mawei/StarPilot
1
0
Fork 0
mirror of https://github.com/firestar5683/StarPilot.git synced 2026-09-11 02:33:51 +08:00
Code Issues Packages Projects Releases Wiki Activity

Team Noah

Browse Source
This commit is contained in:
firestar5683
2026-09-03 23:31:33 -05:00
parent fed4ce6ee0
commit dd7ac353bd
307 changed files with 15168 additions and 6906 deletions
Download Patch File Download Diff File Expand all files Collapse all files
tinygrad_repo/.github/workflows/autogen.yml
Vendored
+4 -43
View File
@@ -35,15 +35,15 @@ jobs:
key: 'autogen'
amd: 'true'
llvm: 'true'
pydeps: 'pyyaml mako'
deps: 'autogen'
- name: Install autogen support packages
run: sudo apt-get install -y --no-install-recommends libclang-20-dev llvm-20-dev hip-dev libusb-1.0-0-dev libdrm-dev liburing-dev
- name: Regenerate autogen files
run: |
find tinygrad/runtime/autogen -type f -name "*.py" -not -path "*/amd/*" -not -name "__init__.py" -not -name "comgr.py" -not -name "metal.py" -not -name "iokit.py" -not -name "corefoundation.py" -not -name "libclang.py" -delete
find tinygrad/runtime/autogen -type f -name "*.py" -not -path "*/amd/*" -not -name "__init__.py" -not -name "metal.py" -not -name "iokit.py" -not -name "corefoundation.py" -not -name "libclang.py" -delete
python3 -c "from tinygrad.runtime.autogen import opencl"
python3 -c "from tinygrad.runtime.autogen import cuda, nvrtc, nvjitlink, nv_570, nv_580, nv_610, nv"
python3 -c "from tinygrad.runtime.autogen import comgr_3, hsa, hip, amd_gpu, sqtt, rocprof, amdgpu_kd, amdgpu_drm"
python3 -c "from tinygrad.runtime.autogen import comgr, comgr_3, hsa, hip, amd_gpu, sqtt, rocprof, amdgpu_kd, amdgpu_drm"
python3 -c "from tinygrad.runtime.autogen.am import *"
python3 -c "from tinygrad.runtime.autogen.nv_regs import *"
python3 -c "from tinygrad.runtime.autogen import libc, kfd, io_uring, pci, vfio"
@@ -54,7 +54,7 @@ jobs:
python3 -c "from tinygrad.runtime.autogen import mesa"
python3 -c "from tinygrad.runtime.autogen import avcodec"
python3 -c "from tinygrad.runtime.autogen import llvm_qcom"
python3 -c "from tinygrad.runtime.autogen import mlx5"
python3 -c "from tinygrad.runtime.autogen import mlx5, bnxt"
python3 -c "from tinygrad.runtime.autogen import ggml_common"
REGEN=1 python3 -c "from tinygrad.runtime.autogen import libclang"
- name: Check for differences
@@ -102,42 +102,3 @@ jobs:
with:
name: autogen-macos-patch
path: autogen-macos.patch
autogen-comgr-2:
name: In-tree Autogen (comgr 2)
runs-on: ubuntu-24.04
timeout-minutes: 15
steps:
- name: Checkout Code
uses: actions/checkout@v6
- name: Setup Environment
uses: ./.github/actions/setup-tinygrad
with:
key: 'autogen-comgr'
- name: Install autogen support packages
run: |
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/6.2 $(lsb_release -cs) 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 -qq update || true
sudo apt-get install -y --no-install-recommends libclang-20-dev comgr
- name: Regenerate autogen files
run: |
rm tinygrad/runtime/autogen/comgr.py
python3 -c "from tinygrad.runtime.autogen import comgr"
- name: Check for differences
run: |
if ! git diff --quiet; then
git diff
git diff > autogen-comgr2.patch
echo "Autogen mismatch detected. Patch available at: ${{ github.server_url }}/${{ github.repository }}/actions/runs/${{ github.run_id }}#artifacts"
exit 1
fi
- name: Upload patch artifact
if: failure()
uses: actions/upload-artifact@v7
with:
name: autogen-comgr2-patch
path: autogen-comgr2.patch
tinygrad_repo/.github/workflows/benchmark.yml
Vendored
+145 -124
View File
@@ -88,13 +88,13 @@ jobs:
fail-fast: false
matrix:
dev: ['METAL', 'AMD', 'NV']
timeout-minutes: 60
timeout-minutes: 30
defaults:
run:
shell: bash -e -o pipefail {0}
env:
DEV: ${{ matrix.dev }}
HCQ2: ${{ matrix.dev == 'AMD' && '1' || '0' }}
HCQ2: '0'
if: github.repository_owner == 'tinygrad'
steps:
- name: Checkout Code
@@ -102,16 +102,11 @@ jobs:
- name: Setup (AMD)
if: ${{ matrix.dev == 'AMD' }}
run: |
./extra/amdpci/setup_python_cap.sh
./extra/hcq/hcq_smi.py amd rmmod
./extra/hcq/hcq_smi.py amd kill_pids
./extra/hcq/hcq_smi.py amd rmmod --expect
./extra/hcq/hcq_smi.py amd kill_pids --sudoless
- name: Setup (NV)
if: ${{ matrix.dev == 'NV' }}
run: sudo lsof -tQ /dev/nvidia* | { xargs -r sudo kill -9 || true; }
- name: Symlink models and datasets
run: |
mkdir -p weights
ln -s /raid/weights/LLaMA-3 weights/LLaMA-3
run: lsof -tQ /dev/nvidia* | { xargs -r kill -9 || true; }
- name: setup staging db
if: github.ref == 'refs/heads/update_benchmark_staging'
run: |
@@ -121,18 +116,14 @@ jobs:
run: python3 test/external/process_replay/reset.py
- name: Run llama3.2
run: BENCHMARK_LOG=llama32_3b-f16 JITBEAM=2 IGNORE_BEAM_CACHE=1 python3 -m tinygrad.llm -m llama3.2:3b-f16 --benchmark --warmup
- name: Run qwen3.6
# qwen3.6:35b-a3b doesn't fit on mac
- name: Run qwen3.8
# qwen3.8:27b doesn't fit on mac
if: ${{ matrix.dev != 'METAL' }}
run: BENCHMARK_LOG=qwen36_35b-a3b JITBEAM=2 IGNORE_BEAM_CACHE=1 python3 -m tinygrad.llm -m qwen3.6:35b-a3b --benchmark --warmup
run: BENCHMARK_LOG=qwen38_27b JITBEAM=2 IGNORE_BEAM_CACHE=1 python3 -m tinygrad.llm -m qwen3.8:27b --benchmark --warmup
- name: Run olmoe
# just metal for now
if: ${{ matrix.dev == 'METAL' }}
run: BENCHMARK_LOG=olmoe JITBEAM=2 IGNORE_BEAM_CACHE=1 python3 -m tinygrad.llm -m olmoe --benchmark --warmup
- name: Run LLaMA-3 8B on 4 GPUs with BEAM
# only run on machines with multiple gpus
if: ${{ matrix.dev != 'METAL' }}
run: BENCHMARK_LOG=llama3_beam_4gpu JITBEAM=2 IGNORE_BEAM_CACHE=1 CAPTURE_PROCESS_REPLAY=0 python3 examples/llama3.py --size 8B --shard 4 --model weights/LLaMA-3/8B-SF-DPO/ --benchmark --temperature 0
- name: Run process replay tests
uses: ./.github/actions/process-replay
@@ -143,13 +134,13 @@ jobs:
fail-fast: false
matrix:
dev: ['METAL', 'AMD', 'NV']
timeout-minutes: 60
timeout-minutes: 10
defaults:
run:
shell: bash -e -o pipefail {0}
env:
DEV: ${{ matrix.dev }}
HCQ2: ${{ matrix.dev == 'AMD' && '1' || '0' }}
HCQ2: '0'
if: github.repository_owner == 'tinygrad'
steps:
- name: Checkout Code
@@ -157,12 +148,11 @@ jobs:
- name: Setup (AMD)
if: ${{ matrix.dev == 'AMD' }}
run: |
./extra/amdpci/setup_python_cap.sh
./extra/hcq/hcq_smi.py amd rmmod
./extra/hcq/hcq_smi.py amd kill_pids
./extra/hcq/hcq_smi.py amd rmmod --expect
./extra/hcq/hcq_smi.py amd kill_pids --sudoless
- name: Setup (NV)
if: ${{ matrix.dev == 'NV' }}
run: sudo lsof -tQ /dev/nvidia* | { xargs -r sudo kill -9 || true; }
run: lsof -tQ /dev/nvidia* | { xargs -r kill -9 || true; }
- name: setup staging db
if: github.ref == 'refs/heads/update_benchmark_staging'
run: |
@@ -182,10 +172,6 @@ jobs:
# slow on metal
if: ${{ matrix.dev != 'METAL' }}
run: time BENCHMARK_LOG=cifar DEFAULT_FLOAT=HALF STEPS=1000 TARGET_EVAL_ACC_PCT=93.0 python3 examples/hlb_cifar10.py
- name: Run full CIFAR training steps w 6 GPUS
# only run on machines with multiple gpus
if: ${{ matrix.dev != 'METAL' }}
run: time BENCHMARK_LOG=cifar_6gpu CAPTURE_PROCESS_REPLAY=0 DEFAULT_FLOAT=HALF STEPS=350 BS=1536 GPUS=6 TARGET_EVAL_ACC_PCT=93.0 python3 examples/hlb_cifar10.py
- name: Run process replay tests
uses: ./.github/actions/process-replay
@@ -196,13 +182,13 @@ jobs:
fail-fast: false
matrix:
dev: ['AMD', 'NV']
timeout-minutes: 60
timeout-minutes: 5
defaults:
run:
shell: bash -e -o pipefail {0}
env:
DEV: ${{ matrix.dev }}
HCQ2: ${{ matrix.dev == 'AMD' && '1' || '0' }}
HCQ2: '0'
if: github.repository_owner == 'tinygrad'
steps:
- name: Checkout Code
@@ -210,12 +196,11 @@ jobs:
- name: Setup (AMD)
if: ${{ matrix.dev == 'AMD' }}
run: |
./extra/amdpci/setup_python_cap.sh
./extra/hcq/hcq_smi.py amd rmmod
./extra/hcq/hcq_smi.py amd kill_pids
./extra/hcq/hcq_smi.py amd rmmod --expect
./extra/hcq/hcq_smi.py amd kill_pids --sudoless
- name: Setup (NV)
if: ${{ matrix.dev == 'NV' }}
run: sudo lsof -tQ /dev/nvidia* | { xargs -r sudo kill -9 || true; }
run: lsof -tQ /dev/nvidia* | { xargs -r kill -9 || true; }
- name: Symlink models and datasets
run: |
mkdir -p extra/datasets
@@ -227,15 +212,8 @@ jobs:
rm -f /tmp/staging.db /tmp/staging.db-shm /tmp/staging.db-wal
- name: reset process replay
run: test/external/process_replay/reset.py
- name: Run MLPerf resnet eval on training data
run: time BENCHMARK_LOG=resnet_eval MODEL=resnet python3 examples/mlperf/model_eval.py
- name: Run 10 MLPerf ResNet50 training steps (1 gpu)
run: BENCHMARK_LOG=resnet_10steps DEFAULT_FLOAT=HALF BENCHMARK=10 BS=256 GPUS=1 MODEL=resnet python3 examples/mlperf/model_train.py
- name: Run 10 MLPerf ResNet50 training steps (6 gpu)
run: BENCHMARK_LOG=resnet_10steps_6gpu CAPTURE_PROCESS_REPLAY=0 DEFAULT_FLOAT=HALF BENCHMARK=10 BS=1536 GPUS=6 MODEL=resnet python3 examples/mlperf/model_train.py
- name: Run 10 MLPerf Bert training steps (6 gpu)
# TODO: remove BERT_LAYERS once scheduler is fast
run: BENCHMARK_LOG=bert_10steps_6gpu CAPTURE_PROCESS_REPLAY=0 DEFAULT_FLOAT=HALF BENCHMARK=10 BS=72 GPUS=6 BERT_LAYERS=2 MODEL=bert python3 examples/mlperf/model_train.py
- name: Run process replay tests
uses: ./.github/actions/process-replay
@@ -246,13 +224,13 @@ jobs:
fail-fast: false
matrix:
dev: ['METAL', 'AMD', 'NV']
timeout-minutes: 60
timeout-minutes: 15
defaults:
run:
shell: bash -e -o pipefail {0}
env:
DEV: ${{ matrix.dev }}
HCQ2: ${{ matrix.dev == 'AMD' && '1' || '0' }}
HCQ2: '0'
if: github.repository_owner == 'tinygrad'
steps:
- name: Checkout Code
@@ -260,12 +238,11 @@ jobs:
- name: Setup (AMD)
if: ${{ matrix.dev == 'AMD' }}
run: |
./extra/amdpci/setup_python_cap.sh
./extra/hcq/hcq_smi.py amd rmmod
./extra/hcq/hcq_smi.py amd kill_pids
./extra/hcq/hcq_smi.py amd rmmod --expect
./extra/hcq/hcq_smi.py amd kill_pids --sudoless
- name: Setup (NV)
if: ${{ matrix.dev == 'NV' }}
run: sudo lsof -tQ /dev/nvidia* | { xargs -r sudo kill -9 || true; }
run: lsof -tQ /dev/nvidia* | { xargs -r kill -9 || true; }
- name: setup staging db
if: github.ref == 'refs/heads/update_benchmark_staging'
run: |
@@ -285,6 +262,59 @@ jobs:
- name: Run process replay tests
uses: ./.github/actions/process-replay
multigpubenchmark:
name: Multi-GPU Benchmarks (DEV=${{ matrix.dev }})
runs-on: [self-hosted, "${{ matrix.dev == 'AMD' && 'tinybox' || 'tinyboxgreen' }}"]
strategy:
fail-fast: false
matrix:
dev: ['AMD', 'NV']
timeout-minutes: 20
defaults:
run:
shell: bash -e -o pipefail {0}
env:
DEV: ${{ matrix.dev }}
HCQ2: '0'
if: github.repository_owner == 'tinygrad'
steps:
- name: Checkout Code
uses: actions/checkout@v6
- name: Setup (AMD)
if: ${{ matrix.dev == 'AMD' }}
run: |
./extra/hcq/hcq_smi.py amd rmmod --expect
