Packages
Fairlearn assesses and mitigates fairness issues in machine learning models by providing metrics to identify which groups are negatively impacted and algorithms to reduce unfairness across various AI tasks.
Install it if fairness assessment or mitigation is part of your model development workflow.
Detects, locates, and identifies faces in images using deep learning, with both Python API and command-line interface for batch processing.
Megatron Core provides GPU-optimized building blocks and parallelism strategies for training large transformer models at scale, including tensor parallelism, pipeline parallelism, and mixed precision support.
Not recommended for simple single-GPU training or inference-only use cases.
Converts HuggingFace tokenizers into OpenVINO models for text processing, enabling tokenization and detokenization within OpenVINO inference pipelines without external tokenizer dependencies.
Extends ONNX Runtime with custom operators for pre- and post-processing in vision, text, and NLP models, available as a C/C++ library with Python, Java, and C# bindings.
Provides helper functions that extend Einops, a tensor manipulation library, with additional utilities for working with tensors using Einstein-like notation.
Composio-core bridges AI agents to external tools and services, providing a framework to integrate actions from platforms like GitHub, Slack, and Gmail with LLM-based agents.
However, no for new projects—the package is explicitly deprecated and the maintainers recommend using 'composio' instead.
Connects to Azure Machine Learning Studio workspaces to download, upload, and manage datasets from Python, with optional conversion to and from Pandas DataFrames.
OmniVoice generates speech from text in over 600 languages using a diffusion-based model, with support for voice cloning from reference audio and voice design via speaker attributes.
Bridges pandas DataFrames and scikit-learn by mapping DataFrame columns to transformations and recombining them into features for machine learning pipelines.
Helion is a Python-embedded domain-specific language for writing machine learning kernels that compile to Triton, with built-in autotuning to optimize GPU kernel performance.
However, the unclear license classification and 10-minute autotuning overhead on first run are real constraints—verify license compatibility for your use case and…
TensorFlow Recommenders provides a Keras-based library for building recommender system models, covering the full workflow from data preparation through training, evaluation, and deployment.
Chatterbox TTS converts text to speech using open-source neural models, supporting English and 23+ languages with optional voice cloning from reference audio clips.
Provides a Conditional Random Field (CRF) layer for TensorFlow 2 Keras models, enabling sequence labeling tasks with built-in support for masking and mixed precision training.
Install only if you can verify compatibility with your TensorFlow and tensorflow-addons versions, and accept the risk that bugs or incompatibilities will not be fixed…
Copulas models multivariate statistical distributions and generates synthetic tabular data that preserves the statistical properties of the original dataset using copula functions.
However, the BUSL-1.1 license restricts commercial use until a future date—verify the license terms against your use case before committing.
TextArena provides a framework of 100+ text-based games with an OpenAI Gym-style interface for training, evaluating, and benchmarking language models through competitive gameplay.
However, maintenance is aging (302 days since last release), so expect slower bug fixes and feature additions.
Petastorm enables direct training of deep learning models from Apache Parquet datasets using TensorFlow, PyTorch, PySpark, or pure Python, with support for selective column access, shuffling, filtering, and distributed parallelism.
However, note the aging maintenance status (224 days since last release)—verify compatibility with your specific framework versions before committing to production use.
Jupyter AI is a JupyterLab extension that integrates agentic AI into computational notebooks, enabling collaboration with frontier AI agents through a native chat UI with file and command execution capabilities.
Install it if you use JupyterLab and want AI agent collaboration; the modular design means you only pay for what you use.
Resizes images or tensors in NumPy or PyTorch with differentiable operations, supporting anti-aliasing, multiple interpolation methods, and flexible scale-factor or output-size specification.
Provides a lightweight Python wrapper to interact with the MinerU Vision-Language Model, supporting multiple deployment backends for document layout detection and content recognition.
Intel-optimized TensorFlow for Windows that accelerates numerical computation and machine learning workloads using oneDNN primitives for Intel architecture.
Install only if you are on Windows; the medium dependency footprint is standard for TensorFlow and the oneDNN environment variable must be set explicitly to activate…
Adds a persistent, sandboxed JavaScript REPL tool to deepagents agents, letting the model write orchestrated JavaScript code instead of issuing serial tool calls.
dlib is a C++ machine learning and computer vision toolkit with Python bindings, providing algorithms for image recognition, classification, and data analysis.
However, ensure your development environment has a working C++ compiler and CMake before attempting installation.
TensorRT compiles and optimizes deep learning models for deployment on NVIDIA GPUs, reducing latency and memory footprint through layer fusion, quantization, and kernel auto-tuning.
However, installation requires CUDA toolkit, system-level build tools, and the TensorRT GA build v11.2.1.2; proprietary licensing requires license review for…
Provides command-line tools and Python APIs to convert deep learning models from multiple frameworks (TensorFlow, PyTorch, ONNX, Caffe, etc.) to OpenVINO IR format, download pre-trained models, and optimize them for inference deployment.
Mistral Workflows is a Python SDK for building AI-powered workflows with built-in fault tolerance, distributed execution, and observability using decorators and Pydantic validation.
However, the Beta status, Python 3.12+ requirement, and temporalio dependency complexity mean you should verify that the maturity level and backend setup align with…
PandasAI lets you ask questions about your data in natural language and get answers without writing SQL or Python code, using an LLM backend to translate questions into data operations.
However, maintenance is aging (311 days since last release), so verify that the library's feature set and bug-fix cadence meet your stability requirements before…
RDT converts raw, mixed-type data into fully numerical form via reversible transformations, then converts it back to the original format.
However, the BUSL-1.1 license restricts commercial use until a future date—verify the license terms match your use case before committing.
Parses ULog files (a self-describing logging format) and provides command-line tools to extract, convert, and display logged data in multiple formats including CSV, KML, and ROS2 bag files.
Install it if you work with PX4 logs or need ULog parsing; skip it otherwise.
PDD is a CLI tool that treats natural-language prompts as source code, automatically generating, testing, and maintaining code in Python, TypeScript, Go, and other languages from `.prompt` files.
Integrates Pinecone vector database with LlamaIndex for storing and retrieving embeddings in RAG and semantic search applications.
Identifies the language of UTF-8 text using Facebook's pretrained fastText lid.176 model, supporting 176 languages with optional memory-efficient compressed mode.
Install it if you need fast, accurate language identification across 176 languages without external service calls.
Provides NVIDIA TensorRT libraries for CUDA 12 environments, enabling optimized deep learning model inference on NVIDIA GPUs.
Aim logs training runs and AI metadata, then provides a web UI and Python API to compare, visualize, and query experiments programmatically.
Install it if you can tolerate the medium dependency footprint and are willing to verify the four known security vulnerabilities do not affect your threat model.
py_trees implements behaviour trees in Python, providing composable decision-making primitives (behaviours, decorators, sequences, selectors, parallels) and a blackboard for data sharing to build medium-sized decision engines.
Install it if you are building a robotics system, game AI, or any medium-complexity state machine that benefits from tree-structured control flow and shared state…
Python SDK for accessing AI21 Labs' language models (including Jamba) via REST API, with support for chat completions, text completion, streaming, and agent orchestration.
However, verify the license terms first, and be aware that the package is aging (last release 281 days ago) with no visible recent maintenance activity—suitable for…
Provides internal AutoML infrastructure and utilities for the Azure Machine Learning SDK's automated machine learning training pipeline.
No, not for direct installation.
MindRoom is a multi-agent runtime that deploys AI agents as persistent Matrix users, giving them access to 100+ integrated tools and persistent memory across chat platforms including Slack, Telegram, Discord, and email via Matrix bridges.
Detects 2D and 3D facial landmarks from images using deep learning, supporting multiple face detection backends and GPU acceleration via PyTorch.
Install it if facial geometry extraction is core to your application.
Provides programmatic access to RF100-VL, a multi-domain object detection benchmark dataset of 100 datasets across seven domains, enabling download and evaluation of vision-language models and object detection systems.
Install it if you are working on vision-language models, few-shot detection, or need a diverse multi-domain evaluation dataset—but note that you must obtain and…