From 3543e3e145141e3d8e810c4839f0c986345a6824 Mon Sep 17 00:00:00 2001
From: cc <52520497+juncaipeng@users.noreply.github.com>
Date: Tue, 10 Mar 2020 15:10:44 +0800
Subject: [PATCH] Add import for post_quant_without_data doc, test=develop,
 test=document_fix (#3137) (#3138)

---
 docs/user_guides/post_quant_no_data.md | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/docs/user_guides/post_quant_no_data.md b/docs/user_guides/post_quant_no_data.md
index 510dd1da40..04ea27a3c2 100644
--- a/docs/user_guides/post_quant_no_data.md
+++ b/docs/user_guides/post_quant_no_data.md
@@ -37,11 +37,13 @@
 对于调用无校准数据训练后量化，首先给出一个例子。
 
 ```python
+from paddle.fluid.contrib.slim.quantization import WeightQuantization
+
 model_dir = path/to/fp32_model_params
 save_model_dir = path/to/save_model_path
 weight_quant = WeightQuantization(model_dir=model_dir)
 weight_quant.quantize_weight_to_int(save_model_dir=save_model_dir,
-                                    weight_bits=8,
+                                    weight_bits=16,
                                     quantizable_op_type=['conv2d', 'depthwise_conv2d', 'mul'])
 ```
 
-- 
GitLab