Question

我已经下载了所有软件程序，以便通过GPU（CUDA / cuDNN）执行Keras。正如您在下面的代码中看到的那样，它似乎起作用。这是一个CNN，当我尝试训练模型时，一个纪元的训练将持续10分钟（与CPU运行时间相同）。

Using TensorFlow backend.
2020-04-08 23:19:03.654388: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cudart64_101.dll
2020-04-08 23:19:08.302815: I tensorflow/core/platform/cpu_feature_guard.cc:142] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2
2020-04-08 23:19:08.357836: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library nvcuda.dll
2020-04-08 23:19:09.085263: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1555] Found device 0 with properties:
pciBusID: 0000:01:00.0 name: GeForce GTX 1050 computeCapability: 6.1
coreClock: 1.493GHz coreCount: 5 deviceMemorySize: 2.00GiB deviceMemoryBandwidth: 104.43GiB/s
2020-04-08 23:19:09.093385: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cudart64_101.dll
2020-04-08 23:19:09.188055: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cublas64_10.dll
2020-04-08 23:19:09.234655: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cufft64_10.dll
2020-04-08 23:19:09.260896: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library curand64_10.dll
2020-04-08 23:19:09.345531: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cusolver64_10.dll
2020-04-08 23:19:09.396772: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cusparse64_10.dll
2020-04-08 23:19:09.540946: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cudnn64_7.dll
2020-04-08 23:19:09.548839: F tensorflow/stream_executor/lib/statusor.cc:34] Attempting to fetch value instead of handling error Internal: failed to get device attribute 13 for device 0: CUDA_ERROR_UNKNOWN: unknown error

(.envWindows) C:\Users\Florian\Desktop\Ptrans - Smart Parking\u_park>python script_training.py
Using TensorFlow backend.
2020-04-08 23:19:29.353347: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cudart64_101.dll
2020-04-08 23:19:31.572314: I tensorflow/core/platform/cpu_feature_guard.cc:142] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2
2020-04-08 23:19:31.628773: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library nvcuda.dll
2020-04-08 23:19:32.276405: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1555] Found device 0 with properties:
pciBusID: 0000:01:00.0 name: GeForce GTX 1050 computeCapability: 6.1
coreClock: 1.493GHz coreCount: 5 deviceMemorySize: 2.00GiB deviceMemoryBandwidth: 104.43GiB/s
2020-04-08 23:19:32.283139: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cudart64_101.dll
2020-04-08 23:19:32.290215: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cublas64_10.dll
2020-04-08 23:19:32.297466: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cufft64_10.dll
2020-04-08 23:19:32.302702: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library curand64_10.dll
2020-04-08 23:19:32.311060: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cusolver64_10.dll
2020-04-08 23:19:32.317673: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cusparse64_10.dll
2020-04-08 23:19:32.332136: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cudnn64_7.dll
2020-04-08 23:19:32.344553: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1697] Adding visible gpu devices: 0
2020-04-08 23:19:33.192037: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1096] Device interconnect StreamExecutor with strength 1 edge matrix:
2020-04-08 23:19:33.196639: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1102]      0
2020-04-08 23:19:33.199622: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1115] 0:   N
2020-04-08 23:19:33.203658: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1241] Created TensorFlow device (/device:GPU:0 with 1337 MB memory) -> physical GPU (device: 0, name: GeForce GTX 1050, pci bus id: 0000:01:00.0, compute capability: 6.1)
[name: "/device:CPU:0"
device_type: "CPU"
memory_limit: 268435456
locality {
}
incarnation: 4888491968132995736
, name: "/device:GPU:0"
device_type: "GPU"
memory_limit: 1401988300
locality {
  bus_id: 1
  links {
  }
}
incarnation: 16691478652768016864
physical_device_desc: "device: 0, name: GeForce GTX 1050, pci bus id: 0000:01:00.0, compute capability: 6.1"
]
2020-04-08 23:19:33.293485: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1555] Found device 0 with properties:
pciBusID: 0000:01:00.0 name: GeForce GTX 1050 computeCapability: 6.1
coreClock: 1.493GHz coreCount: 5 deviceMemorySize: 2.00GiB deviceMemoryBandwidth: 104.43GiB/s
2020-04-08 23:19:33.300364: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cudart64_101.dll
2020-04-08 23:19:33.303574: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cublas64_10.dll
2020-04-08 23:19:33.306443: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cufft64_10.dll
2020-04-08 23:19:33.309228: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library curand64_10.dll
2020-04-08 23:19:33.312697: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cusolver64_10.dll
2020-04-08 23:19:33.316427: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cusparse64_10.dll
2020-04-08 23:19:33.321711: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cudnn64_7.dll
2020-04-08 23:19:33.325715: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1697] Adding visible gpu devices: 0
2020-04-08 23:19:33.334856: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1555] Found device 0 with properties:
pciBusID: 0000:01:00.0 name: GeForce GTX 1050 computeCapability: 6.1
coreClock: 1.493GHz coreCount: 5 deviceMemorySize: 2.00GiB deviceMemoryBandwidth: 104.43GiB/s
2020-04-08 23:19:33.349585: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cudart64_101.dll
2020-04-08 23:19:33.356020: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cublas64_10.dll
2020-04-08 23:19:33.359692: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cufft64_10.dll
2020-04-08 23:19:33.371302: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library curand64_10.dll
2020-04-08 23:19:33.374765: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cusolver64_10.dll
2020-04-08 23:19:33.386556: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cusparse64_10.dll
2020-04-08 23:19:33.390107: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cudnn64_7.dll
2020-04-08 23:19:33.401618: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1697] Adding visible gpu devices: 0
2020-04-08 23:19:33.404151: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1096] Device interconnect StreamExecutor with strength 1 edge matrix:
2020-04-08 23:19:33.412276: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1102]      0
2020-04-08 23:19:33.418029: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1115] 0:   N
2020-04-08 23:19:33.421211: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1241] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 1337 MB memory) -> physical GPU (device: 0, name: GeForce GTX 1050, pci bus id: 0000:01:00.0, compute capability: 6.1)
Found 94493 images belonging to 2 classes.
Found 18647 images belonging to 2 classes.
Found 1888 images belonging to 2 classes.
--- Start fit cycle : NewMNetV2_finetuning_unfreezed ---
Cycle 1
Epoch 1/1
2020-04-08 23:19:46.672284: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cublas64_10.dll
2020-04-08 23:19:47.032341: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cudnn64_7.dll
2020-04-08 23:19:48.284266: W tensorflow/stream_executor/gpu/redzone_allocator.cc:312] Internal: Invoking GPU asm compilation is supported on Cuda non-Windows platforms only
Relying on driver to perform ptx compilation. This message will be only logged once.
2215/2952 [=====================>........] - ETA: 2:35 - loss: 0.1830 - accuracy: 0.9647Traceback (most recent call last):

硬件：

我的电脑->联想军团Y520-15IKBN 80WK

GPU-> Nvidia GeForce GTX 1050

CPU-> i5-7300HQ

长时间执行GPU训练

0 个答案: