fix address.

README.md

@@ -6451,7 +6451,7 @@ We release our pipeline as open-source at [https://github.com/cisnlp/GlotCC](htt
 Replace `eng-Latn` with your specific language.
 
 ```python
-! pip install datasets
+!pip install datasets
 ```
 
 ```python
@@ -6466,7 +6466,7 @@ print(dataset['train'][0])  # First row of eng-Latn
 If you prefer not to use the Hugging Face dataloader or are interested in a specific language, you can download it directly. For example, to download the first file of `eng-Latn`:
 
 ```python
-!wget https://huggingface.co/datasets/cis-lmu/GlotCC-V1/blob/main/v1.0/eng-Latn/eng-Latn_0.parquet
+!wget https://huggingface.co/datasets/cis-lmu/GlotCC-V1/resolve/main/v1.0/eng-Latn/eng-Latn_0.parquet
 ```
 
 Then you can load it with any library that supports Parquet files, such as Pandas: