Datasets:
Tasks:
Text Classification
Modalities:
Text
Formats:
parquet
Sub-tasks:
acceptability-classification
Languages:
English
Size:
10K - 100K
ArXiv:
License:
Commit
•
81398c8
1
Parent(s):
c9bb307
Add complex_NP_island data files
Browse files- README.md +7 -3
- complex_NP_island/train-00000-of-00001.parquet +3 -0
- dataset_infos.json +6 -24
README.md
CHANGED
@@ -212,10 +212,10 @@ dataset_info:
|
|
212 |
dtype: int32
|
213 |
splits:
|
214 |
- name: train
|
215 |
-
num_bytes:
|
216 |
num_examples: 1000
|
217 |
-
download_size:
|
218 |
-
dataset_size:
|
219 |
- config_name: coordinate_structure_constraint_complex_left_branch
|
220 |
features:
|
221 |
- name: sentence_good
|
@@ -1921,6 +1921,10 @@ configs:
|
|
1921 |
data_files:
|
1922 |
- split: train
|
1923 |
path: causative/train-*
|
|
|
|
|
|
|
|
|
1924 |
---
|
1925 |
|
1926 |
# Dataset Card for "blimp"
|
|
|
212 |
dtype: int32
|
213 |
splits:
|
214 |
- name: train
|
215 |
+
num_bytes: 198972
|
216 |
num_examples: 1000
|
217 |
+
download_size: 78211
|
218 |
+
dataset_size: 198972
|
219 |
- config_name: coordinate_structure_constraint_complex_left_branch
|
220 |
features:
|
221 |
- name: sentence_good
|
|
|
1921 |
data_files:
|
1922 |
- split: train
|
1923 |
path: causative/train-*
|
1924 |
+
- config_name: complex_NP_island
|
1925 |
+
data_files:
|
1926 |
+
- split: train
|
1927 |
+
path: complex_NP_island/train-*
|
1928 |
---
|
1929 |
|
1930 |
# Dataset Card for "blimp"
|
complex_NP_island/train-00000-of-00001.parquet
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:4a01cfd34626f006bcf0ae90ca3f340348b20d1f6c0530450f35b7218e1b8db7
|
3 |
+
size 78211
|
dataset_infos.json
CHANGED
@@ -415,62 +415,50 @@
|
|
415 |
"features": {
|
416 |
"sentence_good": {
|
417 |
"dtype": "string",
|
418 |
-
"id": null,
|
419 |
"_type": "Value"
|
420 |
},
|
421 |
"sentence_bad": {
|
422 |
"dtype": "string",
|
423 |
-
"id": null,
|
424 |
"_type": "Value"
|
425 |
},
|
426 |
"field": {
|
427 |
"dtype": "string",
|
428 |
-
"id": null,
|
429 |
"_type": "Value"
|
430 |
},
|
431 |
"linguistics_term": {
|
432 |
"dtype": "string",
|
433 |
-
"id": null,
|
434 |
"_type": "Value"
|
435 |
},
|
436 |
"UID": {
|
437 |
"dtype": "string",
|
438 |
-
"id": null,
|
439 |
"_type": "Value"
|
440 |
},
|
441 |
"simple_LM_method": {
|
442 |
"dtype": "bool",
|
443 |
-
"id": null,
|
444 |
"_type": "Value"
|
445 |
},
|
446 |
"one_prefix_method": {
|
447 |
"dtype": "bool",
|
448 |
-
"id": null,
|
449 |
"_type": "Value"
|
450 |
},
|
451 |
"two_prefix_method": {
|
452 |
"dtype": "bool",
|
453 |
-
"id": null,
|
454 |
"_type": "Value"
|
455 |
},
|
456 |
"lexically_identical": {
|
457 |
"dtype": "bool",
|
458 |
-
"id": null,
|
459 |
"_type": "Value"
|
460 |
},
|
461 |
"pair_id": {
|
