Search Results for "encoder"
Sort By:
Showing 92 open source projects for "encoder"
View related business solutions-
Full-stack observability with actually useful AI | 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.
-
Go From AI Idea to AI App FastAccess Gemini 3 and 200+ models. Build chatbots, agents, or custom models with built-in monitoring and scaling.
-
1
Pytorch-toolbelt
PyTorch extensions for fast R&D prototyping and Kaggle farming
...Extras for Catalyst library (Visualization of batch predictions, additional metrics). By design, both encoder and decoder produces a list of tensors, from fine (high-resolution, indexed 0) to coarse (low-resolution) feature maps. Access to all intermediate feature maps is beneficial if you want to apply deep supervision losses on them or encoder-decoder of object detection task. -
2
ModernBERT
Bringing BERT into modernity via both architecture changes and scaling
ModernBERT is an open-source research project that modernizes the classic BERT encoder architecture by incorporating recent advances in transformer design, training techniques, and efficiency improvements. The goal of the project is to bring BERT-style models up to date with the capabilities of modern large language models while preserving the strengths of bidirectional encoder architectures used for tasks such as classification, retrieval, and semantic search. -
3
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. ... -
4
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. -
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.
-
5
Real-Time Voice Cloning
Clone a voice in 5 seconds to generate arbitrary speech in real-time
Real-Time Voice Cloning is an influential deep-learning repository that demonstrates how to clone a voice from just a few seconds of audio and then generate arbitrary speech in that voice in near real time. It implements the SV2TTS pipeline (“Transfer Learning from Speaker Verification to Multispeaker Text-To-Speech Synthesis”) in three stages: a speaker encoder, a synthesizer, and a vocoder. In the first stage, short audio clips are converted into a fixed-dimensional speaker embedding that captures voice characteristics; this embedding is then used by a Tacotron-style synthesizer to generate spectrograms from text, which a WaveRNN-based vocoder finally turns into audio. The repo includes both a command-line demo and a graphical “toolbox” application where you can load reference voices, type text, and hear the synthesized results interactively. ... -
6
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. -
7
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. -
8
FireRedASR
Open-source industrial-grade ASR models
FireRedASR is an industrial-grade family of open-source automatic speech recognition models designed to provide high-precision speech-to-text performance across languages including Mandarin, English, and various Chinese dialects, achieving new state-of-the-art benchmarks on public test sets. The project includes multiple model variants to meet different application needs, such as high-accuracy end-to-end interaction using an encoder-adapter-LLM framework and efficient real-time recognition using attention-based encoder-decoder architectures, giving developers flexibility in balancing performance and resource constraints. FireRedASR not only excels in traditional speech recognition tasks but also demonstrates strong capability in challenging scenarios like singing lyrics recognition, where accurate transcription is often difficult for conventional models. -
9
Bootstrap Your Own Latent (BYOL)
Usable Implementation of "Bootstrap Your Own Latent" self-supervised
Practical implementation of an astoundingly simple method for self-supervised learning that achieves a new state-of-the-art (surpassing SimCLR) without contrastive learning and having to designate negative pairs. This repository offers a module that one can easily wrap any image-based neural network (residual network, discriminator, policy network) to immediately start benefitting from unlabelled image data. There is now new evidence that batch normalization is key to making this technique... -
Try Google Cloud Risk-Free With $300 in CreditUse your credit across every product. Compute, storage, AI, analytics. When it runs out, 20+ products stay free. You only pay when you choose to.
-
10
ESPnet
End-to-end speech processing toolkit
...ESPnet provides many ready-to-run recipes for popular academic benchmarks, making it straightforward to reproduce published results or serve as baselines for new research. The toolkit also hosts numerous pretrained models and example configs, ranging from Transformer and Conformer architectures to various attention-based encoder-decoder models. -
11
Perception Models
State-of-the-art Image & Video CLIP, Multimodal Large Language Models
Perception Models is a state-of-the-art framework developed by Facebook Research for advanced image and video perception tasks. It introduces two primary components: the Perception Encoder (PE) for visual feature extraction and the Perception Language Model (PLM) for multimodal decoding and reasoning. The PE module is a family of vision encoders designed to excel in image and video understanding, surpassing models like SigLIP2, InternVideo2, and DINOv2 across multiple benchmarks. Meanwhile, PLM integrates with PE to power vision-language modeling, achieving results competitive with leading multimodal systems such as QwenVL2.5 and InternVL3, all while being fully reproducible with open data. ... -
12
Memvid
Video-based AI memory library. Store millions of text chunks in MP4
Memvid encodes text chunks as QR codes within MP4 frames to build a portable “video memory” for AI systems. This innovative approach uses standard video containers and offers millisecond-level semantic search across large corpora with dramatically less storage than vector DBs. It's self-contained—no DB needed—and supports features like PDF indexing, chat integration, and cloud dashboards. -
13
JEPA
PyTorch code and models for V-JEPA self-supervised learning from video
JEPA (Joint-Embedding Predictive Architecture) captures the idea of predicting missing high-level representations rather than reconstructing pixels, aiming for robust, scalable self-supervised learning. A context encoder ingests visible regions and predicts target embeddings for masked regions produced by a separate target encoder, avoiding low-level reconstruction losses that can overfit to texture. This makes learning focus on semantics and structure, yielding features that transfer well with simple linear probes and minimal fine-tuning. The repository provides training recipes, data pipelines, and evaluation utilities for image JEPA variants and often includes ablations that illuminate which masking and architectural choices matter. ... -
14
MoCo (Momentum Contrast)
Self-supervised visual learning using momentum contrast in PyTorch
...It introduces Momentum Contrast (MoCo), a scalable approach to self-supervised learning that enables visual representation learning without labeled data. The core idea of MoCo is to maintain a dynamic dictionary with a momentum-updated encoder, allowing efficient contrastive learning across large batches. The repository includes implementations for both MoCo v1 and MoCo v2, the latter improving training stability and performance through architectural and augmentation enhancements. Training is optimized for distributed multi-GPU environments, using DistributedDataParallel for speed and simplicity. -
15
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. ... -
16
vJEPA-2
PyTorch code and models for VJEPA2 self-supervised learning from video
VJEPA2 is a next-generation self-supervised learning framework for video that extends the “predict in representation space” idea from i-JEPA to the temporal domain. Instead of reconstructing pixels, it predicts the missing high-level embeddings of masked space-time regions using a context encoder and a slowly updated target encoder. This objective encourages the model to learn semantics, motion, and long-range structure without the shortcuts that pixel-level losses can invite. The architecture is designed to scale: spatiotemporal ViT backbones, flexible masking schedules, and efficient sampling let it train on long clips while remaining stable. ... -
17
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. -
18
CTranslate2
Fast inference engine for Transformer models
CTranslate2 is a C++ and Python library for efficient inference with Transformer models. The project implements a custom runtime that applies many performance optimization techniques such as weights quantization, layers fusion, batch reordering, etc., to accelerate and reduce the memory usage of Transformer models on CPU and GPU. The execution is significantly faster and requires less resources than general-purpose deep learning frameworks on supported models and tasks thanks to many... -
19
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. -
20
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. ... -
21
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. ... -
22
Segment Anything
Provides code for running inference with the SegmentAnything Model
...It’s a promptable segmenter: you guide it with points, boxes, or rough masks, and it predicts high-quality object masks consistent with the prompt. The architecture separates a powerful image encoder from a lightweight mask decoder, so the heavy vision work can be computed once and the interactive part stays fast. A bundled automatic mask generator can sweep an image and propose many object masks, which is useful for dataset bootstrapping or bulk annotation. The repository includes ready-to-use weights, Python APIs, and example notebooks demonstrating both interactive and automatic modes. ... -
23
LatentSync
Taming Stable Diffusion for Lip Sync
...In effect, given a source video (with masked or reference frames) and an audio track, LatentSync directly generates frames whose lip motions and expressions align with the audio, producing convincing talking-head or animated lip-sync output. The system leverages a U-Net diffusion backbone, with cross-attention of audio embeddings (via an audio encoder) and reference video frames to guide generation, and applies a set of loss functions (temporal, perceptual, sync-net based) to enforce lip-sync accuracy, visual fidelity, and temporal consistency. Over versions, LatentSync has improved temporal stability and lowered resource requirements — making inference more practical (e.g. 8 GB VRAM for earlier versions, somewhat higher for latest models). -
24
pytorch-cpp
C++ Implementation of PyTorch Tutorials for Everyone
C++ Implementation of PyTorch Tutorials for Everyone. This repository provides tutorial code in C++ for deep learning researchers to learn PyTorch (i.e. Section 1 to 3) Interactive Tutorials are currently running on LibTorch Nightly Version. Libtorch only supports 64bit Windows and an x64 generator needs to be specified. Create all required script module files for pre-learned models/weights during the build. Requires installed python3 with PyTorch and torch-vision. You can choose to only... -
25
x-transformers
A simple but complete full-attention transformer
A simple but complete full-attention transformer with a set of promising experimental features from various papers. Proposes adding learned memory key/values prior to attending. They were able to remove feedforwards altogether and attain a similar performance to the original transformers. I have found that keeping the feedforwards and adding the memory key/values leads to even better performance. Proposes adding learned tokens, akin to CLS tokens, named memory tokens, that is passed through...
