liucongg
Follow
-
-
-
ChatGLM-Finetuning Public
基于ChatGLM-6B、ChatGLM2-6B、ChatGLM3-6B模型，进行下游具体任务微调，涉及Freeze、Lora、P-tuning、全参微调等
-
ChatGPTBook Public
《ChatGPT原理与实战：大型语言模型的算法、技术和私有化》
-
-
ChatGLM-6B Public
Forked from THUDM/ChatGLM-6BChatGLM-6B：开源双语对话语言模型 | An Open Bilingual Dialogue Language Model
-
leetcode-master Public
Forked from youngyangyang04/leetcode-master《代码随想录》LeetCode 刷题攻略：200道经典题目刷题顺序，共60w字的详细图解，视频难点剖析，50余张思维导图，支持C++，Java，Python，Go，JavaScript等多语言版本，从此算法学习不再迷茫！🔥🔥 来看看，你会发现相见恨晚！🚀
-
-
GPT2-NewsTitle Public
Chinese NewsTitle Generation Project by GPT2.带有超级详细注释的中文GPT2新闻标题生成项目。
-
CQG Public
Forked from sion-zcfei/CQGThe codes for ACL2022 paper “CQG: A Simple and Effective Controlled Generation Framework for Multi-hop Question Generation
Python Apache License 2.0 UpdatedMar 6, 2022 -
awesome-DeepLearning Public
Forked from PaddlePaddle/awesome-DeepLearning深度学习入门课、资深课、特色课、学术案例、产业实践案例、深度学习知识百科及面试题库The course, case and knowledge of Deep Learning and AI
-
markdown-nice Public
Forked from mdnice/markdown-nice支持主题设计的 Markdown 编辑器，让排版变 Nice
JavaScript GNU General Public License v3.0 UpdatedFeb 18, 2022 -
BlockShuffleTest Public
BlockShuffle，就是在训练过程中使用分块打乱替代随机打乱的一种方法，即将原始数据按照数据长度进行排序，然后进行batch划分，在对batch训练进行打乱。这样操作，可以减少数据padding长度，缩短训练时长。
-
NLP-Conferences-Code Public
Forked from MLNLP-World/Top-AI-Conferences-Paper-with-CodeNLP-Conferences-Code (ACL、EMNL、NAACL、COLING、AAAI、IJCAI)
-
ccks2021-track3-top1 Public
Forked from wodejiafeiyu/ccks2021-track3-top1 -
GPT2-Chinese Public
Forked from Morizeyao/GPT2-ChineseChinese version of GPT2 training code, using BERT tokenizer.
-
-
2021-GAIIC-Task3-Share Public
全球人工智能技术创新大赛-赛道三：小布助手对话短文本语义匹配
-
UnilmChatchitRobot Public
Unilm for Chinese Chitchat Robot.基于Unilm模型的夸夸式闲聊机器人项目。
-
DeepNER Public
Forked from z814081807/DeepNER天池中药说明书实体识别挑战冠军方案；中文命名实体识别；NER; BERT-CRF & BERT-SPAN & BERT-MRC；Pytorch
-
-
-
-
Deep-Learning-with-TensorFlow-book Public
Forked from dragen1860/Deep-Learning-with-TensorFlow-book深度学习开源书，基于TensorFlow 2.0实战。Open source Deep Learning book, based on TensorFlow 2.0 framework.