Rate this Page

★ ★ ★ ★ ★

recipes/torch_export_aoti_python

Run in Google Colab

Note

Go to the end to download the full example code.

`torch.export` AOTInductor Tutorial for Python runtime (Beta)#

Created On: Aug 23, 2024 | Last Updated: Jan 24, 2025 | Last Verified: Nov 05, 2024

Author: Ankith Gunapal, Bin Bao, Angela Yi

Warning

torch._inductor.aoti_compile_and_package and torch._inductor.aoti_load_package are in Beta status and are subject to backwards compatibility breaking changes. This tutorial provides an example of how to use these APIs for model deployment using Python runtime.

It has been shown previously how AOTInductor can be used to do Ahead-of-Time compilation of PyTorch exported models by creating an artifact that can be run in a non-Python environment. In this tutorial, you will learn an end-to-end example of how to use AOTInductor for Python runtime.

Contents

Prerequisites #

PyTorch 2.6 or later
Basic understanding of torch.export and AOTInductor
Complete the AOTInductor: Ahead-Of-Time Compilation for Torch.Export-ed Models tutorial

What you will learn #

How to use AOTInductor for Python runtime.
How to use torch._inductor.aoti_compile_and_package() along with torch.export.export() to generate a compiled artifact
How to load and run the artifact in a Python runtime using torch._export.aot_load().
When to you use AOTInductor with a Python runtime

Model Compilation #

We will use the TorchVision pretrained ResNet18 model as an example.

The first step is to export the model to a graph representation using torch.export.export(). To learn more about using this function, you can check out the docs or the tutorial.

Once we have exported the PyTorch model and obtained an ExportedProgram, we can apply torch._inductor.aoti_compile_and_package() to AOTInductor to compile the program to a specified device, and save the generated contents into a “.pt2” artifact.

Note

This API supports the same available options that torch.compile() has, such as mode and max_autotune (for those who want to enable CUDA graphs and leverage Triton based matrix multiplications and convolutions)

import os
import torch
import torch._inductor
from torchvision.models import ResNet18_Weights, resnet18

model = resnet18(weights=ResNet18_Weights.DEFAULT)
model.eval()

with torch.inference_mode():
    inductor_configs = {}

    if torch.cuda.is_available():
        device = "cuda"
        inductor_configs["max_autotune"] = True
    else:
        device = "cpu"

    model = model.to(device=device)
    example_inputs = (torch.randn(2, 3, 224, 224, device=device),)

    exported_program = torch.export.export(
        model,
        example_inputs,
    )
    path = torch._inductor.aoti_compile_and_package(
        exported_program,
        package_path=os.path.join(os.getcwd(), "resnet18.pt2"),
        inductor_configs=inductor_configs
    )

Downloading: "https://download.pytorch.org/models/resnet18-f37072fd.pth" to /var/lib/ci-user/.cache/torch/hub/checkpoints/resnet18-f37072fd.pth

  0%|          | 0.00/44.7M [00:00<?, ?B/s]
 90%|█████████ | 40.4M/44.7M [00:00<00:00, 423MB/s]
100%|██████████| 44.7M/44.7M [00:00<00:00, 424MB/s]
/usr/local/lib/python3.10/dist-packages/torch/backends/cuda/__init__.py:131: UserWarning:

Please use the new API settings to control TF32 behavior, such as torch.backends.cudnn.conv.fp32_precision = 'tf32' or torch.backends.cuda.matmul.fp32_precision = 'ieee'. Old settings, e.g, torch.backends.cuda.matmul.allow_tf32 = True, torch.backends.cudnn.allow_tf32 = True, allowTF32CuDNN() and allowTF32CuBLAS() will be deprecated after Pytorch 2.9. Please see https://pytorch.org/docs/main/notes/cuda.html#tensorfloat-32-tf32-on-ampere-and-later-devices (Triggered internally at /pytorch/aten/src/ATen/Context.cpp:80.)

/usr/local/lib/python3.10/dist-packages/torch/_inductor/compile_fx.py:312: UserWarning:

TensorFloat32 tensor cores for float32 matrix multiplication available but not enabled. Consider setting `torch.set_float32_matmul_precision('high')` for better performance.

