openRuyi Lookout
← Packages

python-thinc

pyprojectDepChanges 2

A refreshing functional take on deep learning, compatible with your favorite libraries

Current
8.3.13
Candidate
10.0.0Stale
License
MIT

Build

TargetResultLast successful versionSucceeded at
rva23
8.3.13
rva20
8.3.13
x86_64
8.3.13

RuntimeDeps

DependencyObserved sourceCurrent requirementCandidate requirement
blis
1.3.3
<1.4.0,>=1.3.0✓
<1.4.0,>=1.3.0✓
catalogue
2.0.10
<2.1.0,>=2.0.4✓
<2.1.0,>=2.0.4✓
confection
1.3.3
<2.0.0,>=1.1.0✓
<2.0.0,>=1.3.2✓
cymem
2.0.13
<2.1.0,>=2.0.2✓
<2.1.0,>=2.0.2✓
murmurhash
1.0.15
<1.1.0,>=1.0.2✓
<1.1.0,>=1.0.2✓
numpy
2.4.6
<3.0.0,>=1.21.0✓
<3.0.0,>=1.21.0✓
packaging
25.0
>=20.0✓
>=20.0✓
preshed
3.0.12
<3.1.0,>=3.0.2✓
<3.1.0,>=3.0.2✓
pydantic
2.12.5
<3.0.0,>=2.0.0✓
<3.0.0,>=2.0.0✓
setuptools
83.0.0
any version✓
any version✓
srsly
2.5.3
<3.1.0,>=2.4.0✓
<3.1.0,>=2.4.0✓
wasabi
1.1.3
<1.2.0,>=0.8.1✓
<1.2.0,>=0.8.1✓
Python
3.14.9
<3.15,>=3.10✓
<3.16,>=3.10✓

