Upload batch 362 (20 files, last=huggingface_dataset/Dataset_Card/autoevaluate_autoeval-eval-mathemakitten__winobias_antistereotype_test_v5-mathemak-0d489a-2053267106.md)
Browse files- huggingface_dataset/Dataset_Card/GrainsPolito_BBBicycles.md +53 -0
- huggingface_dataset/Dataset_Card/Phantom-Artist_phantom-diffusion-dataset.md +15 -0
- huggingface_dataset/Dataset_Card/autoevaluate_autoeval-eval-glue-cola-b911f0-1508954844.md +33 -0
- huggingface_dataset/Dataset_Card/autoevaluate_autoeval-eval-mathemakitten__winobias_antistereotype_test_v5-mathemak-0d489a-2053267106.md +34 -0
- huggingface_dataset/Dataset_Card/autoevaluate_autoeval-staging-eval-project-205dcc30-381f-492a-a8e8-fcfbe94b826c-110107.md +33 -0
- huggingface_dataset/Dataset_Card/autoevaluate_autoeval-staging-eval-project-2bc32ae8-3118-4561-b552-cc3a89a73cd5-1816.md +33 -0
- huggingface_dataset/Dataset_Card/autoevaluate_autoeval-staging-eval-project-adversarial_qa-e34332b7-12205628.md +35 -0
- huggingface_dataset/Dataset_Card/autoevaluate_autoeval-staging-eval-project-xsum-ad8ac8a3-10195347.md +31 -0
- huggingface_dataset/Dataset_Card/bigbio_bionlp_st_2011_id.md +59 -0
- huggingface_dataset/Dataset_Card/bio-datasets_re-medical-annotations.md +36 -0
- huggingface_dataset/Dataset_Card/huggingartists_egor-kreed.md +204 -0
- huggingface_dataset/Dataset_Card/irds_mr-tydi_ar_test.md +55 -0
- huggingface_dataset/Dataset_Card/johnowhitaker_vqgan16k_reconstruction.md +28 -0
- huggingface_dataset/Dataset_Card/kashif_App_Flow.md +11 -0
- huggingface_dataset/Dataset_Card/nimaster_autonlp-data-devign_raw_test.md +55 -0
- huggingface_dataset/Dataset_Card/numer_sense.md +210 -0
- huggingface_dataset/Dataset_Card/re_dial.md +450 -0
- huggingface_dataset/Dataset_Card/ronig_protein_binding_sequences.md +10 -0
- huggingface_dataset/Dataset_Card/stanfordnlp_SHP.md +264 -0
- huggingface_dataset/Dataset_Card/turkish_shrinked_ner.md +260 -0
huggingface_dataset/Dataset_Card/GrainsPolito_BBBicycles.md
ADDED
|
@@ -0,0 +1,53 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: cc-by-nc-4.0
|
| 3 |
+
---
|
| 4 |
+
# Dataset Card for BBBicycles
|
| 5 |
+
## Dataset Summary
|
| 6 |
+
Bent & Broken Bicycles (BBBicycles) dataset is a benchmark set for the novel task of **damaged object re-identification**, which aims to identify the same object in multiple images even in the presence of breaks, deformations, and missing parts. You can find an interactive preview [here](https://huggingface.co/spaces/GrainsPolito/BBBicyclesPreview).
|
| 7 |
+
## Dataset Structure
|
| 8 |
+
The final dataset contains:
|
| 9 |
+
|
| 10 |
+
- Total of 39,200 image
|
| 11 |
+
- 2,800 unique IDs
|
| 12 |
+
- 20 models
|
| 13 |
+
- 140 IDs for each model
|
| 14 |
+
|
| 15 |
+
<table border-collapse="collapse">
|
| 16 |
+
<tr>
|
| 17 |
+
<td><b style="font-size:25px">Information for each ID:</b></td>
|
| 18 |
+
<td><b style="font-size:25px">Information for each render:</b></td>
|
| 19 |
+
</tr>
|
| 20 |
+
<tr>
|
| 21 |
+
<td>
|
| 22 |
+
<ul>
|
| 23 |
+
<li>Model</li>
|
| 24 |
+
<li>Type</li>
|
| 25 |
+
<li>Texture type</li>
|
| 26 |
+
<li>Stickers</li>
|
| 27 |
+
</ul>
|
| 28 |
+
</td>
|
| 29 |
+
<td>
|
| 30 |
+
<ul>
|
| 31 |
+
<li>Background</li>
|
| 32 |
+
<li>Viewing Side</li>
|
| 33 |
+
<li>Focal Length</li>
|
| 34 |
+
<li>Presence of dirt</li>
|
| 35 |
+
</ul>
|
| 36 |
+
</td>
|
| 37 |
+
</tr>
|
| 38 |
+
</table>
|
| 39 |
+
|
| 40 |
+
### Citation Information
|
| 41 |
+
```
|
| 42 |
+
@inproceedings{bbb_2022,
|
| 43 |
+
title={Bent & Broken Bicycles: Leveraging synthetic data for damaged object re-identification},
|
| 44 |
+
author={Luca Piano, Filippo Gabriele Pratticò, Alessandro Sebastian Russo, Lorenzo Lanari, Lia Morra, Fabrizio Lamberti},
|
| 45 |
+
booktitle={2022 IEEE Winter Conference on Applications of Computer Vision (WACV)},
|
| 46 |
+
year={2022},
|
| 47 |
+
organization={IEEE}
|
| 48 |
+
}
|
| 49 |
+
|
| 50 |
+
```
|
| 51 |
+
|
| 52 |
+
### Credits
|
| 53 |
+
The authors gratefully acknowledge the financial support of Reale Mutua Assicurazioni.
|
huggingface_dataset/Dataset_Card/Phantom-Artist_phantom-diffusion-dataset.md
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: cc0-1.0
|
| 3 |
+
language:
|
| 4 |
+
- en
|
| 5 |
+
- ja
|
| 6 |
+
size_categories:
|
| 7 |
+
- n<1K
|
| 8 |
+
---
|
| 9 |
+
Images trained for my [phantom diffusion](https://huggingface.co/Phantom-Artist/phantom-diffusion) series.
|
| 10 |
+
|
| 11 |
+
Since they are all AI generated images that are public domain under the US law, I claim it is legal to redistribute them as public domain.
|
| 12 |
+
|
| 13 |
+
However, they might have copyright in your/their original country.
|
| 14 |
+
|
| 15 |
+
Still, many countries including Japan allow us to use them for training an AI under their copyrights law, and because all the artists here are from Japan, I assume it should be allowed to reuse it for training globally.
|
huggingface_dataset/Dataset_Card/autoevaluate_autoeval-eval-glue-cola-b911f0-1508954844.md
ADDED
|
@@ -0,0 +1,33 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
type: predictions
|
| 3 |
+
tags:
|
| 4 |
+
- autotrain
|
| 5 |
+
- evaluation
|
| 6 |
+
datasets:
|
| 7 |
+
- glue
|
| 8 |
+
eval_info:
|
| 9 |
+
task: multi_class_classification
|
| 10 |
+
model: JeremiahZ/bert-base-uncased-cola
|
| 11 |
+
metrics: ['matthews_correlation']
|
| 12 |
+
dataset_name: glue
|
| 13 |
+
dataset_config: cola
|
| 14 |
+
dataset_split: validation
|
| 15 |
+
col_mapping:
|
| 16 |
+
text: sentence
|
| 17 |
+
target: label
|
| 18 |
+
---
|
| 19 |
+
# Dataset Card for AutoTrain Evaluator
|
| 20 |
+
|
| 21 |
+
This repository contains model predictions generated by [AutoTrain](https://huggingface.co/autotrain) for the following task and dataset:
|
| 22 |
+
|
| 23 |
+
* Task: Multi-class Text Classification
|
| 24 |
+
* Model: JeremiahZ/bert-base-uncased-cola
|
| 25 |
+
* Dataset: glue
|
| 26 |
+
* Config: cola
|
| 27 |
+
* Split: validation
|
| 28 |
+
|
| 29 |
+
To run new evaluation jobs, visit Hugging Face's [automatic model evaluator](https://huggingface.co/spaces/autoevaluate/model-evaluator).
|
| 30 |
+
|
| 31 |
+
## Contributions
|
| 32 |
+
|
| 33 |
+
Thanks to [@JeremiahZ](https://huggingface.co/JeremiahZ) for evaluating this model.
|
huggingface_dataset/Dataset_Card/autoevaluate_autoeval-eval-mathemakitten__winobias_antistereotype_test_v5-mathemak-0d489a-2053267106.md
ADDED
|
@@ -0,0 +1,34 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
type: predictions
|
| 3 |
+
tags:
|
| 4 |
+
- autotrain
|
| 5 |
+
- evaluation
|
| 6 |
+
datasets:
|
| 7 |
+
- mathemakitten/winobias_antistereotype_test_v5
|
| 8 |
+
eval_info:
|
| 9 |
+
task: text_zero_shot_classification
|
| 10 |
+
model: inverse-scaling/opt-125m_eval
|
| 11 |
+
metrics: []
|
| 12 |
+
dataset_name: mathemakitten/winobias_antistereotype_test_v5
|
| 13 |
+
dataset_config: mathemakitten--winobias_antistereotype_test_v5
|
| 14 |
+
dataset_split: test
|
| 15 |
+
col_mapping:
|
| 16 |
+
text: text
|
| 17 |
+
classes: classes
|
| 18 |
+
target: target
|
| 19 |
+
---
|
| 20 |
+
# Dataset Card for AutoTrain Evaluator
|
| 21 |
+
|
| 22 |
+
This repository contains model predictions generated by [AutoTrain](https://huggingface.co/autotrain) for the following task and dataset:
|
| 23 |
+
|
| 24 |
+
* Task: Zero-Shot Text Classification
|
| 25 |
+
* Model: inverse-scaling/opt-125m_eval
|
| 26 |
+
* Dataset: mathemakitten/winobias_antistereotype_test_v5
|
| 27 |
+
* Config: mathemakitten--winobias_antistereotype_test_v5
|
| 28 |
+
* Split: test
|
| 29 |
+
|
| 30 |
+
To run new evaluation jobs, visit Hugging Face's [automatic model evaluator](https://huggingface.co/spaces/autoevaluate/model-evaluator).
|
| 31 |
+
|
| 32 |
+
## Contributions
|
| 33 |
+
|
| 34 |
+
Thanks to [@mathemakitten](https://huggingface.co/mathemakitten) for evaluating this model.
|
huggingface_dataset/Dataset_Card/autoevaluate_autoeval-staging-eval-project-205dcc30-381f-492a-a8e8-fcfbe94b826c-110107.md
ADDED
|
@@ -0,0 +1,33 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
type: predictions
|
| 3 |
+
tags:
|
| 4 |
+
- autotrain
|
| 5 |
+
- evaluation
|
| 6 |
+
datasets:
|
| 7 |
+
- glue
|
| 8 |
+
eval_info:
|
| 9 |
+
task: binary_classification
|
| 10 |
+
model: autoevaluate/binary-classification
|
| 11 |
+
metrics: ['matthews_correlation']
|
| 12 |
+
dataset_name: glue
|
| 13 |
+
dataset_config: sst2
|
| 14 |
+
dataset_split: validation
|
| 15 |
+
col_mapping:
|
| 16 |
+
text: sentence
|
| 17 |
+
target: label
|
| 18 |
+
---
|
| 19 |
+
# Dataset Card for AutoTrain Evaluator
|
| 20 |
+
|
| 21 |
+
This repository contains model predictions generated by [AutoTrain](https://huggingface.co/autotrain) for the following task and dataset:
|
| 22 |
+
|
| 23 |
+
* Task: Binary Text Classification
|
| 24 |
+
* Model: autoevaluate/binary-classification
|
| 25 |
+
* Dataset: glue
|
| 26 |
+
* Config: sst2
|
| 27 |
+
* Split: validation
|
| 28 |
+
|
| 29 |
+
To run new evaluation jobs, visit Hugging Face's [automatic model evaluator](https://huggingface.co/spaces/autoevaluate/model-evaluator).
|
| 30 |
+
|
| 31 |
+
## Contributions
|
| 32 |
+
|
| 33 |
+
Thanks to [@lewtun](https://huggingface.co/lewtun) for evaluating this model.
|
huggingface_dataset/Dataset_Card/autoevaluate_autoeval-staging-eval-project-2bc32ae8-3118-4561-b552-cc3a89a73cd5-1816.md
ADDED
|
@@ -0,0 +1,33 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
type: predictions
|
| 3 |
+
tags:
|
| 4 |
+
- autotrain
|
| 5 |
+
- evaluation
|
| 6 |
+
datasets:
|
| 7 |
+
- glue
|
| 8 |
+
eval_info:
|
| 9 |
+
task: binary_classification
|
| 10 |
+
model: autoevaluate/binary-classification
|
| 11 |
+
metrics: ['matthews_correlation']
|
| 12 |
+
dataset_name: glue
|
| 13 |
+
dataset_config: sst2
|
| 14 |
+
dataset_split: validation
|
| 15 |
+
col_mapping:
|
| 16 |
+
text: sentence
|
| 17 |
+
target: label
|
| 18 |
+
---
|
| 19 |
+
# Dataset Card for AutoTrain Evaluator
|
| 20 |
+
|
| 21 |
+
This repository contains model predictions generated by [AutoTrain](https://huggingface.co/autotrain) for the following task and dataset:
|
| 22 |
+
|
| 23 |
+
* Task: Binary Text Classification
|
| 24 |
+
* Model: autoevaluate/binary-classification
|
| 25 |
+
* Dataset: glue
|
| 26 |
+
* Config: sst2
|
| 27 |
+
* Split: validation
|
| 28 |
+
|
| 29 |
+
To run new evaluation jobs, visit Hugging Face's [automatic model evaluator](https://huggingface.co/spaces/autoevaluate/model-evaluator).
|
| 30 |
+
|
| 31 |
+
## Contributions
|
| 32 |
+
|
| 33 |
+
Thanks to [@lewtun](https://huggingface.co/lewtun) for evaluating this model.
