Unakar
Scale up LLM efficiently! Ph.D. candidate focus on LLM Reasoning (Interpretability & RLHF)
Shanghai AI Lab intern @open-mmlab @internlmBeijing, China
Pinned Repositories
AI_Game_KingGlory
王者荣耀AI(在腾讯开悟平台实现) RL algorithm for the game: King Glory (based on kaiwu platform)
Awesome-Scaling-Law
Papers Collections for Scaling Law
Bike-REID
在监控画质下实现对校园自行车的重识别，包含REID模型识别，向量数据库检索，UI展示
ControlNet-Lighter
ControlNet with Fewer Parameters, but Comparable Performance.
DL_intorduction
Efficient_AI
此项目是我个人对MIT 6.5940 课程作业的答案，学习笔记和心得。
Kaggle-Sleep-Detection
利用腕带传感器加速度数据预测入睡苏醒时间
Scaling-Law-from-Scratch
从头开始，探讨LLM pretrain & post-train阶段的 scaling law与数据配比
VFL_with_LDP
中科大密码学大作业：在breast cancer数据集上实现了纵向联邦学习与标签差分隐私结合
Video-CLIP
Adapting CLIP for Cross Model Video Understanding
Unakar's Repositories
Unakar/AI_Game_KingGlory
王者荣耀AI(在腾讯开悟平台实现) RL algorithm for the game: King Glory (based on kaiwu platform)
Unakar/VFL_with_LDP
中科大密码学大作业：在breast cancer数据集上实现了纵向联邦学习与标签差分隐私结合
Unakar/Bike-REID
在监控画质下实现对校园自行车的重识别，包含REID模型识别，向量数据库检索，UI展示
Unakar/DL_intorduction
Unakar/Efficient_AI
此项目是我个人对MIT 6.5940 课程作业的答案，学习笔记和心得。
Unakar/Kaggle-Sleep-Detection
利用腕带传感器加速度数据预测入睡苏醒时间
Unakar/ControlNet-Lighter
ControlNet with Fewer Parameters, but Comparable Performance.
Unakar/Scaling-Law-from-Scratch
从头开始，探讨LLM pretrain & post-train阶段的 scaling law与数据配比
Unakar/Video-CLIP
Adapting CLIP for Cross Model Video Understanding
Unakar/Awesome-Scaling-Law
Papers Collections for Scaling Law
Unakar/learn_differential_privacy
Here's some notes taken by a sophmore student,mainly based on Guatum's CS680 and Dwork's wonderful book.I try to
Unakar/Retrieval-Recs-of-Douban-Data
豆瓣数据的个性化检索与推荐(2024web作业1)
Unakar/TinyLM-from-Scratch
Unakar/awesome-chatgpt-prompts
This repo includes ChatGPT prompt curation to use ChatGPT better.
Unakar/Data_Privacy
差分隐私随机梯度下降(DP-SGD)与纵向联邦学习(Vertical Federated Learning )
Unakar/Difformer
Omni Model for Image & Text Generation
Unakar/llm-action
本项目旨在分享大模型相关技术原理以及实战经验。
Unakar/USTC-CS-Resources
USTC计算机学院 个人学习资料分享
Unakar/VAR
[GPT beats diffusion🔥] [scaling laws in visual generation📈] Official impl. of "Visual Autoregressive Modeling: Scalable Image Generation via Next-Scale Prediction"