nlp-greyfoss / metagradLinks
一个用于学习的仿Pytorch纯Python实现的自动求导工具。
☆51Updated last year
Alternatives and similar repositories for metagrad
Users that are interested in metagrad are comparing it to the libraries listed below
Sorting:
- an implementation of transformer, bert, gpt, and diffusion models for learning purposes☆155Updated 9 months ago
- Inference code for LLaMA models☆122Updated last year
- Implementation of FlashAttention in PyTorch☆155Updated 6 months ago
- 使用单个24G显卡,从0开始训练LLM☆56Updated last week
- mindspore implementation of transformers☆68Updated 2 years ago
- MindSpore online courses: Step into LLM☆475Updated this week
- pytorch distribute tutorials☆142Updated last month
- 记录个人的学习历程。包括但不限于算 法、机器学习、论文写作等。☆106Updated 4 months ago
- 欢迎来到 "LLM-travel" 仓库!探索大语言模型(LLM)的奥秘 🚀。致力于深入理解、探讨以及实现与大模型相关的各种技术、原理和应用。☆329Updated 11 months ago
- from MHA, MQA, GQA to MLA by 苏剑林, with code☆22Updated 4 months ago
- personal chatgpt☆377Updated 7 months ago
- ☆52Updated 2 years ago
- 一些 LLM 方面的从零复现笔记☆205Updated 2 months ago
- DeepSpeed Tutorial☆99Updated 11 months ago
- 怎么训练一个LLM分词器☆151Updated 2 years ago
- 个人总结的大模型、自然语言处理NLP、多模态、计算机视觉CV等方向paper的阅读笔记;收集到或者使用到的一些NLP、CV等领域的优秀开源仓库;其他:如数据集、评测leaderboard等☆49Updated this week
- ☆73Updated last month
- ☆140Updated last week
- pytorch分布式训练☆67Updated last year
- NumPy实现类PyTorch的动态计算图和神经网络框架(MLP, CNN, RNN, Transformer)☆81Updated last year
- simplest online-softmax notebook for explain Flash Attention☆12Updated 6 months ago
- DeepSpeed教程 & 示例注释 & 学习笔记 (大模型高效训练)☆169Updated last year
- ☆36Updated last year
- ☆111Updated 8 months ago
- 对llama3进行全参 微调、lora微调以及qlora微调。☆202Updated 9 months ago
- Train your grpo with zero dataset and low resources, 8bit/4bit/lora/qlora supported, multi-gpu supported ...☆74Updated 2 months ago
- 一个基于HuggingFace开发的大语言模型训练、测试工具。支持各模型的webui、终端预测,低参数量及全参数模型训练(预训练、SFT、RM、PPO、DPO)和融合、量化。☆217Updated last year
- A Transformer Framework Based Couplet Task☆24Updated last year
- CCKS2023-PromptCBLUE: Code implement of TianChi completition☆20Updated last year
- From Llama to Deepseek, grpo/mtp implemented. With pt/sft/lora/qlora included☆26Updated 2 months ago