Datasets:
Tasks:
Text Classification
Modalities:
Text
Formats:
parquet
Sub-tasks:
acceptability-classification
Languages:
English
Size:
10K - 100K
ArXiv:
License:
Commit
•
1def05b
1
Parent(s):
45df30d
Add drop_argument data files
Browse files- README.md +7 -3
- dataset_infos.json +6 -24
- drop_argument/train-00000-of-00001.parquet +3 -0
README.md
CHANGED
@@ -576,10 +576,10 @@ dataset_info:
|
|
576 |
dtype: int32
|
577 |
splits:
|
578 |
- name: train
|
579 |
-
num_bytes:
|
580 |
num_examples: 1000
|
581 |
-
download_size:
|
582 |
-
dataset_size:
|
583 |
- config_name: ellipsis_n_bar_1
|
584 |
features:
|
585 |
- name: sentence_good
|
@@ -1973,6 +1973,10 @@ configs:
|
|
1973 |
data_files:
|
1974 |
- split: train
|
1975 |
path: distractor_agreement_relative_clause/train-*
|
|
|
|
|
|
|
|
|
1976 |
---
|
1977 |
|
1978 |
# Dataset Card for "blimp"
|
|
|
576 |
dtype: int32
|
577 |
splits:
|
578 |
- name: train
|
579 |
+
num_bytes: 109806
|
580 |
num_examples: 1000
|
581 |
+
download_size: 39961
|
582 |
+
dataset_size: 109806
|
583 |
- config_name: ellipsis_n_bar_1
|
584 |
features:
|
585 |
- name: sentence_good
|
|
|
1973 |
data_files:
|
1974 |
- split: train
|
1975 |
path: distractor_agreement_relative_clause/train-*
|
1976 |
+
- config_name: drop_argument
|
1977 |
+
data_files:
|
1978 |
+
- split: train
|
1979 |
+
path: drop_argument/train-*
|
1980 |
---
|
1981 |
|
1982 |
# Dataset Card for "blimp"
|
dataset_infos.json
CHANGED
@@ -1299,62 +1299,50 @@
|
|
1299 |
"features": {
|
1300 |
"sentence_good": {
|
1301 |
"dtype": "string",
|
1302 |
-
"id": null,
|
1303 |
"_type": "Value"
|
1304 |
},
|
1305 |
"sentence_bad": {
|
1306 |
"dtype": "string",
|
1307 |
-
"id": null,
|
1308 |
"_type": "Value"
|
1309 |
},
|
1310 |
"field": {
|
1311 |
"dtype": "string",
|
1312 |
-
"id": null,
|
1313 |
"_type": "Value"
|
1314 |
},
|
1315 |
"linguistics_term": {
|
1316 |
"dtype": "string",
|
1317 |
-
"id": null,
|
1318 |
"_type": "Value"
|
1319 |
},
|
1320 |
"UID": {
|
1321 |
"dtype": "string",
|
1322 |
-
"id": null,
|
1323 |
"_type": "Value"
|
1324 |
},
|
1325 |
"simple_LM_method": {
|
1326 |
"dtype": "bool",
|
1327 |
-
"id": null,
|
1328 |
"_type": "Value"
|
1329 |
},
|
1330 |
"one_prefix_method": {
|
1331 |
"dtype": "bool",
|
1332 |
-
"id": null,
|
1333 |
"_type": "Value"
|
1334 |
},
|
1335 |
"two_prefix_method": {
|
1336 |
"dtype": "bool",
|
1337 |
-
"id": null,
|
1338 |
"_type": "Value"
|
1339 |
},
|
1340 |
"lexically_identical": {
|
1341 |
"dtype": "bool",
|
1342 |
-
"id": null,
|
1343 |
"_type": "Value"
|
1344 |
},
|
1345 |
"pair_id": {
|
1346 |
"dtype": "int32",
|
1347 |
-
"id": null,
|
1348 |
"_type": "Value"
|
1349 |
}
|
1350 |
},
|
1351 |
-
"supervised_keys": null,
|
1352 |
"builder_name": "blimp",
|
|
|
1353 |
