Can't install langchain-docling with uv, only pip on M4 Mac #4
Description
Installing with uv add, python version 3.12.9 fails as below. uv pip install langchain-docling works.
system: Mac M4 Max 64GB
└─[$] git:(main*) uv add langchain-docling
× No solution found when resolving dependencies for split (python_full_version >= '3.12.4' and
│ platform_machine == 'x86_64' and sys_platform == 'darwin'):
╰─▶ Because only the following versions of numpy{platform_machine == 'x86_64' and sys_platform == 'darwin'}
are available:
numpy{platform_machine == 'x86_64' and sys_platform == 'darwin'}<=1.24.4
numpy{platform_machine == 'x86_64' and sys_platform == 'darwin'}==1.25.0
numpy{platform_machine == 'x86_64' and sys_platform == 'darwin'}==1.25.1
numpy{platform_machine == 'x86_64' and sys_platform == 'darwin'}==1.25.2
numpy{platform_machine == 'x86_64' and sys_platform == 'darwin'}==1.26.0
numpy{platform_machine == 'x86_64' and sys_platform == 'darwin'}==1.26.1
numpy{platform_machine == 'x86_64' and sys_platform == 'darwin'}==1.26.2
numpy{platform_machine == 'x86_64' and sys_platform == 'darwin'}==1.26.3
numpy{platform_machine == 'x86_64' and sys_platform == 'darwin'}==1.26.4
numpy{platform_machine == 'x86_64' and sys_platform == 'darwin'}>2.0.0
and docling-ibm-models>=3.3.2 depends on numpy{platform_machine == 'x86_64' and sys_platform
== 'darwin'}>=1.24.4,<2.0.0, we can conclude that docling-ibm-models>=3.3.2 depends on
numpy>=1.24.4,<=1.26.4. (1)
And because only docling-ibm-models<=3.4.0 is available, we can conclude that docling-ibm-models>=3.3.2
depends on numpy>=1.24.4,<=1.26.4. (2)
Because docling-ibm-models>=3.3.0,<=3.3.1 depends on pillow>=10.0.0,<11.0.0 and only the following
versions of docling-ibm-models are available:
docling-ibm-models<=3.3.0
docling-ibm-models==3.3.1
docling-ibm-models==3.3.2
docling-ibm-models>=3.4.0
we can conclude that docling-ibm-models>=3.3.0,<3.3.2 depends on pillow>=10.0.0,<11.0.0. (3)
And because we know from (2) that docling-ibm-models>=3.3.2 depends on numpy>=1.24.4,<=1.26.4, we can
conclude that docling-ibm-models>=3.3.0, all of:
numpy<1.24.4
numpy>1.26.4
, all of:
pillow<10.0.0
pillow>=11.0.0
are incompatible. (4)
Because docling-ibm-models>=3.1.0,<=3.3.1 depends on pillow>=10.0.0,<11.0.0 and only the following
versions of docling-ibm-models are available:
docling-ibm-models<=3.1.0
docling-ibm-models==3.1.1
docling-ibm-models==3.1.2
docling-ibm-models==3.2.0
docling-ibm-models==3.2.1
docling-ibm-models>=3.3.0
we can conclude that docling-ibm-models>=3.1.0,<=3.3.1 depends on pillow>=10.0.0,<11.0.0.
And because we know from (4) that docling-ibm-models>=3.3.0, all of:
numpy<1.24.4
numpy>1.26.4
, all of:
pillow<10.0.0
pillow>=11.0.0
are incompatible, we can conclude that docling-ibm-models>=3.1.0, all of:
numpy<1.24.4
numpy>1.26.4
, all of:
pillow<10.0.0
pillow>=11.0.0
are incompatible. (5)
Because only the following versions of docling are available:
docling<=2.14.0
docling==2.15.0
docling==2.15.1
docling==2.16.0
docling==2.17.0
docling==2.18.0
docling==2.19.0
docling==2.20.0
docling==2.21.0
docling==2.22.0
docling==2.23.0
docling==2.23.1
docling==2.24.0
and docling>=2.14.0,<=2.15.1 depends on docling-ibm-models>=3.1.0, we can conclude that
docling>=2.14.0,<2.16.0 depends on docling-ibm-models>=3.1.0.
And because we know from (5) that docling-ibm-models>=3.1.0, all of:
numpy<1.24.4
numpy>1.26.4
, all of:
pillow<10.0.0
pillow>=11.0.0
are incompatible, we can conclude that docling>=2.14.0,<2.16.0, all of:
numpy<1.24.4
numpy>1.26.4
, all of:
pillow<10.0.0
pillow>=11.0.0
are incompatible. (6)
Because we know from (1) that docling-ibm-models>=3.3.2 depends on numpy>=1.24.4,<=1.26.4 and only
docling-ibm-models<=3.4.0 is available, we can conclude that docling-ibm-models>=3.3.2 depends on
numpy>=1.24.4,<=1.26.4.
And because we know from (3) that docling-ibm-models>=3.3.0,<3.3.2 depends on pillow>=10.0.0,<11.0.0,
we can conclude that docling-ibm-models>=3.3.0, all of:
numpy<1.24.4
numpy>1.26.4
, all of:
pillow<10.0.0
pillow>=11.0.0
are incompatible.
And because docling>=2.16.0,<=2.23.1 depends on docling-ibm-models>=3.3.0, we can conclude that
docling>=2.16.0,<=2.23.1, all of:
numpy<1.24.4
numpy>1.26.4
, all of:
pillow<10.0.0
pillow>=11.0.0
are incompatible.
And because we know from (6) that docling>=2.14.0,<2.16.0, all of:
numpy<1.24.4
numpy>1.26.4
, all of:
pillow<10.0.0
pillow>=11.0.0
are incompatible, we can conclude that docling>=2.14.0,<2.24.0, all of:
numpy<1.24.4
numpy>1.26.4
, all of:
pillow<10.0.0
pillow>=11.0.0
are incompatible. (7)
Because only the following versions of numpy{platform_machine == 'x86_64' and sys_platform == 'darwin'}
are available:
numpy{platform_machine == 'x86_64' and sys_platform == 'darwin'}<=1.24.4
numpy{platform_machine == 'x86_64' and sys_platform == 'darwin'}==1.25.0
numpy{platform_machine == 'x86_64' and sys_platform == 'darwin'}==1.25.1
numpy{platform_machine == 'x86_64' and sys_platform == 'darwin'}==1.25.2
numpy{platform_machine == 'x86_64' and sys_platform == 'darwin'}==1.26.0
numpy{platform_machine == 'x86_64' and sys_platform == 'darwin'}==1.26.1
numpy{platform_machine == 'x86_64' and sys_platform == 'darwin'}==1.26.2
numpy{platform_machine == 'x86_64' and sys_platform == 'darwin'}==1.26.3
numpy{platform_machine == 'x86_64' and sys_platform == 'darwin'}==1.26.4
numpy{platform_machine == 'x86_64' and sys_platform == 'darwin'}>2.0.0
and docling-ibm-models==3.4.0 depends on numpy{platform_machine == 'x86_64' and sys_platform
== 'darwin'}>=1.24.4,<2.0.0, we can conclude that docling-ibm-models==3.4.0 depends on
numpy>=1.24.4,<=1.26.4.
And because only docling-ibm-models<=3.4.0 is available and docling==2.24.0 depends on
docling-ibm-models>=3.4.0, we can conclude that docling==2.24.0 depends on numpy>=1.24.4,<=1.26.4.
And because we know from (7) that docling>=2.14.0,<2.24.0, all of:
numpy<1.24.4
numpy>1.26.4
, all of:
pillow<10.0.0
pillow>=11.0.0
are incompatible, we can conclude that docling>=2.14.0, all of:
numpy<1.24.4
numpy>1.26.4
, all of:
pillow<10.0.0
pillow>=11.0.0
are incompatible.
And because langchain-docling==0.2.0 depends on docling>=2.14.0, we can conclude that
langchain-docling==0.2.0, all of:
numpy<1.24.4
numpy>1.26.4
, all of:
pillow<10.0.0
pillow>=11.0.0
are incompatible.
And because only langchain-docling==0.2.0 is available and your project depends on langchain-docling,
we can conclude that all of:
numpy<1.24.4
numpy>1.26.4
, all of:
pillow<10.0.0
pillow>=11.0.0
, your project are incompatible.
And because your project depends on numpy==2.2.3 and pillow==11.1.0, we can conclude that your
project's requirements are unsatisfiable.
help: If you want to add the package regardless of the failed resolution, provide the --frozen flag to
skip locking and syncing.