Datasets:
Tasks:
Text Classification
Modalities:
Text
Formats:
parquet
Sub-tasks:
acceptability-classification
Languages:
English
Size:
10K - 100K
ArXiv:
License:
Commit
•
514ec2b
1
Parent(s):
22b5a88
Add wh_island data files
Browse files- README.md +7 -3
- dataset_infos.json +6 -24
- wh_island/train-00000-of-00001.parquet +3 -0
README.md
CHANGED
@@ -1696,10 +1696,10 @@ dataset_info:
|
|
1696 |
dtype: int32
|
1697 |
splits:
|
1698 |
- name: train
|
1699 |
-
num_bytes:
|
1700 |
num_examples: 1000
|
1701 |
-
download_size:
|
1702 |
-
dataset_size:
|
1703 |
- config_name: wh_questions_object_gap
|
1704 |
features:
|
1705 |
- name: sentence_good
|
@@ -2133,6 +2133,10 @@ configs:
|
|
2133 |
data_files:
|
2134 |
- split: train
|
2135 |
path: transitive/train-*
|
|
|
|
|
|
|
|
|
2136 |
---
|
2137 |
|
2138 |
# Dataset Card for "blimp"
|
|
|
1696 |
dtype: int32
|
1697 |
splits:
|
1698 |
- name: train
|
1699 |
+
num_bytes: 142340
|
1700 |
num_examples: 1000
|
1701 |
+
download_size: 52808
|
1702 |
+
dataset_size: 142340
|
1703 |
- config_name: wh_questions_object_gap
|
1704 |
features:
|
1705 |
- name: sentence_good
|
|
|
2133 |
data_files:
|
2134 |
- split: train
|
2135 |
path: transitive/train-*
|
2136 |
+
- config_name: wh_island
|
2137 |
+
data_files:
|
2138 |
+
- split: train
|
2139 |
+
path: wh_island/train-*
|
2140 |
---
|
2141 |
|
2142 |
# Dataset Card for "blimp"
|
dataset_infos.json
CHANGED
@@ -4019,62 +4019,50 @@
|
|
4019 |
"features": {
|
4020 |
"sentence_good": {
|
4021 |
"dtype": "string",
|
4022 |
-
"id": null,
|
4023 |
"_type": "Value"
|
4024 |
},
|
4025 |
"sentence_bad": {
|
4026 |
"dtype": "string",
|
4027 |
-
"id": null,
|
4028 |
"_type": "Value"
|
4029 |
},
|
4030 |
"field": {
|
4031 |
"dtype": "string",
|
4032 |
-
"id": null,
|
4033 |
"_type": "Value"
|
4034 |
},
|
4035 |
"linguistics_term": {
|
4036 |
"dtype": "string",
|
4037 |
-
"id": null,
|
4038 |
"_type": "Value"
|
4039 |
},
|
4040 |
"UID": {
|
4041 |
"dtype": "string",
|
4042 |
-
"id": null,
|
4043 |
"_type": "Value"
|
4044 |
},
|
4045 |
"simple_LM_method": {
|
4046 |
"dtype": "bool",
|
4047 |
-
"id": null,
|
4048 |
"_type": "Value"
|
4049 |
},
|
4050 |
"one_prefix_method": {
|
4051 |
"dtype": "bool",
|
4052 |
-
"id": null,
|
4053 |
"_type": "Value"
|
4054 |
},
|
4055 |
"two_prefix_method": {
|
4056 |
"dtype": "bool",
|
4057 |
-
"id": null,
|
4058 |
"_type": "Value"
|
4059 |
},
|
4060 |
"lexically_identical": {
|
4061 |
"dtype": "bool",
|
4062 |
-
"id": null,
|
4063 |
"_type": "Value"
|
4064 |
},
|
4065 |
"pair_id": {
|
4066 |
"dtype": "int32",
|
4067 |
-
"id": null,
|
4068 |
"_type": "Value"
|
4069 |
}
|
4070 |
},
|
4071 |
-
"supervised_keys": null,
|
4072 |
"builder_name": "blimp",
|
|
|
4073 |
"config_name": "wh_island",
|
4074 |
"version": {
|
4075 |
"version_str": "0.1.0",
|
4076 |
-
"description": null,
|
4077 |
-
"datasets_version_to_prepare": null,
|
4078 |
"major": 0,
|
4079 |
"minor": 1,
|
4080 |
"patch": 0
|
@@ -4082,20 +4070,14 @@
|
|
4082 |
"splits": {
|
4083 |
"train": {
|
4084 |
"name": "train",
