解决tensorflow运行显存不够的问题

拼搏现实的明天。 2022-01-05 12:27 772阅读 0赞

在训练模型的过程中遇到程序报错，结果发现时因为电脑显存不够，资源都已经分配完了。根据多方查阅，找到了此问题的解决方法，在这里记录和分享一下~

## 出现问题 ##

**错误：**Error message: OOM when allocating tensor with shape\[8,64,256,256\] and type float on /job:localhost/replica:0/task:0/device:GPU:0 by allocator GPU\_0\_bfc  
     \[\[node moments\_3/SquaredDifference (defined at D:\\Pycharm\\arbitrary\_style\_transfer-master\\train.py:73) \]\]  
Hint: If you want to see a list of allocated tensors when OOM happens, add report\_tensor\_allocations\_upon\_oom to RunOptions for current allocation info.

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM3Mzc0NjQz_size_16_color_FFFFFF_t_70][]

## 解决办法 ##

在main函数中添加如下这段代码

import tensorflow as tf
    import os
    os.environ["CUDA_VISIBLE_DEVICES"] = '0'   #指定第一块GPU可用
    config = tf.ConfigProto()
    config.gpu_options.per_process_gpu_memory_fraction = 0.5  # 程序最多只能占用指定gpu50%的显存
    config.gpu_options.allow_growth = True      #程序按需申请内存
    sess = tf.Session(config=config)

位置就在开头，如图：

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM3Mzc0NjQz_size_16_color_FFFFFF_t_70 1][]

## 查看电脑显存 ##

打开电脑**设置-**>**系统**\->**显示**\->**高级显示设置**\->**显示适配器属性**

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM3Mzc0NjQz_size_16_color_FFFFFF_t_70 2][]           ![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM3Mzc0NjQz_size_16_color_FFFFFF_t_70 3][]

![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM3Mzc0NjQz_size_16_color_FFFFFF_t_70 4][]            ![watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM3Mzc0NjQz_size_16_color_FFFFFF_t_70 5][]

[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM3Mzc0NjQz_size_16_color_FFFFFF_t_70]: /images/20211229/f6cbc04fcd0641be8f88781638959dfb.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM3Mzc0NjQz_size_16_color_FFFFFF_t_70 1]: /images/20211229/96304836c7fe4072b8943c4717ee1d27.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM3Mzc0NjQz_size_16_color_FFFFFF_t_70 2]: https://img-blog.csdnimg.cn/20190623215331494.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM3Mzc0NjQz,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM3Mzc0NjQz_size_16_color_FFFFFF_t_70 3]: /images/20211229/8f9fc54a175f4089b9fe0bf790eabd0b.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM3Mzc0NjQz_size_16_color_FFFFFF_t_70 4]: https://img-blog.csdnimg.cn/20190623215408921.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM3Mzc0NjQz,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzM3Mzc0NjQz_size_16_color_FFFFFF_t_70 5]: /images/20211229/8bc4f45bcd9647649439c5705a99b694.png