Club&Lab for PaddlePaddle contributors
Pinned Loading
Repositories
- ms-swift Public Forked from modelscope/ms-swift
Use PEFT or Full-parameter to CPT/SFT/DPO/GRPO 600+ LLMs (Qwen3.6, DeepSeek-V4, GLM-5.1, InternLM3, Llama4, ...) and 300+ MLLMs (Qwen3-VL, Qwen3-Omni, InternVL3.5, Ovis2.5, GLM4.5v, Gemma4, Llava, Phi4, ...) (AAAI 2025).
- Megatron-LM Public Forked from NVIDIA/Megatron-LM
Ongoing research training transformer models at scale
- MoonEP Public Forked from MoonshotAI/MoonEP
MoonEP: A Perfectly Balanced Expert Parallelism Library via Dynamic Redundant Experts
- flash-linear-attention Public Forked from fla-org/flash-linear-attention
🚀 Efficient implementations of state-of-the-art linear attention models
- TeraMoE Public
TeraMoE: A cross-node expert-parallel MoE training library that uses a cooperative persistent kernel to overlap dispatch, expert compute, and combine.
- PaddleAPITest Public
- FlashKDA Public Forked from MoonshotAI/FlashKDA
FlashKDA: high-performance Kimi Delta Attention kernels
- cudnn-frontend Public Forked from NVIDIA/cudnn-frontend
cuDNN Frontend is NVIDIA's modern, open-source entry point to the cuDNN library and a growing collection of high-performance open-source kernels.
- supersonic-moe Public Forked from Dao-AILab/sonic-moe
An extended sonic-moe implementation, with FP8 support fully developed by agents
Top languages
Loading…
Most used topics
Loading…