内核在运行一些代码后死亡
我尝试运行代码以使用生成器生成示例图像
我试图更新conda和Jupiter,但是它们都不起作用
我一直在关注GPU的内存使用情况,但是它并没有那么多地使用GPU
tensorflow2.0,ubuntu 18.10,cuda 10.0
python 3.5,
def make_generator_model():
model = tf.keras.Sequential()
model.add(layers.Dense(7*7*256, use_bias=False, input_shape=(100,)))
model.add(layers.BatchNormalization())
model.add(layers.LeakyReLU())
model.add(layers.Reshape((7, 7, 256)))
assert model.output_shape == (None, 7, 7, 256) # Note: None is the batch size
model.add(layers.Conv2DTranspose(128, (5, 5), strides=(1, 1), padding='same', use_bias=False))
assert model.output_shape == (None, 7, 7, 128)
model.add(layers.BatchNormalization())
model.add(layers.LeakyReLU())
model.add(layers.Conv2DTranspose(64, (5, 5), strides=(2, 2), padding='same', use_bias=False))
assert model.output_shape == (None, 14, 14, 64)
model.add(layers.BatchNormalization())
model.add(layers.LeakyReLU())
model.add(layers.Conv2DTranspose(1, (5, 5), strides=(2, 2), padding='same', use_bias=False, activation='tanh'))
assert model.output_shape == (None, 28, 28, 1)
return model
generator = make_generator_model()
noise = tf.random.normal([1, 100])
generated_image = generator(noise, training=False)
[I 10:20:06.664 NotebookApp] KernelRestarter:重新启动内核(1/5), 保持随机端口警告:root:kernel 4406ce3b-1b5b-4ef8-aba9-d5fd9ed129e7重新启动2019-04-18 10:20:21.002451:我 tensorflow / stream_executor / platform / default / dso_loader.cc:42] 成功打开动态库libcuda.so.1 2019-04-18 10:20:21.081020:我 tensorflow / core / common_runtime / gpu / gpu_device.cc:1589]找到设备0 具有属性:名称:TITAN Xp主修:6主修:1 memoryClockRate(GHz):1.582 pciBusID:0000:42:00.0 totalMemory: 11.91GiB freeMemory:340.69MiB 2019-04-18 10:20:21.081054:我tensorflow / core / common_runtime / gpu / gpu_device.cc:1712]添加可见 gpu设备:0 2019-04-18 10:20:21.081382:I tensorflow / core / platform / cpu_feature_guard.cc:142]您的CPU支持 TensorFlow二进制文件未编译使用的指令:AVX2 FMA 2019-04-18 10:20:21.107510:I tensorflow / compiler / xla / service / service.cc:168] XLA服务 0x55de6ead0990在平台CUDA上执行计算。设备: 2019-04-18 10:20:21.107562:我 tensorflow / compiler / xla / service / service.cc:175] StreamExecutor 设备(0):TITAN Xp,计算能力6.1 2019-04-18 10:20:21.127890:我 tensorflow / core / platform / profile_utils / cpu_utils.cc:94] CPU频率: 3493050000 Hz 2019-04-18 10:20:21.129460:I tensorflow / compiler / xla / service / service.cc:168] XLA服务 0x55de6eed7eb0在平台Host上执行计算。设备: 2019-04-18 10:20:21.129503:我 tensorflow / compiler / xla / service / service.cc:175] StreamExecutor 设备(0):,2019-04-18 10:20:21.129616:I tensorflow / core / common_runtime / gpu / gpu_device.cc:1712]添加可见 gpu设备:0 2019-04-18 10:20:21.129722:I tensorflow / stream_executor / platform / default / dso_loader.cc:42] 成功打开动态库libcudart.so.10.0 2019-04-18 10:20:21.130785:我 tensorflow / core / common_runtime / gpu / gpu_device.cc:1120]设备 将StreamExecutor与强度1边缘矩阵互连:2019-04-18 10:20:21.130807:我 tensorflow / core / common_runtime / gpu / gpu_device.cc:1126] 0 2019-04-18 10:20:21.130819:我 tensorflow / core / common_runtime / gpu / gpu_device.cc:1139] 0:N 2019-04-18 10:20:21.131090:我 tensorflow / core / common_runtime / gpu / gpu_device.cc:1260]已创建 TensorFlow设备(/ job:localhost /副本:0 /任务:0 /设备:GPU:0与 115 MB内存)->物理GPU(设备:0,名称:TITAN Xp,PCI总线ID: 0000:42:00.0,计算能力:6.1)2019-04-18 10:20:24.168083:I tensorflow / stream_executor / platform / default / dso_loader.cc:42] 成功打开动态库libcublas.so.10.0 2019-04-18 10:20:24.331094:我 tensorflow / stream_executor / platform / default / dso_loader.cc:42] 成功打开动态库libcudnn.so.7 2019-04-18 10:20:24.789774:E tensorflow / stream_executor / cuda / cuda_dnn.cc:329] 无法创建cudnn句柄:CUDNN_STATUS_INTERNAL_ERROR 2019-04-18 10:20:24.791468:E tensorflow / stream_executor / cuda / cuda_dnn.cc:329] 无法创建cudnn句柄:CUDNN_STATUS_INTERNAL_ERROR 2019-04-18 10:20:24.791484:F tensorflow /核心/内核/conv_grad_input_ops.cc:949] 检查失败:stream-> parent()-> GetConvolveBackwardDataAlgorithms( conv_parameters.ShouldIncludeWinogradNonfusedAlgo(stream-> parent()), &algorithms)[I 10:20:27.669 NotebookApp] KernelRestarter:重新启动 内核(1/5),请保留随机端口警告:root:kernel 4406ce3b-1b5b-4ef8-aba9-d5fd9ed129e7重新启动
答案 0 :(得分:0)
根据错误的输出,似乎是内存问题。
“总内存:11.91GiB freeMemory:340.69MiB”
尝试重新启动PC,并在重新打开PC时立即查看有多少RAM可用,然后再次执行代码,查看其是否正常工作。