README.md

metadata

annotations_creators: []
language_creators:
  - crowdsourced
  - expert-generated
language:
  - code
license:
  - other
multilinguality:
  - multilingual
pretty_name: The-Stack-v2
size_categories:
  - unknown
source_datasets: []
task_categories:
  - text-generation
task_ids: []
extra_gated_prompt: >-
  ## Terms of Use for The Stack v2


  The Stack v2 dataset is a collection of source code in over 600 programming
  languages. We ask that you read and acknowledge the following points before
  using the dataset:

  1. Downloading the dataset in bulk requires a an agreement with
  SoftwareHeritage and INRIA. Contact
  [datasets@softwareheritage.org](mailto:datasets@softwareheritage.org?subject=TheStackV2%20request%20for%20dataset%20access%20information)
  for more information.

  2. If you are using the dataset to train models you must adhere to the
  SoftwareHeritage [principles for language model
  training](https://www.softwareheritage.org/2023/10/19/swh-statement-on-llm-for-code/).

  3. The Stack v2 is a collection of source code from repositories with various
  licenses. Any use of all or part of the code gathered in The Stack v2 must
  abide by the terms of the original licenses, including attribution clauses
  when relevant. We facilitate this by providing provenance information for each
  data point.

  4. The Stack v2 is regularly updated to enact validated data removal requests.
  By clicking on "Access repository", you agree to update your own version of
  The Stack v2 to the most recent usable version.


  By clicking on "Access repository" below, you accept that your contact
  information (email address and username) can be shared with the dataset
  maintainers as well.
    
extra_gated_fields:
  Email: text
  I have read the License and agree with its terms: checkbox
dataset_info:
  features:
    - name: blob_id
      dtype: string
    - name: directory_id
      dtype: string
    - name: path
      dtype: string
    - name: content_id
      dtype: string
    - name: detected_licenses
      sequence: string
    - name: license_type
      dtype: string
    - name: repo_name
      dtype: string
    - name: snapshot_id
      dtype: string
    - name: revision_id
      dtype: string
    - name: branch_name
      dtype: string
    - name: visit_date
      dtype: timestamp[ns]
    - name: revision_date
      dtype: timestamp[ns]
    - name: committer_date
      dtype: timestamp[ns]
    - name: github_id
      dtype: int64
    - name: star_events_count
      dtype: int64
    - name: fork_events_count
      dtype: int64
    - name: gha_license_id
      dtype: string
    - name: gha_event_created_at
      dtype: timestamp[ns]
    - name: gha_created_at
      dtype: timestamp[ns]
    - name: gha_language
      dtype: string
    - name: src_encoding
      dtype: string
    - name: language
      dtype: string
    - name: is_vendor
      dtype: bool
    - name: is_generated
      dtype: bool
    - name: length_bytes
      dtype: int64
    - name: extension
      dtype: string
configs:
  - config_name: default
    default: true
    data_files:
      - split: train
        path: data/*/*.parquet
  - config_name: 1C_Enterprise
    data_files:
      - split: train
        path: data/1C_Enterprise/*.parquet
  - config_name: 2-Dimensional_Array
    data_files:
      - split: train
        path: data/2-Dimensional_Array/*.parquet
  - config_name: 4D
    data_files:
      - split: train
        path: data/4D/*.parquet
  - config_name: ABAP
    data_files:
      - split: train
        path: data/ABAP/*.parquet
  - config_name: ABAP_CDS
    data_files:
      - split: train
        path: data/ABAP_CDS/*.parquet
  - config_name: ABNF
    data_files:
      - split: train
        path: data/ABNF/*.parquet
  - config_name: AGS_Script
    data_files:
      - split: train
        path: data/AGS_Script/*.parquet
  - config_name: AIDL
    data_files:
      - split: train
        path: data/AIDL/*.parquet
  - config_name: AL
    data_files:
      - split: train
        path: data/AL/*.parquet
  - config_name: AMPL
    data_files:
      - split: train
        path: data/AMPL/*.parquet
  - config_name: ANTLR
    data_files:
      - split: train
        path: data/ANTLR/*.parquet
  - config_name: API_Blueprint
    data_files:
      - split: train
        path: data/API_Blueprint/*.parquet
  - config_name: APL
    data_files:
      - split: train
        path: data/APL/*.parquet
  - config_name: ASL
    data_files:
      - split: train
        path: data/ASL/*.parquet
  - config_name: ASN.1
    data_files:
      - split: train
        path: data/ASN.1/*.parquet
  - config_name: ASP.NET
    data_files:
      - split: train
        path: data/ASP.NET/*.parquet
  - config_name: ATS
    data_files:
      - split: train
        path: data/ATS/*.parquet
  - config_name: ActionScript
    data_files:
      - split: train
        path: data/ActionScript/*.parquet
  - config_name: Ada
    data_files:
      - split: train
        path: data/Ada/*.parquet
  - config_name: Adobe_Font_Metrics
    data_files:
      - split: train
        path: data/Adobe_Font_Metrics/*.parquet
  - config_name: Agda
    data_files:
      - split: train
        path: data/Agda/*.parquet
  - config_name: Alloy
    data_files:
      - split: train
        path: data/Alloy/*.parquet
  - config_name: Alpine_Abuild
    data_files:
      - split: train
        path: data/Alpine_Abuild/*.parquet
  - config_name: Altium_Designer
    data_files:
      - split: train
        path: data/Altium_Designer/*.parquet
  - config_name: AngelScript
    data_files:
      - split: train
        path: data/AngelScript/*.parquet
  - config_name: Ant_Build_System
    data_files:
      - split: train
        path: data/Ant_Build_System/*.parquet
  - config_name: Antlers
    data_files:
      - split: train
        path: data/Antlers/*.parquet
  - config_name: ApacheConf
    data_files:
      - split: train
        path: data/ApacheConf/*.parquet
  - config_name: Apex
    data_files:
      - split: train
        path: data/Apex/*.parquet
  - config_name: Apollo_Guidance_Computer
    data_files:
      - split: train
        path: data/Apollo_Guidance_Computer/*.parquet
  - config_name: AppleScript
    data_files:
      - split: train
        path: data/AppleScript/*.parquet
  - config_name: Arc
    data_files:
      - split: train
        path: data/Arc/*.parquet
  - config_name: AsciiDoc
    data_files:
      - split: train
        path: data/AsciiDoc/*.parquet
  - config_name: AspectJ
    data_files:
      - split: train
        path: data/AspectJ/*.parquet
  - config_name: Assembly
    data_files:
      - split: train
        path: data/Assembly/*.parquet
  - config_name: Astro
    data_files:
      - split: train
        path: data/Astro/*.parquet
  - config_name: Asymptote
    data_files:
      - split: train
        path: data/Asymptote/*.parquet
  - config_name: Augeas
    data_files:
      - split: train
        path: data/Augeas/*.parquet
  - config_name: AutoHotkey
    data_files:
      - split: train
        path: data/AutoHotkey/*.parquet
  - config_name: AutoIt
    data_files:
      - split: train
        path: data/AutoIt/*.parquet
  - config_name: Avro_IDL
    data_files:
      - split: train
        path: data/Avro_IDL/*.parquet
  - config_name: Awk
    data_files:
      - split: train
        path: data/Awk/*.parquet
  - config_name: BASIC
    data_files:
      - split: train
        path: data/BASIC/*.parquet
  - config_name: Ballerina
    data_files:
      - split: train
        path: data/Ballerina/*.parquet
  - config_name: Batchfile
    data_files:
      - split: train
        path: data/Batchfile/*.parquet
  - config_name: Beef
    data_files:
      - split: train
        path: data/Beef/*.parquet
  - config_name: Befunge
    data_files:
      - split: train
        path: data/Befunge/*.parquet
  - config_name: Berry
    data_files:
      - split: train
        path: data/Berry/*.parquet
  - config_name: BibTeX
    data_files:
      - split: train
        path: data/BibTeX/*.parquet
  - config_name: Bicep
    data_files:
      - split: train
        path: data/Bicep/*.parquet
  - config_name: Bikeshed
    data_files:
      - split: train
        path: data/Bikeshed/*.parquet
  - config_name: Bison
    data_files:
      - split: train
        path: data/Bison/*.parquet
  - config_name: BitBake
    data_files:
      - split: train
        path: data/BitBake/*.parquet
  - config_name: Blade
    data_files:
      - split: train
        path: data/Blade/*.parquet
  - config_name: BlitzBasic
    data_files:
      - split: train
        path: data/BlitzBasic/*.parquet
  - config_name: BlitzMax
    data_files:
      - split: train
        path: data/BlitzMax/*.parquet
  - config_name: Bluespec
    data_files:
      - split: train
        path: data/Bluespec/*.parquet
  - config_name: Boo
    data_files:
      - split: train
        path: data/Boo/*.parquet
  - config_name: Boogie
    data_files:
      - split: train
        path: data/Boogie/*.parquet
  - config_name: Brainfuck
    data_files:
      - split: train
        path: data/Brainfuck/*.parquet
  - config_name: BrighterScript
    data_files:
      - split: train
        path: data/BrighterScript/*.parquet
  - config_name: Brightscript
    data_files:
      - split: train
        path: data/Brightscript/*.parquet
  - config_name: Browserslist
    data_files:
      - split: train
        path: data/Browserslist/*.parquet
  - config_name: C
    data_files:
      - split: train
        path: data/C/*.parquet
  - config_name: C++
    data_files:
      - split: train
        path: data/C++/*.parquet
  - config_name: C-ObjDump
    data_files:
      - split: train
        path: data/C-ObjDump/*.parquet
  - config_name: C-Sharp
    data_files:
      - split: train
        path: data/C-Sharp/*.parquet
  - config_name: C2hs_Haskell
    data_files:
      - split: train
        path: data/C2hs_Haskell/*.parquet
  - config_name: CAP_CDS
    data_files:
      - split: train
        path: data/CAP_CDS/*.parquet
  - config_name: CIL
    data_files:
      - split: train
        path: data/CIL/*.parquet
  - config_name: CLIPS
    data_files:
      - split: train
        path: data/CLIPS/*.parquet
  - config_name: CMake
    data_files:
      - split: train
        path: data/CMake/*.parquet
  - config_name: COBOL
    data_files:
      - split: train
        path: data/COBOL/*.parquet
  - config_name: CODEOWNERS
    data_files:
      - split: train
        path: data/CODEOWNERS/*.parquet
  - config_name: COLLADA
    data_files:
      - split: train
        path: data/COLLADA/*.parquet
  - config_name: CSON
    data_files:
      - split: train
        path: data/CSON/*.parquet
  - config_name: CSS
    data_files:
      - split: train
        path: data/CSS/*.parquet
  - config_name: CSV
    data_files:
      - split: train
        path: data/CSV/*.parquet
  - config_name: CUE
    data_files:
      - split: train
        path: data/CUE/*.parquet
  - config_name: CWeb
    data_files:
      - split: train
        path: data/CWeb/*.parquet
  - config_name: Cabal_Config
    data_files:
      - split: train
        path: data/Cabal_Config/*.parquet
  - config_name: Cadence
    data_files:
      - split: train
        path: data/Cadence/*.parquet
  - config_name: Cairo
    data_files:
      - split: train
        path: data/Cairo/*.parquet
  - config_name: CameLIGO
    data_files:
      - split: train
        path: data/CameLIGO/*.parquet
  - config_name: Cap-n_Proto
    data_files:
      - split: train
        path: data/Cap-n_Proto/*.parquet
  - config_name: CartoCSS
    data_files:
      - split: train
        path: data/CartoCSS/*.parquet
  - config_name: Ceylon
    data_files:
      - split: train
        path: data/Ceylon/*.parquet
  - config_name: Chapel
    data_files:
      - split: train
        path: data/Chapel/*.parquet
  - config_name: Charity
    data_files:
      - split: train
        path: data/Charity/*.parquet
  - config_name: Checksums
    data_files:
      - split: train
        path: data/Checksums/*.parquet
  - config_name: ChucK
    data_files:
      - split: train
        path: data/ChucK/*.parquet
  - config_name: Cirru
    data_files:
      - split: train
        path: data/Cirru/*.parquet
  - config_name: Clarion
    data_files:
      - split: train
        path: data/Clarion/*.parquet
  - config_name: Clarity
    data_files:
      - split: train
        path: data/Clarity/*.parquet
  - config_name: Classic_ASP
    data_files:
      - split: train
        path: data/Classic_ASP/*.parquet
  - config_name: Clean
    data_files:
      - split: train
        path: data/Clean/*.parquet
  - config_name: Click
    data_files:
      - split: train
        path: data/Click/*.parquet
  - config_name: Clojure
    data_files:
      - split: train
        path: data/Clojure/*.parquet
  - config_name: Closure_Templates
    data_files:
      - split: train
        path: data/Closure_Templates/*.parquet
  - config_name: Cloud_Firestore_Security_Rules
    data_files:
      - split: train
        path: data/Cloud_Firestore_Security_Rules/*.parquet
  - config_name: CoNLL-U
    data_files:
      - split: train
        path: data/CoNLL-U/*.parquet
  - config_name: CodeQL
    data_files:
      - split: train
        path: data/CodeQL/*.parquet
  - config_name: CoffeeScript
    data_files:
      - split: train
        path: data/CoffeeScript/*.parquet
  - config_name: ColdFusion
    data_files:
      - split: train
        path: data/ColdFusion/*.parquet
  - config_name: ColdFusion_CFC
    data_files:
      - split: train
        path: data/ColdFusion_CFC/*.parquet
  - config_name: Common_Lisp
    data_files:
      - split: train
        path: data/Common_Lisp/*.parquet
  - config_name: Common_Workflow_Language
    data_files:
      - split: train
        path: data/Common_Workflow_Language/*.parquet
  - config_name: Component_Pascal
    data_files:
      - split: train
        path: data/Component_Pascal/*.parquet
  - config_name: Cool
    data_files:
      - split: train
        path: data/Cool/*.parquet
  - config_name: Coq
    data_files:
      - split: train
        path: data/Coq/*.parquet
  - config_name: Creole
    data_files:
      - split: train
        path: data/Creole/*.parquet
  - config_name: Crystal
    data_files:
      - split: train
        path: data/Crystal/*.parquet
  - config_name: Csound
    data_files:
      - split: train
        path: data/Csound/*.parquet
  - config_name: Csound_Document
    data_files:
      - split: train
        path: data/Csound_Document/*.parquet
  - config_name: Csound_Score
    data_files:
      - split: train
        path: data/Csound_Score/*.parquet
  - config_name: Cuda
    data_files:
      - split: train
        path: data/Cuda/*.parquet
  - config_name: Cue_Sheet
    data_files:
      - split: train
        path: data/Cue_Sheet/*.parquet
  - config_name: Curry
    data_files:
      - split: train
        path: data/Curry/*.parquet
  - config_name: Cycript
    data_files:
      - split: train
        path: data/Cycript/*.parquet
  - config_name: Cython
    data_files:
      - split: train
        path: data/Cython/*.parquet
  - config_name: D
    data_files:
      - split: train
        path: data/D/*.parquet
  - config_name: DIGITAL_Command_Language
    data_files:
      - split: train
        path: data/DIGITAL_Command_Language/*.parquet
  - config_name: DM
    data_files:
      - split: train
        path: data/DM/*.parquet
  - config_name: DNS_Zone
    data_files:
      - split: train
        path: data/DNS_Zone/*.parquet
  - config_name: DTrace
    data_files:
      - split: train
        path: data/DTrace/*.parquet
  - config_name: Dafny
    data_files:
      - split: train
        path: data/Dafny/*.parquet
  - config_name: Darcs_Patch
    data_files:
      - split: train
        path: data/Darcs_Patch/*.parquet
  - config_name: Dart
    data_files:
      - split: train
        path: data/Dart/*.parquet
  - config_name: DataWeave
    data_files:
      - split: train
        path: data/DataWeave/*.parquet
  - config_name: Debian_Package_Control_File
    data_files:
      - split: train
        path: data/Debian_Package_Control_File/*.parquet
  - config_name: DenizenScript
    data_files:
      - split: train
        path: data/DenizenScript/*.parquet
  - config_name: Dhall
    data_files:
      - split: train
        path: data/Dhall/*.parquet
  - config_name: Diff
    data_files:
      - split: train
        path: data/Diff/*.parquet
  - config_name: DirectX_3D_File
    data_files:
      - split: train
        path: data/DirectX_3D_File/*.parquet
  - config_name: Dockerfile
    data_files:
      - split: train
        path: data/Dockerfile/*.parquet
  - config_name: Dogescript
    data_files:
      - split: train
        path: data/Dogescript/*.parquet
  - config_name: Dylan
    data_files:
      - split: train
        path: data/Dylan/*.parquet
  - config_name: E
    data_files:
      - split: train
        path: data/E/*.parquet
  - config_name: E-mail
    data_files:
      - split: train
        path: data/E-mail/*.parquet
  - config_name: EBNF
    data_files:
      - split: train
        path: data/EBNF/*.parquet
  - config_name: ECL
    data_files:
      - split: train
        path: data/ECL/*.parquet
  - config_name: ECLiPSe
    data_files:
      - split: train
        path: data/ECLiPSe/*.parquet
  - config_name: EJS
    data_files:
      - split: train
        path: data/EJS/*.parquet
  - config_name: EQ
    data_files:
      - split: train
        path: data/EQ/*.parquet
  - config_name: Eagle
    data_files:
      - split: train
        path: data/Eagle/*.parquet
  - config_name: Earthly
    data_files:
      - split: train
        path: data/Earthly/*.parquet
  - config_name: Easybuild
    data_files:
      - split: train
        path: data/Easybuild/*.parquet
  - config_name: Ecere_Projects
    data_files:
      - split: train
        path: data/Ecere_Projects/*.parquet
  - config_name: EditorConfig
    data_files:
      - split: train
        path: data/EditorConfig/*.parquet
  - config_name: Edje_Data_Collection
    data_files:
      - split: train
        path: data/Edje_Data_Collection/*.parquet
  - config_name: Eiffel
    data_files:
      - split: train
        path: data/Eiffel/*.parquet
  - config_name: Elixir
    data_files:
      - split: train
        path: data/Elixir/*.parquet
  - config_name: Elm
    data_files:
      - split: train
        path: data/Elm/*.parquet
  - config_name: Elvish
    data_files:
      - split: train
        path: data/Elvish/*.parquet
  - config_name: Emacs_Lisp
    data_files:
      - split: train
        path: data/Emacs_Lisp/*.parquet
  - config_name: EmberScript
    data_files:
      - split: train
        path: data/EmberScript/*.parquet
  - config_name: Erlang
    data_files:
      - split: train
        path: data/Erlang/*.parquet
  - config_name: Euphoria
    data_files:
      - split: train
        path: data/Euphoria/*.parquet
  - config_name: F-Sharp
    data_files:
      - split: train
        path: data/F-Sharp/*.parquet
  - config_name: F-Star
    data_files:
      - split: train
        path: data/F-Star/*.parquet
  - config_name: FIGlet_Font
    data_files:
      - split: train
        path: data/FIGlet_Font/*.parquet
  - config_name: FLUX
    data_files:
      - split: train
        path: data/FLUX/*.parquet
  - config_name: Factor
    data_files:
      - split: train
        path: data/Factor/*.parquet
  - config_name: Fancy
    data_files:
      - split: train
        path: data/Fancy/*.parquet
  - config_name: Fantom
    data_files:
      - split: train
        path: data/Fantom/*.parquet
  - config_name: Faust
    data_files:
      - split: train
        path: data/Faust/*.parquet
  - config_name: Fennel
    data_files:
      - split: train
        path: data/Fennel/*.parquet
  - config_name: Filebench_WML
    data_files:
      - split: train
        path: data/Filebench_WML/*.parquet
  - config_name: Filterscript
    data_files:
      - split: train
        path: data/Filterscript/*.parquet
  - config_name: Fluent
    data_files:
      - split: train
        path: data/Fluent/*.parquet
  - config_name: Formatted
    data_files:
      - split: train
        path: data/Formatted/*.parquet
  - config_name: Forth
    data_files:
      - split: train
        path: data/Forth/*.parquet
  - config_name: Fortran
    data_files:
      - split: train
        path: data/Fortran/*.parquet
  - config_name: Fortran_Free_Form
    data_files:
      - split: train
        path: data/Fortran_Free_Form/*.parquet
  - config_name: FreeBasic
    data_files:
      - split: train
        path: data/FreeBasic/*.parquet
  - config_name: FreeMarker
    data_files:
      - split: train
        path: data/FreeMarker/*.parquet
  - config_name: Frege
    data_files:
      - split: train
        path: data/Frege/*.parquet
  - config_name: Futhark
    data_files:
      - split: train
        path: data/Futhark/*.parquet
  - config_name: G-code
    data_files:
      - split: train
        path: data/G-code/*.parquet
  - config_name: GAML
    data_files:
      - split: train
        path: data/GAML/*.parquet
  - config_name: GAMS
    data_files:
      - split: train
        path: data/GAMS/*.parquet
  - config_name: GAP
    data_files:
      - split: train
        path: data/GAP/*.parquet
  - config_name: GCC_Machine_Description
    data_files:
      - split: train
        path: data/GCC_Machine_Description/*.parquet
  - config_name: GDB
    data_files:
      - split: train
        path: data/GDB/*.parquet
  - config_name: GDScript
    data_files:
      - split: train
        path: data/GDScript/*.parquet
  - config_name: GEDCOM
    data_files:
      - split: train
        path: data/GEDCOM/*.parquet
  - config_name: GLSL
    data_files:
      - split: train
        path: data/GLSL/*.parquet
  - config_name: GN
    data_files:
      - split: train
        path: data/GN/*.parquet
  - config_name: GSC
    data_files:
      - split: train
        path: data/GSC/*.parquet
  - config_name: Game_Maker_Language
    data_files:
      - split: train
        path: data/Game_Maker_Language/*.parquet
  - config_name: Gemfile.lock
    data_files:
      - split: train
        path: data/Gemfile.lock/*.parquet
  - config_name: Gemini
    data_files:
      - split: train
        path: data/Gemini/*.parquet
  - config_name: Genero
    data_files:
      - split: train
        path: data/Genero/*.parquet
  - config_name: Genero_Forms
    data_files:
      - split: train
        path: data/Genero_Forms/*.parquet
  - config_name: Genie
    data_files:
      - split: train
        path: data/Genie/*.parquet
  - config_name: Genshi
    data_files:
      - split: train
        path: data/Genshi/*.parquet
  - config_name: Gentoo_Ebuild
    data_files:
      - split: train
        path: data/Gentoo_Ebuild/*.parquet
  - config_name: Gentoo_Eclass
    data_files:
      - split: train
        path: data/Gentoo_Eclass/*.parquet
  - config_name: Gerber_Image
    data_files:
      - split: train
        path: data/Gerber_Image/*.parquet
  - config_name: Gettext_Catalog
    data_files:
      - split: train
        path: data/Gettext_Catalog/*.parquet
  - config_name: Gherkin
    data_files:
      - split: train
        path: data/Gherkin/*.parquet
  - config_name: Git_Attributes
    data_files:
      - split: train
        path: data/Git_Attributes/*.parquet
  - config_name: Git_Config
    data_files:
      - split: train
        path: data/Git_Config/*.parquet
  - config_name: Git_Revision_List
    data_files:
      - split: train
        path: data/Git_Revision_List/*.parquet
  - config_name: Gleam
    data_files:
      - split: train
        path: data/Gleam/*.parquet
  - config_name: Glyph
    data_files:
      - split: train
        path: data/Glyph/*.parquet
  - config_name: Glyph_Bitmap_Distribution_Format
    data_files:
      - split: train
        path: data/Glyph_Bitmap_Distribution_Format/*.parquet
  - config_name: Gnuplot
    data_files:
      - split: train
        path: data/Gnuplot/*.parquet
  - config_name: Go
    data_files:
      - split: train
        path: data/Go/*.parquet
  - config_name: Go_Checksums
    data_files:
      - split: train
        path: data/Go_Checksums/*.parquet
  - config_name: Go_Module
    data_files:
      - split: train
        path: data/Go_Module/*.parquet
  - config_name: Golo
    data_files:
      - split: train
        path: data/Golo/*.parquet
  - config_name: Gosu
    data_files:
      - split: train
        path: data/Gosu/*.parquet
  - config_name: Grace
    data_files:
      - split: train
        path: data/Grace/*.parquet
  - config_name: Gradle
    data_files:
      - split: train
        path: data/Gradle/*.parquet
  - config_name: Grammatical_Framework
    data_files:
      - split: train
        path: data/Grammatical_Framework/*.parquet
  - config_name: GraphQL
    data_files:
      - split: train
        path: data/GraphQL/*.parquet
  - config_name: Graph_Modeling_Language
    data_files:
      - split: train
        path: data/Graph_Modeling_Language/*.parquet
  - config_name: Graphviz_(DOT)
    data_files:
      - split: train
        path: data/Graphviz_(DOT)/*.parquet
  - config_name: Groovy
    data_files:
      - split: train
        path: data/Groovy/*.parquet
  - config_name: Groovy_Server_Pages
    data_files:
      - split: train
        path: data/Groovy_Server_Pages/*.parquet
  - config_name: HAProxy
    data_files:
      - split: train
        path: data/HAProxy/*.parquet
  - config_name: HCL
    data_files:
      - split: train
        path: data/HCL/*.parquet
  - config_name: HLSL
    data_files:
      - split: train
        path: data/HLSL/*.parquet
  - config_name: HOCON
    data_files:
      - split: train
        path: data/HOCON/*.parquet
  - config_name: HTML
    data_files:
      - split: train
        path: data/HTML/*.parquet
  - config_name: HTML+ECR
    data_files:
      - split: train
        path: data/HTML+ECR/*.parquet
  - config_name: HTML+EEX
    data_files:
      - split: train
        path: data/HTML+EEX/*.parquet
  - config_name: HTML+ERB
    data_files:
      - split: train
        path: data/HTML+ERB/*.parquet
  - config_name: HTML+PHP
    data_files:
      - split: train
        path: data/HTML+PHP/*.parquet
  - config_name: HTML+Razor
    data_files:
      - split: train
        path: data/HTML+Razor/*.parquet
  - config_name: HTTP
    data_files:
      - split: train
        path: data/HTTP/*.parquet
  - config_name: HXML
    data_files:
      - split: train
        path: data/HXML/*.parquet
  - config_name: Hack
    data_files:
      - split: train
        path: data/Hack/*.parquet
  - config_name: Haml
    data_files:
      - split: train
        path: data/Haml/*.parquet
  - config_name: Handlebars
    data_files:
      - split: train
        path: data/Handlebars/*.parquet
  - config_name: Harbour
    data_files:
      - split: train
        path: data/Harbour/*.parquet
  - config_name: Haskell
    data_files:
      - split: train
        path: data/Haskell/*.parquet
  - config_name: Haxe
    data_files:
      - split: train
        path: data/Haxe/*.parquet
  - config_name: HiveQL
    data_files:
      - split: train
        path: data/HiveQL/*.parquet
  - config_name: HolyC
    data_files:
      - split: train
        path: data/HolyC/*.parquet
  - config_name: Hy
    data_files:
      - split: train
        path: data/Hy/*.parquet
  - config_name: HyPhy
    data_files:
      - split: train
        path: data/HyPhy/*.parquet
  - config_name: IDL
    data_files:
      - split: train
        path: data/IDL/*.parquet
  - config_name: IGOR_Pro
    data_files:
      - split: train
        path: data/IGOR_Pro/*.parquet
  - config_name: INI
    data_files:
      - split: train
        path: data/INI/*.parquet
  - config_name: IRC_log
    data_files:
      - split: train
        path: data/IRC_log/*.parquet
  - config_name: Idris
    data_files:
      - split: train
        path: data/Idris/*.parquet
  - config_name: Ignore_List
    data_files:
      - split: train
        path: data/Ignore_List/*.parquet
  - config_name: ImageJ_Macro
    data_files:
      - split: train
        path: data/ImageJ_Macro/*.parquet
  - config_name: Inform_7
    data_files:
      - split: train
        path: data/Inform_7/*.parquet
  - config_name: Inno_Setup
    data_files:
      - split: train
        path: data/Inno_Setup/*.parquet
  - config_name: Io
    data_files:
      - split: train
        path: data/Io/*.parquet
  - config_name: Ioke
    data_files:
      - split: train
        path: data/Ioke/*.parquet
  - config_name: Isabelle
    data_files:
      - split: train
        path: data/Isabelle/*.parquet
  - config_name: Isabelle_ROOT
    data_files:
      - split: train
        path: data/Isabelle_ROOT/*.parquet
  - config_name: J
    data_files:
      - split: train
        path: data/J/*.parquet
  - config_name: JAR_Manifest
    data_files:
      - split: train
        path: data/JAR_Manifest/*.parquet
  - config_name: JFlex
    data_files:
      - split: train
        path: data/JFlex/*.parquet
  - config_name: JSON
    data_files:
      - split: train
        path: data/JSON/*.parquet
  - config_name: JSON5
    data_files:
      - split: train
        path: data/JSON5/*.parquet
  - config_name: JSONLD
    data_files:
      - split: train
        path: data/JSONLD/*.parquet
  - config_name: JSON_with_Comments
    data_files:
      - split: train
        path: data/JSON_with_Comments/*.parquet
  - config_name: JSONiq
    data_files:
      - split: train
        path: data/JSONiq/*.parquet
  - config_name: Janet
    data_files:
      - split: train
        path: data/Janet/*.parquet
  - config_name: Jasmin
    data_files:
      - split: train
        path: data/Jasmin/*.parquet
  - config_name: Java
    data_files:
      - split: train
        path: data/Java/*.parquet
  - config_name: JavaScript
    data_files:
      - split: train
        path: data/JavaScript/*.parquet
  - config_name: JavaScript+ERB
    data_files:
      - split: train
        path: data/JavaScript+ERB/*.parquet
  - config_name: Java_Properties
    data_files:
      - split: train
        path: data/Java_Properties/*.parquet
  - config_name: Java_Server_Pages
    data_files:
      - split: train
        path: data/Java_Server_Pages/*.parquet
  - config_name: Jest_Snapshot
    data_files:
      - split: train
        path: data/Jest_Snapshot/*.parquet
  - config_name: JetBrains_MPS
    data_files:
      - split: train
        path: data/JetBrains_MPS/*.parquet
  - config_name: Jinja
    data_files:
      - split: train
        path: data/Jinja/*.parquet
  - config_name: Jison
    data_files:
      - split: train
        path: data/Jison/*.parquet
  - config_name: Jison_Lex
    data_files:
      - split: train
        path: data/Jison_Lex/*.parquet
  - config_name: Jolie
    data_files:
      - split: train
        path: data/Jolie/*.parquet
  - config_name: Jsonnet
    data_files:
      - split: train
        path: data/Jsonnet/*.parquet
  - config_name: Julia
    data_files:
      - split: train
        path: data/Julia/*.parquet
  - config_name: Jupyter_Notebook
    data_files:
      - split: train
        path: data/Jupyter_Notebook/*.parquet
  - config_name: KRL
    data_files:
      - split: train
        path: data/KRL/*.parquet
  - config_name: Kaitai_Struct
    data_files:
      - split: train
        path: data/Kaitai_Struct/*.parquet
  - config_name: KakouneScript
    data_files:
      - split: train
        path: data/KakouneScript/*.parquet
  - config_name: KiCad_Layout
    data_files:
      - split: train
        path: data/KiCad_Layout/*.parquet
  - config_name: KiCad_Legacy_Layout
    data_files:
      - split: train
        path: data/KiCad_Legacy_Layout/*.parquet
  - config_name: KiCad_Schematic
    data_files:
      - split: train
        path: data/KiCad_Schematic/*.parquet
  - config_name: Kit
    data_files:
      - split: train
        path: data/Kit/*.parquet
  - config_name: Kotlin
    data_files:
      - split: train
        path: data/Kotlin/*.parquet
  - config_name: Kusto
    data_files:
      - split: train
        path: data/Kusto/*.parquet
  - config_name: LFE
    data_files:
      - split: train
        path: data/LFE/*.parquet
  - config_name: LLVM
    data_files:
      - split: train
        path: data/LLVM/*.parquet
  - config_name: LOLCODE
    data_files:
      - split: train
        path: data/LOLCODE/*.parquet
  - config_name: LSL
    data_files:
      - split: train
        path: data/LSL/*.parquet
  - config_name: LTspice_Symbol
    data_files:
      - split: train
        path: data/LTspice_Symbol/*.parquet
  - config_name: LabVIEW
    data_files:
      - split: train
        path: data/LabVIEW/*.parquet
  - config_name: Lark
    data_files:
      - split: train
        path: data/Lark/*.parquet
  - config_name: Lasso
    data_files:
      - split: train
        path: data/Lasso/*.parquet
  - config_name: Latte
    data_files:
      - split: train
        path: data/Latte/*.parquet
  - config_name: Lean
    data_files:
      - split: train
        path: data/Lean/*.parquet
  - config_name: Less
    data_files:
      - split: train
        path: data/Less/*.parquet
  - config_name: Lex
    data_files:
      - split: train
        path: data/Lex/*.parquet
  - config_name: LigoLANG
    data_files:
      - split: train
        path: data/LigoLANG/*.parquet
  - config_name: LilyPond
    data_files:
      - split: train
        path: data/LilyPond/*.parquet
  - config_name: Limbo
    data_files:
      - split: train
        path: data/Limbo/*.parquet
  - config_name: Linker_Script
    data_files:
      - split: train
        path: data/Linker_Script/*.parquet
  - config_name: Linux_Kernel_Module
    data_files:
      - split: train
        path: data/Linux_Kernel_Module/*.parquet
  - config_name: Liquid
    data_files:
      - split: train
        path: data/Liquid/*.parquet
  - config_name: Literate_Agda
    data_files:
      - split: train
        path: data/Literate_Agda/*.parquet
  - config_name: Literate_CoffeeScript
    data_files:
      - split: train
        path: data/Literate_CoffeeScript/*.parquet
  - config_name: Literate_Haskell
    data_files:
      - split: train
        path: data/Literate_Haskell/*.parquet
  - config_name: LiveScript
    data_files:
      - split: train
        path: data/LiveScript/*.parquet
  - config_name: Logos
    data_files:
      - split: train
        path: data/Logos/*.parquet
  - config_name: Logtalk
    data_files:
      - split: train
        path: data/Logtalk/*.parquet
  - config_name: LookML
    data_files:
      - split: train
        path: data/LookML/*.parquet
  - config_name: LoomScript
    data_files:
      - split: train
        path: data/LoomScript/*.parquet
  - config_name: Lua
    data_files:
      - split: train
        path: data/Lua/*.parquet
  - config_name: M
    data_files:
      - split: train
        path: data/M/*.parquet
  - config_name: M4
    data_files:
      - split: train
        path: data/M4/*.parquet
  - config_name: M4Sugar
    data_files:
      - split: train
        path: data/M4Sugar/*.parquet
  - config_name: MATLAB
    data_files:
      - split: train
        path: data/MATLAB/*.parquet
  - config_name: MAXScript
    data_files:
      - split: train
