Datasets:

VelkroLM
/

african-languages-corpus

Error code:   StreamingRowsError
Exception:    CastError
Message:      Couldn't cast
record_id: string
language: string
language_code: string
script: string
text: string
source_dataset: string
source_revision: string
source_url: string
source_record_id: string
license: string
license_tier: string
provenance: struct<u: string, f: string, crawl_id: string, ts: timestamp[s], lang: list<item: string>, prob: lis (... 16 chars omitted)
  child 0, u: string
  child 1, f: string
  child 2, crawl_id: string
  child 3, ts: timestamp[s]
  child 4, lang: list<item: string>
      child 0, item: string
  child 5, prob: list<item: double>
      child 0, item: double
quality: struct<text_chars: int64, lang_prob: double>
  child 0, text_chars: int64
  child 1, lang_prob: double
processing_version: string
sample: list<item: struct<text: string, url: string, lang: string>>
  child 0, item: struct<text: string, url: string, lang: string>
      child 0, text: string
      child 1, url: string
      child 2, lang: string
kept_rows: int64
drop_reasons: struct<>
downloaded_bytes: int64
elapsed_seconds: double
sample_size: int64
output_sha256: string
output_path: string
status: string
total_rows: int64
dropped_rows: int64
to
{'source_url': Value('string'), 'language': Value('string'), 'status': Value('string'), 'downloaded_bytes': Value('int64'), 'total_rows': Value('int64'), 'kept_rows': Value('int64'), 'dropped_rows': Value('int64'), 'drop_reasons': {}, 'sample_size': Value('int64'), 'sample': List({'text': Value('string'), 'url': Value('string'), 'lang': Value('string')}), 'output_path': Value('string'), 'output_sha256': Value('string'), 'elapsed_seconds': Value('float64'), 'license': Value('string'), 'processing_version': Value('string')}
because column names don't match
Traceback:    Traceback (most recent call last):
                File "/src/services/worker/src/worker/utils.py", line 147, in get_rows_or_raise
                  return get_rows(
                      dataset=dataset,
                  ...<4 lines>...
                      column_names=column_names,
                  )
                File "/src/libs/libcommon/src/libcommon/utils.py", line 272, in decorator
                  return func(*args, **kwargs)
                File "/src/services/worker/src/worker/utils.py", line 127, in get_rows
                  rows_plus_one = list(itertools.islice(safe_iter(ds, dataset=dataset), rows_max_number + 1))
                File "/src/services/worker/src/worker/utils.py", line 483, in safe_iter
                  yield from ds.decode(False) if ds.features else ds
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2840, in __iter__
                  for key, example in ex_iterable:
                                      ^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2373, in __iter__
                  for key, pa_table in self._iter_arrow():
                                       ~~~~~~~~~~~~~~~~^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2398, in _iter_arrow
                  for key, pa_table in self.ex_iterable._iter_arrow():
                                       ~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 536, in _iter_arrow
                  for key, pa_table in iterator:
                                       ^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 419, in _iter_arrow
                  for key, pa_table in self.generate_tables_fn(**gen_kwags):
                                       ~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 343, in _generate_tables
                  self._cast_table(pa_table, json_field_paths=json_field_paths),
                  ~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 132, in _cast_table
                  pa_table = table_cast(pa_table, self.info.features.arrow_schema)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2378, in table_cast
                  return cast_table_to_schema(table, schema)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2306, in cast_table_to_schema
                  raise CastError(
                  ...<3 lines>...
                  )
              datasets.table.CastError: Couldn't cast
              record_id: string
              language: string
              language_code: string
              script: string
              text: string
              source_dataset: string
              source_revision: string
              source_url: string
              source_record_id: string
              license: string
              license_tier: string
              provenance: struct<u: string, f: string, crawl_id: string, ts: timestamp[s], lang: list<item: string>, prob: lis (... 16 chars omitted)
                child 0, u: string
                child 1, f: string
                child 2, crawl_id: string
                child 3, ts: timestamp[s]
                child 4, lang: list<item: string>
                    child 0, item: string
                child 5, prob: list<item: double>
                    child 0, item: double
              quality: struct<text_chars: int64, lang_prob: double>
                child 0, text_chars: int64
                child 1, lang_prob: double
              processing_version: string
              sample: list<item: struct<text: string, url: string, lang: string>>
                child 0, item: struct<text: string, url: string, lang: string>
                    child 0, text: string
                    child 1, url: string
                    child 2, lang: string
              kept_rows: int64
              drop_reasons: struct<>
              downloaded_bytes: int64
              elapsed_seconds: double
              sample_size: int64
              output_sha256: string
              output_path: string
              status: string
              total_rows: int64
              dropped_rows: int64
              to
              {'source_url': Value('string'), 'language': Value('string'), 'status': Value('string'), 'downloaded_bytes': Value('int64'), 'total_rows': Value('int64'), 'kept_rows': Value('int64'), 'dropped_rows': Value('int64'), 'drop_reasons': {}, 'sample_size': Value('int64'), 'sample': List({'text': Value('string'), 'url': Value('string'), 'lang': Value('string')}), 'output_path': Value('string'), 'output_sha256': Value('string'), 'elapsed_seconds': Value('float64'), 'license': Value('string'), 'processing_version': Value('string')}
              because column names don't match

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

VelkroLM African Languages Corpus

This repository contains a filtered, provenance-preserving text corpus derived from the language-specific HPLT v3.0 shards discovered at hplt-project.org/datasets/v3.0. It is organized by language and source shard so researchers can load only the languages they need. The upstream HPLT project describes its v3.0 data as multilingual web-corpus material; the upstream release, source metadata, and terms remain authoritative.

