Stars
fastllm是后端无依赖的高性能大模型推理库。同时支持张量并行推理稠密模型和混合模式推理MOE模型，任意10G以上显卡即可推理满血DeepSeek。双路9004/9005服务器+单显卡部署DeepSeek满血满精度原版模型，单并发20tps；INT4量化模型单并发30tps，多并发可达60+。
Community maintained hardware plugin for vLLM on Ascend
A Flexible Framework for Experiencing Cutting-edge LLM Inference Optimizations
The official repo of Qwen (通义千问) chat & pretrained large language model proposed by Alibaba Cloud.
Train a 1B LLM with 1T tokens from scratch by personal
FlashMLA: Efficient MLA decoding kernels
中文文本分类，TextCNN，TextRNN，FastText，TextRCNN，BiLSTM_Attention，DPCNN，Transformer，基于pytorch，开箱即用。
使用Bert，ERNIE，进行中文文本分类
[Paper][ACM MM 2024] Making Large Language Models Perform Better in Knowledge Graph Completion
Differentiable ODE solvers with full GPU support and O(1)-memory backpropagation.
How to Turn Your Knowledge Graph Embeddings into Generative Models
为GPT/GLM等LLM大语言模型提供实用化交互接口，特别优化论文阅读/润色/写作体验，模块化设计，支持自定义快捷按钮&函数插件，支持Python和C++等项目剖析&自译解功能，PDF/LaTex论文翻译&总结功能，支持并行问询多种LLM模型，支持chatglm3等本地模型。接入通义千问, deepseekcoder, 讯飞星火, 文心一言, llama2, rwkv, claude2, m…
Inductive relation prediction by subgraph reasoning, ICML'20
A repository contains more than 12 common statistical machine learning algorithm implementations. 常见10余种机器学习算法原理与实现及视频讲解。@月来客栈 出品
Code for ICLR'22 "Inductive Relation Prediction Using Analogy Subgraph Embeddings"
ACL 2022, SimKGC: Simple Contrastive Knowledge Graph Completion with Pre-trained Language Models
Official implementation of Neural Bellman-Ford Networks (NeurIPS 2021)