Datasets:
Tasks:
Text Classification
Modalities:
Text
Formats:
parquet
Sub-tasks:
acceptability-classification
Languages:
English
Size:
10K - 100K
ArXiv:
License:
Commit
·
094ce12
1
Parent(s):
0595718
Add principle_A_domain_3 data files
Browse files- README.md +7 -3
- dataset_infos.json +6 -24
- principle_A_domain_3/train-00000-of-00001.parquet +3 -0
README.md
CHANGED
@@ -1360,10 +1360,10 @@ dataset_info:
|
|
1360 |
dtype: int32
|
1361 |
splits:
|
1362 |
- name: train
|
1363 |
-
num_bytes:
|
1364 |
num_examples: 1000
|
1365 |
-
download_size:
|
1366 |
-
dataset_size:
|
1367 |
- config_name: principle_A_reconstruction
|
1368 |
features:
|
1369 |
- name: sentence_good
|
@@ -2085,6 +2085,10 @@ configs:
|
|
2085 |
data_files:
|
2086 |
- split: train
|
2087 |
path: principle_A_domain_2/train-*
|
|
|
|
|
|
|
|
|
2088 |
---
|
2089 |
|
2090 |
# Dataset Card for "blimp"
|
|
|
1360 |
dtype: int32
|
1361 |
splits:
|
1362 |
- name: train
|
1363 |
+
num_bytes: 158998
|
1364 |
num_examples: 1000
|
1365 |
+
download_size: 52859
|
1366 |
+
dataset_size: 158998
|
1367 |
- config_name: principle_A_reconstruction
|
1368 |
features:
|
1369 |
- name: sentence_good
|
|
|
2085 |
data_files:
|
2086 |
- split: train
|
2087 |
path: principle_A_domain_2/train-*
|
2088 |
+
- config_name: principle_A_domain_3
|
2089 |
+
data_files:
|
2090 |
+
- split: train
|
2091 |
+
path: principle_A_domain_3/train-*
|
2092 |
---
|
2093 |
|
2094 |
# Dataset Card for "blimp"
|
dataset_infos.json
CHANGED
@@ -3203,62 +3203,50 @@
|
|
3203 |
"features": {
|
3204 |
"sentence_good": {
|
3205 |
"dtype": "string",
|
3206 |
-
"id": null,
|
3207 |
"_type": "Value"
|
3208 |
},
|
3209 |
"sentence_bad": {
|
3210 |
"dtype": "string",
|
3211 |
-
"id": null,
|
3212 |
"_type": "Value"
|
3213 |
},
|
3214 |
"field": {
|
3215 |
"dtype": "string",
|
3216 |
-
"id": null,
|
3217 |
"_type": "Value"
|
3218 |
},
|
3219 |
"linguistics_term": {
|
3220 |
"dtype": "string",
|
3221 |
-
"id": null,
|
3222 |
"_type": "Value"
|
3223 |
},
|
3224 |
"UID": {
|
3225 |
"dtype": "string",
|
3226 |
-
"id": null,
|
3227 |
"_type": "Value"
|
3228 |
},
|
3229 |
"simple_LM_method": {
|
3230 |
"dtype": "bool",
|
3231 |
-
"id": null,
|
3232 |
"_type": "Value"
|
3233 |
},
|
3234 |
"one_prefix_method": {
|
3235 |
"dtype": "bool",
|
3236 |
-
"id": null,
|
3237 |
"_type": "Value"
|
3238 |
},
|
3239 |
"two_prefix_method": {
|
3240 |
"dtype": "bool",
|
3241 |
-
"id": null,
|
3242 |
"_type": "Value"
|
3243 |
},
|
3244 |
"lexically_identical": {
|
3245 |
"dtype": "bool",
|
3246 |
-
"id": null,
|
3247 |
"_type": "Value"
|
3248 |
},
|
3249 |
"pair_id": {
|
3250 |
"dtype": "int32",
|
3251 |
-
"id": null,
|
3252 |
"_type": "Value"
|
3253 |
}
|
3254 |
},
|
3255 |
-
"supervised_keys": null,
|
3256 |
"builder_name": "blimp",
|
|
|
3257 |
