Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/DeepRec-AI/HybridBackend
A high-performance framework for training wide-and-deep recommender systems on heterogeneous cluster
https://github.com/DeepRec-AI/HybridBackend
deep-learning gpu hybrid-parallelism parquet recommender-system
Last synced: 6 days ago
JSON representation
A high-performance framework for training wide-and-deep recommender systems on heterogeneous cluster
- Host: GitHub
- URL: https://github.com/DeepRec-AI/HybridBackend
- Owner: DeepRec-AI
- License: apache-2.0
- Created: 2021-12-01T06:51:24.000Z (over 2 years ago)
- Default Branch: main
- Last Pushed: 2023-07-20T11:42:53.000Z (11 months ago)
- Last Synced: 2023-12-19T17:45:37.414Z (6 months ago)
- Topics: deep-learning, gpu, hybrid-parallelism, parquet, recommender-system
- Language: C++
- Homepage:
- Size: 2.87 MB
- Stars: 144
- Watchers: 13
- Forks: 27
- Open Issues: 9
-
Metadata Files:
- Readme: README.md
- Contributing: CONTRIBUTING.md
- License: LICENSE
- Citation: CITATION.cff
Lists
- awesome-stars - DeepRec-AI/HybridBackend - A high-performance framework for training wide-and-deep recommender systems on heterogeneous cluster (C++)
README
# HybridBackend
[](https://github.com/alibaba/HybridBackend/actions/workflows/cibuild.yaml)
[](https://hybridbackend.readthedocs.io/en/latest/?badge=latest)
[](http://makeapullrequest.com)
[](https://opensource.org/licenses/Apache-2.0)
HybridBackend is a high-performance framework for training wide-and-deep
recommender systems on heterogeneous cluster.
## Features
- Memory-efficient loading of categorical data
- GPU-efficient orchestration of embedding layers
- Communication-efficient training and evaluation at scale
- Easy to use with existing AI workflows
## Usage
A minimal example:
```python
import tensorflow as tf
import hybridbackend.tensorflow as hb
ds = hb.data.Dataset.from_parquet(filenames)
ds = ds.batch(batch_size)
# ...
with tf.device('/gpu:0'):
embs = tf.nn.embedding_lookup_sparse(weights, input_ids)
# ...
```
Please see [documentation](https://hybridbackend.readthedocs.io/en/latest/) for
more information.
## Install
### Method 1: Install from PyPI
`pip install {PACKAGE}`
| `{PACKAGE}` | Dependency | Python | CUDA | GLIBC | Data Opt. | Embedding Opt. | Parallelism Opt. |
| ----------------------------------------------------------------------------------------- | ----------------------------------------------------------------------- | ------ | ---- | ------ | --------- | -------------- | ---------------- |
| [hybridbackend-tf115-cu121](https://pypi.org/project/hybridbackend-tf115-cu121/) | [TensorFlow 1.15](https://github.com/NVIDIA/tensorflow) | 3.8 | 12.1 | >=2.31 | ✓ | ✓ | ✓ |
| [hybridbackend-tf115-cu100](https://pypi.org/project/hybridbackend-tf115-cu100/) | [TensorFlow 1.15](https://github.com/tensorflow/tensorflow/tree/r1.15) | 3.6 | 10.0 | >=2.27 | ✓ | ✓ | ✗ |
| [hybridbackend-tf115-cpu](https://pypi.org/project/hybridbackend-tf115-cpu/) | [TensorFlow 1.15](https://github.com/tensorflow/tensorflow/tree/r1.15) | 3.6 | - | >=2.24 | ✓ | ✗ | ✗ |
### Method 2: Build from source
See [Building Instructions](https://github.com/alibaba/HybridBackend/blob/main/BUILD.md).
We also provide built docker images for latest [DeepRec](https://github.com/alibaba/DeepRec):
`registry.cn-shanghai.aliyuncs.com/pai-dlc/hybridbackend:1.0.0-deeprec-py3.6-cu114-ubuntu18.04`
## License
HybridBackend is licensed under the [Apache 2.0 License](LICENSE).
## Community
- Please see [Contributing Guide](https://github.com/alibaba/HybridBackend/blob/main/CONTRIBUTING.md)
before your first contribution.
- Please [register as an adopter](https://github.com/alibaba/HybridBackend/blob/main/ADOPTERS.md)
if your organization is interested in adoption. We will discuss
[RoadMap](https://github.com/alibaba/HybridBackend/blob/main/ROADMAP.md) with
registered adopters in advance.
- Please cite [HybridBackend](https://ieeexplore.ieee.org/document/9835450) in your publications if it helps:
```text
@inproceedings{zhang2022picasso,
title={PICASSO: Unleashing the Potential of GPU-centric Training for Wide-and-deep Recommender Systems},
author={Zhang, Yuanxing and Chen, Langshi and Yang, Siran and Yuan, Man and Yi, Huimin and Zhang, Jie and Wang, Jiamang and Dong, Jianbo and Xu, Yunlong and Song, Yue and others},
booktitle={2022 IEEE 38th International Conference on Data Engineering (ICDE)},
year={2022},
organization={IEEE}
}
```
## Contact Us
If you would like to share your experiences with others, you are welcome to
contact us in DingTalk:
[](https://qr.dingtalk.com/action/joingroup?code=v1,k1,VouhbeuTwXYEgaLzSOE8o6VF2kTHVJ8lw5h93WbZW8o=&_dt_no_comment=1&origin=11)