Point GitHub links at Colvin0315/MiniSystemOne

README.md

@@ -64,10 +64,10 @@ with a custom input packing format, not a causal LM. The packing is canonical an
 exactly one place in the source repository; re-implementing it here would create a second
 implementation that drifts from the first.
 
-The code is at **`github.com/hyusi2003/MiniSystemOne`** (Apache-2.0).
+The code is at **`github.com/Colvin0315/MiniSystemOne`** (Apache-2.0).
 
 ```bash
-git clone https://github.com/hyusi2003/MiniSystemOne && cd MiniSystemOne
+git clone https://github.com/Colvin0315/MiniSystemOne && cd MiniSystemOne
 pip install -r requirements.txt          # torch 2.6.0+cu124, transformers, datasets
 
 # put decision.pth + tokenizer.json + tokenizer_config.json into model/ and out/decision/
@@ -174,4 +174,4 @@ Latency is **overhead-bound, not compute-bound**: K=2 takes 19.88 ms and K=32 ta
 
 ## License
 
-Apache-2.0 — see the [source repository](https://github.com/hyusi2003/MiniSystemOne).
+Apache-2.0 — see the [source repository](https://github.com/Colvin0315/MiniSystemOne).