https://github.com/keroro824/HashingDeepLearning

Codebase for "SLIDE : In Defense of Smart Algorithms over Hardware Acceleration for Large-Scale Deep Learning Systems"
https://github.com/keroro824/HashingDeepLearning

Last synced: 3 months ago
JSON representation

Codebase for "SLIDE : In Defense of Smart Algorithms over Hardware Acceleration for Large-Scale Deep Learning Systems"

• Host: GitHub
• URL: https://github.com/keroro824/HashingDeepLearning
• Owner: keroro824
• License: mit
• Created: 2019-02-06T03:26:18.000Z (over 5 years ago)
• Default Branch: master
• Last Pushed: 2021-04-13T01:32:24.000Z (about 3 years ago)
• Last Synced: 2024-01-16T05:17:09.918Z (5 months ago)
• Language: C++
• Homepage:
• Size: 22 MB
• Stars: 1,060
• Watchers: 53
• Forks: 168
• Open Issues: 24
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Lists

• awesome-stars - keroro824/HashingDeepLearning - Codebase for "SLIDE : In Defense of Smart Algorithms over Hardware Acceleration for Large-Scale Deep Learning Systems" (C++)
• my-awesome-starred - keroro824/HashingDeepLearning - Codebase for "SLIDE : In Defense of Smart Algorithms over Hardware Acceleration for Large-Scale Deep Learning Systems" (C++)

README

        
# SLIDE

The SLIDE package contains the source code for reproducing the main experiments in this [paper](https://arxiv.org/abs/1903.03129).

For Optimized Code on CPUs (with AVX, BFloat and other memory optimization) from the newer [paper](https://proceedings.mlsys.org/paper/2021/file/3636638817772e42b59d74cff571fbb3-Paper.pdf) please refer [here](https://github.com/RUSH-LAB/SLIDE) 

## Dataset

The Datasets can be downloaded in [Amazon-670K](https://drive.google.com/open?id=0B3lPMIHmG6vGdUJwRzltS1dvUVk). Note that the data is sorted by labels so please shuffle at least the validation/testing data.

## TensorFlow Baselines

We suggest directly get TensorFlow docker image to install [TensorFlow-GPU](https://www.tensorflow.org/install/docker).

For TensorFlow-CPU compiled with AVX2, we recommend using this precompiled [build](https://github.com/lakshayg/tensorflow-build).

Also there is a TensorFlow docker image specifically built for CPUs with AVX-512 instructions, to get it use:

```bash

docker pull clearlinux/stacks-dlrs_2-mkl    

```

`config.py` controls the parameters of TensorFlow training like `learning rate`. `example_full_softmax.py, example_sampled_softmax.py` are example files for `Amazon-670K` dataset with full softmax and sampled softmax respectively.

Run

```bash

python python_examples/example_full_softmax.py

python python_examples/example_sampled_softmax.py

```

## Running SLIDE

### Dependencies

- CMake v3.0 and above

- C++11 Compliant compiler

- Linux: Ubuntu 16.04 and newer

- Transparent Huge Pages must be enabled.

  - SLIDE requires approximately 900 2MB pages, and 10 1GB pages: ([Instructions](https://wiki.debian.org/Hugepages))

### Notes:

- For simplicity, please refer to the our [Docker](https://hub.docker.com/repository/docker/ottovonxu/slide) image with all environments installed. To replicate the experiment without setting Hugepages, please download [Amazon-670K](https://drive.google.com/open?id=0B3lPMIHmG6vGdUJwRzltS1dvUVk) in path ```/home/code/HashingDeepLearning/dataset/Amazon``` 

- Also, note that only Skylake or newer architectures support Hugepages. For older Haswell processors, we need to remove the flag `-mavx512f` from the `OPT_FLAGS` line in Makefile. You can also revert to the commit `2d10d46b5f6f1eda5d19f27038a596446fc17cee` to ignore the HugePages optimization and still use SLIDE (which could lead to a 30% slower performance). 

- This version builds all dependencies (which currently are [ZLIB](https://github.com/madler/zlib/tree/v1.2.11) and [CNPY](https://github.com/sarthakpati/cnpy)).

### Commands

Change the paths in ```./SLIDE/Config_amz.csv``` appropriately.

```bash

git clone https://github.com/sarthakpati/HashingDeepLearning.git

cd HashingDeepLearning

mkdir bin

cd bin

cmake ..

make

./runme ../SLIDE/Config_amz.csv

```