Update README.md

README.md

@@ -27,147 +27,234 @@ license: mit
 language:
   - en
 dataset_info:
-  description: >
-    This repository contains the SuperGLUE benchmark datasets.
-    Each dataset is available as a separate configuration and includes
-    detailed preprocessing to ensure consistency across train, validation,
-    and test splits.
-  total_datasets: 8
-  datasets:
-    multirc:
-      name: MultiRC
-      sample_counts:
-        train: 456
-        validation: 83
-        test: 166
-      features:
-        idx: int
-        version: float
-        passage:
-          __type__: dict
-          schema:
-            questions:
-              __type__: list
-              schema:
-                idx: int
-                answers:
-                  __type__: list
-                  schema:
-                    idx: int
-                    label: int
-                    text: str
-                question: str
-            text: str
-    boolq:
-      name: BoolQ
-      sample_counts:
-        train: 9427
-        validation: 3270
-        test: 3245
-      features:
-        question: str
-        passage: str
-        idx: int
-        label: bool
-    cb:
-      name: CB
-      sample_counts:
-        train: 250
-        validation: 56
-        test: 250
-      features:
-        premise: str
-        hypothesis: str
-        label: str
-        idx: int
-    wic:
-      name: WiC
-      sample_counts:
-        train: 5428
-        validation: 638
-        test: 1400
-      features:
-        word: str
-        sentence1: str
-        sentence2: str
-        idx: int
-        label: bool
-        start1: int
-        start2: int
-        end1: int
-        end2: int
-        version: float
-    record:
-      name: ReCoRD
-      sample_counts:
-        train: 65709
-        validation: 7481
-        test: 7484
-      features:
-        source: str
-        passage:
-          __type__: dict
-          schema:
-            entities:
-              __type__: list
-              schema:
-                end: int
-                start: int
-            text: str
-        qas:
+- config_name: multirc
+  features:
+  - name: idx
+    dtype: int32
+  - name: version
+    dtype: float32
+  - name: passage
+    dtype:
+      __type__: dict
+      schema:
+        text: str
+        questions:
           __type__: list
           schema:
-            query: str
-            idx: int
+            question: str
             answers:
               __type__: list
               schema:
-                end: int
-                start: int
                 text: str
-        idx: int
-    copa:
-      name: COPA
-      sample_counts:
-        train: 400
-        validation: 100
-        test: 500
-      features:
-        premise: str
-        choice1: str
-        choice2: str
-        question: str
-        label: int
-        idx: int
-    rte:
-      name: RTE
-      sample_counts:
-        train: 2490
-        validation: 277
-        test: 3000
-      features:
-        premise: str
-        hypothesis: str
-        label: str
-        idx: int
-    wsc:
-      name: WSC
-      sample_counts:
-        train: 554
-        validation: 104
-        test: 146
-      features:
+                idx: int
+                label: int
+            idx: int
+  splits:
+  - name: train
+    num_bytes: 3094341
+    num_examples: 456
+  - name: validation
+    num_bytes: 550086
+    num_examples: 83
+  - name: test
+    num_bytes: 983819
+    num_examples: 166
+  download_size: 4628246
+  dataset_size: 4628246
+- config_name: boolq
+  features:
+  - name: question
+    dtype: string
+  - name: passage
+    dtype: string
+  - name: idx
+    dtype: int32
+  - name: label
+    dtype: bool
+  splits:
+  - name: train
+    num_bytes: 6573302
+    num_examples: 9427
+  - name: validation
+    num_bytes: 2254565
+    num_examples: 3270
+  - name: test
+    num_bytes: 2193238
+    num_examples: 3245
+  download_size: 11021105
+  dataset_size: 11021105
+- config_name: cb
+  features:
+  - name: premise
+    dtype: string
+  - name: hypothesis
+    dtype: string
+  - name: label
+    dtype: string
+  - name: idx
+    dtype: int32
+  splits:
+  - name: train
+    num_bytes: 99499
+    num_examples: 250
+  - name: validation
+    num_bytes: 24598
+    num_examples: 56
+  - name: test
+    num_bytes: 99881
+    num_examples: 250
+  download_size: 223978
+  dataset_size: 223978
+- config_name: wic
+  features:
+  - name: word
+    dtype: string
+  - name: sentence1
+    dtype: string
+  - name: sentence2
+    dtype: string
+  - name: idx
+    dtype: int32
+  - name: label
+    dtype: bool
+  - name: start1
+    dtype: int32
+  - name: start2
+    dtype: int32
+  - name: end1
+    dtype: int32
+  - name: end2
+    dtype: int32
+  - name: version
+    dtype: float32
+  splits:
+  - name: train
+    num_bytes: 1211413
+    num_examples: 5428
+  - name: validation
+    num_bytes: 146302
+    num_examples: 638
+  - name: test
+    num_bytes: 299101
+    num_examples: 1400
+  download_size: 1656816
+  dataset_size: 1656816
+- config_name: record
+  features:
+  - name: source
+    dtype: string
+  - name: passage
+    dtype:
+      __type__: dict
+      schema:
         text: str
-        target:
-          __type__: dict
+        entities:
+          __type__: list
+          schema:
+            start: int
+            end: int
+  - name: qas
+    dtype:
+      __type__: list
+      schema:
+        query: str
+        answers:
+          __type__: list
           schema:
-            span2_text: str
-            span2_index: int
-            span1_index: int
-            span1_text: str
+            start: int
+            end: int
+            text: str
         idx: int
-        label: bool
----
+  - name: idx
+    dtype: int32
+  splits:
+  - name: train
+    num_bytes: 135976939
+    num_examples: 65709
+  - name: validation
+    num_bytes: 15236903
+    num_examples: 7481
+  - name: test
+    num_bytes: 13637660
+    num_examples: 7484
+  download_size: 164851502
+  dataset_size: 164851502
+- config_name: copa
+  features:
+  - name: premise
+    dtype: string
+  - name: choice1
+    dtype: string
+  - name: choice2
+    dtype: string
+  - name: question
+    dtype: string
+  - name: label
+    dtype: int32
+  - name: idx
+    dtype: int32
+  splits:
+  - name: train
+    num_bytes: 72323
+    num_examples: 400
+  - name: validation
+    num_bytes: 18169
+    num_examples: 100
+  - name: test
+    num_bytes: 82735
+    num_examples: 500
+  download_size: 173227
+  dataset_size: 173227
+- config_name: rte
+  features:
+  - name: premise
+    dtype: string
+  - name: hypothesis
+    dtype: string
+  - name: label
+    dtype: string
+  - name: idx
+    dtype: int32
+  splits:
+  - name: train
+    num_bytes: 977550
+    num_examples: 2490
+  - name: validation
+    num_bytes: 104895
+    num_examples: 277
+  - name: test
+    num_bytes: 1056063
+    num_examples: 3000
+  download_size: 2138508
+  dataset_size: 2138508
+- config_name: wsc
+  features:
+  - name: text
+    dtype: string
+  - name: target
+    dtype:
+      __type__: dict
+      schema:
+        span2_index: int
+        span1_index: int
+        span1_text: str
+        span2_text: str
+  - name: idx
+    dtype: int32
+  - name: label
+    dtype: bool
+  splits:
+  - name: train
+    num_bytes: 140984
+    num_examples: 554
+  - name: validation
+    num_bytes: 31167
+    num_examples: 104
+  - name: test
+    num_bytes: 42751
+    num_examples: 146
+  download_size: 214902
+  dataset_size: 214902
 ---
 # SuperGLUE Benchmark Datasets