Update README.md

README.md

@@ -5,71 +5,14 @@ license: odc-by
 configs:
 - config_name: default
   data_files:
-  - split: tables
-    path: tables.jsonl
   - split: papers
     path: papers.jsonl
-  - split: train
-    path: data/train-*
-dataset_info:
-  features:
-  - name: tabid
-    dtype: string
-  - name: table
-    dtype: string
-  - name: row_bib_map
-    list:
-    - name: abstract
-      dtype: string
-    - name: bib_hash_or_arxiv_id
-      dtype: string
-    - name: corpus_id
-      dtype: int64
-    - name: row
-      dtype: int64
-    - name: title
-      dtype: string
-    - name: type
-      dtype: string
-  - name: caption
-    dtype: string
-  - name: in_text_ref
-    list:
-    - name: cite_spans
-      list:
-      - name: end
-        dtype: int64
-      - name: ref_id
-        dtype: string
-      - name: start
-        dtype: int64
-      - name: text
-        dtype: string
-    - name: content_type
-      dtype: string
-    - name: ref_spans
-      list:
-      - name: end
-        dtype: int64
-      - name: ref_id
-        dtype: string
-      - name: start
-        dtype: int64
-      - name: text
-        dtype: string
-    - name: sec_number
-      dtype: string
-    - name: sec_type
-      dtype: string
-    - name: section
-      dtype: string
-    - name: text
-      dtype: string
-  splits:
-  - name: train
-    num_bytes: 24084397
-    num_examples: 2228
-  download_size: 11342543
-  dataset_size: 24084397
+  - split: tables_dataset_viewer_only
+    path: tables_dataset_viewer_only.jsonl
 ---
 
+# ArXivDIGESTables
+
+(In progress)
+
+**Note**: Please use huggingface datasets as a way to preview the data only as `load_dataset` does not work well with the current data format. Instead it is recommended to directly download `tables.jsonl`, `papers.jsonl`, and `full_texts.jsonl.gz`.