Libraries, frameworks and developer tools for building, training, running and connecting AI models.
62 terms
spaCy is an open-source Python library for production natural language processing, with fast trainable pipelines for tagging, parsing and entity recognition.
Stable-Baselines3 is an open-source PyTorch library of tested reinforcement learning algorithms such as PPO, SAC and DQN with a common training API.
Streamlit is an open-source Python framework that turns data and machine learning scripts into interactive web applications without front-end code.
TensorBoard is the open-source visualization toolkit that turns machine learning training logs into charts, graphs and embedding views for diagnosis.
TensorFlow is an open-source machine learning framework from Google for building, training and deploying neural networks from servers to phones and browsers.
TensorFlow Lite Micro is an open-source C++ runtime that runs small quantized neural networks on microcontrollers with only kilobytes of memory.
TensorRT is the NVIDIA inference optimizer and runtime that compiles trained neural networks into fast, GPU-specific engines for deployment.
Unsloth is an open-source Python library that fine-tunes large language models faster and with far less GPU memory using LoRA, QLoRA and custom kernels.
The Vercel AI SDK is an open-source TypeScript toolkit that gives web applications one interface for calling, streaming and tooling many language models.
Weights & Biases is a machine learning platform that tracks experiments, versions datasets and models, and evaluates LLM applications in one shared record.
Windows ML is the Microsoft inference runtime that runs ONNX models locally on Windows devices, using the CPU, GPU or NPU through hardware-specific providers.
XGBoost is an open-source gradient boosting library that builds regularized decision tree ensembles for fast, accurate prediction on tabular data.