Repository profile
datawhalechina / diy-llm
Covers pre-training data, Tokenizer, Transformer, MoE,distributed training, Scaling Laws, inference & alignment .6 progressive code assignments for full-stack LLM learning | 涵盖预训练数据、分词器、Transformer、MoE、分布式训练、缩放定律、推理与对齐，6 项渐进代码作业，掌握 LLM 全栈知识
Jupyter NotebookView on GitHub
Repository overview
Public GitHub metrics and OpenPulse observationsGitHub stars2,640
Public forks247
Activity · 4 weeks
High activity70/100
Growth · 24h+61 stars
Growth · 7d+815 stars
Growth · 30d+1,080 stars
Repository activity
Four complete UTC weeks · 2026-10-04 to 2026-11-01Commits15
Merged pull requests11
Active contributors2
Relative score across tracked public repositories with complete data. It describes recent activity, not project quality. Updated .
AI-attributed commits Pilot
Not yet analyzed · hot-project pilot
Share of checked commits with explicit AI attribution, not the share of AI-generated code.
View basis
Based on explicit public attribution in recent default-branch commits. Not the percentage of AI-generated code; unmarked use cannot be detected.
Star activity
Last 7 days26 Oct8 observed days · gaps preserved02 Nov
Observed star growth
Samples or dated GitHub history
24h
+61 stars
Measured from observations
01 Nov 2026 → 02 Nov 2026
7d
+815 stars
Measured from observations
26 Oct 2026 → 02 Nov 2026
30d
+1,080 stars
GitHub history · as of 2026-10-30
01 Oct 2026 → 31 Oct 2026