From e5c8b88f90763073fc927ee232bda30fcbc05595 Mon Sep 17 00:00:00 2001
From: Laurent Mazare <laurent.mazare@gmail.com>
Date: Sun, 28 Apr 2024 08:30:35 +0200
Subject: [PATCH] Apply the cast before the scaling. (#2135)

---
 candle-nn/src/ops.rs | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/candle-nn/src/ops.rs b/candle-nn/src/ops.rs
index 1dac8c3b..7fc26c3f 100644
--- a/candle-nn/src/ops.rs
+++ b/candle-nn/src/ops.rs
@@ -70,7 +70,7 @@ pub fn dropout(xs: &Tensor, drop_p: f32) -> Result<Tensor> {
     let rand = Tensor::rand(0f32, 1f32, xs.shape(), xs.device())?;
     let scale = 1.0 / (1.0 - drop_p as f64);
     let drop_p = Tensor::new(drop_p, xs.device())?.broadcast_as(xs.shape())?;
-    let mask = (rand.ge(&drop_p)? * scale)?.to_dtype(xs.dtype())?;
+    let mask = (rand.ge(&drop_p)?.to_dtype(xs.dtype())? * scale)?;
     xs * mask
 }