Tensorflow bug(?) with freshly installed JP45

Autonomous Machines Jetson & Embedded Systems Jetson Nano
1

Hi,
hope I’m in the right place to post this.

I re-installed from scratch my Jetson with the latest JetPack 4.5 and Tensorflow 2.3.

The simple Tensorflow sample code (MNIST) works OK.

When I try to run :

import numpy as np
import tensorflow as tf

D = tf.convert_to_tensor(np.array([[1., 2., 3.], [-3., -7., -1.], [0., 5., -2.]]))
print("Shape of D: ", D.shape)

print(tf.linalg.det(D))

I get the following error :
F tensorflow/core/kernels/determinant_op_gpu.cu.cc:136] Non-OK-status: GpuLaunchKernel( DeterminantFromPivotedLUKernel<Scalar, false>, config.block_count, config.thread_per_block, 0, device.stream(), config.virtual_thread_count, n, lu_factor.data(), pivots, nullptr, output.data()) status: Internal: too many resources requested for launch

Can anybody tell me if I’m doing something wrong or if this is some bug for which I should find a workaround ?

3

Hi,

This might be a configure issue.
Could you add the following lines before importing the TensorFlow?

import os
os.environ['CUDA_VISIBLE_DEVICES'] = '1'

Thanks.

4

Hi AastaLLL,
It works!
Thanks.
Could you please explain to me why these lines are needed ?

Full output :

2021-03-02 15:41:12.788215: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcudart.so.10.2
2021-03-02 15:41:17.940475: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcuda.so.1
2021-03-02 15:41:17.952883: E tensorflow/stream_executor/cuda/cuda_driver.cc:314] failed call to cuInit: CUDA_ERROR_NO_DEVICE: no CUDA-capable device is detected
2021-03-02 15:41:17.952983: I tensorflow/stream_executor/cuda/cuda_diagnostics.cc:156] kernel driver does not appear to be running on this host (jetson): /proc/driver/nvidia/version does not exist
2021-03-02 15:41:17.978398: W tensorflow/core/platform/profile_utils/cpu_utils.cc:108] Failed to find bogomips or clock in /proc/cpuinfo; cannot determine CPU frequency
2021-03-02 15:41:17.979513: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x3f9c73b0 initialized for platform Host (this does not guarantee that XLA will be used). Devices:
2021-03-02 15:41:17.979813: I tensorflow/compiler/xla/service/service.cc:176] StreamExecutor device (0): Host, Default Version
Shape of D: (3, 3)
tf.Tensor(-37.99999999999999, shape=(), dtype=float64)