拟合模型时，内核在一段时间后停止工作

Question

我正在尝试运行 TensorFlow 为 image classification 提供的代码。我使用的是 TensorFlow 提供的完全相同的代码，所以我不在这里分享。代码完美运行到适合模型的程度。它打印一次“Epoch”，然后内核关闭并显示“启动内核时发生错误”。作为它产生的错误消息：

2021???????? 21:19:59.749095: I tensorflow/stream_executor/platform/default/dso_loader.cc:53] Successfully opened dynamic library cudart64_110.dll
2021???????? 21:20:02.178383: I tensorflow/stream_executor/platform/default/dso_loader.cc:53] Successfully opened dynamic library nvcuda.dll
2021???????? 21:20:02.198734: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1733] Found device 0 with properties: 
pciBusID: 0000:01:00.0 name: NVIDIA GeForce RTX 3070 Laptop GPU computeCapability: 8.6
coreClock: 1.62GHz coreCount: 40 deviceMemorySize: 8.00GiB deviceMemoryBandwidth: 417.29GiB/s
2021???????? 21:20:02.198906: I tensorflow/stream_executor/platform/default/dso_loader.cc:53] Successfully opened dynamic library cudart64_110.dll
2021???????? 21:20:02.204104: I tensorflow/stream_executor/platform/default/dso_loader.cc:53] Successfully opened dynamic library cublas64_11.dll
2021???????? 21:20:02.204165: I tensorflow/stream_executor/platform/default/dso_loader.cc:53] Successfully opened dynamic library cublasLt64_11.dll
2021???????? 21:20:02.207305: I tensorflow/stream_executor/platform/default/dso_loader.cc:53] Successfully opened dynamic library cufft64_10.dll
2021???????? 21:20:02.208428: I tensorflow/stream_executor/platform/default/dso_loader.cc:53] Successfully opened dynamic library curand64_10.dll
2021???????? 21:20:02.213539: I tensorflow/stream_executor/platform/default/dso_loader.cc:53] Successfully opened dynamic library cusolver64_11.dll
2021???????? 21:20:02.215481: I tensorflow/stream_executor/platform/default/dso_loader.cc:53] Successfully opened dynamic library cusparse64_11.dll
2021???????? 21:20:02.216199: I tensorflow/stream_executor/platform/default/dso_loader.cc:53] Successfully opened dynamic library cudnn64_8.dll
2021???????? 21:20:02.216287: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1871] Adding visible gpu devices: 0
2021???????? 21:20:02.216750: I tensorflow/core/platform/cpu_feature_guard.cc:142] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance‑critical operations: AVX AVX2
To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
2021???????? 21:20:02.217490: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1733] Found device 0 with properties: 
pciBusID: 0000:01:00.0 name: NVIDIA GeForce RTX 3070 Laptop GPU computeCapability: 8.6
coreClock: 1.62GHz coreCount: 40 deviceMemorySize: 8.00GiB deviceMemoryBandwidth: 417.29GiB/s
2021???????? 21:20:02.217546: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1871] Adding visible gpu devices: 0
2021???????? 21:20:02.708850: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1258] Device interconnect StreamExecutor with strength 1 edge matrix:
2021???????? 21:20:02.708874: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1264] 0 
2021???????? 21:20:02.708880: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1277] 0: N 
2021???????? 21:20:02.709035: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1418] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 5484 MB memory) ‑> physical GPU (device: 0, name: NVIDIA GeForce RTX 3070 Laptop GPU, pci bus id: 0000:01:00.0, compute capability: 8.6)
2021???????? 21:20:04.004652: I tensorflow/compiler/mlir/mlir_graph_optimization_pass.cc:176] None of the MLIR Optimization Passes are enabled (registered 2)
2021???????? 21:20:05.150123: I tensorflow/stream_executor/platform/default/dso_loader.cc:53] Successfully opened dynamic library cudnn64_8.dll```

Answer 1

我在 colab 中复制了相同的给定代码。它运行成功，没有任何错误。请在此处找到相关代码gist。

但是，它们只是信息消息，因为它们以I为前缀，如果是错误消息，它们将以E或W为前缀作为警告，如下所示：

2020-12-30 21:30:27.549172: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cupti64_101.dll

2020-12-30 21:30:27.599977: W tensorflow/core/framework/allocator.cc:101] Allocation of 37171200 exceeds 10% of system memory.

2021-12-30 21:30:27.704083: E tensorflow/core/profiler/internal/gpu/cupti_tracer.cc:1307] function cupti_interface_->Subscribe( &subscriber_, (CUpti_CallbackFunc)ApiCallback, this)failed with error CUPTI_ERROR_INSUFFICIENT_PRIVILEGES

您可以使用以下代码超越这些警告：

import os
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'