Tag
Explore by tags
Converts document images—scans, photos, born-digital PDFs—into structured text in two stages: first map layout and reading order, then parse each element (text, tables, formulas, figures) in parallel, each guided by its own task prompt.
A PyTorch DTensor-native SPMD library for training and fine-tuning LLMs, VLMs, diffusion and retrieval models. Integrates with Hugging Face for day-0 model support, provides YAML-driven recipes, DTensor/FSDP2 parallelism and NVIDIA-optimized kernels (Transformer Engine, DeepEP, FlexAttn).
Enables bidirectional checkpoint conversion between Hugging Face and Megatron formats and provides a PyTorch-native training library with tensor/pipeline parallelism, FP8/BF16 mixed precision, SFT and PEFT (LoRA) support for large and multimodal models.
Physics-aware simulated sensor dataset for training and evaluating autonomous-vehicle perception and control models. Includes multimodal sensor streams with physical-scene annotations intended for tasks that require grounding in real-world dynamics.
Provides MS MARCO queries, passages and answers translated into 14 Indic languages while keeping the original English content and per-example translation metadata. Includes train/validation splits, passage selection flags, and translation model parameters for multilingual IR, QA and RAG research.
Extends RAG beyond text: parses PDFs and Office files containing images, tables, equations, and charts, then queries them through one multimodal knowledge graph. Built on LightRAG, it replaces separate parsing and retrieval tools.
1,000,000 US-focused synthetic persona records (6M persona texts) grounded to demographic, geographic and personality distributions. Contains age, sex, education, occupation and ZCTA/city fields; CC BY 4.0 license for LLM training, diversity augmentation, and bias mitigation.
A toolkit and open-weights system for real-time streaming music generation — offers two model sizes (230M / 2.4B), a Python inference library (JAX/MLX), and a C++ engine optimized for Apple Silicon for embedding into DAWs and apps; real-time streaming requires M‑series chips.
Forecasts financial candlesticks (OHLCV K-lines) with a decoder-only transformer pre-trained on 12B+ records from 45 exchanges. A tokenizer turns market data into discrete tokens, enabling price/volatility forecasting and synthetic K-line generation.
Model-compression toolkit for large LLMs/VLMs that integrates quantization (FP8/INT4/etc.), speculative decoding, token pruning and deployment hooks—designed for end-to-end performance on single/multi-GPU inference workflows and research-to-prod model optimization.
Provides a large-scale, multi-speaker Persian speech–text corpus constructed from audiobooks for TTS, ASR, and speaker research. Includes automated alignment and quality scoring, TTS-ready subsets (thousands of hours/1M+ segments) and metadata for speaker IDs and genders — suitable for multi-speaker synthesis and voice cloning research.
Detects, segments, and tracks every instance of an open-vocabulary concept in images and video from a text phrase or visual exemplar, not just one object per prompt. An 848M-param model reaching ~75-80% of human accuracy across 270K concepts.