Autotune Choices Stats:
{"num_choices": 7, "num_triton_choices": 6, "best_kernel": "convolution", "best_time": 0.08806400001049042, "best_triton_pos": 1, "best_triton_time": 0.1157120019197464, "best_triton_kernel": "triton_convolution2d_0", "best_triton_kernel_desc": "ALLOW_TF32=True, BLOCK_K=16, BLOCK_M=64, BLOCK_N=64, GROUPS=1, KERNEL_H=7, KERNEL_W=7, PADDING_H=3, PADDING_W=3, STRIDE_H=2, STRIDE_W=2, UNROLL=False, num_stages=2, num_warps=4"}
AUTOTUNE convolution(2x3x224x224, 64x3x7x7)
strides: [150528, 1, 672, 3], [147, 1, 21, 3]
dtypes: torch.float32, torch.float32
  convolution 0.0881 ms 100.0%
  triton_convolution2d_0 0.1157 ms 76.1% ALLOW_TF32=True, BLOCK_K=16, BLOCK_M=64, BLOCK_N=64, GROUPS=1, KERNEL_H=7, KERNEL_W=7, PADDING_H=3, PADDING_W=3, STRIDE_H=2, STRIDE_W=2, UNROLL=False, num_stages=2, num_warps=4
  triton_convolution2d_4 0.1536 ms 57.3% ALLOW_TF32=True, BLOCK_K=16, BLOCK_M=64, BLOCK_N=64, GROUPS=1, KERNEL_H=7, KERNEL_W=7, PADDING_H=3, PADDING_W=3, STRIDE_H=2, STRIDE_W=2, UNROLL=False, num_stages=2, num_warps=8
  triton_convolution2d_3 0.1812 ms 48.6% ALLOW_TF32=True, BLOCK_K=16, BLOCK_M=128, BLOCK_N=64, GROUPS=1, KERNEL_H=7, KERNEL_W=7, PADDING_H=3, PADDING_W=3, STRIDE_H=2, STRIDE_W=2, UNROLL=False, num_stages=2, num_warps=8
  triton_convolution2d_5 0.2191 ms 40.2% ALLOW_TF32=True, BLOCK_K=16, BLOCK_M=256, BLOCK_N=64, GROUPS=1, KERNEL_H=7, KERNEL_W=7, PADDING_H=3, PADDING_W=3, STRIDE_H=2, STRIDE_W=2, UNROLL=False, num_stages=2, num_warps=8
  triton_convolution2d_2 0.3000 ms 29.4% ALLOW_TF32=True, BLOCK_K=16, BLOCK_M=1024, BLOCK_N=16, GROUPS=1, KERNEL_H=7, KERNEL_W=7, PADDING_H=3, PADDING_W=3, STRIDE_H=2, STRIDE_W=2, UNROLL=False, num_stages=1, num_warps=8
  triton_convolution2d_1 0.5233 ms 16.8% ALLOW_TF32=True, BLOCK_K=16, BLOCK_M=256, BLOCK_N=64, GROUPS=1, KERNEL_H=7, KERNEL_W=7, PADDING_H=3, PADDING_W=3, STRIDE_H=2, STRIDE_W=2, UNROLL=False, num_stages=2, num_warps=4
SingleProcess AUTOTUNE benchmarking takes 0.3344 seconds and 0.0012 seconds precompiling for 7 choices
Autotune Choices Stats:
{"num_choices": 8, "num_triton_choices": 7, "best_kernel": "triton_convolution2d_10", "best_kernel_desc": "ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=64, BLOCK_N=64, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=1, STRIDE_W=1, UNROLL=False, num_stages=2, num_warps=4", "best_time": 0.03788800165057182, "best_triton_pos": 0}
AUTOTUNE convolution(2x64x56x56, 64x64x3x3)
strides: [200704, 1, 3584, 64], [576, 1, 192, 64]
dtypes: torch.float32, torch.float32
  triton_convolution2d_10 0.0379 ms 100.0% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=64, BLOCK_N=64, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=1, STRIDE_W=1, UNROLL=False, num_stages=2, num_warps=4
  triton_convolution2d_6 0.0389 ms 97.4% ALLOW_TF32=True, BLOCK_K=16, BLOCK_M=64, BLOCK_N=64, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=1, STRIDE_W=1, UNROLL=False, num_stages=2, num_warps=4
  triton_convolution2d_11 0.0389 ms 97.4% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=64, BLOCK_N=64, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=1, STRIDE_W=1, UNROLL=False, num_stages=2, num_warps=8
  triton_convolution2d_9 0.0430 ms 88.1% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=128, BLOCK_N=64, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=1, STRIDE_W=1, UNROLL=False, num_stages=2, num_warps=8
  triton_convolution2d_7 0.0594 ms 63.8% ALLOW_TF32=True, BLOCK_K=16, BLOCK_M=256, BLOCK_N=64, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=1, STRIDE_W=1, UNROLL=False, num_stages=2, num_warps=4
  triton_convolution2d_12 0.0614 ms 61.7% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=256, BLOCK_N=64, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=1, STRIDE_W=1, UNROLL=False, num_stages=2, num_warps=8
  convolution 0.0799 ms 47.4%
  triton_convolution2d_8 0.1178 ms 32.2% ALLOW_TF32=True, BLOCK_K=16, BLOCK_M=1024, BLOCK_N=16, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=1, STRIDE_W=1, UNROLL=False, num_stages=1, num_warps=8
SingleProcess AUTOTUNE benchmarking takes 0.1029 seconds and 0.0008 seconds precompiling for 8 choices
Autotune Choices Stats:
{"num_choices": 8, "num_triton_choices": 7, "best_kernel": "triton_convolution2d_38", "best_kernel_desc": "ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=64, BLOCK_N=64, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=2, STRIDE_W=2, UNROLL=False, num_stages=2, num_warps=4", "best_time": 0.035840000957250595, "best_triton_pos": 0}
