fix typo

Author: w86763777

Readme.md

@@ -85,7 +85,7 @@ The FID is calculated by 50k generated images and CIFAR10 train set.
 ![fid_curve](https://drive.google.com/uc?export=view&id=1P4e_DEyW4wvFubPSu5t_i2gVRoecGqs5)
 
 ## Change Log
-- 2021-01-10
+- 2022-01-10
     - Update pytorch to 1.10.1 and CUDA 11.3
     - Use `pytorch_gan_metrics` to calculate FID and Inception Score
     - Use 50k generated images and CIFAR10 train set to calculate FID
@@ -95,4 +95,4 @@ The FID is calculated by 50k generated images and CIFAR10 train set.
     - Update pytorch to 1.8.1
     - Move metrics to submodule.
     - Evaluate FID on CIFAR10 test set instead of training set.
-    - Fix `cifar10.test.npz` download link and sample images.
+    - Fix `cifar10.test.npz` download link and sample images.