gdfishhannah
commented
5 years ago Make sure you install the tensorflow-gpu instead of the tensorflow.
Open zhangguowei0916 opened 6 years ago
gdfishhannah
commented
5 years ago Make sure you install the tensorflow-gpu instead of the tensorflow.
I try to train on GPU but it only use the CPU and say : 2018-07-30 16:34:20.042900: W d:\nwani\l\tensorflow_1497951990344\work\tensorflow-1.1.0\tensorflow\core\platform\cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE instructions, but these are available on your machine and could speed up CPU computations. 2018-07-30 16:34:20.043657: W d:\nwani\l\tensorflow_1497951990344\work\tensorflow-1.1.0\tensorflow\core\platform\cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE2 instructions, but these are available on your machine and could speed up CPU computations. 2018-07-30 16:34:20.044608: W d:\nwani\l\tensorflow_1497951990344\work\tensorflow-1.1.0\tensorflow\core\platform\cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE3 instructions, but these are available on your machine and could speed up CPU computations. 2018-07-30 16:34:20.045558: W d:\nwani\l\tensorflow_1497951990344\work\tensorflow-1.1.0\tensorflow\core\platform\cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE4.1 instructions, but these are available on your machine and could speed up CPU computations. 2018-07-30 16:34:20.046279: W d:\nwani\l\tensorflow_1497951990344\work\tensorflow-1.1.0\tensorflow\core\platform\cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE4.2 instructions, but these are available on your machine and could speed up CPU computations. 2018-07-30 16:34:20.047699: W d:\nwani\l\tensorflow_1497951990344\work\tensorflow-1.1.0\tensorflow\core\platform\cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use AVX instructions, but these are available on your machine and could speed up CPU computations. 2018-07-30 16:34:20.048379: W d:\nwani\l\tensorflow_1497951990344\work\tensorflow-1.1.0\tensorflow\core\platform\cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use AVX2 instructions, but these are available on your machine and could speed up CPU computations. 2018-07-30 16:34:20.049305: W d:\nwani\l\tensorflow_1497951990344\work\tensorflow-1.1.0\tensorflow\core\platform\cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use FMA instructions, but these are available on your machine and could speed up CPU computations. INFO:tensorflow:Restoring parameters from D:\zgw\workspace\SSD-Tensorflow-master\checkpoints\tfmodel/model.ckpt-3445 INFO:tensorflow:Starting Session. INFO:tensorflow:Starting Queues.
I have change data_format from NCHW to NHWC in train.py and in ssd_vgg300.py because with NCHW it said: InvalidArgumentError (see above for traceback): CPU BiasOp only supports NHWC.
I have installed cdnn tf-cpu ... WHEN i train own data with frcnn/yolo3 and tensorflow, gpu will be used 30% ,but ssd is invalid. anybody know why.
Thanks