pretrainedmodels
Pretrained models for Pytorch
Decision gist · record as of 2026-08-14
No. The package is abandoned, with no updates since 2018-10-29 and no commits since 2022-04-22. Modern PyTorch and torchvision have superseded it with actively maintained model zoos, better API design, and current weight URLs. Install only if you must reproduce a specific 2018-era result or need a model not available elsewhere; otherwise, use torchvision or timm instead.AI-flagged interpretation of the facts on this page — verify before relying
Before you install
- Requires PyTorch to be installed separately; no explicit Python version requirement stated in the fact sheet.
- High install friction with no runtime dependencies listed but requiring manual setup.
- Maintenance is abandoned—last release was 2018-10-29, with no commits since 2022-04-22.
License · maintenance · safety
permissive license (permissive) — Licensed under MIT (permissive), so you can use, modify, and distribute freely with minimal restrictions. No commercial or proprietary constraints.
last release 2018-10-29 (2846 days) · last repo commit 2022-04-22 · 9,098 stars
0 known vulnerabilities (OSV.dev, 2026-08-14) · 85,116 downloads/mo, #13,949 on PyPI
Alternatives
Verify before relying
pip install pretrainedmodels
import pretrainedmodels
model = pretrainedmodels.resnet50(pretrained=True)
output = model(input_tensor)- Whether the package still downloads model weights successfully given the age of the project and potential URL changes.
- Compatibility with modern PyTorch versions beyond what was tested in 2018.
- Whether all listed models remain fully functional without maintenance.
- Current status of model weight URLs and hosting infrastructure.
What it is and what it does
Pretrainedmodels is a PyTorch package that wraps pretrained convolutional neural network models trained on ImageNet. It provides a consistent API across many architectures—ResNet, DenseNet, Inception, NASNet, SENet, DualPathNet, Xception, and others—allowing you to load models with weights already initialized from large-scale training. The package exposes model attributes like input size, normalization statistics, and intermediate layer access for transfer learning and feature extraction.
The package was designed to help reproduce research results and enable transfer learning setups by offering a torchvision-like interface. However, it has been abandoned since late 2018 and receives no maintenance. While it may still work for loading older model checkpoints, it is not updated for modern PyTorch conventions or security practices, and model weight URLs may be stale.
Use it for
- Load a pretrained ResNet or DenseNet for transfer learning on a custom image classification task without retraining from scratch.
- Extract intermediate features from a pretrained model for use in a downstream task like object detection or segmentation.
- Reproduce results from research papers published around 2017–2018 that relied on specific pretrained architectures and their exact weight initializations.
- Access multiple pretrained architectures through a single unified API rather than hunting for individual model implementations.
Worth the install?
AI-flagged interpretation of the facts on this page. Verify before relying on it.
No.
The package is abandoned, with no updates since 2018-10-29 and no commits since 2022-04-22. Modern PyTorch and torchvision have superseded it with actively maintained model zoos, better API design, and current weight URLs. Install only if you must reproduce a specific 2018-era result or need a model not available elsewhere; otherwise, use torchvision or timm instead.
Install
pretrainedmodels on PyPI
Before you install
High install friction with no runtime dependencies listed but requiring manual setup. Maintenance is abandoned—last release was 2018-10-29, with no commits since 2022-04-22. Use only if you need models from that era and cannot migrate to actively maintained alternatives.
Requires PyTorch to be installed separately; no explicit Python version requirement stated in the fact sheet.
License in practice
Licensed under MIT (permissive), so you can use, modify, and distribute freely with minimal restrictions. No commercial or proprietary constraints.
Quickstart
pip install pretrainedmodels
import pretrainedmodels
model = pretrainedmodels.resnet50(pretrained=True)
output = model(input_tensor)Verify before relying
- Whether the package still downloads model weights successfully given the age of the project and potential URL changes.
- Compatibility with modern PyTorch versions beyond what was tested in 2018.
- Whether all listed models remain fully functional without maintenance.
- Current status of model weight URLs and hosting infrastructure.
Package facts
| License | permissive license permissive |
| Python support | Not specified |
| Install friction | High. Source build required |
| Runtime dependencies | None |
| Maintenance | Abandoned 2,846 days since the last release |
| Last repo commit | |
| First released | |
| Downloads | 85,116 / month, #13,949 on PyPI 30-day window, as of 2026-08-14 |
| Known vulnerabilities | None known OSV.dev, checked 2026-08-14 |
| Classifiers | Development Status :: 3 - AlphaIntended Audience :: DevelopersLicense :: OSI Approved :: MIT LicenseProgramming Language :: Python :: 3.6Topic :: Software Development :: Build Tools |
Evidence: pretrainedmodels-0.7.4.tar.gz
Tags
Let your AI agent find packages like this
Example. Real query, live index.
You found this page by searching. An agent finds it by wishing: SkillFed indexes 14,416 PyPI packages by what they can do, searchable in plain language.
wish › “pretrained pytorch models”
- pretrainedmodelsProvides a unified interface to load pretrained PyTorch convolutional…
- efficientnet-pytorchLoads and runs pretrained EfficientNet image classification models in…
- timmTimm provides a large collection of pretrained PyTorch image…
Give your agent the search over MCP, or paste the wish link into any chat.
More Build Tools packages
Provides reusable utilities for Python packaging interoperability, including version handling, specifiers, markers, requirements, tags, and metadata parsing according to standards like PEP 440 and PEP 425.
Wraps any iterable to display a real-time progress bar in the terminal or Jupyter notebook, showing iteration count, elapsed time, and estimated time remaining.
pip is the standard installer for Python packages, enabling you to download and install packages from the Python Package Index and other indexes into your Python environment.
Hatchling is a standards-compliant Python build backend that handles packaging, metadata, and distribution of Python projects when configured in a project's pyproject.toml file.
Generates Python gRPC service stubs and message classes from Protocol Buffer definitions, enabling developers to build gRPC clients and servers.
pre-commit is a framework for installing and running git hooks written in any language before commits are made, automating code quality and validation checks across multi-language projects.
Install it if your team needs consistent, automated validation at commit time.
See also facenet-pytorch · Keras-Applications · pytorchcv · efficientnet-pytorch · timm · segmentation-models-pytorch · torchvision · efficientnet · face_recognition_models