Datasets:
Tasks:
Text Classification
Modalities:
Text
Formats:
parquet
Sub-tasks:
acceptability-classification
Languages:
English
Size:
10K - 100K
ArXiv:
License:
Commit
·
044d155
1
Parent(s):
33491b0
Add wh_vs_that_no_gap_long_distance data files
Browse files
README.md
CHANGED
@@ -1836,10 +1836,10 @@ dataset_info:
|
|
1836 |
dtype: int32
|
1837 |
splits:
|
1838 |
- name: train
|
1839 |
-
num_bytes:
|
1840 |
num_examples: 1000
|
1841 |
-
download_size:
|
1842 |
-
dataset_size:
|
1843 |
- config_name: wh_vs_that_with_gap
|
1844 |
features:
|
1845 |
- name: sentence_good
|
@@ -2153,6 +2153,10 @@ configs:
|
|
2153 |
data_files:
|
2154 |
- split: train
|
2155 |
path: wh_vs_that_no_gap/train-*
|
|
|
|
|
|
|
|
|
2156 |
---
|
2157 |
|
2158 |
# Dataset Card for "blimp"
|
|
|
1836 |
dtype: int32
|
1837 |
splits:
|
1838 |
- name: train
|
1839 |
+
num_bytes: 247039
|
1840 |
num_examples: 1000
|
1841 |
+
download_size: 95504
|
1842 |
+
dataset_size: 247039
|
1843 |
- config_name: wh_vs_that_with_gap
|
1844 |
features:
|
1845 |
- name: sentence_good
|
|
|
2153 |
data_files:
|
2154 |
- split: train
|
2155 |
path: wh_vs_that_no_gap/train-*
|
2156 |
+
- config_name: wh_vs_that_no_gap_long_distance
|
2157 |
+
data_files:
|
2158 |
+
- split: train
|
2159 |
+
path: wh_vs_that_no_gap_long_distance/train-*
|
2160 |
---
|
2161 |
|
2162 |
# Dataset Card for "blimp"
|
dataset_infos.json
CHANGED
@@ -4359,62 +4359,50 @@
|
|
4359 |
"features": {
|
4360 |
"sentence_good": {
|
4361 |
"dtype": "string",
|
4362 |
-
"id": null,
|
4363 |
"_type": "Value"
|
4364 |
},
|
4365 |
"sentence_bad": {
|
4366 |
"dtype": "string",
|
4367 |
-
"id": null,
|
4368 |
"_type": "Value"
|
4369 |
},
|
4370 |
"field": {
|
4371 |
"dtype": "string",
|
4372 |
-
"id": null,
|
4373 |
"_type": "Value"
|
4374 |
},
|
4375 |
"linguistics_term": {
|
4376 |
"dtype": "string",
|
4377 |
-
"id": null,
|
4378 |
"_type": "Value"
|
4379 |
},
|
4380 |
"UID": {
|
4381 |
"dtype": "string",
|
4382 |
-
"id": null,
|
4383 |
"_type": "Value"
|
4384 |
},
|
4385 |
"simple_LM_method": {
|
4386 |
"dtype": "bool",
|
4387 |
-
"id": null,
|
4388 |
"_type": "Value"
|
4389 |
},
|
4390 |
"one_prefix_method": {
|
4391 |
"dtype": "bool",
|
4392 |
-
"id": null,
|
4393 |
"_type": "Value"
|
4394 |
},
|
4395 |
"two_prefix_method": {
|
4396 |
"dtype": "bool",
|
4397 |
-
"id": null,
|
4398 |
"_type": "Value"
|
4399 |
},
|
4400 |
"lexically_identical": {
|
4401 |
"dtype": "bool",
|
4402 |
-
"id": null,
|
4403 |
"_type": "Value"
|
4404 |
},
|
4405 |
"pair_id": {
|
4406 |
"dtype": "int32",
|
4407 |
-
"id": null,
|
4408 |
"_type": "Value"
|
4409 |
}
|
4410 |
},
|
4411 |
-
"supervised_keys": null,
|
4412 |
"builder_name": "blimp",
|
|
|
4413 |
"config_name": "wh_vs_that_no_gap_long_distance",
|
4414 |
"version": {
