Dataset Preview
The full dataset viewer is not available (click to read why). Only showing a preview of the rows.
The dataset generation failed
Error code: DatasetGenerationError
Exception: CastError
Message: Couldn't cast
text: string
meta: struct<redpajama_set_name: string>
child 0, redpajama_set_name: string
__index_level_0__: int64
to
{'text': Value('string'), 'meta': {'redpajama_set_name': Value('string')}}
because column names don't match
Traceback: Traceback (most recent call last):
File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 1872, in _prepare_split_single
for key, table in generator:
^^^^^^^^^
File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 609, in wrapped
for item in generator(*args, **kwargs):
^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/packaged_modules/json/json.py", line 265, in _generate_tables
self._cast_table(pa_table, json_field_paths=json_field_paths),
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/packaged_modules/json/json.py", line 120, in _cast_table
pa_table = table_cast(pa_table, self.info.features.arrow_schema)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/table.py", line 2272, in table_cast
return cast_table_to_schema(table, schema)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/table.py", line 2218, in cast_table_to_schema
raise CastError(
datasets.table.CastError: Couldn't cast
text: string
meta: struct<redpajama_set_name: string>
child 0, redpajama_set_name: string
__index_level_0__: int64
to
{'text': Value('string'), 'meta': {'redpajama_set_name': Value('string')}}
because column names don't match
The above exception was the direct cause of the following exception:
Traceback (most recent call last):
File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 1342, in compute_config_parquet_and_info_response
parquet_operations, partial, estimated_dataset_info = stream_convert_to_parquet(
^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 907, in stream_convert_to_parquet
builder._prepare_split(split_generator=splits_generators[split], file_format="parquet")
File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 1739, in _prepare_split
for job_id, done, content in self._prepare_split_single(
^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.12/site-packages/datasets/builder.py", line 1922, in _prepare_split_single
raise DatasetGenerationError("An error occurred while generating the dataset") from e
datasets.exceptions.DatasetGenerationError: An error occurred while generating the datasetNeed help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
text string | meta dict |
|---|---|
\section{Introduction}
Spallation reactions, i.e. proton-induced reactions on heavy targets at a few hundred MeV, have been the subject of many studies since 1950. They are known to be a valuable tool for the study of the de-excitation of hot nuclei because, contrarily to reactions between heavy ions, they lead to th... | {
"redpajama_set_name": "RedPajamaArXiv"
} |
\section{Introduction}
The uniqueness problem of meromorphic mappings under a condition
on the inverse images of divisors was first studied by Nevanlinna \cite{Ne}. He
showed that for two nonconstant meromorphic functions $f$ and $g$ on the
complex plane $\mathbb{C}$, if they have the same inverse images for five
disti... | {
"redpajama_set_name": "RedPajamaArXiv"
} |
\section{Introduction}
The entanglement entropy of a subregion in a relativistic quantum field theory is UV divergent because of short range correlations across the entangling surface. This is evident from the continuum limit of the earliest calculations on a lattice \cite{sorkin,srednicki}, or in calculations using t... | {
"redpajama_set_name": "RedPajamaArXiv"
} |
\section{#1}}
\baselineskip=20pt
\newfont{\elevenmib}{cmmib10 scaled\magstep1}
\newcommand{\preprint}{
\begin{flushleft}
\end{flushleft}\vspace{-1.3cm}
\begin{flushright}\normalsize
\end{flushright}}
\newcommand{\Title}[1]{{\baselineskip=26pt
\begin{center} \Large \bf #1 \\ \ \\ \end{center... | {
"redpajama_set_name": "RedPajamaArXiv"
} |
"\\section{Examples of Hard Negatives}\n\nFig.~\\ref{fig:sample_outputs} compares the outputs of \\e(...TRUNCATED) | {
"redpajama_set_name": "RedPajamaArXiv"
} |
"\\section{Introduction}\n\\label{intro}\nUnknown intent detection is a realistic and challenging ta(...TRUNCATED) | {
"redpajama_set_name": "RedPajamaArXiv"
} |
"\\section{Introduction}\n\n\nRecently, the advancement of electric battery technology pushes forwar(...TRUNCATED) | {
"redpajama_set_name": "RedPajamaArXiv"
} |
"\\section{Introduction}\n\nWhenever we measure anything using a particular number system, the\ncorr(...TRUNCATED) | {
"redpajama_set_name": "RedPajamaArXiv"
} |
"\\section{Introduction}\nWithin the next few years, powerful lasers may be able to realise\nintensi(...TRUNCATED) | {
"redpajama_set_name": "RedPajamaArXiv"
} |
"\\section{Introduction} \n\nThe field of quantum computation has seen fervent activity in the deve(...TRUNCATED) | {
"redpajama_set_name": "RedPajamaArXiv"
} |
End of preview.
