Packages
-
paddlex
PaddleX is a low-code framework for training,…
permissive · active · 2.5M/mo
-
rapidocr-onnxruntime
Performs optical character recognition (OCR) on…
permissive · active · 2.5M/mo
-
nvidia-cudnn-cu11
Provides cuDNN runtime libraries for…
unclear · aging · 2.5M/mo
-
decord
Decord decodes video and audio files with…
permissive · dormant · 2.4M/mo
-
paddlepaddle
PaddlePaddle is a deep learning framework for…
permissive · active · 2.4M/mo
-
assemblyai
Python SDK for transcribing and understanding…
permissive · active · 2.4M/mo
-
neptune-api
A generated Python client library for the…
permissive · abandoned · 2.4M/mo
-
qwen-vl-utils
Provides helper functions to process images and…
permissive · aging · 2.4M/mo
-
lean-lsp-mcp
Provides an MCP server that exposes the Lean…
permissive · active · 2.4M/mo
-
neptune-fetcher
Queries and fetches experiment metadata, runs,…
permissive · aging · 2.4M/mo
-
llama-index-legacy
LlamaIndex Legacy is a data framework for…
permissive · active · 2.3M/mo
-
langchain-groq
Connects Groq's inference API to LangChain,…
permissive · active · 2.3M/mo
-
braintrust-langchain
Provides a LangChain callback handler to log…
permissive · active · 2.3M/mo
-
chex
Chex provides utilities for writing reliable…
permissive · active · 2.3M/mo
-
nvidia-nvvm
Provides the NVVM compiler IR library for…
unclear · active · 2.3M/mo
-
pyannote-audio
Speaker diarization toolkit that identifies and…
unclear · active · 2.3M/mo
-
llama-index-embeddings-bedrock
Provides Amazon Bedrock embedding models…
permissive · active · 2.3M/mo
-
nvidia-cuda-runtime-cu11
Provides CUDA 11 runtime native libraries for…
unclear · abandoned · 2.3M/mo
-
unsloth
Unsloth accelerates training and fine-tuning of…
permissive · active · 2.3M/mo
-
outlines
Outlines constrains LLM generation to produce…
permissive · active · 2.3M/mo
-
agno
Agno is a framework and runtime for building,…
permissive · active · 2.3M/mo
-
dvc-studio-client
Client library for posting experiment metrics,…
permissive · active · 2.3M/mo
-
nvidia-cuda-nvrtc-cu11
Provides NVIDIA CUDA NVRTC (runtime…
unclear · abandoned · 2.2M/mo
-
tecton
Tecton is a Python SDK for defining, testing,…
unclear · aging · 2.2M/mo
-
ml-collections
Provides dict-like configuration data…
permissive · active · 2.2M/mo
-
xformers
XFormers provides reusable, composable building…
permissive · active · 2.2M/mo
-
databricks-ai-bridge
Provides a shared Python API layer for building…
unclear · active · 2.2M/mo
-
tensorflow-hub
Downloads and loads pre-trained TensorFlow…
permissive · dormant · 2.2M/mo
-
pyannote-pipeline
Defines and optimizes tunable hyperparameter…
unclear · aging · 2.2M/mo
-
cog
Cog packages machine learning models into…
permissive · active · 2.1M/mo
-
grain
Grain is a Python library for reading,…
permissive · active · 2.1M/mo
-
torch-audiomentations
Provides PyTorch-native audio data augmentation…
permissive · aging · 2.1M/mo
-
fvcore
fvcore provides shared computer vision…
permissive · active · 2.1M/mo
-
asteroid-filterbanks
Provides PyTorch-based filterbank…
permissive · dormant · 2.1M/mo
-
fastsafetensors
Loads safetensors model files significantly…
permissive · active · 2.1M/mo
-
nvidia-cuda-crt
Provides NVIDIA CUDA C Runtime libraries for…
unclear · active · 2.1M/mo
-
tensorboard-plugin-wit
Adds the What-If Tool, an interactive visual…
permissive · abandoned · 2.1M/mo
-
optimum
Optimum provides optimization tools to export…
permissive · active · 2.1M/mo
-
torchviz
Generates visual diagrams of PyTorch neural…
permissive · dormant · 2.1M/mo
-
flash-attn-4
GPU-accelerated attention mechanism…
permissive · active · 2.0M/mo
-
pinecone-plugin-assistant
Provides Python bindings to Pinecone's…
permissive · active · 2.0M/mo
-
tokenspeed-mla
Provides optimized MLA (Multi-head Latent…
permissive · active · 2.0M/mo
-
openvino
OpenVINO converts and optimizes deep learning…
permissive · active · 1.9M/mo
-
livekit-plugins-elevenlabs
Integrates ElevenLabs text-to-speech into the…
permissive · active · 1.9M/mo
-
composio
Composio provides a Python SDK that connects AI…
permissive · active · 1.9M/mo
-
onnxconverter-common
Provides common utilities and functions for…
permissive · active · 1.9M/mo
-
azure-ai-formrecognizer
Extracts text, structure, and key-value pairs…
permissive · active · 1.9M/mo
-
nvidia-cuda-cccl
Provides NVIDIA CUDA C++ Core Compute Libraries…
unclear · active · 1.9M/mo
-
openrouter
A Python SDK for accessing 400+ AI models…
permissive · active · 1.9M/mo
-
tensorflow-cpu
TensorFlow CPU provides a machine learning and…
permissive · active · 1.8M/mo
-
comfy-aimdo
A PyTorch VRAM allocator that dynamically…
unclear · active · 1.8M/mo
-
replicate
A Python client for running machine learning…
permissive · aging · 1.8M/mo
-
vector-quantize-pytorch
Implements vector quantization layers for…
permissive · active · 1.8M/mo
-
torchsde
Solves stochastic differential equations (SDEs)…
permissive · abandoned · 1.8M/mo
-
cma
CMA-ES is a derivative-free numerical…
permissive · active · 1.8M/mo
-
openhands-ai
OpenHands is a self-hosted developer control…
permissive · active · 1.8M/mo
-
langchain-huggingface
Connects Hugging Face models and embeddings to…
permissive · active · 1.8M/mo
-
dbos
DBOS adds durable workflows and queues to…
permissive · active · 1.8M/mo
-
tensorflow-datasets
Provides access to many public datasets as…
permissive · active · 1.8M/mo
-
speechbrain
SpeechBrain is a PyTorch-based toolkit for…
permissive · active · 1.8M/mo
-
milvus-lite
Milvus Lite is a pure-Python local vector…
permissive · active · 1.8M/mo
-
unitycatalog-langchain
Exposes Unity Catalog functions as LangChain…
permissive · active · 1.8M/mo
-
lap
Solves the linear assignment problem using the…
permissive · active · 1.7M/mo
-
aistudio-sdk
Python client library for the AIStudio API,…
permissive · active · 1.7M/mo
-
unitycatalog-ai
Provides Python APIs to create, retrieve, and…
permissive · active · 1.7M/mo
-
graphiti-core
Graphiti builds and queries temporal context…
permissive · active · 1.6M/mo
-
azureml-core
Provides core APIs and utilities for managing…
unclear · active · 1.6M/mo
-
ragas
Ragas provides objective metrics, test data…
permissive · active · 1.6M/mo
-
onnx2tf
Converts ONNX model files to LiteRT,…
permissive · active · 1.6M/mo
-
langchain-chroma
Connects LangChain applications to Chroma, a…
permissive · active · 1.6M/mo
-
lm-eval
Unified framework for evaluating generative…
permissive · active · 1.6M/mo
-
fiddle
Fiddle is a Python-first configuration library…
permissive · active · 1.6M/mo
-
browsergym-core
BrowserGym provides a gymnasium environment for…
permissive · active · 1.5M/mo
-
nvidia-cufft-cu11
Provides NVIDIA CUFFT native runtime libraries…
unclear · abandoned · 1.5M/mo
-
nvidia-cusparse-cu11
Provides CUSPARSE native runtime libraries for…
unclear · abandoned · 1.5M/mo
-
nvidia-cusolver-cu11
Provides NVIDIA CUDA solver native runtime…
unclear · abandoned · 1.5M/mo
-
nemo-toolkit
nemo-toolkit provides a PyTorch framework for…
permissive · active · 1.5M/mo
-
nvidia-curand-cu11
Provides NVIDIA CURAND native runtime libraries…
unclear · abandoned · 1.5M/mo
-
nvidia-cuda-cupti-cu11
Provides NVIDIA CUDA profiling runtime…
unclear · abandoned · 1.5M/mo
-
agent-framework
Builds and orchestrates AI agents powered by…
permissive · active · 1.5M/mo
-
nvidia-nvtx-cu11
