Scalable Data Ablations
Collection
Datasets and models for EMNLP paper "Scalable Data Ablation Approximations for Language Models through Modular Training and Merging"
•
34 items
•
Updated
•
1
added
stringlengths 24
24
| attributes
dict | created
stringlengths 24
24
| id
stringlengths 40
40
| metadata
dict | source
stringclasses 1
value | text
stringlengths 0
10.5M
| version
stringclasses 1
value |
|---|---|---|---|---|---|---|---|
2023-04-25T05:49:46.922Z | {
"bff_duplicate_paragraph_spans_decontamination": []
} | 2006-01-09T00:00:00.000Z | 600f7d0e70e779b5c95464411000c5998ea252ba | {"gutenberg_metadata_available":true,"issued_or_updated_available":true,"length":53715,"pipeline":"g(...TRUNCATED) | gutenberg | "THE PRISONER OF ZENDA\n\nby Anthony Hope\n\n\nCONTENTS\n\n 1 The Rassendylls--With a Word on t(...TRUNCATED) | v0 |
2023-04-25T05:49:46.973Z | {
"bff_duplicate_paragraph_spans_decontamination": []
} | 2008-06-25T00:00:00.000Z | b90ec5c54b09987a559cd2129e04c1105ee4c1cc | {"gutenberg_metadata_available":true,"issued_or_updated_available":true,"length":5176,"pipeline":"gu(...TRUNCATED) | gutenberg | "THE HUNTING OF THE SNARK\n\n an Agony in Eight Fits\n\n\n (...TRUNCATED) | v0 |
2023-04-25T05:49:46.999Z | {
"bff_duplicate_paragraph_spans_decontamination": []
} | 1971-12-01T00:00:00.000Z | e243c6571caed82824c88cf48abfd019634c6db3 | {"gutenberg_metadata_available":true,"issued_or_updated_available":true,"length":6534,"pipeline":"gu(...TRUNCATED) | gutenberg | "This is a retranscription of one of the first Project\nGutenberg Etexts, offically dated November 2(...TRUNCATED) | v0 |
2023-04-25T05:49:47.056Z | {
"bff_duplicate_paragraph_spans_decontamination": []
} | 2008-06-25T00:00:00.000Z | 569bf2fd58a30f84bc610263289ffc2f73869f07 | {"gutenberg_metadata_available":true,"issued_or_updated_available":true,"length":29751,"pipeline":"g(...TRUNCATED) | gutenberg | "[Illustration]\n\n\nTHROUGH THE LOOKING-GLASS\n\nAnd What Alice Found There\n\nBy Lewis Carroll\n\n(...TRUNCATED) | v0 |
2023-04-25T05:49:47.110Z | {
"bff_duplicate_paragraph_spans_decontamination": []
} | 1993-04-01T00:00:00.000Z | 4cfb0517ed685fcfa9614f4698b2e7b84c14394f | {"gutenberg_metadata_available":true,"issued_or_updated_available":true,"length":2593,"pipeline":"gu(...TRUNCATED) | gutenberg | "The Number \"e\"\n\nBelow is the value of 'e' to about 100,000 places, computed on the NCSA\nCray Y(...TRUNCATED) | v0 |
2023-04-25T05:49:47.934Z | {
"bff_duplicate_paragraph_spans_decontamination": []
} | 1993-05-01T00:00:00.000Z | 370bf6d8aca0e86dcf412278fca5578b9a8a76a6 | {"gutenberg_metadata_available":true,"issued_or_updated_available":true,"length":101351,"pipeline":"(...TRUNCATED) | gutenberg | "[1 is not a prime by many definitions of prime]\n\n1\n2\n3\n5\n7\n11\n13\n17\n19\n23\n29\n31\n37\n4(...TRUNCATED) | v0 |
2023-04-25T05:49:48.253Z | {
"bff_duplicate_paragraph_spans_decontamination": []
} | 1993-05-01T00:00:00.000Z | a0d77923cd120285375f0afd739ecbc907ae990f | {"gutenberg_metadata_available":true,"issued_or_updated_available":true,"length":31089,"pipeline":"g(...TRUNCATED) | gutenberg | "The dawn of amateur radio in the U.K. and Greece: a personal view\n\n\n Norman F. Joly.\n\n COPYR(...TRUNCATED) | v0 |
2023-04-25T05:49:48.415Z | {
"bff_duplicate_paragraph_spans_decontamination": []
} | 1993-09-01T00:00:00.000Z | c457f9428564eb4a8b7dd3ad2cebb64154cc09fd | {"gutenberg_metadata_available":true,"issued_or_updated_available":true,"length":89793,"pipeline":"r(...TRUNCATED) | gutenberg | "THE ONLINE WORLD\n\nThe Online World book's text on paper, disk and in any other\nelectronic form i(...TRUNCATED) | v0 |
2023-04-25T05:49:48.608Z | {
"bff_duplicate_paragraph_spans_decontamination": []
} | 1993-09-01T00:00:00.000Z | 6454d606db9b5110b377e674a7b2958b428c33cb | {"gutenberg_metadata_available":true,"issued_or_updated_available":true,"length":91499,"pipeline":"g(...TRUNCATED) | gutenberg | "[Illustration]\n\n\nFrom the Earth to the Moon\n\nby Jules Verne\n\n\nContents: From the Earth to t(...TRUNCATED) | v0 |
2023-04-25T05:49:48.738Z | {
"bff_duplicate_paragraph_spans_decontamination": []
} | 2008-06-25T00:00:00.000Z | b3d6f3c9802e875b627cb548e3cde89cf4b683dd | {"gutenberg_metadata_available":true,"issued_or_updated_available":true,"length":47267,"pipeline":"g(...TRUNCATED) | gutenberg | "Peter Pan\n\n[PETER AND WENDY]\n\nby J. M. Barrie [James Matthew Barrie]\n\nA Millennium Fulcrum Ed(...TRUNCATED) | v0 |
End of preview. Expand
in Dataset Viewer.
Dataset Card for Dataset Name
Pre-training corpus for seed models in "Scalable Data Ablation Approximations for Language Models through Modular Training and Merging", to be presented at EMNLP 2024.
Dataset Details
Dataset Description
- Curated by: [More Information Needed]
- Funded by [optional]: [More Information Needed]
- Shared by [optional]: [More Information Needed]
- Language(s) (NLP): [More Information Needed]
- License: [More Information Needed]
Dataset Sources [optional]
- Repository: [More Information Needed]
- Paper [optional]: [More Information Needed]
- Demo [optional]: [More Information Needed]
Uses
Direct Use
[More Information Needed]
Out-of-Scope Use
[More Information Needed]
Dataset Structure
[More Information Needed]
Dataset Creation
Curation Rationale
[More Information Needed]
Source Data
Data Collection and Processing
[More Information Needed]
Who are the source data producers?
[More Information Needed]
Annotations [optional]
Annotation process
[More Information Needed]
Who are the annotators?
[More Information Needed]
Personal and Sensitive Information
[More Information Needed]
Bias, Risks, and Limitations
[More Information Needed]
Recommendations
Users should be made aware of the risks, biases and limitations of the dataset. More information needed for further recommendations.
Citation [optional]
BibTeX:
[More Information Needed]
APA:
[More Information Needed]
Glossary [optional]
[More Information Needed]
More Information [optional]
[More Information Needed]
Dataset Card Authors [optional]
[More Information Needed]
Dataset Card Contact
[More Information Needed]
- Downloads last month
- 127
Size of the auto-converted Parquet files (First 5GB per split):
6.02 GB
Number of rows (First 5GB per split):
1,034,098
Estimated number of rows:
3,446,798