|
huggingface_dataset/Dataset_Card/autoevaluate_autoeval-staging-eval-project-adversarial_qa-e34332b7-12205628.md
ADDED
|
@@ -0,0 +1,35 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
type: predictions
|
| 3 |
+
tags:
|
| 4 |
+
- autotrain
|
| 5 |
+
- evaluation
|
| 6 |
+
datasets:
|
| 7 |
+
- adversarial_qa
|
| 8 |
+
eval_info:
|
| 9 |
+
task: extractive_question_answering
|
| 10 |
+
model: deepset/tinybert-6l-768d-squad2
|
| 11 |
+
metrics: []
|
| 12 |
+
dataset_name: adversarial_qa
|
| 13 |
+
dataset_config: adversarialQA
|
| 14 |
+
dataset_split: validation
|
| 15 |
+
col_mapping:
|
| 16 |
+
context: context
|
| 17 |
+
question: question
|
| 18 |
+
answers-text: answers.text
|
| 19 |
+
answers-answer_start: answers.answer_start
|
| 20 |
+
---
|
| 21 |
+
# Dataset Card for AutoTrain Evaluator
|
| 22 |
+
|
| 23 |
+
This repository contains model predictions generated by [AutoTrain](https://huggingface.co/autotrain) for the following task and dataset:
|
| 24 |
+
|
| 25 |
+
* Task: Question Answering
|
| 26 |
+
* Model: deepset/tinybert-6l-768d-squad2
|
| 27 |
+
* Dataset: adversarial_qa
|
| 28 |
+
* Config: adversarialQA
|
| 29 |
+
* Split: validation
|
| 30 |
+
|
| 31 |
+
To run new evaluation jobs, visit Hugging Face's [automatic model evaluator](https://huggingface.co/spaces/autoevaluate/model-evaluator).
|
| 32 |
+
|
| 33 |
+
## Contributions
|
| 34 |
+
|
| 35 |
+
Thanks to [@ceyda](https://huggingface.co/ceyda) for evaluating this model.
|
huggingface_dataset/Dataset_Card/autoevaluate_autoeval-staging-eval-project-xsum-ad8ac8a3-10195347.md
ADDED
|
@@ -0,0 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
type: predictions
|
| 3 |
+
tags:
|
| 4 |
+
- autotrain
|
| 5 |
+
- evaluation
|
| 6 |
+
datasets:
|
| 7 |
+
- xsum
|
| 8 |
+
eval_info:
|
| 9 |
+
task: summarization
|
| 10 |
+
model: t5-large
|
| 11 |
+
metrics: []
|
| 12 |
+
dataset_name: xsum
|
| 13 |
+
dataset_config: default
|
| 14 |
+
dataset_split: test
|
| 15 |
+
col_mapping:
|
| 16 |
+
text: document
|
| 17 |
+
target: summary
|
| 18 |
+
---
|
| 19 |
+
# Dataset Card for AutoTrain Evaluator
|
| 20 |
+
|
| 21 |
+
This repository contains model predictions generated by [AutoTrain](https://huggingface.co/autotrain) for the following task and dataset:
|
| 22 |
+
|
| 23 |
+
* Task: Summarization
|
| 24 |
+
* Model: t5-large
|
| 25 |
+
* Dataset: xsum
|
| 26 |
+
|
| 27 |
+
To run new evaluation jobs, visit Hugging Face's [automatic model evaluator](https://huggingface.co/spaces/autoevaluate/model-evaluator).
|
| 28 |
+
|
| 29 |
+
## Contributions
|
| 30 |
+
|
| 31 |
+
Thanks to [@abhijeet](https://huggingface.co/abhijeet) for evaluating this model.
|
huggingface_dataset/Dataset_Card/bigbio_bionlp_st_2011_id.md
ADDED
|
@@ -0,0 +1,59 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
|
| 2 |
+
---
|
| 3 |
+
language:
|
| 4 |
+
- en
|
| 5 |
+
bigbio_language:
|
| 6 |
+
- English
|
| 7 |
+
license: other
|
| 8 |
+
multilinguality: monolingual
|
| 9 |
+
bigbio_license_shortname: GENIA_PROJECT_LICENSE
|
| 10 |
+
pretty_name: BioNLP 2011 ID
|
| 11 |
+
homepage: https://github.com/openbiocorpora/bionlp-st-2011-id
|
| 12 |
+
bigbio_pubmed: True
|
| 13 |
+
bigbio_public: True
|
| 14 |
+
bigbio_tasks:
|
| 15 |
+
- EVENT_EXTRACTION
|
| 16 |
+
- COREFERENCE_RESOLUTION
|
| 17 |
+
- NAMED_ENTITY_RECOGNITION
|
| 18 |
+
---
|
| 19 |
+
|
| 20 |
+
|
| 21 |
+
# Dataset Card for BioNLP 2011 ID
|
| 22 |
+
|
| 23 |
+
## Dataset Description
|
| 24 |
+
|
| 25 |
+
- **Homepage:** https://github.com/openbiocorpora/bionlp-st-2011-id
|
| 26 |
+
- **Pubmed:** True
|
| 27 |
+
- **Public:** True
|
| 28 |
+
- **Tasks:** EE,COREF,NER
|
| 29 |
+
|
| 30 |
+
|
| 31 |
+
The dataset of the Infectious Diseases (ID) task of
|
| 32 |
+
BioNLP Shared Task 2011.
|
| 33 |
+
|
| 34 |
+
|
| 35 |
+
|
| 36 |
+
## Citation Information
|
| 37 |
+
|
| 38 |
+
```
|
| 39 |
+
@inproceedings{pyysalo-etal-2011-overview,
|
| 40 |
+
title = "Overview of the Infectious Diseases ({ID}) task of {B}io{NLP} Shared Task 2011",
|
| 41 |
+
author = "Pyysalo, Sampo and
|
| 42 |
+
Ohta, Tomoko and
|
| 43 |
+
Rak, Rafal and
|
| 44 |
+
Sullivan, Dan and
|
| 45 |
+
Mao, Chunhong and
|
| 46 |
+
Wang, Chunxia and
|
| 47 |
+
Sobral, Bruno and
|
| 48 |
+
Tsujii, Jun{'}ichi and
|
| 49 |
+
Ananiadou, Sophia",
|
| 50 |
+
booktitle = "Proceedings of {B}io{NLP} Shared Task 2011 Workshop",
|
| 51 |
+
month = jun,
|
| 52 |
+
year = "2011",
|
| 53 |
+
address = "Portland, Oregon, USA",
|
| 54 |
+
publisher = "Association for Computational Linguistics",
|
| 55 |
+
url = "https://aclanthology.org/W11-1804",
|
| 56 |
+
pages = "26--35",
|
| 57 |
+
}
|
| 58 |
+
|
| 59 |
+
```
|
huggingface_dataset/Dataset_Card/bio-datasets_re-medical-annotations.md
ADDED
|
@@ -0,0 +1,36 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# Dataset Card for re-medical-annotations
|
| 2 |
+
|
| 3 |
+
## Dataset Description
|
| 4 |
+
|
| 5 |
+
### Dataset Summary
|
| 6 |
+
|
| 7 |
+
HuggingFace Dataset from the Inception Medical Annotations project.
|
| 8 |
+
|
| 9 |
+
This dataset can be used locally with any archive downloaded from Inception that contains relation annotations.
|
| 10 |
+
|
| 11 |
+
Loading this dataset requires `dkpro-cassis>=0.7.2`.
|
| 12 |
+
|
| 13 |
+
**Example**: load the dataset from the "RE Temporality POC"
|
| 14 |
+
|
| 15 |
+
```
|
| 16 |
+
import datasets
|
| 17 |
+
|
| 18 |
+
ds = datasets.load_dataset(
|
| 19 |
+
"bio-datasets/re-medical-annotations",
|
| 20 |
+
data_dir=<Inception Archive path>,
|
| 21 |
+
labels = ["bound"],
|
| 22 |
+
)
|
| 23 |
+
```
|
| 24 |
+
|
| 25 |
+
## Dataset Structure
|
| 26 |
+
|
| 27 |
+
### Data Fields
|
| 28 |
+
|
| 29 |
+
- `text (str)`: text of the sentence
|
| 30 |
+
- `subj_start (int)`: start char of the relation subject mention
|
| 31 |
+
- `subj_end (int)`: end char of the relation subject mention, exclusive
|
| 32 |
+
- `subj_type (str)`: NER label of the relation subject
|
| 33 |
+
- `obj_start (int)`: start char of the relation object mention
|
| 34 |
+
- `obj_end (int)`: end char of the relation object mention, exclusive
|
| 35 |
+
- `obj_type (str)`: NER label of the relation object
|
| 36 |
+
- `relation (str)`: the relation label of this instance
|
huggingface_dataset/Dataset_Card/huggingartists_egor-kreed.md
ADDED
|
@@ -0,0 +1,204 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
language:
|
| 3 |
+
- en
|
| 4 |
+
tags:
|
| 5 |
+
- huggingartists
|
| 6 |
+
- lyrics
|
| 7 |
+
---
|
| 8 |
+
|
| 9 |
+
# Dataset Card for "huggingartists/egor-kreed"
|
| 10 |
+
|
| 11 |
+
## Table of Contents
|
| 12 |
+
- [Dataset Description](#dataset-description)
|
| 13 |
+
- [Dataset Summary](#dataset-summary)
|
| 14 |
+
- [Supported Tasks and Leaderboards](#supported-tasks-and-leaderboards)
|
| 15 |
+
- [Languages](#languages)
|
| 16 |
+
- [How to use](#how-to-use)
|
| 17 |
+
- [Dataset Structure](#dataset-structure)
|
| 18 |
+
- [Data Fields](#data-fields)
|
| 19 |
+
- [Data Splits](#data-splits)
|
| 20 |
+
- [Dataset Creation](#dataset-creation)
|
| 21 |
+
- [Curation Rationale](#curation-rationale)
|
| 22 |
+
- [Source Data](#source-data)
|
| 23 |
+
- [Annotations](#annotations)
|
| 24 |
+
- [Personal and Sensitive Information](#personal-and-sensitive-information)
|
| 25 |
+
- [Considerations for Using the Data](#considerations-for-using-the-data)
|
| 26 |
+
- [Social Impact of Dataset](#social-impact-of-dataset)
|
| 27 |
+
- [Discussion of Biases](#discussion-of-biases)
|
| 28 |
+
- [Other Known Limitations](#other-known-limitations)
|
| 29 |
+
- [Additional Information](#additional-information)
|
| 30 |
+
- [Dataset Curators](#dataset-curators)
|
| 31 |
+
- [Licensing Information](#licensing-information)
|
| 32 |
+
- [Citation Information](#citation-information)
|
| 33 |
+
- [About](#about)
|
| 34 |
+
|
| 35 |
+
## Dataset Description
|
| 36 |
+
|
| 37 |
+
- **Homepage:** [https://github.com/AlekseyKorshuk/huggingartists](https://github.com/AlekseyKorshuk/huggingartists)
|
| 38 |
+
- **Repository:** [https://github.com/AlekseyKorshuk/huggingartists](https://github.com/AlekseyKorshuk/huggingartists)
|
| 39 |
+
- **Paper:** [More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
| 40 |
+
- **Point of Contact:** [More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
| 41 |
+
- **Size of the generated dataset:** 0.321207 MB
|
| 42 |
+
|
| 43 |
+
|
| 44 |
+
<div class="inline-flex flex-col" style="line-height: 1.5;">
|
| 45 |
+
<div class="flex">
|
| 46 |
+
<div style="display:DISPLAY_1; margin-left: auto; margin-right: auto; width: 92px; height:92px; border-radius: 50%; background-size: cover; background-image: url('https://images.genius.com/f52808edb2078f52ddab162623f0c6e3.1000x1000x1.jpg')">
|
| 47 |
+
</div>
|
| 48 |
+
</div>
|
| 49 |
+
<a href="https://huggingface.co/huggingartists/egor-kreed">
|
| 50 |
+
<div style="text-align: center; margin-top: 3px; font-size: 16px; font-weight: 800">🤖 HuggingArtists Model 🤖</div>
|
| 51 |
+
</a>
|
| 52 |
+
<div style="text-align: center; font-size: 16px; font-weight: 800">ЕГОР КРИД (EGOR KREED)</div>
|
| 53 |
+
<a href="https://genius.com/artists/egor-kreed">
|
| 54 |
+
<div style="text-align: center; font-size: 14px;">@egor-kreed</div>
|
| 55 |
+
</a>
|
| 56 |
+
</div>
|
| 57 |
+
|
| 58 |
+
### Dataset Summary
|
| 59 |
+
|
| 60 |
+
The Lyrics dataset parsed from Genius. This dataset is designed to generate lyrics with HuggingArtists.
|
| 61 |
+
Model is available [here](https://huggingface.co/huggingartists/egor-kreed).
|
| 62 |
+
|
| 63 |
+
### Supported Tasks and Leaderboards
|
| 64 |
+
|
| 65 |
+
[More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
| 66 |
+
|
| 67 |
+
### Languages
|
| 68 |
+
|
| 69 |
+
en
|
| 70 |
+
|
| 71 |
+
## How to use
|
| 72 |
+
|
| 73 |
+
How to load this dataset directly with the datasets library:
|
| 74 |
+
|
| 75 |
+
```python
|
| 76 |
+
from datasets import load_dataset
|
| 77 |
+
|
| 78 |
+
dataset = load_dataset("huggingartists/egor-kreed")
|
| 79 |
+
```
|
| 80 |
+
|
| 81 |
+
## Dataset Structure
|
| 82 |
+
|
| 83 |
+
An example of 'train' looks as follows.
|
| 84 |
+
```
|
| 85 |
+
This example was too long and was cropped:
|
| 86 |
+
|
| 87 |
+
{
|
| 88 |
+
"text": "Look, I was gonna go easy on you\nNot to hurt your feelings\nBut I'm only going to get this one chance\nSomething's wrong, I can feel it..."
|
| 89 |
+
}
|
| 90 |
+
```
|
| 91 |
+
|
| 92 |
+
### Data Fields
|
| 93 |
+
|
| 94 |
+
The data fields are the same among all splits.
|
| 95 |
+
|
| 96 |
+
- `text`: a `string` feature.
|
| 97 |
+
|
| 98 |
+
|
| 99 |
+
### Data Splits
|
| 100 |
+
|
| 101 |
+
| train |validation|test|
|
| 102 |
+
|------:|---------:|---:|
|
| 103 |
+
|103| -| -|
|
| 104 |
+
|
| 105 |
+
'Train' can be easily divided into 'train' & 'validation' & 'test' with few lines of code:
|
| 106 |
+
|
| 107 |
+
```python
|
| 108 |
+
from datasets import load_dataset, Dataset, DatasetDict
|
| 109 |
+
import numpy as np
|
| 110 |
+
|
| 111 |
+
datasets = load_dataset("huggingartists/egor-kreed")
|
| 112 |
+
|
| 113 |
+
train_percentage = 0.9
|
| 114 |
+
validation_percentage = 0.07
|
| 115 |
+
test_percentage = 0.03
|
| 116 |
+
|
| 117 |
+
train, validation, test = np.split(datasets['train']['text'], [int(len(datasets['train']['text'])*train_percentage), int(len(datasets['train']['text'])*(train_percentage + validation_percentage))])
|
| 118 |
+
|
| 119 |
+
datasets = DatasetDict(
|
| 120 |
+
{
|
| 121 |
+
'train': Dataset.from_dict({'text': list(train)}),
|
| 122 |
+
'validation': Dataset.from_dict({'text': list(validation)}),
|
| 123 |
+
'test': Dataset.from_dict({'text': list(test)})
|
| 124 |
+
}
|
| 125 |
+
)
|
| 126 |
+
```
|
| 127 |
+
|
| 128 |
+
## Dataset Creation
|
| 129 |
+
|
| 130 |
+
### Curation Rationale
|
| 131 |
+
|
| 132 |
+
[More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
| 133 |
+
|
| 134 |
+
### Source Data
|
| 135 |
+
|
| 136 |
+
#### Initial Data Collection and Normalization
|
| 137 |
+
|
| 138 |
+
[More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
| 139 |
+
|
| 140 |
+
#### Who are the source language producers?
|
| 141 |
+
|
| 142 |
+
[More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
| 143 |
+
|
| 144 |
+
### Annotations
|
| 145 |
+
|
| 146 |
+
#### Annotation process
|
| 147 |
+
|
| 148 |
+
[More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
| 149 |
+
|
| 150 |
+
#### Who are the annotators?
|
| 151 |
+
|
| 152 |
+
[More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
| 153 |
+
|
| 154 |
+
### Personal and Sensitive Information
|
| 155 |
+
|
| 156 |
+
[More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
| 157 |
+
|
| 158 |
+
## Considerations for Using the Data
|
| 159 |
+
|
| 160 |
+
### Social Impact of Dataset
|
| 161 |
+
|
| 162 |
+
[More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
| 163 |
+
|
| 164 |
+
### Discussion of Biases
|
| 165 |
+
|
| 166 |
+
[More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
| 167 |
+
|
| 168 |
+
### Other Known Limitations
|
| 169 |
+
|
| 170 |
+
[More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
| 171 |
+
|
| 172 |
+
## Additional Information
|
| 173 |
+
|
| 174 |
+
### Dataset Curators
|
| 175 |
+
|
| 176 |
+
[More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
| 177 |
+
|
| 178 |
+
### Licensing Information
|
| 179 |
+
|
| 180 |
+
[More Information Needed](https://github.com/huggingface/datasets/blob/master/CONTRIBUTING.md#how-to-contribute-to-the-dataset-cards)
|
| 181 |
+
|
| 182 |
+
### Citation Information
|
| 183 |
+
|
| 184 |
+
```
|
| 185 |
+
@InProceedings{huggingartists,
|
| 186 |
+
author={Aleksey Korshuk}
|
| 187 |
+
year=2021
|
| 188 |
+
}
|
| 189 |
+
```
|
| 190 |
+
|
| 191 |
+
|
| 192 |
+
## About
|
| 193 |
+
|
| 194 |
+
*Built by Aleksey Korshuk*
|
| 195 |
+
|
| 196 |
+
[](https://github.com/AlekseyKorshuk)
|
| 197 |
+
|
| 198 |
+
[](https://twitter.com/intent/follow?screen_name=alekseykorshuk)
|
| 199 |
+
|
| 200 |
+
[](https://t.me/joinchat/_CQ04KjcJ-4yZTky)
|
| 201 |
+
|
| 202 |
+
For more details, visit the project repository.
|
| 203 |
+
|
| 204 |
+
[](https://github.com/AlekseyKorshuk/huggingartists)
|
huggingface_dataset/Dataset_Card/irds_mr-tydi_ar_test.md
ADDED
|
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
pretty_name: '`mr-tydi/ar/test`'
|
| 3 |
+
viewer: false
|
| 4 |
+
source_datasets: ['irds/mr-tydi_ar']
|
| 5 |
+
task_categories:
|
| 6 |
+
- text-retrieval
|
| 7 |
+
---
|
| 8 |
+
|
| 9 |
+
# Dataset Card for `mr-tydi/ar/test`
|
| 10 |
+
|
| 11 |
+
The `mr-tydi/ar/test` dataset, provided by the [ir-datasets](https://ir-datasets.com/) package.
|
| 12 |
+
For more information about the dataset, see the [documentation](https://ir-datasets.com/mr-tydi#mr-tydi/ar/test).
|
| 13 |
+
|
| 14 |
+
# Data
|
| 15 |
+
|
| 16 |
+
This dataset provides:
|
| 17 |
+
- `queries` (i.e., topics); count=1,081
|
| 18 |
+
- `qrels`: (relevance assessments); count=1,257
|
| 19 |
+
|
| 20 |
+
- For `docs`, use [`irds/mr-tydi_ar`](https://huggingface.co/datasets/irds/mr-tydi_ar)
|
| 21 |
+
|
| 22 |
+
## Usage
|
| 23 |
+
|
| 24 |
+
```python
|
| 25 |
+
from datasets import load_dataset
|
| 26 |
+
|
| 27 |
+
queries = load_dataset('irds/mr-tydi_ar_test', 'queries')
|
| 28 |
+
for record in queries:
|
| 29 |
+
record # {'query_id': ..., 'text': ...}
|
| 30 |
+
|
| 31 |
+
qrels = load_dataset('irds/mr-tydi_ar_test', 'qrels')
|
| 32 |
+
for record in qrels:
|
| 33 |
+
record # {'query_id': ..., 'doc_id': ..., 'relevance': ..., 'iteration': ...}
|
| 34 |
+
|
| 35 |
+
```
|
| 36 |
+
|
| 37 |
+
Note that calling `load_dataset` will download the dataset (or provide access instructions when it's not public) and make a copy of the
|
| 38 |
+
data in 🤗 Dataset format.
|
| 39 |
+
|
| 40 |
+
## Citation Information
|
| 41 |
+
|
| 42 |
+
```
|
| 43 |
+
@article{Zhang2021MrTyDi,
|
| 44 |
+
title={{Mr. TyDi}: A Multi-lingual Benchmark for Dense Retrieval},
|
| 45 |
+
author={Xinyu Zhang and Xueguang Ma and Peng Shi and Jimmy Lin},
|
| 46 |
+
year={2021},
|
| 47 |
+
journal={arXiv:2108.08787},
|
| 48 |
+
}
|
| 49 |
+
@article{Clark2020TyDiQa,
|
| 50 |
+
title={{TyDi QA}: A Benchmark for Information-Seeking Question Answering in Typologically Diverse Languages},
|
| 51 |
+
author={Jonathan H. Clark and Eunsol Choi and Michael Collins and Dan Garrette and Tom Kwiatkowski and Vitaly Nikolaev and Jennimaria Palomaki},
|
| 52 |
+
year={2020},
|
| 53 |
+
journal={Transactions of the Association for Computational Linguistics}
|
| 54 |
+
}
|
| 55 |
+
```
|
huggingface_dataset/Dataset_Card/johnowhitaker_vqgan16k_reconstruction.md
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
VQGAN is great, but leaves artifacts that are especially visible around things like faces.
|
| 2 |
+
|
| 3 |
+
It's be great to be able to train a model to fix ('devqganify') these flaws.
|
| 4 |
+
|
| 5 |
+
For this purpose, I've made this dataset, which contains >100k examples, each with
|
| 6 |
+
- A 512px image
|
| 7 |
+
- A smaller 256px version of the same image
|
| 8 |
+
- A reconstructed version, which is made by encoding the 256px image with VQGAN (f16, 16384 imagenet version from https://heibox.uni-heidelberg.de/d/a7530b09fed84f80a887/) and then decoding the result.
|
| 9 |
+
|
| 10 |
+
The idea is to train a model to go from the 256px vqgan output back to something as close to the original image as possible, or even to try and output an up-scaled 512px version for extra points.
|
| 11 |
+
|
| 12 |
+
Let me know what you come up with :)
|
| 13 |
+
|
| 14 |
+
Usage:
|
| 15 |
+
```python
|
| 16 |
+
from datasets import load_dataset
|
| 17 |
+
dataset = load_dataset('johnowhitaker/vqgan1024_reconstruction')
|
| 18 |
+
dataset['train'][0]['image_256'] # Original image
|
| 19 |
+
dataset['train'][0]['reconstruction_256'] # Reconstructed version
|
| 20 |
+
````
|
| 21 |
+
|
| 22 |
+
|
| 23 |
+
|
| 24 |
+
Approximate code used to prepare this data (vqgan model was changed for this version): https://colab.research.google.com/drive/1AXzlRMvAIE6krkpFwFnFr2c5SnOsygf-?usp=sharing (let me know if you hit issues)
|
| 25 |
+
|
| 26 |
+
The VQGAN model used for this version: https://heibox.uni-heidelberg.de/d/a7530b09fed84f80a887/
|
| 27 |
+
|
| 28 |
+
See also: https://huggingface.co/datasets/johnowhitaker/vqgan1024_reconstruction (same idea but vqgan with smaller vocab size of 1024)
|
huggingface_dataset/Dataset_Card/kashif_App_Flow.md
ADDED
|
@@ -0,0 +1,11 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
task_categories:
|
| 4 |
+
- time-series-forecasting
|
| 5 |
+
---
|
| 6 |
+
|
| 7 |
+
# App Flow
|
| 8 |
+
|
| 9 |
+
This dataset consists of hourly maximum traffic flow for 128 systems deployed on 16 logic data centers, resulting in 1083 different time series in total.
|
| 10 |
+
The length of each series is more than 4 months. Each time series is divided into two segments for training and testing with a ratio of 32:1.
|
| 11 |
+
This dataset was collected at Ant Group and does not contain any Personal Identifiable Information and is desensitized and encrypted.
|
huggingface_dataset/Dataset_Card/nimaster_autonlp-data-devign_raw_test.md
ADDED
|
@@ -0,0 +1,55 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
languages:
|
| 3 |
+
- en
|
| 4 |
+
task_categories:
|
| 5 |
+
- text-classification
|
| 6 |
+
|
| 7 |
+
---
|
| 8 |
+
# AutoNLP Dataset for project: devign_raw_test
|
| 9 |
+
|
| 10 |
+
## Dataset Descritpion
|
| 11 |
+
|
| 12 |
+
This dataset has been automatically processed by AutoNLP for project devign_raw_test.
|
| 13 |
+
|
| 14 |
+
### Languages
|
| 15 |
+
|
| 16 |
+
The BCP-47 code for the dataset's language is en.
|
| 17 |
+
|
| 18 |
+
## Dataset Structure
|
| 19 |
+
|
| 20 |
+
### Data Instances
|
| 21 |
+
|
| 22 |
+
A sample from this dataset looks as follows:
|
| 23 |
+
|
| 24 |
+
```json
|
| 25 |
+
[
|
| 26 |
+
{
|
| 27 |
+
"text": "void ff_avg_h264_qpel16_mc32_msa ( uint8_t * dst , const uint8_t * src , ptrdiff_t stride ) { avc_lu[...]",
|
| 28 |
+
"target": 0
|
| 29 |
+
},
|
| 30 |
+
{
|
| 31 |
+
"text": "static void sd_cardchange ( void * opaque , bool load ) { SDState * sd = opaque ; qemu_set_irq ( sd [...]",
|
| 32 |
+
"target": 0
|
| 33 |
+
}
|
| 34 |
+
]
|
| 35 |
+
```
|
| 36 |
+
|
| 37 |
+
### Dataset Fields
|
| 38 |
+
|
| 39 |
+
The dataset has the following fields (also called "features"):
|
| 40 |
+
|
| 41 |
+
```json
|
| 42 |
+
{
|
| 43 |
+
"text": "Value(dtype='string', id=None)",
|
| 44 |
+
"target": "ClassLabel(num_classes=2, names=['0', '1'], id=None)"
|
| 45 |
+
}
|
| 46 |
+
```
|
| 47 |
+
|
| 48 |
+
### Dataset Splits
|
| 49 |
+
|
| 50 |
+
This dataset is split into a train and validation split. The split sizes are as follow:
|
| 51 |
+
|
| 52 |
+
| Split name | Num samples |
|
| 53 |
+
| ------------ | ------------------- |
|
| 54 |
+
| train | 21188 |
|
| 55 |
+
| valid | 5298 |
|
huggingface_dataset/Dataset_Card/numer_sense.md
ADDED
|
@@ -0,0 +1,210 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
annotations_creators:
|
| 3 |
+
- expert-generated
|
| 4 |
+
language_creators:
|
| 5 |
+
- crowdsourced
|
| 6 |
+
language:
|
| 7 |
+
- en
|
| 8 |
+
license:
|
| 9 |
+
- mit
|
| 10 |
+
multilinguality:
|
| 11 |
+
- monolingual
|
| 12 |
+
size_categories:
|
| 13 |
+
- 10K<n<100K
|
| 14 |
+
source_datasets:
|
| 15 |
+
- extended|other
|
| 16 |
+
task_categories:
|
| 17 |
+
- text-generation
|
| 18 |
+
- fill-mask
|
| 19 |
+
task_ids:
|
| 20 |
+
- slot-filling
|
| 21 |
+
paperswithcode_id: numersense
|
| 22 |
+
pretty_name: NumerSense
|
| 23 |
+
dataset_info:
|
| 24 |
+
features:
|
| 25 |
+
- name: sentence
|
| 26 |
+
dtype: string
|
| 27 |
+
- name: target
|
| 28 |
+
dtype: string
|
| 29 |
+
splits:
|
| 30 |
+
- name: train
|
| 31 |
+
num_bytes: 825865
|
| 32 |
+
num_examples: 10444
|
| 33 |
+
- name: test_core
|
| 34 |
+
num_bytes: 62652
|
| 35 |
+
num_examples: 1132
|
| 36 |
+
- name: test_all
|
| 37 |
+
num_bytes: 184180
|
| 38 |
+
num_examples: 3146
|
| 39 |
+
download_size: 985463
|
| 40 |
+
dataset_size: 1072697
|
| 41 |
+
---
|
| 42 |
+
|
| 43 |
+
# Dataset Card for [Dataset Name]
|
| 44 |
+
|
| 45 |
+
## Table of Contents
|
| 46 |
+
- [Dataset Description](#dataset-description)
|
| 47 |
+
- [Dataset Summary](#dataset-summary)
|
| 48 |
+
- [Supported Tasks and Leaderboards](#supported-tasks-and-leaderboards)
|
| 49 |
+
- [Languages](#languages)
|
| 50 |
+
- [Dataset Structure](#dataset-structure)
|
| 51 |
+
- [Data Instances](#data-instances)
|
| 52 |
+
- [Data Fields](#data-fields)
|
| 53 |
+
- [Data Splits](#data-splits)
|
| 54 |
+
- [Dataset Creation](#dataset-creation)
|
| 55 |
+
- [Curation Rationale](#curation-rationale)
|
| 56 |
+
- [Source Data](#source-data)
|
| 57 |
+
- [Annotations](#annotations)
|
| 58 |
+
- [Personal and Sensitive Information](#personal-and-sensitive-information)
|
| 59 |
+
- [Considerations for Using the Data](#considerations-for-using-the-data)
|
| 60 |
+
- [Social Impact of Dataset](#social-impact-of-dataset)
|
| 61 |
+
- [Discussion of Biases](#discussion-of-biases)
|
| 62 |
+
- [Other Known Limitations](#other-known-limitations)
|
| 63 |
+
- [Additional Information](#additional-information)
|
| 64 |
+
- [Dataset Curators](#dataset-curators)
|
| 65 |
+
- [Licensing Information](#licensing-information)
|
| 66 |
+
- [Citation Information](#citation-information)
|
| 67 |
+
- [Contributions](#contributions)
|
| 68 |
+
|
| 69 |
+
## Dataset Description
|
| 70 |
+
|
| 71 |
+
- **Homepage:** https://inklab.usc.edu/NumerSense/
|
| 72 |
+
- **Repository:** https://github.com/INK-USC/NumerSense
|
| 73 |
+
- **Paper:** https://arxiv.org/abs/2005.00683
|
| 74 |
+
- **Leaderboard:** https://inklab.usc.edu/NumerSense/#exp
|
| 75 |
+
- **Point of Contact:** Author emails listed in [paper](https://arxiv.org/abs/2005.00683)
|
| 76 |
+
|
| 77 |
+
### Dataset Summary
|
| 78 |
+
|
| 79 |
+
NumerSense is a new numerical commonsense reasoning probing task, with a diagnostic dataset consisting of 3,145
|
| 80 |
+
masked-word-prediction probes. The general idea is to mask numbers between 0-10 in sentences mined from a commonsense
|
| 81 |
+
corpus and evaluate whether a language model can correctly predict the masked value.
|
| 82 |
+
|
| 83 |
+
### Supported Tasks and Leaderboards
|
| 84 |
+
|
| 85 |
+
The dataset supports the task of slot-filling, specifically as an evaluation of numerical common sense. A leaderboard
|
| 86 |
+
is included on the [dataset webpage](https://inklab.usc.edu/NumerSense/#exp) with included benchmarks for GPT-2,
|
| 87 |
+
RoBERTa, BERT, and human performance. Leaderboards are included for both the core set and the adversarial set
|
| 88 |
+
discussed below.
|
| 89 |
+
|
| 90 |
+
### Languages
|
| 91 |
+
|
| 92 |
+
This dataset is in English.
|
| 93 |
+
|
| 94 |
+
## Dataset Structure
|
| 95 |
+
|
| 96 |
+
### Data Instances
|
| 97 |
+
|
| 98 |
+
Each instance consists of a sentence with a masked numerical value between 0-10 and (in the train set) a target.
|
| 99 |
+
Example from the training set:
|
| 100 |
+
|
| 101 |
+
```
|
| 102 |
+
sentence: Black bears are about <mask> metres tall.
|
| 103 |
+
target: two
|
| 104 |
+
```
|
| 105 |
+
|
| 106 |
+
### Data Fields
|
| 107 |
+
|
| 108 |
+
Each value of the training set consists of:
|
| 109 |
+
- `sentence`: The sentence with a number masked out with the `<mask>` token.
|
| 110 |
+
- `target`: The ground truth target value. Since the test sets do not include the ground truth, the `target` field
|
| 111 |
+
values are empty strings in the `test_core` and `test_all` splits.
|
| 112 |
+
|
| 113 |
+
### Data Splits
|
| 114 |
+
|
| 115 |
+
The dataset includes the following pre-defined data splits:
|
| 116 |
+
|
| 117 |
+
- A train set with >10K labeled examples (i.e. containing a ground truth value)
|
| 118 |
+
- A core test set (`test_core`) with 1,132 examples (no ground truth provided)
|
| 119 |
+
- An expanded test set (`test_all`) encompassing `test_core` with the addition of adversarial examples for a total of
|
| 120 |
+
3,146 examples. See section 2.2 of [the paper] for a discussion of how these examples are constructed.
|
| 121 |
+
|
| 122 |
+
## Dataset Creation
|
| 123 |
+
|
| 124 |
+
### Curation Rationale
|
| 125 |
+
|
| 126 |
+
The purpose of this dataset is "to study whether PTLMs capture numerical commonsense knowledge, i.e., commonsense
|
| 127 |
+
knowledge that provides an understanding of the numeric relation between entities." This work is motivated by the
|
| 128 |
+
prior research exploring whether language models possess _commonsense knowledge_.
|
| 129 |
+
|
| 130 |
+
### Source Data
|
| 131 |
+
|
| 132 |
+
#### Initial Data Collection and Normalization
|
| 133 |
+
|
| 134 |
+
The dataset is an extension of the [Open Mind Common Sense](https://huggingface.co/datasets/open_mind_common_sense)
|
| 135 |
+
corpus. A query was performed to discover sentences containing numbers between 0-12, after which the resulting
|
| 136 |
+
sentences were manually evaluated for inaccuracies, typos, and the expression of commonsense knowledge. The numerical
|
| 137 |
+
values were then masked.
|
| 138 |
+
|
| 139 |
+
#### Who are the source language producers?
|
| 140 |
+
|
| 141 |
+
The [Open Mind Common Sense](https://huggingface.co/datasets/open_mind_common_sense) corpus, from which this dataset
|
| 142 |
+
is sourced, is a crowdsourced dataset maintained by the MIT Media Lab.
|
| 143 |
+
|
| 144 |
+
### Annotations
|
| 145 |
+
|
| 146 |
+
#### Annotation process
|
| 147 |
+
|
| 148 |
+
No annotations are present in this dataset beyond the `target` values automatically sourced from the masked
|
| 149 |
+
sentences, as discussed above.
|
| 150 |
+
|
| 151 |
+
#### Who are the annotators?
|
| 152 |
+
|
| 153 |
+
The curation and inspection was done in two rounds by graduate students.
|
| 154 |
+
|
| 155 |
+
### Personal and Sensitive Information
|
| 156 |
+
|
| 157 |
+
[More Information Needed]
|
| 158 |
+
|
| 159 |
+
## Considerations for Using the Data
|
| 160 |
+
|
| 161 |
+
### Social Impact of Dataset
|
| 162 |
+
|
| 163 |
+
The motivation of measuring a model's ability to associate numerical values with real-world concepts appears
|
| 164 |
+
relatively innocuous. However, as discussed in the following section, the source dataset may well have biases encoded
|
| 165 |
+
from crowdworkers, particularly in terms of factoid coverage. A model's ability to perform well on this benchmark
|
| 166 |
+
should therefore not be considered evidence that it is more unbiased or objective than a human performing similar
|
| 167 |
+
tasks.
|
| 168 |
+
|
| 169 |
+
[More Information Needed]
|
| 170 |
+
|
| 171 |
+
### Discussion of Biases
|
| 172 |
+
|
| 173 |
+
This dataset is sourced from a crowdsourced commonsense knowledge base. While the information contained in the graph
|
| 174 |
+
is generally considered to be of high quality, the coverage is considered to very low as a representation of all
|
| 175 |
+
possible commonsense knowledge. The representation of certain factoids may also be skewed by the demographics of the
|
| 176 |
+
crowdworkers. As one possible example, the term "homophobia" is connected with "Islam" in the ConceptNet knowledge
|
| 177 |
+
base, but not with any other religion or group, possibly due to the biases of crowdworkers contributing to the
|
| 178 |
+
project.
|
| 179 |
+
|
| 180 |
+
### Other Known Limitations
|
| 181 |
+
|
| 182 |
+
[More Information Needed]
|
| 183 |
+
|
| 184 |
+
## Additional Information
|
| 185 |
+
|
| 186 |
+
### Dataset Curators
|
| 187 |
+
|
| 188 |
+
This dataset was collected by Bill Yuchen Lin, Seyeon Lee, Rahul Khanna, and Xiang Ren, Computer Science researchers
|
| 189 |
+
at the at the University of Southern California.
|
| 190 |
+
|
| 191 |
+
### Licensing Information
|
| 192 |
+
|
| 193 |
+
The data is hosted in a GitHub repositor with the
|
| 194 |
+
[MIT License](https://github.com/INK-USC/NumerSense/blob/main/LICENSE).
|
| 195 |
+
|
| 196 |
+
### Citation Information
|
| 197 |
+
|
| 198 |
+
```
|
| 199 |
+
@inproceedings{lin2020numersense,
|
| 200 |
+
title={Birds have four legs?! NumerSense: Probing Numerical Commonsense Knowledge of Pre-trained Language Models},
|
| 201 |
+
author={Bill Yuchen Lin and Seyeon Lee and Rahul Khanna and Xiang Ren},
|
| 202 |
+
booktitle={Proceedings of EMNLP},
|
| 203 |
+
year={2020},
|
| 204 |
+
note={to appear}
|
| 205 |
+
}
|
| 206 |
+
```
|
| 207 |
+
|
| 208 |
+
### Contributions
|
| 209 |
+
|
| 210 |
+
Thanks to [@joeddav](https://github.com/joeddav) for adding this dataset.
|
huggingface_dataset/Dataset_Card/re_dial.md
ADDED
|
@@ -0,0 +1,450 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
annotations_creators:
|
| 3 |
+
- crowdsourced
|
| 4 |
+
language_creators:
|
| 5 |
+
- crowdsourced
|
| 6 |
+
language:
|
| 7 |
+
- en
|
| 8 |
+
license:
|
| 9 |
+
- cc-by-4.0
|
| 10 |
+
multilinguality:
|
| 11 |
+
- monolingual
|
| 12 |
+
size_categories:
|
| 13 |
+
- 10K<n<100K
|
| 14 |
+
source_datasets:
|
| 15 |
+
- original
|
| 16 |
+
task_categories:
|
| 17 |
+
- other
|
| 18 |
+
- text-classification
|
| 19 |
+
task_ids:
|
| 20 |
+
- sentiment-classification
|
| 21 |
+
paperswithcode_id: redial
|
| 22 |
+
pretty_name: ReDial (Recommendation Dialogues)
|
| 23 |
+
tags:
|
| 24 |
+
- dialogue-sentiment-classification
|
| 25 |
+
dataset_info:
|
| 26 |
+
features:
|
| 27 |
+
- name: movieMentions
|
| 28 |
+
list:
|
| 29 |
+
- name: movieId
|
| 30 |
+
dtype: string
|
| 31 |
+
- name: movieName
|
| 32 |
+
dtype: string
|
| 33 |
+
- name: respondentQuestions
|
| 34 |
+
list:
|
| 35 |
+
- name: movieId
|
| 36 |
+
dtype: string
|
| 37 |
+
- name: suggested
|
| 38 |
+
dtype: int32
|
| 39 |
+
- name: seen
|
| 40 |
+
dtype: int32
|
| 41 |
+
- name: liked
|
| 42 |
+
dtype: int32
|
| 43 |
+
- name: messages
|
| 44 |
+
list:
|
| 45 |
+
- name: timeOffset
|
| 46 |
+
dtype: int32
|
| 47 |
+
- name: text
|
| 48 |
+
dtype: string
|
| 49 |
+
- name: senderWorkerId
|
| 50 |
+
dtype: int32
|
| 51 |
+
- name: messageId
|
| 52 |
+
dtype: int32
|
| 53 |
+
- name: conversationId
|
| 54 |
+
dtype: int32
|
| 55 |
+
- name: respondentWorkerId
|
| 56 |
+
dtype: int32
|
| 57 |
+
- name: initiatorWorkerId
|
| 58 |
+
dtype: int32
|
| 59 |
+
- name: initiatorQuestions
|
| 60 |
+
list:
|
| 61 |
+
- name: movieId
|
| 62 |
+
dtype: string
|
| 63 |
+
- name: suggested
|
| 64 |
+
dtype: int32
|
| 65 |
+
- name: seen
|
| 66 |
+
dtype: int32
|
| 67 |
+
- name: liked
|
| 68 |
+
dtype: int32
|
| 69 |
+
splits:
|
| 70 |
+
- name: train
|
| 71 |
+
num_bytes: 13496125
|
| 72 |
+
num_examples: 10006
|
| 73 |
+
- name: test
|
| 74 |
+
num_bytes: 1731449
|
| 75 |
+
num_examples: 1342
|
| 76 |
+
download_size: 5765261
|
| 77 |
+
dataset_size: 15227574
|
| 78 |
+
---
|
| 79 |
+
|
| 80 |
+
# Dataset Card for ReDial (Recommendation Dialogues)
|
| 81 |
+
|
| 82 |
+
## Table of Contents
|
| 83 |
+
- [Dataset Description](#dataset-description)
|
| 84 |
+
- [Dataset Summary](#dataset-summary)
|
| 85 |
+
- [Supported Tasks and Leaderboards](#supported-tasks-and-leaderboards)
|
| 86 |
+
- [Languages](#languages)
|
| 87 |
+
- [Dataset Structure](#dataset-structure)
|
| 88 |
+
- [Data Instances](#data-instances)
|
| 89 |
+
- [Data Fields](#data-fields)
|
| 90 |
+
- [Data Splits](#data-splits)
|
| 91 |
+
- [Dataset Creation](#dataset-creation)
|
| 92 |
+
- [Curation Rationale](#curation-rationale)
|
| 93 |
+
- [Source Data](#source-data)
|
| 94 |
+
- [Annotations](#annotations)
|
| 95 |
+
- [Personal and Sensitive Information](#personal-and-sensitive-information)
|
| 96 |
+
- [Considerations for Using the Data](#considerations-for-using-the-data)
|
| 97 |
+
- [Social Impact of Dataset](#social-impact-of-dataset)
|
| 98 |
+
- [Discussion of Biases](#discussion-of-biases)
|
| 99 |
+
- [Other Known Limitations](#other-known-limitations)
|
| 100 |
+
- [Additional Information](#additional-information)
|
| 101 |
+
- [Dataset Curators](#dataset-curators)
|
| 102 |
+
- [Licensing Information](#licensing-information)
|
| 103 |
+
- [Citation Information](#citation-information)
|
| 104 |
+
- [Contributions](#contributions)
|
| 105 |
+
|
| 106 |
+
## Dataset Description
|
| 107 |
+
|
| 108 |
+
- **Homepage:** [ReDial Dataset](https://redialdata.github.io/website/)
|
| 109 |
+
- **Repository:** [ReDialData](https://github.com/ReDialData/website/tree/data)
|
| 110 |
+
- **Paper:** [Towards Deep Conversational Recommendations](https://proceedings.neurips.cc/paper/2018/file/800de15c79c8d840f4e78d3af937d4d4-Paper.pdf)
|
| 111 |
+
- **Point of Contact:** [ReDial Google Group](https://groups.google.com/forum/embed/?place=forum/redial-dataset&showpopout=true#!forum/redial-dataset)
|
| 112 |
+
|
| 113 |
+
### Dataset Summary
|
| 114 |
+
|
| 115 |
+
ReDial (Recommendation Dialogues) is an annotated dataset of dialogues, where users
|
| 116 |
+
recommend movies to each other. The dataset was collected by a team of researchers working at
|
| 117 |
+
Polytechnique Montréal, MILA – Quebec AI Institute, Microsoft Research Montréal, HEC Montreal, and Element AI.
|
| 118 |
+
|
| 119 |
+
The dataset allows research at the intersection of goal-directed dialogue systems
|
| 120 |
+
(such as restaurant recommendation) and free-form (also called “chit-chat”) dialogue systems.
|
| 121 |
+
|
| 122 |
+
### Supported Tasks and Leaderboards
|
| 123 |
+
|
| 124 |
+
[More Information Needed]
|
| 125 |
+
|
| 126 |
+
### Languages
|
| 127 |
+
|
| 128 |
+
The text in the dataset is in English.
|
| 129 |
+
|
| 130 |
+
## Dataset Structure
|
| 131 |
+
|
| 132 |
+
### Data Instances
|
| 133 |
+
|
| 134 |
+
JSON-formatted example of a typical instance in the dataset.
|
| 135 |
+
|
| 136 |
+
```
|
| 137 |
+
{
|
| 138 |
+
"movieMentions":{
|
| 139 |
+
"203371":"Final Fantasy: The Spirits Within (2001)",
|
| 140 |
+
"84779":"The Triplets of Belleville (2003)",
|
| 141 |
+
"122159":"Mary and Max (2009)",
|
| 142 |
+
"151313":"A Scanner Darkly (2006)",
|
| 143 |
+
"191602":"Waking Life (2001)",
|
| 144 |
+
"165710":"The Boss Baby (2017)"
|
| 145 |
+
},
|
| 146 |
+
"respondentQuestions":{
|
| 147 |
+
"203371":{
|
| 148 |
+
"suggested":1,
|
| 149 |
+
"seen":0,
|
| 150 |
+
"liked":1
|
| 151 |
+
},
|
| 152 |
+
"84779":{
|
| 153 |
+
"suggested":0,
|
| 154 |
+
"seen":1,
|
| 155 |
+
"liked":1
|
| 156 |
+
},
|
| 157 |
+
"122159":{
|
| 158 |
+
"suggested":0,
|
| 159 |
+
"seen":1,
|
| 160 |
+
"liked":1
|
| 161 |
+
},
|
| 162 |
+
"151313":{
|
| 163 |
+
"suggested":0,
|
| 164 |
+
"seen":1,
|
| 165 |
+
"liked":1
|
| 166 |
+
},
|
| 167 |
+
"191602":{
|
| 168 |
+
"suggested":0,
|
| 169 |
+
"seen":1,
|
| 170 |
+
"liked":1
|
| 171 |
+
},
|
| 172 |
+
"165710":{
|
| 173 |
+
"suggested":1,
|
| 174 |
+
"seen":0,
|
| 175 |
+
"liked":1
|
| 176 |
+
}
|
| 177 |
+
},
|
| 178 |
+
"messages":[
|
| 179 |
+
{
|
| 180 |
+
"timeOffset":0,
|
| 181 |
+
"text":"Hi there, how are you? I'm looking for movie recommendations",
|
| 182 |
+
"senderWorkerId":0,
|
| 183 |
+
"messageId":1021
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"timeOffset":15,
|
| 187 |
+
"text":"I am doing okay. What kind of movies do you like?",
|
| 188 |
+
"senderWorkerId":1,
|
| 189 |
+
"messageId":1022
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"timeOffset":66,
|
| 193 |
+
"text":"I like animations like @84779 and @191602",
|
| 194 |
+
"senderWorkerId":0,
|
| 195 |
+
"messageId":1023
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"timeOffset":86,
|
| 199 |
+
"text":"I also enjoy @122159",
|
| 200 |
+
"senderWorkerId":0,
|
| 201 |
+
"messageId":1024
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"timeOffset":95,
|
| 205 |
+
"text":"Anything artistic",
|
| 206 |
+
"senderWorkerId":0,
|
| 207 |
+
"messageId":1025
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"timeOffset":135,
|
| 211 |
+
"text":"You might like @165710 that was a good movie.",
|
| 212 |
+
"senderWorkerId":1,
|
| 213 |
+
"messageId":1026
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"timeOffset":151,
|
| 217 |
+
"text":"What's it about?",
|
| 218 |
+
"senderWorkerId":0,
|
| 219 |
+
"messageId":1027
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"timeOffset":207,
|
| 223 |
+
"text":"It has Alec Baldwin it is about a baby that works for a company and gets adopted it is very funny",
|
| 224 |
+
"senderWorkerId":1,
|
| 225 |
+
"messageId":1028
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"timeOffset":238,
|
| 229 |
+
"text":"That seems like a nice comedy",
|
| 230 |
+
"senderWorkerId":0,
|
| 231 |
+
"messageId":1029
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"timeOffset":272,
|
| 235 |
+
"text":"Do you have any animated recommendations that are a bit more dramatic? Like @151313 for example",
|
| 236 |
+
"senderWorkerId":0,
|
| 237 |
+
"messageId":1030
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"timeOffset":327,
|
| 241 |
+
"text":"I like comedies but I prefer films with a little more depth",
|
| 242 |
+
"senderWorkerId":0,
|
| 243 |
+
"messageId":1031
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"timeOffset":467,
|
| 247 |
+
"text":"That is a tough one but I will remember something",
|
| 248 |
+
"senderWorkerId":1,
|
| 249 |
+
"messageId":1032
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"timeOffset":509,
|
| 253 |
+
"text":"@203371 was a good one",
|
| 254 |
+
"senderWorkerId":1,
|
| 255 |
+
"messageId":1033
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"timeOffset":564,
|
| 259 |
+
"text":"Ooh that seems cool! Thanks for the input. I'm ready to submit if you are.",
|
| 260 |
+
"senderWorkerId":0,
|
| 261 |
+
"messageId":1034
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"timeOffset":571,
|
| 265 |
+
"text":"It is animated, sci fi, and has action",
|
| 266 |
+
"senderWorkerId":1,
|
| 267 |
+
"messageId":1035
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"timeOffset":579,
|
| 271 |
+
"text":"Glad I could help",
|
| 272 |
+
"senderWorkerId":1,
|
| 273 |
+
"messageId":1036
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"timeOffset":581,
|
| 277 |
+
"text":"Nice",
|
| 278 |
+
"senderWorkerId":0,
|
| 279 |
+
"messageId":1037
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"timeOffset":591,
|
| 283 |
+
"text":"Take care, cheers!",
|
| 284 |
+
"senderWorkerId":0,
|
| 285 |
+
"messageId":1038
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"timeOffset":608,
|
| 289 |
+
"text":"bye",
|
| 290 |
+
"senderWorkerId":1,
|
| 291 |
+
"messageId":1039
|
| 292 |
+
}
|
| 293 |
+
],
|
| 294 |
+
"conversationId":"391",
|
| 295 |
+
"respondentWorkerId":1,
|
| 296 |
+
"initiatorWorkerId":0,
|
| 297 |
+
"initiatorQuestions":{
|
| 298 |
+
"203371":{
|
| 299 |
+
"suggested":1,
|
| 300 |
+
"seen":0,
|
| 301 |
+
"liked":1
|
| 302 |
+
},
|
| 303 |
+
"84779":{
|
| 304 |
+
"suggested":0,
|
| 305 |
+
"seen":1,
|
| 306 |
+
"liked":1
|
| 307 |
+
},
|
| 308 |
+
"122159":{
|
| 309 |
+
"suggested":0,
|
| 310 |
+
"seen":1,
|
| 311 |
+
"liked":1
|
| 312 |
+
},
|
| 313 |
+
"151313":{
|
| 314 |
+
"suggested":0,
|
| 315 |
+
"seen":1,
|
| 316 |
+
"liked":1
|
| 317 |
+
},
|
| 318 |
+
"191602":{
|
| 319 |
+
"suggested":0,
|
| 320 |
+
"seen":1,
|
| 321 |
+
"liked":1
|
| 322 |
+
},
|
| 323 |
+
"165710":{
|
| 324 |
+
"suggested":1,
|
| 325 |
+
"seen":0,
|
| 326 |
+
"liked":1
|
| 327 |
+
}
|
| 328 |
+
}
|
| 329 |
+
}
|
| 330 |
+
```
|
| 331 |
+
|
| 332 |
+
### Data Fields
|
| 333 |
+
|
| 334 |
+
The dataset is published in the “jsonl” format, i.e., as a text file where each line corresponds to a Dialogue given as a valid JSON document.
|
| 335 |
+
|
| 336 |
+
A Dialogue contains these fields:
|
| 337 |
+
|
| 338 |
+
**conversationId:** an integer
|
| 339 |
+
**initiatorWorkerId:** an integer identifying to the worker initiating the conversation (the recommendation seeker)
|
| 340 |
+
**respondentWorkerId:** an integer identifying the worker responding to the initiator (the recommender)
|
| 341 |
+
**messages:** a list of Message objects
|
| 342 |
+
**movieMentions:** a dict mapping movie IDs mentioned in this dialogue to movie names
|
| 343 |
+
**initiatorQuestions:** a dictionary mapping movie IDs to the labels supplied by the initiator. Each label is a bool corresponding to whether the initiator has said he saw the movie, liked it, or suggested it.
|
| 344 |
+
**respondentQuestions:** a dictionary mapping movie IDs to the labels supplied by the respondent. Each label is a bool corresponding to whether the initiator has said he saw the movie, liked it, or suggested it.
|
| 345 |
+
Each Message contains these fields:
|
| 346 |
+
|
| 347 |
+
**messageId:** a unique ID for this message
|
| 348 |
+
**text:** a string with the actual message. The string may contain a token starting with @ followed by an integer. This is a movie ID which can be looked up in the movieMentions field of the Dialogue object.
|
| 349 |
+
**timeOffset:** time since start of dialogue in seconds
|
| 350 |
+
**senderWorkerId:** the ID of the worker sending the message, either initiatorWorkerId or respondentWorkerId.
|
| 351 |
+
|
| 352 |
+
The labels in initiatorQuestions and respondentQuestions have the following meaning:
|
| 353 |
+
*suggested:* 0 if it was mentioned by the seeker, 1 if it was a suggestion from the recommender
|
| 354 |
+
*seen:* 0 if the seeker has not seen the movie, 1 if they have seen it, 2 if they did not say
|
| 355 |
+
*liked:* 0 if the seeker did not like the movie, 1 if they liked it, 2 if they did not say
|
| 356 |
+
|
| 357 |
+
### Data Splits
|
| 358 |
+
|
| 359 |
+
The dataset contains a total of 11348 dialogues, 10006 for training and model selection, and 1342 for testing.
|
| 360 |
+
|
| 361 |
+
## Dataset Creation
|
| 362 |
+
|
| 363 |
+
### Curation Rationale
|
| 364 |
+
|
| 365 |
+
The dataset allows research at the intersection of goal-directed dialogue systems (such as restaurant recommendation) and free-form (also called “chit-chat”) dialogue systems.
|
| 366 |
+
|
| 367 |
+
In the dataset, users talk about which movies they like and which ones they do not like, which ones they have seen or not etc., and labels which we ensured agree between the two participants. This allows to research how sentiment is expressed in dialogues, which differs a lot from e.g. review websites.
