在张量流中构建模型时，我尝试使用网格搜索尝试不同的超参数。因此，有时会收到以下警告：
W tensorflow/core/common_runtime/bfc_allocator.cc:211] Allocator (GPU_0_bfc) ran out of memory trying to allocate 3.50GiB. The caller indicates that this is not a failure, but may mean that there could be performance gains if more memory were available.
因此，我想跳过训练导致此警告的模型。
此外，我发现类似的警告导致以下错误：
F tensorflow/stream_executor/cuda/cuda_dnn.cc:231] Check failed: status == CUDNN_STATUS_SUCCESS (7 vs. 0)Failed to set cuDNN stream.
请注意，我有从二进制文件构建的tensorflow-gpu 1.12；窗户10; GPU：GTX 1080 Ti和cuda v9.0
答案 0 :(得分：0)
将cuDNN从v7.4.2.24更新到v7.5.0.56后，我的问题得到解决