mirror of
https://github.com/saymrwulf/onnxruntime.git
synced 2026-07-29 20:14:01 +00:00
### Description Previously, Flash Attention only worked on Linux systems. This PR will make it work and enable it to be built and run on Windows. Limitations of Flash Attention in Windows: Requires CUDA 12. ### Motivation and Context This will significantly increase the performance of Windows-based LLM's with hardware sm>=80. To illustrate the improvement of Flash Attention over Memory Efficient Attention, here are some average benchmark numbers for the GQA operator, run with configurations based on several recent models (Llama, Mixtral, Phi-3). The benchmarks were obtained on RTX4090 GPU using the test script located at (onnxruntime/test/python/transformers/benchmark_gqa_windows.py). * Clarifying Note: These benchmarks are just for the GQA operator, not the entire model. ### Memory Efficient Attention Kernel Benchmarks: | Model Name | Max Sequence Length | Inference Interval (ms) | Throughput (samples/second) | |----------------------------------------|---------------------|-------------------------|-----------------------------| | Llama3-8B (Average Prompt) | 8192 | 0.19790525 | 13105.63425 | | Llama3-8B (Average Token) | 8192 | 0.207775538 | 12025.10172 | | Llama3-70B (Average Prompt) | 8192 | 0.216049167 | 11563.31185 | | Llama3-70B (Average Token) | 8192 | 0.209730731 | 12284.38149 | | Mixtral-8x22B-v0.1 (Average Prompt) | 32768 | 0.371928785 | 7031.440056 | | Mixtral-8x22B-v0.1 (Average Token) | 32768 | 0.2996659 | 7607.947159 | | Phi-3-mini-128k (Average Prompt) | 131072 | 0.183195867 | 15542.0852 | | Phi-3-mini-128k (Average Token) | 131072 | 0.198215688 | 12874.53494 | | Phi-3-small-128k (Average Prompt) | 65536 | 2.9884929 | 2332.584142 | | Phi-3-small-128k (Average Token) | 65536 | 0.845072406 | 2877.85822 | | Phi-3-medium-128K (Average Prompt) | 32768 | 0.324974429 | 8094.909517 | | Phi-3-medium-128K (Average Token) | 32768 | 0.263662567 | 8978.463687 | ### Flash Attention Kernel Benchmarks: | Model Name | Max Sequence Length | Inference Interval (ms) | Throughput (samples/second) | |--------------------------------------|---------------------|-------------------------|-----------------------------| | Llama3-8B (Average Prompt) | 8192 | 0.163566292 | 16213.69057 | | Llama3-8B (Average Token) | 8192 | 0.161643692 | 16196.14715 | | Llama3-70B (Average Prompt) | 8192 | 0.160510375 | 17448.67753 | | Llama3-70B (Average Token) | 8192 | 0.169427308 | 14702.62043 | | Mixtral-8x22B-v0.1 (Average Prompt) | 32768 | 0.164121964 | 15618.51301 | | Mixtral-8x22B-v0.1 (Average Token) | 32768 | 0.1715865 | 14524.32273 | | Phi-3-mini-128k (Average Prompt) | 131072 | 0.167527167 | 14576.725 | | Phi-3-mini-128k (Average Token) | 131072 | 0.175940594 | 15762.051 | | Phi-3-small-128k (Average Prompt) | 65536 | 0.162719733 | 17824.494 | | Phi-3-small-128k (Average Token) | 65536 | 0.14977525 | 16749.19858 | | Phi-3-medium-128K (Average Prompt) | 32768 | 0.156490786 | 17679.2513 | | Phi-3-medium-128K (Average Token) | 32768 | 0.165333833 | 14932.26079 | Flash Attention is consistently faster for every configuration we benchmarked, with improvements in our trials ranging from ~20% to ~650%. In addition to these improvements in performance, Flash Attention has better memory usage. For example, Memory Efficient Attention cannot handle a max sequence length higher than 32,768, but Flash Attention can handle max sequence lengths at least as high as 131,072. --------- Co-authored-by: Tianlei Wu <tlwu@microsoft.com>
