zoukankan      html  css  js  c++  java
  • 已经安装cuda但是tensorflow仍然使用cpu加速的问题

    安装了keras、theano之后,一直以为自己用的GPU,今天找到一个小程序测试一下,竟然一直在用CPU(黑人问号)

    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')
    

    解决方案:

    pip list 查看有没有安装#tensorflow和tensorboard

    pip uninstall tensorflow

    pip uninstall keras

    pip uninastall tensorflow-gpu

    pip install tendorflow-gpu

    pip install keras

    把页面关了,重启

  • 相关阅读:
    python
    python
    gitlab
    nodejs
    java
    ElasticSearch 安装与配置 (windows)
    shell脚本批量注释
    C获取系统中CPU核数
    linux内核内存管理
    perf: interrupt took too long (3136 > 3126), lowering kernel.perf_event_max_sample_rate to 63000
  • 原文地址:https://www.cnblogs.com/kongle666/p/9456417.html
Copyright © 2011-2022 走看看