AUTOTUNE convolution(2x64x56x56, 128x64x3x3)
strides: [200704, 1, 3584, 64], [576, 1, 192, 64]
dtypes: torch.float32, torch.float32
  triton_convolution2d_38 0.0358 ms 100.0% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=64, BLOCK_N=64, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=2, STRIDE_W=2, UNROLL=False, num_stages=2, num_warps=4
  triton_convolution2d_39 0.0420 ms 85.4% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=64, BLOCK_N=128, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=2, STRIDE_W=2, UNROLL=False, num_stages=2, num_warps=8
  triton_convolution2d_34 0.0492 ms 72.9% ALLOW_TF32=True, BLOCK_K=16, BLOCK_M=64, BLOCK_N=128, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=2, STRIDE_W=2, UNROLL=False, num_stages=2, num_warps=4
  triton_convolution2d_40 0.0604 ms 59.3% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=256, BLOCK_N=64, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=2, STRIDE_W=2, UNROLL=False, num_stages=2, num_warps=8
  triton_convolution2d_35 0.0614 ms 58.3% ALLOW_TF32=True, BLOCK_K=16, BLOCK_M=256, BLOCK_N=64, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=2, STRIDE_W=2, UNROLL=False, num_stages=2, num_warps=4
  triton_convolution2d_37 0.0655 ms 54.7% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=128, BLOCK_N=128, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=2, STRIDE_W=2, UNROLL=False, num_stages=2, num_warps=8
  convolution 0.0737 ms 48.6%
  triton_convolution2d_36 0.1147 ms 31.3% ALLOW_TF32=True, BLOCK_K=16, BLOCK_M=1024, BLOCK_N=16, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=2, STRIDE_W=2, UNROLL=False, num_stages=1, num_warps=8
SingleProcess AUTOTUNE benchmarking takes 0.1212 seconds and 0.0002 seconds precompiling for 8 choices
Autotune Choices Stats:
{"num_choices": 8, "num_triton_choices": 7, "best_kernel": "triton_convolution2d_45", "best_kernel_desc": "ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=64, BLOCK_N=64, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=1, STRIDE_W=1, UNROLL=False, num_stages=2, num_warps=4", "best_time": 0.048128001391887665, "best_triton_pos": 0}
AUTOTUNE convolution(2x128x28x28, 128x128x3x3)
strides: [100352, 1, 3584, 128], [1152, 1, 384, 128]
dtypes: torch.float32, torch.float32
  triton_convolution2d_45 0.0481 ms 100.0% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=64, BLOCK_N=64, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=1, STRIDE_W=1, UNROLL=False, num_stages=2, num_warps=4
  triton_convolution2d_46 0.0758 ms 63.5% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=64, BLOCK_N=128, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=1, STRIDE_W=1, UNROLL=False, num_stages=2, num_warps=8
  convolution 0.0788 ms 61.0%
  triton_convolution2d_41 0.0881 ms 54.7% ALLOW_TF32=True, BLOCK_K=16, BLOCK_M=64, BLOCK_N=128, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=1, STRIDE_W=1, UNROLL=False, num_stages=2, num_warps=4
  triton_convolution2d_42 0.1075 ms 44.8% ALLOW_TF32=True, BLOCK_K=16, BLOCK_M=256, BLOCK_N=64, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=1, STRIDE_W=1, UNROLL=False, num_stages=2, num_warps=4
  triton_convolution2d_47 0.1126 ms 42.7% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=256, BLOCK_N=64, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=1, STRIDE_W=1, UNROLL=False, num_stages=2, num_warps=8
  triton_convolution2d_44 0.1208 ms 39.8% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=128, BLOCK_N=128, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=1, STRIDE_W=1, UNROLL=False, num_stages=2, num_warps=8
  triton_convolution2d_43 0.2212 ms 21.8% ALLOW_TF32=True, BLOCK_K=16, BLOCK_M=1024, BLOCK_N=16, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=1, STRIDE_W=1, UNROLL=False, num_stages=1, num_warps=8
SingleProcess AUTOTUNE benchmarking takes 0.1570 seconds and 0.0002 seconds precompiling for 8 choices
Autotune Choices Stats:
{"num_choices": 8, "num_triton_choices": 7, "best_kernel": "triton_convolution2d_49", "best_kernel_desc": "ALLOW_TF32=True, BLOCK_K=16, BLOCK_M=256, BLOCK_N=64, GROUPS=1, KERNEL_H=1, KERNEL_W=1, PADDING_H=0, PADDING_W=0, STRIDE_H=2, STRIDE_W=2, UNROLL=True, num_stages=2, num_warps=4", "best_time": 0.023552000522613525, "best_triton_pos": 0}
