Download Latest Version v8.3.6_ Support Python 3.13 source code.tar.gz (1.7 MB)
Email in envelope

Get an email when there's a new version of Thinc

Home / release-v8.3.3
Name Modified Size InfoDownloads / Week
Parent folder
thinc-8.3.3-cp312-cp312-macosx_11_0_arm64.whl 2024-12-16 761.0 kB
0
thinc-8.3.3-cp39-cp39-macosx_10_9_x86_64.whl 2024-12-16 848.0 kB
0
thinc-8.3.3-cp39-cp39-macosx_11_0_arm64.whl 2024-12-16 780.8 kB
0
thinc-8.3.3-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl 2024-12-16 3.7 MB
0
thinc-8.3.3-cp39-cp39-musllinux_1_2_x86_64.whl 2024-12-16 4.7 MB
0
thinc-8.3.3-cp39-cp39-win_amd64.whl 2024-12-16 1.5 MB
0
thinc-8.3.3-cp310-cp310-macosx_10_9_x86_64.whl 2024-12-16 843.9 kB
0
thinc-8.3.3-cp310-cp310-macosx_11_0_arm64.whl 2024-12-16 779.4 kB
0
thinc-8.3.3-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl 2024-12-16 3.7 MB
0
thinc-8.3.3-cp310-cp310-musllinux_1_2_x86_64.whl 2024-12-16 4.7 MB
0
thinc-8.3.3-cp310-cp310-win_amd64.whl 2024-12-16 1.5 MB
0
thinc-8.3.3-cp311-cp311-macosx_10_9_x86_64.whl 2024-12-16 839.3 kB
0
thinc-8.3.3-cp311-cp311-macosx_11_0_arm64.whl 2024-12-16 774.2 kB
0
thinc-8.3.3-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl 2024-12-16 3.9 MB
0
thinc-8.3.3-cp311-cp311-musllinux_1_2_x86_64.whl 2024-12-16 4.9 MB
0
thinc-8.3.3-cp311-cp311-win_amd64.whl 2024-12-16 1.5 MB
0
thinc-8.3.3-cp312-cp312-macosx_10_13_x86_64.whl 2024-12-16 825.1 kB
0
thinc-8.3.3-cp312-cp312-manylinux_2_17_x86_64.manylinux2014_x86_64.whl 2024-12-16 3.7 MB
0
thinc-8.3.3-cp312-cp312-musllinux_1_2_x86_64.whl 2024-12-16 4.7 MB
0
thinc-8.3.3-cp312-cp312-win_amd64.whl 2024-12-16 1.5 MB
0
thinc-8.3.3.tar.gz 2024-12-16 193.9 kB
0
README.md 2024-12-16 452 Bytes
0
v8.3.3_ Fix Blis crashes, widen numpy pin source code.tar.gz 2024-12-16 1.7 MB
0
v8.3.3_ Fix Blis crashes, widen numpy pin source code.zip 2024-12-16 1.8 MB
0
Totals: 24 Items   50.3 MB 0
  • Update blis pin to v1.1. This updates the vendored blis code to 1.1, which should fix crashes from the previously vendored v0.9 code on Windows.
  • Widen numpy pin, allowing versions across v1 and v2. Previously I had thought that if I build against numpy v2, I couldn't also have v1 as a runtime dependency. This is actually incorrect, so we can widen the numpy pin
  • Set flag on loading PyTorch models to improve safety of loading PyTorch models.
Source: README.md, updated 2024-12-16