(linear_regression_tf.py:8295): Gdk-CRITICAL **: 21:22:35.526: gdk_cursor_new_for_display: assertion 'GDK_IS_DISPLAY (display)' failed (linear_regression_tf.py:8295): Gdk-CRITICAL **: 21:22:35.522: gdk_cursor_new_for_display: assertion 'GDK_IS_DISPLAY (display)' failed It took 25s, nearly half the time, to complete the training. Running the same script on CPU by specifying the environment variable to hide CUDA: CUDA_VISIBLE_DEVICES="" python3 linear_regression_tf.py 21:21:43.782819: I tensorflow/stream_executor/dso_:153] successfully opened CUDA library libcublas.so.10.0 locally 21:21:42.374771: I tensorflow/core/common_runtime/gpu/gpu_:1115] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 1518 MB memory) -> physical GPU (device: 0, name: NVIDIA Tegra X1, pci bus id: 0000:00:00.0, compute capability: 5.3) 21:21:42.374481: I tensorflow/core/common_runtime/gpu/gpu_:984] Device interconnect StreamExecutor with strength 1 edge matrix: Name: NVIDIA Tegra X1 major: 5 minor: 3 memor圜lockRate(GHz): 0.9216 21:21:41.289806: I tensorflow/core/common_runtime/gpu/gpu_:1433] Found device 0 with properties: 21:21:41.289441: I tensorflow/compiler/xla/service/:168] StreamExecutor device (0): NVIDIA Tegra X1, Compute Capability 5.3 21:21:41.289392: I tensorflow/compiler/xla/service/:161] XLA service 0x3d4e3d50 executing computations on platform CUDA. 21:21:41.289107: I tensorflow/stream_executor/cuda/cuda_gpu_:965] ARM64 does not support NUMA - returning NUMA node zero 21:21:41.202159: I tensorflow/compiler/xla/service/:161] XLA service 0x3f3f6740 executing computations on platform Host. 21:21:41.201561: W tensorflow/core/platform/profile_utils/cpu_:98] Failed to find bogomips in /proc/cpuinfo cannot determine CPU frequency WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow/python/framework/op_def_library.py:263: colocate_with (from ) is deprecated and will be removed in a future version.Ĭolocations handled automatically by placer. (linear_regression_tf.py:8106): Gdk-CRITICAL **: 21:21:38.164: gdk_cursor_new_for_display: assertion 'GDK_IS_DISPLAY (display)' failed Running the code normally on CUDA, the training part took: 43s: (linear_regression_tf.py:8106): Gdk-CRITICAL **: 21:21:38.160: gdk_cursor_new_for_display: assertion 'GDK_IS_DISPLAY (display)' failed Switch Ubuntu to runlevel 3 to free up some memory before the tests. Then I tried to run this sample, which is modified from from the book “TensorFlow for Deep Learning Companion Code” I followed the steps to install TensorFlow: (similar to the steps mentioned in another topic here) Anyone has experience with TensorFlow on Nano? How is the performance? I just tried a simple script and it runs nearly twice as fast on CPU when compared with running it on CUDA.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |