Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Package request: tensorflow-text #19578

Open
2 tasks done
mfansler opened this issue Jul 7, 2022 · 6 comments · May be fixed by #29008
Open
2 tasks done

Package request: tensorflow-text #19578

mfansler opened this issue Jul 7, 2022 · 6 comments · May be fixed by #29008

Comments

@mfansler
Copy link
Member

mfansler commented Jul 7, 2022

Package name

tensorflow-text

Package version

Newest

Package website

Website: https://www.tensorflow.org/text
Repository: https://github.com/tensorflow/text

Package availability

PyPI: https://pypi.org/project/tensorflow-text/

Additional comments

Multiple users have attempted and failed to build this in the past: #17613, #16321, #16161, #16215. That is, expect non-trivial workload.

Package is not available

  • The package is not available on conda-forge.

No previous issues or open PRs

  • No previous issue exists and no PR has been opened.
@ngam
Copy link
Contributor

ngam commented Jul 8, 2022

It's a rough cookie this one along with anything tensorflow... But we will eventually get to it. If someone wants to give this a go, or any of these tensorflow- derivatives, an important first step is to fully understand the complex build process in https://github.com/conda-forge/tensorflow-feedstock

A likely best solution is to have a maintainer from tensorflow-feedstock have a go at this at some point...

@Anselmoo
Copy link
Member

Anselmoo commented Feb 2, 2023

It's a rough cookie this one along with anything tensorflow... But we will eventually get to it. If someone wants to give this a go, or any of these tensorflow- derivatives, an important first step is to fully understand the complex build process in https://github.com/conda-forge/tensorflow-feedstock

A likely best solution is to have a maintainer from tensorflow-feedstock have a go at this at some point...

A PR was initialised conda-forge/tensorflow-feedstock#294 (comment) however, it further increase the compiling time.

@Anselmoo
Copy link
Member

Anselmoo commented Feb 2, 2023

This might be a workaround ...

IF "%PY_VER%"=="2.7" (
	%PYTHON% -m pip install --no-deps https://pypi.org/packages/cp27/c/catboost/catboost-%PKG_VERSION%-cp27-none-win_amd64.whl
)

IF "%PY_VER%"=="3.5" (
	%PYTHON% -m pip install --no-deps https://pypi.org/packages/cp35/c/catboost/catboost-%PKG_VERSION%-cp35-none-win_amd64.whl
)

IF "%PY_VER%"=="3.6" (
	%PYTHON% -m pip install --no-deps https://pypi.org/packages/cp36/c/catboost/catboost-%PKG_VERSION%-cp36-none-win_amd64.whl
)

IF "%PY_VER%"=="3.7" (
	%PYTHON% -m pip install --no-deps https://pypi.org/packages/cp37/c/catboost/catboost-%PKG_VERSION%-cp37-none-win_amd64.whl
)

IF "%PY_VER%"=="3.8" (
	%PYTHON% -m pip install --no-deps https://pypi.org/packages/cp38/c/catboost/catboost-%PKG_VERSION%-cp38-none-win_amd64.whl
)

IF "%PY_VER%"=="3.9" (
	%PYTHON% -m pip install --no-deps https://pypi.org/packages/cp39/c/catboost/catboost-%PKG_VERSION%-cp39-none-win_amd64.whl
)

IF "%PY_VER%"=="3.10" (
	%PYTHON% -m pip install --no-deps https://pypi.org/packages/cp310/c/catboost/catboost-%PKG_VERSION%-cp310-none-win_amd64.whl
)

if errorlevel 1 exit 1

https://github.com/conda-forge/catboost-feedstock/blob/036182496875e787b7d3cbb1c7e051c1c05b395b/recipe/bld.bat

@ehfd
Copy link
Member

ehfd commented Jan 28, 2025

In the era where keras-nlp became keras-hub... I guess not a big process yet?

@ehfd
Copy link
Member

ehfd commented Jan 29, 2025

@hmaarrfk @ngam Is there any way to inherit the build artifacts of tensorflow from conda in order to build tensorflow-text?
https://github.com/tensorflow/text#installation

Following up from conda-forge/tensorflow-feedstock#294 (comment).

After all, conda-forge/tensorflow-feedstock#206 (from conda-forge/tensorflow-feedstock#276 (comment)) is still incomplete.

@hmaarrfk
Copy link
Contributor

You can only depend on conda-forge to get recipes merged.

You may add temporary channels to troubleshoot things. But at the end, they must be removed.

@ehfd ehfd linked a pull request Feb 2, 2025 that will close this issue
10 tasks
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Development

Successfully merging a pull request may close this issue.

5 participants