|
| 368 |
+
|
| 369 |
+
The dialogues and the movies they mention form a curious bi-partite graph structure, which is related to how users talk about the movie (e.g. genre information).
|
| 370 |
+
|
| 371 |
+
Ignoring label information, this dataset can also be viewed as a limited domain chit-chat dialogue dataset.
|
| 372 |
+
|
| 373 |
+
### Source Data
|
| 374 |
+
|
| 375 |
+
#### Initial Data Collection and Normalization
|
| 376 |
+
|
| 377 |
+
Describe the data collection process. Describe any criteria for data selection or filtering. List any key words or search terms used. If possible, include runtime information for the collection process.
|
| 378 |
+
|
| 379 |
+
If data was collected from other pre-existing datasets, link to source here and to their [Hugging Face version](https://huggingface.co/datasets/dataset_name).
|
| 380 |
+
|
| 381 |
+
If the data was modified or normalized after being collected (e.g. if the data is word-tokenized), describe the process and the tools used.
|
| 382 |
+
|
| 383 |
+
#### Who are the source language producers?
|
| 384 |
+
|
| 385 |
+
Here we formalize the setup of a conversation involving recommendations for the purposes of data collection. To provide some additional structure to our data (and models) we define one person in the dialogue as the recommendation seeker and the other as the recommender.
|
| 386 |
+
|
| 387 |
+
To obtain data in this form, we developed an interface and pairing mechanism mediated by Amazon Mechanical Turk (AMT).
|
| 388 |
+
|
| 389 |
+
We pair up AMT workers and give each of them a role. The movie seeker has to explain what kind of movie he/she likes, and asks for movie suggestions. The recommender tries to understand the seeker’s movie tastes, and recommends movies. All exchanges of information and recommendations are made using natural language.
|
| 390 |
+
|
| 391 |
+
We add additional instructions to improve the data quality and guide the workers to dialogue the way we expect them to. Thus we ask to use formal language and that conversations contain roughly ten messages minimum. We also require that at least four different movies are mentioned in every conversation. Finally, we also ask to converse only about movies, and notably not to mention Mechanical Turk or the task itself.
|
| 392 |
+
|
| 393 |
+
In addition, we ask that every movie mention is tagged using the ‘@’ symbol. When workers type ‘@’, the following characters are used to find matching movie names, and workers can choose a movie from that list. This allows us to detect exactly what movies are mentioned and when. We gathered entities from DBpedia that were of type http://dbpedia.org/ontology/Film to obtain a list of movies, but also allow workers to add their own movies to the list if it is not present already. We obtained the release dates from the movie titles (e.g. http://dbpedia.org/page/American_Beauty_(1999_film), or, if the movie title does not contain that information, from an additional SPARQL request. Note that the year or release date of a movie can be essential to differentiate movies with the same name, but released at different dates.
|
| 394 |
+
|
| 395 |
+
We will refer to these additional labels as movie dialogue forms. Both workers have to answer these forms even though it really concerns the seeker’s movie tastes. Ideally, the two participants would give the same answer to every form, but it is possible that their answers do not coincide (because of carelessness, or dialogue ambiguity). The movie dialogue forms therefore allow us to evaluate sub-components of an overall neural dialogue system more systematically, for example one can train and evaluate a sentiment analysis model directly using these labels. %which could produce a reward for the dialogue agent.
|
| 396 |
+
|
| 397 |
+
In each conversation, the number of movies mentioned varies, so we have different numbers of movie dialogue form answers for each conversation. The distribution of the different classes of the movie dialogue form is shown in Table 1a. The liked/disliked/did not say label is highly imbalanced. This is standard for recommendation data, since people are naturally more likely to talk about movies that they like, and the recommender’s objective is to recommend movies that the seeker is likely to like.
|
| 398 |
+
|
| 399 |
+
### Annotations
|
| 400 |
+
|
| 401 |
+
#### Annotation process
|
| 402 |
+
|
| 403 |
+
Mentioned in above sub-section.
|
| 404 |
+
|
| 405 |
+
#### Who are the annotators?
|
| 406 |
+
|
| 407 |
+
For the AMT HIT we collect data in English and chose to restrict the data collection to countries where English is the main language. The fact that we pair workers together slows down the data collection since we ask that at least two persons are online at the same time to do the task, so a good amount of workers is required to make the collection possible. Meanwhile, the task is quite demanding, and we have to select qualified workers. HIT reward and qualification requirement were decisive to get good conversation quality while still ensuring that people could get paired together. We launched preliminary HITs to find a compromise and finally set the reward to $0.50 per person for each completed conversation (so each conversation costs us $1, plus taxes), and ask that workers meet the following requirements: (1)~Approval percentage greater than 95, (2)~Number of approved HITs greater than 1000, (3)~Their location must be in United States, Canada, United Kingdom, Australia, or New Zealand.
|
| 408 |
+
|
| 409 |
+
### Personal and Sensitive Information
|
| 410 |
+
|
| 411 |
+
Workers had to confirm a consent form before every task that explains what the data is being collected for and how it is going to be used.
|
| 412 |
+
|
| 413 |
+
## Considerations for Using the Data
|
| 414 |
+
|
| 415 |
+
### Social Impact of Dataset
|
| 416 |
+
|
| 417 |
+
[More Information Needed]
|
| 418 |
+
|
| 419 |
+
### Discussion of Biases
|
| 420 |
+
|
| 421 |
+
[More Information Needed]
|
| 422 |
+
|
| 423 |
+
### Other Known Limitations
|
| 424 |
+
|
| 425 |
+
[More Information Needed]
|
| 426 |
+
|
| 427 |
+
## Additional Information
|
| 428 |
+
|
| 429 |
+
### Dataset Curators
|
| 430 |
+
|
| 431 |
+
The dataset collection was funded by Google, IBM, and NSERC, with editorial support from Microsoft Research.
|
| 432 |
+
|
| 433 |
+
### Licensing Information
|
| 434 |
+
|
| 435 |
+
The data is published under the CC BY 4.0 License.
|
| 436 |
+
|
| 437 |
+
### Citation Information
|
| 438 |
+
|
| 439 |
+
```
|
| 440 |
+
@inproceedings{li2018conversational,
|
| 441 |
+
title={Towards Deep Conversational Recommendations},
|
| 442 |
+
author={Li, Raymond and Kahou, Samira Ebrahimi and Schulz, Hannes and Michalski, Vincent and Charlin, Laurent and Pal, Chris},
|
| 443 |
+
booktitle={Advances in Neural Information Processing Systems 31 (NIPS 2018)},
|
| 444 |
+
year={2018}
|
| 445 |
+
}
|
| 446 |
+
```
|
| 447 |
+
|
| 448 |
+
### Contributions
|
| 449 |
+
|
| 450 |
+
Thanks to [@bhavitvyamalik](https://github.com/bhavitvyamalik) for adding this dataset.
|
huggingface_dataset/Dataset_Card/ronig_protein_binding_sequences.md
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: mit
|
| 3 |
+
pretty_name: Sequence Based Protein - Peptide Binding Dataset
|
| 4 |
+
---
|
| 5 |
+
# Sequence Based Protein - Peptide Binding Dataset
|
| 6 |
+
- Data sources:
|
| 7 |
+
- [Huang Laboratory](http://huanglab.phys.hust.edu.cn)
|
| 8 |
+
- [Propedia](http://bioinfo.dcc.ufmg.br/propedia/)
|
| 9 |
+
- Dataset size: 15,764 sets of Protein-Peptide sequences that bind, the protein sequence contains only the relevant chain.
|
| 10 |
+
- Train / Val split: the dataset is split to 80% train 10% val and 10% test.
|
huggingface_dataset/Dataset_Card/stanfordnlp_SHP.md
ADDED
|
@@ -0,0 +1,264 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
task_categories:
|
| 3 |
+
- text-generation
|
| 4 |
+
- question-answering
|
| 5 |
+
tags:
|
| 6 |
+
- human feedback
|
| 7 |
+
- rlhf
|
| 8 |
+
- preferences
|
| 9 |
+
- reddit
|
| 10 |
+
- preference model
|
| 11 |
+
- RL
|
| 12 |
+
- NLG
|
| 13 |
+
- evaluation
|
| 14 |
+
size_categories:
|
| 15 |
+
- 100K<n<1M
|
| 16 |
+
language:
|
| 17 |
+
- en
|
| 18 |
+
---
|
| 19 |
+
# 🚢 Stanford Human Preferences Dataset (SHP)
|
| 20 |
+
|
| 21 |
+
## Summary
|
| 22 |
+
|
| 23 |
+
SHP is a dataset of **385K collective human preferences** over responses to questions/instructions in 18 different subject areas, from cooking to legal advice.
|
| 24 |
+
The preferences are meant to reflect the helpfulness of one response over another, and are intended to be used for training RLHF reward models and NLG evaluation models (e.g., [SteamSHP](https://huggingface.co/stanfordnlp/SteamSHP-flan-t5-xl)).
|
| 25 |
+
|
| 26 |
+
Each example is a Reddit post with a question/instruction and a pair of top-level comments for that post, where one comment is more preferred by Reddit users (collectively).
|
| 27 |
+
SHP exploits the fact that if comment A was written *after* comment B but has a higher score nonetheless, then A is ostensibly more preferred to B.
|
| 28 |
+
If A had been written before B, then we could not conclude this, since its higher score could have been the result of more visibility.
|
| 29 |
+
We chose data where the preference label is intended to reflect which response is more *helpful* rather than which is less *harmful*, the latter being the focus of much past work.
|
| 30 |
+
|
| 31 |
+
How is SHP different from [Anthropic's HH-RLHF dataset](https://huggingface.co/datasets/Anthropic/hh-rlhf)?
|
| 32 |
+
Most notably, all the data in SHP is naturally occurring and human-written, whereas the responses in HH-RLHF are machine-written, giving us two very different distributions that can complement each other.
|
| 33 |
+
|
| 34 |
+
| Dataset | Size | Input | Label | Domains | Data Format | Length |
|
| 35 |
+
| -------------------- | ---- | -------------------------- | ---------------------------- | ------------------------- | ------------------------------------- | --------------- |
|
| 36 |
+
| SHP | 385K | Naturally occurring human-written responses | Collective Human Preference | 18 (labelled) | Question/Instruction + Response (Single-turn) | up to 10.1K T5 tokens |
|
| 37 |
+
| HH-RLHF | 91K | Dialogue with LLM | Individual Human Preference | not labelled | Live Chat (Multi-turn) | up to 1.5K T5 tokens |
|
| 38 |
+
|
| 39 |
+
How is SHP different from other datasets that have scraped Reddit, like [ELI5](https://huggingface.co/datasets/eli5#source-data)?
|
| 40 |
+
SHP uses the timestamp information to infer preferences, while ELI5 only provides comments and scores -- the latter are not enough to infer preferences since comments made earlier tend to get higher scores from more visibility.
|
| 41 |
+
It also contains data from more domains:
|
| 42 |
+
|
| 43 |
+
| Dataset | Size | Comments + Scores | Preferences | Number of Domains |
|
| 44 |
+
| -------------------- | ---- | ------------------ | -------------| ------------------ |
|
| 45 |
+
| SHP | 385K | Yes | Yes | 18 |
|
| 46 |
+
| ELI5 | 270K | Yes | No | 3 |
|
| 47 |
+
|
| 48 |
+
|
| 49 |
+
## Data Structure
|
| 50 |
+
|
| 51 |
+
There are 18 directories, one for each subreddit, and each directory contains a JSONL file for the training, validation, and test data.
|
| 52 |
+
Here's how to get the data using Huggingface's `datasets` library:
|
| 53 |
+
|
| 54 |
+
```python
|
| 55 |
+
from datasets import load_dataset
|
| 56 |
+
|
| 57 |
+
# Load all the data
|
| 58 |
+
dataset = load_dataset("stanfordnlp/shp")
|
| 59 |
+
|
| 60 |
+
# Load one of the subreddits
|
| 61 |
+
dataset = load_dataset("stanfordnlp/shp", data_dir="askculinary")
|
| 62 |
+
```
|
| 63 |
+
|
| 64 |
+
Here's an example from `askculinary/train.json`:
|
| 65 |
+
```
|
| 66 |
+
{
|
| 67 |
+
`post_id`:"qt3nxl",
|
| 68 |
+
`domain`:"askculinary_train",
|
| 69 |
+
`upvote_ratio`:0.98,
|
| 70 |
+
`history`:"What's the best way to disassemble raspberries? Like this, but down to the individual seeds: https:\/\/i.imgur.com\/Z0c6ZKE.jpg I've been pulling them apart with tweezers and it's really time consuming. I have about 10 pounds to get through this weekend.",
|
| 71 |
+
`c_root_id_A`:"hkh25sc",
|
| 72 |
+
`c_root_id_B`:"hkh25lp",
|
| 73 |
+
`created_at_utc_A`:1636822112,
|
| 74 |
+
`created_at_utc_B`:1636822110,
|
| 75 |
+
`score_A`:340,
|
| 76 |
+
`score_B`:166,
|
| 77 |
+
`human_ref_A`:"Pectinex, perhaps? It's an enzyme that breaks down cellulose. With citrus, you let it sit in a dilute solution of pectinex overnight to break down the connective tissues. You end up with perfect citrus supremes. If you let the raspberries sit for a shorter time, I wonder if it would separate the seeds the same way...? Here's an example: https:\/\/www.chefsteps.com\/activities\/perfect-citrus-supreme",
|
| 78 |
+
`human_ref_B`:"Raspberry juice will make a bright stain at first, but in a matter of weeks it will start to fade away to almost nothing. It is what is known in the natural dye world as a fugitive dye, it will fade even without washing or exposure to light. I hope she gets lots of nice photos of these stains on her dress, because soon that will be all she has left of them!",
|
| 79 |
+
`labels`:1,
|
| 80 |
+
`seconds_difference`:2.0,
|
| 81 |
+
`score_ratio`:2.0481927711
|
| 82 |
+
}
|
| 83 |
+
```
|
| 84 |
+
|
| 85 |
+
where the fields are:
|
| 86 |
+
- ```post_id```: the ID of the Reddit post (string)
|
| 87 |
+
- ```domain```: the subreddit and split the example is drawn from, separated by an underscore (string)
|
| 88 |
+
- ```upvote_ratio```: the percent of votes received by the post that were positive (aka upvotes) (float)
|
| 89 |
+
- ```history```: the post title concatented to the post body (string)
|
| 90 |
+
- ```c_root_id_A```: the ID of comment A (string)
|
| 91 |
+
- ```c_root_id_B```: the ID of comment B (string)
|
| 92 |
+
- ```created_at_utc_A```: utc timestamp of when comment A was created (integer)
|
| 93 |
+
- ```created_at_utc_B```: utc timestamp of when comment B was created (integer)
|
| 94 |
+
- ```score_A```: (# positive votes - # negative votes + 1) received by comment A (integer)
|
| 95 |
+
- ```score_B```: (# positive votes - # negative votes + 1) received by comment B (integer)
|
| 96 |
+
- ```human_ref_A```: text of comment A (string)
|
| 97 |
+
- ```human_ref_B```: text of comment B (string)
|
| 98 |
+
- ```labels```: the preference label -- it is 1 if A is preferred to B; 0 if B is preferred to A. This was randomized such that the label distribution is roughly 50/50. (integer)
|
| 99 |
+
- ```seconds_difference```: how many seconds after the less preferred comment the more preferred one was created (will always be >= 0) (integer)
|
| 100 |
+
- ```score_ratio```: the ratio of the more preferred comment's score to the less preferred comment's score (will be >= 1) (float)
|
| 101 |
+
|
| 102 |
+
|
| 103 |
+
## Dataset Design
|
| 104 |
+
|
| 105 |
+
### Domain Selection
|
| 106 |
+
|
| 107 |
+
The data is sourced from Reddit, which is a public forum organized into topic-specific fora called *subreddits*.
