github/tinygrad
1
1
Fork 0
mirror of https://github.com/tinygrad/tinygrad.git synced 2026-04-29 03:00:14 -04:00
Code Issues Packages Projects Releases Wiki Activity

hotfix: mac uses python 3.12

Browse Source
This commit is contained in:
George Hotz
2024-11-11 23:23:48 +08:00
parent aaa8059aec
commit 9648372ee6
1 changed files with 26 additions and 26 deletions
Download Patch File Download Diff File Expand all files Collapse all files

52
.github/workflows/benchmark.yml vendored
View File

@@ -47,55 +47,55 @@ jobs:
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
run: python3.12 test/external/process_replay/reset.py
- name: Run Stable Diffusion
run: JIT=2 python3 examples/stable_diffusion.py --seed 0 --noshow --timing | tee sd.txt
run: JIT=2 python3.12 examples/stable_diffusion.py --seed 0 --noshow --timing | tee sd.txt
- name: Run Stable Diffusion with fp16
run: JIT=2 python3 examples/stable_diffusion.py --fp16 --seed 0 --noshow --timing | tee sd_fp16.txt
run: JIT=2 python3.12 examples/stable_diffusion.py --fp16 --seed 0 --noshow --timing | tee sd_fp16.txt
- name: Run SDXL
run: JIT=2 python3 examples/sdxl.py --seed 0 --noshow --timing | tee sdxl.txt
run: JIT=2 python3.12 examples/sdxl.py --seed 0 --noshow --timing | tee sdxl.txt
- name: Run model inference benchmark
run: METAL=1 python3 test/external/external_model_benchmark.py
run: METAL=1 python3.12 test/external/external_model_benchmark.py
- name: Test speed vs torch
run: BIG=2 MPS=1 python3 test/test_speed_v_torch.py | tee torch_speed.txt
run: BIG=2 MPS=1 python3.12 test/test_speed_v_torch.py | tee torch_speed.txt
- name: Test tensor cores
run: METAL=1 python3 test/test_linearizer.py TestLinearizer.test_tensor_cores TestLinearizer.test_tensor_cores_padded
run: METAL=1 python3.12 test/test_linearizer.py TestLinearizer.test_tensor_cores TestLinearizer.test_tensor_cores_padded
- name: Run Tensor Core GEMM
run: |
DEBUG=2 python3 extra/gemm/simple_matmul.py | tee matmul.txt
DEBUG=2 HALF=1 python3 extra/gemm/simple_matmul.py | tee matmul_half.txt
DEBUG=2 python3.12 extra/gemm/simple_matmul.py | tee matmul.txt
DEBUG=2 HALF=1 python3.12 extra/gemm/simple_matmul.py | tee matmul_half.txt
- name: Fuzz Padded Tensor Core GEMM
run: METAL=1 M_START=6 M_STOP=10 M_STEP=1 N_START=6 N_STOP=10 N_STEP=1 K_START=6 K_STOP=24 K_STEP=1 TC_OPT=2 DEBUG=2 python3 ./extra/gemm/fuzz_matmul.py
run: METAL=1 M_START=6 M_STOP=10 M_STEP=1 N_START=6 N_STOP=10 N_STEP=1 K_START=6 K_STOP=24 K_STEP=1 TC_OPT=2 DEBUG=2 python3.12 ./extra/gemm/fuzz_matmul.py
- name: Run LLaMA
run: |
JIT=0 python3 examples/llama.py --gen 1 --prompt "Hello." --count 10 --temperature 0 --timing | tee llama_unjitted.txt
JIT=1 python3 examples/llama.py --gen 1 --prompt "Hello." --count 10 --temperature 0 --timing | tee llama_jitted.txt
JIT=0 python3.12 examples/llama.py --gen 1 --prompt "Hello." --count 10 --temperature 0 --timing | tee llama_unjitted.txt
JIT=1 python3.12 examples/llama.py --gen 1 --prompt "Hello." --count 10 --temperature 0 --timing | tee llama_jitted.txt
- name: Run LLaMA with BEAM
run: JITBEAM=2 IGNORE_BEAM_CACHE=1 python3 examples/llama.py --gen 1 --prompt "Hello." --count 10 --temperature 0 --timing | tee llama_beam.txt
run: JITBEAM=2 IGNORE_BEAM_CACHE=1 python3.12 examples/llama.py --gen 1 --prompt "Hello." --count 10 --temperature 0 --timing | tee llama_beam.txt
- name: Run quantized LLaMA
