Search Results for "encoder"
Sort By:
Showing 37 open source projects for "encoder"
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.
-
Forever Free Full-Stack Observability | Grafana CloudBuilt on open standards like Prometheus and OpenTelemetry, Grafana Cloud includes Kubernetes Monitoring, Application Observability, Incident Response, plus the AI-powered Grafana Assistant. Get started with our generous free tier today.
-
1
FastVLM
This repository contains the official implementation of FastVLM
...The repository documents model variants, showcases head-to-head numbers against known baselines, and explains how the encoder integrates with common LLM backbones. Apple’s research brief frames FastVLM as targeting real-time or latency-sensitive scenarios, where lowering visual token pressure is critical to interactive UX. In short, it’s a practical recipe to make VLMs fast without exotic token-selection heuristics. -
2
Janus
Unified Multimodal Understanding and Generation Models
Janus is a sophisticated open-source project from DeepSeek AI that aims to unify both visual understanding and image generation in a single model architecture. Rather than having separate systems for “look and describe” and “prompt and generate”, Janus uses an autoregressive transformer framework with a decoupled visual encoder—allowing it to ingest images for comprehension and to produce images from text prompts with shared internal representations. The design tackles long-standing conflicts in multimodal models: namely that the visual encoder has to serve both analysis (understanding) and synthesis (generation) roles. By splitting those pathways but keeping one unified core transformer, Janus maintains flexibility and achieves strong performance across tasks previously requiring distinct architectures. ... -
3
IndexTTS2
Industrial-level controllable zero-shot text-to-speech system
IndexTTS is a modern, zero-shot text-to-speech (TTS) system engineered to deliver high-quality, natural-sounding speech synthesis with few requirements and strong voice-cloning capabilities. It builds on state-of-the-art models such as XTTS and other modern neural TTS backbones, improving them with a conformer-based speech conditional encoder and upgrading the decoder to a high-quality vocoder (BigVGAN2), leading to clearer and more natural audio output. The system supports zero-shot voice cloning — meaning it can mimic a target speaker’s voice from a short reference sample — making it versatile for multi-voice uses. Compared to many open-source TTS tools, IndexTTS emphasizes efficiency and controllability: it offers faster inference, simpler training pipelines, and controllable speech parameters (like duration, pitch, and prosody), which is critical for production use. -
4
MedGemma
Collection of Gemma 3 variants that are trained for performance
...It includes multiple variants such as a 4 billion-parameter multimodal model that can process both medical images and text and a 27 billion-parameter text-only (and multimodal) model that offers deeper clinical reasoning and understanding at higher capacity, making it suitable for complex tasks like medical question answering, summarization of clinical notes, or generating reports from radiology images. The multimodal versions pair a SigLIP-based image encoder pre-trained on diverse de-identified medical imaging data. -
Gemini 3 and 200+ AI Models on One PlatformBuild generative AI apps with Vertex AI. Switch between models without switching platforms.
-
5
DeepSeek-OCR 2
Visual Causal Flow
DeepSeek-OCR-2 is the second-generation optical character recognition system developed to improve document understanding by introducing a “visual causal flow” mechanism, enabling the encoder to reorder visual tokens in a way that better reflects semantic structure rather than strict raster scan order. It is designed to handle complex layouts and noisy documents by giving the model causal reasoning capabilities that mimic human visual scanning behavior, enhancing OCR performance on documents with rich spatial structure. ... -
6
DeepSeek VL
Towards Real-World Vision-Language Understanding
...It enables understanding and generation across visual and textual modalities—meaning it can process an image + a prompt, answer questions about images, caption, classify, or reason about visuals in context. The model is likely used internally as the visual encoder backbone for agent use cases, to ground perception in downstream tasks (e.g. answering questions about a screenshot). The repository includes model weights (or pointers to them), evaluation metrics on standard vision + language benchmarks, and configuration or architecture files. It also supports inference tools for forwarding image + prompt through the model to produce text output. ... -
7
Kimi K2.5
Moonshot's most powerful AI model
...Designed for agentic workflows, it features an Agent Swarm mechanism that decomposes complex problems into coordinated sub-agents executing in parallel. With a 256K context length and MoonViT vision encoder, the model excels across reasoning, coding, long-context comprehension, image, and video benchmarks. Kimi K2.5 is available via Moonshot’s API (OpenAI/Anthropic-compatible) and supports deployment through vLLM, SGLang, and KTransformers. -
8
GLM-OCR
Accurate × Fast × Comprehensive
GLM-OCR is an open-source multimodal optical character recognition (OCR) model built on a GLM-V encoder–decoder foundation that brings robust, accurate document understanding to complex real-world layouts and modalities. Designed to handle text recognition, table parsing, formula extraction, and general information retrieval from documents containing mixed content, GLM-OCR excels across major benchmarks while remaining highly efficient with a relatively compact parameter size (~0.9B), enabling deployment in high-concurrency services and edge environments. ... -
9
FLUX.2
Official inference repo for FLUX.2 models
...It supports high-resolution output (up to ~4 megapixels), which allows for photography-quality images, detailed product shots, infographics or UI mockups rather than just low-resolution drafts. FLUX.2 is built with a modern architecture (a flow-matching transformer + a revamped VAE + a strong vision-language encoder), enabling strong prompt adherence, correct rendering of text/typography in images, reliable lighting, layout, and physical realism, and consistent style/character/product identity across multiple generations or edits. -
$300 in Free Credit Towards Top Cloud ServicesStart your project in minutes. After credits run out, 20+ products include free monthly usage. Only pay when you're ready to scale.
-
10
Qwen-2.5-VL
Qwen2.5-VL is the multimodal large language model series
Qwen2.5 is a series of large language models developed by the Qwen team at Alibaba Cloud, designed to enhance natural language understanding and generation across multiple languages. The models are available in various sizes, including 0.5B, 1.5B, 3B, 7B, 14B, 32B, and 72B parameters, catering to diverse computational requirements. Trained on a comprehensive dataset of up to 18 trillion tokens, Qwen2.5 models exhibit significant improvements in instruction following, long-text generation... -
11
Universal Sentence Encoder
Encoder of greater-than-word length text trained on a variety of data
The Universal Sentence Encoder (USE) is a pre-trained deep learning model designed to encode sentences into fixed-length embeddings for use in various natural language processing (NLP) tasks. It leverages Transformer and Deep Averaging Network (DAN) architectures to generate embeddings that capture the semantic meaning of sentences. The model is designed for tasks like sentiment analysis, semantic textual similarity, and clustering, and provides high-quality sentence representations in a computationally efficient manner. -
12
Step3-VL-10B
Multimodal model achieving SOTA performance
...It achieves this efficiency and strong performance through unified pre-training on a massive 1.2 trillion-token multimodal corpus that jointly optimizes a language-aligned perception encoder with a powerful decoder, creating deep synergy between image processing and text understanding. -
13
NVIDIA Isaac GR00T
NVIDIA Isaac GR00T N1.5 is the world's first open foundation model
...The vision-language model remains frozen during both pretraining and finetuning, preserving language understanding and improving generalization. Streamlined MLP connection between vision encoder and LLM with added layer normalization. -
14
Stable-Dreamfusion
Text-to-3D & Image-to-3D & Mesh Exportation with NeRF + Diffusion
...Different from Imagen, Stable-Diffusion is a latent diffusion model, which diffuses in a latent space instead of the original image space. Therefore, we need the loss to propagate back from the VAE's encoder part too, which introduces extra time costs in training. We use the multi-resolution grid encoder to implement the NeRF backbone (implementation from torch-ngp), which enables much faster rendering. -
15
Stable Diffusion
A latent text-to-image diffusion model
Stable Diffusion is a widely used open-source latent text-to-image diffusion model developed by the CompVis group for generating high-quality images from natural language prompts. The model operates by conditioning a diffusion process on text embeddings produced by a CLIP text encoder, enabling detailed and controllable image synthesis. It was trained on large-scale image datasets and later fine-tuned to produce 512×512 images with strong visual fidelity. Because the system runs efficiently on consumer hardware compared to earlier generative models, it helped popularize local AI image generation workflows. The repository includes reference scripts and model configurations that allow researchers and developers to reproduce, modify, or extend the architecture. ... -
16
MAE (Masked Autoencoders)
PyTorch implementation of MAE
...It trains a Vision Transformer (ViT) by randomly masking a high percentage of image patches (typically 75%) and reconstructing the missing content from the remaining visible patches. This forces the model to learn semantic structure and global context without supervision. The encoder processes only the visible patches, while a lightweight decoder reconstructs the full image—making pretraining computationally efficient. After pretraining, the encoder serves as a powerful backbone for downstream tasks like image classification, segmentation, and detection, achieving top performance with minimal fine-tuning. The repository provides pretrained models, fine-tuning scripts, evaluation protocols, and visualization tools for reconstruction quality and learned features. -
17
fairseq-lua
Facebook AI Research Sequence-to-Sequence Toolkit
fairseq-lua is the original Lua/Torch7 version of Facebook AI Research’s sequence modeling toolkit, designed for neural machine translation (NMT) and sequence generation. 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... -
18
Denoiser
Real Time Speech Enhancement in the Waveform Domain (Interspeech 2020)
Denoiser is a real-time speech enhancement model operating directly on raw waveforms, designed to clean noisy audio while running efficiently on CPU. It uses a causal encoder-decoder architecture with skip connections, optimized with losses defined both in the time domain and frequency domain to better suppress noise while preserving speech. Unlike models that operate on spectrograms alone, this design enables lower latency and coherent waveform output. The implementation includes data augmentation techniques applied to the raw waveforms (e.g. noise mixing, reverberation) to improve model robustness and generalization to diverse noise types. ... -
19
SG2Im
Code for "Image Generation from Scene Graphs", Johnson et al, CVPR 201
...This separation lets the model reason about geometry and composition before committing to texture and color, improving spatial fidelity. The repository includes training code, datasets, and evaluation scripts so researchers can reproduce baselines and extend components such as the graph encoder or image generator. In practice, sg2im demonstrates how structured semantics can guide generative models to produce controllable, compositional imagery. -
20
Retrieval-Based Conversational Model
Dual LSTM Encoder for Dialog Response Generation
Retrieval-Based Conversational Model in Tensorflow is a project implementing a retrieval-based conversational model using a dual LSTM encoder architecture in TensorFlow, illustrating how neural networks can be trained to select appropriate responses from a fixed set of candidate replies rather than generate them from scratch. The core idea is to embed both the conversation context and potential replies into vector representations, then score how well each candidate fits the current dialogue, choosing the best match accordingly. ... -
21
Ministral 3 8B Instruct 2512
Compact 8B multimodal instruct model optimized for edge deployment
Ministral 3 8B Instruct 2512 is a balanced, efficient model in the Ministral 3 family, offering strong multimodal capabilities within a compact footprint. It combines an 8.4B-parameter language model with a 0.4B vision encoder, enabling both text reasoning and image understanding. This FP8 instruct-fine-tuned variant is optimized for chat, instruction following, and structured outputs, making it ideal for daily assistant tasks and lightweight agentic workflows. Designed for edge deployment, the model can run on a wide range of hardware and fits locally on a single 12GB GPU, with the option for even smaller quantized configurations. ... -
22
Qwen-Image-Edit
An advanced bilingual image editing with semantic control
Qwen-Image-Edit is the image editing extension of Qwen-Image, a 20B parameter model that combines advanced visual and text-rendering capabilities for creative and precise editing. It leverages both Qwen2.5-VL for semantic control and a VAE Encoder for appearance control, enabling users to edit at both the content and detail level. The model excels at semantic edits like style transfer, object rotation, and novel view synthesis, while also handling precise appearance edits such as adding or removing elements without altering surrounding regions. A standout feature is its bilingual text editing in English and Chinese, which preserves original font, size, and style during modifications. ... -
23
Mistral Large 3 675B Base 2512
Frontier-scale 675B multimodal base model for custom AI training
...The model is engineered for reliability, long-context comprehension, and stable performance across many enterprise, scientific, and knowledge-intensive workloads. Its architecture includes a powerful language MoE and a 2.5B-parameter vision encoder, enabling multimodal understanding out of the box. Mistral Large 3 Base supports deployment on-premises using FP8 or NVFP4 formats, enabling high-performance workflows on B200, H200, H100, or A100 hardware. -
24
Mistral Large 3 675B Instruct 2512 Eagle
Speculative-decoding accelerator for the 675B Mistral Large 3
...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. The Eagle variant is specialized rather than standalone, serving as a performance accelerator for production-grade assistants, agentic workflows, long-context applications, and retrieval-augmented reasoning pipelines. It supports the same multilingual, system-prompt-aligned, and function-calling behavior as the main instruct model when used in the recommended server-client configuration. -
25
Mistral Large 3 675B Instruct 2512 NVFP4
Quantized 675B multimodal instruct model optimized for NVFP4
...It retains the same instruction-tuned behavior as the FP8 model, making it ideal for production assistants, agentic workflows, scientific tasks, and long-context enterprise systems. The model integrates a 673B-parameter MoE language backbone with a 2.5B-parameter vision encoder, enabling rich multimodal analysis across text and images. Designed for efficient deployment, it runs on a single H100 or A100 node in NVFP4 while delivering performance similar to FP8 for short- and mid-context workloads.