AUTOTUNE convolution(2x64x56x56, 128x64x1x1)
strides: [200704, 1, 3584, 64], [64, 1, 1, 1]
dtypes: torch.float32, torch.float32
  triton_convolution2d_49 0.0236 ms 100.0% ALLOW_TF32=True, BLOCK_K=16, BLOCK_M=256, BLOCK_N=64, GROUPS=1, KERNEL_H=1, KERNEL_W=1, PADDING_H=0, PADDING_W=0, STRIDE_H=2, STRIDE_W=2, UNROLL=True, num_stages=2, num_warps=4
  triton_convolution2d_51 0.0236 ms 100.0% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=128, BLOCK_N=128, GROUPS=1, KERNEL_H=1, KERNEL_W=1, PADDING_H=0, PADDING_W=0, STRIDE_H=2, STRIDE_W=2, UNROLL=True, num_stages=2, num_warps=8
  triton_convolution2d_52 0.0236 ms 100.0% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=64, BLOCK_N=64, GROUPS=1, KERNEL_H=1, KERNEL_W=1, PADDING_H=0, PADDING_W=0, STRIDE_H=2, STRIDE_W=2, UNROLL=True, num_stages=2, num_warps=4
  triton_convolution2d_53 0.0236 ms 100.0% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=64, BLOCK_N=128, GROUPS=1, KERNEL_H=1, KERNEL_W=1, PADDING_H=0, PADDING_W=0, STRIDE_H=2, STRIDE_W=2, UNROLL=True, num_stages=2, num_warps=8
  triton_convolution2d_54 0.0236 ms 100.0% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=256, BLOCK_N=64, GROUPS=1, KERNEL_H=1, KERNEL_W=1, PADDING_H=0, PADDING_W=0, STRIDE_H=2, STRIDE_W=2, UNROLL=True, num_stages=2, num_warps=8
  triton_convolution2d_48 0.0246 ms 95.8% ALLOW_TF32=True, BLOCK_K=16, BLOCK_M=64, BLOCK_N=128, GROUPS=1, KERNEL_H=1, KERNEL_W=1, PADDING_H=0, PADDING_W=0, STRIDE_H=2, STRIDE_W=2, UNROLL=True, num_stages=2, num_warps=4
  triton_convolution2d_50 0.0256 ms 92.0% ALLOW_TF32=True, BLOCK_K=16, BLOCK_M=1024, BLOCK_N=16, GROUPS=1, KERNEL_H=1, KERNEL_W=1, PADDING_H=0, PADDING_W=0, STRIDE_H=2, STRIDE_W=2, UNROLL=True, num_stages=1, num_warps=8
  convolution 0.0573 ms 41.1%
SingleProcess AUTOTUNE benchmarking takes 0.1035 seconds and 0.0002 seconds precompiling for 8 choices
Autotune Choices Stats:
{"num_choices": 8, "num_triton_choices": 7, "best_kernel": "triton_convolution2d_73", "best_kernel_desc": "ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=64, BLOCK_N=64, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=2, STRIDE_W=2, UNROLL=False, num_stages=2, num_warps=4", "best_time": 0.048128001391887665, "best_triton_pos": 0}
AUTOTUNE convolution(2x128x28x28, 256x128x3x3)
strides: [100352, 1, 3584, 128], [1152, 1, 384, 128]
dtypes: torch.float32, torch.float32
  triton_convolution2d_73 0.0481 ms 100.0% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=64, BLOCK_N=64, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=2, STRIDE_W=2, UNROLL=False, num_stages=2, num_warps=4
  convolution 0.0829 ms 58.0%
  triton_convolution2d_75 0.1147 ms 42.0% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=256, BLOCK_N=64, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=2, STRIDE_W=2, UNROLL=False, num_stages=2, num_warps=8
  triton_convolution2d_70 0.1178 ms 40.9% ALLOW_TF32=True, BLOCK_K=16, BLOCK_M=256, BLOCK_N=64, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=2, STRIDE_W=2, UNROLL=False, num_stages=2, num_warps=4
  triton_convolution2d_72 0.1239 ms 38.8% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=128, BLOCK_N=128, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=2, STRIDE_W=2, UNROLL=False, num_stages=2, num_warps=8
  triton_convolution2d_74 0.1290 ms 37.3% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=64, BLOCK_N=256, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=2, STRIDE_W=2, UNROLL=False, num_stages=2, num_warps=8
  triton_convolution2d_69 0.1321 ms 36.4% ALLOW_TF32=True, BLOCK_K=16, BLOCK_M=64, BLOCK_N=256, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=2, STRIDE_W=2, UNROLL=False, num_stages=2, num_warps=4
  triton_convolution2d_71 0.1792 ms 26.9% ALLOW_TF32=True, BLOCK_K=16, BLOCK_M=1024, BLOCK_N=16, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=2, STRIDE_W=2, UNROLL=False, num_stages=1, num_warps=8
SingleProcess AUTOTUNE benchmarking takes 0.1732 seconds and 0.0002 seconds precompiling for 8 choices
Autotune Choices Stats:
{"num_choices": 8, "num_triton_choices": 7, "best_kernel": "convolution", "best_time": 0.08601599931716919, "best_triton_pos": 1, "best_triton_time": 0.09011200070381165, "best_triton_kernel": "triton_convolution2d_80", "best_triton_kernel_desc": "ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=64, BLOCK_N=64, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=1, STRIDE_W=1, UNROLL=False, num_stages=2, num_warps=4"}