run: |
python3 examples/llama.py --gen 1 --prompt "Hello." --count 10 --temperature 0 --timing --quantize int8 | tee llama_int8.txt
python3 examples/llama.py --gen 1 --prompt "Hello." --count 10 --temperature 0 --timing --quantize nf4 | tee llama_nf4.txt
python3.12 examples/llama.py --gen 1 --prompt "Hello." --count 10 --temperature 0 --timing --quantize int8 | tee llama_int8.txt
python3.12 examples/llama.py --gen 1 --prompt "Hello." --count 10 --temperature 0 --timing --quantize nf4 | tee llama_nf4.txt
- name: Run LLaMA 7B on 4 (virtual) GPUs
run: python3 examples/llama.py --gen 1 --size 7B --shard 4 --prompt "Hello." --count 10 --temperature 0 --timing | tee llama_four_gpu.txt
run: python3.12 examples/llama.py --gen 1 --size 7B --shard 4 --prompt "Hello." --count 10 --temperature 0 --timing | tee llama_four_gpu.txt
- name: Run GPT2
run: |
JIT=0 python3 examples/gpt2.py --prompt "Hello." --count 10 --temperature 0 --timing | tee gpt2_unjitted.txt
JIT=1 python3 examples/gpt2.py --prompt "Hello." --count 10 --temperature 0 --timing | tee gpt2_jitted.txt
JIT=0 python3.12 examples/gpt2.py --prompt "Hello." --count 10 --temperature 0 --timing | tee gpt2_unjitted.txt
JIT=1 python3.12 examples/gpt2.py --prompt "Hello." --count 10 --temperature 0 --timing | tee gpt2_jitted.txt
- name: Run GPT2 w HALF
run: HALF=1 python3 examples/gpt2.py --count 10 --temperature 0 --timing | tee gpt2_half.txt
run: HALF=1 python3.12 examples/gpt2.py --count 10 --temperature 0 --timing | tee gpt2_half.txt
- name: Run GPT2 w HALF/BEAM
run: 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
run: HALF=1 JITBEAM=2 IGNORE_BEAM_CACHE=1 CAST_BEFORE_VIEW=0 python3.12 examples/gpt2.py --count 10 --temperature 0 --timing | tee gpt2_half_beam.txt
- name: Train MNIST
run: time PYTHONPATH=. TARGET_EVAL_ACC_PCT=96.0 python3 examples/beautiful_mnist.py | tee beautiful_mnist.txt
run: time PYTHONPATH=. TARGET_EVAL_ACC_PCT=96.0 python3.12 examples/beautiful_mnist.py | tee beautiful_mnist.txt
- name: Run 10 CIFAR training steps
run: JIT=2 STEPS=10 python3 examples/hlb_cifar10.py | tee train_cifar.txt
run: JIT=2 STEPS=10 python3.12 examples/hlb_cifar10.py | tee train_cifar.txt
- name: Run 10 CIFAR training steps w HALF
run: JIT=2 STEPS=10 DEFAULT_FLOAT=HALF python3 examples/hlb_cifar10.py | tee train_cifar_half.txt
run: JIT=2 STEPS=10 DEFAULT_FLOAT=HALF python3.12 examples/hlb_cifar10.py | tee train_cifar_half.txt
#- name: Run 10 CIFAR training steps w BF16
# run: STEPS=10 DEFAULT_FLOAT=BFLOAT16 python3 examples/hlb_cifar10.py | tee train_cifar_bf16.txt
# run: STEPS=10 DEFAULT_FLOAT=BFLOAT16 python3.12 examples/hlb_cifar10.py | tee train_cifar_bf16.txt
- name: Run 10 CIFAR training steps w winograd
run: JIT=2 WINO=1 STEPS=10 python3 examples/hlb_cifar10.py | tee train_cifar_wino.txt
run: JIT=2 WINO=1 STEPS=10 python3.12 examples/hlb_cifar10.py | tee train_cifar_wino.txt
- uses: actions/upload-artifact@v4
with:
name: Speed (Mac)
@@ -123,7 +123,7 @@ jobs:
train_cifar_bf16.txt
train_cifar_wino.txt
- name: Run process replay tests
run: 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
run: cp test/external/process_replay/process_replay.py ./process_replay.py && git fetch origin master && git -c advice.detachedHead=false checkout origin/master && PYTHONPATH=. python3.12 process_replay.py
testnvidiabenchmark:
name: tinybox green Benchmark