|
4085 |
-
"num_bytes":
|
4086 |
"num_examples": 1000,
|
4087 |
-
"dataset_name":
|
4088 |
-
}
|
4089 |
-
},
|
4090 |
-
"download_checksums": {
|
4091 |
-
"https://raw.githubusercontent.com/alexwarstadt/blimp/master/data/wh_island.jsonl": {
|
4092 |
-
"num_bytes": 448630,
|
4093 |
-
"checksum": "77824ca8c0008fb7a204e785ca2dcee1b2803f666bb545bd95cb8a51e04f1a4c"
|
4094 |
}
|
4095 |
},
|
4096 |
-
"download_size":
|
4097 |
-
"dataset_size":
|
4098 |
-
"size_in_bytes":
|
4099 |
},
|
4100 |
"wh_questions_object_gap": {
|
4101 |
"description": "\nBLiMP is a challenge set for evaluating what language models (LMs) know about\nmajor grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each\ncontaining 1000 minimal pairs isolating specific contrasts in syntax,\nmorphology, or semantics. The data is automatically generated according to\nexpert-crafted grammars.\n",
|
|
|
4019 |
"features": {
|
4020 |
"sentence_good": {
|
4021 |
"dtype": "string",
|
|
|
4022 |
"_type": "Value"
|
4023 |
},
|
4024 |
"sentence_bad": {
|
4025 |
"dtype": "string",
|
|
|
4026 |
"_type": "Value"
|
4027 |
},
|
4028 |
"field": {
|
4029 |
"dtype": "string",
|
|
|
4030 |
"_type": "Value"
|
4031 |
},
|
4032 |
"linguistics_term": {
|
4033 |
"dtype": "string",
|
|
|
4034 |
"_type": "Value"
|
4035 |
},
|
4036 |
"UID": {
|
4037 |
"dtype": "string",
|
|
|
4038 |
"_type": "Value"
|
4039 |
},
|
4040 |
"simple_LM_method": {
|
4041 |
"dtype": "bool",
|
|
|
4042 |
"_type": "Value"
|
4043 |
},
|
4044 |
"one_prefix_method": {
|
4045 |
"dtype": "bool",
|
|
|
4046 |
"_type": "Value"
|
4047 |
},
|
4048 |
"two_prefix_method": {
|
4049 |
"dtype": "bool",
|
|
|
4050 |
"_type": "Value"
|
4051 |
},
|
4052 |
"lexically_identical": {
|
4053 |
"dtype": "bool",
|
|
|
4054 |
"_type": "Value"
|
4055 |
},
|
4056 |
"pair_id": {
|
4057 |
"dtype": "int32",
|
|
|
4058 |
"_type": "Value"
|
4059 |
}
|
4060 |
},
|
|
|
4061 |
"builder_name": "blimp",
|
4062 |
+
"dataset_name": "blimp",
|
4063 |
"config_name": "wh_island",
|
4064 |
"version": {
|
4065 |
"version_str": "0.1.0",
|
|
|
|
|
4066 |
"major": 0,
|
4067 |
"minor": 1,
|
4068 |
"patch": 0
|
|
|
4070 |
"splits": {
|
4071 |
"train": {
|
4072 |
"name": "train",
|
4073 |
+
"num_bytes": 142340,
|
4074 |
"num_examples": 1000,
|
4075 |
+
"dataset_name": null
|
|
|
|
|
|
|
|
|
|
|
|
|
4076 |
}
|
4077 |
},
|
4078 |
+
"download_size": 52808,
|
4079 |
+
"dataset_size": 142340,
|
4080 |
+
"size_in_bytes": 195148
|
4081 |
},
|
4082 |
"wh_questions_object_gap": {
|
4083 |
"description": "\nBLiMP is a challenge set for evaluating what language models (LMs) know about\nmajor grammatical phenomena in English. BLiMP consists of 67 sub-datasets, each\ncontaining 1000 minimal pairs isolating specific contrasts in syntax,\nmorphology, or semantics. The data is automatically generated according to\nexpert-crafted grammars.\n",
|
wh_island/train-00000-of-00001.parquet
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:63e455edde12fd76bca6e9287cb80920833b705ac656bc5ad105831b3327f219
|
3 |
+
size 52808
|