Search Results for "binary"

KittenTTS is an open-source, ultra-lightweight, and high-quality text-to-speech model featuring just 15 million parameters and a binary size under 25 MB. It is designed for real-time CPU-based deployment across diverse platforms. Ultra-lightweight, model size less than 25MB. CPU-optimized, runs without GPU on any device. High-quality voices, several premium voice options available. Fast inference, optimized for real-time speech synthesis.

...The repository describes how they use recursive proof decomposition by prompting DeepSeek-V3 to break complex theorems into subgoals, synthesize proof sketches, and then combine them to bootstrap training data. They then fine-tune via reinforcement learning with binary correct/incorrect feedback to integrate informal reasoning with formal proof behavior. The repo releases two model sizes (7B and 671B) and provides evaluation performance (e.g. pass rates on MiniF2F, results on ProverBench) as well as prompt / usage examples for proof generation in Lean 4. It also includes a PDF of the paper or project overview and sample formalization datasets. ...

VGGSfM is an advanced structure-from-motion (SfM) framework jointly developed by Meta AI Research (GenAI) and the University of Oxford’s Visual Geometry Group (VGG). It reconstructs 3D geometry, dense depth, and camera poses directly from unordered or sequential images and videos. The system combines learned feature matching and geometric optimization to generate high-quality camera calibrations, sparse/dense point clouds, and depth maps in standard COLMAP format. Version 2.0 adds support...

pplx-decider-v1-27b is Perplexity AI's open-weight decision model, fine-tuned from Qwen3.8-27B to perform structured classification and probabilistic decision-making across text and image inputs. Unlike conventional generative language models, it focuses on selecting answers from predefined choices or estimating the probability of binary outcomes. The model supports Choice decisions for multiclass classification and Noul decisions for yes/no judgments, returning calibrated probabilities that help applications evaluate confidence. It is designed for intent detection, request routing, factual verification, content moderation, document analysis, and automated decision workflows. ...

...Unlike conventional generative models, Quyet evaluates a supplied state, such as text, JSON, or conversation history, and answers multiple typed questions using predefined options. It supports three decision formats: Choice for categorical classification, Score for ordered scales, and Noul for binary judgments. The model calculates probabilities from next-token option predictions and applies calibrated temperatures to improve confidence estimates. It primarily supports English and is additionally tuned for Vietnamese. Quyet handles states up to 6,000 tokens within an 8,000-token prompt and supports up to 10 options per question.

...The model accepts text, structured data, and images within a 262K-token context window and supports three decision formats: Choice for categorical selection, Noul for binary judgments, and Score for ordinal evaluation. It returns probability distributions in a single forward pass, with optional reasoning for uncertain questions. Rune v3 achieved a 57.44 Decision Index score and 83.4% accuracy on a custom 1,196-item image evaluation. Temperature calibration improves the reliability of its confidence estimates. ...

...Built on Google's Gemma 4 12B IT architecture and fine-tuned on 30,000 questions, it returns calibrated probabilities for each possible answer rather than generating natural-language explanations. The model supports three decision formats: Noul for binary yes/no questions, Choice for multiclass selection, and Score for probability-based evaluation. It achieved 87.57% accuracy on DecisionBench Medium, 86.15% on JevBench, 63.10% on MMAU, and 53.10% on MVBench. Jev-Omni is designed for classification, automated decision-making, multimodal assessment, and applications requiring measurable confidence. ...