python - 从 SysFS 读取成功的 NUMA 节点为负值(-1),但必须至少有一个 NUMA 节点,因此返回 NUMA 节点零
问题描述
我从 tensorflow.org https://www.tensorflow.org/tutorials/images/transfer_learning_with_hub开始了关于使用 TF hub 进行迁移学习的教程
当我运行它时,结果如下:
2020-02-10 10:49:00.817444: W tensorflow/stream_executor/platform/default/dso_loader.cc:55] Could not load dynamic library 'libnvinfer.so.6'; dlerror: libnvinfer.so.6: cannot open shared object file: No such file or directory
2020-02-10 10:49:00.817538: W tensorflow/stream_executor/platform/default/dso_loader.cc:55] Could not load dynamic library 'libnvinfer_plugin.so.6'; dlerror: libnvinfer_plugin.so.6: cannot open shared object file: No such file or directory
2020-02-10 10:49:00.817565: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:30] Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.
Using TensorFlow backend.
2020-02-10 10:49:01.671573: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcuda.so.1
2020-02-10 10:49:01.686150: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:981] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-02-10 10:49:01.686802: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1555] Found device 0 with properties:
pciBusID: 0000:01:00.0 name: GeForce GTX 1070 computeCapability: 6.1
coreClock: 1.7465GHz coreCount: 15 deviceMemorySize: 7.92GiB deviceMemoryBandwidth: 238.66GiB/s
2020-02-10 10:49:01.687014: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcudart.so.10.1
2020-02-10 10:49:01.688374: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcublas.so.10
2020-02-10 10:49:01.689491: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcufft.so.10
2020-02-10 10:49:01.689755: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcurand.so.10
2020-02-10 10:49:01.691268: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcusolver.so.10
2020-02-10 10:49:01.692212: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcusparse.so.10
2020-02-10 10:49:01.696243: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcudnn.so.7
2020-02-10 10:49:01.696421: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:981] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-02-10 10:49:01.696934: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:981] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-02-10 10:49:01.697350: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1697] Adding visible gpu devices: 0
2020-02-10 10:49:01.723450: I tensorflow/core/platform/profile_utils/cpu_utils.cc:94] CPU Frequency: 3391195000 Hz
2020-02-10 10:49:01.724086: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x5cc4620 initialized for platform Host (this does not guarantee that XLA will be used). Devices:
2020-02-10 10:49:01.724110: I tensorflow/compiler/xla/service/service.cc:176] StreamExecutor device (0): Host, Default Version
2020-02-10 10:49:01.801614: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:981] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-02-10 10:49:01.802276: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x5636b50 initialized for platform CUDA (this does not guarantee that XLA will be used). Devices:
2020-02-10 10:49:01.802298: I tensorflow/compiler/xla/service/service.cc:176] StreamExecutor device (0): GeForce GTX 1070, Compute Capability 6.1
2020-02-10 10:49:01.802464: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:981] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-02-10 10:49:01.802908: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1555] Found device 0 with properties:
pciBusID: 0000:01:00.0 name: GeForce GTX 1070 computeCapability: 6.1
coreClock: 1.7465GHz coreCount: 15 deviceMemorySize: 7.92GiB deviceMemoryBandwidth: 238.66GiB/s
2020-02-10 10:49:01.802962: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcudart.so.10.1
2020-02-10 10:49:01.802983: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcublas.so.10
2020-02-10 10:49:01.802999: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcufft.so.10
2020-02-10 10:49:01.803022: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcurand.so.10
2020-02-10 10:49:01.803041: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcusolver.so.10
2020-02-10 10:49:01.803057: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcusparse.so.10
2020-02-10 10:49:01.803074: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcudnn.so.7
2020-02-10 10:49:01.803126: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:981] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-02-10 10:49:01.803610: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:981] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-02-10 10:49:01.804030: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1697] Adding visible gpu devices: 0
2020-02-10 10:49:01.804066: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcudart.so.10.1
2020-02-10 10:49:01.804854: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1096] Device interconnect StreamExecutor with strength 1 edge matrix:
2020-02-10 10:49:01.804869: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1102] 0
2020-02-10 10:49:01.804878: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1115] 0: N
2020-02-10 10:49:01.804969: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:981] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-02-10 10:49:01.805430: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:981] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-02-10 10:49:01.805876: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1241] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 7090 MB memory) -> physical GPU (device: 0, name: GeForce GTX 1070, pci bus id: 0000:01:00.0, compute capability: 6.1)
我认为问题来自 gpu 配置,但我不知道如何解决它。如果有人知道该怎么做...
解决方案
推荐阅读
- ruby-on-rails - 外键约束失败 - Rails 数据库迁移
- ios - 适用于 iOS 的 Firebase 动态链接适用于我的设备和模拟器,但发布后不适用于用户
- javascript - 使用 Gmail API 获取 gmail 附件
- python - 尝试创建一个快速排序功能来更改原始列表,但它不会对其进行任何更改
- nightwatch.js - 如何在 Nightwatch 中根据环境设置 retryAssertionTimeout
- domain-name - Punycode 域名 (UName) 是否存储使用的 IDN 表?
- python - Pytest cov 不读取 pyproject.toml
- python - 用 ReferenceFrame 向量替换 SymPy sp.Symbol
- neural-network - PyTorch - 正确计算神经切线内核(每个数据点的雅可比)
- ios - UIStackView 的子视图具有固有大小,但仍然忽略内容拥抱优先级