CODEWITHSUNDEEP
tensorflowubuntu-16.04
kane_choigo
kane_choigo

Reputation: 11

Am I using tensorflow GPU?

I'm using TensorFlow-GPU 1.14 on Ubuntu 16.04.

As I'm not familiar with TensorFlow, I wonder I'm using GPU practically or not.

I have

GeForce GTX 1060

Nvidia-driver 418

CUDA 10.0

cuDNN v7.6.5

And when I execute my codes I always get this message,

WARNING:tensorflow:From /home/mine/catkin_ws/src/PROJECT/project6_3/src/ddpg.py:26: The name tf.InteractiveSession is deprecated. Please use tf.compat.v1.InteractiveSession instead.

2020-06-24 20:29:13.827441: I tensorflow/core/platform/cpu_feature_guard.cc:142] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2 FMA
2020-06-24 20:29:13.834067: I tensorflow/stream_executor/platform/default/dso_loader.cc:42] Successfully opened dynamic library libcuda.so.1
2020-06-24 20:29:13.930412: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:1005] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-06-24 20:29:13.931260: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x6a40a50 executing computations on platform CUDA. Devices:
2020-06-24 20:29:13.931277: I tensorflow/compiler/xla/service/service.cc:175]   StreamExecutor device (0): GeForce GTX 1060 6GB, Compute Capability 6.1
2020-06-24 20:29:13.959129: I tensorflow/core/platform/profile_utils/cpu_utils.cc:94] CPU Frequency: 3000000000 Hz
2020-06-24 20:29:13.959392: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x6ab1d70 executing computations on platform Host. Devices:
2020-06-24 20:29:13.959409: I tensorflow/compiler/xla/service/service.cc:175]   StreamExecutor device (0): <undefined>, <undefined>
2020-06-24 20:29:13.959576: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:1005] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-06-24 20:29:13.960326: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1640] Found device 0 with properties: 
name: GeForce GTX 1060 6GB major: 6 minor: 1 memoryClockRate(GHz): 1.759
pciBusID: 0000:01:00.0
2020-06-24 20:29:13.961867: I tensorflow/stream_executor/platform/default/dso_loader.cc:42] Successfully opened dynamic library libcudart.so.10.0
2020-06-24 20:29:13.988711: I tensorflow/stream_executor/platform/default/dso_loader.cc:42] Successfully opened dynamic library libcublas.so.10.0
2020-06-24 20:29:14.002012: I tensorflow/stream_executor/platform/default/dso_loader.cc:42] Successfully opened dynamic library libcufft.so.10.0
2020-06-24 20:29:14.006381: I tensorflow/stream_executor/platform/default/dso_loader.cc:42] Successfully opened dynamic library libcurand.so.10.0
2020-06-24 20:29:14.038179: I tensorflow/stream_executor/platform/default/dso_loader.cc:42] Successfully opened dynamic library libcusolver.so.10.0
2020-06-24 20:29:14.057922: I tensorflow/stream_executor/platform/default/dso_loader.cc:42] Successfully opened dynamic library libcusparse.so.10.0
2020-06-24 20:29:14.114149: I tensorflow/stream_executor/platform/default/dso_loader.cc:42] Successfully opened dynamic library libcudnn.so.7
2020-06-24 20:29:14.114248: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:1005] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-06-24 20:29:14.115060: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:1005] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-06-24 20:29:14.115765: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1763] Adding visible gpu devices: 0
2020-06-24 20:29:14.116472: I tensorflow/stream_executor/platform/default/dso_loader.cc:42] Successfully opened dynamic library libcudart.so.10.0
2020-06-24 20:29:14.118350: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1181] Device interconnect StreamExecutor with strength 1 edge matrix:
2020-06-24 20:29:14.118378: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1187]      0 
2020-06-24 20:29:14.118386: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1200] 0:   N 
2020-06-24 20:29:14.119144: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:1005] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-06-24 20:29:14.119963: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:1005] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-06-24 20:29:14.120610: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1326] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 4889 MB memory) -> physical GPU (device: 0, name: GeForce GTX 1060 6GB, pci bus id: 0000:01:00.0, compute capability: 6.1)
WARNING:tensorflow:From /home/mine/catkin_ws/src/PROJECT/project6_3/src/actor_network_bn.py:75: The name tf.placeholder is deprecated. Please use tf.compat.v1.placeholder instead.

WARNING:tensorflow:From /home/mine/catkin_ws/src/PROJECT/project6_3/src/actor_network_bn.py:177: The name tf.random_uniform is deprecated. Please use tf.random.uniform instead.

WARNING:tensorflow:From /home/mine/catkin_ws/src/PROJECT/project6_3/src/actor_network_bn.py:178: The name tf.get_variable is deprecated. Please use tf.compat.v1.get_variable instead.

Anyone who knows what this message means?

Am I using GPU, properly? (When I checked how much my GPU was being used with following commands

$ watch -d -n 0.5 nvidia-smi

it always returns 1407 Mib/ 6000 Mib of usage.)

And additionally, should I modify my codes following WARNING messages? (it works well currently on some level)

Thanks in advance. :)

Upvotes: 1

Views: 83

Answers (1)

user11530462
user11530462

Reputation:

Am I using Tensorflow GPU ?

If you have executed below code and if it returns device_type='GPU' means, there is no issue with Tensorflow GPU installation and you are good to use.

import tensorflow as tf
tf.config.list_physical_devices('GPU')

Output:

[PhysicalDevice(name='/physical_device:GPU:0', device_type='GPU')]

2020-06-24 20:29:14.120610: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1326] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 4889 MB memory) -> physical GPU (device: 0, name: GeForce GTX 1060 6GB, pci bus id: 0000:01:00.0, compute capability: 6.1)

If you have check above log from stack trace Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 4889 MB memory), that means you are using GPU.

2020-06-24 20:29:13.961867: I tensorflow/stream_executor/platform/default/dso_loader.cc:42] Successfully opened dynamic library libcudart.so.10.0
2020-06-24 20:29:13.988711: I tensorflow/stream_executor/platform/default/dso_loader.cc:42] Successfully opened dynamic library libcublas.so.10.0
2020-06-24 20:29:14.002012: I tensorflow/stream_executor/platform/default/dso_loader.cc:42] Successfully opened dynamic library libcufft.so.10.0
2020-06-24 20:29:14.006381: I tensorflow/stream_executor/platform/default/dso_loader.cc:42] Successfully opened dynamic library libcurand.so.10.0
2020-06-24 20:29:14.038179: I tensorflow/stream_executor/platform/default/dso_loader.cc:42] Successfully opened dynamic library libcusolver.so.10.0

They are just the Information as they are prefixed with I. If there would be any warnings then it prefixed with W and for error they prefixed with E.

And you are seeing WARNING:tensorflow: they are conveying you to replace modules with newer one(i.e compat) since those are deprecated and to execute same code in TF2.x.

Upvotes: 0

Related Questions