tinygrad/.github/workflows/benchmark.yml

name: Benchmarks

on:
  push:
    branches:
      - master
      - update_benchmark

jobs:
  testmacbenchmark:
    name: Mac Benchmark
    runs-on: [self-hosted, macOS]
    env:
      PYTHONPATH: .
    steps:
    - name: Checkout Code
      uses: actions/checkout@v3
    - name: Run model inference benchmark
      run: python3 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
    - name: Run Stable Diffusion
      run: |
        ln -s ~/tinygrad/weights/sd-v1-4.ckpt weights/sd-v1-4.ckpt
        ln -s ~/tinygrad/weights/bpe_simple_vocab_16e6.txt.gz weights/bpe_simple_vocab_16e6.txt.gz
        time python3 examples/stable_diffusion.py --noshow
    - name: Run LLaMA
      run: |
        ln -s ~/tinygrad/weights/LLaMA weights/LLaMA
        python3 examples/llama.py --prompt "Hello." --count 10 --temperature 0 --timing | tee llama_unjitted.txt
        JIT=1 python3 examples/llama.py --prompt "Hello." --count 10 --temperature 0 --timing | tee llama_jitted.txt
    - name: Run 10 CIFAR training steps
      run: |
        ln -s ~/tinygrad/extra/datasets/cifar-10-python.tar.gz extra/datasets/cifar-10-python.tar.gz
        STEPS=10 python3 examples/hlb_cifar10.py | tee train_cifar.txt
    - uses: actions/upload-artifact@v3
      with:
        name: Speed (Mac)
        path: |
          onnx_inference_speed.csv
          torch_speed.txt
          train_cifar.txt
          llama_unjitted.txt
          llama_jitted.txt

  testamdbenchmark:
    name: AMD Benchmark
    runs-on: [self-hosted, Linux]
    env:
      PYTHONPATH: .
    steps:
    - name: Checkout Code
      uses: actions/checkout@v3
    - name: Run model inference benchmark
      run: python3 test/external/external_model_benchmark.py
    - name: Test speed vs torch
      run: BIG=2 TORCHCUDA=1 python3 test/test_speed_v_torch.py | tee torch_speed.txt
    - name: Run Stable Diffusion
      run: |
        ln -s ~/tinygrad/weights/sd-v1-4.ckpt weights/sd-v1-4.ckpt
        ln -s ~/tinygrad/weights/bpe_simple_vocab_16e6.txt.gz weights/bpe_simple_vocab_16e6.txt.gz
        time DEBUG=1 python3 examples/stable_diffusion.py --noshow
    - name: Run LLaMA
      run: |
        ln -s ~/tinygrad/weights/LLaMA weights/LLaMA
        python3 examples/llama.py --prompt "Hello." --count 10 --temperature 0 --timing | tee llama_unjitted.txt
        JIT=1 python3 examples/llama.py --prompt "Hello." --count 10 --temperature 0 --timing | tee llama_jitted.txt
    - name: Run 10 CIFAR training steps
      run: |
        ln -s ~/tinygrad/extra/datasets/cifar-10-python.tar.gz extra/datasets/cifar-10-python.tar.gz
        STEPS=10 python3 examples/hlb_cifar10.py | tee train_cifar.txt
    - uses: actions/upload-artifact@v3
      with:
        name: Speed (AMD)
        path: |
          onnx_inference_speed.csv
          torch_speed.txt
          train_cifar.txt
          llama_unjitted.txt
          llama_jitted.txt