WARNING:root:Failure to load the inference.so custom c++ tensorflow ops
Opened this issue · 8 comments
Hi,
I have tensorflow 2.15.0 and tensorflow_decision_forests 1.8.1 and according to the website, I know that these are compatible, but for some reason when I try import tensorflow_decision_forests as tfdf I keep getting the following error:
import tensorflow_decision_forests as tfdf
WARNING:root:Failure to load the inference.so custom c++ tensorflow ops. This error is likely caused the version of TensorFlow and TensorFlow Decision Forests are not compatible. Full error:/global/scratch/users/tmatsika/bf_python/lib/python3.11/site-packages/tensorflow_decision_forests/tensorflow/ops/inference/inference.so: cannot open shared object file: No such file or directory
Traceback (most recent call last):
File "", line 1, in
File "/global/scratch/users/tmatsika/bf_python/lib/python3.11/site-packages/tensorflow_decision_forests/init.py", line 64, in
from tensorflow_decision_forests import keras
File "/global/scratch/users/tmatsika/bf_python/lib/python3.11/site-packages/tensorflow_decision_forests/keras/init.py", line 53, in
from tensorflow_decision_forests.keras import core
File "/global/scratch/users/tmatsika/bf_python/lib/python3.11/site-packages/tensorflow_decision_forests/keras/core.py", line 62, in
from tensorflow_decision_forests.keras import core_inference
File "/global/scratch/users/tmatsika/bf_python/lib/python3.11/site-packages/tensorflow_decision_forests/keras/core_inference.py", line 36, in
from tensorflow_decision_forests.tensorflow.ops.inference import api as tf_op
File "/global/scratch/users/tmatsika/bf_python/lib/python3.11/site-packages/tensorflow_decision_forests/tensorflow/ops/inference/api.py", line 179, in
from tensorflow_decision_forests.tensorflow.ops.inference import op
File "/global/scratch/users/tmatsika/bf_python/lib/python3.11/site-packages/tensorflow_decision_forests/tensorflow/ops/inference/op.py", line 15, in
from tensorflow_decision_forests.tensorflow.ops.inference.op_dynamic import *
File "/global/scratch/users/tmatsika/bf_python/lib/python3.11/site-packages/tensorflow_decision_forests/tensorflow/ops/inference/op_dynamic.py", line 24, in
raise e
File "/global/scratch/users/tmatsika/bf_python/lib/python3.11/site-packages/tensorflow_decision_forests/tensorflow/ops/inference/op_dynamic.py", line 21, in
ops = tf.load_op_library(resource_loader.get_path_to_datafile("inference.so"))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/global/scratch/users/tmatsika/bf_python/lib/python3.11/site-packages/tensorflow/python/framework/load_library.py", line 54, in load_op_library
lib_handle = py_tf.TF_LoadLibrary(library_filename)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
tensorflow.python.framework.errors_impl.NotFoundError: /global/scratch/users/tmatsika/bf_python/lib/python3.11/site-packages/tensorflow_decision_forests/tensorflow/ops/inference/inference.so: cannot open shared object file: No such file or directory
I have looked at the other issues and I didn't find a solution that worked for me. I'm using a system with linux arm64 architecture.
I also have the same problem, did you figure it out?
No, I'm hoping that the developers will know how to fix the issue. I think the issue is that TensorFlow Decision Forests doesn't work on linux arm64 architecture, but it does work on Mac arm so I think it's something the developers may need to update so that it works on linux arm64.
I find out that there is an extension called tensorflow in vscode, do you know if it also contains tensorflow decision forest?
Hi, while I believe that it is possible to compile TF-DF on ARM64 Linux, it's not something we've looked into in more detail. I might have a closer look, but absolutely no promises, and I don't think we have the bandwidth to officially support another architecture at this time for TF-DF.
I find out that there is an extension called tensorflow in vscode, do you know if it also contains tensorflow decision forest?
Very likely not.
I still have this problem on 1.9
I have this exact problem but on Windows, have you found a solution to this yet?
I also encountered the same problem, on windows, is there a suitable solution now?
I also encountered the same problem, it turns out to be a compatibility issue. If you install tensorflow using conda and tensorflowjs using pip, this issue still happens even when you have a compatible version.
My workaround is uninstall tensorflow using conda remove tensorflow and reinstalling it using pip install tensorflow==2.15.0 tensorflowjs
Edit: I'm running the command in Ubuntu 22 WSL 2 and miniforge env