AUTOTUNE convolution(2x256x14x14, 256x256x3x3)
strides: [50176, 1, 3584, 256], [2304, 1, 768, 256]
dtypes: torch.float32, torch.float32
  convolution 0.0860 ms 100.0%
  triton_convolution2d_80 0.0901 ms 95.5% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=64, BLOCK_N=64, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=1, STRIDE_W=1, UNROLL=False, num_stages=2, num_warps=4
  triton_convolution2d_77 0.2099 ms 41.0% ALLOW_TF32=True, BLOCK_K=16, BLOCK_M=256, BLOCK_N=64, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=1, STRIDE_W=1, UNROLL=False, num_stages=2, num_warps=4
  triton_convolution2d_78 0.2150 ms 40.0% ALLOW_TF32=True, BLOCK_K=16, BLOCK_M=512, BLOCK_N=16, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=1, STRIDE_W=1, UNROLL=False, num_stages=1, num_warps=8
  triton_convolution2d_82 0.2202 ms 39.1% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=256, BLOCK_N=64, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=1, STRIDE_W=1, UNROLL=False, num_stages=2, num_warps=8
  triton_convolution2d_79 0.2324 ms 37.0% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=128, BLOCK_N=128, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=1, STRIDE_W=1, UNROLL=False, num_stages=2, num_warps=8
  triton_convolution2d_81 0.2509 ms 34.3% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=64, BLOCK_N=256, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=1, STRIDE_W=1, UNROLL=False, num_stages=2, num_warps=8
  triton_convolution2d_76 0.2601 ms 33.1% ALLOW_TF32=True, BLOCK_K=16, BLOCK_M=64, BLOCK_N=256, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=1, STRIDE_W=1, UNROLL=False, num_stages=2, num_warps=4
SingleProcess AUTOTUNE benchmarking takes 0.2313 seconds and 0.0002 seconds precompiling for 8 choices
Autotune Choices Stats:
{"num_choices": 8, "num_triton_choices": 7, "best_kernel": "triton_convolution2d_87", "best_kernel_desc": "ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=64, BLOCK_N=64, GROUPS=1, KERNEL_H=1, KERNEL_W=1, PADDING_H=0, PADDING_W=0, STRIDE_H=2, STRIDE_W=2, UNROLL=True, num_stages=2, num_warps=4", "best_time": 0.023552000522613525, "best_triton_pos": 0}
AUTOTUNE convolution(2x128x28x28, 256x128x1x1)
strides: [100352, 1, 3584, 128], [128, 1, 1, 1]
dtypes: torch.float32, torch.float32
  triton_convolution2d_87 0.0236 ms 100.0% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=64, BLOCK_N=64, GROUPS=1, KERNEL_H=1, KERNEL_W=1, PADDING_H=0, PADDING_W=0, STRIDE_H=2, STRIDE_W=2, UNROLL=True, num_stages=2, num_warps=4
  triton_convolution2d_84 0.0256 ms 92.0% ALLOW_TF32=True, BLOCK_K=16, BLOCK_M=256, BLOCK_N=64, GROUPS=1, KERNEL_H=1, KERNEL_W=1, PADDING_H=0, PADDING_W=0, STRIDE_H=2, STRIDE_W=2, UNROLL=True, num_stages=2, num_warps=4
  triton_convolution2d_86 0.0266 ms 88.5% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=128, BLOCK_N=128, GROUPS=1, KERNEL_H=1, KERNEL_W=1, PADDING_H=0, PADDING_W=0, STRIDE_H=2, STRIDE_W=2, UNROLL=True, num_stages=2, num_warps=8
  triton_convolution2d_83 0.0276 ms 85.2% ALLOW_TF32=True, BLOCK_K=16, BLOCK_M=64, BLOCK_N=256, GROUPS=1, KERNEL_H=1, KERNEL_W=1, PADDING_H=0, PADDING_W=0, STRIDE_H=2, STRIDE_W=2, UNROLL=True, num_stages=2, num_warps=4
  triton_convolution2d_88 0.0276 ms 85.2% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=64, BLOCK_N=256, GROUPS=1, KERNEL_H=1, KERNEL_W=1, PADDING_H=0, PADDING_W=0, STRIDE_H=2, STRIDE_W=2, UNROLL=True, num_stages=2, num_warps=8
  triton_convolution2d_89 0.0276 ms 85.2% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=256, BLOCK_N=64, GROUPS=1, KERNEL_H=1, KERNEL_W=1, PADDING_H=0, PADDING_W=0, STRIDE_H=2, STRIDE_W=2, UNROLL=True, num_stages=2, num_warps=8
  triton_convolution2d_85 0.0338 ms 69.7% ALLOW_TF32=True, BLOCK_K=16, BLOCK_M=1024, BLOCK_N=16, GROUPS=1, KERNEL_H=1, KERNEL_W=1, PADDING_H=0, PADDING_W=0, STRIDE_H=2, STRIDE_W=2, UNROLL=True, num_stages=1, num_warps=8
  convolution 0.0614 ms 38.3%
SingleProcess AUTOTUNE benchmarking takes 0.1026 seconds and 0.0002 seconds precompiling for 8 choices
Autotune Choices Stats:
{"num_choices": 8, "num_triton_choices": 7, "best_kernel": "convolution", "best_time": 0.09011200070381165, "best_triton_pos": 1, "best_triton_time": 0.09318400174379349, "best_triton_kernel": "triton_convolution2d_108", "best_triton_kernel_desc": "ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=64, BLOCK_N=64, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=2, STRIDE_W=2, UNROLL=False, num_stages=2, num_warps=4"}
AUTOTUNE convolution(2x256x14x14, 512x256x3x3)
strides: [50176, 1, 3584, 256], [2304, 1, 768, 256]
dtypes: torch.float32, torch.float32
  convolution 0.0901 ms 100.0%
  triton_convolution2d_108 0.0932 ms 96.7% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=64, BLOCK_N=64, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=2, STRIDE_W=2, UNROLL=False, num_stages=2, num_warps=4
  triton_convolution2d_106 0.2099 ms 42.9% ALLOW_TF32=True, BLOCK_K=16, BLOCK_M=512, BLOCK_N=16, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=2, STRIDE_W=2, UNROLL=False, num_stages=1, num_warps=8
  triton_convolution2d_110 0.2232 ms 40.4% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=256, BLOCK_N=64, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=2, STRIDE_W=2, UNROLL=False, num_stages=2, num_warps=8
  triton_convolution2d_105 0.2355 ms 38.3% ALLOW_TF32=True, BLOCK_K=16, BLOCK_M=256, BLOCK_N=64, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=2, STRIDE_W=2, UNROLL=False, num_stages=2, num_warps=4
  triton_convolution2d_107 0.2396 ms 37.6% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=128, BLOCK_N=128, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=2, STRIDE_W=2, UNROLL=False, num_stages=2, num_warps=8
  triton_convolution2d_109 0.2529 ms 35.6% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=64, BLOCK_N=256, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=2, STRIDE_W=2, UNROLL=False, num_stages=2, num_warps=8
  triton_convolution2d_104 0.2611 ms 34.5% ALLOW_TF32=True, BLOCK_K=16, BLOCK_M=64, BLOCK_N=256, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=2, STRIDE_W=2, UNROLL=False, num_stages=2, num_warps=4
SingleProcess AUTOTUNE benchmarking takes 0.2318 seconds and 0.0002 seconds precompiling for 8 choices
Autotune Choices Stats:
{"num_choices": 8, "num_triton_choices": 7, "best_kernel": "convolution", "best_time": 0.10035199671983719, "best_triton_pos": 1, "best_triton_time": 0.17612800002098083, "best_triton_kernel": "triton_convolution2d_115", "best_triton_kernel_desc": "ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=64, BLOCK_N=64, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=1, STRIDE_W=1, UNROLL=False, num_stages=2, num_warps=4"}
AUTOTUNE convolution(2x512x7x7, 512x512x3x3)
strides: [25088, 1, 3584, 512], [4608, 1, 1536, 512]
dtypes: torch.float32, torch.float32
  convolution 0.1004 ms 100.0%
  triton_convolution2d_115 0.1761 ms 57.0% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=64, BLOCK_N=64, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=1, STRIDE_W=1, UNROLL=False, num_stages=2, num_warps=4
  triton_convolution2d_113 0.2161 ms 46.4% ALLOW_TF32=True, BLOCK_K=16, BLOCK_M=128, BLOCK_N=16, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=1, STRIDE_W=1, UNROLL=False, num_stages=1, num_warps=8
  triton_convolution2d_117 0.2806 ms 35.8% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=128, BLOCK_N=64, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=1, STRIDE_W=1, UNROLL=False, num_stages=2, num_warps=8
  triton_convolution2d_112 0.3512 ms 28.6% ALLOW_TF32=True, BLOCK_K=16, BLOCK_M=128, BLOCK_N=64, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=1, STRIDE_W=1, UNROLL=False, num_stages=2, num_warps=4
  triton_convolution2d_114 0.4588 ms 21.9% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=128, BLOCK_N=128, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=1, STRIDE_W=1, UNROLL=False, num_stages=2, num_warps=8
  triton_convolution2d_116 0.4946 ms 20.3% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=64, BLOCK_N=256, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=1, STRIDE_W=1, UNROLL=False, num_stages=2, num_warps=8
  triton_convolution2d_111 0.5161 ms 19.4% ALLOW_TF32=True, BLOCK_K=16, BLOCK_M=64, BLOCK_N=256, GROUPS=1, KERNEL_H=3, KERNEL_W=3, PADDING_H=1, PADDING_W=1, STRIDE_H=1, STRIDE_W=1, UNROLL=False, num_stages=2, num_warps=4
SingleProcess AUTOTUNE benchmarking takes 0.2456 seconds and 0.0002 seconds precompiling for 8 choices
Autotune Choices Stats:
{"num_choices": 8, "num_triton_choices": 7, "best_kernel": "triton_convolution2d_122", "best_kernel_desc": "ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=64, BLOCK_N=64, GROUPS=1, KERNEL_H=1, KERNEL_W=1, PADDING_H=0, PADDING_W=0, STRIDE_H=2, STRIDE_W=2, UNROLL=True, num_stages=2, num_warps=4", "best_time": 0.023552000522613525, "best_triton_pos": 0}
AUTOTUNE convolution(2x256x14x14, 512x256x1x1)
strides: [50176, 1, 3584, 256], [256, 1, 1, 1]
dtypes: torch.float32, torch.float32
  triton_convolution2d_122 0.0236 ms 100.0% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=64, BLOCK_N=64, GROUPS=1, KERNEL_H=1, KERNEL_W=1, PADDING_H=0, PADDING_W=0, STRIDE_H=2, STRIDE_W=2, UNROLL=True, num_stages=2, num_warps=4
  triton_convolution2d_120 0.0338 ms 69.7% ALLOW_TF32=True, BLOCK_K=16, BLOCK_M=512, BLOCK_N=16, GROUPS=1, KERNEL_H=1, KERNEL_W=1, PADDING_H=0, PADDING_W=0, STRIDE_H=2, STRIDE_W=2, UNROLL=True, num_stages=1, num_warps=8
  triton_convolution2d_119 0.0358 ms 65.7% ALLOW_TF32=True, BLOCK_K=16, BLOCK_M=256, BLOCK_N=64, GROUPS=1, KERNEL_H=1, KERNEL_W=1, PADDING_H=0, PADDING_W=0, STRIDE_H=2, STRIDE_W=2, UNROLL=True, num_stages=2, num_warps=4
  triton_convolution2d_121 0.0369 ms 63.9% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=128, BLOCK_N=128, GROUPS=1, KERNEL_H=1, KERNEL_W=1, PADDING_H=0, PADDING_W=0, STRIDE_H=2, STRIDE_W=2, UNROLL=True, num_stages=2, num_warps=8
  triton_convolution2d_118 0.0379 ms 62.2% ALLOW_TF32=True, BLOCK_K=16, BLOCK_M=64, BLOCK_N=256, GROUPS=1, KERNEL_H=1, KERNEL_W=1, PADDING_H=0, PADDING_W=0, STRIDE_H=2, STRIDE_W=2, UNROLL=True, num_stages=2, num_warps=4
  triton_convolution2d_124 0.0379 ms 62.2% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=256, BLOCK_N=64, GROUPS=1, KERNEL_H=1, KERNEL_W=1, PADDING_H=0, PADDING_W=0, STRIDE_H=2, STRIDE_W=2, UNROLL=True, num_stages=2, num_warps=8
  triton_convolution2d_123 0.0389 ms 60.5% ALLOW_TF32=True, BLOCK_K=32, BLOCK_M=64, BLOCK_N=256, GROUPS=1, KERNEL_H=1, KERNEL_W=1, PADDING_H=0, PADDING_W=0, STRIDE_H=2, STRIDE_W=2, UNROLL=True, num_stages=2, num_warps=8
  convolution 0.0748 ms 31.5%
SingleProcess AUTOTUNE benchmarking takes 0.1045 seconds and 0.0002 seconds precompiling for 8 choices
Autotune Choices Stats:
{"num_choices": 13, "num_triton_choices": 12, "best_kernel": "triton_mm_142", "best_kernel_desc": "ACC_TYPE='tl.float32', ALLOW_TF32=False, BLOCK_K=32, BLOCK_M=16, BLOCK_N=32, EVEN_K=True, GROUP_M=8, USE_FAST_ACCUM=False, num_stages=5, num_warps=2", "best_time": 0.025599999353289604, "best_triton_pos": 0}
AUTOTUNE addmm(2x1000, 2x512, 512x1000)
strides: [0, 1], [512, 1], [1, 512]
dtypes: torch.float32, torch.float32, torch.float32
  triton_mm_142 0.0256 ms 100.0% ACC_TYPE='tl.float32', ALLOW_TF32=False, BLOCK_K=32, BLOCK_M=16, BLOCK_N=32, EVEN_K=True, GROUP_M=8, USE_FAST_ACCUM=False, num_stages=5, num_warps=2
  triton_mm_140 0.0297 ms 86.2% ACC_TYPE='tl.float32', ALLOW_TF32=False, BLOCK_K=128, BLOCK_M=16, BLOCK_N=32, EVEN_K=True, GROUP_M=8, USE_FAST_ACCUM=False, num_stages=2, num_warps=2
  triton_mm_141 0.0389 ms 65.8% ACC_TYPE='tl.float32', ALLOW_TF32=False, BLOCK_K=32, BLOCK_M=16, BLOCK_N=64, EVEN_K=True, GROUP_M=8, USE_FAST_ACCUM=False, num_stages=5, num_warps=4
  triton_mm_145 0.0389 ms 65.8% ACC_TYPE='tl.float32', ALLOW_TF32=False, BLOCK_K=64, BLOCK_M=16, BLOCK_N=64, EVEN_K=True, GROUP_M=8, USE_FAST_ACCUM=False, num_stages=3, num_warps=4
  triton_mm_148 0.0389 ms 65.8% ACC_TYPE='tl.float32', ALLOW_TF32=False, BLOCK_K=32, BLOCK_M=16, BLOCK_N=64, EVEN_K=True, GROUP_M=8, USE_FAST_ACCUM=False, num_stages=3, num_warps=4
  triton_mm_149 0.0389 ms 65.8% ACC_TYPE='tl.float32', ALLOW_TF32=False, BLOCK_K=32, BLOCK_M=16, BLOCK_N=64, EVEN_K=True, GROUP_M=8, USE_FAST_ACCUM=False, num_stages=4, num_warps=4
  triton_mm_139 0.0399 ms 64.1% ACC_TYPE='tl.float32', ALLOW_TF32=False, BLOCK_K=16, BLOCK_M=16, BLOCK_N=32, EVEN_K=True, GROUP_M=8, USE_FAST_ACCUM=False, num_stages=1, num_warps=2
  triton_mm_144 0.0399 ms 64.1% ACC_TYPE='tl.float32', ALLOW_TF32=False, BLOCK_K=32, BLOCK_M=16, BLOCK_N=64, EVEN_K=True, GROUP_M=8, USE_FAST_ACCUM=False, num_stages=2, num_warps=4
  addmm 0.0430 ms 59.5%
  triton_mm_143 0.0450 ms 56.9% ACC_TYPE='tl.float32', ALLOW_TF32=False, BLOCK_K=16, BLOCK_M=16, BLOCK_N=64, EVEN_K=True, GROUP_M=8, USE_FAST_ACCUM=False, num_stages=2, num_warps=4
SingleProcess AUTOTUNE benchmarking takes 0.3680 seconds and 0.0002 seconds precompiling for 13 choices

