Adding go emotions dataset (#865)

prepare/cards/go_emotions.py

@@ -0,0 +1,56 @@
+from datasets import load_dataset_builder
+from unitxt import add_to_catalog
+from unitxt.blocks import (
+    AddFields,
+    LoadHF,
+    TaskCard,
+)
+from unitxt.operators import MapInstanceValues
+from unitxt.test_utils.card import test_card
+
+dataset_name = "go_emotions"
+subset = "simplified"
+
+ds_builder = load_dataset_builder(dataset_name, subset)
+classes = ds_builder.info.features["labels"].feature.names
+
+mappers = {str(i): cls for i, cls in enumerate(classes)}
+
+card = TaskCard(
+    loader=LoadHF(path=dataset_name, name=subset),
+    preprocess_steps=[
+        MapInstanceValues(mappers={"labels": mappers}, process_every_value=True),
+        AddFields(
+            fields={
+                "classes": classes,
+                "text_type": "text",
+                "type_of_classes": "emotions",
+            }
+        ),
+    ],
+    task="tasks.classification.multi_label",
+    templates="templates.classification.multi_label.all",
+    __tags__={
+        "annotations_creators": "crowdsourced",
+        "arxiv": "2005.00547",
+        "croissant": True,
+        "language": "en",
+        "language_creators": "found",
+        "license": "apache-2.0",
+        "multilinguality": "monolingual",
+        "region": "us",
+        "size_categories": "10K<n<100K",
+        "source_datasets": "original",
+        "task_categories": "text-classification",
+        "task_ids": ["multi-class-classification", "multi-label-classification"],
+    },
+    __description__=(
+        "Dataset Card for Go Emotions\n"
+        "Dataset Summary\n"
+        "The GoEmotions dataset contains 58k carefully curated Reddit comments labeled for 27 emotion categories or Neutral.\n"
+        "This card contains the simplified version of the dataset with predefined train/val/test splits.\n"
+        "See full description on the dataset page: https://huggingface.co/datasets/go_emotions."
+    ),
+)
+test_card(card, debug=False)
+add_to_catalog(card, f"cards.{dataset_name}.{subset}", overwrite=True)

src/unitxt/catalog/cards/go_emotions/simplified.json

@@ -0,0 +1,103 @@
+{
+    "type": "task_card",
+    "loader": {
+        "type": "load_hf",
+        "path": "go_emotions",
+        "name": "simplified"
+    },
+    "preprocess_steps": [
+        {
+            "type": "map_instance_values",
+            "mappers": {
+                "labels": {
+                    "0": "admiration",
+                    "1": "amusement",
+                    "2": "anger",
+                    "3": "annoyance",
+                    "4": "approval",
+                    "5": "caring",
+                    "6": "confusion",
+                    "7": "curiosity",
+                    "8": "desire",
+                    "9": "disappointment",
+                    "10": "disapproval",
+                    "11": "disgust",
+                    "12": "embarrassment",
+                    "13": "excitement",
+                    "14": "fear",
+                    "15": "gratitude",
+                    "16": "grief",
+                    "17": "joy",
+                    "18": "love",
+                    "19": "nervousness",
+                    "20": "optimism",
+                    "21": "pride",
+                    "22": "realization",
+                    "23": "relief",
+                    "24": "remorse",
+                    "25": "sadness",
+                    "26": "surprise",
+                    "27": "neutral"
+                }
+            },
+            "process_every_value": true
+        },
+        {
+            "type": "add_fields",
+            "fields": {
+                "classes": [
+                    "admiration",
+                    "amusement",
+                    "anger",
+                    "annoyance",
+                    "approval",
+                    "caring",
+                    "confusion",
+                    "curiosity",
+                    "desire",
+                    "disappointment",
+                    "disapproval",
+                    "disgust",
+                    "embarrassment",
+                    "excitement",
+                    "fear",
+                    "gratitude",
+                    "grief",
+                    "joy",
+                    "love",
+                    "nervousness",
+                    "optimism",
+                    "pride",
+                    "realization",
+                    "relief",
+                    "remorse",
+                    "sadness",
+                    "surprise",
+                    "neutral"
+                ],
+                "text_type": "text",
+                "type_of_classes": "emotions"
+            }
+        }
+    ],
+    "task": "tasks.classification.multi_label",
+    "templates": "templates.classification.multi_label.all",
+    "__tags__": {
+        "annotations_creators": "crowdsourced",
+        "arxiv": "2005.00547",
+        "croissant": true,
+        "language": "en",
+        "language_creators": "found",
+        "license": "apache-2.0",
+        "multilinguality": "monolingual",
+        "region": "us",
+        "size_categories": "10K<n<100K",
+        "source_datasets": "original",
+        "task_categories": "text-classification",
+        "task_ids": [
+            "multi-class-classification",
+            "multi-label-classification"
+        ]
+    },
+    "__description__": "Dataset Card for Go Emotions\nDataset Summary\nThe GoEmotions dataset contains 58k carefully curated Reddit comments labeled for 27 emotion categories or Neutral.\nThis card contains the simplified version of the dataset with predefined train/val/test splits.\nSee full description on the dataset page: https://huggingface.co/datasets/go_emotions."
+}

src/unitxt/operators.py

@@ -138,7 +138,7 @@ class MapInstanceValues(InstanceOperator):
         replaces '1' with 'hi' and '2' with 'bye' in field 'a' in all instances of all streams:
         instance {"a":"1", "b": 2} becomes {"a":"hi", "b": 2}.
 
-        MapInstanceValues(mappers={"a": {"1": "hi", "2": "bye"}}, process_every_element=True)
+        MapInstanceValues(mappers={"a": {"1": "hi", "2": "bye"}}, process_every_value=True)
         Assuming field 'a' is a list of values, potentially including "1"-s and "2"-s, this replaces
         each such "1" with "hi" and "2" -- with "bye" in all instances of all streams:
         instance {"a": ["1", "2"], "b": 2} becomes {"a": ["hi", "bye"], "b": 2}.