AI Model2026
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.
AI Dataset2026
Provides 104.9M curated image–text pairs with precomputed embeddings, structured annotations and pre-encoded VAE latents for text-to-image pretraining and retrieval. Combines filtered web sources and synthetic samples with multi-model re-captioning, deduplication and safety filters; Apache-2.0.
AI Model2026
An instruct-focused LLM (104B total, 7.4B active) optimized for fast, token-efficient inference in agent workflows. Uses hybrid linear attention plus a sparse MoE to raise throughput and cut token use; suited for high-frequency production agents, with some trade-offs in very deep reasoning.
AI Dataset2026
Contains ~1,973 distilled roleplay conversations with character-perspective chain-of-thought traces (<think> blocks) for fine-tuning persona-focused chat models. Includes teacher provenance, safety/review flags, and filters for NSFW/borderline samples — suited for SFT and character retention tests.
AI Model2026
A trillion-parameter LLM optimized for long-context, low-latency text generation and agentic coding workflows. Combines MLA+Linear Attention and a post-training 'fast thinking' token-suppression strategy to reduce token overhead and improve multi-step execution reliability for production agents.
AI Model2026
Image-to-video (I2V) diffusion model merge tuned for prompt-conditioned motion and evolution. Uses layer-scaled weight merges (not straight averaging) with BF16 and FP8 checkpoint options; prompt engineering is required for predictable motion and audio. Avoid large distilled Loras for best results.
AI Model2026
Distills DeepSeek‑V4's multi-step structured reasoning into a Qwen3.5‑9B model for fast image-text-to-text reasoning and agentic tool workflows. Trades larger teacher size for inference efficiency and improved procedural reasoning — good for low-latency research, evaluation, and agent integration.
AI Model2026
Unified 4B vision-language model for document understanding that converts images or text into template-driven structured JSON or clean Markdown. Key features: multimodal inputs (image+text), template-based extraction, reasoning vs non-reasoning modes, and vLLM/OpenAI-compatible deployment for OCR, invoice/forms extraction, and RAG preprocessing.
AI Model2026
An uncensored, fine-tuned and GGUF-quantized variant of Qwen3.6-27B tailored for long-context, coding, vision and creative-writing use. Offers multiple NEO-CODE Di-Matrix quants (IQ2/IQ4/Q6/Q8), mmproj vision support and recommended inference settings for local servers.
AI Model2026
Multilingual 2B speech–language model for ASR and bidirectional speech translation (EN, FR, DE, ES, PT, JA), providing punctuation/truecasing, keyword biasing, and a dual-head CTC encoder to boost transcription accuracy.
AI Model2026
Draft model for speculative decoding that uses a lightweight block-diffusion drafter to propose multiple tokens in parallel; designed to pair with google/gemma-4-31B-it and accelerate autoregressive text generation (official benchmarks report up to ~5.8× throughput).
AI Model2026
DavidAU/Qwen3.6-40B-Claude-4.6-Opus-Deckard-Heretic-Uncensored-Thinking-NEO-CODE-Di-IMatrix-MAX-GGUF
A 40B GGUF-quantized Qwen3.6 variant fine-tuned with Claude 4.6 Opus and Deckard/Heretic datasets for multimodal image-text-to-text tasks. Offers 256K context, custom NEO-CODE Di-IMatrix quants for long conversations and coding, optimized for local inference and creative/coding use cases; safety alignment removed.