From cf0fae9a55e97148ac1ac27f251f4fc4e2a0c968 Mon Sep 17 00:00:00 2001
From: Daniel Han-Chen <danielhanchen@gmail.com>
Date: Sat, 3 Feb 2024 23:01:56 +1100
Subject: [PATCH] Update llama.py

---
 unsloth/models/llama.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/unsloth/models/llama.py b/unsloth/models/llama.py
index e2edd9c4da..bc401dbbe4 100644
--- a/unsloth/models/llama.py
+++ b/unsloth/models/llama.py
@@ -160,7 +160,7 @@ def LlamaAttention_fast_forward_inference(
     sin = self.rotary_emb.sin_cached[seq_len]
     h = head_dim // 2
 
-    RH_Q = self.RH_Q
+    # RH_Q = self.RH_Q
     RH_Q[:,:,:,:h] = Qn[:,:,:,h:]; RH_Q[:,:,:,h:] = Qn[:,:,:,:h]; torch.neg(RH_Q[:,:,:,:h], out = RH_Q[:,:,:,:h]);
     Qn *= cos; Qn.addcmul_(RH_Q, sin);