The result of aoti_compile_and_package() is an artifact “resnet18.pt2” which can be loaded and executed in Python and C++.

The artifact itself contains a bunch of AOTInductor generated code, such as a generated C++ runner file, a shared library compiled from the C++ file, and CUDA binary files, aka cubin files, if optimizing for CUDA.

Structure-wise, the artifact is a structured .zip file, with the following specification:

We can use the following command to inspect the artifact contents:

$ unzip -l resnet18.pt2

Archive:  resnet18.pt2
  Length      Date    Time    Name
---------  ---------- -----   ----
        1  01-08-2025 16:40   version
        3  01-08-2025 16:40   archive_format
    10088  01-08-2025 16:40   data/aotinductor/model/cagzt6akdaczvxwtbvqe34otfe5jlorktbqlojbzqjqvbfsjlge4.cubin
    17160  01-08-2025 16:40   data/aotinductor/model/c6oytfjmt5w4c7onvtm6fray7clirxt7q5xjbwx3hdydclmwoujz.cubin
    16616  01-08-2025 16:40   data/aotinductor/model/c7ydp7nocyz323hij4tmlf2kcedmwlyg6r57gaqzcsy3huneamu6.cubin
    17776  01-08-2025 16:40   data/aotinductor/model/cyqdf46ordevqhiddvpdpp3uzwatfbzdpl3auj2nx23uxvplnne2.cubin
    10856  01-08-2025 16:40   data/aotinductor/model/cpzfebfgrusqslui7fxsuoo4tvwulmrxirc5tmrpa4mvrbdno7kn.cubin
    14608  01-08-2025 16:40   data/aotinductor/model/c5ukeoz5wmaszd7vczdz2qhtt6n7tdbl3b6wuy4rb2se24fjwfoy.cubin
    11376  01-08-2025 16:40   data/aotinductor/model/csu3nstcp56tsjfycygaqsewpu64l5s6zavvz7537cm4s4cv2k3r.cubin
    10984  01-08-2025 16:40   data/aotinductor/model/cp76lez4glmgq7gedf2u25zvvv6rksv5lav4q22dibd2zicbgwj3.cubin
    14736  01-08-2025 16:40   data/aotinductor/model/c2bb5p6tnwz4elgujqelsrp3unvkgsyiv7xqxmpvuxcm4jfl7pc2.cubin
    11376  01-08-2025 16:40   data/aotinductor/model/c6eopmb2b4ngodwsayae4r5q6ni3jlfogfbdk3ypg56tgpzhubfy.cubin
    11624  01-08-2025 16:40   data/aotinductor/model/chmwe6lvoekzfowdbiizitm3haiiuad5kdm6sd2m6mv6dkn2zk32.cubin
    15632  01-08-2025 16:40   data/aotinductor/model/c3jop5g344hj3ztsu4qm6ibxyaaerlhkzh2e6emak23rxfje6jam.cubin
    25472  01-08-2025 16:40   data/aotinductor/model/chaiixybeiuuitm2nmqnxzijzwgnn2n7uuss4qmsupgblfh3h5hk.cubin
   139389  01-08-2025 16:40   data/aotinductor/model/cvk6qzuybruhwxtfblzxiov3rlrziv5fkqc4mdhbmantfu3lmd6t.cpp
       27  01-08-2025 16:40   data/aotinductor/model/cvk6qzuybruhwxtfblzxiov3rlrziv5fkqc4mdhbmantfu3lmd6t_metadata.json
 47195424  01-08-2025 16:40   data/aotinductor/model/cvk6qzuybruhwxtfblzxiov3rlrziv5fkqc4mdhbmantfu3lmd6t.so
---------                     -------
 47523148                     18 files

Model Inference in Python #

To load and run the artifact in Python, we can use torch._inductor.aoti_load_package().

import os
import torch
import torch._inductor

model_path = os.path.join(os.getcwd(), "resnet18.pt2")

compiled_model = torch._inductor.aoti_load_package(model_path)
example_inputs = (torch.randn(2, 3, 224, 224, device=device),)

with torch.inference_mode():
    output = compiled_model(example_inputs)

When to use AOTInductor with a Python Runtime #

There are mainly two reasons why one would use AOTInductor with a Python Runtime:

torch._inductor.aoti_compile_and_package generates a singular serialized artifact. This is useful for model versioning for deployments and tracking model performance over time.
With torch.compile() being a JIT compiler, there is a warmup cost associated with the first compilation. Your deployment needs to account for the compilation time taken for the first inference. With AOTInductor, the compilation is done ahead of time using torch.export.export and torch._inductor.aoti_compile_and_package. At deployment time, after loading the model, running inference does not have any additional cost.

The section below shows the speedup achieved with AOTInductor for first inference

We define a utility function timed to measure the time taken for inference

import time
def timed(fn):
    # Returns the result of running `fn()` and the time it took for `fn()` to run,
    # in seconds. We use CUDA events and synchronization for accurate
    # measurement on CUDA enabled devices.
    if torch.cuda.is_available():
        start = torch.cuda.Event(enable_timing=True)
        end = torch.cuda.Event(enable_timing=True)
        start.record()
    else:
        start = time.time()

    result = fn()
    if torch.cuda.is_available():
        end.record()
        torch.cuda.synchronize()
    else:
        end = time.time()

    # Measure time taken to execute the function in miliseconds
    if torch.cuda.is_available():
        duration = start.elapsed_time(end)
    else:
        duration = (end - start) * 1000

    return result, duration

Lets measure the time for first inference using AOTInductor

torch._dynamo.reset()

model = torch._inductor.aoti_load_package(model_path)
example_inputs = (torch.randn(1, 3, 224, 224, device=device),)

with torch.inference_mode():
    _, time_taken = timed(lambda: model(example_inputs))
    print(f"Time taken for first inference for AOTInductor is {time_taken:.2f} ms")

Time taken for first inference for AOTInductor is 3.46 ms

Lets measure the time for first inference using torch.compile

torch._dynamo.reset()

model = resnet18(weights=ResNet18_Weights.DEFAULT).to(device)
model.eval()

model = torch.compile(model)
example_inputs = torch.randn(1, 3, 224, 224, device=device)

with torch.inference_mode():
    _, time_taken = timed(lambda: model(example_inputs))
    print(f"Time taken for first inference for torch.compile is {time_taken:.2f} ms")

Time taken for first inference for torch.compile is 3778.10 ms

We see that there is a drastic speedup in first inference time using AOTInductor compared to torch.compile

Conclusion #

In this recipe, we have learned how to effectively use the AOTInductor for Python runtime by compiling and loading a pretrained ResNet18 model. This process demonstrates the practical application of generating a compiled artifact and running it within a Python environment. We also looked at the advantage of using AOTInductor in model deployments, with regards to speed up in first inference time.

Total running time of the script: (0 minutes 28.184 seconds)

`torch.export` AOTInductor Tutorial for Python runtime (Beta)#

Prerequisites #

What you will learn #

Model Compilation #

Model Inference in Python #

When to use AOTInductor with a Python Runtime #

Conclusion #

Docs

Tutorials

Resources

torch.export AOTInductor Tutorial for Python runtime (Beta)#

Prerequisites#

What you will learn#

Model Compilation#

Model Inference in Python#

When to use AOTInductor with a Python Runtime#

Conclusion#

Docs

Tutorials

Resources

`torch.export` AOTInductor Tutorial for Python runtime (Beta)#

Prerequisites #

What you will learn #

Model Compilation #

Model Inference in Python #

When to use AOTInductor with a Python Runtime #

Conclusion #