Hugging Face's logo

Datasets:
issai
/
Base_Transfer

Modalities:
Tabular
Text
Formats:
parquet
Languages:
Kazakh
Size:
10M - 100M
Libraries:
Datasets
Dask
Dataset card Data Studio Files Community
1
Base_Transfer / README.md
asspunchman's picture
asspunchman
Update README.md
50da6c6 verified
preview code
|
raw
history blame contribute delete
14.9 kB
metadata
language:
  - kk
configs:
  - config_name: all
    data_files:
      - split: train
        path: data/*.parquet
    default: true
  - config_name: doc_tq_kaz
    data_files:
      - split: train
        path: data/doc_tq_kaz-00000-of-00001.parquet
  - config_name: doc_tq_kaz_2
    data_files:
      - split: train
        path: data/doc_tq_kaz_2-00000-of-00001.parquet
  - config_name: kk_wikipedia
    data_files:
      - split: train
        path: data/kk_wikipedia_cleaned_v2_split-00000-of-00001.parquet
  - config_name: kk_news
    data_files:
      - split: train
        path: data/kk_news_cleaned_v2_split-00000-of-00001.parquet
  - config_name: kk_news180924
    data_files:
      - split: train
        path: data/kk_news180924_cleaned_v2_split-00000-of-00001.parquet
  - config_name: kk_fineweb
    data_files:
      - split: train
        path: data/kk_fineweb_cleaned_v2_split-00000-of-00001.parquet
  - config_name: kk_ocr_fixed
    data_files:
      - split: train
        path: data/kk_ocr_fixed_cleaned_v2_split-00000-of-00001.parquet
  - config_name: kk_tilqazyna
    data_files:
      - split: train
        path: data/kk_tilqazyna_cleaned_v2_split-00000-of-00001.parquet
  - config_name: kk_various
    data_files:
      - split: train
        path: data/kk_various_cleaned_v2_split-00000-of-00001.parquet
  - config_name: kk_moh
    data_files:
      - split: train
        path: data/kk_moh_cleaned_v2_split-00000-of-00001.parquet
  - config_name: pdf_tq_kaz
    data_files:
      - split: train
        path: data/pdf_tq_kaz-00000-of-00001.parquet
  - config_name: pdf_tq_kaz_2
    data_files:
      - split: train
        path: data/pdf_tq_kaz_2-00000-of-00001.parquet
  - config_name: synthetic_vlad_kk
    data_files:
      - split: train
        path: data/synthetic_vlad_kk-00000-of-00001.parquet
  - config_name: txt_tq_kaz
    data_files:
      - split: train
        path: data/txt_tq_kaz-00000-of-00001.parquet
  - config_name: txt_tq_kaz_2
    data_files:
      - split: train
        path: data/txt_tq_kaz_2-00000-of-00001.parquet
  - config_name: fineweb_2022_27_0
    data_files:
      - split: train
        path: >-
          data/huggingfacefw_fineweb_edu_cc_main_2022_27_0_kk-00000-of-00001.parquet
  - config_name: fineweb_2022_27_1
    data_files:
      - split: train
        path: >-
          data/huggingfacefw_fineweb_edu_cc_main_2022_27_1_kk-00000-of-00001.parquet
  - config_name: fineweb_2022_27_2
    data_files:
      - split: train
        path: >-
          data/huggingfacefw_fineweb_edu_cc_main_2022_27_2_kk-00000-of-00001.parquet
  - config_name: fineweb_2022_27_3
    data_files:
      - split: train
        path: >-
          data/huggingfacefw_fineweb_edu_cc_main_2022_27_3_kk-00000-of-00001.parquet
  - config_name: fineweb_2022_27_4
    data_files:
      - split: train
        path: >-
          data/huggingfacefw_fineweb_edu_cc_main_2022_27_4_kk-00000-of-00001.parquet
  - config_name: fineweb_2022_27_5
    data_files:
      - split: train
        path: >-
          data/huggingfacefw_fineweb_edu_cc_main_2022_27_5_kk-00000-of-00001.parquet
  - config_name: fineweb_2022_27_6
    data_files:
      - split: train
        path: >-
          data/huggingfacefw_fineweb_edu_cc_main_2022_27_6_kk-00000-of-00001.parquet
  - config_name: fineweb_2022_27_7
    data_files:
      - split: train
        path: >-
          data/huggingfacefw_fineweb_edu_cc_main_2022_27_7_kk-00000-of-00001.parquet
  - config_name: fineweb_2022_27_8
    data_files:
      - split: train
        path: >-
          data/huggingfacefw_fineweb_edu_cc_main_2022_27_8_kk-00000-of-00001.parquet
  - config_name: fineweb_2022_33_0
    data_files:
      - split: train
        path: >-
          data/huggingfacefw_fineweb_edu_cc_main_2022_33_0_kk-00000-of-00001.parquet
  - config_name: fineweb_2022_33_1
    data_files:
      - split: train
        path: >-
          data/huggingfacefw_fineweb_edu_cc_main_2022_33_1_kk-00000-of-00001.parquet
  - config_name: fineweb_2022_33_2
    data_files:
      - split: train
        path: >-
          data/huggingfacefw_fineweb_edu_cc_main_2022_33_2_kk-00000-of-00001.parquet
  - config_name: fineweb_2022_33_3
    data_files:
      - split: train
        path: >-
          data/huggingfacefw_fineweb_edu_cc_main_2022_33_3_kk-00000-of-00001.parquet
  - config_name: fineweb_2022_33_4
    data_files:
      - split: train
        path: >-
          data/huggingfacefw_fineweb_edu_cc_main_2022_33_4_kk-00000-of-00001.parquet
  - config_name: fineweb_2022_33_5
    data_files:
      - split: train
        path: >-
          data/huggingfacefw_fineweb_edu_cc_main_2022_33_5_kk-00000-of-00001.parquet
  - config_name: fineweb_2022_33_6
    data_files:
      - split: train
        path: >-
          data/huggingfacefw_fineweb_edu_cc_main_2022_33_6_kk-00000-of-00001.parquet
  - config_name: fineweb_2022_40_0
    data_files:
      - split: train
        path: >-
          data/huggingfacefw_fineweb_edu_cc_main_2022_40_0_kk-00000-of-00001.parquet
  - config_name: fineweb_2022_40_1
    data_files:
      - split: train
        path: >-
          data/huggingfacefw_fineweb_edu_cc_main_2022_40_1_kk-00000-of-00001.parquet
  - config_name: fineweb_2022_40_2
    data_files:
      - split: train
        path: >-
          data/huggingfacefw_fineweb_edu_cc_main_2022_40_2_kk-00000-of-00001.parquet
  - config_name: fineweb_2022_40_3
    data_files:
      - split: train
        path: >-
          data/huggingfacefw_fineweb_edu_cc_main_2022_40_3_kk-00000-of-00001.parquet
  - config_name: fineweb_2022_40_4
    data_files:
      - split: train
        path: >-
          data/huggingfacefw_fineweb_edu_cc_main_2022_40_4_kk-00000-of-00001.parquet
  - config_name: fineweb_2022_40_6
    data_files:
      - split: train
        path: >-
          data/huggingfacefw_fineweb_edu_cc_main_2022_40_6_kk-00000-of-00001.parquet
  - config_name: fineweb_2022_40_7
    data_files:
      - split: train
        path: >-
          data/huggingfacefw_fineweb_edu_cc_main_2022_40_7_kk-00000-of-00001.parquet
  - config_name: fineweb_2022_40_9
    data_files:
      - split: train
        path: >-
          data/huggingfacefw_fineweb_edu_cc_main_2022_40_9_kk-00000-of-00001.parquet
  - config_name: fineweb_2022_49_0
    data_files:
      - split: train
        path: >-
          data/huggingfacefw_fineweb_edu_cc_main_2022_49_0_kk-00000-of-00001.parquet
  - config_name: fineweb_2022_49_2
    data_files:
      - split: train
        path: >-
          data/huggingfacefw_fineweb_edu_cc_main_2022_49_2_kk-00000-of-00001.parquet
  - config_name: fineweb_2022_49_3
    data_files:
      - split: train
        path: >-
          data/huggingfacefw_fineweb_edu_cc_main_2022_49_3_kk-00000-of-00001.parquet
  - config_name: fineweb_2022_49_4
    data_files:
      - split: train
        path: >-
          data/huggingfacefw_fineweb_edu_cc_main_2022_49_4_kk-00000-of-00001.parquet
  - config_name: fineweb_2022_49_5
    data_files:
      - split: train
        path: >-
          data/huggingfacefw_fineweb_edu_cc_main_2022_49_5_kk-00000-of-00001.parquet
  - config_name: fineweb_2022_49_6
    data_files:
      - split: train
        path: >-
          data/huggingfacefw_fineweb_edu_cc_main_2022_49_6_kk-00000-of-00001.parquet
  - config_name: fineweb_2024_26_0
    data_files:
      - split: train
        path: >-
          data/huggingfacefw_fineweb_edu_cc_main_2024_26_0_kk-00000-of-00001.parquet
  - config_name: fineweb_2024_26_1
    data_files:
      - split: train
        path: >-
          data/huggingfacefw_fineweb_edu_cc_main_2024_26_1_kk-00000-of-00001.parquet
  - config_name: fineweb_2024_26_2
    data_files:
      - split: train
        path: >-
          data/huggingfacefw_fineweb_edu_cc_main_2024_26_2_kk-00000-of-00001.parquet
  - config_name: fineweb_2024_26_3
    data_files:
      - split: train
        path: >-
          data/huggingfacefw_fineweb_edu_cc_main_2024_26_3_kk-00000-of-00001.parquet
  - config_name: fineweb_2024_26_4
    data_files:
      - split: train
        path: >-
          data/huggingfacefw_fineweb_edu_cc_main_2024_26_4_kk-00000-of-00001.parquet
  - config_name: fineweb_2024_26_5
    data_files:
      - split: train
        path: >-
          data/huggingfacefw_fineweb_edu_cc_main_2024_26_5_kk-00000-of-00001.parquet
  - config_name: fineweb_2024_26_6
    data_files:
      - split: train
        path: >-
          data/huggingfacefw_fineweb_edu_cc_main_2024_26_6_kk-00000-of-00001.parquet
  - config_name: fineweb_2024_26_7
    data_files:
      - split: train
        path: >-
          data/huggingfacefw_fineweb_edu_cc_main_2024_26_7_kk-00000-of-00001.parquet
  - config_name: fineweb_2024_26_8
    data_files:
      - split: train
        path: >-
          data/huggingfacefw_fineweb_edu_cc_main_2024_26_8_kk-00000-of-00001.parquet

