diff --git a/README.md b/README.md
index 5cfc5fa..64314ac 100644
--- a/README.md
+++ b/README.md
@@ -1,6 +1,6 @@
 # llm.c
 
-LLM training in simple, pure C/CUDA. There is no need for 245MB of PyTorch or 107MB of cPython. Training GPT-2 (CPU, fp32) is ~1,000 lines of clean code in a single file, and training it on GPU is ~2,000 lines (adds CUDA kernels). It compiles and runs instantly, and exactly matches the PyTorch reference implementation. I chose GPT-2 as the first working example because it is the grand-daddy of LLMs, the first time the modern stack was put together.
+LLM training in simple, pure C/CUDA. There is no need for 245MB of PyTorch or 107MB of cPython. Training GPT-2 (CPU, fp32) is ~1,000 lines of clean code in the single file [train_gpt2.c](train_gpt2.c), and training it on GPU is ~2,000 lines (adds CUDA kernels) in [train_gpt2.cu](train_gpt2.cu). The code compiles and runs instantly, it exactly matches the PyTorch reference implementation, and it ~matches the speed of (compiled) PyTorch (fp32, no flash attention). I chose GPT-2 as the first working example because it is the grand-daddy of LLMs, the first time the modern stack was put together.
 
 Currently, we are working on: