Fix typo.

README.md

@@ -180,7 +180,7 @@ See the commands listed in [CONFIG.md](https://github.com/facebookresearch/moco-
 
 ### Transfer Learning
 
-See the instruction in the [transfer](https://github.com/facebookresearch/moco-v3/tree/main/transfer) dir.
+See the instructions in the [transfer](https://github.com/facebookresearch/moco-v3/tree/main/transfer) dir.
 
 ### License
 

main_moco.py

@@ -115,7 +115,7 @@ parser.add_argument('--optimizer', default='lars', type=str,
 parser.add_argument('--warmup-epochs', default=10, type=int, metavar='N',
                     help='number of warmup epochs')
 parser.add_argument('--crop-min', default=0.08, type=float,
-                    help='minimum scale for random cropping (default: 0.2)')
+                    help='minimum scale for random cropping (default: 0.08)')
 
 
 def main():

transfer/README.md

@@ -46,7 +46,7 @@ The following results are based on ImageNet-1k self-supervised pre-training, fol
 </tr>
 </tbody></table>
 
-Similar to the end-to-end fine-tuning experiment on ImageNet, the transfer learning results are also obtained using the [DeiT](https://github.com/facebookresearch/deit) repo, with the default model [deit_base_patch16_224]. 
+Similar to the end-to-end fine-tuning experiment on ImageNet, the transfer learning results are also obtained using the [DEiT](https://github.com/facebookresearch/deit) repo, with the default model [deit_base_patch16_224]. 
 
 ### Preparation: Transfer learning with ViT