mnist_cnn by keras-PlaidML GPU in macOS
@siero5335
2018/6/25
Install Keras and PlaidML
install.packages(keras)
install_keras()CPU-based Keras and TensorFlow was installed.
I installed PlaidML in r-tensorflow env (Site CPU-based Keras and TensorFlow was installed) according to the following site.
Installation Instructions of PlaidML (macOS) In PlaidML installation, Keras ver 2.0.8 might be install with PlaidML.
In this time, after install the PlaidML, Keras ver was updated to 2.2.0.
Next, setup PlaidML to use my computing device:
plaidml-setup (in terminal)
Reference (Japanese): Macでも高速に機械学習できるかもしれないPlaidMLを試してみた
plaidml-setup PlaidML Setup (0.3.4)
Thanks for using PlaidML!
Some Notes: * Bugs and other issues: https://github.com/plaidml/plaidml * Questions: https://stackoverflow.com/questions/tagged/plaidml * Say hello: https://groups.google.com/forum/#!forum/plaidml-dev * PlaidML is licensed under the GNU AGPLv3
Default Config Devices: No devices.
Experimental Config Devices: llvm_cpu.0 : CPU (LLVM) opencl_intel_intel(r)_hd_graphics_530.0 : Intel Inc. Intel(R) HD Graphics 530 (OpenCL) opencl_cpu.0 : Intel CPU (OpenCL) metal_amd_radeon_pro_450.0 : AMD Radeon Pro 450 (Metal) opencl_amd_amd_radeon_pro_450_compute_engine.0 : AMD AMD Radeon Pro 450 Compute Engine (OpenCL) metal_intel(r)_hd_graphics_530.0 : Intel(R) HD Graphics 530 (Metal)
Using experimental devices can cause poor performance, crashes, and other nastiness.
Enable experimental device support? (y,n)[n]:yMultiple devices detected (You can override by setting PLAIDML_DEVICE_IDS). Please choose a default device:
1 : llvm_cpu.0 2 : opencl_intel_intel(r)_hd_graphics_530.0 3 : opencl_cpu.0 4 : metal_amd_radeon_pro_450.0 5 : opencl_amd_amd_radeon_pro_450_compute_engine.0 6 : metal_intel(r)_hd_graphics_530.0
Default device? (1,2,3,4,5,6)[1]:5Selected device: opencl_amd_amd_radeon_pro_450_compute_engine.0
PlaidML sends anonymous usage statistics to help guide improvements. We’d love your help making it better.
Enable telemetry reporting? (y,n)[y]:yAlmost done. Multiplying some matrices… Tile code: function (B[X,Z], C[Z,Y]) -> (A) { A[x,y : X,Y] = +(B[x,z] * C[z,y]); } Whew. That worked.
Save settings to /Users/XX/.plaidml? (y,n)[y]:ySuccess!
Load keras
library("keras"); library("sessioninfo")Set PlaidML as backend
use_backend(backend = "plaidml")mnist_cnn_PlaidML
Following is almost same as mnist_cnn exsample.
batch_size <- 128
num_classes <- 10
epochs <- 5img_rows <- 28
img_cols <- 28mnist <- dataset_mnist()
x_train <- mnist$train$x
y_train <- mnist$train$y
x_test <- mnist$test$x
y_test <- mnist$test$yx_train <- array_reshape(x_train, c(nrow(x_train), img_rows, img_cols, 1))
x_test <- array_reshape(x_test, c(nrow(x_test), img_rows, img_cols, 1))
input_shape <- c(img_rows, img_cols, 1)
x_train <- x_train / 255
x_test <- x_test / 255
y_train <- to_categorical(y_train, 10)
y_test <- to_categorical(y_test, 10)
model <- keras_model_sequential() %>%
layer_conv_2d(filters = 32, kernel_size = c(3,3), activation = 'relu',
input_shape = input_shape) %>%
layer_conv_2d(filters = 64, kernel_size = c(3,3), activation = 'relu') %>%
layer_max_pooling_2d(pool_size = c(2, 2)) %>%
layer_dropout(rate = 0.25) %>%
layer_flatten() %>%
layer_dense(units = 128, activation = 'relu') %>%
layer_dropout(rate = 0.5) %>%
layer_dense(units = num_classes, activation = 'softmax')
summary(model)## ___________________________________________________________________________
## Layer (type) Output Shape Param #
## ===========================================================================
## conv2d_1 (Conv2D) (None, 26, 26, 32) 320
## ___________________________________________________________________________
## conv2d_2 (Conv2D) (None, 24, 24, 64) 18496
## ___________________________________________________________________________
## max_pooling2d_1 (MaxPooling2D) (None, 12, 12, 64) 0
## ___________________________________________________________________________
## dropout_1 (Dropout) (None, 12, 12, 64) 0
## ___________________________________________________________________________
## flatten_1 (Flatten) (None, 9216) 0
## ___________________________________________________________________________
## dense_1 (Dense) (None, 128) 1179776
## ___________________________________________________________________________
## dropout_2 (Dropout) (None, 128) 0
## ___________________________________________________________________________
## dense_2 (Dense) (None, 10) 1290
## ===========================================================================
## Total params: 1,199,882
## Trainable params: 1,199,882
## Non-trainable params: 0
## ___________________________________________________________________________model %>% compile(
loss = loss_categorical_crossentropy,
optimizer = optimizer_adadelta(),
metrics = c('accuracy')
)Train model using plaidml (GPU)
model %>% fit(
x_train, y_train,
batch_size = batch_size,
epochs = epochs,
validation_split = 0.2
)
GPU was worked
Keras-PlaidML GPU was about 3-4 times fater than Keras-TensorFlow-CPU in my environment.
Finally I got motivated deep learning with R installed on macOS (AMD GPU) !!
session_info()## ─ Session info ──────────────────────────────────────────────────────────
## setting value
## version R version 3.5.0 (2018-04-23)
## os macOS High Sierra 10.13.5
## system x86_64, darwin15.6.0
## ui X11
## language (EN)
## collate ja_JP.UTF-8
## tz Asia/Tokyo
## date 2018-06-25
##
## ─ Packages ──────────────────────────────────────────────────────────────
## package * version date source
## backports 1.1.2 2017-12-13 CRAN (R 3.5.0)
## base64enc 0.1-3 2015-07-28 CRAN (R 3.5.0)
## clisymbols 1.2.0 2017-05-21 CRAN (R 3.5.0)
## digest 0.6.15 2018-01-28 CRAN (R 3.5.0)
## evaluate 0.10.1 2017-06-24 CRAN (R 3.5.0)
## htmltools 0.3.6 2017-04-28 CRAN (R 3.5.0)
## jsonlite 1.5 2017-06-01 CRAN (R 3.5.0)
## keras * 2.1.6 2018-04-29 CRAN (R 3.5.0)
## knitr 1.20 2018-02-20 CRAN (R 3.5.0)
## lattice 0.20-35 2017-03-25 CRAN (R 3.5.0)
## magrittr 1.5 2014-11-22 CRAN (R 3.5.0)
## Matrix 1.2-14 2018-04-13 CRAN (R 3.5.0)
## R6 2.2.2 2017-06-17 CRAN (R 3.5.0)
## Rcpp 0.12.17 2018-05-18 CRAN (R 3.5.0)
## reticulate 1.8.0.9000 2018-06-24 Github (rstudio/reticulate@ea5e9b5)
## rmarkdown 1.10 2018-06-11 CRAN (R 3.5.0)
## rprojroot 1.3-2 2018-01-03 CRAN (R 3.5.0)
## sessioninfo * 1.0.0 2017-06-21 CRAN (R 3.5.0)
## stringi 1.2.3 2018-06-12 CRAN (R 3.5.0)
## stringr 1.3.1 2018-05-10 cran (@1.3.1)
## tensorflow 1.8 2018-06-14 CRAN (R 3.5.0)
## tfruns 1.3 2018-02-18 CRAN (R 3.5.0)
## whisker 0.3-2 2013-04-28 CRAN (R 3.5.0)
## withr 2.1.2 2018-03-15 CRAN (R 3.5.0)
## yaml 2.1.19 2018-05-01 cran (@2.1.19)
## zeallot 0.1.0 2018-01-28 CRAN (R 3.5.0)