Thank You NVIDIA - Everything is working fine on wsl2 and windows 10

Well Done. Good Job.

On Windows 10 -

(base) C:\WINDOWS\system32>pip show numpy
Name: numpy
Version: 1.20.2
Summary: NumPy is the fundamental package for array computing with Python.
Home-page: https://www.numpy.org
Author: Travis E. Oliphant et al.
Author-email: None
License: BSD
Location: c:\programdata\anaconda3\lib\site-packages
Requires:
Required-by: tifffile, tensorflow, tensorboard, tables, statsmodels, spherical, seaborn, scipy, scikit-learn, scikit-image, quaternionic, PyWavelets, pytools, pyrr, pyopencl, pyerfa, patsy, pandas, opt-einsum, numpy-quaternion, numexpr, numba, mkl-random, mkl-fft, matplotlib, Keras-Preprocessing, imageio, imagecodecs, h5py, cupy-cuda112, Bottleneck, bokeh, blis, bkcharts, astropy

everything is installed and working now fine

On WSL2

mabd@LAPTOP-T8DQ9UK0:~$ pip3 show numpy
Name: numpy
Version: 1.19.5
Summary: NumPy is the fundamental package for array computing with Python.
Home-page: https://www.numpy.org
Author: Travis E. Oliphant et al.
Author-email: None
License: BSD
Location: /home/mabd/.local/lib/python3.6/site-packages
Requires:
Required-by: torchvision, torch, tensorflow-gpu, tensorboard, scipy, pytools, opt-einsum, numba, Keras, Keras-Preprocessing, h5py, cupy-cuda112

everything: tensorflow- pycuda -cupy - tensorflow+directml
and torch torchvision as you can see

The Key Solution use cudatools 11.1 in wsl2 not cudatools 11.3
here my
mabd@LAPTOP-T8DQ9UK0:~$ ./deviceQuery
./deviceQuery Starting…

CUDA Device Query (Runtime API) version (CUDART static linking)

Detected 1 CUDA Capable device(s)

Device 0: “NVIDIA GeForce GTX 1660 Ti with Max-Q Design”
CUDA Driver Version / Runtime Version 11.3 / 11.1
CUDA Capability Major/Minor version number: 7.5
Total amount of global memory: 6144 MBytes (6442450944 bytes)
(024) Multiprocessors, (064) CUDA Cores/MP: 1536 CUDA Cores
GPU Max Clock rate: 1335 MHz (1.34 GHz)
Memory Clock rate: 6001 Mhz
Memory Bus Width: 192-bit
L2 Cache Size: 1572864 bytes
Maximum Texture Dimension Size (x,y,z) 1D=(131072), 2D=(131072, 65536), 3D=(16384, 16384, 16384)
Maximum Layered 1D Texture Size, (num) layers 1D=(32768), 2048 layers
Maximum Layered 2D Texture Size, (num) layers 2D=(32768, 32768), 2048 layers
Total amount of constant memory: 65536 bytes
Total amount of shared memory per block: 49152 bytes
Total shared memory per multiprocessor: 65536 bytes
Total number of registers available per block: 65536
Warp size: 32
Maximum number of threads per multiprocessor: 1024
Maximum number of threads per block: 1024
Max dimension size of a thread block (x,y,z): (1024, 1024, 64)
Max dimension size of a grid size (x,y,z): (2147483647, 65535, 65535)
Maximum memory pitch: 2147483647 bytes
Texture alignment: 512 bytes
Concurrent copy and kernel execution: Yes with 2 copy engine(s)
Run time limit on kernels: Yes
Integrated GPU sharing Host Memory: No
Support host page-locked memory mapping: Yes
Alignment requirement for Surfaces: Yes
Device has ECC support: Disabled
Device supports Unified Addressing (UVA): Yes
Device supports Managed Memory: Yes
Device supports Compute Preemption: Yes
Supports Cooperative Kernel Launch: Yes
Supports MultiDevice Co-op Kernel Launch: Yes
Device PCI Domain ID / Bus ID / location ID: 0 / 1 / 0
Compute Mode:
< Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) >

deviceQuery, CUDA Driver = CUDART, CUDA Driver Version = 11.3, CUDA Runtime Version = 11.1, NumDevs = 1
Result = PASS
mabd@LAPTOP-T8DQ9UK0:~$

mabd@LAPTOP-T8DQ9UK0:~$ ./concurrentKernels
[./concurrentKernels] - Starting…
GPU Device 0: “Turing” with compute capability 7.5

Detected Compute SM 7.5 hardware with 24 multi-processors
Expected time for serial execution of 8 kernels = 0.080s
Expected time for concurrent execution of 8 kernels = 0.010s
Measured time for sample = 0.013s
Test passed
mabd@LAPTOP-T8DQ9UK0:~$

mabd@LAPTOP-T8DQ9UK0:~$ ./bandwidthTest
[CUDA Bandwidth Test] - Starting…
Running on…

Device 0: NVIDIA GeForce GTX 1660 Ti with Max-Q Design
Quick Mode

Host to Device Bandwidth, 1 Device(s)
PINNED Memory Transfers
Transfer Size (Bytes) Bandwidth(GB/s)
32000000 6.7

Device to Host Bandwidth, 1 Device(s)
PINNED Memory Transfers
Transfer Size (Bytes) Bandwidth(GB/s)
32000000 6.6

Device to Device Bandwidth, 1 Device(s)
PINNED Memory Transfers
Transfer Size (Bytes) Bandwidth(GB/s)
32000000 251.4

Result = PASS

NOTE: The CUDA Samples are not meant for performance measurements. Results may vary when GPU Boost is enabled.

