Update README.md

README.md

@@ -6,12 +6,12 @@ configs:
   - split: train
     path: bytelevel/*.parquet
 - config_name: bytelevel-subset
-  data_files:
-  - split: ngram
-    path: bytelevel-subset-ngram/*.parquet
-- config_name: ngram
   data_files:
   - split: train
+    path: bytelevel-subset/*.parquet
+- config_name: bytelevel-llm-data
+  data_files:
+  - split: ngram
     path: ngram/train-*
 language:
 - en
@@ -20,29 +20,6 @@ tags:
 pretty_name: FineWebEDU 20B
 size_categories:
 - 10B<n<100B
-dataset_info:
-  config_name: ngram
-  features:
-  - name: id
-    dtype: string
-  - name: input_ids
-    sequence: int64
-  - name: num_tokens
-    dtype: int64
-  - name: Entropy
-    sequence: float64
-  - name: Surprisal
-    sequence: float64
-  - name: Space Probability
-    sequence: float64
-  - name: EOS Probability
-    sequence: float64
-  splits:
-  - name: train
-    num_bytes: 19186769240
-    num_examples: 100000
-  download_size: 8752067839
-  dataset_size: 19186769240
 ---
 
 # FineWebEDU 20B