This publication is deliberately honest about scale. The completed HPLT subset is 3,382,250 retained records and 3.59 GiB of compressed JSONL, not 500 GB. The 500 GB aspiration can only be approached by adding additional legally reusable sources such as larger language shards, parallel corpora, and speech/audio collections after separate license and quality review. The corpus is not inflated with duplicate mirrors or synthetic copies.

Language coverage

Language	Code	Region	Shards	Retained records	Compressed output	Retention
Hausa	`hau`	Nigeria	6	737,138	773.3 MiB	99.10%
Yoruba	`yor`	Nigeria	6	163,086	187.9 MiB	95.23%
Igbo	`ibo`	Nigeria	5	171,474	203.8 MiB	99.21%
Fulfulde	`fuv`	Nigeria / Sahel	6	9,811	12.0 MiB	98.39%
Kanuri	`knc`	Nigeria / Lake Chad	10	1,498	2.1 MiB	65.16%
Amharic	`amh`	Africa	6	568,631	1020.5 MiB	99.54%
Somali	`som`	Africa	6	1,415,029	1076.7 MiB	99.38%
Tigrinya	`tir`	Africa	6	67,424	121.9 MiB	99.70%
Wolof	`wol`	Africa	5	4,512	6.5 MiB	89.24%
Lingala	`lin`	Africa	6	13,280	25.9 MiB	97.93%
Luganda	`lug`	Africa	6	49,189	40.5 MiB	99.17%
Shona	`sna`	Africa	6	181,178	199.8 MiB	99.00%

File layout

Each language directory contains compressed JSONL output and a matching audit file for every processed source shard.

Hausa/10_1.jsonl.zst
Hausa/10_1.audit.json
Yoruba/8_1.jsonl.zst
...
aggregate_stats.json
manifest.json
README.md

Each retained JSON object contains record_id, language, language_code, script, text, source_dataset, source_revision, source_url, source_record_id, license, license_tier, provenance, quality, and processing_version.

Filtering policy

The pipeline used one active download worker and streamed each Zstandard shard without loading the corpus into memory. It removed empty or malformed rows, obvious credential or spam artifacts, excessive URL spam, low-confidence language rows when a confidence field was provided, and records that violated the minimum quality policy. It did not delete embedded English names or code-switching from otherwise valid African-language text. Parallel corpora and bilingual records must be handled as aligned data and are not made monolingual by deleting one side.

The filter is conservative rather than a claim of perfect language identification. Every shard includes a 200-record seeded reservoir sample in its audit JSON. Random samples should be inspected by downstream users before training, especially because web corpora may contain copyrighted material, duplicated articles, named entities, translated pages, and domain-specific text.

Provenance and licensing

All records in this repository identify the HPLT v3.0 source family and direct source shard URL. The HPLT source page and its accompanying metadata must be cited together with this derivative. The repository metadata uses CC0 for this HPLT-derived subset based on the HPLT release terms; this does not erase rights or restrictions that may attach to the underlying web documents. Users are responsible for complying with applicable law, upstream notices, privacy obligations, and the terms of any source material.

The complete processing manifest is in manifest.json, aggregate counts are in aggregate_stats.json, and each *.audit.json file records source URL, counts, drop reasons, random sample, output checksum, and processing version.

Loading example

from datasets import load_dataset

ds = load_dataset(
    "VelkroLM/african-languages-corpus",
    data_files={"train": "Hausa/*.jsonl.zst"},
    streaming=True,
)
for row in ds["train"].take(3):
    print(row["text"])

To load another language, replace Hausa with Yoruba, Igbo, Amharic, Somali, or another directory listed above. The compressed JSONL files are intentionally shardable for streaming.

Reproducibility

Processing version: 0.2.0.
Audit manifest: manifest.json and the local append-only source manifest.
Random validation: 200 seeded reservoir samples per shard where the shard contained enough records.
Download policy: resumable HTTP Range requests, one active source connection, checksum verification, and deletion of local raw staging only after successful output verification.
Publication policy: source originals are not overwritten; filtered derivatives are separate repositories or directories.

Citation

Please cite both the derivative repository and the HPLT project. For a specific record, retain its source_url, source_dataset, and source_record_id fields.

References

Build date: 2026-08-22

Downloads last month: 109

Total file size:

3.86 GB

Collection including VelkroLM/african-languages-corpus

VelkroLM African Languages Corpus

Collection

Audited African-language corpora and source catalog. • 4 items • Updated 5 days ago