Optional RuntimeDeps

DependencyObserved sourceCurrent requirementCandidate requirement
cupyUnmapped
—
>=5.0.0b4
>=5.0.0b4
cupy-cuda100Unmapped
—
>=5.0.0b4
>=5.0.0b4
cupy-cuda101Unmapped
—
>=5.0.0b4
>=5.0.0b4
cupy-cuda102Unmapped
—
>=5.0.0b4
>=5.0.0b4
cupy-cuda110Unmapped
—
>=5.0.0b4
>=5.0.0b4
cupy-cuda111Unmapped
—
>=5.0.0b4
>=5.0.0b4
cupy-cuda112Unmapped
—
>=5.0.0b4
>=5.0.0b4
cupy-cuda113Unmapped
—
>=5.0.0b4
>=5.0.0b4
cupy-cuda114Unmapped
—
>=5.0.0b4
>=5.0.0b4
cupy-cuda115Unmapped
—
>=5.0.0b4
>=5.0.0b4
cupy-cuda116Unmapped
—
>=5.0.0b4
>=5.0.0b4
cupy-cuda117Unmapped
—
>=5.0.0b4
>=5.0.0b4
cupy-cuda11xUnmapped
—
>=11.0.0
>=11.0.0
cupy-cuda12xUnmapped
—
>=11.5.0
>=11.5.0
cupy-cuda80Unmapped
—
>=5.0.0b4
>=5.0.0b4
cupy-cuda90Unmapped
—
>=5.0.0b4
>=5.0.0b4
cupy-cuda91Unmapped
—
>=5.0.0b4
>=5.0.0b4
cupy-cuda92Unmapped
—
>=5.0.0b4
>=5.0.0b4
cupy-wheelUnmapped
—
>=11.0.0
>=11.0.0
ml_datasetsUnmapped
—
<0.3.0,>=0.2.0
<0.3.0,>=0.2.0
mxnetUnmapped
—
<1.6.0,>=1.5.1
<1.6.0,>=1.5.1
tensorflowUnmapped
—
<2.6.0,>=2.0.0
<2.6.0,>=2.0.0
thinc-apple-opsUnmapped
—
<2.0.0,>=1.0.0
<2.0.0,>=1.0.0
torch
2.13.0
>=1.6.0?
>=1.6.0?
Dependency conditions
  • cupy>=5.0.0b4Unmapped
    Applies when (any)
    extra == "cuda"
  • cupy-cuda100>=5.0.0b4Unmapped
    Applies when (any)
    extra == "cuda100"
  • cupy-cuda101>=5.0.0b4Unmapped
    Applies when (any)
    extra == "cuda101"
  • cupy-cuda102>=5.0.0b4Unmapped
    Applies when (any)
    extra == "cuda102"
  • cupy-cuda110>=5.0.0b4Unmapped
    Applies when (any)
    extra == "cuda110"
  • cupy-cuda111>=5.0.0b4Unmapped
    Applies when (any)
    extra == "cuda111"
  • cupy-cuda112>=5.0.0b4Unmapped
    Applies when (any)
    extra == "cuda112"
  • cupy-cuda113>=5.0.0b4Unmapped
    Applies when (any)
    extra == "cuda113"
  • cupy-cuda114>=5.0.0b4Unmapped
    Applies when (any)
    extra == "cuda114"
  • cupy-cuda115>=5.0.0b4Unmapped
    Applies when (any)
    extra == "cuda115"
  • cupy-cuda116>=5.0.0b4Unmapped
    Applies when (any)
    extra == "cuda116"
  • cupy-cuda117>=5.0.0b4Unmapped
    Applies when (any)
    extra == "cuda117"
  • cupy-cuda11x>=11.0.0Unmapped
    Applies when (any)
    extra == "cuda11x"
  • cupy-cuda12x>=11.5.0Unmapped
    Applies when (any)
    extra == "cuda12x"
  • cupy-cuda80>=5.0.0b4Unmapped
    Applies when (any)
    extra == "cuda80"
  • cupy-cuda90>=5.0.0b4Unmapped
    Applies when (any)
    extra == "cuda90"
  • cupy-cuda91>=5.0.0b4Unmapped
    Applies when (any)
    extra == "cuda91"
  • cupy-cuda92>=5.0.0b4Unmapped
    Applies when (any)
    extra == "cuda92"
  • cupy-wheel>=11.0.0Unmapped
    Applies when (any)
    extra == "cuda-autodetect"
  • ml_datasets<0.3.0,>=0.2.0Unmapped
    Applies when (any)
    extra == "datasets"
  • mxnet<1.6.0,>=1.5.1Unmapped
    Applies when (any)
    extra == "mxnet"
  • tensorflow<2.6.0,>=2.0.0Unmapped
    Applies when (any)
    extra == "tensorflow"
  • thinc-apple-ops<2.0.0,>=1.0.0Unmapped
    Applies when (any)
    extra == "apple"
  • torch>=1.6.0?
    Applies when (any)
    extra == "torch"

Description

Thinc is a lightweight deep learning library that offers an elegant, type-checked, functional-programming API for composing models, with support for layers defined in other frameworks such as PyTorch, TensorFlow and MXNet. You can use Thinc as an interface layer, a standalone toolkit or a flexible way to develop new models. Previous versions of Thinc have been running quietly in production in thousands of companies, via both spaCy and Prodigy. We wrote the new version to let users compose, configure and deploy custom models built with their favorite framework.

Changelog

PR · 3

Checks
MonitorObservationLast checked
Source
Stale
Repository refresh: git fetch exited 128 (authentication)
Version
Stale
Source last confirmed at 2026-11-11T11:23:08+00:00; repository refresh overdue.
Build
Checked
GitHub
Checked
EOL
Not configured
No reliable monitor identity/configuration.
—
LicenseDiff
Checked
Compared same-provider, same-project SPDX metadata for this version pair.
Last attempted
Dependencies
Checked
Last attempted
Advisory
Checked
Current source component only; local patches, build options, bundled dependencies and binary artifacts are not evaluated.
Yanked
Checked