mirror of
https://github.com/tinygrad/tinygrad.git
synced 2026-04-07 03:00:26 -04:00
more process replay tooling (#5407)
* replays * what's in there * can it be up there * sha is enough * insert sha as the key * fix str * update reset utils * that nested try/except was terrible * github_context can go
This commit is contained in:
qazal
31
.github/workflows/benchmark.yml
vendored
31
.github/workflows/benchmark.yml
vendored
@@ -1,4 +1,8 @@
|
||||
name: Benchmarks
|
||||
env:
|
||||
RUN_PROCESS_REPLAY: "1"
|
||||
ASSERT_PROCESS_REPLAY: "0"
|
||||
PYTHONPATH: .
|
||||
|
||||
on:
|
||||
push:
|
||||
@@ -21,8 +25,6 @@ jobs:
|
||||
run:
|
||||
shell: bash -o pipefail {0}
|
||||
if: github.repository_owner == 'tinygrad'
|
||||
env:
|
||||
PYTHONPATH: .
|
||||
steps:
|
||||
- name: Checkout Code
|
||||
uses: actions/checkout@v4
|
||||
@@ -34,9 +36,6 @@ jobs:
|
||||
ln -s ~/tinygrad/weights/bpe_simple_vocab_16e6.txt.gz weights/bpe_simple_vocab_16e6.txt.gz
|
||||
ln -s ~/tinygrad/weights/LLaMA weights/LLaMA
|
||||
ln -s ~/tinygrad/extra/datasets/cifar-10-python.tar.gz extra/datasets/cifar-10-python.tar.gz
|
||||
# - name: Setup process replay
|
||||
# if: github.event.inputs.run_process_replay == 'true' || contains(github.event.head_commit.message, '[run_process_replay]') || contains(github.event.pull_request.title, '[run_process_replay]')
|
||||
# run: echo "RUN_PROCESS_REPLAY=1" >> $GITHUB_ENV
|
||||
- name: Run Stable Diffusion
|
||||
run: JIT=2 python3 examples/stable_diffusion.py --seed 0 --noshow --timing | tee sd.txt
|
||||
- name: Run Stable Diffusion with fp16
|
||||
@@ -87,7 +86,7 @@ jobs:
|
||||
run: JIT=2 WINO=1 STEPS=10 python3 examples/hlb_cifar10.py | tee train_cifar_wino.txt
|
||||
- name: Run process replay tests
|
||||
if: env.RUN_PROCESS_REPLAY == '1'
|
||||
run: cp test/external/replay_codegen.py ./replay_codegen.py && git fetch origin master && git checkout origin/master && PYTHONPATH=. python3 replay_codegen.py
|
||||
run: cp test/external/process_replay/process_replay.py ./process_replay.py && git fetch origin master && git checkout origin/master && PYTHONPATH=. python3 process_replay.py
|
||||
- uses: actions/upload-artifact@v4
|
||||
with:
|
||||
name: Speed (Mac)
|
||||
@@ -122,8 +121,6 @@ jobs:
|
||||
run:
|
||||
shell: bash -o pipefail {0}
|
||||
if: github.repository_owner == 'tinygrad'
|
||||
env:
|
||||
PYTHONPATH: .
|
||||
steps:
|
||||
- name: Checkout Code
|
||||
uses: actions/checkout@v4
|
||||
@@ -193,6 +190,9 @@ jobs:
|
||||
run: NV=1 HALF=1 python3 examples/gpt2.py --count 10 --temperature 0 --timing | tee gpt2_half.txt
|
||||
- name: Run GPT2 w HALF/BEAM
|
||||
run: NV=1 HALF=1 JITBEAM=2 IGNORE_BEAM_CACHE=1 CAST_BEFORE_VIEW=0 python3 examples/gpt2.py --count 10 --temperature 0 --timing | tee gpt2_half_beam.txt
|
||||
- name: Run process replay tests
|
||||
if: env.RUN_PROCESS_REPLAY == '1'
|
||||
run: cp test/external/process_replay/process_replay.py ./process_replay.py && git fetch origin master && git checkout origin/master && PYTHONPATH=. python3 process_replay.py
|
||||
- uses: actions/upload-artifact@v4
|
||||
with:
|
||||
name: Speed (NVIDIA)
|
||||
@@ -227,8 +227,6 @@ jobs:
|
||||
run:
|
||||
shell: bash -o pipefail {0}
|
||||
if: github.repository_owner == 'tinygrad'
|
||||
env:
|
||||
PYTHONPATH: .
