gpo.zugaina.org

Search Portage & Overlays:

dev-python/tensorizer

Fast PyTorch module/model/tensor serialization + deserialization

Screenshots

  • tensorizer-2.12.1
    ~amd64 ~arm64
    python_single_target_python3_12 python_single_target_python3_13 python_single_target_python3_14

    View      Download      Browse     License: MIT   
    Overlay: stuff
  • tensorizer-2.10.1
    ~amd64 ~arm64
    python_single_target_python3_12 python_single_target_python3_13 python_single_target_python3_14

    View      Download      Browse     License: MIT   
    Overlay: stuff

ChangeLog

commit 7c378c750524ed172319a0944a8e810547e1ff98
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Sat Jun 20 15:49:22 2026 +0200

dev-python/tensorizer: Keyword 2.10.1 for ~arm64

commit 65be050bbee562f0e3438f69f5312c564a62f96b
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Tue Jun 16 22:30:36 2026 +0200

dev-python/tensorizer: Keyword 2.12.1 for ~arm64

Pure-Python (PEP517), architecture-independent.

commit 648a12c4fec0bf4941259123e1df9fcfce9b5229
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Wed May 13 14:35:34 2026 +0200

dev-python/tensorizer: disable py3.11

commit 5ced0b7b26f0c0621e907ccb9f5fff231b56cb4a
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Sun May 10 18:40:21 2026 +0200

dev-python/tensorizer: switch to DISTUTILS_SINGLE_IMPL

sci-ml/pytorch is SINGLE_IMPL; multi-impl consumer with bare
$ produces python_targets_python3_*(-)? that the
single-impl child can't expose. Both versions converted in one pass.

commit df8d8f1699efb446b9cc1fc1ab37a702f58aaa9c
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Sun May 10 15:13:27 2026 +0200

dev-python/tensorizer: add 2.12.1

commit 4c4291ca193640b45579a5896b9bcdbd08d7505b
Author: Ivan S. Titov <iohann.s.titov@gmail.com>
Date: Thu May 7 00:50:34 2026 +0200

dev-python/tensorizer: new package

CoreWeave's fast PyTorch tensor serialization library. Required by
dev-python/vllm at exact version 2.10.1. Pure Python (despite the
name suggesting native code).

NonsolvableDepsInStable is policy noise — ~amd64 only because pytorch
is ~amd64 only too.