nvidia-nemo

🕵️ NeMo Anonymizer: Detect and protect PII through context-aware replacement and rewriting

Speech synthesis (TTS) in low-resource languages by training from scratch with Fastpitch and fine-tuning with HifiGan

LLM tutorial materials include but not limited to NVIDIA NeMo, TensorRT-LLM, Triton Inference Server, and NeMo Guardrails.

Extremely fast and accurate audio transcriber surpassing Whisper. Fast on GPU or CPU.

Automatic transcriber made with the Nvidia NeMo AI toolkit. Used to transcribe speech to text in real-time from any source. Requires CUDA capable GPU to run on the local machine, if setup using virtual audio cables can transcribe the audio that is being played in real-time without any other requirements.

The simplest & most comprehensible tutorial on speaker identification with NVIDIA's `Nemo`.

🤖 NemoClawd — NVIDIA NeMo Agent Toolkit + ClawdBot in one bridge-first AI workspace. TypeScript orchestrates NeMo Python workflows. Enterprise AI Digital Project Manager.

FastAPI-based Hindi ASR app using NVIDIA NeMo + ONNX, with Docker support for easy deployment.

Module for russian speech recognition using NVIDIA Nemo.

🎬 AI subtitle generator: convert video to SRT subtitles locally with NVIDIA NeMo Parakeet-TDT speech-to-text. GPU-accelerated, word-level timestamps, VAD, LLM correction — a fast offline Whisper alternative.

Implementation of a Kazakh Speech-to-Text Model using the NVIDIA NeMo toolkit for efficient transcription of spoken Kazakh speech into text.

Fully-local speech-to-text dictation. Hold a hotkey, talk, and the transcript lands in the field you're already in — an NVIDIA Parakeet streaming server plus native macOS and Windows clients. Your voice never leaves your LAN.

Make the agents you ship faster, more accurate, and safer.

CI/CD templates for NeMo-FW libraries

NeMo SDG-PGMs is a user-friendly Python library for building probabilistic graphical models (PGMs) to generate synthetic data

🛡️ NeMo Safe Synthesizer: Create private, safe versions of sensitive tabular datasets.

Fork of NVIDIA/NeMo-Agent-Toolkit; source of the ReAct agent LiteLLM/Anthropic TypeError fix merged upstream as PR #1251.