|
4415 |
"version_str": "0.1.0",
|
4416 |
-
"description": null,
|
4417 |
-
"datasets_version_to_prepare": null,
|
4418 |
"major": 0,
|
4419 |
"minor": 1,
|
4420 |
"patch": 0
|
@@ -4422,20 +4410,14 @@
|
|
4422 |
"splits": {
|
4423 |
"train": {
|
4424 |
"name": "train",
|
4425 |
-
"num_bytes":
|
4426 |
"num_examples": 1000,
|
4427 |
-
"dataset_name":
|
4428 |
-
}
|
4429 |
-
},
|
4430 |
-
"download_checksums": {
|
4431 |
-
"https://raw.githubusercontent.com/alexwarstadt/blimp/master/data/wh_vs_that_no_gap_long_distance.jsonl": {
|
4432 |
-
"num_bytes": 441429,
|
4433 |
-
"checksum": "968faf59602a67f9cfd7c2e91a314a3f7342fd5daf41c21cadd25288fc85fdcf"
|
4434 |
}
|
4435 |
},
|
4436 |
-
"download_size":
|
4437 |
-
"dataset_size":
|
4438 |
-
"size_in_bytes":
|
4439 |
},
|
4440 |
"wh_vs_that_with_gap": {
|
4441 |
"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",
|
|
|
4359 |
"features": {
|
4360 |
"sentence_good": {
|
4361 |
"dtype": "string",
|
|
|
4362 |
"_type": "Value"
|
4363 |
},
|
4364 |
"sentence_bad": {
|
4365 |
"dtype": "string",
|
|
|
4366 |
"_type": "Value"
|
4367 |
},
|
4368 |
"field": {
|
4369 |
"dtype": "string",
|
|
|
4370 |
"_type": "Value"
|
4371 |
},
|
4372 |
"linguistics_term": {
|
4373 |
"dtype": "string",
|
|
|
4374 |
"_type": "Value"
|
4375 |
},
|
4376 |
"UID": {
|
4377 |
"dtype": "string",
|
|
|
4378 |
"_type": "Value"
|
4379 |
},
|
4380 |
"simple_LM_method": {
|
4381 |
"dtype": "bool",
|
|
|
4382 |
"_type": "Value"
|
4383 |
},
|
4384 |
"one_prefix_method": {
|
4385 |
"dtype": "bool",
|
|
|
4386 |
"_type": "Value"
|
4387 |
},
|
4388 |
"two_prefix_method": {
|
4389 |
"dtype": "bool",
|
|
|
4390 |
"_type": "Value"
|
4391 |
},
|
4392 |
"lexically_identical": {
|
4393 |
"dtype": "bool",
|
|
|
4394 |
"_type": "Value"
|
4395 |
},
|
4396 |
"pair_id": {
|
4397 |
"dtype": "int32",
|
|
|
4398 |
"_type": "Value"
|
4399 |
}
|
4400 |
},
|
|
|
4401 |
"builder_name": "blimp",
|
4402 |
+
"dataset_name": "blimp",
|
4403 |
"config_name": "wh_vs_that_no_gap_long_distance",
|
4404 |
"version": {
|
4405 |
"version_str": "0.1.0",
|
|
|
|
|
4406 |
"major": 0,
|
4407 |
"minor": 1,
|
4408 |
"patch": 0
|
|
|
4410 |
"splits": {
|
4411 |
"train": {
|
4412 |
"name": "train",
|
4413 |
+
"num_bytes": 247039,
|
4414 |
"num_examples": 1000,
|
4415 |
+
"dataset_name": null
|
|
|
|
|
|
|
|
|
|
|
|
|
4416 |
}
|
4417 |
},
|
4418 |
+
"download_size": 95504,
|
4419 |
+
"dataset_size": 247039,
|
4420 |
+
"size_in_bytes": 342543
|
4421 |
},
|
4422 |
"wh_vs_that_with_gap": {
|
4423 |
"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_vs_that_no_gap_long_distance/train-00000-of-00001.parquet
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e1aa6fd1c7406b3bea5206e954b75f1dec7dd17c1019d526abe6005e715b38b3
|
3 |
+
size 95504
|