FedKTL

FLNK 开源

让联邦学习系统持续吸收新知识，延长寿命、降低成本

Federated Learning with New Knowledge -- explore to incorporate various new knowledg···

lora-speedrun 开源

固定任务硬件，比谁把 LoRA 微调跑得最快

Speedrunning LoRA fine-tuning: frozen task, frozen hardware, public wall-clock leade···

whichllm 开源

一条命令，找出你电脑上跑得最快的本地大模型

Find the local LLM that actually runs and performs best on your hardware. Ranked by ···

trainer 开源

在 Kubernetes 上轻松跑起分布式模型训练和微调

Distributed AI Model Training and LLM Fine-Tuning on Kubernetes

LLM-SR 开源

用大模型从数据中自动发现科学方程，告别手动推导。

[ICLR 2025 Oral] This is the official repo for the paper "LLM-SR" on Scientific Equa···

DAEDAL 开源

免训练，让扩散语言模型按需生成任意长度文本

[ICLR 2026] Official repository of "Beyond Fixed: Training-Free Variable-Length Deno···

LECO 开源

用LoRA轻量擦除扩散模型中的指定概念，不动原模型

Low-rank adaptation for Erasing COncepts from diffusion models.

maestro 开源

一行命令微调多模态模型，快速搞定视觉语言任务

streamline the fine-tuning process for multimodal models: PaliGemma 2, Florence-2, a···

MirrorMetrics 开源

用 InsightFace 人脸识别，直观诊断 LoRA 过拟合与数据质量。

MirrorMetrics: How to evaluate Stable Diffusion LoRAs. A visual diagnostic tool to d···