Initial commit

Megatron-Energon/pyproject.toml

+[build-system]
+requires = ["hatchling", "hatch-vcs"]
+build-backend = "hatchling.build"
+
+[project]
+name = "megatron-energon"
+dynamic = ["version"]
+authors = [
+  { name="Lukas Vögtle", email="lvoegtle@nvidia.com" },
+  { name="Philipp Fischer", email="pfischer@nvidia.com" },
+]
+description = "Megatron's multi-modal data loader"
+readme = "README.md"
+license = "BSD-3-Clause"
+requires-python = ">=3.10"
+classifiers = [
+  "Programming Language :: Python :: 3",
+  "Programming Language :: Python :: 3.10",
+  "Programming Language :: Python :: 3.11",
+  "Programming Language :: Python :: 3.12",
+  "Operating System :: OS Independent",
+]
+dependencies = [
+  "braceexpand",
+  "click",
+  "dataslots; python_version<'3.10'",
+  "mfusepy",
+  "multi-storage-client>=0.18.0,<0.26.0",
+  "numpy",
+  "pillow>=10.0.1",  # WEBP vulnerability fixed starting from 10.0.1
+  "pyyaml",
+  "rapidyaml>=0.10.0",
+  "s3fs",
+  "torch",
+  "tqdm",
+  "webdataset",
+]
+
+[project.optional-dependencies]
+dev = [
+  "ruff",
+  "sphinxcontrib-napoleon",
+  "sphinx",
+  "myst-parser",
+  "soundfile",
+  "sphinx-rtd-theme",
+  "sphinx-click",
+]
+transforms = [
+  "torchvision",  # Needed for megatron.energon.transforms
+]
+# Storage services for MSC
+s3 = [
+  "multi-storage-client[boto3]",
+]
+aistore = [
+  "multi-storage-client[aistore]",
+]
+azure-storage-blob = [
+  "multi-storage-client[azure-storage-blob]",
+]
+google-cloud-storage = [
+  "multi-storage-client[google-cloud-storage]",
+]
+oci = [
+  "multi-storage-client[oci]",
+]
+
+# Dependencies for video decoding
+av_decode = [
+  # needed for efficient audio and video file decoding
+  "bitstring>=4.2.3",
+  "sortedcontainers>=2.4.0",
+  "filetype>=1.2.0",
+  "ebmlite>=3.3.1",
+  "av>=14.4.0",
+]
+# If using guess_content=True for decoding
+guess_content = [
+  "filetype>=1.0.0",
+]
+
+[project.urls]
+Homepage = "https://github.com/NVIDIA/Megatron-Energon"
+
+[tool.hatch.build.targets.wheel]
+packages = ["src/megatron"]
+
+[tool.hatch.build.targets.sdist]
+packages = ["src/megatron"]
+
+[project.scripts]
+energon = "megatron.energon.cli.main:main"
+
+[tool.hatch.version]
+source = "vcs"
+
+[tool.ruff.lint]
+extend-select = ["I"]
+ignore = ["E741", "E731"]
+
+[tool.ruff]
+line-length = 100
+target-version = "py310"
+include = ["**/*.py", "**/*.pyi"]
+exclude = [
+  ".idea",
+  "docs",
+]
+

Megatron-Energon/src/megatron/energon/av/fastseek/__init__.py

+# Copyright (c) 2025, NVIDIA CORPORATION.
+# SPDX-License-Identifier: BSD-3-Clause
+from .fastseek import Fastseek as Fastseek
+from .keyframeinfo import KeyframeInfo as KeyframeInfo

Megatron-Energon/src/megatron/energon/av/fastseek/containers/__init__.py

+# Copyright (c) 2025, NVIDIA CORPORATION.
+# SPDX-License-Identifier: BSD-3-Clause