VideoDataset
A GPU-accelerated library that enables random frame access and efficient video decoding for data loading.
Overview
VideoDataset is a high-performance video decoding multi-framework supporting library. It aims to provide framework-integrated solutions for working with video decoding tasks.
Key Features:
- GPU-accelerated video decoding using NvCodec library
- Support for common video formats (H.264, H.265, etc.)
- Easy integration with multi-frameworks and multi-formats.
Installation
Prerequisites
- NVIDIA GPU with CUDA support and CUDA Toolkit 12.0+ installed
- Python 3.10 or later
Install from PyPI
pip install agibot-videodataset
Building from Source
pip install git+https://github.com/AgiBot-World/VideoDataset.git
Quick Start
The complete example can be found in the quickstart documentation.
Documentation
Please refer to full documentation here.
Also, a sphinx-based documentation can be generated by running the following command:
make dev-doc doc
It will generate the documentation in the docs/_build/html directory and serve it on http://localhost:8000.
Performance
VideoDataset is optimized for high-throughput video processing. Benchmark results show:
- GPU Decoding: A decoding throughput of 20,000 FPS is achieved in a multiprocessing scenario.
- Random Access: Minimal overhead for non-sequential frame access.
- GPU Decoder Utilization: Over 90% GPU decoder utilization is achieved in a multiprocessing scenario.
See the benchmark documentation for detailed performance analysis.
Comparison with other CPU decoding solutions
In addition, we conducted a comprehensive benchmark comparing it against mainstream CPU software decoding solutions, including OpenCV, Torchvision (PyAV), Torchvision (VideoReader), and TorchCodec (CPU).The results demonstrate that VideoDataset achieves a 3 to 4 times improvement in decoding throughput.
Furthermore, it also demonstrates outstanding performance in reducing CPU utilization.
Development Status
- GPU acceleration via NvCodec
- Random frame access
- PyTorch integration
- Compatibility with LeRobot
- Asynchronous pipeline optimization
License
MIT License, for more details, see the LICENSE file.
1 maintainer
Metadata
1 maintainer
Release files for agibot-videodataset 0.1.5
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | |
|---|---|---|---|
| agibot_videodataset-0.1.5.tar.gz | 249.5 kB | Details |
Built distributions (wheels)
Total release size: 482.0 MB
Release files / agibot_videodataset-0.1.5.tar.gz
| Download URL | agibot_videodataset-0.1.5.tar.gz |
|---|---|
| Size | 249.5 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
f1f79263a42bc346b18228f940752d7e836d303376d4718f5adc8392a2cdd133
|
|
BLAKE2b-256 checksum How to use checksums |
de2323e5f8976643c9ebb59ae887bae5ae8995a3e5b8670ae4aa6ffcd9887f62
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.7
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Jan 8, 2026.
Transparency logRelease files / agibot_videodataset-0.1.5-cp314-cp314t-manylinux_2_28_x86_64.whl
| Download URL | agibot_videodataset-0.1.5-cp314-cp314t-manylinux_2_28_x86_64.whl |
|---|---|
| Size | 41.7 MB |
| Tags | CPython 3.14 CPython 3.14 free-threading Linux glibc 2.28+ x86-64 |
|
SHA-256 checksum How to use checksums |
cf7a04c21378aab738ed8fc2912c6f0960a4eab7e6b5ec69d2fb795273e80902
|
|
BLAKE2b-256 checksum How to use checksums |
4bbb9065bdd373abcbda08c22292584198b13bafd0661780b807cf570b2c9d50
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.7
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Jan 8, 2026.
Transparency logRelease files / agibot_videodataset-0.1.5-cp314-cp314t-manylinux_2_28_aarch64.whl
| Download URL | agibot_videodataset-0.1.5-cp314-cp314t-manylinux_2_28_aarch64.whl |
|---|---|
| Size | 38.5 MB |
| Tags | CPython 3.14 CPython 3.14 free-threading Linux glibc 2.28+ ARM64 |
|
SHA-256 checksum How to use checksums |
0a4f01deb6570e13634d47deb7738142f33438a01266776105427859e38e3c94
|
|
BLAKE2b-256 checksum How to use checksums |
44a83e0796bd7ff6921e0c471d12071dccc59cc4950aee9a549a0af8f25a01e7
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.7
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Jan 8, 2026.
Transparency logRelease files / agibot_videodataset-0.1.5-cp314-cp314-manylinux_2_28_x86_64.whl
| Download URL | agibot_videodataset-0.1.5-cp314-cp314-manylinux_2_28_x86_64.whl |
|---|---|
| Size | 41.7 MB |
| Tags | CPython 3.14 Linux glibc 2.28+ x86-64 |
|
SHA-256 checksum How to use checksums |
82807f755cbb04a14f0e798b70772a089cdbbf9b75681ee0432548671e7b9435
|
|
BLAKE2b-256 checksum How to use checksums |
90b0383ee205cde2266c071ff8de52863e1a96c39175d7ed161cdacbe900c961
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.7
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Jan 8, 2026.
Transparency logRelease files / agibot_videodataset-0.1.5-cp314-cp314-manylinux_2_28_aarch64.whl
| Download URL | agibot_videodataset-0.1.5-cp314-cp314-manylinux_2_28_aarch64.whl |
|---|---|
| Size | 38.6 MB |
| Tags | CPython 3.14 Linux glibc 2.28+ ARM64 |
|
SHA-256 checksum How to use checksums |
56ae4209f3f63fc7d50296bbd990487016edf43718842acfbc82bbdd7efd787c
|
|
BLAKE2b-256 checksum How to use checksums |
bda92c545eb149318ae86d2a681e0551bebe63352a492a9233e5ce27ae30ffb9
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.7
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Jan 8, 2026.
Transparency logRelease files / agibot_videodataset-0.1.5-cp313-cp313-manylinux_2_28_x86_64.whl
| Download URL | agibot_videodataset-0.1.5-cp313-cp313-manylinux_2_28_x86_64.whl |
|---|---|
| Size | 41.7 MB |
| Tags | CPython 3.13 Linux glibc 2.28+ x86-64 |
|
SHA-256 checksum How to use checksums |
0b00ac555d8c24bda9fe4c99150bbd1a8d1002d1bafe1ea7b479b39e155fe037
|
|
BLAKE2b-256 checksum How to use checksums |
83d4340b5f5c21055fbb3e49e6efafb517d2c83bf19c6c9be7d19193e338154a
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.7
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Jan 8, 2026.
Transparency logRelease files / agibot_videodataset-0.1.5-cp313-cp313-manylinux_2_28_aarch64.whl
| Download URL | agibot_videodataset-0.1.5-cp313-cp313-manylinux_2_28_aarch64.whl |
|---|---|
| Size | 38.6 MB |
| Tags | CPython 3.13 Linux glibc 2.28+ ARM64 |
|
SHA-256 checksum How to use checksums |
eab00df982ffb6bb286fce7932be931e91b6385c64313b1c0479313754d1de94
|
|
BLAKE2b-256 checksum How to use checksums |
56237377286a0419a7b7531a46bc9f61ba72563c73c87753f27e5830d38a88d6
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.7
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Jan 8, 2026.
Transparency logRelease files / agibot_videodataset-0.1.5-cp312-cp312-manylinux_2_28_x86_64.whl
| Download URL | agibot_videodataset-0.1.5-cp312-cp312-manylinux_2_28_x86_64.whl |
|---|---|
| Size | 41.7 MB |
| Tags | CPython 3.12 Linux glibc 2.28+ x86-64 |
|
SHA-256 checksum How to use checksums |
ea8a8a5b3c2d104d70a6f530fc6a9622e4b7e0822ff3cb7f78b56ac44e94731b
|
|
BLAKE2b-256 checksum How to use checksums |
cf2edf5063b0d652e91061ee634c2ea5c11983b9c21ffa9ef45adebeab6adb54
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.7
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Jan 8, 2026.