Make sure you add the complete list of parquet files. The key is that we're keeping the working "all" subset while adding the individual ones.

Some troubleshooting points to consider:

Be consistent with file paths: If your working config uses "data/*.parquet", then make sure all your individual files use "data/" prefix too. Check for typos in filenames: Ensure the filenames exactly match what's in your repository. Make incremental changes: Add just a few subsets first, see if that works, then add more. Try without the "all" subset: If there are issues, try removing the "all" subset and just having individual ones. If you're still facing issues, you might want to check if there's a limit on how many subsets you can define in a dataset. Since you have a large number of files, try starting with a smaller subset of them to see if that works.

Dataset Card

The Base Train Dataset is a large-scale collection of corpora designed for Kazakh language model training.

Overview

The dataset comprises a mix of original Kazakh datasets and machine-translated content, with a significant portion derived from the FineWeb Edu corpus.

Data can be very huge, so if you need a specific file, you can download it using the code snippet below.:

from datasets import load_dataset

dataset = load_dataset(
    "issai/Base_Transfer",
    data_files={"train": "data/kk_ocr_fixed_cleaned_v2_split-00000-of-00001.parquet"},
    split="train"
)

File Summary

File Name Row Count
HuggingFaceFW_fineweb-edu_CC-MAIN-2022-27_train_8-kk 303439
HuggingFaceFW_fineweb-edu_CC-MAIN-2022-40_train_9-kk 1071921
HuggingFaceFW_fineweb-edu_CC-MAIN-2022-33_train_6-kk 1118775
HuggingFaceFW_fineweb-edu_CC-MAIN-2022-33_train_1-kk 1905224
HuggingFaceFW_fineweb-edu_CC-MAIN-2022-49_train_0-kk 1838387
HuggingFaceFW_fineweb-edu_CC-MAIN-2022-49_train_2-kk 1881636
HuggingFaceFW_fineweb-edu_CC-MAIN-2022-49_train_1-kk 1862602
HuggingFaceFW_fineweb-edu_CC-MAIN-2022-49_train_3-kk 1905713
HuggingFaceFW_fineweb-edu_CC-MAIN-2022-33_train_0-kk 1954983
HuggingFaceFW_fineweb-edu_CC-MAIN-2022-33_train_4-kk 1966187
HuggingFaceFW_fineweb-edu_CC-MAIN-2022-33_train_3-kk 1962158
HuggingFaceFW_fineweb-edu_CC-MAIN-2022-40_train_2-kk 1965651
HuggingFaceFW_fineweb-edu_CC-MAIN-2022-49_train_5-kk 1904798
HuggingFaceFW_fineweb-edu_CC-MAIN-2022-27_train_0-kk 1967683
HuggingFaceFW_fineweb-edu_CC-MAIN-2022-33_train_2-kk 1962805
HuggingFaceFW_fineweb-edu_CC-MAIN-2022-49_train_4-kk 1904738
HuggingFaceFW_fineweb-edu_CC-MAIN-2022-27_train_3-kk 1985097
HuggingFaceFW_fineweb-edu_CC-MAIN-2022-27_train_4-kk 1985361
HuggingFaceFW_fineweb-edu_CC-MAIN-2022-40_train_8-kk 1977335
HuggingFaceFW_fineweb-edu_CC-MAIN-2022-27_train_6-kk 1992677
HuggingFaceFW_fineweb-edu_CC-MAIN-2022-40_train_6-kk 1979823
HuggingFaceFW_fineweb-edu_CC-MAIN-2022-33_train_5-kk 1969247
HuggingFaceFW_fineweb-edu_CC-MAIN-2022-27_train_7-kk 1996366
HuggingFaceFW_fineweb-edu_CC-MAIN-2022-27_train_5-kk 1984027
HuggingFaceFW_fineweb-edu_CC-MAIN-2022-40_train_1-kk 1964816
HuggingFaceFW_fineweb-edu_CC-MAIN-2022-40_train_5-kk 1973355
HuggingFaceFW_fineweb-edu_CC-MAIN-2022-40_train_7-kk 1975429
HuggingFaceFW_fineweb-edu_CC-MAIN-2022-40_train_4-kk 1971413
HuggingFaceFW_fineweb-edu_CC-MAIN-2022-40_train_0-kk 1964920
HuggingFaceFW_fineweb-edu_CC-MAIN-2022-40_train_3-kk 1972242
HuggingFaceFW_fineweb-edu_CC-MAIN-2022-27_train_1-kk 1967275
HuggingFaceFW_fineweb-edu_CC-MAIN-2022-49_train_6-kk 1919322
HuggingFaceFW_fineweb-edu_CC-MAIN-2024-26_train_0-kk 1541581
HuggingFaceFW_fineweb-edu_CC-MAIN-2024-26_train_1-kk 1866939
HuggingFaceFW_fineweb-edu_CC-MAIN-2024-42_train_7-kk 1312480
HuggingFaceFW_fineweb-edu_CC-MAIN-2024-42_train_3-kk 1289695
HuggingFaceFW_fineweb-edu_CC-MAIN-2024-46_train_1-kk 1310115
HuggingFaceFW_fineweb-edu_CC-MAIN-2024-26_train_8-kk 1671792
HuggingFaceFW_fineweb-edu_CC-MAIN-2024-26_train_2-kk 1849038
HuggingFaceFW_fineweb-edu_CC-MAIN-2022-27_train_2-kk 3821369
HuggingFaceFW_fineweb-edu_CC-MAIN-2024-26_train_6-kk 1848854
HuggingFaceFW_fineweb-edu_CC-MAIN-2024-30_train_1-kk 1797537
HuggingFaceFW_fineweb-edu_CC-MAIN-2024-26_train_3-kk 1886754
HuggingFaceFW_fineweb-edu_CC-MAIN-2024-30_train_3-kk 1833826
HuggingFaceFW_fineweb-edu_CC-MAIN-2024-26_train_4-kk 1851013
HuggingFaceFW_fineweb-edu_CC-MAIN-2024-30_train_6-kk 1797638
HuggingFaceFW_fineweb-edu_CC-MAIN-2024-42_train_1-kk 1747944
HuggingFaceFW_fineweb-edu_CC-MAIN-2024-30_train_2-kk 1790560
HuggingFaceFW_fineweb-edu_CC-MAIN-2024-42_train_5-kk 1811397
pdf_tq_kaz_2 4957
train_kk_news180924_cleaned_v2_split 21307
HuggingFaceFW_fineweb-edu_CC-MAIN-2024-42_train_2-kk 1783420
doc_tq_kaz_2 4928
train_kk_moh_cleaned_v2_split 5867
HuggingFaceFW_fineweb-edu_CC-MAIN-2024-26_train_5-kk 1887322
HuggingFaceFW_fineweb-edu_CC-MAIN-2024-30_train_4-kk 1844746
HuggingFaceFW_fineweb-edu_CC-MAIN-2024-30_train_7-kk 1842195
HuggingFaceFW_fineweb-edu_CC-MAIN-2024-30_train_0-kk 1878396
HuggingFaceFW_fineweb-edu_CC-MAIN-2024-30_train_5-kk 1834744
HuggingFaceFW_fineweb-edu_CC-MAIN-2024-26_train_7-kk 1878031
doc_tq_kaz 5844
synthetic_vlad_kk 129300
HuggingFaceFW_fineweb-edu_CC-MAIN-2024-42_train_6-kk 1779623
txt_tq_kaz_2 20845
pdf_tq_kaz 9520
HuggingFaceFW_fineweb-edu_CC-MAIN-2024-42_train_0-kk 1828313
HuggingFaceFW_fineweb-edu_CC-MAIN-2024-42_train_4-kk 1821469
HuggingFaceFW_fineweb-edu_CC-MAIN-2024-46_train_4-kk 1832893
HuggingFaceFW_fineweb-edu_CC-MAIN-2024-46_train_0-kk 1844265
txt_tq_kaz 671
HuggingFaceFW_fineweb-edu_CC-MAIN-2024-46_train_2-kk 1843752
HuggingFaceFW_fineweb-edu_CC-MAIN-2024-46_train_3-kk 1833394
HuggingFaceFW_fineweb-edu_CC-MAIN-2024-46_train_5-kk 1841828
train_kk_various_cleaned_v2_split 55752
train_kk_tilqazyna_cleaned_v2_split 74170
train_kk_ocr_fixed_cleaned_v2_split 170820
HuggingFaceFW_fineweb-edu_CC-MAIN-2024-46_train_6-kk 1841071
HuggingFaceFW_fineweb-edu_CC-MAIN-2024-51_train_6-kk 1169812
HuggingFaceFW_fineweb-edu_CC-MAIN-2024-51_train_0-kk 1925577
HuggingFaceFW_fineweb-edu_CC-MAIN-2024-51_train_4-kk 1928149
HuggingFaceFW_fineweb-edu_CC-MAIN-2024-51_train_2-kk 1927534
HuggingFaceFW_fineweb-edu_CC-MAIN-2024-51_train_1-kk 1929222
HuggingFaceFW_fineweb-edu_CC-MAIN-2024-51_train_3-kk 1928565
HuggingFaceFW_fineweb-edu_CC-MAIN-2024-51_train_5-kk 1922570
train_kk_wikipedia_cleaned_v2_split 6711027
train_kk_fineweb_cleaned_v2_split 3918479
train_kk_news_cleaned_v2_split 6661343

Total Files: 87 Total Rows: 150,025,658

License & Citation

This dataset is intended for research purposes. Please cite appropriately if used in publications.