"config_name": "principle_A_domain_3",
|
3258 |
"version": {
|
3259 |
"version_str": "0.1.0",
|
3260 |
-
"description": null,
|
3261 |
-
"datasets_version_to_prepare": null,
|
3262 |
"major": 0,
|
3263 |
"minor": 1,
|
3264 |
"patch": 0
|
@@ -3266,20 +3254,14 @@
|
|
3266 |
"splits": {
|
3267 |
"train": {
|
3268 |
"name": "train",
|
3269 |
-
"num_bytes":
|
3270 |
"num_examples": 1000,
|
3271 |
-
"dataset_name":
|
3272 |
-
}
|
3273 |
-
},
|
3274 |
-
"download_checksums": {
|
3275 |
-
"https://raw.githubusercontent.com/alexwarstadt/blimp/master/data/principle_A_domain_3.jsonl": {
|
3276 |
-
"num_bytes": 513886,
|
3277 |
-
"checksum": "a0d66b9d8541ab1df08f629161c6c4e8bca3aa2693eaa342edd2097e6b579bdd"
|
3278 |
}
|
3279 |
},
|
3280 |
-
"download_size":
|
3281 |
-
"dataset_size":
|
3282 |
-
"size_in_bytes":
|
3283 |
},
|
3284 |
"principle_A_reconstruction": {
|
3285 |
"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",
|
|
|
3203 |
"features": {
|
3204 |
"sentence_good": {
|
3205 |
"dtype": "string",
|
|
|
3206 |
"_type": "Value"
|
3207 |
},
|
3208 |
"sentence_bad": {
|
3209 |
"dtype": "string",
|
|
|
3210 |
"_type": "Value"
|
3211 |
},
|
3212 |
"field": {
|
3213 |
"dtype": "string",
|
|
|
3214 |
"_type": "Value"
|
3215 |
},
|
3216 |
"linguistics_term": {
|
3217 |
"dtype": "string",
|
|
|
3218 |
"_type": "Value"
|
3219 |
},
|
3220 |
"UID": {
|
3221 |
"dtype": "string",
|
|
|
3222 |
"_type": "Value"
|
3223 |
},
|
3224 |
"simple_LM_method": {
|
3225 |
"dtype": "bool",
|
|
|
3226 |
"_type": "Value"
|
3227 |
},
|
3228 |
"one_prefix_method": {
|
3229 |
"dtype": "bool",
|
|
|
3230 |
"_type": "Value"
|
3231 |
},
|
3232 |
"two_prefix_method": {
|
3233 |
"dtype": "bool",
|
|
|
3234 |
"_type": "Value"
|
3235 |
},
|
3236 |
"lexically_identical": {
|
3237 |
"dtype": "bool",
|
|
|
3238 |
"_type": "Value"
|
3239 |
},
|
3240 |
"pair_id": {
|
3241 |
"dtype": "int32",
|
|
|
3242 |
"_type": "Value"
|
3243 |
}
|
3244 |
},
|
|
|
3245 |
"builder_name": "blimp",
|
3246 |
+
"dataset_name": "blimp",
|
3247 |
"config_name": "principle_A_domain_3",
|
3248 |
"version": {
|
3249 |
"version_str": "0.1.0",
|
|
|
|
|
3250 |
"major": 0,
|
3251 |
"minor": 1,
|
3252 |
"patch": 0
|
|
|
3254 |
"splits": {
|
3255 |
"train": {
|
3256 |
"name": "train",
|
3257 |
+
"num_bytes": 158998,
|
3258 |
"num_examples": 1000,
|
3259 |
+
"dataset_name": null
|
|
|
|
|
|
|
|
|
|
|
|
|
3260 |
}
|
3261 |
},
|
3262 |
+
"download_size": 52859,
|
3263 |
+
"dataset_size": 158998,
|
3264 |
+
"size_in_bytes": 211857
|
3265 |
},
|
3266 |
"principle_A_reconstruction": {
|
3267 |
"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",
|
principle_A_domain_3/train-00000-of-00001.parquet
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:25227d39f6168b88f5b3ee0b35d0f6dd8105545016d76764d11c27f6031844d5
|
3 |
+
size 52859
|