The projects tests the performance of MXNet and Tensorflow using MNIST data.
├── LICENSE
├── README.md <- Top-level README for developers using this project
├── mnist_mx_gluon.py <- MXNet + Gluon
├── mnist_mx_gluon_mgpu.py <- MXNet + Gluon + multi-gpu
├── mnist_mx_gluon_hvd.py <- MXNet + Gluon + Horovod
├── mnist_mx_keras.py <- MXNet + Keras
├── mnist_mx_keras_mgpu.py <- MXNet + Keras + multi-gpu (multi_gpu_model())
├── mnist_mx_sym.py <- MXNet + symbol/module API
├── mnist_mx_sym_mgpu.py <- MXNet + symbol/module API + multi-gpu
├── mnist_tf_keras.py <- Tensorflow + Keras
└── mnist_tf_keras_mirrored.py <- Tensorflow + Keras + multi-gpu (MirroredStrategy())
Version : 1.5.1
Directory : /home/graemer/Projekte/MXNet/apache-mxnet-src-1.5.1-incubating/python/mxnet
Num GPUs : 2
system : Linux
node : e5lx
release : 5.4.3-arch1-1
version : #1 SMP PREEMPT Fri, 13 Dec 2019 09:39:02 +0000
Architektur: x86_64
CPU Operationsmodus: 32-bit, 64-bit
Byte-Reihenfolge: Little Endian
Adressgrößen: 46 bits physical, 48 bits virtual
CPU(s): 32
Liste der Online-CPU(s): 0-31
Thread(s) pro Kern: 2
Kern(e) pro Socket: 8
Sockel: 2
NUMA-Knoten: 2
Anbieterkennung: GenuineIntel
Prozessorfamilie: 6
Modell: 79
Modellname: Intel(R) Xeon(R) CPU E5-2620 v4 @ 2.10GHz
Stepping: 1
CPU MHz: 1197.887
Maximale Taktfrequenz der CPU: 3000,0000
Minimale Taktfrequenz der CPU: 1200,0000
BogoMIPS: 4192.42
Virtualisierung: VT-x
L1d Cache: 512 KiB
L1i Cache: 512 KiB
L2 Cache: 4 MiB
L3 Cache: 40 MiB
NUMA-Knoten0 CPU(s): 0-7,16-23
NUMA-Knoten1 CPU(s): 8-15,24-31
multigpu [[Horovod]] (2 GPUs): horovodrun -np 2 -H localhost:2 python mnist_mx_gluon_hvd.py
| framework | duration | accuracy |
|---|---|---|
| mnist_mx_gluon.py | 14.2s | 0.9926 |
| mnist_mx_keras.py | 16.4s | 0.9931 |
| mnist_mx_sym.py | 16.8s | 0.9924 |
| mnist_tf_keras.py | 17.6s | 0.9932 |
| mnist_mx_gluon_ds.py | 17.6s | 0.9902 |
| mnist_mx_gluon_mgpu.py | 7.2s | 0.9926 |
| mnist_mx_gluon_hvd.py | 9.0s | 0.9963 |
| mnist_mx_sym_mgpu.py | 10.9s | 0.9923 |
| mnist_mx_keras_mgpu.py | 11.4s | 0.9927 |
| mnist_tf_keras_mgpu.py | 12.7s | 0.9921 |
| mnist_mx_gluon_ds_mgpu.py | 15.2s | 0.9906 |