Update README.md

README.md

@@ -17,30 +17,29 @@ This is the repository card of {repo_id} that has been pushed on the Hub. It was
 from kernels import get_kernel
 
 kernel_module = get_kernel("REPO_ID") # <- change the ID if needed
-silu_and_mul = kernel_module.silu_and_mul
+flash_attn_combine = kernel_module.flash_attn_combine
 
-silu_and_mul(...)
+flash_attn_combine(...)
 ```
 
 ## Available functions
 
-- `silu_and_mul`
-- `mul_and_silu`
-- `gelu_and_mul`
-- `gelu_tanh_and_mul`
-- `fatrelu_and_mul`
-- `gelu_fast`
-- `gelu_new`
-- `gelu_quick`
-- `gelu_tanh`
-- `silu`
-- `gelu`
-- `layers`
+- `flash_attn_combine`
+- `flash_attn_func`
+- `flash_attn_qkvpacked_func`
+- `flash_attn_varlen_func`
+- `flash_attn_with_kvcache`
+- `get_scheduler_metadata`
 
 ## Supported backends
 
 - cuda
 
+## CUDA Capabilities
+
+- 9.0a
+- 8.0
+
 ## Benchmarks
 
 [TODO: provide benchmarks if available]