Now, we can talk. I will make some benchmarks and see the difference.
Stay Tuned

i am using Windows Build 21387
my python on windows
(base) C:\WINDOWS\system32>python --version
Python 3.8.10

on wsl 2

mabd@LAPTOP-T8DQ9UK0:~$ python3 --version
Python 3.6.9

here is nvidia-smi from windows10

(base) C:\WINDOWS\system32>nvidia-smi
Sat May 29 16:32:33 2021
±----------------------------------------------------------------------------+
| NVIDIA-SMI 470.14 Driver Version: 470.14 CUDA Version: 11.3 |
|-------------------------------±---------------------±---------------------+
| GPU Name TCC/WDDM | Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |
| | | MIG M. |
|===============================+======================+======================|
| 0 NVIDIA GeForce … WDDM | 00000000:01:00.0 Off | N/A |
| N/A 49C P8 3W / N/A | 257MiB / 6144MiB | 0% Default |
| | | N/A |
±------------------------------±---------------------±---------------------+

±----------------------------------------------------------------------------+
| Processes: |
| GPU GI CI PID Type Process name GPU Memory |
| ID ID Usage |
|=============================================================================|
| 0 N/A N/A 1784 C+G Insufficient Permissions N/A |
| 0 N/A N/A 6412 C+G …ge\Application\msedge.exe N/A |
| 0 N/A N/A 8588 C+G …artMenuExperienceHost.exe N/A |
| 0 N/A N/A 8688 C+G …5n1h2txyewy\SearchApp.exe N/A |
±----------------------------------------------------------------------------+

and here is again from wsl2

mabd@LAPTOP-T8DQ9UK0:~$ ./deviceQuery
./deviceQuery Starting…

CUDA Device Query (Runtime API) version (CUDART static linking)

Detected 1 CUDA Capable device(s)

Device 0: “NVIDIA GeForce GTX 1660 Ti with Max-Q Design”
CUDA Driver Version / Runtime Version 11.3 / 11.1
CUDA Capability Major/Minor version number: 7.5
Total amount of global memory: 6144 MBytes (6442450944 bytes)
(024) Multiprocessors, (064) CUDA Cores/MP: 1536 CUDA Cores
GPU Max Clock rate: 1335 MHz (1.34 GHz)
Memory Clock rate: 6001 Mhz
Memory Bus Width: 192-bit
L2 Cache Size: 1572864 bytes
Maximum Texture Dimension Size (x,y,z) 1D=(131072), 2D=(131072, 65536), 3D=(16384, 16384, 16384)
Maximum Layered 1D Texture Size, (num) layers 1D=(32768), 2048 layers
Maximum Layered 2D Texture Size, (num) layers 2D=(32768, 32768), 2048 layers
Total amount of constant memory: 65536 bytes
Total amount of shared memory per block: 49152 bytes
Total shared memory per multiprocessor: 65536 bytes
Total number of registers available per block: 65536
Warp size: 32
Maximum number of threads per multiprocessor: 1024
Maximum number of threads per block: 1024
Max dimension size of a thread block (x,y,z): (1024, 1024, 64)
Max dimension size of a grid size (x,y,z): (2147483647, 65535, 65535)
Maximum memory pitch: 2147483647 bytes
Texture alignment: 512 bytes
Concurrent copy and kernel execution: Yes with 2 copy engine(s)
Run time limit on kernels: Yes
Integrated GPU sharing Host Memory: No
Support host page-locked memory mapping: Yes
Alignment requirement for Surfaces: Yes
Device has ECC support: Disabled
Device supports Unified Addressing (UVA): Yes
Device supports Managed Memory: Yes
Device supports Compute Preemption: Yes
Supports Cooperative Kernel Launch: Yes
Supports MultiDevice Co-op Kernel Launch: Yes
Device PCI Domain ID / Bus ID / location ID: 0 / 1 / 0
Compute Mode:
< Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) >

deviceQuery, CUDA Driver = CUDART, CUDA Driver Version = 11.3, CUDA Runtime Version = 11.1, NumDevs = 1
Result = PASS
mabd@LAPTOP-T8DQ9UK0:~$

as you can see

CUDA Driver = CUDART, CUDA Driver Version = 11.3, CUDA Runtime Version = 11.1, NumDevs = 1
the 11.3 from thw windows and the 11.1 from the wsl2

desktop dokker is buggy and makes problems with wsl+CUDA
I installed dokker on wsl not the desktop and service start/stop is working
but i didn’t use it for two reasons
1- i have torch + torchvision + nvidia card i.e i don’t need simulator
2- i am ruuning out of my monthly internet quota, sorry guys.

now i shall write something in details it is not that straightforward
i had to compile openssl + cmake + llvm from sources
and i had to correct some errors of nvcc confilcts
–threads not defined not to mention the ryzen/amd – nvidia early conflict
but now the nvidia card is working fine and the memory can be used without going first
through the amd gpu/apu.
ciao, ciao.

usually, i have three running at least three open windows
1- wsl2
2- command prompt from windows 10 running nvidia-smi for testing what happens
3- windows 10 Task Manger GPU “nVIDIA” pannel
i use the windows 10 to calibrate the GPU usage and the GPU memory.

Here how a first run from TensorFlow from inside WSL2 UBUNTU 18.04

mabd@LAPTOP-T8DQ9UK0:~$ python3
Python 3.6.9 (default, Jan 26 2021, 15:33:00)
[GCC 8.4.0] on linux
Type “help”, “copyright”, “credits” or “license” for more information.

import tensorflow-gpu
File “”, line 1
import tensorflow-gpu
^
SyntaxError: invalid syntax

import tensorflow as tf
2021-05-31 01:55:54.689597: I tensorflow/stream_executor/platform/default/dso_loader.cc:53] Successfully opened dynamic library libcudart.so.11.0

print(tf.constant(‘Hello’))
2021-05-31 01:57:05.413109: I tensorflow/stream_executor/platform/default/dso_loader.cc:53] Successfully opened dynamic library libcuda.so.1
2021-05-31 01:57:05.646824: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:923] could not open file to read NUMA node: /sys/bus/pci/devices/0000:01:00.0/numa_node
Your kernel may have been built without NUMA support.
2021-05-31 01:57:05.647549: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1733] Found device 0 with properties:
pciBusID: 0000:01:00.0 name: NVIDIA GeForce GTX 1660 Ti with Max-Q Design computeCapability: 7.5
coreClock: 1.335GHz coreCount: 24 deviceMemorySize: 6.00GiB deviceMemoryBandwidth: 268.26GiB/s
2021-05-31 01:57:05.647718: I tensorflow/stream_executor/platform/default/dso_loader.cc:53] Successfully opened dynamic library libcudart.so.11.0
2021-05-31 01:57:05.661688: I tensorflow/stream_executor/platform/default/dso_loader.cc:53] Successfully opened dynamic library libcublas.so.11
2021-05-31 01:57:05.661800: I tensorflow/stream_executor/platform/default/dso_loader.cc:53] Successfully opened dynamic library libcublasLt.so.11
2021-05-31 01:57:05.664940: I tensorflow/stream_executor/platform/default/dso_loader.cc:53] Successfully opened dynamic library libcufft.so.10
2021-05-31 01:57:05.665421: I tensorflow/stream_executor/platform/default/dso_loader.cc:53] Successfully opened dynamic library libcurand.so.10
2021-05-31 01:57:05.671247: I tensorflow/stream_executor/platform/default/dso_loader.cc:53] Successfully opened dynamic library libcusolver.so.11
2021-05-31 01:57:05.673878: I tensorflow/stream_executor/platform/default/dso_loader.cc:53] Successfully opened dynamic library libcusparse.so.11
2021-05-31 01:57:05.674023: W tensorflow/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library ‘libcudnn.so.8’; dlerror: libcudnn.so.8: cannot open shared object file: No such file or directory
2021-05-31 01:57:05.674072: W tensorflow/core/common_runtime/gpu/gpu_device.cc:1766] Cannot dlopen some GPU libraries. Please make sure the missing libraries mentioned above are installed properly if you would like to use GPU. Follow the guide at GPU support  |  TensorFlow for how to download and setup the required libraries for your platform.
Skipping registering GPU devices…
2021-05-31 01:57:05.675334: I tensorflow/core/platform/cpu_feature_guard.cc:142] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations: AVX2 FMA
To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
2021-05-31 01:57:05.679373: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1258] Device interconnect StreamExecutor with strength 1 edge matrix:
2021-05-31 01:57:05.679426: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1264]
tf.Tensor(b’Hello’, shape=(), dtype=string)

print(tf.version)
2.5.0

print(tf.add([1.0,2.0],[3.0,4.0]))
tf.Tensor([4. 6.], shape=(2,), dtype=float32)

quit()
mabd@LAPTOP-T8DQ9UK0:~$

as you can see all libraries are installed except ‘libcudnn.so.8’ i know, i know it needs to be intalled
seperately


and here my tensorflow-gpu version

mabd@LAPTOP-T8DQ9UK0:~$ pip show tensorflow-gpu
Name: tensorflow-gpu
Version: 2.5.0
Summary: TensorFlow is an open source machine learning framework for everyone.
Home-page: https://www.tensorflow.org/
Author: Google Inc.
Author-email: packages@tensorflow.org
License: Apache 2.0
Location: /home/mabd/.local/lib/python3.6/site-packages
Requires: gast, tensorflow-estimator, flatbuffers, astunparse, protobuf, h5py, opt-einsum, six, typing-extensions, google-pasta, absl-py, termcolor, wrapt, numpy, wheel, keras-preprocessing, tensorboard, keras-nightly, grpcio
Required-by:
mabd@LAPTOP-T8DQ9UK0:~$


it is ok thhis the last TensFlow
here some additional installed ML package

mabd@LAPTOP-T8DQ9UK0:~$ pip show theano
Name: Theano
Version: 1.0.5
Summary: Optimizing compiler for evaluating mathematical expressions on CPUs and GPUs.
Home-page: http://deeplearning.net/software/theano/
Author: LISA laboratory, University of Montreal
Author-email: theano-dev@googlegroups.com
License: BSD
Location: /usr/local/lib/python3.6/dist-packages
Requires: scipy, six, numpy
Required-by:
mabd@LAPTOP-T8DQ9UK0:~$


i needed to compile numba from sources :
that is where i had to install openssl - cmake - llvm

mabd@LAPTOP-T8DQ9UK0:~$ pip show numba
Name: numba
Version: 0.48.0
Summary: compiling Python code using LLVM
Home-page: http://numba.github.com
Author: Anaconda, Inc.
Author-email: numba-users@continuum.io
License: BSD
Location: /home/mabd/.local/lib/python3.6/site-packages
Requires: llvmlite, setuptools, numpy
Required-by: cusignal
mabd@LAPTOP-T8DQ9UK0:~$


i needed numba for cusignal

mabd@LAPTOP-T8DQ9UK0:~$ pip show cusignal
Name: cusignal
Version: 21.6.0a0+29.g5138415
Summary: cuSignal - GPU Signal Processing
Home-page: GitHub - rapidsai/cusignal: cuSignal - RAPIDS Signal Processing Library
Author: NVIDIA Corporation
Author-email: None
License: Apache 2.0
Location: /usr/local/lib/python3.6/dist-packages/cusignal-21.6.0a0+29.g5138415-py3.6.egg
Requires: numba
Required-by:


and then to install spacy

mabd@LAPTOP-T8DQ9UK0:~$ pip show spacy
Name: spacy
Version: 3.0.6
Summary: Industrial-strength Natural Language Processing (NLP) in Python
Home-page: https://spacy.io
Author: Explosion
Author-email: contact@explosion.ai
License: MIT
Location: /home/mabd/.local/lib/python3.6/site-packages
Requires: spacy-legacy, setuptools, cymem, thinc, preshed, wasabi, srsly, tqdm, pydantic, typing-extensions, typer, murmurhash, blis, numpy, requests, pathy, jinja2, packaging, catalogue
Required-by: spacy-transformers
mabd@LAPTOP-T8DQ9UK0:~$


i had to install cupy-cuda112
and here cuda 11.2 was important

mabd@LAPTOP-T8DQ9UK0:~$ pip show cupy-cuda112
Name: cupy-cuda112
Version: 8.6.0
Summary: CuPy: A NumPy-compatible array library accelerated by CUDA
Home-page: https://cupy.dev/
Author: Seiya Tokui
Author-email: tokui@preferred.jp
License: MIT License
Location: /home/mabd/.local/lib/python3.6/site-packages
Requires: numpy, fastrlock
Required-by:
mabd@LAPTOP-T8DQ9UK0:~$

spacy needed
Name: cupy-cuda112
Version: 8.6.0
i had to downgrade from initial
Name: cupy-cuda112
Version: 9.0


here mu full python 3.6 dependent tree

mabd@LAPTOP-T8DQ9UK0:~$ pipdeptree
asn1crypto==0.24.0
Automat==0.6.0
blinker==1.4
cloud-init==21.1
colorama==0.3.7
command-not-found==0.3
configobj==5.0.6

  • six [required: Any, installed: 1.15.0]
    constantly==15.1.0
    cryptography==2.1.4
    cupy-cuda112==8.6.0
  • fastrlock [required: >=0.3, installed: 0.6]
  • numpy [required: >=1.15, installed: 1.19.5]
    cusignal==21.6.0a0+29.g5138415
  • numba [required: Any, installed: 0.48.0]
    • llvmlite [required: >=0.31.0dev0,<0.32.0, installed: 0.31.0]
    • numpy [required: >=1.15, installed: 1.19.5]
    • setuptools [required: Any, installed: 57.0.0]
      Cython==0.29.23
      distro-info==0.18ubuntu0.18.04.1
      flake8==3.5.0
  • mccabe [required: >=0.6.0,<0.7.0, installed: 0.6.1]
  • pycodestyle [required: >=2.0.0,<2.4.0, installed: 2.3.1]
  • pyflakes [required: >=1.5.0,<1.7.0, installed: 1.6.0]
    httplib2==0.9.2
    hyperlink==17.3.1
    hypothesis==6.13.10
  • attrs [required: >=19.2.0, installed: 21.2.0]
  • sortedcontainers [required: >=2.1.0,<3.0.0, installed: 2.4.0]
    incremental==16.10.1
    jsonpatch==1.16
    jsonpointer==1.10
    jsonschema==2.6.0
    Keras==2.4.3
  • h5py [required: Any, installed: 3.1.0]
    • cached-property [required: Any, installed: 1.5.2]
    • numpy [required: >=1.12, installed: 1.19.5]
  • numpy [required: >=1.9.1, installed: 1.19.5]
  • pyyaml [required: Any, installed: 3.12]
  • scipy [required: >=0.14, installed: 1.5.4]
    • numpy [required: >=1.14.5, installed: 1.19.5]
      keyring==10.6.0
  • secretstorage [required: Any, installed: 2.3.1]
    keyrings.alt==3.0
    language-selector==0.1
    ml-datasets==0.2.0
  • catalogue [required: >=0.2.0,<3.0.0, installed: 2.0.4]
    • typing-extensions [required: >=3.6.4, installed: 3.7.4.3]
    • zipp [required: >=0.5, installed: 3.4.1]
  • numpy [required: >=1.7.0, installed: 1.19.5]
  • srsly [required: >=1.0.1,<3.0.0, installed: 2.4.1]
    • catalogue [required: >=2.0.1,<2.1.0, installed: 2.0.4]
      • typing-extensions [required: >=3.6.4, installed: 3.7.4.3]
      • zipp [required: >=0.5, installed: 3.4.1]
  • tqdm [required: >=4.10.0,<5.0.0, installed: 4.61.0]
    mock==2.0.0
  • pbr [required: >=0.11, installed: 5.6.0]
  • six [required: >=1.9, installed: 1.15.0]
    netifaces==0.10.4
    PAM==0.4.2
    pexpect==4.2.1
    pipdeptree==2.0.0
  • pip [required: >=6.0.0, installed: 21.1.2]
    pycrypto==2.6.1
    pycuda==2021.1
  • appdirs [required: >=1.4.0, installed: 1.4.4]
  • mako [required: Any, installed: 1.1.4]
    • MarkupSafe [required: >=0.9.2, installed: 2.0.1]
  • pytools [required: >=2011.2, installed: 2021.2.6]
    • appdirs [required: >=1.4.0, installed: 1.4.4]
    • dataclasses [required: >=0.7, installed: 0.8]
    • numpy [required: >=1.6.0, installed: 1.19.5]
      Pygments==2.2.0
      pygobject==3.26.1
      PyJWT==1.5.3
      pyOpenSSL==17.5.0
      pyserial==3.4
      pytest-timeout==1.4.2
  • pytest [required: >=3.6.0, installed: 6.2.4]
    • attrs [required: >=19.2.0, installed: 21.2.0]
    • importlib-metadata [required: >=0.12, installed: 4.3.1]
      • typing-extensions [required: >=3.6.4, installed: 3.7.4.3]
      • zipp [required: >=0.5, installed: 3.4.1]
    • iniconfig [required: Any, installed: 1.1.1]
    • packaging [required: Any, installed: 20.9]
      • pyparsing [required: >=2.0.2, installed: 2.4.7]
    • pluggy [required: >=0.12,<1.0.0a1, installed: 0.13.1]
      • importlib-metadata [required: >=0.12, installed: 4.3.1]
        • typing-extensions [required: >=3.6.4, installed: 3.7.4.3]
        • zipp [required: >=0.5, installed: 3.4.1]
    • py [required: >=1.8.2, installed: 1.10.0]
    • toml [required: Any, installed: 0.10.2]
      python-apt==1.6.5+ubuntu0.5
      python-debian==0.1.32
      pyxdg==0.25
      requests-unixsocket==0.1.5
      service-identity==16.0.0
      sos==4.1
      spacy-lookups-data==1.0.0
  • setuptools [required: Any, installed: 57.0.0]
    spacy-transformers==1.0.2
  • dataclasses [required: >=0.6,<1.0, installed: 0.8]
  • spacy [required: >=3.0.0,<3.1.0, installed: 3.0.6]
    • blis [required: >=0.4.0,<0.8.0, installed: 0.7.4]
      • numpy [required: >=1.15.0, installed: 1.19.5]
    • catalogue [required: >=2.0.3,<2.1.0, installed: 2.0.4]
      • typing-extensions [required: >=3.6.4, installed: 3.7.4.3]
      • zipp [required: >=0.5, installed: 3.4.1]
    • cymem [required: >=2.0.2,<2.1.0, installed: 2.0.5]
    • jinja2 [required: Any, installed: 2.10]
    • murmurhash [required: >=0.28.0,<1.1.0, installed: 1.0.5]
    • numpy [required: >=1.15.0, installed: 1.19.5]
    • packaging [required: >=20.0, installed: 20.9]
      • pyparsing [required: >=2.0.2, installed: 2.4.7]
    • pathy [required: >=0.3.5, installed: 0.5.2]
      • dataclasses [required: >=0.6,<1.0, installed: 0.8]
      • smart-open [required: >=2.2.0,<4.0.0, installed: 3.0.0]
        • requests [required: Any, installed: 2.25.1]
          • certifi [required: >=2017.4.17, installed: 2018.1.18]
          • chardet [required: >=3.0.2,<5, installed: 3.0.4]
          • idna [required: >=2.5,<3, installed: 2.6]
          • urllib3 [required: >=1.21.1,<1.27, installed: 1.22]
      • typer [required: >=0.3.0,<1.0.0, installed: 0.3.2]
        • click [required: >=7.1.1,<7.2.0, installed: 7.1.2]
    • preshed [required: >=3.0.2,<3.1.0, installed: 3.0.5]
      • cymem [required: >=2.0.2,<2.1.0, installed: 2.0.5]
      • murmurhash [required: >=0.28.0,<1.1.0, installed: 1.0.5]
    • pydantic [required: >=1.7.1,<1.8.0, installed: 1.7.4]
      • dataclasses [required: >=0.6, installed: 0.8]
    • requests [required: >=2.13.0,<3.0.0, installed: 2.25.1]
      • certifi [required: >=2017.4.17, installed: 2018.1.18]
      • chardet [required: >=3.0.2,<5, installed: 3.0.4]
      • idna [required: >=2.5,<3, installed: 2.6]
      • urllib3 [required: >=1.21.1,<1.27, installed: 1.22]
    • setuptools [required: Any, installed: 57.0.0]
    • spacy-legacy [required: >=3.0.4,<3.1.0, installed: 3.0.5]
    • srsly [required: >=2.4.1,<3.0.0, installed: 2.4.1]
      • catalogue [required: >=2.0.1,<2.1.0, installed: 2.0.4]
        • typing-extensions [required: >=3.6.4, installed: 3.7.4.3]
        • zipp [required: >=0.5, installed: 3.4.1]
    • thinc [required: >=8.0.3,<8.1.0, installed: 8.0.3]
      • blis [required: >=0.4.0,<0.8.0, installed: 0.7.4]
        • numpy [required: >=1.15.0, installed: 1.19.5]
      • catalogue [required: >=2.0.3,<2.1.0, installed: 2.0.4]
        • typing-extensions [required: >=3.6.4, installed: 3.7.4.3]
        • zipp [required: >=0.5, installed: 3.4.1]
      • contextvars [required: >=2.4,<3, installed: 2.4]
        • immutables [required: >=0.9, installed: 0.15]
      • cymem [required: >=2.0.2,<2.1.0, installed: 2.0.5]
      • dataclasses [required: >=0.6,<1.0, installed: 0.8]
      • murmurhash [required: >=0.28.0,<1.1.0, installed: 1.0.5]
      • numpy [required: >=1.15.0, installed: 1.19.5]
      • preshed [required: >=3.0.2,<3.1.0, installed: 3.0.5]
        • cymem [required: >=2.0.2,<2.1.0, installed: 2.0.5]
        • murmurhash [required: >=0.28.0,<1.1.0, installed: 1.0.5]
      • pydantic [required: >=1.7.1,<1.8.0, installed: 1.7.4]
        • dataclasses [required: >=0.6, installed: 0.8]
      • setuptools [required: Any, installed: 57.0.0]
      • srsly [required: >=2.4.0,<3.0.0, installed: 2.4.1]
        • catalogue [required: >=2.0.1,<2.1.0, installed: 2.0.4]
          • typing-extensions [required: >=3.6.4, installed: 3.7.4.3]
          • zipp [required: >=0.5, installed: 3.4.1]
      • typing-extensions [required: >=3.7.4.1,<4.0.0.0, installed: 3.7.4.3]
      • wasabi [required: >=0.8.1,<1.1.0, installed: 0.8.2]
    • tqdm [required: >=4.38.0,<5.0.0, installed: 4.61.0]
    • typer [required: >=0.3.0,<0.4.0, installed: 0.3.2]
      • click [required: >=7.1.1,<7.2.0, installed: 7.1.2]
    • typing-extensions [required: >=3.7.4,<4.0.0.0, installed: 3.7.4.3]
    • wasabi [required: >=0.8.1,<1.1.0, installed: 0.8.2]
  • spacy-alignments [required: >=0.7.2,<1.0.0, installed: 0.8.3]
  • srsly [required: >=2.4.0,<3.0.0, installed: 2.4.1]
    • catalogue [required: >=2.0.1,<2.1.0, installed: 2.0.4]
      • typing-extensions [required: >=3.6.4, installed: 3.7.4.3]
      • zipp [required: >=0.5, installed: 3.4.1]
  • torch [required: >=1.5.0, installed: 1.6.0+cu101]
    • future [required: Any, installed: 0.18.2]
    • numpy [required: Any, installed: 1.19.5]
  • transformers [required: >=3.4.0,<4.6.0, installed: 4.5.1]
    • dataclasses [required: Any, installed: 0.8]
    • filelock [required: Any, installed: 3.0.12]
    • importlib-metadata [required: Any, installed: 4.3.1]
      • typing-extensions [required: >=3.6.4, installed: 3.7.4.3]
      • zipp [required: >=0.5, installed: 3.4.1]
    • numpy [required: >=1.17, installed: 1.19.5]
    • packaging [required: Any, installed: 20.9]
      • pyparsing [required: >=2.0.2, installed: 2.4.7]
    • regex [required: !=2019.12.17, installed: 2021.4.4]
    • requests [required: Any, installed: 2.25.1]
      • certifi [required: >=2017.4.17, installed: 2018.1.18]
      • chardet [required: >=3.0.2,<5, installed: 3.0.4]
      • idna [required: >=2.5,<3, installed: 2.6]
      • urllib3 [required: >=1.21.1,<1.27, installed: 1.22]
    • sacremoses [required: Any, installed: 0.0.45]
      • click [required: Any, installed: 7.1.2]
      • joblib [required: Any, installed: 1.0.1]
      • regex [required: Any, installed: 2021.4.4]
      • six [required: Any, installed: 1.15.0]
      • tqdm [required: Any, installed: 4.61.0]
    • tokenizers [required: >=0.10.1,<0.11, installed: 0.10.3]
    • tqdm [required: >=4.27, installed: 4.61.0]
      ssh-import-id==5.7
      systemd-python==234
      tensorflow-gpu==2.5.0
  • absl-py [required: ~=0.10, installed: 0.12.0]
    • six [required: Any, installed: 1.15.0]
  • astunparse [required: ~=1.6.3, installed: 1.6.3]
    • six [required: >=1.6.1,<2.0, installed: 1.15.0]
    • wheel [required: >=0.23.0,<1.0, installed: 0.36.2]
  • flatbuffers [required: ~=1.12.0, installed: 1.12]
  • gast [required: ==0.4.0, installed: 0.4.0]
  • google-pasta [required: ~=0.2, installed: 0.2.0]
    • six [required: Any, installed: 1.15.0]
  • grpcio [required: ~=1.34.0, installed: 1.34.1]
    • six [required: >=1.5.2, installed: 1.15.0]
  • h5py [required: ~=3.1.0, installed: 3.1.0]
    • cached-property [required: Any, installed: 1.5.2]
    • numpy [required: >=1.12, installed: 1.19.5]
  • keras-nightly [required: ~=2.5.0.dev, installed: 2.5.0.dev2021032900]
  • keras-preprocessing [required: ~=1.1.2, installed: 1.1.2]
    • numpy [required: >=1.9.1, installed: 1.19.5]
    • six [required: >=1.9.0, installed: 1.15.0]
  • numpy [required: ~=1.19.2, installed: 1.19.5]
  • opt-einsum [required: ~=3.3.0, installed: 3.3.0]
    • numpy [required: >=1.7, installed: 1.19.5]
  • protobuf [required: >=3.9.2, installed: 3.17.1]
    • six [required: >=1.9, installed: 1.15.0]
  • six [required: ~=1.15.0, installed: 1.15.0]
  • tensorboard [required: ~=2.5, installed: 2.5.0]
    • absl-py [required: >=0.4, installed: 0.12.0]
      • six [required: Any, installed: 1.15.0]
    • google-auth [required: >=1.6.3,<2, installed: 1.30.1]
      • cachetools [required: >=2.0.0,<5.0, installed: 4.2.2]
      • pyasn1-modules [required: >=0.2.1, installed: 0.2.1]
      • rsa [required: >=3.1.4,<5, installed: 4.7.2]
        • pyasn1 [required: >=0.1.3, installed: 0.4.2]
      • setuptools [required: >=40.3.0, installed: 57.0.0]
      • six [required: >=1.9.0, installed: 1.15.0]
    • google-auth-oauthlib [required: >=0.4.1,<0.5, installed: 0.4.4]
      • google-auth [required: >=1.0.0, installed: 1.30.1]
        • cachetools [required: >=2.0.0,<5.0, installed: 4.2.2]
        • pyasn1-modules [required: >=0.2.1, installed: 0.2.1]
        • rsa [required: >=3.1.4,<5, installed: 4.7.2]
          • pyasn1 [required: >=0.1.3, installed: 0.4.2]
        • setuptools [required: >=40.3.0, installed: 57.0.0]
        • six [required: >=1.9.0, installed: 1.15.0]
      • requests-oauthlib [required: >=0.7.0, installed: 1.3.0]
        • oauthlib [required: >=3.0.0, installed: 3.1.0]
        • requests [required: >=2.0.0, installed: 2.25.1]
          • certifi [required: >=2017.4.17, installed: 2018.1.18]
          • chardet [required: >=3.0.2,<5, installed: 3.0.4]
          • idna [required: >=2.5,<3, installed: 2.6]
          • urllib3 [required: >=1.21.1,<1.27, installed: 1.22]
    • grpcio [required: >=1.24.3, installed: 1.34.1]
      • six [required: >=1.5.2, installed: 1.15.0]
    • markdown [required: >=2.6.8, installed: 3.3.4]
      • importlib-metadata [required: Any, installed: 4.3.1]
        • typing-extensions [required: >=3.6.4, installed: 3.7.4.3]
        • zipp [required: >=0.5, installed: 3.4.1]
    • numpy [required: >=1.12.0, installed: 1.19.5]
    • protobuf [required: >=3.6.0, installed: 3.17.1]
      • six [required: >=1.9, installed: 1.15.0]
    • requests [required: >=2.21.0,<3, installed: 2.25.1]
      • certifi [required: >=2017.4.17, installed: 2018.1.18]
      • chardet [required: >=3.0.2,<5, installed: 3.0.4]
      • idna [required: >=2.5,<3, installed: 2.6]
      • urllib3 [required: >=1.21.1,<1.27, installed: 1.22]
    • setuptools [required: >=41.0.0, installed: 57.0.0]
    • tensorboard-data-server [required: >=0.6.0,<0.7.0, installed: 0.6.1]
    • tensorboard-plugin-wit [required: >=1.6.0, installed: 1.8.0]
    • werkzeug [required: >=0.11.15, installed: 2.0.1]
      • dataclasses [required: Any, installed: 0.8]
    • wheel [required: >=0.26, installed: 0.36.2]
  • tensorflow-estimator [required: >=2.5.0rc0,<2.6.0, installed: 2.5.0]
  • termcolor [required: ~=1.1.0, installed: 1.1.0]
  • typing-extensions [required: ~=3.7.4, installed: 3.7.4.3]
  • wheel [required: ~=0.35, installed: 0.36.2]
  • wrapt [required: ~=1.12.1, installed: 1.12.1]
    Theano==1.0.5
  • numpy [required: >=1.9.1, installed: 1.19.5]
  • scipy [required: >=0.14, installed: 1.5.4]
    • numpy [required: >=1.14.5, installed: 1.19.5]
  • six [required: >=1.9.0, installed: 1.15.0]
    torchvision==0.7.0+cu101
  • numpy [required: Any, installed: 1.19.5]
  • pillow [required: >=4.1.1, installed: 8.2.0]
  • torch [required: ==1.6.0, installed: 1.6.0+cu101]
    • future [required: Any, installed: 0.18.2]
    • numpy [required: Any, installed: 1.19.5]
      Twisted==17.9.0
      ubuntu-advantage-tools==27.0
      ubuntu-drivers-common==0.0.0
      ufw==0.36
      unattended-upgrades==0.1
      xkit==0.0.0
      zope.interface==4.3.2
      mabd@LAPTOP-T8DQ9UK0:~$

