
            license='https://www.apache.org/licenses/LICENSE-2.0',
            classifiers=[
                'License :: OSI Approved :: Apache Software License',
            ],
    
    * Regex should be using a raw string
    
    * update README for 1.13.120 release
    
    ---------
    
    Co-authored-by: Ben <bstriner@users.noreply.github.com>
    Co-authored-by: STerliakov <terlya.stas@gmail.com>
    3 people authored May 9, 2023
    Copy the full SHA
    6e2ffa8 View commit details
  3. Update launch script for preset env vars and ld_preload fixing (#1603)

    * avoid environment variable configuration in the launch script modifies system values
    
    * fix preload issue with preset LD_PRELOAD values
    jingxu10 authored May 9, 2023
    Copy the full SHA
    6a341a3 View commit details

Commits on May 12, 2023

  1. Copy the full SHA
    25b7212 View commit details
Showing with 3,995 additions and 2,470 deletions.
  1. +0 −33 .github/workflows/publish.yml
  2. +0 −202 LICENSE.txt
  3. +3 −4 README.md
  4. +3 −0 csrc/cpu/CMakeLists.txt
  5. +52 −18 csrc/cpu/aten/Conv.cpp
  6. +10 −5 csrc/cpu/aten/Conv.h
  7. +55 −22 csrc/cpu/aten/ConvTranspose.cpp
  8. +8 −4 csrc/cpu/aten/ConvTranspose.h
  9. +32 −15 csrc/cpu/aten/EmbeddingBag.cpp
  10. +7 −0 csrc/cpu/aten/EmbeddingBag.h
  11. +7 −0 csrc/cpu/aten/GroupNorm.cpp
  12. +9 −14 csrc/cpu/aten/RNN.cpp
  13. +100 −38 csrc/cpu/aten/ROIAlign.cpp
  14. +85 −53 csrc/cpu/aten/ROIAlign.h
  15. +2 −2 csrc/cpu/aten/WeightPack.cpp
  16. +405 −241 csrc/cpu/aten/kernels/GroupNormKrnl.cpp
  17. +1 −2 csrc/cpu/aten/kernels/MergedEmbeddingBagBackwardKrnl.cpp
  18. +104 −115 csrc/cpu/aten/kernels/ROIAlignKrnl.cpp
  19. +20 −12 csrc/cpu/isa/cpu_feature.cpp
  20. +2 −1 csrc/jit/cpu/kernels/ConvPacked.cpp
  21. +0 −2 csrc/jit/cpu/kernels/LinearSwishCustomized.cpp
  22. +0 −2 csrc/jit/cpu/kernels/Matmul.cpp
  23. +1 −1 docker/Dockerfile.compile
  24. +4 −4 docker/Dockerfile.prebuilt
  25. +3 −2 docs/tutorials/blogs_publications.md
  26. +1 −1 docs/tutorials/examples.md
  27. +1 −1 docs/tutorials/features.rst
  28. +5 −5 docs/tutorials/features/codeless_optimization.md
  29. +1 −1 docs/tutorials/features/graph_capture.md
  30. +9 −65 docs/tutorials/features/graph_optimization.md
  31. +19 −19 docs/tutorials/features/hypertune.md
  32. +1 −1 docs/tutorials/features/int8_overview.md
  33. +3 −3 docs/tutorials/features/nhwc.md
  34. +13 −11 docs/tutorials/installation.md
  35. +11 −4 docs/tutorials/performance_tuning/known_issues.md
  36. +93 −57 docs/tutorials/performance_tuning/launch_script.md
  37. +38 −38 docs/tutorials/performance_tuning/torchserve.md
  38. +17 −0 docs/tutorials/releases.md
  39. +13 −0 examples/cpu/features/graph_optimization/folding.py
  40. +19 −0 examples/cpu/features/graph_optimization/fp32_bf16.py
  41. +53 −0 examples/cpu/features/graph_optimization/int8.py
  42. +4 −4 examples/cpu/features/hypertune/example.yaml
  43. +1 −1 examples/cpu/features/hypertune/run_hypertune_example.sh
  44. +3 −1 examples/cpu/features/int8_recipe_tuning/imagenet_autotune.py
  45. +4 −3 examples/cpu/inference/python/int8_calibration_static.py
  46. +5 −4 intel_extension_for_pytorch/__init__.py
  47. +0 −1 intel_extension_for_pytorch/cpu/__init__.py
  48. +29 −9 intel_extension_for_pytorch/cpu/auto_ipex.py
  49. +41 −41 intel_extension_for_pytorch/cpu/hypertune/README.md
  50. +13 −12 intel_extension_for_pytorch/cpu/hypertune/__main__.py
  51. +65 −58 intel_extension_for_pytorch/cpu/hypertune/conf/config.py
  52. +2 −2 intel_extension_for_pytorch/cpu/hypertune/example/example.yaml
  53. +72 −54 intel_extension_for_pytorch/cpu/hypertune/objective.py
  54. +3 −3 intel_extension_for_pytorch/cpu/hypertune/strategy/grid.py
  55. +7 −7 intel_extension_for_pytorch/cpu/hypertune/strategy/random.py
  56. +38 −38 intel_extension_for_pytorch/cpu/hypertune/strategy/strategy.py
  57. +0 −942 intel_extension_for_pytorch/cpu/launch.py
  58. +4 −0 intel_extension_for_pytorch/cpu/launch/__init__.py
  59. +348 −0 intel_extension_for_pytorch/cpu/launch/__main__.py
  60. +328 −0 intel_extension_for_pytorch/cpu/launch/cpu_info.py
  61. +227 −0 intel_extension_for_pytorch/cpu/launch/launcher_base.py
  62. +245 −0 intel_extension_for_pytorch/cpu/launch/launcher_distributed.py
  63. +258 −0 intel_extension_for_pytorch/cpu/launch/launcher_multi_instances.py
  64. +0 −1 intel_extension_for_pytorch/csrc/CMakeLists.txt
  65. +0 −20 intel_extension_for_pytorch/csrc/cpu/isa_help/CMakeLists.txt
  66. +2 −2 intel_extension_for_pytorch/frontend.py
  67. +1 −1 intel_extension_for_pytorch/nn/functional/_tensor_method.py
  68. +3 −3 intel_extension_for_pytorch/nn/modules/merged_embeddingbag.py
  69. +8 −2 intel_extension_for_pytorch/nn/utils/_weight_cast.py
  70. +37 −7 intel_extension_for_pytorch/nn/utils/_weight_prepack.py
  71. +2 −2 intel_extension_for_pytorch/quantization/README.md
  72. +4 −1 intel_extension_for_pytorch/quantization/_autotune.py
  73. +5 −1 intel_extension_for_pytorch/quantization/_quantization_state.py
  74. +19 −21 scripts/compile_bundle.sh
  75. +10 −3 setup.py
  76. +1 −0 tests/cpu/isa/CMakeLists.txt
  77. +10 −10 tests/cpu/test_code_free_optimization.py
  78. +71 −8 tests/cpu/test_cpu_ops.py
  79. +22 −0 tests/cpu/test_emb.py
  80. +2 −3 tests/cpu/test_ipex_optimize.py
  81. +540 −196 tests/cpu/test_launcher.py
  82. +2 −2 tests/cpu/test_merged_embeddingbag.py
  83. +49 −0 tests/cpu/test_roialign.py
  84. +168 −0 tests/cpu/test_torch_compile.py
  85. +47 −7 tests/cpu/test_weight_prepack.py
  86. +83 −0 tests/cpu/utils/cpuinfo.py
  87. +1 −1 third_party/ideep
  88. +1 −1 version.txt
33 changes: 0 additions & 33 deletions .github/workflows/publish.yml

