Update repo refs.

Author: jayy-wu

mkdocs.yml

@@ -1,9 +1,9 @@
 site_name: Croqtile
-site_url: https://codes1gn.github.io/croqtile-tutorial/
+site_url: https://LancerLab.github.io/croqtile-tutorial/
 docs_dir: docs
 
-repo_name: codes1gn/croqtile-tutorial
-repo_url: https://github.com/codes1gn/croqtile-tutorial
+repo_name: LancerLab/croqtile-tutorial
+repo_url: https://github.com/LancerLab/croqtile-tutorial
 
 theme:
   name: material

optimization/fused-moe-fp8.md

@@ -5,7 +5,7 @@ description: "Optimizing a fused Mixture-of-Experts end-to-end kernel from 8.09
 
 # Fused MoE FP8: 8.09 → 13.18 TFLOPS (1.63×)
 
-In this tutorial, we walk through the process of optimizing a **fused Mixture-of-Experts (MoE) end-to-end kernel** written in [Croqtile](https://codes1gn.github.io/croktile-tutorial/). The target workload is real-world inference from **Qwen3.5-35B-A3B**, a model that routes each token through 8 of 256 experts. Our goal: make the full pipeline — routing, quantization, grouped GEMM, and scatter — as fast as possible on an NVIDIA H800 PCIe (SM90), using only Croqtile `.co` source code.
+In this tutorial, we walk through the process of optimizing a **fused Mixture-of-Experts (MoE) end-to-end kernel** written in [Croqtile](https://github.com/LancerLab/croqtile). The target workload is real-world inference from **Qwen3.5-35B-A3B**, a model that routes each token through 8 of 256 experts. Our goal: make the full pipeline — routing, quantization, grouped GEMM, and scatter — as fast as possible on an NVIDIA H800 PCIe (SM90), using only Croqtile `.co` source code.
 
 ### The Pipeline at a Glance
 

tutorial/ch00-installation.md

@@ -20,7 +20,7 @@ Most Linux distributions ship Flex and an older Bison. If your system Bison is b
 Clone the repository and run the automated setup:
 
 ```bash
-git clone https://github.com/codes1gn/croqtile.git
+git clone https://github.com/LancerLab/croqtile.git
 cd croqtile
 make setup-core
 ```

zh/optimization/fused-moe-fp8.md

@@ -5,7 +5,7 @@ description: "在 H800 PCIe 上将 fused Mixture-of-Experts 端到端 kernel 从
 
 # Fused MoE FP8：8.09 → 13.18 TFLOPS（1.63×）
 
-本教程逐步展示如何优化用 [鳄霸](https://codes1gn.github.io/croktile-tutorial/) 编写的 **fused Mixture-of-Experts（MoE）端到端 kernel**。目标 workload 来自真实推理场景 **Qwen3.5-35B-A3B**：每个 token 从 256 个 expert 中路由到 8 个。目标：在 NVIDIA H800 PCIe（SM90）上让整个流水线——路由、量化、grouped GEMM 与 scatter——尽可能快，且仅使用鳄霸 `.co` 源码。
+本教程逐步展示如何优化用 [鳄霸](https://github.com/LancerLab/croqtile) 编写的 **fused Mixture-of-Experts（MoE）端到端 kernel**。目标 workload 来自真实推理场景 **Qwen3.5-35B-A3B**：每个 token 从 256 个 expert 中路由到 8 个。目标：在 NVIDIA H800 PCIe（SM90）上让整个流水线——路由、量化、grouped GEMM 与 scatter——尽可能快，且仅使用鳄霸 `.co` 源码。
 
 ### 流水线一览
 

zh/tutorial/ch00-installation.md

@@ -20,7 +20,7 @@
 克隆仓库并运行自动化安装：
 
 ```bash
-git clone https://github.com/codes1gn/croqtile.git
+git clone https://github.com/LancerLab/croqtile.git
 cd croqtile
 make setup-core
 ```