SlimPajama-6B & SlimPajama-30B
Pre-training data sampled from cerebras/SlimPajama-627B, at two scales: ~6B tokens and ~30B tokens. The data is formatted in JSONL and is compatible with LLaMA-Factory training pipelines.
Repository Structure
├── train/
│ ├── 6B/ # ~6B tokens, 7 JSONL files by source
│ │ ├── RedPajamaCommonCrawl-6B.jsonl
│ │ ├── RedPajamaC4-6B.jsonl
│ │ ├── RedPajamaGithub-6B.jsonl
│ │ ├── RedPajamaBook-6B.jsonl
│ │ ├── RedPajamaArXiv-6B.jsonl
│ │ ├── RedPajamaWikipedia-6B.jsonl
│ │ └── RedPajamaStackExchange-6B.jsonl
│ └── 30B/ # ~30B tokens, 7 JSONL files by source
│ ├── RedPajamaCommonCrawl-30B.jsonl
│ ├── RedPajamaC4-30B.jsonl
│ ├── RedPajamaGithub-30B.jsonl
│ ├── RedPajamaBook-30B.jsonl
│ ├── RedPajamaArXiv-30B.jsonl
│ ├── RedPajamaWikipedia-30B.jsonl
│ └── RedPajamaStackExchange-30B.jsonl
└── val/
└── validation.jsonl # validation set
Each JSONL record has the following schema:
{"text": "...", "meta": {"redpajama_set_name": "RedPajamaC4"}, "__index_level_0__": 12345}
Data Source
Both subsets originate from cerebras/SlimPajama-627B, a 627B-token cleaned and deduplicated version of RedPajama. SlimPajama-627B contains data from 7 sources: CommonCrawl, C4, GitHub, Books, ArXiv, Wikipedia, and StackExchange.
SlimPajama-6B
Source & Sampling
Obtained from DKYoon/SlimPajama-6B.
Since the original data was shuffled before chunking, only train/chunk1 (of 10 total) was downloaded and further sampled at 10%, resulting in roughly 6B tokens.
- Training set: 5,489,000 records
- Validation set: 9,347 records
SlimPajama-30B
Source & Sampling
Sampled directly from cerebras/SlimPajama-627B train/chunk1. The sampling targets ~30B tokens while preserving the source proportions of SlimPajama-6B.
Sampling procedure:
- Load all records from
train/chunk1and tokenize with the Qwen-2.5 tokenizer. - For each source, compute a target token count based on the SlimPajama-6B domain weights.
- Randomly shuffle records within each source (seed=42) and greedily select records until the target token count is reached.
Token Statistics (Qwen-2.5 Tokenizer)
| Data Source | Records | Tokens | Proportion % |
|---|---|---|---|
| RedPajamaCommonCrawl | 8,779,572 | 14,973,812,443 | 54.10% |
| RedPajamaC4 | 16,214,931 | 7,943,593,272 | 28.70% |
| RedPajamaGithub | 952,170 | 1,162,483,837 | 4.20% |
| RedPajamaBook | 8,145 | 1,024,205,301 | 3.70% |
| RedPajamaArXiv | 50,277 | 941,068,936 | 3.40% |
| RedPajamaWikipedia | 1,026,174 | 858,018,818 | 3.10% |
| RedPajamaStackExchange | 1,252,229 | 774,984,699 | 2.80% |
| Total | 28,283,498 | 27,678,167,306 | 100% |
Acknowledgement
This dataset is built upon cerebras/SlimPajama-627B and DKYoon/SlimPajama-6B. We thank the authors for their awesome work.
- Downloads last month
- 30