PlaidML, macOS Machine Learning with GPU
usage
- 20210101
pip install -U plaidml-keras
plaidml-setup
# 需要指定版本
pip install h5py==2.10.0
# test
plaidbench keras mobilenet
<< EELOG
Running 1024 examples with mobilenet, batch size 1, on backend plaid
INFO:plaidml:Opening device "metal_amd_radeon_rx_570.0"
Compiling network... Warming up... Running...
Example finished, elapsed: 1.952s (compile), 5.352s (execution)
-----------------------------------------------------------------------------------------
Network Name Inference Latency Time / FPS
-----------------------------------------------------------------------------------------
mobilenet 5.23 ms 0.00 ms / 1000000000.00 fps
Correctness: PASS, max_error: 1.675534622336272e-05, max_abs_error: 7.674098014831543e-07, fail_ratio: 0.0
EELOG
- issue
'str' object has no attribute 'decode'
$ plaidbench keras mobilenet
Running 1024 examples with mobilenet, batch size 1, on backend plaid
INFO:plaidml:Opening device "metal_amd_radeon_rx_570.0"
Downloading data from https://github.com/fchollet/deep-learning-models/releases/download/v0.6/mobilenet_1_0_224_tf.h5
17227776/17225924 [==============================] - 3s 0us/step
'str' object has no attribute 'decode'
Set --print-stacktraces to see the entire traceback
---solution
pip install h5py==2.10.0end
EOF