tensorflow.python.framework.errors_impl.ResourceExhaustedError: OOM when allocating tensor with shape[1,1424,2128,64]
1、開始以為是 是顯卡顯存不足，沒有想到是gpu問題
ref: https://www.cnblogs.com/heiao10duan/p/9083303.html
Dense(1024, activation='relu'), #將batch_size從4096改為1024因為我的顯存不夠
2、開始以為是沒有選中獨立顯卡
# import os #key: https://blog.csdn.net/littlehaes/article/details/82317220
# os.environ["CUDA_VISIBLE_DEVICES"] = "2" #ref: https://www.cnblogs.com/ying-chease/p/9473938.html
#ref: https://www.cnblogs.com/darkknightzh/archive/2004/01/13/8576825.html
# ref:http://www.voidcn.com/article/p-xhpiftyw-bte.html
# ref:https://blog.csdn.net/weixin_43301333/article/details/108854504
# os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2' # ref:https://blog.csdn.net/zqx951102/article/details/88897117
# ref:https://www.imooc.com/article/details/id/289425
3、
后來通過list_local_devices發現是沒有裝TensorFlow-gpu， install TensorFlow-gpu
# from tensorflow.python.client import device_lib
# print(device_lib.list_local_devices())
# https://blog.csdn.net/littlehaes/article/details/82317220
但是注意，安裝命令應該使用：
conda install -c anaconda tensorflow-gpu
而不是
pip3 uninstall tensorflow-gpu
測試是否安裝成功
方法一：
import tensorflow as tf
import tensorflow.compat.v1 as tfs
sess = tfs.Session(config=tfs.ConfigProto(log_device_placement=True))
輸出中看到：
2020-12-10 08:38:41.553101: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1241] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 3001 MB memory) -> physical GPU (device: 0, name: Quadro P1000, pci bus id: 0000:01:00.0, compute capability: 6.1)
Device mapping:
方法二：
from tensorflow.python.client import device_lib
print(device_lib.list_local_devices())
輸出中看到：
ncarnation: 5727256362115513404
, name: "/device:GPU:0"
device_type: "GPU"
memory_limit: 3147143577
locality {
bus_id: 1
links {
}
}
4、安裝成功后，代碼運行報錯
from tensorflow.python._pywrap_tensorflow_internal import *
ImportError: DLL load failed: 找不到指定的模塊。
https://www.jb51.net/article/178940.htm
