github/tinygrad
1
1
Fork 0
mirror of https://github.com/tinygrad/tinygrad.git synced 2026-01-09 23:18:04 -05:00
Code Issues Packages Projects Releases Wiki Activity

refactor into subactions (#8946)

Browse Source 
* refactor into subactions

* this work?

* add shell

* move install opencl

* valid?

* support mac os x

* refactor other osx

* fix linux/osx

* fixes

* cleanups

* used everywhere

* no quotes

* quotes on true

* bugfixes

* this run?

* hardcode

* that

* process replay action

* fix checkout

* restore to branch

* fix caching

* fix osx python cache

* does replace function exist

* Revert "does replace function exist"

This reverts commit 622177c5a0.

* Revert "fix osx python cache"

This reverts commit e70d55cd93.

* user on osx to fix untar issue

* that
This commit is contained in:
George Hotz
2025-02-07 18:06:44 +08:00
committed by GitHub
parent 133cacadde
commit 9ed2d0dfa2
3 changed files with 285 additions and 304 deletions
Download Patch File Download Diff File Expand all files Collapse all files

13
.github/actions/process-replay/action.yml vendored Normal file
View File

@@ -0,0 +1,13 @@
name: Run process replay tests
description: Verify process replay compared to master
runs:
using: "composite"
steps:
- name: Run process replay tests
shell: bash
run: |
export PR_TITLE=$(jq -r .pull_request.title "$GITHUB_EVENT_PATH")
export COMMIT_MESSAGE=$(git show -s --format=%B ${{ github.event.pull_request.head.sha }})
export CURRENT_HEAD=$(git rev-parse HEAD)
cp test/external/process_replay/process_replay.py ./process_replay.py && git fetch origin master && git -c advice.detachedHead=false checkout origin/master && PYTHONPATH=. python3 process_replay.py
git checkout $CURRENT_HEAD # restore to branch

212
.github/actions/setup-tinygrad/action.yml vendored Normal file
View File

@@ -0,0 +1,212 @@
name: Setup Python & Install
description: Sets up Python and installs project dependencies.
inputs:
python-version:
description: 'Python version to use'
required: false
default: '3.12'
key:
description: 'Key for the python cache'
required: false
default: '' # if you don't set a key, it doesn't cache
deps:
description: 'Extra dependency groups (comma separated)'
required: false
default: ''
opencl:
description: "Install OpenCL?"
required: false
default: 'false'
amd:
description: "Install AMD?"
required: false
default: 'false'
cuda:
description: "Install CUDA?"
required: false
default: 'false'
webgpu:
description: "Install webgpu?"
required: false
default: 'false'
llvm:
description: "Install LLVM?"
required: false
default: 'false'
runs:
using: "composite"
steps:
- name: Set up Python ${{ inputs.python-version }}
uses: actions/setup-python@v5
with:
python-version: ${{ inputs.python-version }}
# **** Caching packages ****
# TODO: key should include input.deps, but it can't since it can't contain commas
- name: Cache Python packages (Linux)
if: inputs.key != '' && runner.os == 'Linux'
uses: actions/cache@v4
with:
path: ${{ env.Python3_ROOT_DIR }}/lib/python${{ inputs.python-version }}/site-packages
key: python-package-${{ inputs.key }}-${{ hashFiles('**/setup.py') }}
- name: Cache Python packages (macOS)
if: inputs.key != '' && runner.os == 'macOS'
uses: actions/cache@v4
with:
path: /Users/runner/Library/Python/${{ inputs.python-version }}/lib/python/site-packages
key: osx-python-package-${{ inputs.key }}-${{ hashFiles('**/setup.py') }}
- name: Cache Python packages (Windows)
if: inputs.key != '' && runner.os == 'Windows'
uses: actions/cache@v4
with:
path: ${{ env.Python3_ROOT_DIR }}\Lib\site-packages
key: windows-python-package-${{ inputs.key }}-${{ hashFiles('**/setup.py') }}
# **** Caching downloads ****
- name: Cache downloads (Linux)
if: inputs.key != '' && runner.os == 'Linux'
uses: actions/cache@v4
with:
path: ~/.cache/tinygrad/downloads/
key: downloads-cache-${{ inputs.key }}-${{ env.DOWNLOAD_CACHE_VERSION }}
- name: Cache downloads (macOS)
if: inputs.key != '' && runner.os == 'macOS'
uses: actions/cache@v4
with:
path: ~/Library/Caches/tinygrad/downloads/
key: osx-downloads-cache-${{ inputs.key }}-${{ env.DOWNLOAD_CACHE_VERSION }}
# **** Python deps ****
- name: Install dependencies (with extra)
if: inputs.deps != ''
shell: bash
run: pip install ${{ (runner.os == 'macOS' && '--user') || (runner.os != 'macOS' && '') }} -e ".[${{ inputs.deps }}]" --extra-index-url https://download.pytorch.org/whl/cpu --extra-index-url https://aiinfra.pkgs.visualstudio.com/PublicPackages/_packaging/Triton-Nightly/pypi/simple/
- name: Install dependencies (without extra)
if: inputs.deps == ''
shell: bash
run: pip install ${{ (runner.os == 'macOS' && '--user') || (runner.os != 'macOS' && '') }} -e .
# **** OpenCL ****
- name: Install OpenCL
if: inputs.opencl == 'true'
shell: bash
run: |
echo 'Acquire::http::Pipeline-Depth "5";' | sudo tee -a /etc/apt/apt.conf.d/99parallel
echo "deb [ allow-insecure=yes ] https://apt.repos.intel.com/oneapi all main" | sudo tee /etc/apt/sources.list.d/oneAPI.list
sudo apt update || true
sudo apt install --allow-unauthenticated -y --no-install-recommends opencl-headers \
intel-oneapi-runtime-openmp=2023.2.1-16 intel-oneapi-runtime-compilers-common=2023.2.1-16 intel-oneapi-runtime-compilers=2023.2.1-16 \
intel-oneapi-runtime-dpcpp-sycl-opencl-cpu=2023.2.1-16 intel-oneapi-runtime-tbb-common=2021.10.0-49541 \
intel-oneapi-runtime-tbb=2021.10.0-49541 intel-oneapi-runtime-opencl=2023.2.1-16
# **** AMD ****
- name: Install AMD (Linux)
if: inputs.amd == 'true' && runner.os == 'Linux'
shell: bash
run: |
echo 'Acquire::http::Pipeline-Depth "5";' | sudo tee -a /etc/apt/apt.conf.d/99parallel
wget https://repo.radeon.com/rocm/rocm.gpg.key -O - | gpg --dearmor | sudo tee /etc/apt/keyrings/rocm.gpg > /dev/null
sudo tee /etc/apt/sources.list.d/rocm.list <<'EOF'
deb [arch=amd64 signed-by=/etc/apt/keyrings/rocm.gpg] https://repo.radeon.com/rocm/apt/6.1.2 jammy main
EOF
echo -e 'Package: *\nPin: release o=repo.radeon.com\nPin-Priority: 600' | sudo tee /etc/apt/preferences.d/rocm-pin-600
sudo apt update || true
sudo apt install --no-install-recommends --allow-unauthenticated -y hsa-rocr comgr hsa-rocr-dev liburing-dev libc6-dev
curl -s https://api.github.com/repos/Qazalin/remu/releases/latest | \
jq -r '.assets[] | select(.name == "libremu.so").browser_download_url' | \
sudo xargs curl -L -o /usr/local/lib/libremu.so
sudo tee --append /etc/ld.so.conf.d/rocm.conf <<'EOF'
/opt/rocm/lib
/opt/rocm/lib64
EOF
sudo ldconfig
- name: Install AMD comgr+remu (macOS)
if: inputs.amd == 'true' && runner.os == 'macOS'
shell: bash
run: |
sudo mkdir -p /usr/local/lib
curl -s -H "Authorization: token $GH_TOKEN" curl -s https://api.github.com/repos/nimlgen/amdcomgr_dylib/releases/latest | \
jq -r '.assets[] | select(.name == "libamd_comgr.dylib").browser_download_url' | \
sudo xargs curl -L -o /usr/local/lib/libamd_comgr.dylib
curl -s -H "Authorization: token $GH_TOKEN" curl -s https://api.github.com/repos/Qazalin/remu/releases/latest | \
jq -r '.assets[] | select(.name == "libremu.dylib").browser_download_url' | \
sudo xargs curl -L -o /usr/local/lib/libremu.dylib
# **** CUDA ****
- name: Install packages (cuda)
if: inputs.cuda == 'true'
shell: bash
run: |
echo 'Acquire::http::Pipeline-Depth "5";' | sudo tee -a /etc/apt/apt.conf.d/99parallel
sudo apt update -y || true
sudo apt install -y --no-install-recommends git g++ cmake ninja-build llvm-15-dev zlib1g-dev libglew-dev \
flex bison libfl-dev libboost-thread-dev libboost-filesystem-dev nvidia-cuda-toolkit-gcc libzstd-dev
- name: Cache gpuocelot
if: inputs.cuda == 'true'
id: cache-build
uses: actions/cache@v4
env:
cache-name: cache-gpuocelot-build
with:
path: ${{ github.workspace }}/gpuocelot/ocelot
key: ubuntu22.04-gpuocelot-4524e34adb7eaccc6f71262f2e21d7052bb17c2f-rebuild-9
- name: Clone/compile gpuocelot
if: inputs.cuda == 'true' && steps.cache-build.outputs.cache-hit != 'true'
shell: bash
run: |
git clone --recurse-submodules https://github.com/gpuocelot/gpuocelot.git ${{ github.workspace }}/gpuocelot
cd ${{ github.workspace }}/gpuocelot/ocelot
git checkout 4524e34adb7eaccc6f71262f2e21d7052bb17c2f
mkdir build
cd build
cmake .. -Wno-dev -G Ninja -DOCELOT_BUILD_TOOLS=OFF -DCMAKE_BUILD_ALWAYS=0 -DBUILD_TESTS_CUDA=OFF
ninja
- name: Install gpuocelot
if: inputs.cuda == 'true'
shell: bash
run: |
cd ${{ github.workspace }}/gpuocelot/ocelot/build
sudo cp libgpuocelot.so /usr/lib/libgpuocelot.so
# **** WebGPU ****
- name: Install WebGPU dawn (Linux)
if: inputs.webgpu == 'true' && runner.os == 'Linux'
shell: bash
run: |
sudo curl -L https://github.com/wpmed92/pydawn/releases/download/v0.1.6/libwebgpu_dawn.so -o /usr/local/lib/libwebgpu_dawn.so
- name: Install dependencies for software-based vulkan
if: inputs.webgpu == 'true' && runner.os == 'Linux'
shell: bash
run: |
sudo apt update -y || true
sudo apt install -y libegl1-mesa libgl1-mesa-dri libxcb-xfixes0-dev mesa-vulkan-drivers
- name: Install WebGPU dawn (macOS)
if: inputs.webgpu == 'true' && runner.os == 'macOS'
shell: bash
run: |
sudo mkdir -p /usr/local/lib
sudo curl -L https://github.com/wpmed92/pydawn/releases/download/v0.1.6/libwebgpu_dawn.dylib -o /usr/local/lib/libwebgpu_dawn.dylib
# **** LLVM ****
- name: Install LLVM (Linux)
if: inputs.llvm == 'true' && runner.os == 'Linux'
shell: bash
run: |
echo 'Acquire::http::Pipeline-Depth "5";' | sudo tee -a /etc/apt/apt.conf.d/99parallel
sudo apt update -y || true
sudo apt install -y --no-install-recommends llvm-14-dev
- name: Install LLVM (macOS)
if: inputs.llvm == 'true' && runner.os == 'macOS'
shell: bash
run: |
brew install llvm

364
.github/workflows/test.yml vendored
View File

@@ -20,12 +20,15 @@ jobs:
steps:
- name: Checkout Code
uses: actions/checkout@v4
- name: Set up Python 3.12
uses: actions/setup-python@v5
- name: Setup Environment
uses: ./.github/actions/setup-tinygrad
with:
python-version: 3.12
- name: Install docs dependencies (no cache)
run: pip install -e '.[docs]'
deps: docs
opencl: 'true'
amd: 'true'
cuda: 'true'
webgpu: 'true'
llvm: 'true'
- name: Install capstone for CLANG disassembly
run: pip install capstone
- name: Use as an external package
@@ -48,7 +51,8 @@ jobs:
cp $GITHUB_WORKSPACE/examples/beautiful_mnist.py .
PYTHONPATH=$GITHUB_WORKSPACE BS=2 STEPS=10 python beautiful_mnist.py
- name: Test Docs Build
run: python -m mkdocs build --strict
# NOTE: the default is GPU here, but you can't use it since it lacks 'cl_khr_fp16'
run: CLANG=1 python -m mkdocs build --strict
- name: Test Docs
run: |
python docs/abstractions2.py
@@ -57,47 +61,6 @@ jobs:
run: awk '/```python/{flag=1;next}/```/{flag=0}flag' docs/quickstart.md > quickstart.py && PYTHONPATH=. python quickstart.py
- name: Test DEBUG
run: DEBUG=100 python3 -c "from tinygrad import Tensor; N = 1024; a, b = Tensor.rand(N, N), Tensor.rand(N, N); c = (a.reshape(N, 1, N) * b.T.reshape(1, N, N)).sum(axis=2); print((c.numpy() - (a.numpy() @ b.numpy())).mean())"
- name: Install OpenCL
run: |
echo 'Acquire::http::Pipeline-Depth "5";' | sudo tee -a /etc/apt/apt.conf.d/99parallel
echo "deb [ allow-insecure=yes ] https://apt.repos.intel.com/oneapi all main" | sudo tee /etc/apt/sources.list.d/oneAPI.list
sudo apt update || true
sudo apt install --allow-unauthenticated -y --no-install-recommends opencl-headers \
intel-oneapi-runtime-openmp=2023.2.1-16 intel-oneapi-runtime-compilers-common=2023.2.1-16 intel-oneapi-runtime-compilers=2023.2.1-16 \
intel-oneapi-runtime-dpcpp-sycl-opencl-cpu=2023.2.1-16 intel-oneapi-runtime-tbb-common=2021.10.0-49541 \
intel-oneapi-runtime-tbb=2021.10.0-49541 intel-oneapi-runtime-opencl=2023.2.1-16
- name: Install packages (cuda)
run: |
echo 'Acquire::http::Pipeline-Depth "5";' | sudo tee -a /etc/apt/apt.conf.d/99parallel
sudo apt update -y || true
sudo apt install -y --no-install-recommends git g++ cmake ninja-build llvm-15-dev zlib1g-dev libglew-dev \
flex bison libfl-dev libboost-thread-dev libboost-filesystem-dev nvidia-cuda-toolkit-gcc libzstd-dev
- name: Install packages (webgpu)
run: |
sudo curl -L https://github.com/wpmed92/pydawn/releases/download/v0.1.6/libwebgpu_dawn.so -o /usr/local/lib/libwebgpu_dawn.so
- name: Install packages (amd)
run: |
echo 'Acquire::http::Pipeline-Depth "5";' | sudo tee -a /etc/apt/apt.conf.d/99parallel
wget https://repo.radeon.com/rocm/rocm.gpg.key -O - | gpg --dearmor | sudo tee /etc/apt/keyrings/rocm.gpg > /dev/null
sudo tee /etc/apt/sources.list.d/rocm.list <<'EOF'
deb [arch=amd64 signed-by=/etc/apt/keyrings/rocm.gpg] https://repo.radeon.com/rocm/apt/6.1.2 jammy main
EOF
echo -e 'Package: *\nPin: release o=repo.radeon.com\nPin-Priority: 600' | sudo tee /etc/apt/preferences.d/rocm-pin-600
sudo apt update || true
sudo apt install --no-install-recommends --allow-unauthenticated -y hsa-rocr comgr hsa-rocr-dev liburing-dev libc6-dev
curl -s https://api.github.com/repos/Qazalin/remu/releases/latest | \
jq -r '.assets[] | select(.name == "libremu.so").browser_download_url' | \
sudo xargs curl -L -o /usr/local/lib/libremu.so
sudo tee --append /etc/ld.so.conf.d/rocm.conf <<'EOF'
/opt/rocm/lib
/opt/rocm/lib64
EOF
sudo ldconfig
- name: Install packages (llvm)
run: |
echo 'Acquire::http::Pipeline-Depth "5";' | sudo tee -a /etc/apt/apt.conf.d/99parallel
sudo apt update -y || true
sudo apt install -y --no-install-recommends llvm-14-dev
- name: Compile EfficientNet to C and test it
run: |
CLANG=1 PYTHONPATH="." python examples/compile_efficientnet.py > recognize.c
@@ -153,17 +116,11 @@ jobs:
steps:
- name: Checkout Code
uses: actions/checkout@v4
- name: Set up Python 3.12
uses: actions/setup-python@v5
- name: Setup Environment
uses: ./.github/actions/setup-tinygrad
with:
python-version: 3.12
- name: Cache python packages
uses: actions/cache@v4
with:
path: ${{ env.Python3_ROOT_DIR }}/lib/python3.12/site-packages