"config_name": "drop_argument",
|
1354 |
"version": {
|
1355 |
"version_str": "0.1.0",
|
1356 |
-
"description": null,
|
1357 |
-
"datasets_version_to_prepare": null,
|
1358 |
"major": 0,
|
1359 |
"minor": 1,
|
1360 |
"patch": 0
|
@@ -1362,20 +1350,14 @@
|
|
1362 |
"splits": {
|
1363 |
"train": {
|
1364 |
"name": "train",
|
1365 |
-
"num_bytes":
|
1366 |
"num_examples": 1000,
|
1367 |
-
"dataset_name":
|
1368 |
-
}
|
1369 |
-
},
|
1370 |
-
"download_checksums": {
|
1371 |
-
"https://raw.githubusercontent.com/alexwarstadt/blimp/master/data/drop_argument.jsonl": {
|
1372 |
-
"num_bytes": 304196,
|
1373 |
-
"checksum": "e262339a262a57a8f56dc41941ee1dc32c62eb97054b27d33cd8c356a6e62064"
|
1374 |
}
|
1375 |
},
|
1376 |
-
"download_size":
|
1377 |
-
"dataset_size":
|
1378 |
-
"size_in_bytes":
|
1379 |
},
|
1380 |
"ellipsis_n_bar_1": {
|
1381 |
"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",
|
|
|
1299 |
"features": {
|
1300 |
"sentence_good": {
|
1301 |
"dtype": "string",
|
|
|
1302 |
"_type": "Value"
|
1303 |
},
|
1304 |
"sentence_bad": {
|
1305 |
"dtype": "string",
|
|
|
1306 |
"_type": "Value"
|
1307 |
},
|
1308 |
"field": {
|
1309 |
"dtype": "string",
|
|
|
1310 |
"_type": "Value"
|
1311 |
},
|
1312 |
"linguistics_term": {
|
1313 |
"dtype": "string",
|
|
|
1314 |
"_type": "Value"
|
1315 |
},
|
1316 |
"UID": {
|
1317 |
"dtype": "string",
|
|
|
1318 |
"_type": "Value"
|
1319 |
},
|
1320 |
"simple_LM_method": {
|
1321 |
"dtype": "bool",
|
|
|
1322 |
"_type": "Value"
|
1323 |
},
|
1324 |
"one_prefix_method": {
|
1325 |
"dtype": "bool",
|
|
|
1326 |
"_type": "Value"
|
1327 |
},
|
1328 |
"two_prefix_method": {
|
1329 |
"dtype": "bool",
|
|
|
1330 |
"_type": "Value"
|
1331 |
},
|
1332 |
"lexically_identical": {
|
1333 |
"dtype": "bool",
|
|
|
1334 |
"_type": "Value"
|
1335 |
},
|
1336 |
"pair_id": {
|
1337 |
"dtype": "int32",
|
|
|
1338 |
"_type": "Value"
|
1339 |
}
|
1340 |
},
|
|
|
1341 |
"builder_name": "blimp",
|
1342 |
+
"dataset_name": "blimp",
|
1343 |
"config_name": "drop_argument",
|
1344 |
"version": {
|
1345 |
"version_str": "0.1.0",
|
|
|
|
|
1346 |
"major": 0,
|
1347 |
"minor": 1,
|
1348 |
"patch": 0
|
|
|
1350 |
"splits": {
|
1351 |
"train": {
|
1352 |
"name": "train",
|
1353 |
+
"num_bytes": 109806,
|
1354 |
"num_examples": 1000,
|
1355 |
+
"dataset_name": null
|
|
|
|
|
|
|
|
|
|
|
|
|
1356 |
}
|
1357 |
},
|
1358 |
+
"download_size": 39961,
|
1359 |
+
"dataset_size": 109806,
|
1360 |
+
"size_in_bytes": 149767
|
1361 |
},
|
1362 |
"ellipsis_n_bar_1": {
|
1363 |
"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",
|
drop_argument/train-00000-of-00001.parquet
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:4acf04febbe94ca4d8e0d12d2e02c808fb028c3501f4e66e62b1a1e2bfc8fdf3
|
3 |
+
size 39961
|