161 lines
3.6 KiB
TOML
161 lines
3.6 KiB
TOML
# Configuration for lintrunner https://github.com/suo/lintrunner
|
|
# You can install the dependencies and initialize with
|
|
#
|
|
# ```sh
|
|
# pip install lintrunner lintrunner-adapters
|
|
# lintrunner init
|
|
# ```
|
|
#
|
|
# This will install lintrunner on your system and download all the necessary
|
|
# dependencies to run linters locally.
|
|
# If you want to see what lintrunner init will install, run
|
|
# `lintrunner init --dry-run`.
|
|
#
|
|
# To lint local changes:
|
|
#
|
|
# ```bash
|
|
# lintrunner
|
|
# ```
|
|
#
|
|
# To lint all files:
|
|
#
|
|
# ```bash
|
|
# lintrunner --all-files
|
|
# ```
|
|
#
|
|
# To format files:
|
|
#
|
|
# ```bash
|
|
# lintrunner f --all-files
|
|
# ```
|
|
#
|
|
# To read more about lintrunner, see [wiki](https://github.com/pytorch/pytorch/wiki/lintrunner).
|
|
# To update an existing linting rule or create a new one, modify this file or create a
|
|
# new adapter following examples in https://github.com/justinchuby/lintrunner-adapters.
|
|
|
|
merge_base_with = 'origin/main'
|
|
|
|
[[linter]]
|
|
code = 'RUFF'
|
|
include_patterns = [
|
|
'**/*.py',
|
|
'**/*.pyi',
|
|
]
|
|
exclude_patterns = [
|
|
'cmake/external/**',
|
|
# ignore generated flatbuffers code
|
|
'onnxruntime/core/flatbuffers/ort_flatbuffers_py/**',
|
|
'orttraining/orttraining/python/training/optim/_ds_code_store.py',
|
|
]
|
|
command = [
|
|
'python',
|
|
'-m',
|
|
'lintrunner_adapters',
|
|
'run',
|
|
'ruff_linter',
|
|
'--config=pyproject.toml',
|
|
'@{{PATHSFILE}}'
|
|
]
|
|
init_command = [
|
|
'python',
|
|
'-m',
|
|
'lintrunner_adapters',
|
|
'run',
|
|
'pip_init',
|
|
'--dry-run={{DRYRUN}}',
|
|
'--requirement=requirements-lintrunner.txt',
|
|
]
|
|
is_formatter = true
|
|
|
|
|
|
[[linter]]
|
|
code = 'BLACK-ISORT'
|
|
include_patterns = [
|
|
'**/*.py',
|
|
]
|
|
exclude_patterns = [
|
|
'cmake/**',
|
|
'orttraining/*',
|
|
'onnxruntime/core/flatbuffers/**',
|
|
'orttraining/orttraining/python/training/optim/_ds_code_store.py',
|
|
]
|
|
command = [
|
|
'python',
|
|
'-m',
|
|
'lintrunner_adapters',
|
|
'run',
|
|
'black_isort_linter',
|
|
'--',
|
|
'@{{PATHSFILE}}'
|
|
]
|
|
init_command = [
|
|
'python',
|
|
'-m',
|
|
'lintrunner_adapters',
|
|
'run',
|
|
'pip_init',
|
|
'--dry-run={{DRYRUN}}',
|
|
'--requirement=requirements-lintrunner.txt',
|
|
]
|
|
is_formatter = true
|
|
|
|
[[linter]]
|
|
code = 'RUSTFMT'
|
|
include_patterns = ['**/*.rs']
|
|
command = [
|
|
'python',
|
|
'-m',
|
|
'lintrunner_adapters',
|
|
'run',
|
|
'rustfmt_linter',
|
|
'--binary=rustfmt',
|
|
'--config-path=rust/rustfmt.toml',
|
|
'--',
|
|
'@{{PATHSFILE}}'
|
|
]
|
|
is_formatter = true
|
|
|
|
[[linter]]
|
|
code = 'CLANGFORMAT'
|
|
include_patterns = [
|
|
'**/*.h',
|
|
'**/*.cc',
|
|
'**/*.hpp',
|
|
'**/*.cpp',
|
|
'**/*.m',
|
|
'**/*.mm',
|
|
]
|
|
exclude_patterns = [
|
|
'java/**', # FIXME: Enable clang-format for java
|
|
'js/**',
|
|
'onnxruntime/contrib_ops/cuda/bert/tensorrt_fused_multihead_attention/**', # Contains data chunks
|
|
'onnxruntime/core/flatbuffers/schema/*.fbs.h', # Generated code
|
|
'onnxruntime/test/flatbuffers/*.fbs.h', # Generated code
|
|
'onnxruntime/core/graph/contrib_ops/quantization_defs.cc',
|
|
'onnxruntime/core/mlas/**', # Contains assembly code
|
|
'onnxruntime/core/mickey/cutlass_ext/**', # CUTLASS based libs recommends NO automatic code formatting
|
|
'onnxruntime/core/mickey/gemm/**', # CUTLASS based libs recommends NO automatic code formatting
|
|
'winml/lib/Api.Image/shaders/**', # Contains data chunks
|
|
'onnxruntime/contrib_ops/cuda/bert/flash_attention/flash_fwd_launch_template.h', # Bool Switches hang Clang
|
|
]
|
|
command = [
|
|
'python',
|
|
'-m',
|
|
'lintrunner_adapters',
|
|
'run',
|
|
'clangformat_linter',
|
|
'--binary=clang-format',
|
|
'--fallback',
|
|
'--',
|
|
'@{{PATHSFILE}}'
|
|
]
|
|
init_command = [
|
|
'python',
|
|
'-m',
|
|
'lintrunner_adapters',
|
|
'run',
|
|
'pip_init',
|
|
'--dry-run={{DRYRUN}}',
|
|
'--requirement=requirements-lintrunner.txt',
|
|
]
|
|
is_formatter = true
|