hhaAndroid/mmdetection-mini
mmdetection最小学习版
LLM&MLLM Infra, RL
mmdetection最小学习版
mini dataloader
多模态 MM +Chat 合集
Miles is an enterprise-facing reinforcement learning framework for LLM and VLM post-training, forked from and co-evolving with slime.
Scalable toolkit for efficient model reinforcement
Ongoing research training transformer models at scale
DeepSpec: a full-stack codebase for training and evaluating speculative decoding algorithms
自动监控开源项目 PR 动态
yolov5的注释版本
A lightweight framework for building LLM-based agents
Ray is an AI compute engine. Ray consists of a core distributed runtime and a set of AI Libraries for accelerating ML workloads.
AReaL inference pipeline example - AI generated demo
GatedDeltaNet implementation with variable length sequence support
verl: Volcano Engine Reinforcement Learning for LLMs
SGLang is a fast serving framework for large language models and vision language models.
🤗 Transformers: the model-definition framework for state-of-the-art machine learning models in text, vision, audio, and multimodal models, for both inference and training.
A Scientific Multimodal Foundation Model
OpenMMLab Detection Toolbox and Benchmark
Use PEFT or Full-parameter to CPT/SFT/DPO/GRPO 500+ LLMs (Qwen3, Qwen3-MoE, Llama4, GLM4.5, InternLM3, DeepSeek-R1, ...) and 200+ MLLMs (Qwen2.5-VL, Qwen2.5-Omni, Qwen2-Audio, InternVL3, Ovis2.5, Llava, GLM4v, Phi4, ...) (AAAI 2025).
slime is a LLM post-training framework aiming at scaling RL.
Code segment are often used in deep learning algorithms(pytorch/numpy)
Towards Economical Inference: Enabling DeepSeek's Multi-Head Latent Attention in Any Transformer-based LLMs
A native PyTorch Library for large model training
Aligning and Prompting Everything All at Once for Universal Visual Perception
深度学习500问,以问答形式对常用的概率知识、线性代数、机器学习、深度学习、计算机视觉等热点问题进行阐述,以帮助自己及有需要的读者。 全书分为15个章节,近20万字。由于水平有限,书中不妥之处恳请广大读者批评指正。 未完待续............ 如有意合作,联系[email protected] 版权所有,违权必究 Tan 2018.06
DeepSpeed is a deep learning optimization library that makes distributed training and inference easy, efficient, and effective.
Example models using DeepSpeed