davidenunes / tensorflow-wheels Goto Github PK

Hi @davidenunes,

I have a AMD FX-6100 processor with NVIDIA GTX 750Ti Graphics card which is not supported by TensorFlow 2.3.0. If you compile a wheel which supports my configuration please let me know . Thank you :)

I am using a very old cpu
Intel Pentium G3260 (2) @ 3.300GHz

I have a whl for windows but need one for linux.

when i execute code, i have:
Ignoring visible gpu device (device: 0, name: GeForce 940M, pci bus id: 0000:01:00.0, compute capability: 5.0) with Cuda compute capability 5.0. The minimum required Cuda capability is 6.1.

i need compile tensorflow 2.2.0, python 3.8.2, cuda 10.2, cudnn 7.6, nccl 2.6 and Compute Capability 5.0
TF_CUDA_COMPUTE_CAPABILITIES=5.0

Can I request a tensorflow build with minimal requirements:

Tensorflow: 1.13.?
Python: 3.6
Linux x86_64

No GPU
No AVX
No FMA
No MKL

Cheers,

Just wanted to leave a quick note, that the latest package seems to be fedora33 compatible, after installing python 3.8 (default here is python 3.9)

thanks for your work!

Hi,

I don't manage to compile the tensorflow wheel version I need for my computer. Perhaps you could help !

I would like:
Tensorflow 2.1 or newer (newer is better !)
Python 3.7 or newer (newer is better !)
Ubuntu 20.04
CPU-only (no need for CUDA/cuDNN)
All supports SSE4.x, AVX, AVX2, FMA, XLA

Do you think you could do that for me ? Thanks.

Hello,
Thanks for your page.

would it be possible to build for
centos 7.2
TF 2.4
AVX

can sponsor for this.

thanks vm.

Hi,

Any chance you could provide a 2.1 build for amdfam10 (Python 3.7). I don't seem to be able to get it to build as I only have a limited RAM?

Using Ubuntu 18.04 x86_64 as os.

You kindly built 1.14.1 a while back, but I need to update version for a new project.

I downloaded your build for TF 2.3 | GPU | python3.8 | cuda11 |cudnn 8 | XLA,SSE4.X,AVX,AVX2,FMA | Arch Linux x86_64 and installed it. However, when I try and import tensorflow I get the following error

>>> import tensorflow Traceback (most recent call last): File "/home/mfp/virtualenv/py38/lib/python3.8/site-packages/tensorflow/python/pywrap_tensorflow.py", line 64, in <module> from tensorflow.python._pywrap_tensorflow_internal import * ImportError: /lib/x86_64-linux-gnu/libm.so.6: version GLIBC_2.29' not found (required by /home/mfp/virtualenv/py38/lib/python3.8/site-packages/tensorflow/python/_pywrap_tensorflow_internal.so)

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "", line 1, in
File "/home/mfp/virtualenv/py38/lib/python3.8/site-packages/tensorflow/init.py", line 41, in
from tensorflow.python.tools import module_util as _module_util
File "/home/mfp/virtualenv/py38/lib/python3.8/site-packages/tensorflow/python/init.py", line 40, in
from tensorflow.python.eager import context
File "/home/mfp/virtualenv/py38/lib/python3.8/site-packages/tensorflow/python/eager/context.py", line 35, in
from tensorflow.python import pywrap_tfe
File "/home/mfp/virtualenv/py38/lib/python3.8/site-packages/tensorflow/python/pywrap_tfe.py", line 28, in
from tensorflow.python import pywrap_tensorflow
File "/home/mfp/virtualenv/py38/lib/python3.8/site-packages/tensorflow/python/pywrap_tensorflow.py", line 83, in
raise ImportError(msg)
ImportError: Traceback (most recent call last):
File "/home/mfp/virtualenv/py38/lib/python3.8/site-packages/tensorflow/python/pywrap_tensorflow.py", line 64, in
from tensorflow.python._pywrap_tensorflow_internal import *
ImportError: /lib/x86_64-linux-gnu/libm.so.6: version `GLIBC_2.29' not found (required by /home/mfp/virtualenv/py38/lib/python3.8/site-packages/tensorflow/python/_pywrap_tensorflow_internal.so)

Failed to load the native TensorFlow runtime.

See https://www.tensorflow.org/install/errors

for some common reasons and solutions. Include the entire stack trace
above this error message when asking for help.
`

I am running on Ubuntu 18.04 LTS with cuda 11 and cuDNN 8 installed. I looked into this error some and the default glibc is 2.27 for Ubuntu 18, and there does not appear to be any simple method of upgrading to 2.29 unless I move to Ubuntu 19 or greater. I cannot do that though. Any ideas on how to fix this, or is there a different package I should be using? Thanks for your help.

Hey hi, i'm a bio engineering student writing my graduate thesis this year. Since Computer vision is not my main field of study, i'm struggling since a while to make my script work !

I'm working on a remote serve so i can't build the wheel myself.

I would need something to working with FMA AVX2 and AVX512F :

-Debian 9.0
-Python 3.7.5+ (or 3.6.8)
-TF 1.13.1

Would it be possible ?

I need it not only to improve my tensorflow permormances but also to simply make it work.

Faithfully,

Quentin

Download

Hi,
Maybe you can compile tensorflow 2.2 with cuda 11.0 and cuddn 8.0, cuda compute compatibiliti 5.0. And python 3.8 ?

I use manjaro 20. My cpu use microarchitecture skylake.

Not sure why but says it is not a supported wheel on this platform.

Hello,

I need tensorflow 1.13.1 SSE running on Ubuntu 18.04, CPU, and Python 3.6

Hey David, any way you can do the following build on Ubuntu 19?
2.0 | GPU | 3.7 | 10.1 | 7.6 | XLA,SSE,AVX,FMA

I am struggling building it from source on my machine.

Hey mate,

I am trying to reach you on the other socmed platforms (since I don't have a twitter account) but got no luck.
Thus, I am reaching you thru here instead, if you don't mind. :)

I have tried more than 7 times for 4 whole days just to compile TF with GPU support.
I am using GTX 870m (yeah it's a 7 years old laptop haha) and I have tried with Ubuntu 20 and Ubuntu 18 but I could not get these OS to properly install my GPU driver. Thus, I am using Ubuntu 16 instead and its quite stable especially for my very old laptop.

I tried to follow the steps on the TF Docs (build from source, https://www.tensorflow.org/install/source#gpu_support_2) but keep on failing.

I wonder if you can help me, please? I could not compile the wheel :(

Hey, hope you might be able to help me, because I've been banging my head against the wall the whole day.

I want a TF2.3 with CUDA 11.1, CUDNN 8.0.5 and RT 7.2.1., system is fresh ubuntu 20.04
Tried these steps:

• Downloaded the newest versions (CUDA 11.1.1- 455.32 , Cudnn 8.0.5.398 and RT 7.2.1.6)
• After installing everything, trying to do the bazel build fails on cudart (it looks for version 11.1 and SONAME information in libcudart.so.11.1.74 is 11.0) . Overriding the bazel compiler options with the method described here (https://towardsdatascience.com/how-to-compile-tensorflow-2-3-with-cuda-11-1-8cbecffcb8d3) makes the build complete but when I launch TF, it still pops error about being unable to load libcudart.so.11.1. Creating appropriately named softlink pointing at libcudart.so.11.1.74 doesn't help, still the same error at TF import
• wiped the env and tried changing the internal SONAME in libcudart.so.11.1.74 from 11.0 to 11.1 using patchelf. Then the bazel build works without touching the compiler, however after installing the wheel, the same problem remains - TF still can't load libcudart.so.11.1

Any ideas how to deal with that?
Thanks in advance!

It would be great if you got a build up with AVX2 support for any TF version over 2.0. I tried to build it myself but I can't seem to get it working.

got stuck building this configuration due to missing avx and/or cublas10 - problem - so reaching out to you, thanks for your time looking if you can support.
Unfortually due to shared server ubuntu, cuda and cudnn are fixed.

Thanks for sharing these

I'm after a 1.14 build for amdfam10 (python3.6) - do you have one at all?

If not any pointer on how to compile it myself (which compiler options to set) would be helpful.

@davidenunes Hello,
I need a Tensorflow GPU 1.13.2 wheel file for ubuntu 16.04 which supports CUDA 10.0, cuDNN 7.5. (Without SSE,AVX,FMA) with python 3.5