Jing Qiu 3108aa614d
ultralytics 8.4.38 Unify args naming for standalone export functions (#24120)
Signed-off-by: Jing Qiu <61612323+Laughing-q@users.noreply.github.com>
Signed-off-by: Onuralp SEZER <onuralp@ultralytics.com>
Signed-off-by: Glenn Jocher <glenn.jocher@ultralytics.com>
Co-authored-by: Ultralytics Assistant <135830346+UltralyticsAssistant@users.noreply.github.com>
Co-authored-by: Onuralp SEZER <onuralp@ultralytics.com>
Co-authored-by: UltralyticsAssistant <web@ultralytics.com>
Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
Co-authored-by: Lakshantha Dissanayake <lakshantha@ultralytics.com>
2026-04-16 14:00:28 +02:00

47 lines
1.6 KiB
Python

# Ultralytics 🚀 AGPL-3.0 License - https://ultralytics.com/license
from __future__ import annotations
import json
from pathlib import Path
import torch
from ultralytics.utils import LOGGER, TORCH_VERSION
def torch2torchscript(
model: torch.nn.Module,
im: torch.Tensor,
output_file: Path | str,
optimize: bool = False,
metadata: dict | None = None,
prefix: str = "",
) -> str:
"""Export a PyTorch model to TorchScript format.
Args:
model (torch.nn.Module): The PyTorch model to export (may be NMS-wrapped).
im (torch.Tensor): Example input tensor for tracing.
output_file (Path | str): Path to save the exported TorchScript model.
optimize (bool): Whether to optimize for mobile deployment.
metadata (dict | None): Optional metadata to embed in the TorchScript archive.
prefix (str): Prefix for log messages.
Returns:
(str): Path to the exported ``.torchscript`` file.
"""
LOGGER.info(f"\n{prefix} starting export with torch {TORCH_VERSION}...")
output_file = str(output_file)
ts = torch.jit.trace(model, im, strict=False)
extra_files = {"config.txt": json.dumps(metadata or {})} # torch._C.ExtraFilesMap()
if optimize: # https://pytorch.org/tutorials/recipes/mobile_interpreter.html
LOGGER.info(f"{prefix} optimizing for mobile...")
from torch.utils.mobile_optimizer import optimize_for_mobile
optimize_for_mobile(ts)._save_for_lite_interpreter(output_file, _extra_files=extra_files)
else:
ts.save(output_file, _extra_files=extra_files)
return output_file