Transparency logRelease files / agibot_videodataset-0.1.5-cp312-cp312-manylinux_2_28_aarch64.whl
| Download URL | agibot_videodataset-0.1.5-cp312-cp312-manylinux_2_28_aarch64.whl |
|---|---|
| Size | 38.5 MB |
| Tags | CPython 3.12 Linux glibc 2.28+ ARM64 |
|
SHA-256 checksum How to use checksums |
4253e354c53567317d5e9a4a4a3815b1697b4b99139a6283a8d21f2514c3302d
|
|
BLAKE2b-256 checksum How to use checksums |
1e9c74a1a5862263f821f2135e12f20e1a30b1adad8a4d30c4fe60b1ff2cea09
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.7
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Jan 8, 2026.
Transparency logRelease files / agibot_videodataset-0.1.5-cp311-cp311-manylinux_2_28_x86_64.whl
| Download URL | agibot_videodataset-0.1.5-cp311-cp311-manylinux_2_28_x86_64.whl |
|---|---|
| Size | 41.7 MB |
| Tags | CPython 3.11 Linux glibc 2.28+ x86-64 |
|
SHA-256 checksum How to use checksums |
85994de74fd3a3cdcdf48cd62943022d34cfe30ed91401447ed4417ef4bb8468
|
|
BLAKE2b-256 checksum How to use checksums |
49a6fee126efb6076e102b5e71f0efce94b0cc36c5aefaa186679aa34ebe28ac
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.7
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Jan 8, 2026.
Transparency logRelease files / agibot_videodataset-0.1.5-cp311-cp311-manylinux_2_28_aarch64.whl
| Download URL | agibot_videodataset-0.1.5-cp311-cp311-manylinux_2_28_aarch64.whl |
|---|---|
| Size | 38.5 MB |
| Tags | CPython 3.11 Linux glibc 2.28+ ARM64 |
|
SHA-256 checksum How to use checksums |
0aa44a076ced368a9a8fed00ece4da2cb9eebd0795589e10fa30e6f2fca7ef08
|
|
BLAKE2b-256 checksum How to use checksums |
e0331db431db90aa5151f4b5f53f6483fc592864ef9ea5c7e99624c8f087bfed
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.7
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Jan 8, 2026.
Transparency logRelease files / agibot_videodataset-0.1.5-cp310-cp310-manylinux_2_28_x86_64.whl
| Download URL | agibot_videodataset-0.1.5-cp310-cp310-manylinux_2_28_x86_64.whl |
|---|---|
| Size | 41.7 MB |
| Tags | CPython 3.10 Linux glibc 2.28+ x86-64 |
|
SHA-256 checksum How to use checksums |
394ad01cd01571af369fc545134bdfb6ec0bcf11b45ac38f0530304e733c34da
|
|
BLAKE2b-256 checksum How to use checksums |
762196d9f7cb5cb5a24375b62c5d62ae5671dcfa42c9bbe0d09548710079ff2a
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.7
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Jan 8, 2026.
Transparency logRelease files / agibot_videodataset-0.1.5-cp310-cp310-manylinux_2_28_aarch64.whl
| Download URL | agibot_videodataset-0.1.5-cp310-cp310-manylinux_2_28_aarch64.whl |
|---|---|
| Size | 38.5 MB |
| Tags | CPython 3.10 Linux glibc 2.28+ ARM64 |
|
SHA-256 checksum How to use checksums |
c59d4f4c6d84b9481b4c07f2ea708214687b5e8af1e214a6fc4772f159433973
|
|
BLAKE2b-256 checksum How to use checksums |
fe1caa892bf2925f9c6b42fad146930a4a740b71f7547f2396fbf79e1c002c98
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/6.1.0 CPython/3.13.7
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Jan 8, 2026.
Transparency log