./extra/hcq/hcq_smi.py amd kill_pids --sudoless
- name: Setup (NV)
if: ${{ matrix.dev == 'NV' }}
run: lsof -tQ /dev/nvidia* | { xargs -r kill -9 || true; }
- name: Symlink models and datasets
run: |
mkdir -p weights
mkdir -p extra/datasets
ln -s /raid/weights/LLaMA-3 weights/LLaMA-3
ln -s /raid/datasets/imagenet extra/datasets/imagenet
- name: setup staging db
if: github.ref == 'refs/heads/update_benchmark_staging'
run: |
echo "CACHEDB=/tmp/staging.db" >> $GITHUB_ENV
rm -f /tmp/staging.db /tmp/staging.db-shm /tmp/staging.db-wal
- name: reset process replay
run: python3 test/external/process_replay/reset.py
- name: Run LLaMA-3 8B on 4 GPUs with BEAM
run: BENCHMARK_LOG=llama3_beam_4gpu JITBEAM=2 IGNORE_BEAM_CACHE=1 CAPTURE_PROCESS_REPLAY=0 python3 examples/llama3.py --size 8B --shard 4 --model weights/LLaMA-3/8B-SF-DPO/ --benchmark --temperature 0
- name: Run full CIFAR training steps w 6 GPUS
run: time BENCHMARK_LOG=cifar_6gpu CAPTURE_PROCESS_REPLAY=0 DEFAULT_FLOAT=HALF STEPS=350 BS=1536 GPUS=6 TARGET_EVAL_ACC_PCT=93.0 python3 examples/hlb_cifar10.py
- name: Run MLPerf resnet eval on training data
run: time BENCHMARK_LOG=resnet_eval MODEL=resnet python3 examples/mlperf/model_eval.py
- name: Run 10 MLPerf ResNet50 training steps (6 gpu)
run: BENCHMARK_LOG=resnet_10steps_6gpu CAPTURE_PROCESS_REPLAY=0 DEFAULT_FLOAT=HALF BENCHMARK=10 BS=1536 GPUS=6 MODEL=resnet python3 examples/mlperf/model_train.py
- name: Run 10 MLPerf Bert training steps (6 gpu)
# TODO: remove BERT_LAYERS once scheduler is fast
run: BENCHMARK_LOG=bert_10steps_6gpu CAPTURE_PROCESS_REPLAY=0 DEFAULT_FLOAT=HALF BENCHMARK=10 BS=72 GPUS=6 BERT_LAYERS=2 MODEL=bert python3 examples/mlperf/model_train.py
- name: Run process replay tests
uses: ./.github/actions/process-replay
tests:
name: Tests (DEV=${{ matrix.dev }})
runs-on: [self-hosted, "${{ matrix.dev == 'METAL' && 'macOS' || matrix.dev == 'AMD' && 'tinybox' || 'tinyboxgreen' }}"]
@@ -292,7 +322,7 @@ jobs:
fail-fast: false
matrix:
dev: ['METAL', 'AMD', 'NV']
timeout-minutes: 60
timeout-minutes: 10
defaults:
run:
shell: bash -e -o pipefail {0}
@@ -305,12 +335,11 @@ jobs:
- name: Setup (AMD)
if: ${{ matrix.dev == 'AMD' }}
run: |
./extra/amdpci/setup_python_cap.sh
./extra/hcq/hcq_smi.py amd rmmod
./extra/hcq/hcq_smi.py amd kill_pids
./extra/hcq/hcq_smi.py amd rmmod --expect
./extra/hcq/hcq_smi.py amd kill_pids --sudoless
- name: Setup (NV)
if: ${{ matrix.dev == 'NV' }}
run: sudo lsof -tQ /dev/nvidia* | { xargs -r sudo kill -9 || true; }
run: lsof -tQ /dev/nvidia* | { xargs -r kill -9 || true; }
- name: setup staging db
if: github.ref == 'refs/heads/update_benchmark_staging'
run: |
@@ -374,7 +403,7 @@ jobs:
run: python test/external/external_benchmark_multitensor_allreduce.py
- name: HEVC Decode Benchmark
if: ${{ matrix.dev == 'NV' }}
run: VALIDATE=1 MAX_FRAMES=100 ASSERT_FPS=1400 JITBEAM=1 PYTHONPATH=. python3 extra/hevc/decode.py
run: IGNORE_BEAM_CACHE=1 VALIDATE=1 MAX_FRAMES=100 ASSERT_FPS=1400 JITBEAM=1 PYTHONPATH=. python3 extra/hevc/decode.py
- uses: actions/upload-artifact@v7
if: ${{ matrix.dev != 'AMD' }}
with:
@@ -387,7 +416,7 @@ jobs:
testusbgpu:
name: UsbGPU Benchmark
runs-on: [self-hosted, macOS]
timeout-minutes: 10
timeout-minutes: 3
defaults:
run:
shell: bash -e -o pipefail {0}
@@ -402,32 +431,66 @@ jobs:
rm -f /tmp/staging.db /tmp/staging.db-shm /tmp/staging.db-wal
- name: Kill stale pids
run: |
PYTHONPATH=. ./extra/hcq/hcq_smi.py amd kill_pids
PYTHONPATH=. ./extra/hcq/hcq_smi.py nv kill_pids
# since sudo is required for usbgpu on macos, do not write bytecode, as some of the files are owned by root
./extra/hcq/hcq_smi.py amd kill_pids --sudoless
./extra/hcq/hcq_smi.py nv kill_pids --sudoless
- name: UsbGPU boot time
run: sudo -E PYTHONDONTWRITEBYTECODE=1 PYTHONPATH=. GMMU=0 DEBUG=2 AM_RESET=1 DEV=USB+AMD time python3.11 test/test_tiny.py TestTiny.test_plus
run: GMMU=0 DEBUG=2 AM_RESET=1 DEV=USB+AMD time python3.11 test/test_tiny.py TestTiny.test_plus
- name: UsbGPU tiny tests
run: sudo -E PYTHONDONTWRITEBYTECODE=1 PYTHONPATH=. GMMU=0 DEV=USB+AMD python3.11 test/test_tiny.py
run: GMMU=0 DEV=USB+AMD python3.11 test/test_tiny.py
- name: UsbGPU copy speeds
run: sudo -E PYTHONDONTWRITEBYTECODE=1 PYTHONPATH=. GMMU=0 DEV=USB+AMD python3.11 test/external/external_test_usb_asm24.py TestDevCopySpeeds
#- name: UsbGPU openpilot test
# run: sudo -E PYTHONPATH=. GMMU=0 DEV=USB+AMD GRAPH_ONE_KERNEL=1 python3.11 examples/openpilot/compile3.py https://github.com/commaai/openpilot/raw/9118973ed03c1ae1d40cf69a29507ec2cc78efd7/selfdrive/modeld/models/supercombo.onnx
run: SIZE=64000000 PYTHONPATH=. GMMU=0 DEV=USB+AMD python3.11 test/external/external_test_usb_asm24.py TestDevCopySpeeds
- name: UsbGPU (USB4/TB) install script
run: PYTHONPATH=. sh extra/setup_tinygpu_osx.sh
run: sh extra/setup_tinygpu_osx.sh
- name: UsbGPU (USB4/TB) boot time
run: PYTHONPATH=. DEBUG=3 DEV=PCI+NV:NAK time python3.11 test/test_tiny.py TestTiny.test_plus
run: DEBUG=3 DEV=PCI+NV:NAK time python3.11 test/test_tiny.py TestTiny.test_plus
- name: UsbGPU (USB4/TB) tiny tests
run: PYTHONPATH=. DEV=PCI+NV:NAK python3.11 test/test_tiny.py
run: DEV=PCI+NV:NAK python3.11 test/test_tiny.py
testcommalatest:
name: comma Benchmark (0.11.2)
testcomma:
strategy:
matrix:
dev: ['QCOM', 'QCOM:IR3']
version: ['0.11.0', '0.11.2']
model: ['vision', 'policy', 'supercombo', 'dmonitoring']
# exclude non-existent models
exclude: [{ version: '0.11.0', model: supercombo }, { version: '0.11.2', model: vision }, { version: '0.11.2', model: policy }]
include:
- version: '0.11.0'
model: vision
url: https://github.com/commaai/openpilot/raw/v0.11.0/selfdrive/modeld/models/driving_vision.onnx
timing: 18
- version: '0.11.0'
model: policy
url: https://github.com/commaai/openpilot/raw/v0.11.0/selfdrive/modeld/models/driving_policy.onnx
timing: 3.4
- version: '0.11.0'
model: dmonitoring
url: https://github.com/commaai/openpilot/raw/v0.11.0/selfdrive/modeld/models/dmonitoring_model.onnx
timing: 12.5
- version: '0.11.2'
model: supercombo
url: https://gitlab.com/commaai/openpilot-lfs.git/gitlab-lfs/objects/433f85f956837606ad1f1cbee4aa7e2158ad23c768dea914b20436c97232741b
timing: 28
- dev: QCOM:IR3
version: '0.11.2'
model: supercombo
timing: 29
- version: '0.11.2'
model: dmonitoring
url: https://gitlab.com/commaai/openpilot-lfs.git/gitlab-lfs/objects/3e7b31dfbc0a5234f1baf196513b77fc6af12204b8a8ffe8ee0417e48352f316
timing: 12.5
fail-fast: false
name: openpilot ${{ matrix.version }} compile3 ${{ matrix.model }} (DEV=${{ matrix.dev }})
runs-on: [self-hosted, Linux, comma]
timeout-minutes: 12
timeout-minutes: 5
defaults:
run:
shell: bash -e -o pipefail {0}
if: github.repository_owner == 'tinygrad'
env:
DEV: ${{ matrix.dev }}
ASSERT_MIN_STEP_TIME: ${{ matrix.timing }}
BENCHMARK_LOG: ${{ matrix.dev == 'QCOM:IR3' && 'ir3_' || '' }}openpilot_${{ matrix.version }}_${{ matrix.model }}
steps:
- name: Checkout Code
uses: actions/checkout@v6
@@ -438,45 +501,10 @@ jobs:
rm -f /tmp/staging.db /tmp/staging.db-shm /tmp/staging.db-wal
- name: reset process replay
run: test/external/process_replay/reset.py
- name: openpilot compile3 0.11.2 supercombo
run: BENCHMARK_LOG=openpilot_0_11_2_supercombo PYTHONPATH="." ASSERT_MIN_STEP_TIME=26 DEV=QCOM FLOAT16=1 IMAGE=1 NOLOCALS=1 taskset -c 4-7 python3 examples/openpilot/compile3.py https://gitlab.com/commaai/openpilot-lfs.git/gitlab-lfs/objects/433f85f956837606ad1f1cbee4aa7e2158ad23c768dea914b20436c97232741b
- name: openpilot compile3 0.11.2 supercombo (from pickle)
run: BENCHMARK_LOG=openpilot_0_11_2_supercombo_run_pickle RUN_PICKLE=1 PYTHONPATH="." ASSERT_MIN_STEP_TIME=26 DEV=QCOM taskset -c 4-7 python3 examples/openpilot/compile3.py
- name: IR3 openpilot compile3 0.11.2 supercombo
run: BENCHMARK_LOG=ir3_openpilot_0_11_2_supercombo PYTHONPATH="." ASSERT_MIN_STEP_TIME=41 DEV=QCOM:IR3 FLOAT16=1 IMAGE=1 NOLOCALS=1 taskset -c 4-7 python3 examples/openpilot/compile3.py https://gitlab.com/commaai/openpilot-lfs.git/gitlab-lfs/objects/433f85f956837606ad1f1cbee4aa7e2158ad23c768dea914b20436c97232741b
- name: openpilot compile3 0.11.2 dmonitoring
run: BENCHMARK_LOG=openpilot_0_11_2_dmonitoring PYTHONPATH="." ASSERT_MIN_STEP_TIME=11 DEV=QCOM FLOAT16=1 IMAGE=1 NOLOCALS=1 taskset -c 4-7 python3 examples/openpilot/compile3.py https://gitlab.com/commaai/openpilot-lfs.git/gitlab-lfs/objects/3e7b31dfbc0a5234f1baf196513b77fc6af12204b8a8ffe8ee0417e48352f316
- name: Run process replay tests
uses: ./.github/actions/process-replay
testcommaold:
name: comma Benchmark (0.11.0)
runs-on: [self-hosted, Linux, comma]
timeout-minutes: 12
defaults:
run:
shell: bash -e -o pipefail {0}
if: github.repository_owner == 'tinygrad'
steps:
- name: Checkout Code
uses: actions/checkout@v6
- name: setup staging db
if: github.ref == 'refs/heads/update_benchmark_staging'
run: |
echo "CACHEDB=/tmp/staging.db" >> $GITHUB_ENV
rm -f /tmp/staging.db /tmp/staging.db-shm /tmp/staging.db-wal
- name: reset process replay
run: test/external/process_replay/reset.py
- name: openpilot compile3 0.11.0 driving_vision
run: BENCHMARK_LOG=openpilot_0_11_0_vision PYTHONPATH="." ASSERT_MIN_STEP_TIME=17 DEV=QCOM FLOAT16=1 IMAGE=1 NOLOCALS=1 taskset -c 4-7 python3 examples/openpilot/compile3.py https://github.com/commaai/openpilot/raw/v0.11.0/selfdrive/modeld/models/driving_vision.onnx
- name: openpilot compile3 0.11.0 driving_vision (from pickle)
run: BENCHMARK_LOG=openpilot_0_11_0_vision_run_pickle RUN_PICKLE=1 PYTHONPATH="." ASSERT_MIN_STEP_TIME=17 DEV=QCOM taskset -c 4-7 python3 examples/openpilot/compile3.py
- name: IR3 openpilot compile3 0.11.0 driving_vision
run: BENCHMARK_LOG=ir3_openpilot_0_11_0_vision PYTHONPATH="." ASSERT_MIN_STEP_TIME=18 DEV=QCOM:IR3 FLOAT16=1 IMAGE=1 NOLOCALS=1 taskset -c 4-7 python3 examples/openpilot/compile3.py https://github.com/commaai/openpilot/raw/v0.11.0/selfdrive/modeld/models/driving_vision.onnx
- name: openpilot compile3 0.11.0 driving_policy
run: BENCHMARK_LOG=openpilot_0_11_0_policy PYTHONPATH="." ASSERT_MIN_STEP_TIME=3.2 DEV=QCOM FLOAT16=1 IMAGE=1 NOLOCALS=1 taskset -c 4-7 python3 examples/openpilot/compile3.py https://github.com/commaai/openpilot/raw/v0.11.0/selfdrive/modeld/models/driving_policy.onnx
- name: openpilot compile3 0.11.0 dmonitoring
run: BENCHMARK_LOG=openpilot_0_11_0_dmonitoring PYTHONPATH="." ASSERT_MIN_STEP_TIME=11 DEV=QCOM FLOAT16=1 IMAGE=1 NOLOCALS=1 taskset -c 4-7 python3 examples/openpilot/compile3.py https://github.com/commaai/openpilot/raw/v0.11.0/selfdrive/modeld/models/dmonitoring_model.onnx
- name: compile
run: FLOAT16=1 IMAGE=1 taskset -c 4-7 python3 examples/openpilot/compile3.py ${{ matrix.url }}
- name: run pickle
run: BENCHMARK_LOG="${BENCHMARK_LOG}_run_pickle" RUN_PICKLE=1 taskset -c 4-7 python3 examples/openpilot/compile3.py
- name: Run process replay tests
uses: ./.github/actions/process-replay
@@ -489,15 +517,6 @@ jobs:
shell: bash -e -o pipefail {0}
if: github.repository_owner == 'tinygrad'
steps:
- name: Checkout Code
uses: actions/checkout@v6
- name: setup staging db
if: github.ref == 'refs/heads/update_benchmark_staging'
run: |
echo "CACHEDB=/tmp/staging.db" >> $GITHUB_ENV
rm -f /tmp/staging.db /tmp/staging.db-shm /tmp/staging.db-wal
- name: reset process replay
run: test/external/process_replay/reset.py
- name: Checkout Code
uses: actions/checkout@v6
- name: setup staging db
@@ -521,7 +540,7 @@ jobs:
testcommausbgpubenchmark:
name: UsbGPU Benchmark (comma)
runs-on: [self-hosted, Linux, comma4]
timeout-minutes: 20
timeout-minutes: 10
defaults:
run:
shell: bash -e -o pipefail {0}
@@ -541,7 +560,7 @@ jobs:
- name: openpilot run_pickle big_driving_supercombo
run: BENCHMARK_LOG=usbgpu_openpilot_big_driving_supercombo_run_pickle RUN_PICKLE=1 PICKLE_OOB=1 PYTHONPATH="." GMMU=0 DEV=USB+AMD ASSERT_MIN_STEP_TIME=50 python3 examples/openpilot/compile3.py - openpilot.pkl
- name: Test copy speeds
run: SIZE=64e6 PYTHONPATH=. GMMU=0 DEV=USB+AMD python3 test/external/external_test_usb_asm24.py TestDevCopySpeeds
run: SIZE=64000000 PYTHONPATH=. GMMU=0 DEV=USB+AMD python3 test/external/external_test_usb_asm24.py TestDevCopySpeeds
driverbenchmarks:
name: PCI Driver Benchmark (DEV=${{ matrix.dev }})
@@ -550,7 +569,7 @@ jobs:
fail-fast: false
matrix:
dev: ['AMD', 'NV']
timeout-minutes: 20
timeout-minutes: 5
defaults:
run:
shell: bash -e -o pipefail {0}
@@ -562,9 +581,8 @@ jobs:
uses: actions/checkout@v6
- name: Setup
run: |
./extra/amdpci/setup_python_cap.sh
./extra/hcq/hcq_smi.py ${{ matrix.dev == 'AMD' && 'amd' || 'nv' }} rmmod
./extra/hcq/hcq_smi.py ${{ matrix.dev == 'AMD' && 'amd' || 'nv' }} kill_pids
./extra/hcq/hcq_smi.py ${{ matrix.dev }} rmmod --expect
./extra/hcq/hcq_smi.py ${{ matrix.dev }} kill_pids --sudoless
mkdir -p extra/datasets
ln -s /raid/datasets/imagenet extra/datasets/imagenet
- name: setup staging db
@@ -599,6 +617,9 @@ jobs:
run: |
GRAPH_ONE_KERNEL=1 NSZ=8192 python3 test/speed/external_test_copy_speed.py TestCopySpeed.testCopyDefaulttoCPUJit
GRAPH_ONE_KERNEL=1 NSZ=8192 python3 test/speed/external_test_copy_speed.py TestCopySpeed.testCopyCPUtoDefaultJit
- name: HEVC Decode Benchmark
if: ${{ matrix.dev == 'NV' }}
run: IGNORE_BEAM_CACHE=1 VALIDATE=1 MAX_FRAMES=100 ASSERT_FPS=1400 JITBEAM=1 PYTHONPATH=. python3 extra/hevc/decode.py
- name: Run 10 MLPerf ResNet50 training steps (1 gpu)
if: ${{ matrix.dev == 'NV' }}
run: BENCHMARK_LOG=resnet_10steps MNISTMOCK=1 DEFAULT_FLOAT=HALF BENCHMARK=10 BS=256 GPUS=1 MODEL=resnet python3 examples/mlperf/model_train.py
@@ -621,12 +642,12 @@ jobs:
llvmspeed:
name: LLVM Speed
runs-on: [self-hosted, Linux, tinyboxrandom]
timeout-minutes: 20
timeout-minutes: 10
if: github.repository_owner == 'tinygrad'
steps:
- name: Checkout Code
uses: actions/checkout@v6
- name: Speed Test
run: DEV=CPU:LLVM THREADS=0 python3 test/speed/external_test_speed_v_torch.py
run: DEV=CPU:LLVM python3 test/speed/external_test_speed_v_torch.py
- name: Speed Test (BEAM=2)
run: BEAM=2 DEV=CPU:LLVM THREADS=0 python3 test/speed/external_test_speed_v_torch.py
run: IGNORE_BEAM_CACHE=1 BEAM=2 DEV=CPU:LLVM python3 test/speed/external_test_speed_v_torch.py
tinygrad_repo/.github/workflows/docs.yml
Vendored
+1 -1
View File
@@ -8,7 +8,7 @@ permissions:
contents: write
jobs:
deploy:
runs-on: ubuntu-latest
runs-on: ubuntu-24.04
steps:
- uses: actions/checkout@v6
- name: Configure Git Credentials
tinygrad_repo/.github/workflows/platform.yml
Vendored
+1 -33
View File
@@ -166,7 +166,7 @@ jobs:
uses: ./.github/actions/setup-tinygrad
with:
key: windows-${{ matrix.dev }}-minimal
deps: testing_unit
deps: testing_minimal
pydeps: ${{ matrix.dev == 'WEBGPU' && 'dawn-python' || '' }}
- name: Set env
shell: bash
@@ -179,35 +179,3 @@ jobs:
- name: Run test_tiny
shell: bash
run: python -m pytest -n=auto test/test_tiny.py --durations=20
qcomclcompiletests:
name: Compile-only (QCOM CL)
runs-on: ubuntu-24.04-arm
timeout-minutes: 15
steps:
- name: Checkout Code
uses: actions/checkout@v6
- name: Setup Environment
uses: ./.github/actions/setup-tinygrad
with:
key: compile-qcomcl
deps: testing_unit
tinydreno: 'true'
- name: Set env
shell: bash
run: printf "DEV=NULL:QCOMCL:a630\nNULL_ALLOW_COPYOUT=1" >> $GITHUB_ENV
- name: Run test_ops
shell: bash
run: |
python -c "from tinygrad import Device; assert Device.DEFAULT == 'NULL'"
DEBUG=4 python3 test/backend/test_ops.py TestOps.test_add
python -m pytest -n=auto test/backend/test_ops.py --durations=20
- name: Run test_ops (IMAGE)
shell: bash
env:
IMAGE: 1
DEV: "NULL:QCOMCL:a630,IMAGE_PITCH_ALIGNMENT=64"
run: |
DEBUG=4 python test/backend/test_ops.py TestOps.test_gemm | grep read_imagef
python -m pytest -n=auto test/backend/test_ops.py --durations=20
tinygrad_repo/.github/workflows/python-publish.yml
Vendored
+1 -1
View File
@@ -10,7 +10,7 @@ on:
jobs:
deploy:
runs-on: ubuntu-latest
runs-on: ubuntu-24.04
steps:
- uses: actions/checkout@v6
- name: Set up Python
tinygrad_repo/.github/workflows/szdiff.yml
Vendored
+3 -3
View File
@@ -10,7 +10,7 @@ concurrency:
jobs:
checkbranch:
name: Check PR Branch status
runs-on: ubuntu-latest
runs-on: ubuntu-24.04
outputs:
branchstat: ${{ steps.brstat.outputs.stat}}
steps:
@@ -44,7 +44,7 @@ jobs:
permissions:
contents: read
pull-requests: write
runs-on: ubuntu-latest
runs-on: ubuntu-24.04
needs: checkbranch
if: needs.checkbranch.outputs.branchstat == 'false'
steps:
@@ -87,7 +87,7 @@ jobs:
name: Core Library Line Difference
permissions:
pull-requests: write
runs-on: ubuntu-latest
runs-on: ubuntu-24.04
needs: checkbranch
if: needs.checkbranch.outputs.branchstat == 'true'
steps:
tinygrad_repo/.github/workflows/test.yml
Vendored
+66 -53
View File
@@ -21,7 +21,7 @@ concurrency:
jobs:
docs:
name: Docs
runs-on: &linux ${{ github.repository == 'tinygrad/tinygrad' && github.event_name == 'pull_request' && github.event.pull_request.author_association == 'COLLABORATOR' && 'namespace-profile-tinygrad' || 'ubuntu-24.04' }}
runs-on: ${{ github.repository == 'tinygrad/tinygrad' && github.event_name == 'pull_request' && github.event.pull_request.author_association == 'COLLABORATOR' && 'namespace-profile-tinygrad' || 'ubuntu-24.04' }}
timeout-minutes: 10
env:
CHECK_OOB: 0
@@ -31,8 +31,7 @@ jobs:
- name: Setup Environment
uses: ./.github/actions/setup-tinygrad
with:
deps: docs
pydeps: "capstone torch"
deps: "docs testing_minimal"
- name: Build wheel and show size
run: |
uv build --wheel
@@ -61,7 +60,7 @@ jobs:
torchbackend:
name: Torch Backend Tests
runs-on: *linux
runs-on: ${{ github.repository == 'tinygrad/tinygrad' && github.event_name == 'pull_request' && github.event.pull_request.author_association == 'COLLABORATOR' && 'namespace-profile-tinygrad' || 'ubuntu-24.04' }}
timeout-minutes: 15
steps:
- name: Checkout Code
@@ -73,10 +72,7 @@ jobs:
deps: testing_unit
pydeps: "pillow torchvision expecttest"
llvm: 'true'
- name: Install ninja
run: |
sudo apt update || true
sudo apt install -y --no-install-recommends ninja-build
ninja: 'true'
- name: Test ResNet-18
run: DEBUG=2 python3 extra/torch_backend/example.py
- name: Test one op in torch tests
@@ -86,9 +82,26 @@ jobs:
- name: Custom tests
run: DEV=CPU:LLVM GPUS=4 TINY_BACKEND=1 python3 -m pytest -nauto extra/torch_backend/test.py extra/torch_backend/test_inplace.py extra/torch_backend/test_multigpu.py extra/torch_backend/test_kernel_fusion.py --durations=20
torchbackendtrain:
name: Torch Backend Training
runs-on: ${{ github.repository == 'tinygrad/tinygrad' && github.event_name == 'pull_request' && github.event.pull_request.author_association == 'COLLABORATOR' && 'namespace-profile-tinygrad' || 'ubuntu-24.04' }}
timeout-minutes: 15
steps:
- name: Checkout Code
uses: actions/checkout@v6
- name: Setup Environment
uses: ./.github/actions/setup-tinygrad
with:
key: torch-backend-pillow-torchvision-et-pt
deps: testing_unit
llvm: 'true'
ninja: 'true'
- name: Test beautiful_mnist in torch with TINY_BACKEND
run: STEPS=20 DEV=CPU TARGET_EVAL_ACC_PCT=90.0 MAX_BUFFER_SIZE=0 TINY_BACKEND=1 python3 examples/other_mnist/beautiful_mnist_torch.py
bepython:
name: Python Backend
runs-on: *linux
runs-on: ${{ github.repository == 'tinygrad/tinygrad' && github.event_name == 'pull_request' && github.event.pull_request.author_association == 'COLLABORATOR' && 'namespace-profile-tinygrad' || 'ubuntu-24.04' }}
timeout-minutes: 15
steps:
- name: Checkout Code
@@ -126,7 +139,7 @@ jobs:
linter:
name: Linters
runs-on: *linux
runs-on: ${{ github.repository == 'tinygrad/tinygrad' && github.event_name == 'pull_request' && github.event.pull_request.author_association == 'COLLABORATOR' && 'namespace-profile-tinygrad' || 'ubuntu-24.04' }}
timeout-minutes: 10
steps:
@@ -157,7 +170,7 @@ jobs:
nulltest:
name: Null Tests
runs-on: *linux
runs-on: ${{ github.repository == 'tinygrad/tinygrad' && github.event_name == 'pull_request' && github.event.pull_request.author_association == 'COLLABORATOR' && 'namespace-profile-tinygrad' || 'ubuntu-24.04' }}
timeout-minutes: 15
steps:
@@ -191,7 +204,7 @@ jobs:
unittest:
name: Unit Tests
runs-on: *linux
runs-on: ${{ github.repository == 'tinygrad/tinygrad' && github.event_name == 'pull_request' && github.event.pull_request.author_association == 'COLLABORATOR' && 'namespace-profile-tinygrad' || 'ubuntu-24.04' }}
timeout-minutes: 15
steps:
@@ -220,7 +233,7 @@ jobs:
- name: Run process replay tests
uses: ./.github/actions/process-replay
- name: Repo line count <= 26000 lines
run: MAX_LINE_COUNT=26000 python sz.py
run: MAX_LINE_COUNT=26500 python sz.py
spec:
strategy:
@@ -228,7 +241,7 @@ jobs:
matrix:
group: [1, 2]
name: SPEC=2 (${{ matrix.group }})
runs-on: *linux
runs-on: ${{ github.repository == 'tinygrad/tinygrad' && github.event_name == 'pull_request' && github.event.pull_request.author_association == 'COLLABORATOR' && 'namespace-profile-tinygrad' || 'ubuntu-24.04' }}
timeout-minutes: 15
steps:
- name: Checkout Code
@@ -244,7 +257,7 @@ jobs:
fuzzing:
name: Fuzzing
runs-on: *linux
runs-on: ${{ github.repository == 'tinygrad/tinygrad' && github.event_name == 'pull_request' && github.event.pull_request.author_association == 'COLLABORATOR' && 'namespace-profile-tinygrad' || 'ubuntu-24.04' }}
timeout-minutes: 10
steps:
- name: Checkout Code
@@ -260,7 +273,7 @@ jobs:
testopenclimage:
name: CL IMAGE Tests
runs-on: *linux
runs-on: ${{ github.repository == 'tinygrad/tinygrad' && github.event_name == 'pull_request' && github.event.pull_request.author_association == 'COLLABORATOR' && 'namespace-profile-tinygrad' || 'ubuntu-24.04' }}
timeout-minutes: 15
steps:
- name: Checkout Code
@@ -280,7 +293,7 @@ jobs:
testopenpilot:
name: openpilot Compile Tests
runs-on: *linux
runs-on: ${{ github.repository == 'tinygrad/tinygrad' && github.event_name == 'pull_request' && github.event.pull_request.author_association == 'COLLABORATOR' && 'namespace-profile-tinygrad' || 'ubuntu-24.04' }}
timeout-minutes: 15
steps:
- name: Checkout Code
@@ -309,7 +322,7 @@ jobs:
testonnxcpu:
name: ONNX (CPU) Tests
runs-on: *linux
runs-on: ${{ github.repository == 'tinygrad/tinygrad' && github.event_name == 'pull_request' && github.event.pull_request.author_association == 'COLLABORATOR' && 'namespace-profile-tinygrad' || 'ubuntu-24.04' }}
timeout-minutes: 20
steps:
@@ -328,7 +341,7 @@ jobs:
testoptim:
name: Optimization Tests
runs-on: *linux
runs-on: ${{ github.repository == 'tinygrad/tinygrad' && github.event_name == 'pull_request' && github.event.pull_request.author_association == 'COLLABORATOR' && 'namespace-profile-tinygrad' || 'ubuntu-24.04' }}
timeout-minutes: 20
steps:
- name: Checkout Code
@@ -360,7 +373,7 @@ jobs:
testllm:
name: Test LLM
runs-on: *linux
runs-on: ${{ github.repository == 'tinygrad/tinygrad' && github.event_name == 'pull_request' && github.event.pull_request.author_association == 'COLLABORATOR' && 'namespace-profile-tinygrad' || 'ubuntu-24.04' }}
timeout-minutes: 15
env:
CHECK_OOB: 0
@@ -377,17 +390,17 @@ jobs:
run: |
parallel --link --tagstring '[{1}]' '{2}' \
::: llama 'llama q4' qwen3.5 qwen \
::: $'echo "What\'s a male chicken called? Answer with only one word." | python3 -m tinygrad.llm --model llama3.2:1b | tee /dev/stderr | grep -i rooster' \
$'echo "What\'s a male chicken called? Answer with only one word." | python3 -m tinygrad.llm --model llama3.2:1b-q4 | tee /dev/stderr | grep -i rooster' \
$'echo "What\'s a male chicken called? Answer with only one word." | python3 -m tinygrad.llm --model qwen3.5:0.8b | tee /dev/stderr | grep -i rooster' \
$'echo "What\'s a female chicken called? Answer with only one word." | python3 -m tinygrad.llm --model qwen3:0.6b | tee /dev/stderr | grep -i hen'
::: $'echo "What\'s a male chicken called? Answer with only one word." | python3 -m tinygrad.llm --no_chat_template --model llama3.2:1b | tee /dev/stderr | grep -i rooster' \
$'echo "What\'s a male chicken called? Answer with only one word." | python3 -m tinygrad.llm --no_chat_template --model llama3.2:1b-q4 | tee /dev/stderr | grep -i rooster' \
$'echo "What\'s a male chicken called? Answer with only one word." | python3 -m tinygrad.llm --no_chat_template --model qwen3.5:0.8b | tee /dev/stderr | grep -i rooster' \
$'echo "What\'s a female chicken called? Answer with only one word." | python3 -m tinygrad.llm --no_chat_template --model qwen3:0.6b | tee /dev/stderr | grep -i hen'
# NOTE: qwen is dumb and only knows about female chickens
# ****** Models Tests ******
testmodels:
name: Models
runs-on: *linux
runs-on: ${{ github.repository == 'tinygrad/tinygrad' && github.event_name == 'pull_request' && github.event.pull_request.author_association == 'COLLABORATOR' && 'namespace-profile-tinygrad' || 'ubuntu-24.04' }}
timeout-minutes: 15
steps:
- name: Checkout Code
@@ -407,7 +420,7 @@ jobs:
testdsp:
name: Linux (DSP)
runs-on: *linux
runs-on: ${{ github.repository == 'tinygrad/tinygrad' && github.event_name == 'pull_request' && github.event.pull_request.author_association == 'COLLABORATOR' && 'namespace-profile-tinygrad' || 'ubuntu-24.04' }}
timeout-minutes: 15
steps:
- name: Checkout Code
@@ -435,7 +448,7 @@ jobs:
- 'WEBGPU'
name: Linux (DEV=${{ matrix.dev }})
runs-on: *linux
runs-on: ${{ github.repository == 'tinygrad/tinygrad' && github.event_name == 'pull_request' && github.event.pull_request.author_association == 'COLLABORATOR' && 'namespace-profile-tinygrad' || 'ubuntu-24.04' }}
timeout-minutes: 20
steps:
- name: Checkout Code
@@ -461,7 +474,7 @@ jobs:
testamdasm:
name: AMD ASM IDE
runs-on: *linux
runs-on: ${{ github.repository == 'tinygrad/tinygrad' && github.event_name == 'pull_request' && github.event.pull_request.author_association == 'COLLABORATOR' && 'namespace-profile-tinygrad' || 'ubuntu-24.04' }}
timeout-minutes: 20
env:
DEV: MOCKKFD+AMD
@@ -491,7 +504,7 @@ jobs:
- name: Run AMD renderer tests (AMD:LLVM)
run: DEV=MOCKKFD+AMD:LLVM python -m pytest -n=auto test/amd/ --durations 20
- name: Run SQTT profiling tests
run: PROFILE=1 SQTT=1 python3 -m pytest -n=auto test/amd/test_sqtt_profiler.py
run: VIZ=-2 python3 -m pytest -n=auto test/amd/test_sqtt_profiler.py
- name: Run AMD emulated tests on NULL backend
env:
AMD: 0
@@ -507,7 +520,7 @@ jobs:
hcq2:
name: hcq2
runs-on: *linux
runs-on: ${{ github.repository == 'tinygrad/tinygrad' && github.event_name == 'pull_request' && github.event.pull_request.author_association == 'COLLABORATOR' && 'namespace-profile-tinygrad' || 'ubuntu-24.04' }}
timeout-minutes: 5
steps:
- name: Checkout Code
@@ -521,16 +534,15 @@ jobs:
- name: Run HCQ2 tests
run: HCQ_RUNTIME_DEV=PYTHON HCQ2=1 DEV=MOCKKFD+AMD FORWARD_ONLY=1 PYTHONPATH=. python test/test_tiny.py
- name: Run HCQ2 multi-device tests
run: |
HCQ_RUNTIME_DEV=PYTHON HCQ2=1 DEV=MOCKKFD+AMD FORWARD_ONLY=1 PYTHONPATH=. python test/unit/test_multitensor.py \
TestMultiTensor.test_simple_add TestMultiTensor.test_shard_reduce \
TestMultiTensor.test_backward_sum TestMultiTensor.test_matmul_shard_0_0
run: HCQ_RUNTIME_DEV=PYTHON HCQ2=1 DEV=MOCKKFD+AMD FORWARD_ONLY=1 PYTHONPATH=. python -m pytest -n=auto test/backend/test_multitensor.py
- name: Run HCQ2 JIT tests
run: HCQ_RUNTIME_DEV=PYTHON HCQ2=1 DEV=MOCKKFD+AMD FORWARD_ONLY=1 PYTHONPATH=. python test/unit/test_jit.py
- name: Run HCQ2 unit tests
run: HCQ_RUNTIME_DEV=PYTHON HCQ2=1 DEV=MOCKKFD+AMD FORWARD_ONLY=1 PYTHONPATH=. python -m pytest test/device/test_hcq2.py
testmockam:
name: Linux (am)
runs-on: *linux
runs-on: ${{ github.repository == 'tinygrad/tinygrad' && github.event_name == 'pull_request' && github.event.pull_request.author_association == 'COLLABORATOR' && 'namespace-profile-tinygrad' || 'ubuntu-24.04' }}
timeout-minutes: 15
env:
DEV: MOCKPCI+AMD
@@ -566,7 +578,7 @@ jobs:
arch: [gfx1100, gfx1201, gfx950]
name: Linux (${{ matrix.backend }} ${{ matrix.arch }})
runs-on: *linux
runs-on: ${{ github.repository == 'tinygrad/tinygrad' && github.event_name == 'pull_request' && github.event.pull_request.author_association == 'COLLABORATOR' && 'namespace-profile-tinygrad' || 'ubuntu-24.04' }}
timeout-minutes: 15
env:
DEV: MOCKKFD+AMD:${{ matrix.backend == 'amdllvm' && 'LLVM' || '' }}:${{ matrix.arch }}
@@ -589,7 +601,7 @@ jobs:
if: ${{ matrix.backend == 'amd' && matrix.arch == 'gfx950' }}
run: PYTHONPATH=. DEV=NULL:HIP:gfx950 MXFP4=1 LLAMA_LAYERS=2 BENCHMARK=3 NULL_ALLOW_COPYOUT=1 NO_HIPCC=1 ROCM_PATH=/opt/rocm JITBEAM=0 examples/mlperf/training_submission_v6.0/tinycorp/benchmarks/llama31_8b/implementations/tinybox_8xMI350X/profile.sh
- name: Run pytest (amd)
run: python -m pytest -n=auto test/backend/test_ops.py test/backend/test_dtype.py test/backend/test_dtype_alu.py test/backend/test_linearizer.py test/backend/test_randomness.py test/backend/test_jit.py test/backend/test_graph.py test/backend/test_multitensor.py test/device/test_hcq.py test/external/external_test_am.py test/backend/test_asm_gemm.py::TestAsmGEMM test/opt/test_tensor_cores.py --durations=20
run: python -m pytest -n=auto test/backend/test_ops.py test/backend/test_dtype.py test/backend/test_dtype_alu.py test/backend/test_linearizer.py test/backend/test_randomness.py test/backend/test_jit.py test/backend/test_graph.py test/backend/test_multitensor.py test/device/test_hcq.py test/external/external_test_am.py test/backend/test_asm_gemm.py::TestAsmGEMM --durations=20
- name: Run disk copy tests
run: python -m pytest test/unit/test_disk_tensor.py -k test_copy_from_disk
- name: Run TRANSCENDENTAL math
@@ -604,7 +616,7 @@ jobs:
backend: [ptx, nv]
name: Linux (${{ matrix.backend }})
runs-on: *linux
runs-on: ${{ github.repository == 'tinygrad/tinygrad' && github.event_name == 'pull_request' && github.event.pull_request.author_association == 'COLLABORATOR' && 'namespace-profile-tinygrad' || 'ubuntu-24.04' }}
timeout-minutes: 20
env:
FORWARD_ONLY: 1
@@ -638,10 +650,17 @@ jobs:
strategy:
fail-fast: false
matrix:
backend: [ir3, nak]
name: Compile-only (${{ matrix.backend }})
runs-on: *linux
dev:
- 'NULL:IR3:a630'
- 'NULL:QCOMCL:a630'
- 'NULL:NAK:sm_120'
name: Compile-only (DEV=${{ matrix.dev }})
runs-on: ${{ github.repository == 'tinygrad/tinygrad' && github.event_name == 'pull_request' && github.event.pull_request.author_association == 'COLLABORATOR' && 'namespace-profile-tinygrad' || 'ubuntu-24.04' }}
timeout-minutes: 15
env:
NULL_ALLOW_COPYOUT: 1
DEV: ${{ matrix.dev }}${{ contains(matrix.dev, 'a630') && ',IMAGE_PITCH_ALIGNMENT=64' || '' }}
IMAGE: ${{ contains(matrix.dev, 'a630') && '1' || '0' }}
steps:
- name: Checkout Code
uses: actions/checkout@v6
@@ -650,21 +669,15 @@ jobs:
with:
key: compile-${{ matrix.backend }}
deps: "testing_unit mesa"
- name: Set env
qemu: ${{ contains(matrix.dev, 'QCOMCL') }}
- name: Test IMAGE
shell: bash
run: printf "NULL_ALLOW_COPYOUT=1\n${{ matrix.backend == 'ir3' && 'DEV=NULL:IR3:a630' || matrix.backend == 'nak' && 'DEV=NULL:NAK:sm_120' }}" >> $GITHUB_ENV
if: contains(matrix.dev, 'a630')
run: DEBUG=7 python3 test/backend/test_ops.py TestOps.test_gemm | grep isam
- name: Run test_ops
shell: bash
run: |
python -c "from tinygrad import Device; assert Device.DEFAULT == 'NULL'"
DEBUG=4 python3 test/backend/test_ops.py TestOps.test_add
python -m pytest -n=auto test/backend/test_ops.py --durations=20
- name: Run test_ops (IMAGE)
if: matrix.backend == 'ir3'
shell: bash
env:
IMAGE: 1
DEV: "NULL:IR3:a630,IMAGE_PITCH_ALIGNMENT=64"
run: |
DEBUG=4 python3 test/backend/test_ops.py TestOps.test_gemm | grep image_load
python -m pytest -n=auto test/backend/test_ops.py --durations=20
# QCOMCL compiles in qemu, too slow for parallel workers
${{ contains(matrix.dev, 'QCOMCL') && 'PARALLEL=0' || '' }} python -m pytest -n=auto test/backend/test_ops.py --durations=20