Update README.md

Author: csyonghe

README.md

@@ -17,7 +17,7 @@ The Slang system is designed to provide developers of real-time graphics applica
 
 * The Slang compiler can generate code for a wide variety of targets and APIs: D3D12, Vulkan, D3D11, OpenGL, CUDA, and CPU. Slang code can be broadly portable, but still take advantage of the unique features of each platform.
 
-* [Automatic differentiation](https://shader-slang.com/slang/user-guide/autodiff.html) as a first-class language feature. Slang can automatically generate both forward and backward derivative propagation code for complex functions that involve arbitrary control flow and dynamic dispatch. This allows users to easily make existing rendering codebases differentiable, or to use Slang as the kernel language in a PyTorch driven machine learning framework via [`slangpy`](https://shader-slang.com/slang/user-guide/a1-02-slangpy.html).
+* [Automatic differentiation](https://shader-slang.com/slang/user-guide/autodiff.html) as a first-class language feature. Slang can automatically generate both forward and backward derivative propagation code for complex functions that involve arbitrary control flow and dynamic dispatch. This allows users to easily make existing rendering codebases differentiable, or to use Slang as the kernel language in a PyTorch driven machine learning framework via [`slangtorch`](https://shader-slang.com/slang/user-guide/a1-02-slangpy.html).
 
 * Generics and interfaces allow shader specialization to be expressed cleanly without resort to preprocessor techniques or string-pasting. Unlike C++ templates, Slang's generics are checked ahead of time and don't produce cascading error messages that are difficult to diagnose. The same generic shader can be specialized for a variety of different types to produce specialized code ahead of time, or on the fly, completely under application control.