diff --git a/doc/fluid/advanced_usage/paddle_slim/paddle_slim_en.md b/doc/fluid/advanced_usage/paddle_slim/paddle_slim_en.md
index 4600bad90085924b6ca10a75c4d44e83ae4fc1ef..11e72a091203ead44a7a86f7b26ee74141351cdd 100644
--- a/doc/fluid/advanced_usage/paddle_slim/paddle_slim_en.md
+++ b/doc/fluid/advanced_usage/paddle_slim/paddle_slim_en.md
@@ -40,7 +40,7 @@ Refer to [demos](https://github.com/PaddlePaddle/models/blob/develop/PaddleSlim/
 ### Optimized Effects
 
 - For MobileNetV1 model with less redundant information, the convolution core pruning strategy can still reduce the size of the model and maintain as little accuracy loss as possible.
-- The distillation strategy can increase the accuracy of the origin model dramatically.
+- The distillation strategy can increase the accuracy of the original model dramatically.
 - Combination of the quantization strategy and the distillation strategy can reduce the size ande increase the accuracy of the model at the same time.
 
 Refer to [Performance Data and ModelZoo](https://github.com/PaddlePaddle/models/blob/develop/PaddleSlim/docs/model_zoo.md) for more details