|
| 108 |
+
For example, the `askculinary` subreddit is where users ask cooking-related questions and are answered by other users.
|
| 109 |
+
|
| 110 |
+
SHP contains a train, validation, and test split for comments scraped from 18 different subreddits. We chose subreddits based on:
|
| 111 |
+
1. whether they were well-known (subscriber count >= 100K)
|
| 112 |
+
2. whether posts were expected to pose a question or instruction
|
| 113 |
+
3. whether responses were valued based on how *helpful* they were
|
| 114 |
+
4. whether comments had to be rooted in some objectivity, instead of being entirely about personal experiences (e.g., `askscience` vs. `AskAmericans`)
|
| 115 |
+
|
| 116 |
+
The train/validation/test splits were created by splitting the post IDs of a subreddit in 90%/5%/5% proportions respectively, so that no post would appear in multiple splits.
|
| 117 |
+
Since different posts have different numbers of comments, the number of preferences in each split is not exactly 90%/5%/5%:
|
| 118 |
+
|
| 119 |
+
| subreddit | train | validation | test | total |
|
| 120 |
+
| ------------------ | -------: | ---------: | ---: | ----: |
|
| 121 |
+
| askacademia | 31450 | 2095 | 1708 | 35253 |
|
| 122 |
+
| askanthropology | 3910 | 203 | 268 | 4381 |
|
| 123 |
+
| askbaking | 44007 | 2096 | 1544 | 47647 |
|
| 124 |
+
| askcarguys | 3227 | 159 | 117 | 3503 |
|
| 125 |
+
| askculinary | 45710 | 2094 | 2563 | 50367 |
|
| 126 |
+
| askdocs | 6449 | 315 | 455 | 7219 |
|
| 127 |
+
| askengineers | 57096 | 3154 | 2638 | 62888 |
|
| 128 |
+
| askhistorians | 3264 | 113 | 164 | 3541 |
|
| 129 |
+
| askhr | 8295 | 641 | 395 | 9331 |
|
| 130 |
+
| askphilosophy | 10307 | 608 | 677 | 11592 |
|
| 131 |
+
| askphysics | 7364 | 409 | 587 | 8360 |
|
| 132 |
+
| askscience | 13316 | 899 | 977 | 15192 |
|
| 133 |
+
| asksciencefiction | 29382 | 1576 | 1987 | 32945 |
|
| 134 |
+
| asksocialscience | 2706 | 147 | 188 | 3041 |
|
| 135 |
+
| askvet | 3300 | 170 | 224 | 3694 |
|
| 136 |
+
| changemyview | 38173 | 1637 | 1836 | 41646 |
|
| 137 |
+
| explainlikeimfive | 19592 | 1014 | 1070 | 21676 |
|
| 138 |
+
| legaladvice | 21170 | 1106 | 1011 | 23287 |
|
| 139 |
+
| ALL | 348718 | 18436 | 18409 | 385563 |
|
| 140 |
+
|
| 141 |
+
### Data Selection
|
| 142 |
+
|
| 143 |
+
The score of a post/comment is 1 plus the number of upvotes (approvals) it gets from users, minus the number of downvotes (disapprovals) it gets.
|
| 144 |
+
The value of a score is relative; in subreddits(posts) with more traffic, there will be more higher-scoring posts(comments).
|
| 145 |
+
Within a post, comments posted earlier will tend to have a higher score simply due to having more exposure, which is why using timestamp information is essential when inferring preferences.
|
| 146 |
+
|
| 147 |
+
Given a post P and two comments (A,B) we only included the preference A > B in the dataset if
|
| 148 |
+
1. A was written *no later than* B and A has a higher score than B.
|
| 149 |
+
2. The post is a self-post (i.e., a body of text and not a link to another page) made before 2023, was not edited, and is not NSFW (over 18).
|
| 150 |
+
3. Neither comment was made by a deleted user, a moderator, or the post creator. The post was not made by a deleted user or moderator.
|
| 151 |
+
4. The post has a score >= 10 and each comment has a score >= 2 (upvoted at least once).
|
| 152 |
+
|
| 153 |
+
A post with `n` comments could have up to (`n` choose `2`) preferences in the data.
|
| 154 |
+
Since the number of comments per post is Pareto-distributed, to prevent a relatively small number of posts from dominating the data, we limited the scraping to 50 comments per post.
|
| 155 |
+
This means that each post could have up to (`50` choose `2`) comments in the dataset, though this is a much smaller number in practice, since all the criteria above need to be met.
|
| 156 |
+
|
| 157 |
+
Reddit makes it very difficult to get anything beyond the top 1000 posts for each subreddit.
|
| 158 |
+
We started with the top-scoring 1000 posts (of all time) and searched for the 25 most similar posts to each one using Reddit's search function to get up to 7500 unique post IDs per subreddit.
|
| 159 |
+
|
| 160 |
+
|
| 161 |
+
### Preprocessing
|
| 162 |
+
|
| 163 |
+
We tried to keep preprocessing to a minimum. Subreddit-specific abbreviations were expanded (e.g., "CMV" to "Change my view that").
|
| 164 |
+
In hyperlinks, only the referring text was kept and the URL was removed (if the URL was written out, then it was kept).
|
| 165 |
+
|
| 166 |
+
|
| 167 |
+
## Building a Preference Model
|
| 168 |
+
|
| 169 |
+
### Finetuning
|
| 170 |
+
|
| 171 |
+
If you want to finetune a model to predict human preferences (e.g., for NLG evaluation or an RLHF reward model), here are some helpful tips:
|
| 172 |
+
|
| 173 |
+
1. **Preprocess the data.** The total input length should fit under the model's token limit (usually 512 tokens).
|
| 174 |
+
Although models like FLAN-T5 use positional embeddings, we found that the loss would not converge if we finetuned it on inputs over 512 tokens.
|
| 175 |
+
To avoid this, truncate the post text (in the `history` field) as much as possible, such that the whole input is under 512 tokens (do not truncate the comment(s) however).
|
| 176 |
+
If this is still over 512 tokens, simply skip the example.
|
| 177 |
+
2. **Use a sufficiently large model.**
|
| 178 |
+
Finetuning a single FLAN-T5-xl model across all the training data should give you a test accuracy between 72-73% (across all domains on examples where the entire input fits within the token limit), ranging from 65-80% on individual subreddits.
|
| 179 |
+
3. **Do in-domain prediction.** Out-of-domain performance will be poor if the subreddits are unrelated (e.g., if you fine-tune on `askculinary` preferences and test on `askcarguys` preferences).
|
| 180 |
+
4. **Train for fewer epochs.** The InstructGPT paper paper suggests training a reward model for only 1 epoch.
|
| 181 |
+
Since the same comment appears in multiple preferences, it is easy to overfit to the data.
|
| 182 |
+
5. **Training on less data may help.**
|
| 183 |
+
Preferences with a large `score_ratio` (e.g., comment A having 2x the score of comment B) will provide a stronger signal for finetuning the model, so you may only want to consider preferences above a certain `score_ratio`.
|
| 184 |
+
The number of preferences per post is Pareto-distributed, so to prevent the model from over-fitting to certain posts, you may want to limit the number of preferences from a particular post.
|
| 185 |
+
|
| 186 |
+
### Evaluating
|
| 187 |
+
|
| 188 |
+
Since it is easier to predict strongly-held preferences than weakly-held ones, instead of reporting a single accuracy value, we recommend reporting a performance curve as a function of the `score_ratio`.
|
| 189 |
+
For example, here is the accuracy curve for a FLAN-T5-xl model trained on the askculinary data using the suggestions above.
|
| 190 |
+
The orange line is from finetuning only on preferences with a 2+ score ratio and using no more than 5 preferences from each post to prevent overfitting:
|
| 191 |
+
|
| 192 |
+
|
| 193 |
+
|
| 194 |
+
We see that finetuning on less -- but higher quality -- data leads to higher accuracies on test data with a score ratio below 3.5, with no real downsides!
|
| 195 |
+
Note that any examples whose inputs did not fit within the token limit were left out of the experiment, since the model could not be expected to handle them.
|
| 196 |
+
|
| 197 |
+
### SteamSHP - An Open-Source Preference Model
|
| 198 |
+
|
| 199 |
+
We have finetuned two FLAN-T5 models on both the SHP dataset and the helpfulness data from Anthropic's HH-RLHF. They are
|
| 200 |
+
- [SteamSHP-XL](https://huggingface.co/stanfordnlp/SteamSHP-flan-t5-xl), a 3B parameter model that achieves 72.8% on the test data.
|
| 201 |
+
- [SteamSHP-Large](https://huggingface.co/stanfordnlp/SteamSHP-flan-t5-large), a 780M parameter model that achieves 72.0% on the test data.
|
| 202 |
+
|
| 203 |
+
We encourage you to use SteamSHP for NLG evaluation, for building reward models for RLHF, or for another purpose you deem fit!
|
| 204 |
+
|
| 205 |
+
|
| 206 |
+
## Biases and Limitations
|
| 207 |
+
|
| 208 |
+
### Biases
|
| 209 |
+
|
| 210 |
+
Although we filtered out posts with NSFW (over 18) content, chose subreddits that were well-moderated and had policies against harassment and bigotry, some of the data may contain discriminatory or harmful language.
|
| 211 |
+
The data does not reflect the views of the dataset creators.
|
| 212 |
+
Reddit users on these subreddits are also not representative of the broader population.
|
| 213 |
+
Although subreddit-specific demographic information is not available, Reddit users overall are disproportionately male and from developed, Western, and English-speaking countries ([Pew Research](https://www.pewresearch.org/internet/2013/07/03/6-of-online-adults-are-reddit-users/)).
|
| 214 |
+
Please keep this in mind before using any models trained on this data.
|
| 215 |
+
|
| 216 |
+
### Limitations
|
| 217 |
+
|
| 218 |
+
The preference label in SHP is intended to reflect how *helpful* one response is relative to another, given an instruction/question.
|
| 219 |
+
SHP is not intended for use in harm-minimization, as it was not designed to include the toxic content that would be necessary to learn a good toxicity detector.
|
| 220 |
+
If you are looking for data where the preference label denotes less harm, we would recommend the harmfulness split of [Anthropic's HH-RLHF](https://huggingface.co/datasets/Anthropic/hh-rlhf).
|
| 221 |
+
|
| 222 |
+
Another limitation is that the more preferred response in SHP is not necessarily the more factual one.
|
| 223 |
+
Though some comments do provide citations to justify their response, most do not.
|
| 224 |
+
There are exceptions to this, such as the `askhistorians` subreddit, which is heavily moderated and answers are expected to provide citations.
|
| 225 |
+
|
| 226 |
+
Note that the collective preference label in SHP is not necessarily what we would get if we asked users to independently vote on each comment before taking an unweighted sum.
|
| 227 |
+
This is because comment scores on Reddit are public and are known to influence user preferences; a high score increases the likelihood of getting more positive votes [(Muchnik et al., 2013)](https://pubmed.ncbi.nlm.nih.gov/23929980/).
|
| 228 |
+
Whether this "herding effect" temporarily or permanently shifts a user's preference is unclear.
|
| 229 |
+
Therefore, while SHP does reflect collective human preferences, models trained on SHP may not generalize to settings where individual preferences are aggregated differently (e.g., users vote independently without ever seeing the current comment score, users vote after conferring, etc.).
|
| 230 |
+
Thanks to Greg Stoddard for pointing this out.
|
| 231 |
+
|
| 232 |
+
|
| 233 |
+
## License
|
| 234 |
+
|
| 235 |
+
Last updated: 03/01/2023
|
| 236 |
+
|
| 237 |
+
This dataset was made by scraping Reddit in accordance with the [Reddit API Terms of Use](https://docs.google.com/a/reddit.com/forms/d/e/1FAIpQLSezNdDNK1-P8mspSbmtC2r86Ee9ZRbC66u929cG2GX0T9UMyw/viewform), without any direct communication or written agreements with Reddit.
|
| 238 |
+
According to the Terms of Use, "User Content" is owned by the users themselves -- not by Reddit -- and Reddit grants a "non-exclusive, non-transferable, non-sublicensable, and revocable license to copy and display the User Content".
|
| 239 |
+
|
| 240 |
+
Datasets made by scraping Reddit are widely used in the research community: for example, Facebook AI Research used data scraped from Reddit to make the [ELI5](https://huggingface.co/datasets/eli5#source-data) dataset in 2019, which was made available without a license.
|
| 241 |
+
Anthropic AI has also [attested to scraping Reddit](https://arxiv.org/pdf/2112.00861.pdf) for preferences using a different methodology, though this data was not made public.
|
| 242 |
+
The [PushShift Reddit dataset](https://arxiv.org/abs/2001.08435), which makes entire dumps of Reddit available on a regular schedule, is also made available without a license (to our knowledge).
|
| 243 |
+
|
| 244 |
+
We take no responsibility for and we do not expressly or implicitly endorse any downstream use of this dataset.
|
| 245 |
+
We reserve the right to modify the SHP dataset and this license at any point in the future.
|
| 246 |
+
|
| 247 |
+
|
| 248 |
+
## Contact
|
| 249 |
+
|
| 250 |
+
Please contact kawin@stanford.edu if you have any questions about the data.
|
| 251 |
+
This dataset was created by Kawin Ethayarajh, Heidi (Chenyu) Zhang, Yizhong Wang, and Dan Jurafsky.
|
| 252 |
+
|
| 253 |
+
## Citation
|
| 254 |
+
|
| 255 |
+
We will have a paper out soon, but until then, please cite:
|
| 256 |
+
|
| 257 |
+
```
|
| 258 |
+
@online{SHP,
|
| 259 |
+
author = {Ethayarajh, Kawin and Zhang, Heidi and Wang, Yizhong and Jurafsky, Dan},
|
| 260 |
+
title = {Stanford Human Preferences Dataset},
|
| 261 |
+
year = {2023},
|
| 262 |
+
url = {https://huggingface.co/datasets/stanfordnlp/SHP}
|
| 263 |
+
}
|
| 264 |
+
```
|
huggingface_dataset/Dataset_Card/turkish_shrinked_ner.md
ADDED
|
@@ -0,0 +1,260 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
annotations_creators:
|
| 3 |
+
- machine-generated
|
| 4 |
+
language_creators:
|
| 5 |
+
- expert-generated
|
| 6 |
+
language:
|
| 7 |
+
- tr
|
| 8 |
+
license:
|
| 9 |
+
- cc-by-4.0
|
| 10 |
+
multilinguality:
|
| 11 |
+
- monolingual
|
| 12 |
+
size_categories:
|
| 13 |
+
- 100K<n<1M
|
| 14 |
+
source_datasets:
|
| 15 |
+
- extended|other-turkish_ner
|
| 16 |
+
task_categories:
|
| 17 |
+
- token-classification
|
| 18 |
+
task_ids:
|
| 19 |
+
- named-entity-recognition
|
| 20 |
+
pretty_name: TurkishShrinkedNer
|
| 21 |
+
dataset_info:
|
| 22 |
+
features:
|
| 23 |
+
- name: id
|
| 24 |
+
dtype: string
|
| 25 |
+
- name: tokens
|
| 26 |
+
sequence: string
|
| 27 |
+
- name: ner_tags
|
| 28 |
+
sequence:
|
| 29 |
+
class_label:
|
| 30 |
+
names:
|
| 31 |
+
'0': O
|
| 32 |
+
'1': B-academic
|
| 33 |
+
'2': I-academic
|
| 34 |
+
'3': B-academic_person
|
| 35 |
+
'4': I-academic_person
|
| 36 |
+
'5': B-aircraft
|
| 37 |
+
'6': I-aircraft
|
| 38 |
+
'7': B-album_person
|
| 39 |
+
'8': I-album_person
|
| 40 |
+
'9': B-anatomy
|
| 41 |
+
'10': I-anatomy
|
| 42 |
+
'11': B-animal
|
| 43 |
+
'12': I-animal
|
| 44 |
+
'13': B-architect_person
|
| 45 |
+
'14': I-architect_person
|
| 46 |
+
'15': B-capital
|
| 47 |
+
'16': I-capital
|
| 48 |
+
'17': B-chemical
|
| 49 |
+
'18': I-chemical
|
| 50 |
+
'19': B-clothes
|
| 51 |
+
'20': I-clothes
|
| 52 |
+
'21': B-country
|
| 53 |
+
'22': I-country
|
| 54 |
+
'23': B-culture
|
| 55 |
+
'24': I-culture
|
| 56 |
+
'25': B-currency
|
| 57 |
+
'26': I-currency
|
| 58 |
+
'27': B-date
|
| 59 |
+
'28': I-date
|
| 60 |
+
'29': B-food
|
| 61 |
+
'30': I-food
|
| 62 |
+
'31': B-genre
|
| 63 |
+
'32': I-genre
|
| 64 |
+
'33': B-government
|
| 65 |
+
'34': I-government
|
| 66 |
+
'35': B-government_person
|
| 67 |
+
'36': I-government_person
|
| 68 |
+
'37': B-language
|
| 69 |
+
'38': I-language
|
| 70 |
+
'39': B-location
|
| 71 |
+
'40': I-location
|
| 72 |
+
'41': B-material
|
| 73 |
+
'42': I-material
|
| 74 |
+
'43': B-measure
|
| 75 |
+
'44': I-measure
|
| 76 |
+
'45': B-medical
|
| 77 |
+
'46': I-medical
|
| 78 |
+
'47': B-military
|
| 79 |
+
'48': I-military
|
| 80 |
+
'49': B-military_person
|
| 81 |
+
'50': I-military_person
|
| 82 |
+
'51': B-nation
|
| 83 |
+
'52': I-nation
|
| 84 |
+
'53': B-newspaper
|
| 85 |
+
'54': I-newspaper
|
| 86 |
+
'55': B-organization
|
| 87 |
+
'56': I-organization
|
| 88 |
+
'57': B-organization_person
|
| 89 |
+
'58': I-organization_person
|
| 90 |
+
'59': B-person
|
| 91 |
+
'60': I-person
|
| 92 |
+
'61': B-production_art_music
|
| 93 |
+
'62': I-production_art_music
|
| 94 |
+
'63': B-production_art_music_person
|
| 95 |
+
'64': I-production_art_music_person
|
| 96 |
+
'65': B-quantity
|
| 97 |
+
'66': I-quantity
|
| 98 |
+
'67': B-religion
|
| 99 |
+
'68': I-religion
|
| 100 |
+
'69': B-science
|
| 101 |
+
'70': I-science
|
| 102 |
+
'71': B-shape
|
| 103 |
+
'72': I-shape
|
| 104 |
+
'73': B-ship
|
| 105 |
+
'74': I-ship
|
| 106 |
+
'75': B-software
|
| 107 |
+
'76': I-software
|
| 108 |
+
'77': B-space
|
| 109 |
+
'78': I-space
|
| 110 |
+
'79': B-space_person
|
| 111 |
+
'80': I-space_person
|
| 112 |
+
'81': B-sport
|
| 113 |
+
'82': I-sport
|
| 114 |
+
'83': B-sport_name
|
| 115 |
+
'84': I-sport_name
|
| 116 |
+
'85': B-sport_person
|
| 117 |
+
'86': I-sport_person
|
| 118 |
+
'87': B-structure
|
| 119 |
+
'88': I-structure
|
| 120 |
+
'89': B-subject
|
| 121 |
+
'90': I-subject
|
| 122 |
+
'91': B-tech
|
| 123 |
+
'92': I-tech
|
| 124 |
+
'93': B-train
|
| 125 |
+
'94': I-train
|
| 126 |
+
'95': B-vehicle
|
| 127 |
+
'96': I-vehicle
|
| 128 |
+
splits:
|
| 129 |
+
- name: train
|
| 130 |
+
num_bytes: 200728389
|
| 131 |
+
num_examples: 614515
|
| 132 |
+
download_size: 0
|
| 133 |
+
dataset_size: 200728389
|
| 134 |
+
---
|
| 135 |
+
|
| 136 |
+
# Dataset Card for turkish_shrinked_ner
|
| 137 |
+
|
| 138 |
+
## Table of Contents
|
| 139 |
+
- [Dataset Description](#dataset-description)
|
| 140 |
+
- [Dataset Summary](#dataset-summary)
|
| 141 |
+
- [Supported Tasks and Leaderboards](#supported-tasks-and-leaderboards)
|
| 142 |
+
- [Languages](#languages)
|
| 143 |
+
- [Dataset Structure](#dataset-structure)
|
| 144 |
+
- [Data Instances](#data-instances)
|
| 145 |
+
- [Data Fields](#data-fields)
|
| 146 |
+
- [Data Splits](#data-splits)
|
| 147 |
+
- [Dataset Creation](#dataset-creation)
|
| 148 |
+
- [Curation Rationale](#curation-rationale)
|
| 149 |
+
- [Source Data](#source-data)
|
| 150 |
+
- [Annotations](#annotations)
|
| 151 |
+
- [Personal and Sensitive Information](#personal-and-sensitive-information)
|
| 152 |
+
- [Considerations for Using the Data](#considerations-for-using-the-data)
|
| 153 |
+
- [Social Impact of Dataset](#social-impact-of-dataset)
|
| 154 |
+
- [Discussion of Biases](#discussion-of-biases)
|
| 155 |
+
- [Other Known Limitations](#other-known-limitations)
|
| 156 |
+
- [Additional Information](#additional-information)
|
| 157 |
+
- [Dataset Curators](#dataset-curators)
|
| 158 |
+
- [Licensing Information](#licensing-information)
|
| 159 |
+
- [Citation Information](#citation-information)
|
| 160 |
+
- [Contributions](#contributions)
|
| 161 |
+
|
| 162 |
+
## Dataset Description
|
| 163 |
+
|
| 164 |
+
- **Homepage:** https://www.kaggle.com/behcetsenturk/shrinked-twnertc-turkish-ner-data-by-kuzgunlar
|
| 165 |
+
- **Repository:** [Needs More Information]
|
| 166 |
+
- **Paper:** [Needs More Information]
|
| 167 |
+
- **Leaderboard:** [Needs More Information]
|
| 168 |
+
- **Point of Contact:** https://www.kaggle.com/behcetsenturk
|
| 169 |
+
|
| 170 |
+
### Dataset Summary
|
| 171 |
+
|
| 172 |
+
Shrinked processed version (48 entity type) of the turkish_ner.
|
| 173 |
+
|
| 174 |
+
Original turkish_ner dataset: Automatically annotated Turkish corpus for named entity recognition and text categorization using large-scale gazetteers. The constructed gazetteers contains approximately 300K entities with thousands of fine-grained entity types under 25 different domains.
|
| 175 |
+
|
| 176 |
+
Shrinked entity types are: academic, academic_person, aircraft, album_person, anatomy, animal, architect_person, capital, chemical, clothes, country, culture, currency, date, food, genre, government, government_person, language, location, material, measure, medical, military, military_person, nation, newspaper, organization, organization_person, person, production_art_music, production_art_music_person, quantity, religion, science, shape, ship, software, space, space_person, sport, sport_name, sport_person, structure, subject, tech, train, vehicle
|
| 177 |
+
|
| 178 |
+
### Supported Tasks and Leaderboards
|
| 179 |
+
|
| 180 |
+
[Needs More Information]
|
| 181 |
+
|
| 182 |
+
### Languages
|
| 183 |
+
|
| 184 |
+
Turkish
|
| 185 |
+
|
| 186 |
+
## Dataset Structure
|
| 187 |
+
|
| 188 |
+
### Data Instances
|
| 189 |
+
|
| 190 |
+
[Needs More Information]
|
| 191 |
+
|
| 192 |
+
### Data Fields
|
| 193 |
+
|
| 194 |
+
[Needs More Information]
|
| 195 |
+
|
| 196 |
+
### Data Splits
|
| 197 |
+
|
| 198 |
+
There's only the training set.
|
| 199 |
+
|
| 200 |
+
## Dataset Creation
|
| 201 |
+
|
| 202 |
+
### Curation Rationale
|
| 203 |
+
|
| 204 |
+
[Needs More Information]
|
| 205 |
+
|
| 206 |
+
### Source Data
|
| 207 |
+
|
| 208 |
+
#### Initial Data Collection and Normalization
|
| 209 |
+
|
| 210 |
+
[Needs More Information]
|
| 211 |
+
|
| 212 |
+
#### Who are the source language producers?
|
| 213 |
+
|
| 214 |
+
[Needs More Information]
|
| 215 |
+
|
| 216 |
+
### Annotations
|
| 217 |
+
|
| 218 |
+
#### Annotation process
|
| 219 |
+
|
| 220 |
+
[Needs More Information]
|
| 221 |
+
|
| 222 |
+
#### Who are the annotators?
|
| 223 |
+
|
| 224 |
+
[Needs More Information]
|
| 225 |
+
|
| 226 |
+
### Personal and Sensitive Information
|
| 227 |
+
|
| 228 |
+
[Needs More Information]
|
| 229 |
+
|
| 230 |
+
## Considerations for Using the Data
|
| 231 |
+
|
| 232 |
+
### Social Impact of Dataset
|
| 233 |
+
|
| 234 |
+
[Needs More Information]
|
| 235 |
+
|
| 236 |
+
### Discussion of Biases
|
| 237 |
+
|
| 238 |
+
[Needs More Information]
|
| 239 |
+
|
| 240 |
+
### Other Known Limitations
|
| 241 |
+
|
| 242 |
+
[Needs More Information]
|
| 243 |
+
|
| 244 |
+
## Additional Information
|
| 245 |
+
|
| 246 |
+
### Dataset Curators
|
| 247 |
+
|
| 248 |
+
Behcet Senturk
|
| 249 |
+
|
| 250 |
+
### Licensing Information
|
| 251 |
+
|
| 252 |
+
Creative Commons Attribution 4.0 International
|
| 253 |
+
|
| 254 |
+
### Citation Information
|
| 255 |
+
|
| 256 |
+
[Needs More Information]
|
| 257 |
+
|
| 258 |
+
### Contributions
|
| 259 |
+
|
| 260 |
+
Thanks to [@bhctsntrk](https://github.com/bhctsntrk) for adding this dataset.
|