Category
Explore by categories
A lightweight 'drafter' assistant for Gemma 4 31B that generates speculative token drafts to enable up-to-2× decoding speedups while preserving final output quality; compatible with Hugging Face Transformers and any-to-any pipelines.
Acts as the assistant (drafter) checkpoint for Gemma 4 26B A4B on Hugging Face, used in Speculative Decoding to pre-draft tokens and speed up generation. Designed for long-context, multimodal workflows where lower latency and on-device or edge inference matter.
Unifies video, audio, image and text understanding for enterprise Q&A, summarization, transcription and document intelligence. The NVFP4 quantized variant reduces footprint to ~20.9GB for more efficient single‑GPU deployment and is tuned for NVIDIA runtimes (vLLM, TensorRT).
Provides unquantized BF16 weights of Qwen3.6-27B with the base model's MTP head grafted in for high-fidelity, uncensored text (and multimodal) generation. Includes deployment guidance and hardware-tuned variants for A100/H100 and Blackwell-class GPUs.
Generates anime-style images from natural-language prompts with a full fine-tune family built on Z-Image Base — available as Base, 8-step and 4-step distillations, plus AIO and GGUF variants for 8GB/low-VRAM workflows (BF16/FP8 formats).
Provides multiple GGUF-quantized exports of Carnice V2 (a merged BF16 SFT of Qwen3.6-27B) optimized for llama.cpp and Hermes-style agent traces, with quant tiers targeted at 16–24GB local GPUs and agentic inference.
Generates expressive, scene-aware speech from XML-style prompts and supports zero-shot voice cloning from 10–20s references. Produces emotional acting, ambient SFX, multilingual output, and continuous long-form narration; requires large model weights and gated Gemma text-encoder access.
GGUF-format, DS4-optimized quantized weights for DeepSeek-V4-Flash, offering q2 (≈80.8 GiB) and q4 (≈153.3 GiB) variants plus an optional small MTP file for speculative decoding. Built for the DS4 inference engine; MIT-licensed.
Open-source Mixture-of-Experts LLM designed for extremely long-context (up to 1M tokens) text generation and agentic workflows; uses a hybrid attention + MTP design to reduce KV-cache footprint while enabling 42B active parameters and FP8 mixed-precision training.
Unified omnimodal foundation model for text, image, video and audio understanding and agentic workflows, with support for up to 1M-token context. Combines a sparse MoE LLM backbone, dedicated vision/audio encoders, multi-token prediction, and a hybrid sliding-window + global attention design to reduce KV-cache overhead.
Provides a GGUF-quantized build of NVIDIA's Nemotron 3 Nano Omni 30B (Reasoning) for local inference — enables multimodal (video/audio/image/text) reasoning, transcription, and document understanding on compatible runtimes such as llama.cpp, Ollama, vLLM, and TensorRT-LLM.
Multilingual on-device translation model compressed to 1.25-bit via the Sherry quantization, supporting 33 languages and 1,056 directions in a 440MB package for offline mobile translation and demos.