Search Results for "decoding"
Sort By:
Showing 16 open source projects for "decoding"
View related business solutions-
MongoDB Atlas runs apps anywhereMongoDB Atlas gives you the freedom to build and run modern applications anywhere—across AWS, Azure, and Google Cloud. With global availability in over 115 regions, Atlas lets you deploy close to your users, meet compliance needs, and scale with confidence across any geography.
-
Build Securely on AWS with Proven FrameworksMoving to the cloud brings new challenges. How can you manage a larger attack surface while ensuring great network performance? Turn to Fortinet’s Tested Reference Architectures, blueprints for designing and securing cloud environments built by cybersecurity experts. Learn more and explore use cases in this white paper.
-
1
DFlash
Block Diffusion for Ultra-Fast Speculative Decoding
DFlash is an open-source framework for ultra-fast speculative decoding using a lightweight block diffusion model to draft text in parallel with a target large language model, dramatically improving inference speed without sacrificing generation quality. It acts as a “drafter” that proposes likely continuations which the main model then verifies, enabling significant throughput gains compared to traditional autoregressive decoding methods that generate token by token. -
2
FlashMLA
FlashMLA: Efficient Multi-head Latent Attention Kernels
FlashMLA is a high-performance decoding kernel library designed especially for Multi-Head Latent Attention (MLA) workloads, targeting NVIDIA Hopper GPU architectures. It provides optimized kernels for MLA decoding, including support for variable-length sequences, helping reduce latency and increase throughput in model inference systems using that attention style. The library supports both BF16 and FP16 data types, and includes a paged KV cache implementation with a block size of 64 to efficiently manage memory during decoding. -
3
MiniCPM4.1
Achieving 3+ generation speedup on reasoning tasks
MiniCPM4.1 is an enhanced iteration of the MiniCPM4 architecture, introducing improvements in reasoning capabilities, inference speed, and hybrid operation modes that allow dynamic switching between deep reasoning and standard generation. It builds upon the same efficiency-focused philosophy but further optimizes decoding performance, achieving substantial speed gains in reasoning-intensive tasks while maintaining high-quality outputs. One of its key innovations is the hybrid reasoning mode, which allows developers to control whether the model engages in deeper reasoning processes or faster responses depending on the use case. The model also supports both dense and sparse attention mechanisms, enabling more efficient computation depending on the selected inference framework. ... -
4
Tiktoken
tiktoken is a fast BPE tokeniser for use with OpenAI's models
...It also offers extension mechanisms so that custom encodings can be registered. Internally, it includes the core tokenizer logic (often implemented in Rust or efficient lower-level code), APIs for encoding, decoding, and counting tokens, and binding layers to Python (and sometimes other languages) for easy use. -
$300 Free Credits for Your Google Cloud ProjectsLaunch your next project with $300 in free Google Cloud credits—no strings attached. Test, build, and deploy without risk. Use your credits across the entire Google Cloud platform to find what works best for your needs. After your credits are used, continue with always-free tier services. Only pay when you're ready to scale. Sign up in minutes and start exploring.
-
5
LTX-Video
Official repository for LTX-Video
LTX-Video is a sophisticated multimedia processing framework from Lightricks designed to handle high-quality video editing, compositing, and transformation tasks with performance and scalability. It provides runtime components that efficiently decode, encode, and manipulate video streams, frame buffers, and audio tracks while exposing a rich API for building customized editing features like transitions, effects, color grading, and keyframe automation. The toolkit is built with both real-time... -
6
stable-diffusion.cpp
Diffusion model(SD,Flux,Wan,Qwen Image,Z-Image,...) inference
stable-diffusion.cpp is a lightweight, high-performance implementation of Stable Diffusion and related generative models written entirely in portable C/C++, designed to run on virtually any device without heavy dependencies. It enables text-to-image and image-to-image generation, supports a growing set of models like SD1.x, SD2.x, SDXL, SD-Turbo, Qwen Image, and more, and is continually updated with support for cutting-edge model variants including video and image editing models. The project... -
7
Colibrì
Run GLM-5.2 (744B MoE) on a 25GB-RAM consumer machine
...The runtime is implemented in pure C, requires no Python or BLAS during inference, and can operate without a GPU. Compressed attention caches, expert caching, optional hot tiers, and speculative decoding reduce memory pressure and improve repeated use. A planning tool calculates safe disk, RAM, and VRAM placement before loading the model, while a diagnostic command checks system readiness. Colibri includes terminal chat, an OpenAI-compatible text API, and a browser client, but disk-bound generation can be slow on cold caches. -
8
GLM-4.5
GLM-4.5: Open-source LLM for intelligent agents by Z.ai
GLM-4.5 is a cutting-edge open-source large language model designed by Z.ai for intelligent agent applications. The flagship GLM-4.5 model has 355 billion total parameters with 32 billion active parameters, while the compact GLM-4.5-Air version offers 106 billion total parameters and 12 billion active parameters. Both models unify reasoning, coding, and intelligent agent capabilities, providing two modes: a thinking mode for complex reasoning and tool usage, and a non-thinking mode for... -
9
Watermark Anything
Official implementation of Watermark Anything with Localized Messages
Watermark Anything (WAM) is an advanced deep learning framework for embedding and detecting localized watermarks in digital images. Developed by Facebook Research, it provides a robust, flexible system that allows users to insert one or multiple watermarks within selected image regions while maintaining visual quality and recoverability. Unlike traditional watermarking methods that rely on uniform embedding, WAM supports spatially localized watermarks, enabling targeted protection of... -
Ship Agents FasterGemini Enterprise Agent Platform lets you rapidly build, scale, govern and optimize production-ready agents grounded in your organization's data. The platform enables developers to build custom or pre-built agents for virtually any use case. New customers get $300 in free credits.