key: uops-packages-${{ hashFiles('**/setup.py') }}-3.12
- name: Install dependencies
run: pip install -e '.[testing]' --extra-index-url https://download.pytorch.org/whl/cpu
key: uops
deps: testing
- name: Test IMAGE=2 support
run: |
IMAGE=2 PYTHON=1 python3 test/test_ops.py TestOps.test_gemm
@@ -230,17 +187,12 @@ jobs:
steps:
- name: Checkout Code
uses: actions/checkout@v4
- name: Set up Python 3.10
uses: actions/setup-python@v5
- name: Setup Environment
uses: ./.github/actions/setup-tinygrad
with:
python-version: "3.10"
- name: Cache python packages
uses: actions/cache@v4
with:
path: ${{ env.Python3_ROOT_DIR }}/lib/python3.10/site-packages
key: linting-packages-${{ hashFiles('**/setup.py') }}-3.10
- name: Install dependencies
run: pip install -e '.[linting,testing]' --extra-index-url https://download.pytorch.org/whl/cpu
key: linting
python-version: '3.10'
deps: linting,testing
- name: Lint bad-indentation and trailing-whitespace with pylint
run: python -m pylint --disable=all -e W0311 -e C0303 --jobs=0 --indent-string=' ' --recursive=y .
- name: Lint with ruff
@@ -277,31 +229,13 @@ jobs:
steps:
- name: Checkout Code
uses: actions/checkout@v4
- name: Install OpenCL
run: |
echo 'Acquire::http::Pipeline-Depth "5";' | sudo tee -a /etc/apt/apt.conf.d/99parallel
echo "deb [ allow-insecure=yes ] https://apt.repos.intel.com/oneapi all main" | sudo tee /etc/apt/sources.list.d/oneAPI.list
sudo apt update || true
sudo apt install --allow-unauthenticated -y --no-install-recommends \
intel-oneapi-runtime-openmp=2023.2.1-16 intel-oneapi-runtime-compilers-common=2023.2.1-16 intel-oneapi-runtime-compilers=2023.2.1-16 \
intel-oneapi-runtime-dpcpp-sycl-opencl-cpu=2023.2.1-16 intel-oneapi-runtime-tbb-common=2021.10.0-49541 \
intel-oneapi-runtime-tbb=2021.10.0-49541 intel-oneapi-runtime-opencl=2023.2.1-16
- name: Set up Python 3.11
uses: actions/setup-python@v5
- name: Setup Environment
uses: ./.github/actions/setup-tinygrad
with:
python-version: 3.11
- name: Cache python packages
uses: actions/cache@v4
with:
path: ${{ env.Python3_ROOT_DIR }}/lib/python3.11/site-packages
key: testing-packages-${{ hashFiles('**/setup.py') }}
- name: Cache downloads
uses: actions/cache@v4
with:
path: ~/.cache/tinygrad/downloads/
key: downloads-cache-${{ matrix.task }}-${{ env.DOWNLOAD_CACHE_VERSION }}
- name: Install Dependencies
run: pip install -e '.[testing,testing_tf]' --extra-index-url https://download.pytorch.org/whl/cpu
key: alt-${{ matrix.task }}
deps: testing,testing_tf
python-version: '3.11'
opencl: 'true'
- if: ${{ matrix.task == 'optimage' }}
name: Run Kernel Count Test
run: PYTHONPATH="." GPU=1 python -m pytest -n=auto test/external/external_test_opt.py
@@ -360,10 +294,7 @@ jobs:
name: Run handcode_opt
run: PYTHONPATH=. MODEL=resnet GPU=1 DEBUG=1 BS=4 HALF=0 python3 examples/handcode_opt.py
- name: Run process replay tests
run: |
export PR_TITLE=$(jq -r .pull_request.title "$GITHUB_EVENT_PATH")
export COMMIT_MESSAGE=$(git show -s --format=%B ${{ github.event.pull_request.head.sha }})
cp test/external/process_replay/process_replay.py ./process_replay.py && git fetch origin master && git -c advice.detachedHead=false checkout origin/master && PYTHONPATH=. python3 process_replay.py
uses: ./.github/actions/process-replay
testdsp:
name: DSP Tests
@@ -372,12 +303,8 @@ jobs:
steps:
- name: Checkout Code
uses: actions/checkout@v4
- name: Set up Python 3.12
uses: actions/setup-python@v5
with:
python-version: 3.12
- name: Install Dependencies
run: pip install --user -e .
- name: Setup Environment
uses: ./.github/actions/setup-tinygrad
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@v3
- name: Build QEMU Docker with cache
@@ -399,29 +326,13 @@ jobs:
steps:
- name: Checkout Code
uses: actions/checkout@v4
- name: Set up Python 3.11
uses: actions/setup-python@v5
- name: Setup Environment
uses: ./.github/actions/setup-tinygrad
with:
python-version: 3.11
- name: Cache python packages
uses: actions/cache@v4
with:
path: ~/.local/lib/python3.11/site-packages
key: webgpu-testing-user3-packages-${{ hashFiles('**/setup.py') }}
- name: Install Dependencies
run: pip install --user -e '.[testing]' --extra-index-url https://download.pytorch.org/whl/cpu
