From 263e5022a51aae4abb3f2d6ae38e05a9d9cac9d1 Mon Sep 17 00:00:00 2001
From: raxcl <raxcl@qq.com>
Date: Sun, 20 Aug 2023 11:30:23 +0800
Subject: [PATCH] Update web_demo.py
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

更好的适配笔记本用户
---
 web_demo.py | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/web_demo.py b/web_demo.py
index 1af24c9..3631f7e 100644
--- a/web_demo.py
+++ b/web_demo.py
@@ -8,6 +8,11 @@ model = AutoModel.from_pretrained("THUDM/chatglm2-6b", trust_remote_code=True).c
 # 多显卡支持，使用下面两行代替上面一行，将num_gpus改为你实际的显卡数量
 # from utils import load_model_on_gpus
 # model = load_model_on_gpus("THUDM/chatglm2-6b", num_gpus=2)
+
+# 量化int4模型示例 替换第八行即可 quantize 对应gpt内存大小（适用于gpu不够的笔记本用户，如gpu为4G，采用前者，gpu为8G，采用后者）
+# model = AutoModel.from_pretrained("THUDM/chatglm2-6b", trust_remote_code=True).half().quantize(4).cuda()
+# model = AutoModel.from_pretrained("THUDM/chatglm2-6b", trust_remote_code=True).half().quantize(8).cuda()
+
 model = model.eval()
 
 """Override Chatbot.postprocess"""