Skip to content
Change the repository type filter

All

    Repositories list

    • BreezeApp

      Public
      BreezeAPP 是一款為 Android 和 iOS 平台開發的純手機 AI 應用程式。從 App Store下載,即可在不連網的狀態下享受多項 AI 功能。源碼由聯發創新基地(MediaTek Research)提供。我們旨在推廣兩個概念: 人人都可以在自己的手機上自由選擇並運行不同的LLM - one is free to choose one's own LLM to run on a phone,以及任何app開發者都可以輕鬆寫作創意的純手機AI應用 - any dev can create purely phone-based AI apps easily。
      Java
      Apache License 2.0
      10110120Updated Jun 23, 2025Jun 23, 2025
    • A method that directly addresses the modality gap by aligning speech token with the corresponding text transcription during the tokenization stage.
      Python
      78000Updated Jun 23, 2025Jun 23, 2025
    • Breeze ASR 25 是一款先進的自動語音辨識(ASR)模型,基於 Whisper-large-v2 微調而成,特別針對台灣華語以及華語與英語混用的情境進行優化。Breeze ASR 25 is an advanced ASR model fine-tuned from Whisper-large-v2, optimized for Taiwanese Mandarin and Mandarin-English code-switching scenarios.
      Python
      MIT License
      0700Updated Jun 21, 2025Jun 21, 2025
    • Python
      Apache License 2.0
      3119421Updated Jun 21, 2025Jun 21, 2025
    • Jupyter Notebook
      0000Updated Jun 20, 2025Jun 20, 2025
    • Roo-Code

      Public
      TypeScript
      Apache License 2.0
      0005Updated Jun 16, 2025Jun 16, 2025
    • Generative Fusion Decoding (GFD) is a novel framework for integrating Large Language Models (LLMs) into multi-modal text recognition systems like ASR and OCR, improving performance and efficiency by enabling seamless fusion without requiring re-training.
      Python
      Apache License 2.0
      118222Updated May 25, 2025May 25, 2025
    • HTML
      1000Updated May 23, 2025May 23, 2025
    • JavaScript
      1000Updated May 20, 2025May 20, 2025
    • 1510Updated May 20, 2025May 20, 2025
    • TMMBench

      Public
      TMMBench - Taiwan Multi-modal Model Benchmark
      Python
      1000Updated Apr 14, 2025Apr 14, 2025
    • Repository for the paper "Towards a Foundation Model for Communication Systems"
      Apache License 2.0
      0000Updated Apr 9, 2025Apr 9, 2025
    • Python
      0000Updated Mar 9, 2025Mar 9, 2025
    • MR-Models

      Public
      聯發創新基地(MediaTek Research) 致力於研究基礎模型。我們將研究體現在適合繁體中文使用者的模型上,並在使用權許可的情況下,提供模型給學術界研究或產業界使用。
      Other
      2023550Updated Feb 26, 2025Feb 26, 2025
    • Python
      3500Updated Feb 25, 2025Feb 25, 2025
    • Python
      2400Updated Feb 18, 2025Feb 18, 2025
    • Repository for the paper "Exact, Tractable Gauss-Newton Optimization in Deep Reversible Architectures Reveal Poor Generalization" published at NeurIPS 2024
      Python
      Apache License 2.0
      0300Updated Feb 5, 2025Feb 5, 2025
    • Function Calling Leaderboard for Traditional Chinese (zh-tw)
      Python
      Apache License 2.0
      42900Updated Nov 25, 2024Nov 25, 2024
    • RAD-Bench

      Public
      RAD-Bench: A Benchmark for Evaluating Large Language Models with Context Utilization in Multi-Turn Dialogues
      Python
      Apache License 2.0
      0420Updated Aug 15, 2024Aug 15, 2024
    • TCEval

      Public
      Python
      3210Updated May 14, 2024May 14, 2024
    • FishLeg

      Public
      Coming soon...
      Python
      Apache License 2.0
      81831Updated Apr 3, 2024Apr 3, 2024
    • A framework for few-shot evaluation of autoregressive language models.
      Python
      MIT License
      2.5k000Updated Jan 5, 2024Jan 5, 2024
    • Repo for NeurIPS 23 Diffusion workshop paper titled "Score Normalization for a Faster Diffusion Exponential Integrator Sampler" from MediaTek Research UK
      Python
      MIT License
      0400Updated Dec 12, 2023Dec 12, 2023
    • Zero-shot Domain-sensitive Speech Recognition with Prompt-conditioning Fine-tuning (ASRU2023)
      Python
      Apache License 2.0
      32710Updated Oct 10, 2023Oct 10, 2023
    • Official code for the paper "Image generation with shortest path diffusion" accepted at ICML 2023.
      Python
      Apache License 2.0
      52310Updated Jul 10, 2023Jul 10, 2023
    • Code for Orthogonal Convolutions
      Python
      0100Updated Feb 21, 2020Feb 21, 2020
    • HTML
      0200Updated Jul 11, 2019Jul 11, 2019