462 |
"dtype": "int32",
|
463 |
-
"id": null,
|
464 |
"_type": "Value"
|
465 |
}
|
466 |
},
|
467 |
-
"supervised_keys": null,
|
468 |
"builder_name": "blimp",
|
|
|
469 |
"config_name": "complex_NP_island",
|
470 |
"version": {
|
471 |
"version_str": "0.1.0",
|
472 |
-
"description": null,
|
473 |
-
"datasets_version_to_prepare": null,
|
474 |
"major": 0,
|
475 |
"minor": 1,
|
476 |
"patch": 0
|
@@ -478,20 +466,14 @@
|
|
478 |
"splits": {
|
479 |
"train": {
|
480 |
"name": "train",
|
481 |
-
"num_bytes":
|
482 |
"num_examples": 1000,
|
483 |
-
"dataset_name":
|
484 |
-
}
|
485 |
-
},
|
486 |
-
"download_checksums": {
|
487 |
-
"https://raw.githubusercontent.com/alexwarstadt/blimp/master/data/complex_NP_island.jsonl": {
|
488 |
-
"num_bytes": 392362,
|
489 |
-
"checksum": "1c92472731dd9a21eb222d64bee756042a3763c3797b5283d5b84ab8e3cbb74c"
|
490 |
}
|
491 |
},
|
492 |
-
"download_size":
|
493 |
-
"dataset_size":
|
494 |
-
"size_in_bytes":
|
495 |
},
|
496 |
"coordinate_structure_constraint_complex_left_branch": {
|
497 |
"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",
|
|
|
415 |
"features": {
|
416 |
"sentence_good": {
|
417 |
"dtype": "string",
|
|
|
418 |
"_type": "Value"
|
419 |
},
|
420 |
"sentence_bad": {
|
421 |
"dtype": "string",
|
|
|
422 |
"_type": "Value"
|
423 |
},
|
424 |
"field": {
|
425 |
"dtype": "string",
|
|
|
426 |
"_type": "Value"
|
427 |
},
|
428 |
"linguistics_term": {
|
429 |
"dtype": "string",
|
|
|
430 |
"_type": "Value"
|
431 |
},
|
432 |
"UID": {
|
433 |
"dtype": "string",
|
|
|
434 |
"_type": "Value"
|
435 |
},
|
436 |
"simple_LM_method": {
|
437 |
"dtype": "bool",
|
|
|
438 |
"_type": "Value"
|
439 |
},
|
440 |
"one_prefix_method": {
|
441 |
"dtype": "bool",
|
|
|
442 |
"_type": "Value"
|
443 |
},
|
444 |
"two_prefix_method": {
|
445 |
"dtype": "bool",
|
|
|
446 |
"_type": "Value"
|
447 |
},
|
448 |
"lexically_identical": {
|
449 |
"dtype": "bool",
|
|
|
450 |
"_type": "Value"
|
451 |
},
|
452 |
"pair_id": {
|
453 |
"dtype": "int32",
|
|
|
454 |
"_type": "Value"
|
455 |
}
|
456 |
},
|
|
|
457 |
"builder_name": "blimp",
|
458 |
+
"dataset_name": "blimp",
|
459 |
"config_name": "complex_NP_island",
|
460 |
"version": {
|
461 |
"version_str": "0.1.0",
|
|
|
|
|
462 |
"major": 0,
|
463 |
"minor": 1,
|
464 |
"patch": 0
|
|
|
466 |
"splits": {
|
467 |
"train": {
|
468 |
"name": "train",
|
469 |
+
"num_bytes": 198972,
|
470 |
"num_examples": 1000,
|
471 |
+
"dataset_name": null
|
|
|
|
|
|
|
|
|
|
|
|
|
472 |
}
|
473 |
},
|
474 |
+
"download_size": 78211,
|
475 |
+
"dataset_size": 198972,
|
476 |
+
"size_in_bytes": 277183
|
477 |
},
|
478 |
"coordinate_structure_constraint_complex_left_branch": {
|
479 |
"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",
|