Hi all,
as many of you I was very excited to learn about TensorFlow Decision Forests, and especially that its API lies within Keras. However, I am unable to install it. My pip install returns the following message:
“ERROR: Could not find a version that satisfies the requirement tensorflow_decision_forests (from versions: none)
ERROR: No matching distribution found for tensorflow_decision_forests”
Would you have any pointers here that could help me? Thanks very much in advance.
Hi Mathieu,
Thanks for the response. Indeed I hadn’t noted that it works only in Linux (I’m using MacOS). Is there any expectation that it will be ported to other systems as well?
Many thanks for your kind sharing of this troubleshooting guide. As Mathieu said, it only works in Linux (and I am using MacOS). But when I use it in Colab, if I find any problem I will definitely refer to the Yggdrasil troubleshooting guide as you indicated.
Thanks for the idea, @ashutosh1919. Actually I have started using it in Colab. For the benefit of others who might be running in cloud environments where machine may or may not be linux, I am posting below a code snippet that checks if the machine is linux, and if it s it installs tfdf.
from sys import platform
if platform != "linux" and platform != "linux":
print("'tensorflow_decision_forests' is currently only available for Linux.")
try:
import tensorflow_decision_forests
except ModuleNotFoundError:
!pip install tensorflow_decision_forests
import tensorflow_decision_forests as tfdf
I too have just spent considerable time building a new TF 2.5 cuda* environment for Windows (10, 64-bit) only to discover that the TFDF installation fails because there is “no matching distribution” (because, as I discover here, it is currently Linux only).