|
||||
steps:
|
||||
- name: Checkout Code
|
||||
uses: actions/checkout@v4
|
||||
@@ -262,6 +260,9 @@ jobs:
|
||||
run: NV=1 DEFAULT_FLOAT=HALF BENCHMARK=10 BS=256 GPUS=1 MODEL=resnet python3 examples/mlperf/model_train.py | tee train_resnet_one_gpu.txt
|
||||
- name: Run 10 MLPerf ResNet50 training steps (6 gpu)
|
||||
run: NV=1 DEFAULT_FLOAT=HALF BENCHMARK=10 BS=1536 GPUS=6 MODEL=resnet python3 examples/mlperf/model_train.py | tee train_resnet.txt
|
||||
- name: Run process replay tests
|
||||
if: env.RUN_PROCESS_REPLAY == '1'
|
||||
run: cp test/external/process_replay/process_replay.py ./process_replay.py && git fetch origin master && git checkout origin/master && PYTHONPATH=. python3 process_replay.py
|
||||
- uses: actions/upload-artifact@v4
|
||||
with:
|
||||
name: Speed (NVIDIA Training)
|
||||
@@ -283,8 +284,6 @@ jobs:
|
||||
run:
|
||||
shell: bash -o pipefail {0}
|
||||
if: github.repository_owner == 'tinygrad'
|
||||
env:
|
||||
PYTHONPATH: .
|
||||
steps:
|
||||
- name: Checkout Code
|
||||
uses: actions/checkout@v4
|
||||
@@ -349,6 +348,9 @@ jobs:
|
||||
run: AMD=1 HALF=1 python3 examples/gpt2.py --count 10 --temperature 0 --timing | tee gpt2_half.txt
|
||||
- name: Run GPT2 w HALF/BEAM
|
||||
run: AMD=1 HALF=1 JITBEAM=2 IGNORE_BEAM_CACHE=1 CAST_BEFORE_VIEW=0 python3 examples/gpt2.py --count 10 --temperature 0 --timing | tee gpt2_half_beam.txt
|
||||
- name: Run process replay tests
|
||||
if: env.RUN_PROCESS_REPLAY == '1'
|
||||
run: cp test/external/process_replay/process_replay.py ./process_replay.py && git fetch origin master && git checkout origin/master && PYTHONPATH=. python3 process_replay.py
|
||||
- uses: actions/upload-artifact@v4
|
||||
with:
|
||||
name: Speed (AMD)
|
||||
@@ -381,8 +383,6 @@ jobs:
|
||||
run:
|
||||
shell: bash -o pipefail {0}
|
||||
if: github.repository_owner == 'tinygrad'
|
||||
env:
|
||||
PYTHONPATH: .
|
||||
steps:
|
||||
- name: Checkout Code
|
||||
uses: actions/checkout@v4
|
||||
@@ -416,6 +416,9 @@ jobs:
|
||||
run: AMD=1 DEFAULT_FLOAT=HALF BENCHMARK=10 BS=256 GPUS=1 MODEL=resnet python3 examples/mlperf/model_train.py | tee train_resnet_one_gpu.txt
|
||||
- name: Run 10 MLPerf ResNet50 training steps (6 gpu)
|
||||
run: AMD=1 DEFAULT_FLOAT=HALF BENCHMARK=10 BS=1536 GPUS=6 MODEL=resnet python3 examples/mlperf/model_train.py | tee train_resnet.txt
|
||||
- name: Run process replay tests
|
||||
if: env.RUN_PROCESS_REPLAY == '1'
|
||||
run: cp test/external/process_replay/process_replay.py ./process_replay.py && git fetch origin master && git checkout origin/master && PYTHONPATH=. python3 process_replay.py
|
||||
- uses: actions/upload-artifact@v4
|
||||
with:
|
||||
name: Speed (AMD Training)
|
||||
|
||||
Reference in New Issue
Block a user