update hf datasets example

README.md

@@ -22,8 +22,8 @@ dataset_info:
     sequence: string
   splits:
     - name: test
-      num_bytes: 89609915
-      num_examples: 502939
+      num_bytes: 226515502
+      num_examples: 3679
 train-eval-index:
   - config: default
     task: sentence-similarity
@@ -67,10 +67,16 @@ A dataset in a [nixietune](https://github.com/nixiesearch/nixietune) compatible
 
 ## Usage
 
+To use with HF datasets:
+
+```bash
+pip install datasets zstandard
+```
+
 ```python
 from datasets import load_dataset
 
-data = load_dataset('nixiesearch/beir-eval-hardneg')
+data = load_dataset('nixiesearch/beir-eval-hard-negatives')
 print(data["test"].features)
 ```