From 82d6aacf15775a736a1e31141e1c69051cb41885 Mon Sep 17 00:00:00 2001
From: intellinjun <jun.lin@intel.com>
Date: Mon, 3 Jun 2024 22:33:25 -0700
Subject: [PATCH] enable qwen2-70b

Signed-off-by: intellinjun <jun.lin@intel.com>
---
 neural_speed/models/qwen/qwen.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/neural_speed/models/qwen/qwen.h b/neural_speed/models/qwen/qwen.h
index 2dacfab02..2bca01fab 100644
--- a/neural_speed/models/qwen/qwen.h
+++ b/neural_speed/models/qwen/qwen.h
@@ -50,7 +50,7 @@ static const model_scratch qwen_mem_req(int n_layers, float scratch_size_ratio =
           static_cast<unsigned long long>(scratch_size_ratio * 2048) * MB,
           static_cast<unsigned long long>(scratch_size_ratio * 4096) * MB,
       };
-    case 70:
+    case 80:
       return {
           static_cast<unsigned long long>(scratch_size_ratio * 10*4096) * MB,
           static_cast<unsigned long long>(scratch_size_ratio * 10*2048) * MB,