- name: Install dawn (WebGPU)
run: |
sudo curl -L https://github.com/wpmed92/pydawn/releases/download/v0.1.6/libwebgpu_dawn.so -o /usr/lib/libwebgpu_dawn.so
- name: Install dependencies for software-based vulkan
run: |
sudo apt update -y || true
sudo apt install -y libegl1-mesa libgl1-mesa-dri libxcb-xfixes0-dev mesa-vulkan-drivers
- name: Cache downloads
uses: actions/cache@v4
with:
path: ~/.cache/tinygrad/downloads/
key: downloads-cache-webgpu-${{ env.DOWNLOAD_CACHE_VERSION }}
key: webgpu
deps: testing
python-version: '3.11'
webgpu: 'true'
- name: Check Device.DEFAULT (WEBGPU) and print some source
run: |
WEBGPU=1 python -c "from tinygrad import Device; assert Device.DEFAULT == 'WEBGPU', Device.DEFAULT"
@@ -432,10 +343,7 @@ jobs:
--ignore=test/test_copy_speed.py --ignore=test/test_rearrange_einops.py --ignore=test/test_speed_v_torch.py --ignore=test/test_transcendental.py \
--ignore=test/test_fuzz_shape_ops.py --ignore=test/test_linearizer_failures.py --durations=20
- name: Run process replay tests
run: |
export PR_TITLE=$(jq -r .pull_request.title "$GITHUB_EVENT_PATH")
export COMMIT_MESSAGE=$(git show -s --format=%B ${{ github.event.pull_request.head.sha }})
cp test/external/process_replay/process_replay.py ./process_replay.py && git fetch origin master && git -c advice.detachedHead=false checkout origin/master && PYTHONPATH=. python3 process_replay.py
uses: ./.github/actions/process-replay
testmetal:
name: Metal Tests
@@ -445,28 +353,13 @@ jobs:
steps:
- name: Checkout Code
uses: actions/checkout@v4
- name: Setup Environment
uses: ./.github/actions/setup-tinygrad
with:
fetch-depth: 2 # NOTE: this fetches the HEAD commit of the PR
- name: Set up Python 3.11
uses: actions/setup-python@v5
with:
python-version: 3.11
- name: Cache python packages
uses: actions/cache@v4
with:
path: /Users/runner/Library/Python/3.11/lib/python/site-packages
key: metal-m1-testing-user3-packages-${{ hashFiles('**/setup.py') }}
- name: Install Dependencies
run: pip install --user -e '.[webgpu,testing]' --extra-index-url https://download.pytorch.org/whl/cpu
- name: Install dawn (WebGPU)
run: |
sudo mkdir -p /usr/local/lib
sudo curl -L https://github.com/wpmed92/pydawn/releases/download/v0.1.6/libwebgpu_dawn.dylib -o /usr/local/lib/libwebgpu_dawn.dylib
- name: Cache downloads
uses: actions/cache@v4
with:
path: ~/Library/Caches/tinygrad/downloads/
key: downloads-cache-metal-only-${{ env.DOWNLOAD_CACHE_VERSION }}
key: metal
deps: testing
python-version: '3.11'
webgpu: 'true'
- name: Check Device.DEFAULT (METAL) and print some source
run: |
METAL=1 python -c "from tinygrad import Device; assert Device.DEFAULT == 'METAL', Device.DEFAULT"
@@ -501,10 +394,7 @@ jobs:
- name: Run WEBGPU Efficientnet
run: node test/web/test_webgpu.js
- name: Run process replay tests
run: |
export PR_TITLE=$(jq -r .pull_request.title "$GITHUB_EVENT_PATH")
export COMMIT_MESSAGE=$(git show -s --format=%B ${{ github.event.pull_request.head.sha }})
cp test/external/process_replay/process_replay.py ./process_replay.py && git fetch origin master && git -c advice.detachedHead=false checkout origin/master && PYTHONPATH=. python3 process_replay.py
uses: ./.github/actions/process-replay
tests:
strategy:
@@ -519,86 +409,16 @@ jobs:
steps:
- name: Checkout Code
uses: actions/checkout@v4
- name: Setup Environment
uses: ./.github/actions/setup-tinygrad
with:
fetch-depth: 2 # NOTE: this fetches the HEAD commit of the PR
- name: Set up Python 3.12
uses: actions/setup-python@v5
with:
python-version: 3.12
- name: Cache python packages
uses: actions/cache@v4
with:
path: ${{ env.Python3_ROOT_DIR }}/lib/python3.12/site-packages
key: ${{ matrix.backend }}-packages-${{ hashFiles('**/setup.py') }}
- name: Cache downloads
uses: actions/cache@v4
with:
path: ~/.cache/tinygrad/downloads/
key: downloads-cache-${{ matrix.backend }}-${{ env.DOWNLOAD_CACHE_VERSION }}
key: ${{ matrix.backend }}
deps: testing${{matrix.backend=='ptx'&&',cuda'||matrix.backend=='triton'&&',triton'||''}}
opencl: ${{ matrix.backend == 'gpu' && 'true' }}
amd: ${{ matrix.backend == 'amd' && 'true' }}
cuda: ${{ (matrix.backend == 'ptx' || matrix.backend == 'triton' || matrix.backend == 'nv') && 'true' }}
- name: Set env
run: printf "${{ matrix.backend == 'llvm' && 'LLVM=1' || matrix.backend == 'clang' && 'CLANG=1' || matrix.backend == 'gpu' && 'GPU=1' || matrix.backend == 'PTX' && 'FORWARD_ONLY=1\nJIT=1\nOPT=2\nCUDA=1\nPTX=1\nMOCKGPU=1' || matrix.backend == 'triton' && 'FORWARD_ONLY=1\nJIT=1\nOPT=2\nNV=1\nMOCKGPU=1\nTRITON=1\nTRITON_PTXAS_PATH=/usr/bin/ptxas' || matrix.backend == 'amd' && 'AMD=1\nMOCKGPU=1\nFORWARD_ONLY=1' || matrix.backend == 'nv' && 'NV=1\nMOCKGPU=1\nFORWARD_ONLY=1' }}" >> $GITHUB_ENV
- name: Install OpenCL
if: matrix.backend == 'gpu'
run: |
echo 'Acquire::http::Pipeline-Depth "5";' | sudo tee -a /etc/apt/apt.conf.d/99parallel
echo "deb [ allow-insecure=yes ] https://apt.repos.intel.com/oneapi all main" | sudo tee /etc/apt/sources.list.d/oneAPI.list
sudo apt update || true
sudo apt install --allow-unauthenticated -y --no-install-recommends opencl-headers \
intel-oneapi-runtime-openmp=2023.2.1-16 intel-oneapi-runtime-compilers-common=2023.2.1-16 intel-oneapi-runtime-compilers=2023.2.1-16 \
intel-oneapi-runtime-dpcpp-sycl-opencl-cpu=2023.2.1-16 intel-oneapi-runtime-tbb-common=2021.10.0-49541 \
intel-oneapi-runtime-tbb=2021.10.0-49541 intel-oneapi-runtime-opencl=2023.2.1-16
- name: Install packages (cuda)
if: matrix.backend == 'ptx' || matrix.backend == 'triton' || matrix.backend == 'nv'
run: |
echo 'Acquire::http::Pipeline-Depth "5";' | sudo tee -a /etc/apt/apt.conf.d/99parallel
sudo apt update -y || true
sudo apt install -y --no-install-recommends git g++ cmake ninja-build llvm-15-dev zlib1g-dev libglew-dev \
flex bison libfl-dev libboost-thread-dev libboost-filesystem-dev nvidia-cuda-toolkit-gcc libzstd-dev
- name: Cache gpuocelot
if: matrix.backend == 'ptx' || matrix.backend == 'triton' || matrix.backend == 'nv'
id: cache-build
uses: actions/cache@v4
env:
cache-name: cache-gpuocelot-build
with:
path: ${{ github.workspace }}/gpuocelot/ocelot
key: ubuntu22.04-gpuocelot-4524e34adb7eaccc6f71262f2e21d7052bb17c2f-rebuild-9
- name: Clone/compile gpuocelot
if: (matrix.backend == 'ptx' || matrix.backend == 'triton' || matrix.backend == 'nv') && steps.cache-build.outputs.cache-hit != 'true'
run: |
git clone --recurse-submodules https://github.com/gpuocelot/gpuocelot.git ${{ github.workspace }}/gpuocelot
cd ${{ github.workspace }}/gpuocelot/ocelot
git checkout 4524e34adb7eaccc6f71262f2e21d7052bb17c2f
mkdir build
cd build
cmake .. -Wno-dev -G Ninja -DOCELOT_BUILD_TOOLS=OFF -DCMAKE_BUILD_ALWAYS=0 -DBUILD_TESTS_CUDA=OFF
ninja
- name: Install gpuocelot
if: matrix.backend == 'ptx' || matrix.backend == 'triton' || matrix.backend == 'nv'
run: |
cd ${{ github.workspace }}/gpuocelot/ocelot/build
sudo cp libgpuocelot.so /usr/lib/libgpuocelot.so
- name: Install packages (amd)
if: matrix.backend == 'amd'
run: |
echo 'Acquire::http::Pipeline-Depth "5";' | sudo tee -a /etc/apt/apt.conf.d/99parallel
wget https://repo.radeon.com/rocm/rocm.gpg.key -O - | gpg --dearmor | sudo tee /etc/apt/keyrings/rocm.gpg > /dev/null
sudo tee /etc/apt/sources.list.d/rocm.list <<'EOF'
deb [arch=amd64 signed-by=/etc/apt/keyrings/rocm.gpg] https://repo.radeon.com/rocm/apt/6.1.2 jammy main
EOF
echo -e 'Package: *\nPin: release o=repo.radeon.com\nPin-Priority: 600' | sudo tee /etc/apt/preferences.d/rocm-pin-600
sudo apt update || true
sudo apt install --no-install-recommends --allow-unauthenticated -y hsa-rocr comgr hsa-rocr-dev liburing-dev libc6-dev
curl -s https://api.github.com/repos/Qazalin/remu/releases/latest | \
jq -r '.assets[] | select(.name == "libremu.so").browser_download_url' | \
sudo xargs curl -L -o /usr/local/lib/libremu.so
sudo tee --append /etc/ld.so.conf.d/rocm.conf <<'EOF'
/opt/rocm/lib
/opt/rocm/lib64
EOF
sudo ldconfig
- name: Install dependencies
run: pip install -e '.[testing${{matrix.backend=='ptx'&&',cuda'||matrix.backend=='triton'&&',triton'||''}}]' --extra-index-url https://download.pytorch.org/whl/cpu --extra-index-url https://aiinfra.pkgs.visualstudio.com/PublicPackages/_packaging/Triton-Nightly/pypi/simple/
- name: Check Device.DEFAULT and print some source
run: |
PYTHONPATH=${{ github.workspace }} python3 -c "from tinygrad import Device; assert Device.DEFAULT in ['LLVM','CLANG','CUDA','GPU','AMD','NV'], Device.DEFAULT"
@@ -618,10 +438,7 @@ jobs:
- name: Run TRANSCENDENTAL math
run: TRANSCENDENTAL=2 python -m pytest -n=auto test/test_ops.py::TestOps::test_sin test/test_ops.py::TestOps::test_cos test/test_ops.py::TestOps::test_tan test/test_ops.py::TestOps::test_exp test/test_ops.py::TestOps::test_log --durations=20
- name: Run process replay tests
run: |
export PR_TITLE=$(jq -r .pull_request.title "$GITHUB_EVENT_PATH")
export COMMIT_MESSAGE=$(git show -s --format=%B ${{ github.event.pull_request.head.sha }})
cp test/external/process_replay/process_replay.py ./process_replay.py && git fetch origin master && git -c advice.detachedHead=false checkout origin/master && PYTHONPATH=. python3 process_replay.py
uses: ./.github/actions/process-replay
osxtests:
strategy:
@@ -633,33 +450,13 @@ jobs:
steps:
- name: Checkout Code
uses: actions/checkout@v4
- name: Setup Environment
uses: ./.github/actions/setup-tinygrad
with:
fetch-depth: 2 # NOTE: this fetches the HEAD commit of the PR
- name: Set up Python 3.12
uses: actions/setup-python@v5
with:
python-version: 3.12
- name: Cache python packages
uses: actions/cache@v4
with:
path: /Users/runner/Library/Python/3.12/lib/python/site-packages
key: mockgpu-osx-${{ matrix.backend }}-packages-${{ hashFiles('**/setup.py') }}
- name: Install llvm
run: |
brew install llvm
- name: Install comgr
run: |
sudo mkdir -p /usr/local/lib
curl -s -H "Authorization: token $GH_TOKEN" curl -s https://api.github.com/repos/nimlgen/amdcomgr_dylib/releases/latest | \
jq -r '.assets[] | select(.name == "libamd_comgr.dylib").browser_download_url' | \
sudo xargs curl -L -o /usr/local/lib/libamd_comgr.dylib
- name: Install remu
run: |
curl -s -H "Authorization: token $GH_TOKEN" curl -s https://api.github.com/repos/Qazalin/remu/releases/latest | \
jq -r '.assets[] | select(.name == "libremu.dylib").browser_download_url' | \
sudo xargs curl -L -o /usr/local/lib/libremu.dylib
- name: Install dependencies
run: pip install --user -e '.[testing]' --extra-index-url https://download.pytorch.org/whl/cpu
key: macos-other
deps: testing
amd: 'true'
llvm: 'true'
- name: Check Device.DEFAULT and print some source (AMD)
env:
PYTHONPATH: ${{ github.workspace }}
@@ -710,27 +507,11 @@ jobs:
steps:
- name: Checkout Code
uses: actions/checkout@v4
- name: Setup Environment
uses: ./.github/actions/setup-tinygrad
with:
fetch-depth: 2 # NOTE: this fetches the HEAD commit of the PR
- name: Set up Python 3.12
uses: actions/setup-python@v5
with:
python-version: 3.12
- name: Cache python packages
uses: actions/cache@v4
with:
path: ${{ env.Python3_ROOT_DIR }}\Lib\site-packages
key: windows-packages-${{ hashFiles('**/setup.py') }}
- name: Set env
shell: bash
run: |
if [ "${{ matrix.backend }}" = "clang" ]; then
echo "CLANG=1" >> $GITHUB_ENV
elif [ "${{ matrix.backend }}" = "llvm" ]; then
echo "LLVM=1" >> $GITHUB_ENV
fi
- name: Install dependencies
run: pip install -e '.[testing]' --extra-index-url https://download.pytorch.org/whl/cpu
key: windows
deps: testing
- name: Check Device.DEFAULT and print some source (llvm)
shell: bash
run: |
@@ -749,28 +530,3 @@ jobs:
shell: bash
run: |
CLANG=1 python -m pytest -n=auto test/test_tiny.py test/test_ops.py --durations=20
#testunicorn:
# name: ARM64 unicorn Test
# runs-on: ubuntu-latest
# timeout-minutes: 20
# steps:
# - name: Checkout Code
# uses: actions/checkout@v4
# - name: Set up Python 3.11
# uses: actions/setup-python@v5
# with:
# python-version: 3.11
# - name: Cache python packages
# uses: actions/cache@v4
# with:
# path: ${{ env.Python3_ROOT_DIR }}/lib/python3.11/site-packages
# key: testing-arm-packages-${{ hashFiles('**/setup.py') }}
# - name: Install cross-assembler
# run: |
# sudo apt update -y
# sudo apt install -y --no-install-recommends gcc-aarch64-linux-gnu
# - name: Install dependencies
# run: pip install -e '.[testing,arm]' --extra-index-url https://download.pytorch.org/whl/cpu
# - name: Test arm
# run: CI=1 ARM64=1 CLANG=1 python -m pytest -n=auto test/ -k 'not (test_nn.py and (test_conv_transpose2d or test_conv2d))' --ignore=test/models --ignore=test/test_speed_v_torch.py --ignore=test/test_net_speed.py --ignore=test/test_specific_conv.py --ignore=test/unit/test_disk_tensor.py