From edd6440456689c25cdaff09f6d722eaa9f088b70 Mon Sep 17 00:00:00 2001 From: orangeH25 <18085625039@163.com> Date: Tue, 9 Sep 2025 08:26:44 +0000 Subject: [PATCH] docs: fix URL in basic dtype section --- docs/source/quantization_overview.rst | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/docs/source/quantization_overview.rst b/docs/source/quantization_overview.rst index bdb675eff8..f5c82bfe5f 100644 --- a/docs/source/quantization_overview.rst +++ b/docs/source/quantization_overview.rst @@ -21,7 +21,7 @@ Any quantization algorithm will be using some components from the above stack, f Basic DTypes ~~~~~~~~~~~~ -`dtype `__ is a bit of overloaded term, by basic dtype, we mean the dtypes that makes sense without any extra metadata (e.g. makes sense when people call ``torch.empty(.., dtype)``), for more details please check out `this post `__. +`dtype `__ is a bit of overloaded term, by basic dtype, we mean the dtypes that makes sense without any extra metadata (e.g. makes sense when people call ``torch.empty(.., dtype)``), for more details please check out `this post `__. No matter what quantization we are doing, in the end we will be using some low precision dtypes to represent the quantized data or quantization parameters, the low precision dtypes relevant for torchao are: