remove single-threaded reference in docs

metascroy · metascroy · commit 74d221e50bd9 · 2024-09-19T09:08:19.000-07:00
diff --git a/docs/quantization.md b/docs/quantization.md
@@ -178,7 +178,6 @@ python torchchat.py export llama3 --device cpu --dtype float32 --quantize '{"lin
 ```
 
 Note: only the ExecuTorch C++ runner in torchchat when built using the instructions in the setup can run the exported *.pte file.
-Also note that the ExecuTorch op that wraps the new torchao kernel is currently single threaded.
 
 ## Quantization Profiles
 