-
10
MiniCPM4
Ultra-Efficient LLMs on End Device
MiniCPM4 is part of the MiniCPM family of ultra-efficient large language models designed specifically for high performance on edge devices and resource-constrained environments. Unlike traditional large-scale models that require extensive computational resources, MiniCPM4 focuses on delivering competitive reasoning and language capabilities while maintaining significantly lower latency and higher efficiency. It achieves this through optimized architectures, scalable training strategies, and... -
11
Chinese-LLaMA-Alpaca-2 v2.0
Chinese LLaMA & Alpaca large language model + local CPU/GPU training
This project has open-sourced the Chinese LLaMA model and the Alpaca large model with instruction fine-tuning to further promote the open research of large models in the Chinese NLP community. Based on the original LLaMA , these models expand the Chinese vocabulary and use Chinese data for secondary pre-training, which further improves the basic semantic understanding of Chinese. At the same time, the Chinese Alpaca model further uses Chinese instruction data for fine-tuning, which... -
12
fairseq-lua
Facebook AI Research Sequence-to-Sequence Toolkit
...It introduced early attention-based architectures and training pipelines that later evolved into the modern PyTorch-based fairseq. The framework implements sequence-to-sequence models with attention, beam search decoding, and distributed training, providing a research platform for exploring translation, summarization, and language modeling. Its modular design made it easy to prototype new architectures by modifying encoders, decoders, or attention mechanisms. Although now deprecated in favor of the PyTorch rewrite, fairseq-lua played a key role in advancing large-scale NMT systems, such as early versions of Facebook’s production translation models. ... -
13
Image GPT
Large-scale autoregressive pixel model for image generation by OpenAI
Image-GPT is the official research code and models from OpenAI’s paper Generative Pretraining from Pixels. The project adapts GPT-2 to the image domain, showing that the same transformer architecture can model sequences of pixels without altering its fundamental structure. It provides scripts to download pretrained checkpoints of different model sizes (small, medium, large) trained on large-scale datasets and includes utilities for handling color quantization with a 9-bit palette.... -
14
Mistral Large 3 675B Instruct 2512 Eagle
Speculative-decoding accelerator for the 675B Mistral Large 3
Mistral Large 3 675B Instruct 2512 Eagle is the dedicated speculative-decoding draft model for the full Mistral Large 3 Instruct system, designed to significantly speed up generation while preserving high output quality. It works alongside the primary 675B instruct model, enabling faster response times by predicting several tokens ahead using Mistral’s Eagle speculative method. Built on the same frontier-scale multimodal Mixture-of-Experts architecture, it complements a system featuring 41B active parameters and a 2.5B-parameter vision encoder. ... -
15
GigaChat 3 Ultra
High-performance MoE model with MLA, MTP, and multilingual reasoning
...The model also employs Multi-Token Prediction, enabling multi-step token generation in a single pass for up to 40% faster output through speculative and parallel decoding techniques. Its training corpus incorporates ten languages, enriched with books, academic sources, code datasets, mathematical tasks, and more than 5.5 trillion tokens of high-quality synthetic data. This combination significantly boosts reasoning, coding, and multilingual performance across modern benchmarks. Designed for high-performance deployment, GigaChat 3 Ultra supports major inference engines and offers optimized BF16 and FP8 execution paths for cluster-grade hardware. -
16
Hy3 preview
Efficient MoE model for reasoning, coding, and AI agent workflows
...The model features 295B total parameters with only 21B activated during inference, plus a dedicated 3.8B Multi-Token Prediction (MTP) layer that accelerates generation through speculative decoding. Architecturally, it uses 192 routed experts with top-8 activation, a dense-MoE hybrid design, and a native 256K-token context window. Hy3-preview is optimized for efficient deployment while maintaining strong benchmark performance across reasoning, coding, and agent evaluations. It supports function calling, integration with popular agent frameworks such as OpenClaw and OpenCode, and deployment through Transformers.
- Previous
- You're on page 1
- Next