This file was deleted.

202 changes: 0 additions & 202 deletions LICENSE.txt

This file was deleted.

7 changes: 3 additions & 4 deletions README.md
Original file line number Diff line number Diff line change
@@ -32,8 +32,7 @@ More installation methods can be found at [CPU Installation Guide](https://intel
You can install Intel® Extension for PyTorch\* for GPU via command below.

```python
python -m pip install torch==1.13.0a0 -f https://developer.intel.com/ipex-whl-stable-xpu
python -m pip install intel_extension_for_pytorch==1.13.10+xpu -f https://developer.intel.com/ipex-whl-stable-xpu
python -m pip install torch==1.13.0a0+git6c9b55e intel_extension_for_pytorch==1.13.120+xpu -f https://developer.intel.com/ipex-whl-stable-xpu
```

**Note:** The patched PyTorch 1.13.0a0 is required to work with Intel® Extension for PyTorch\* on Intel® graphics card for now.
@@ -84,11 +83,11 @@ with torch.no_grad():

## Model Zoo

Use cases that had already been optimized by Intel engineers are available at [Model Zoo for Intel® Architecture](https://github.com/IntelAI/models/tree/pytorch-r1.13-models). A bunch of PyTorch use cases for benchmarking are also available on the [Github page](https://github.com/IntelAI/models/tree/pytorch-r1.13-models/benchmarks#pytorch-use-cases). You can get performance benefits out-of-box by simply running scipts in the Model Zoo.
Use cases that had already been optimized by Intel engineers are available at [Model Zoo for Intel® Architecture](https://github.com/IntelAI/models/tree/pytorch-r2.0-models). A bunch of PyTorch use cases for benchmarking are also available on the [Github page](https://github.com/IntelAI/models/tree/pytorch-r2.0-models/benchmarks#pytorch-use-cases). You can get performance benefits out-of-box by simply running scipts in the Model Zoo.

## License

_Apache License_, Version _2.0_. As found in [LICENSE](https://github.com/intel/intel-extension-for-pytorch/blob/master/LICENSE.txt) file.
_Apache License_, Version _2.0_. As found in [LICENSE](https://github.com/intel/intel-extension-for-pytorch/blob/master/LICENSE) file.

## Security

3 changes: 3 additions & 0 deletions csrc/cpu/CMakeLists.txt
Original file line number Diff line number Diff line change
@@ -73,6 +73,9 @@ set(IPEX_CPU_CPP_SRCS ${IPEX_CPU_CPP_DYNDISP_SRCS} ${IPEX_CPU_CPP_ISA_SRCS_GEN}

list(REMOVE_ITEM IPEX_CPU_CPP_SRCS ${IPEX_CPU_CPP_ISA_SRCS_ORIGIN})

# Disable XCR check to support Virtual Machines. Some hypervisor can't simulate XCR0 correctly.
# add_definitions (-DENABLE_XCR_CHECK)

add_library(${PLUGIN_NAME_CPU} SHARED ${IPEX_CPU_CPP_SRCS})

set_target_properties(${PLUGIN_NAME_CPU} PROPERTIES ONEDNN_INCLUDE_DIR "${IPEX_CPU_CPP_THIRD_PARTY_ROOT}/ideep/mkl-dnn/include")
Loading