我已經基於Theano example下面的代碼:啓用CNMeM的效果如何,但Theano中的'cuDNN不可用'?
from theano import function, config, shared, sandbox
import theano.tensor as T
import numpy
import time
vlen = 10 * 30 * 768 # 10 x #cores x # threads per core
iters = 1000
rng = numpy.random.RandomState(22)
x = shared(numpy.asarray(rng.rand(vlen), config.floatX))
f = function([], T.exp(x))
print(f.maker.fgraph.toposort())
t0 = time.time()
for i in range(iters):
r = f()
t1 = time.time()
print("Looping %d times took %f seconds" % (iters, t1 - t0))
print("Result is %s" % (r,))
if numpy.any([isinstance(x.op, T.Elemwise) for x in f.maker.fgraph.toposort()]):
print('Used the cpu')
else:
print('Used the gpu')
現在,當我測試兩種模式的代碼:
GPU模式下,我得到這個:
$ THEANO_FLAGS=mode=FAST_RUN,device=gpu,floatX=float32 python gpu.py
Using gpu device 0: Tesla C2075 (CNMeM is enabled with initial size: 95.0% of memory, cuDNN not available)
[GpuElemwise{exp,no_inplace}(<CudaNdarrayType(float32, vector)>), HostFromGpu(GpuElemwise{exp,no_inplace}.0)]
Looping 1000 times took 0.475526 seconds
Result is [ 1.23178029 1.61879349 1.52278066 ..., 2.20771813 2.29967761
1.62323296]
Used the gpu
CPU模式,我得到這個:
$ THEANO_FLAGS=mode=FAST_RUN,device=cpu,floatX=float32 python gpu.py
[Elemwise{exp,no_inplace}(<TensorType(float32, vector)>)]
Looping 1000 times took 5.221368 seconds
Result is [ 1.23178029 1.61879337 1.52278066 ..., 2.20771813 2.29967761
1.62323284]
Used the cpu
注意兩件事,GPU是ind比CPU更快(0.47秒vs 5秒)。但同時在GPU上,我得到了cuDNN not available消息。
我的問題是這樣的。缺少cuDNN有什麼作用?它有害嗎?
自從我上次使用theano以來已經有一段時間了,但我認爲你可能會使用一般的CUDA庫,因此獲得GPU加速的好處,但不是更專業化的cudNN庫。所以如果你能夠安裝並運行cudNN,你可能會獲得一些額外的加速。 – Marius