Provides a C-based API for annotating events,…
unclear · abandoned · 1.5M/mo
-
torch-geometric
PyTorch Geometric is a library for building and…
permissive · active · 1.5M/mo
-
nvidia-cuda-tileiras
Provides a portable intermediate representation…
unclear · active · 1.5M/mo
-
spacy-curated-transformers
Integrates curated transformer models (ALBERT,…
permissive · active · 1.5M/mo
-
unitycatalog-openai
Exposes Unity Catalog functions as tools for…
permissive · active · 1.5M/mo
-
sagemaker-schema-inference-artifacts
Provides pre-built schema inference artifacts…
permissive · dormant · 1.5M/mo
-
nvidia-nccl-cu11
Provides NVIDIA's collective communication…
unclear · dormant · 1.5M/mo
-
whisperx
WhisperX performs fast automatic speech…
permissive · active · 1.5M/mo
-
neptune-scale
Neptune-scale is a Python client library for…
permissive · aging · 1.5M/mo
-
databricks-openai
Integrates Databricks AI features, particularly…
permissive · active · 1.4M/mo
-
bedrock-agentcore-starter-toolkit
Provides a Python CLI and SDK for scaffolding,…
permissive · active · 1.4M/mo
-
pythainlp
PyThaiNLP provides Thai-language natural…
permissive · active · 1.4M/mo
-
nvidia-cuda-nvdisasm
nvdisasm disassembles NVIDIA CUDA cubin files…
unclear · active · 1.4M/mo
-
humming-kernels
Humming is a JIT-compiled GEMM kernel library…
unclear · active · 1.4M/mo
-
langchain-fireworks
Integrates Fireworks.ai language models with…
permissive · active · 1.4M/mo
-
mlx-lm
MLX LM loads, generates text with, fine-tunes,…
permissive · active · 1.4M/mo
-
stable-baselines3
Stable Baselines3 provides PyTorch…
permissive · active · 1.4M/mo
-
tf-keras-nightly
TF-Keras is a pure-TensorFlow implementation of…
permissive · active · 1.4M/mo
-
stagehand
Stagehand is a Python SDK for building browser…
permissive · active · 1.4M/mo
-
x-transformers
x-transformers provides modular transformer…
permissive · active · 1.4M/mo
-
neptune-query
Read-only API for querying experiment metadata,…
permissive · abandoned · 1.4M/mo
-
google-cloud-pipeline-components
Provides predefined Kubeflow Pipelines…
permissive · active · 1.4M/mo
-
supervision
Supervision provides utilities for loading,…
permissive · active · 1.4M/mo
-
smg-grpc-proto
Provides pre-compiled Python gRPC stubs for…
permissive · active · 1.4M/mo
-
unsloth-zoo
Unsloth Zoo provides utilities for fine-tuning…
copyleft · active · 1.4M/mo
-
imgaug
Augments images and related data (heatmaps,…
permissive · dormant · 1.4M/mo
-
chonkie
Chonkie splits text into semantically…
permissive · active · 1.3M/mo
-
botframework-connector
Enables Python bots to send and receive…
permissive · abandoned · 1.3M/mo
-
cmaes
Provides CMA-ES (Covariance Matrix Adaptation…
permissive · active · 1.3M/mo
-
botbuilder-schema
Provides serialized data schemas and message…
permissive · abandoned · 1.3M/mo
-
sagemaker-serve
Provides model serving and deployment…
permissive · active · 1.3M/mo
-
truss
Truss is a CLI tool for packaging ML models…
permissive · active · 1.3M/mo
-
mlx
MLX is an array framework for machine learning…
permissive · active · 1.3M/mo
-
langchain-mistralai
Connects Mistral AI language models to…
permissive · active · 1.3M/mo
-
sagemaker-train
Trains and deploys machine learning models on…
unclear · active · 1.3M/mo
-
torchrl
TorchRL is a PyTorch-native toolkit for…
unclear · active · 1.3M/mo
-
livekit-plugins-cartesia
Integrates Cartesia's voice AI services…
permissive · active · 1.3M/mo
-
sagemaker-mlops
Provides high-level orchestration for Amazon…
permissive · active · 1.3M/mo
-
lhotse
Lhotse prepares multimodal (speech, audio,…
permissive · active · 1.3M/mo
-
nvidia-cuda-nvcc-cu12
Provides the NVIDIA CUDA nvcc compiler for…
unclear · aging · 1.3M/mo