RF-Inversion – 無需 ControlNet 的圖片編輯

GitHub – LituRout/RF-Inversion: Rectified Flow Inversion (RF-Inversion)

（附 ComfyUI Node）Rectified Flow Inversion (RF-Inversion). Contribute to LituRout/RF-Inversion development by creating an account on GitHub.

---

Diffusion – 生成 3D 遊戲的擴散式世界模型

構成了一種以安全且樣本高效的方式訓練強化學習 💎

---

DH-Live實時直播數字人

DH-Live实时直播数字人，全新商用项目开源，超低显存占用，人人可用！

---

DiffSynth-Studio 擴散模型的魔力

DiffSynth Studio 是個擴散引擎

重構了文字編碼器、UNet、VAE等架構，保持與開源社群模型的相容性，同時增強運算效能。

---

Llama 3.2 Vision Instruct 詳細的安裝教學

Llama 3.2 Vision Instruct - Installation & Usage Tutorial

---

Open-LLM-VTuber

透過免持語音互動、語音中斷、Live2D 臉部辨識和跨平臺本地運行的長期記憶與任何 LLM 交談

LLM 推理後端、語音辨識和語音合成器均設計為可交換。此專案可以配置為在 macOS、Linux 和 Windows 上離線運行。也支援線上 LLM/ASR/TTS 選項。

---

Llama3 微調神器

微调Llama 3 1，用神器Unsloth

---

Speech-to-speech 語音到語音開源模組

GitHub – huggingface/speech-to-speech: Speech To Speech: an effort for an open-sourced and modular GPT4-o

Speech To Speech: an effort for an open-sourced and modular GPT4-o – huggingface/speech-to-speech

---

CogVideo 文字 > 影片產生

GitHub – THUDM/CogVideo: Text-to-video generation: CogVideoX (2024) and CogVideo (ICLR 2023)

Text-to-video generation: CogVideoX (2024) and CogVideo (ICLR 2023) – THUDM/CogVideo

---

RagFlow：終極 RAG 引擎 – 語意搜尋、嵌入、向量搜尋 + 支援圖形！

RagFlow: Ultimate RAG Engine - Semantic Search, Embeddings, Vector Search + Supports Graph!

---