**********************************************8

since i haven’t installed conda i didn;t installed yet tensorflow-directml
i have both of tensorflow and tensorflow-directml (conda env)
in my windows system the tensorflow under python 3.8.8
and the tensorflow-directml under conda env python-3.6

of course my wsl2 python (UBUNTU 18.04)
is 3.6
mabd@LAPTOP-T8DQ9UK0:~$ python3 --version
Python 3.6.9
i.e. i can install tensorflow-directml BUT IT shall BREAK my TENSORFLOW-GPU
there is a conflict they use different tensorboard library…

NOW I NEED a GLOBAL Profilling
that is important
i undrstand we have WSL2->python 3.6 → cuda 11.2 → Microsoft 3D DRiver → Win cuda 11.3
this is the chain.




HI, nVIDIA, Hello
Do you have a GLOBAL profiling or should i do double profiling
one from WSL2 and then another one from Windows 10
and how can i profile WSL2 cuda from Windows 10 cuda.
i am not running cuda inside MSVC-2019 ( i know how to do this)




as a change different type of questions, i hope to hear from you some answers.

Initially, i have written the fastest categorial quaternion/octonion library on the planet but
it is cpu-multithreading. i can beat Mathematica flat by order of magnitude and precession
and i am interested to rewrite it in CUDA - Win10/Linux even if i shall lose the precession but it
should be faster. I have calculed some of my functions to 1.6 Trillion (10^12) terms over a laptop.
now, you can imagine how far i can go with CUDA not to mention the CLOUD


nVIDIA now is a computing company not anymore a gaming company as i heared you CEO in his last nice Talk. We shall have some fun.


i.e i should actually recompile tensorflow-gpu from the source not python3 -m pip install
shall i need nvcc or some special additional nVIDIA stuff of cuda 11.2

you know i have to install openGL separetly and as far as i know (pleas correct me)
there is no yet Vulkan-mesa drive support under WSL2.

OR by kernel it is meant WSL kernel and so no way forget !!

To solve this problem i should download your linux-64 cudnn 1.7 Giga !!!
tell me first they are huge and there is no need to download them and then they fail.

[ok now reedited i hope]

oops - thanx