Datasets:
Tasks:
Text Classification
Modalities:
Text
Formats:
parquet
Sub-tasks:
acceptability-classification
Languages:
English
Size:
10K - 100K
ArXiv:
License:
Commit
·
055bdd3
1
Parent(s):
b665065
Add npi_present_1 data files
Browse files- README.md +7 -3
- dataset_infos.json +6 -24
- npi_present_1/train-00000-of-00001.parquet +3 -0
README.md
CHANGED
|
@@ -1052,10 +1052,10 @@ dataset_info:
|
|
| 1052 |
dtype: int32
|
| 1053 |
splits:
|
| 1054 |
- name: train
|
| 1055 |
-
num_bytes:
|
| 1056 |
num_examples: 1000
|
| 1057 |
-
download_size:
|
| 1058 |
-
dataset_size:
|
| 1059 |
- config_name: npi_present_2
|
| 1060 |
features:
|
| 1061 |
- name: sentence_good
|
|
@@ -2041,6 +2041,10 @@ configs:
|
|
| 2041 |
data_files:
|
| 2042 |
- split: train
|
| 2043 |
path: matrix_question_npi_licensor_present/train-*
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2044 |
---
|
| 2045 |
|
| 2046 |
# Dataset Card for "blimp"
|
|
|
|
| 1052 |
dtype: int32
|
| 1053 |
splits:
|
| 1054 |
- name: train
|
| 1055 |
+
num_bytes: 138465
|
| 1056 |
num_examples: 1000
|
| 1057 |
+
download_size: 51981
|
| 1058 |
+
dataset_size: 138465
|
| 1059 |
- config_name: npi_present_2
|
| 1060 |
features:
|
| 1061 |
- name: sentence_good
|
|
|
|
| 2041 |
data_files:
|
| 2042 |
- split: train
|
| 2043 |
path: matrix_question_npi_licensor_present/train-*
|
| 2044 |
+
- config_name: npi_present_1
|
| 2045 |
+
data_files:
|
| 2046 |
+
- split: train
|
| 2047 |
+
path: npi_present_1/train-*
|
| 2048 |
---
|
| 2049 |
|
| 2050 |
# Dataset Card for "blimp"
|
dataset_infos.json
CHANGED
|
@@ -2455,62 +2455,50 @@
|
|
| 2455 |
"features": {
|
| 2456 |
"sentence_good": {
|
| 2457 |
"dtype": "string",
|
| 2458 |
-
"id": null,
|
| 2459 |
"_type": "Value"
|
| 2460 |
},
|
| 2461 |
"sentence_bad": {
|
| 2462 |
"dtype": "string",
|
| 2463 |
-
"id": null,
|
| 2464 |
"_type": "Value"
|
| 2465 |
},
|
| 2466 |
"field": {
|
| 2467 |
"dtype": "string",
|
| 2468 |
-
"id": null,
|
| 2469 |
"_type": "Value"
|
| 2470 |
},
|
| 2471 |
"linguistics_term": {
|
| 2472 |
"dtype": "string",
|
| 2473 |
-
"id": null,
|
| 2474 |
"_type": "Value"
|
| 2475 |
},
|
| 2476 |
"UID": {
|
| 2477 |
"dtype": "string",
|
| 2478 |
-
"id": null,
|
| 2479 |
"_type": "Value"
|
| 2480 |
},
|
| 2481 |
"simple_LM_method": {
|
| 2482 |
"dtype": "bool",
|
| 2483 |
-
"id": null,
|
| 2484 |
"_type": "Value"
|
| 2485 |
},
|
| 2486 |
"one_prefix_method": {
|
| 2487 |
"dtype": "bool",
|
| 2488 |
-
"id": null,
|
| 2489 |
"_type": "Value"
|
| 2490 |
},
|
| 2491 |
"two_prefix_method": {
|
| 2492 |
"dtype": "bool",
|
| 2493 |
-
"id": null,
|
| 2494 |
"_type": "Value"
|
| 2495 |
},
|
| 2496 |
"lexically_identical": {
|
| 2497 |
"dtype": "bool",
|
| 2498 |
-
"id": null,
|
| 2499 |
"_type": "Value"
|
| 2500 |
},
|
| 2501 |
"pair_id": {
|
| 2502 |
"dtype": "int32",
|
| 2503 |
-
"id": null,
|
| 2504 |
"_type": "Value"
|
| 2505 |
}
|
| 2506 |
},
|
| 2507 |
-
"supervised_keys": null,
|
| 2508 |
"builder_name": "blimp",
|
|
|
|
| 2509 |
"config_name": "npi_present_1",
|
| 2510 |
"version": {
|
| 2511 |
"version_str": "0.1.0",
|
| 2512 |
-
"description": null,
|
| 2513 |
-
"datasets_version_to_prepare": null,
|
| 2514 |
"major": 0,
|
| 2515 |
"minor": 1,
|
| 2516 |
"patch": 0
|
|
@@ -2518,20 +2506,14 @@
|
|
| 2518 |
"splits": {
|
| 2519 |
"train": {
|
| 2520 |
"name": "train",
|
| 2521 |
-
"num_bytes":
|
| 2522 |
"num_examples": 1000,
|
| 2523 |
-
"dataset_name":
|
| 2524 |
-
}
|
| 2525 |
-
},
|
| 2526 |
-
"download_checksums": {
|
| 2527 |
-
"https://raw.githubusercontent.com/alexwarstadt/blimp/master/data/npi_present_1.jsonl": {
|
| 2528 |
-
"num_bytes": 438013,
|
| 2529 |
-
"checksum": "9aa85dc5c2b0d85ea51f21db50d6d327ab5b45d1669e386d54b1360d3ac954af"
|
| 2530 |
}
|
| 2531 |
},
|
| 2532 |
-
"download_size":
|
| 2533 |
-
"dataset_size":
|
| 2534 |
-
"size_in_bytes":
|
| 2535 |
},
|
| 2536 |
"npi_present_2": {
|
| 2537 |
"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",
|
|
|
|
| 2455 |
"features": {
|
| 2456 |
"sentence_good": {
|
| 2457 |
"dtype": "string",
|
|
|
|
| 2458 |
"_type": "Value"
|
| 2459 |
},
|
| 2460 |
"sentence_bad": {
|
| 2461 |
"dtype": "string",
|
|
|
|
| 2462 |
"_type": "Value"
|
| 2463 |
},
|
| 2464 |
"field": {
|
| 2465 |
"dtype": "string",
|
|
|
|
| 2466 |
"_type": "Value"
|
| 2467 |
},
|
| 2468 |
"linguistics_term": {
|
| 2469 |
"dtype": "string",
|
|
|
|
| 2470 |
"_type": "Value"
|
| 2471 |
},
|
| 2472 |
"UID": {
|
| 2473 |
"dtype": "string",
|
|
|
|
| 2474 |
"_type": "Value"
|
| 2475 |
},
|
| 2476 |
"simple_LM_method": {
|
| 2477 |
"dtype": "bool",
|
|
|
|
| 2478 |
"_type": "Value"
|
| 2479 |
},
|
| 2480 |
"one_prefix_method": {
|
| 2481 |
"dtype": "bool",
|
|
|
|
| 2482 |
"_type": "Value"
|
| 2483 |
},
|
| 2484 |
"two_prefix_method": {
|
| 2485 |
"dtype": "bool",
|
|
|
|
| 2486 |
"_type": "Value"
|
| 2487 |
},
|
| 2488 |
"lexically_identical": {
|
| 2489 |
"dtype": "bool",
|
|
|
|
| 2490 |
"_type": "Value"
|
| 2491 |
},
|
| 2492 |
"pair_id": {
|
| 2493 |
"dtype": "int32",
|
|
|
|
| 2494 |
"_type": "Value"
|
| 2495 |
}
|
| 2496 |
},
|
|
|
|
| 2497 |
"builder_name": "blimp",
|
| 2498 |
+
"dataset_name": "blimp",
|
| 2499 |
"config_name": "npi_present_1",
|
| 2500 |
"version": {
|
| 2501 |
"version_str": "0.1.0",
|
|
|
|
|
|
|
| 2502 |
"major": 0,
|
| 2503 |
"minor": 1,
|
| 2504 |
"patch": 0
|
|
|
|
| 2506 |
"splits": {
|
| 2507 |
"train": {
|
| 2508 |
"name": "train",
|
| 2509 |
+
"num_bytes": 138465,
|
| 2510 |
"num_examples": 1000,
|
| 2511 |
+
"dataset_name": null
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2512 |
}
|
| 2513 |
},
|
| 2514 |
+
"download_size": 51981,
|
| 2515 |
+
"dataset_size": 138465,
|
| 2516 |
+
"size_in_bytes": 190446
|
| 2517 |
},
|
| 2518 |
"npi_present_2": {
|
| 2519 |
"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",
|
npi_present_1/train-00000-of-00001.parquet
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cca3e48e8a08528c06eaec33e9a53aa2875048642f6b38f8c03355c008a55c4f
|
| 3 |
+
size 51981
|