        path: data/MAXScript/*.parquet
  - config_name: MLIR
    data_files:
      - split: train
        path: data/MLIR/*.parquet
  - config_name: MQL4
    data_files:
      - split: train
        path: data/MQL4/*.parquet
  - config_name: MQL5
    data_files:
      - split: train
        path: data/MQL5/*.parquet
  - config_name: MTML
    data_files:
      - split: train
        path: data/MTML/*.parquet
  - config_name: MUF
    data_files:
      - split: train
        path: data/MUF/*.parquet
  - config_name: Macaulay2
    data_files:
      - split: train
        path: data/Macaulay2/*.parquet
  - config_name: Makefile
    data_files:
      - split: train
        path: data/Makefile/*.parquet
  - config_name: Mako
    data_files:
      - split: train
        path: data/Mako/*.parquet
  - config_name: Markdown
    data_files:
      - split: train
        path: data/Markdown/*.parquet
  - config_name: Marko
    data_files:
      - split: train
        path: data/Marko/*.parquet
  - config_name: Mask
    data_files:
      - split: train
        path: data/Mask/*.parquet
  - config_name: Mathematica
    data_files:
      - split: train
        path: data/Mathematica/*.parquet
  - config_name: Maven_POM
    data_files:
      - split: train
        path: data/Maven_POM/*.parquet
  - config_name: Max
    data_files:
      - split: train
        path: data/Max/*.parquet
  - config_name: Mercury
    data_files:
      - split: train
        path: data/Mercury/*.parquet
  - config_name: Meson
    data_files:
      - split: train
        path: data/Meson/*.parquet
  - config_name: Metal
    data_files:
      - split: train
        path: data/Metal/*.parquet
  - config_name: Microsoft_Developer_Studio_Project
    data_files:
      - split: train
        path: data/Microsoft_Developer_Studio_Project/*.parquet
  - config_name: Microsoft_Visual_Studio_Solution
    data_files:
      - split: train
        path: data/Microsoft_Visual_Studio_Solution/*.parquet
  - config_name: MiniD
    data_files:
      - split: train
        path: data/MiniD/*.parquet
  - config_name: MiniYAML
    data_files:
      - split: train
        path: data/MiniYAML/*.parquet
  - config_name: Mint
    data_files:
      - split: train
        path: data/Mint/*.parquet
  - config_name: Mirah
    data_files:
      - split: train
        path: data/Mirah/*.parquet
  - config_name: Modelica
    data_files:
      - split: train
        path: data/Modelica/*.parquet
  - config_name: Modula-2
    data_files:
      - split: train
        path: data/Modula-2/*.parquet
  - config_name: Modula-3
    data_files:
      - split: train
        path: data/Modula-3/*.parquet
  - config_name: Module_Management_System
    data_files:
      - split: train
        path: data/Module_Management_System/*.parquet
  - config_name: Monkey
    data_files:
      - split: train
        path: data/Monkey/*.parquet
  - config_name: Monkey_C
    data_files:
      - split: train
        path: data/Monkey_C/*.parquet
  - config_name: Moocode
    data_files:
      - split: train
        path: data/Moocode/*.parquet
  - config_name: MoonScript
    data_files:
      - split: train
        path: data/MoonScript/*.parquet
  - config_name: Motoko
    data_files:
      - split: train
        path: data/Motoko/*.parquet
  - config_name: Motorola_68K_Assembly
    data_files:
      - split: train
        path: data/Motorola_68K_Assembly/*.parquet
  - config_name: Move
    data_files:
      - split: train
        path: data/Move/*.parquet
  - config_name: Muse
    data_files:
      - split: train
        path: data/Muse/*.parquet
  - config_name: Mustache
    data_files:
      - split: train
        path: data/Mustache/*.parquet
  - config_name: Myghty
    data_files:
      - split: train
        path: data/Myghty/*.parquet
  - config_name: NASL
    data_files:
      - split: train
        path: data/NASL/*.parquet
  - config_name: NCL
    data_files:
      - split: train
        path: data/NCL/*.parquet
  - config_name: NEON
    data_files:
      - split: train
        path: data/NEON/*.parquet
  - config_name: NL
    data_files:
      - split: train
        path: data/NL/*.parquet
  - config_name: NPM_Config
    data_files:
      - split: train
        path: data/NPM_Config/*.parquet
  - config_name: NSIS
    data_files:
      - split: train
        path: data/NSIS/*.parquet
  - config_name: NWScript
    data_files:
      - split: train
        path: data/NWScript/*.parquet
  - config_name: Nasal
    data_files:
      - split: train
        path: data/Nasal/*.parquet
  - config_name: Nearley
    data_files:
      - split: train
        path: data/Nearley/*.parquet
  - config_name: Nemerle
    data_files:
      - split: train
        path: data/Nemerle/*.parquet
  - config_name: NetLinx
    data_files:
      - split: train
        path: data/NetLinx/*.parquet
  - config_name: NetLinx+ERB
    data_files:
      - split: train
        path: data/NetLinx+ERB/*.parquet
  - config_name: NetLogo
    data_files:
      - split: train
        path: data/NetLogo/*.parquet
  - config_name: NewLisp
    data_files:
      - split: train
        path: data/NewLisp/*.parquet
  - config_name: Nextflow
    data_files:
      - split: train
        path: data/Nextflow/*.parquet
  - config_name: Nginx
    data_files:
      - split: train
        path: data/Nginx/*.parquet
  - config_name: Nim
    data_files:
      - split: train
        path: data/Nim/*.parquet
  - config_name: Ninja
    data_files:
      - split: train
        path: data/Ninja/*.parquet
  - config_name: Nit
    data_files:
      - split: train
        path: data/Nit/*.parquet
  - config_name: Nix
    data_files:
      - split: train
        path: data/Nix/*.parquet
  - config_name: Nu
    data_files:
      - split: train
        path: data/Nu/*.parquet
  - config_name: NumPy
    data_files:
      - split: train
        path: data/NumPy/*.parquet
  - config_name: Nunjucks
    data_files:
      - split: train
        path: data/Nunjucks/*.parquet
  - config_name: OCaml
    data_files:
      - split: train
        path: data/OCaml/*.parquet
  - config_name: ObjDump
    data_files:
      - split: train
        path: data/ObjDump/*.parquet
  - config_name: ObjectScript
    data_files:
      - split: train
        path: data/ObjectScript/*.parquet
  - config_name: Object_Data_Instance_Notation
    data_files:
      - split: train
        path: data/Object_Data_Instance_Notation/*.parquet
  - config_name: Objective-C
    data_files:
      - split: train
        path: data/Objective-C/*.parquet
  - config_name: Objective-C++
    data_files:
      - split: train
        path: data/Objective-C++/*.parquet
  - config_name: Objective-J
    data_files:
      - split: train
        path: data/Objective-J/*.parquet
  - config_name: Odin
    data_files:
      - split: train
        path: data/Odin/*.parquet
  - config_name: Omgrofl
    data_files:
      - split: train
        path: data/Omgrofl/*.parquet
  - config_name: Opa
    data_files:
      - split: train
        path: data/Opa/*.parquet
  - config_name: Opal
    data_files:
      - split: train
        path: data/Opal/*.parquet
  - config_name: OpenCL
    data_files:
      - split: train
        path: data/OpenCL/*.parquet
  - config_name: OpenEdge_ABL
    data_files:
      - split: train
        path: data/OpenEdge_ABL/*.parquet
  - config_name: OpenQASM
    data_files:
      - split: train
        path: data/OpenQASM/*.parquet
  - config_name: OpenRC_runscript
    data_files:
      - split: train
        path: data/OpenRC_runscript/*.parquet
  - config_name: OpenSCAD
    data_files:
      - split: train
        path: data/OpenSCAD/*.parquet
  - config_name: OpenStep_Property_List
    data_files:
      - split: train
        path: data/OpenStep_Property_List/*.parquet
  - config_name: OpenType_Feature_File
    data_files:
      - split: train
        path: data/OpenType_Feature_File/*.parquet
  - config_name: Open_Policy_Agent
    data_files:
      - split: train
        path: data/Open_Policy_Agent/*.parquet
  - config_name: Org
    data_files:
      - split: train
        path: data/Org/*.parquet
  - config_name: Ox
    data_files:
      - split: train
        path: data/Ox/*.parquet
  - config_name: Oxygene
    data_files:
      - split: train
        path: data/Oxygene/*.parquet
  - config_name: Oz
    data_files:
      - split: train
        path: data/Oz/*.parquet
  - config_name: P4
    data_files:
      - split: train
        path: data/P4/*.parquet
  - config_name: PEG.js
    data_files:
      - split: train
        path: data/PEG.js/*.parquet
  - config_name: PHP
    data_files:
      - split: train
        path: data/PHP/*.parquet
  - config_name: PLSQL
    data_files:
      - split: train
        path: data/PLSQL/*.parquet
  - config_name: PLpgSQL
    data_files:
      - split: train
        path: data/PLpgSQL/*.parquet
  - config_name: POV-Ray_SDL
    data_files:
      - split: train
        path: data/POV-Ray_SDL/*.parquet
  - config_name: Pan
    data_files:
      - split: train
        path: data/Pan/*.parquet
  - config_name: Papyrus
    data_files:
      - split: train
        path: data/Papyrus/*.parquet
  - config_name: Parrot
    data_files:
      - split: train
        path: data/Parrot/*.parquet
  - config_name: Parrot_Assembly
    data_files:
      - split: train
        path: data/Parrot_Assembly/*.parquet
  - config_name: Parrot_Internal_Representation
    data_files:
      - split: train
        path: data/Parrot_Internal_Representation/*.parquet
  - config_name: Pascal
    data_files:
      - split: train
        path: data/Pascal/*.parquet
  - config_name: Pawn
    data_files:
      - split: train
        path: data/Pawn/*.parquet
  - config_name: Pep8
    data_files:
      - split: train
        path: data/Pep8/*.parquet
  - config_name: Perl
    data_files:
      - split: train
        path: data/Perl/*.parquet
  - config_name: Pic
    data_files:
      - split: train
        path: data/Pic/*.parquet
  - config_name: Pickle
    data_files:
      - split: train
        path: data/Pickle/*.parquet
  - config_name: PicoLisp
    data_files:
      - split: train
        path: data/PicoLisp/*.parquet
  - config_name: PigLatin
    data_files:
      - split: train
        path: data/PigLatin/*.parquet
  - config_name: Pike
    data_files:
      - split: train
        path: data/Pike/*.parquet
  - config_name: PlantUML
    data_files:
      - split: train
        path: data/PlantUML/*.parquet
  - config_name: Pod
    data_files:
      - split: train
        path: data/Pod/*.parquet
  - config_name: Pod_6
    data_files:
      - split: train
        path: data/Pod_6/*.parquet
  - config_name: PogoScript
    data_files:
      - split: train
        path: data/PogoScript/*.parquet
  - config_name: Pony
    data_files:
      - split: train
        path: data/Pony/*.parquet
  - config_name: Portugol
    data_files:
      - split: train
        path: data/Portugol/*.parquet
  - config_name: PostCSS
    data_files:
      - split: train
        path: data/PostCSS/*.parquet
  - config_name: PostScript
    data_files:
      - split: train
        path: data/PostScript/*.parquet
  - config_name: PowerBuilder
    data_files:
      - split: train
        path: data/PowerBuilder/*.parquet
  - config_name: PowerShell
    data_files:
      - split: train
        path: data/PowerShell/*.parquet
  - config_name: Prisma
    data_files:
      - split: train
        path: data/Prisma/*.parquet
  - config_name: Processing
    data_files:
      - split: train
        path: data/Processing/*.parquet
  - config_name: Procfile
    data_files:
      - split: train
        path: data/Procfile/*.parquet
  - config_name: Proguard
    data_files:
      - split: train
        path: data/Proguard/*.parquet
  - config_name: Prolog
    data_files:
      - split: train
        path: data/Prolog/*.parquet
  - config_name: Promela
    data_files:
      - split: train
        path: data/Promela/*.parquet
  - config_name: Propeller_Spin
    data_files:
      - split: train
        path: data/Propeller_Spin/*.parquet
  - config_name: Protocol_Buffer
    data_files:
      - split: train
        path: data/Protocol_Buffer/*.parquet
  - config_name: Protocol_Buffer_Text_Format
    data_files:
      - split: train
        path: data/Protocol_Buffer_Text_Format/*.parquet
  - config_name: Public_Key
    data_files:
      - split: train
        path: data/Public_Key/*.parquet
  - config_name: Pug
    data_files:
      - split: train
        path: data/Pug/*.parquet
  - config_name: Puppet
    data_files:
      - split: train
        path: data/Puppet/*.parquet
  - config_name: PureBasic
    data_files:
      - split: train
        path: data/PureBasic/*.parquet
  - config_name: PureScript
    data_files:
      - split: train
        path: data/PureScript/*.parquet
  - config_name: Pure_Data
    data_files:
      - split: train
        path: data/Pure_Data/*.parquet
  - config_name: Python
    data_files:
      - split: train
        path: data/Python/*.parquet
  - config_name: Python_traceback
    data_files:
      - split: train
        path: data/Python_traceback/*.parquet
  - config_name: Q-Sharp
    data_files:
      - split: train
        path: data/Q-Sharp/*.parquet
  - config_name: QML
    data_files:
      - split: train
        path: data/QML/*.parquet
  - config_name: QMake
    data_files:
      - split: train
        path: data/QMake/*.parquet
  - config_name: Qt_Script
    data_files:
      - split: train
        path: data/Qt_Script/*.parquet
  - config_name: Quake
    data_files:
      - split: train
        path: data/Quake/*.parquet
  - config_name: R
    data_files:
      - split: train
        path: data/R/*.parquet
  - config_name: RAML
    data_files:
      - split: train
        path: data/RAML/*.parquet
  - config_name: RDoc
    data_files:
      - split: train
        path: data/RDoc/*.parquet
  - config_name: REALbasic
    data_files:
      - split: train
        path: data/REALbasic/*.parquet
  - config_name: REXX
    data_files:
      - split: train
        path: data/REXX/*.parquet
  - config_name: RMarkdown
    data_files:
      - split: train
        path: data/RMarkdown/*.parquet
  - config_name: RPC
    data_files:
      - split: train
        path: data/RPC/*.parquet
  - config_name: RPGLE
    data_files:
      - split: train
        path: data/RPGLE/*.parquet
  - config_name: RPM_Spec
    data_files:
      - split: train
        path: data/RPM_Spec/*.parquet
  - config_name: RUNOFF
    data_files:
      - split: train
        path: data/RUNOFF/*.parquet
  - config_name: Racket
    data_files:
      - split: train
        path: data/Racket/*.parquet
  - config_name: Ragel
    data_files:
      - split: train
        path: data/Ragel/*.parquet
  - config_name: Raku
    data_files:
      - split: train
        path: data/Raku/*.parquet
  - config_name: Rascal
    data_files:
      - split: train
        path: data/Rascal/*.parquet
  - config_name: Raw_token_data
    data_files:
      - split: train
        path: data/Raw_token_data/*.parquet
  - config_name: ReScript
    data_files:
      - split: train
        path: data/ReScript/*.parquet
  - config_name: Readline_Config
    data_files:
      - split: train
        path: data/Readline_Config/*.parquet
  - config_name: Reason
    data_files:
      - split: train
        path: data/Reason/*.parquet
  - config_name: ReasonLIGO
    data_files:
      - split: train
        path: data/ReasonLIGO/*.parquet
  - config_name: Rebol
    data_files:
      - split: train
        path: data/Rebol/*.parquet
  - config_name: Record_Jar
    data_files:
      - split: train
        path: data/Record_Jar/*.parquet
  - config_name: Red
    data_files:
      - split: train
        path: data/Red/*.parquet
  - config_name: Redcode
    data_files:
      - split: train
        path: data/Redcode/*.parquet
  - config_name: Redirect_Rules
    data_files:
      - split: train
        path: data/Redirect_Rules/*.parquet
  - config_name: Regular_Expression
    data_files:
      - split: train
        path: data/Regular_Expression/*.parquet
  - config_name: Ren-Py
    data_files:
      - split: train
        path: data/Ren-Py/*.parquet
  - config_name: RenderScript
    data_files:
      - split: train
        path: data/RenderScript/*.parquet
  - config_name: Rich_Text_Format
    data_files:
      - split: train
        path: data/Rich_Text_Format/*.parquet
  - config_name: Ring
    data_files:
      - split: train
        path: data/Ring/*.parquet
  - config_name: Riot
    data_files:
      - split: train
        path: data/Riot/*.parquet
  - config_name: RobotFramework
    data_files:
      - split: train
        path: data/RobotFramework/*.parquet
  - config_name: Roff
    data_files:
      - split: train
        path: data/Roff/*.parquet
  - config_name: Roff_Manpage
    data_files:
      - split: train
        path: data/Roff_Manpage/*.parquet
  - config_name: Rouge
    data_files:
      - split: train
        path: data/Rouge/*.parquet
  - config_name: RouterOS_Script
    data_files:
      - split: train
        path: data/RouterOS_Script/*.parquet
  - config_name: Ruby
    data_files:
      - split: train
        path: data/Ruby/*.parquet
  - config_name: Rust
    data_files:
      - split: train
        path: data/Rust/*.parquet
  - config_name: SAS
    data_files:
      - split: train
        path: data/SAS/*.parquet
  - config_name: SCSS
    data_files:
      - split: train
        path: data/SCSS/*.parquet
  - config_name: SELinux_Policy
    data_files:
      - split: train
        path: data/SELinux_Policy/*.parquet
  - config_name: SMT
    data_files:
      - split: train
        path: data/SMT/*.parquet
  - config_name: SPARQL
    data_files:
      - split: train
        path: data/SPARQL/*.parquet
  - config_name: SQF
    data_files:
      - split: train
        path: data/SQF/*.parquet
  - config_name: SQL
    data_files:
      - split: train
        path: data/SQL/*.parquet
  - config_name: SQLPL
    data_files:
      - split: train
        path: data/SQLPL/*.parquet
  - config_name: SRecode_Template
    data_files:
      - split: train
        path: data/SRecode_Template/*.parquet
  - config_name: SSH_Config
    data_files:
      - split: train
        path: data/SSH_Config/*.parquet
  - config_name: STAR
    data_files:
      - split: train
        path: data/STAR/*.parquet
  - config_name: STL
    data_files:
      - split: train
        path: data/STL/*.parquet
  - config_name: STON
    data_files:
      - split: train
        path: data/STON/*.parquet
  - config_name: SVG
    data_files:
      - split: train
        path: data/SVG/*.parquet
  - config_name: SWIG
    data_files:
      - split: train
        path: data/SWIG/*.parquet
  - config_name: Sage
    data_files:
      - split: train
        path: data/Sage/*.parquet
  - config_name: SaltStack
    data_files:
      - split: train
        path: data/SaltStack/*.parquet
  - config_name: Sass
    data_files:
      - split: train
        path: data/Sass/*.parquet
  - config_name: Scala
    data_files:
      - split: train
        path: data/Scala/*.parquet
  - config_name: Scaml
    data_files:
      - split: train
        path: data/Scaml/*.parquet
  - config_name: Scheme
    data_files:
      - split: train
        path: data/Scheme/*.parquet
  - config_name: Scilab
    data_files:
      - split: train
        path: data/Scilab/*.parquet
  - config_name: Self
    data_files:
      - split: train
        path: data/Self/*.parquet
  - config_name: ShaderLab
    data_files:
      - split: train
        path: data/ShaderLab/*.parquet
  - config_name: Shell
    data_files:
      - split: train
        path: data/Shell/*.parquet
  - config_name: ShellCheck_Config
    data_files:
      - split: train
        path: data/ShellCheck_Config/*.parquet
  - config_name: ShellSession
    data_files:
      - split: train
        path: data/ShellSession/*.parquet
  - config_name: Shen
    data_files:
      - split: train
        path: data/Shen/*.parquet
  - config_name: Sieve
    data_files:
      - split: train
        path: data/Sieve/*.parquet
  - config_name: Singularity
    data_files:
      - split: train
        path: data/Singularity/*.parquet
  - config_name: Slash
    data_files:
      - split: train
        path: data/Slash/*.parquet
  - config_name: Slice
    data_files:
      - split: train
        path: data/Slice/*.parquet
  - config_name: Slim
    data_files:
      - split: train
        path: data/Slim/*.parquet
  - config_name: SmPL
    data_files:
      - split: train
        path: data/SmPL/*.parquet
  - config_name: Smali
    data_files:
      - split: train
        path: data/Smali/*.parquet
  - config_name: Smalltalk
    data_files:
      - split: train
        path: data/Smalltalk/*.parquet
  - config_name: Smarty
    data_files:
      - split: train
        path: data/Smarty/*.parquet
  - config_name: Solidity
    data_files:
      - split: train
        path: data/Solidity/*.parquet
  - config_name: Soong
    data_files:
      - split: train
        path: data/Soong/*.parquet
  - config_name: SourcePawn
    data_files:
      - split: train
        path: data/SourcePawn/*.parquet
  - config_name: Spline_Font_Database
    data_files:
      - split: train
        path: data/Spline_Font_Database/*.parquet
  - config_name: Squirrel
    data_files:
      - split: train
        path: data/Squirrel/*.parquet
  - config_name: Stan
    data_files:
      - split: train
        path: data/Stan/*.parquet
  - config_name: Standard_ML
    data_files:
      - split: train
        path: data/Standard_ML/*.parquet
  - config_name: Starlark
    data_files:
      - split: train
        path: data/Starlark/*.parquet
  - config_name: Stata
    data_files:
      - split: train
        path: data/Stata/*.parquet
  - config_name: StringTemplate
    data_files:
      - split: train
        path: data/StringTemplate/*.parquet
  - config_name: Stylus
    data_files:
      - split: train
        path: data/Stylus/*.parquet
  - config_name: SubRip_Text
    data_files:
      - split: train
        path: data/SubRip_Text/*.parquet
  - config_name: SugarSS
    data_files:
      - split: train
        path: data/SugarSS/*.parquet
  - config_name: SuperCollider
    data_files:
      - split: train
        path: data/SuperCollider/*.parquet
  - config_name: Svelte
    data_files:
      - split: train
        path: data/Svelte/*.parquet
  - config_name: Swift
    data_files:
      - split: train
        path: data/Swift/*.parquet
  - config_name: SystemVerilog
    data_files:
      - split: train
        path: data/SystemVerilog/*.parquet
  - config_name: TI_Program
    data_files:
      - split: train
        path: data/TI_Program/*.parquet
  - config_name: TLA
    data_files:
      - split: train
        path: data/TLA/*.parquet
  - config_name: TOML
    data_files:
      - split: train
        path: data/TOML/*.parquet
  - config_name: TSQL
    data_files:
      - split: train
        path: data/TSQL/*.parquet
  - config_name: TSV
    data_files:
      - split: train
        path: data/TSV/*.parquet
  - config_name: TSX
    data_files:
      - split: train
        path: data/TSX/*.parquet
  - config_name: TXL
    data_files:
      - split: train
        path: data/TXL/*.parquet
  - config_name: Talon
    data_files:
      - split: train
        path: data/Talon/*.parquet
  - config_name: Tcl
    data_files:
      - split: train
        path: data/Tcl/*.parquet
  - config_name: Tcsh
    data_files:
      - split: train
        path: data/Tcsh/*.parquet
  - config_name: TeX
    data_files:
      - split: train
        path: data/TeX/*.parquet
  - config_name: Tea
    data_files:
      - split: train
        path: data/Tea/*.parquet
  - config_name: Terra
    data_files:
      - split: train
        path: data/Terra/*.parquet
  - config_name: Texinfo
    data_files:
      - split: train
        path: data/Texinfo/*.parquet
  - config_name: Text
    data_files:
      - split: train
        path: data/Text/*.parquet
  - config_name: TextMate_Properties
    data_files:
      - split: train
        path: data/TextMate_Properties/*.parquet
  - config_name: Textile
    data_files:
      - split: train
        path: data/Textile/*.parquet
  - config_name: Thrift
    data_files:
      - split: train
        path: data/Thrift/*.parquet
  - config_name: Turing
    data_files:
      - split: train
        path: data/Turing/*.parquet
  - config_name: Turtle
    data_files:
      - split: train
        path: data/Turtle/*.parquet
  - config_name: Twig
    data_files:
      - split: train
        path: data/Twig/*.parquet
  - config_name: TypeScript
    data_files:
      - split: train
        path: data/TypeScript/*.parquet
  - config_name: Type_Language
    data_files:
      - split: train
        path: data/Type_Language/*.parquet
  - config_name: Unified_Parallel_C
    data_files:
      - split: train
        path: data/Unified_Parallel_C/*.parquet
  - config_name: Unity3D_Asset
    data_files:
      - split: train
        path: data/Unity3D_Asset/*.parquet
  - config_name: Unix_Assembly
    data_files:
      - split: train
        path: data/Unix_Assembly/*.parquet
  - config_name: Uno
    data_files:
      - split: train
        path: data/Uno/*.parquet
  - config_name: UnrealScript
    data_files:
      - split: train
        path: data/UnrealScript/*.parquet
  - config_name: UrWeb
    data_files:
      - split: train
        path: data/UrWeb/*.parquet
  - config_name: V
    data_files:
      - split: train
        path: data/V/*.parquet
  - config_name: VBA
    data_files:
      - split: train
        path: data/VBA/*.parquet
  - config_name: VBScript
    data_files:
      - split: train
        path: data/VBScript/*.parquet
  - config_name: VCL
    data_files:
      - split: train
        path: data/VCL/*.parquet
  - config_name: VHDL
    data_files:
      - split: train
        path: data/VHDL/*.parquet
  - config_name: Vala
    data_files:
      - split: train
        path: data/Vala/*.parquet
  - config_name: Valve_Data_Format
    data_files:
      - split: train
        path: data/Valve_Data_Format/*.parquet
  - config_name: Velocity_Template_Language
    data_files:
      - split: train
        path: data/Velocity_Template_Language/*.parquet
  - config_name: Verilog
    data_files:
      - split: train
        path: data/Verilog/*.parquet
  - config_name: Vim_Help_File
    data_files:
      - split: train
        path: data/Vim_Help_File/*.parquet
  - config_name: Vim_Script
    data_files:
      - split: train
        path: data/Vim_Script/*.parquet
  - config_name: Vim_Snippet
    data_files:
      - split: train
        path: data/Vim_Snippet/*.parquet
  - config_name: Visual_Basic_.NET
    data_files:
      - split: train
        path: data/Visual_Basic_.NET/*.parquet
  - config_name: Volt
    data_files:
      - split: train
        path: data/Volt/*.parquet
  - config_name: Vue
    data_files:
      - split: train
        path: data/Vue/*.parquet
  - config_name: Vyper
    data_files:
      - split: train
        path: data/Vyper/*.parquet
  - config_name: Wavefront_Material
    data_files:
      - split: train
        path: data/Wavefront_Material/*.parquet
  - config_name: Wavefront_Object
    data_files:
      - split: train
        path: data/Wavefront_Object/*.parquet
  - config_name: WebAssembly
    data_files:
      - split: train
        path: data/WebAssembly/*.parquet
  - config_name: WebIDL
    data_files:
      - split: train
        path: data/WebIDL/*.parquet
  - config_name: WebVTT
    data_files:
      - split: train
        path: data/WebVTT/*.parquet
  - config_name: Web_Ontology_Language
    data_files:
      - split: train
        path: data/Web_Ontology_Language/*.parquet
  - config_name: Wget_Config
    data_files:
      - split: train
        path: data/Wget_Config/*.parquet
  - config_name: Whiley
    data_files:
      - split: train
        path: data/Whiley/*.parquet
  - config_name: Wikitext
    data_files:
      - split: train
        path: data/Wikitext/*.parquet
  - config_name: Win32_Message_File
    data_files:
      - split: train
        path: data/Win32_Message_File/*.parquet
  - config_name: Windows_Registry_Entries
    data_files:
      - split: train
        path: data/Windows_Registry_Entries/*.parquet
  - config_name: Witcher_Script
    data_files:
      - split: train
        path: data/Witcher_Script/*.parquet
  - config_name: Wollok
    data_files:
      - split: train
        path: data/Wollok/*.parquet
  - config_name: World_of_Warcraft_Addon_Data
    data_files:
      - split: train
        path: data/World_of_Warcraft_Addon_Data/*.parquet
  - config_name: Wren
    data_files:
      - split: train
        path: data/Wren/*.parquet
  - config_name: X10
    data_files:
      - split: train
        path: data/X10/*.parquet
  - config_name: XC
    data_files:
      - split: train
        path: data/XC/*.parquet
  - config_name: XCompose
    data_files:
      - split: train
        path: data/XCompose/*.parquet
  - config_name: XML
    data_files:
      - split: train
        path: data/XML/*.parquet
  - config_name: XML_Property_List
    data_files:
      - split: train
        path: data/XML_Property_List/*.parquet
  - config_name: XPages
    data_files:
      - split: train
        path: data/XPages/*.parquet
  - config_name: XProc
    data_files:
      - split: train
        path: data/XProc/*.parquet
  - config_name: XQuery
    data_files:
      - split: train
        path: data/XQuery/*.parquet
  - config_name: XS
    data_files:
      - split: train
        path: data/XS/*.parquet
  - config_name: XSLT
    data_files:
      - split: train
        path: data/XSLT/*.parquet
  - config_name: X_BitMap
    data_files:
      - split: train
        path: data/X_BitMap/*.parquet
  - config_name: X_Font_Directory_Index
    data_files:
      - split: train
        path: data/X_Font_Directory_Index/*.parquet
  - config_name: X_PixMap
    data_files:
      - split: train
        path: data/X_PixMap/*.parquet
  - config_name: Xojo
    data_files:
      - split: train
        path: data/Xojo/*.parquet
  - config_name: Xonsh
    data_files:
      - split: train
        path: data/Xonsh/*.parquet
  - config_name: Xtend
    data_files:
      - split: train
        path: data/Xtend/*.parquet
  - config_name: YAML
    data_files:
      - split: train
        path: data/YAML/*.parquet
  - config_name: YANG
    data_files:
      - split: train
        path: data/YANG/*.parquet
  - config_name: YARA
    data_files:
      - split: train
        path: data/YARA/*.parquet
  - config_name: YASnippet
    data_files:
      - split: train
        path: data/YASnippet/*.parquet
  - config_name: Yacc
    data_files:
      - split: train
        path: data/Yacc/*.parquet
  - config_name: Yul
    data_files:
      - split: train
        path: data/Yul/*.parquet
  - config_name: ZAP
    data_files:
      - split: train
        path: data/ZAP/*.parquet
  - config_name: ZIL
    data_files:
      - split: train
        path: data/ZIL/*.parquet
  - config_name: Zeek
    data_files:
      - split: train
        path: data/Zeek/*.parquet
  - config_name: ZenScript
    data_files:
      - split: train
        path: data/ZenScript/*.parquet
  - config_name: Zephir
    data_files:
      - split: train
        path: data/Zephir/*.parquet
  - config_name: Zig
    data_files:
      - split: train
        path: data/Zig/*.parquet
  - config_name: Zimpl
    data_files:
      - split: train
        path: data/Zimpl/*.parquet
  - config_name: cURL_Config
    data_files:
      - split: train
        path: data/cURL_Config/*.parquet
  - config_name: desktop
    data_files:
      - split: train
        path: data/desktop/*.parquet
  - config_name: dircolors
    data_files:
      - split: train
        path: data/dircolors/*.parquet
  - config_name: eC
    data_files:
      - split: train
        path: data/eC/*.parquet
  - config_name: edn
    data_files:
      - split: train
        path: data/edn/*.parquet
  - config_name: fish
    data_files:
      - split: train
        path: data/fish/*.parquet
  - config_name: hoon
    data_files:
      - split: train
        path: data/hoon/*.parquet
  - config_name: jq
    data_files:
      - split: train
        path: data/jq/*.parquet
  - config_name: kvlang
    data_files:
      - split: train
        path: data/kvlang/*.parquet
  - config_name: mIRC_Script
    data_files:
      - split: train
        path: data/mIRC_Script/*.parquet
  - config_name: mcfunction
    data_files:
      - split: train
        path: data/mcfunction/*.parquet
  - config_name: mupad
    data_files:
      - split: train
        path: data/mupad/*.parquet
  - config_name: nanorc
    data_files:
      - split: train
        path: data/nanorc/*.parquet
  - config_name: nesC
    data_files:
      - split: train
        path: data/nesC/*.parquet
  - config_name: ooc
    data_files:
      - split: train
        path: data/ooc/*.parquet
  - config_name: q
    data_files:
      - split: train
        path: data/q/*.parquet
  - config_name: reStructuredText
    data_files:
      - split: train
        path: data/reStructuredText/*.parquet
  - config_name: robots.txt
    data_files:
      - split: train
        path: data/robots.txt/*.parquet
  - config_name: sed
    data_files:
      - split: train
        path: data/sed/*.parquet
  - config_name: wdl
    data_files:
      - split: train
        path: data/wdl/*.parquet
  - config_name: wisp
    data_files:
      - split: train
        path: data/wisp/*.parquet
  - config_name: xBase
    data_files:
      - split: train
        path: data/xBase/*.parquet

The Stack v2

Dataset Description

• Homepage: https://www.bigcode-project.org/
• Repository: https://github.com/bigcode-project
• Paper: Link
• Point of Contact: contact@bigcode-project.org

The dataset consists of 4 versions:

• bigcode/the-stack-v2: the full "The Stack v2" dataset <-- you are here
• bigcode/the-stack-v2-dedup: based on the bigcode/the-stack-v2 but further near-deduplicated
• bigcode/the-stack-v2-train-full-ids: based on the bigcode/the-stack-v2-dedup dataset but further filtered with heuristics and spanning 600+ programming languages. The data is grouped into repositories.
• bigcode/the-stack-v2-train-smol-ids: based on the bigcode/the-stack-v2-dedup dataset but further filtered with heuristics and spanning 17 programming languages. The data is grouped into repositories.

These datasets only contain the SWHIDs to download the code files and not the content of the files itself. See examples below to see how to download content. We are working on making the training datasets available in the coming weeks.

The Stack v2 is significantly larger than v1:

	The Stack v1	The Stack v2
full	6.4TB	67.5TB
dedup	2.9TB	32.1TB
train (full)	~200B tokens	~900B tokens

Changelog

Release	Description
v2.1.0	Removed repositories that opted out before 2024-04-09. Removed unreachable/private repositories (according to SWH)
v2.0.1	Removed repositories that opted out before 2023-10-20. StarCoder2 was trained on this version
v2.0	Initial release of the Stack v2

Dataset Summary

The Stack v2 contains over 3B files in 600+ programming and markup languages. The dataset was created as part of the BigCode Project, an open scientific collaboration working on the responsible development of Large Language Models for Code (Code LLMs). The Stack serves as a pre-training dataset for Code LLMs, i.e., code-generating AI systems which enable the synthesis of programs from natural language descriptions as well as other from code snippets.

This dataset is derived from the Software Heritage archive, the largest public archive of software source code and accompanying development history. Software Heritage is an open, non profit initiative to collect, preserve, and share the source code of all publicly available software, launched by Inria, in partnership with UNESCO. We acknowledge Software Heritage for providing access to this invaluable resource. For more details, visit the Software Heritage website.

Languages

The dataset contains 658 languages. The full list can be found in the language stats table.

How to use it

from datasets import load_dataset

# full dataset (file IDs only)
ds = load_dataset("bigcode/the-stack-v2", split="train")

# specific language (e.g. Dockerfiles)
ds = load_dataset("bigcode/the-stack-v2", "Dockerfile", split="train")

# dataset streaming (will only download the data as needed)
ds = load_dataset("bigcode/the-stack-v2", streaming=True, split="train")
for sample in iter(ds): 
    print(sample) 

Downloading the file contents

The file contents are stored in the Software Heritage S3 bucket to ensure data compliance. Downloading data in bulk requires an agreement with SoftwareHeritage and INRIA as stated in the dataset agreement. Make sure to configure your environment with your AWS credentials.

pip install smart_open[s3]

import os
import boto3
from smart_open import open
from datasets import load_dataset

session = boto3.Session(
    aws_access_key_id=os.environ["AWS_ACCESS_KEY_ID"],
    aws_secret_access_key=os.environ["AWS_SECRET_ACCESS_KEY"])
s3 = session.client("s3")

def download_contents(blob_id, src_encoding):
    s3_url = f"s3://softwareheritage/content/{blob_id}"
    
    with open(s3_url, "rb", compression=".gz", transport_params={"client": s3}) as fin:
        content = fin.read().decode(src_encoding)
    
    return {"content": content}

ds = load_dataset("bigcode/the-stack-v2", split="train", streaming=True)
ds = ds.map(lambda row: download_contents(row["blob_id"], row["src_encoding"]))
for row in ds:
    print(row["content"])
    break

Dataset Structure

Data Fields

• blob_id (string): Software Heritage (SWH) ID of the file on AWS S3.
• directory_id (string): SWH ID of the root directory of the repository.
• path (string): The file path within the repository.
• content_id (string): SWH content ID.
• detected_licenses (string[]): List of licenses (SPDX) detected by ScanCode.
• license_type (string): Inferred license type (permissive or no_license).
• repo_name (string): Repository name on GitHub.
• snapshot_id (string): SWH snapshot ID.
• revision_id (string): SWH revision (commit) ID.
• branch_name (string): Repository branch name.
• visit_date (timestamp[ns]): SWH crawl (snapshot) timestamp.
• revision_date (timestamp[ns]): SWH revision (commit) timestamp.
• committer_date (timestamp[ns]): SWH revision (commit) timestamp reported by the committer.
• github_id (int64): GitHub identifier for the repository.
• star_events_count (int64): number of stars calculated from GHArchive events.
• fork_events_count (int64): number of forks calculated from GHArchive events.
• gha_license_id (string): GHArchive SPDX license identifier, None if the repo is missing.
• gha_event_created_at (timestamp[ns]): Timestamp of the latest event on GHArchive for this repository.
• gha_created_at (timestamp[ns]): Timestamp of repository creation on GitHub, None if the repo is missing.
• gha_language (string): Repository's primary programming language on GitHub, None if the repo is missing.
• src_encoding (string): Original encoding of the file content befre converting to UTF-8.
• language (string): Programming language of the file, detected by go-enry / linguist.
• is_vendor (bool): Indicator of vendor file (external library), detected by go-enry.
• is_generated (bool): Indicator of generated file (external library), detected by go-enry.
• length_bytes (int64): Length of the file content in UTF-8 bytes.
• extension (string): File extension.

Data Splits

The dataset has no splits and all data is loaded as train split by default. If you want to setup a custom train-test split beware that dataset contains a lot of near-duplicates which can cause leakage into the test split.

Dataset Creation

For more information on the dataset creation pipeline please refer to the technical report.

Curation Rationale

One of the challenges faced by researchers working on code LLMs is the lack of openness and transparency around the development of these systems. Most prior works described the high-level data collection process but did not release the training data. It is therefore difficult for other researchers to fully reproduce these models and understand what kind of pre-training data leads to high-performing code LLMs. By releasing an open large-scale code dataset we hope to make training of code LLMs more reproducible.

Source Data

Data Collection

3.28B unique files belonging to 104.2M github repositories were collected by traversing the Software Heritage 2023-09-06 graph dataset. Additional repository-level metadata was collected from GitHub Archive data up to 2023-09-14. The total uncompressed size of all files is 67.53TB. Near-deduplication was implemented in the pre-processing pipeline on top of exact deduplication. Roughly 40% of permissively licensed files were (near-)duplicates.

The following are not stored:

• Files that cannot contribute to training code: binary, empty, could not be decoded
• Files larger than 10MB

Training Datasets: For the training datasets the programming languages were filtered further to 17 and 600+ for the the-stack-v2-smol-ids and the-stack-v2-full-ids dataset, respecively. In addition, heuristics were applied to further increase the quality of the dataset. The code files are also grouped into repositories to allow to pretrain with full repository context. For more details see the technical report.

License detection

We extract repository-level license information from GH Archive for all repositories with matching names in the SWH dataset. When the repo-level license is not available, i.e., for 96.93% of repositories, we use the ScanCode Toolkit to detect file-level licenses as follows:

• Find all filenames that could contain a license (e.g., LICENSE, MIT.txt, Apache2.0) or contain a reference to the license (e.g., README.md, GUIDELINES);
• Apply ScanCode's license detection to the matching files and gather the SPDX IDs of the detected licenses;
• Propagate the detected licenses to all files that have the same base path within the repository as the license file.

The licenses we consider permissive are listed here. This list was compiled from the licenses approved by the Blue Oak Council, as well as licenses categorized as "Permissive" or "Public Domain" by ScanCode.

Who are the source language producers?

The source (code) language producers are users of GitHub that created unique repository names up until 2023-09-06 (cutoff date).

Personal and Sensitive Information

The released dataset may contain sensitive information such as emails, IP addresses, and API/ssh keys that have previously been published to public repositories on GitHub. Deduplication has helped to reduce the amount of sensitive data that may exist. In the event that the dataset contains personal information, researchers should only use public, non-personal information in support of conducting and publishing their open-access research. Personal information should not be used for spamming purposes, including sending unsolicited emails or selling of personal information. Complaints, removal requests, and "do not contact" requests can be sent to contact@bigcode-project.org.

Opting out of The Stack v2

We are giving developers the ability to have their code removed from the dataset upon request. The process for submitting and enacting removal requests will keep evolving throughout the project as we receive feedback and build up more data governance tools.

You can check if your code is in The Stack v2 with the following "Am I In The Stack?" Space. If you'd like to have your data removed from the dataset follow the instructions on GitHub.

Considerations for Using the Data

Social Impact of Dataset

The Stack v2 is an output of the BigCode Project. BigCode aims to be responsible by design and by default. The project is conducted in the spirit of Open Science, focused on the responsible development of LLMs for code.

With the release of The Stack v2, we aim to increase access, reproducibility, and transparency of code LLMs in the research community. Work to de-risk and improve on the implementation of ethical best practices of code LLMs is conducted in various BigCode working groups. The Legal, Ethics, and Governance working group has explored topics such as licensing (including copyleft and the intended use of permissively licensed code), attribution of generated code to original code, rights to restrict processing, the inclusion of Personally Identifiable Information (PII), and risks of malicious code, among other topics. This work is ongoing as of October 25th, 2022.

We expect code LLMs to enable people from diverse backgrounds to write higher quality code and develop low-code applications. Mission-critical software could become easier to maintain as professional developers are guided by code-generating systems on how to write more robust and efficient code. While the social impact is intended to be positive, the increased accessibility of code LLMs comes with certain risks such as over-reliance on the generated code and long-term effects on the software development job market.

A broader impact analysis relating to Code LLMs can be found in section 7 of this paper. An in-depth risk assessments for Code LLMs can be found in section 4 of this paper.

Discussion of Biases

The code collected from GitHub does not contain demographic information or proxy information about the demographics. However, it is not without risks, as the comments within the code may contain harmful or offensive language, which could be learned by the models.

Widely adopted programming languages like C and Javascript are overrepresented compared to niche programming languages like Julia and Scala. Some programming languages such as SQL, Batchfile, TypeScript are less likely to be permissively licensed (4% vs the average 10%). This may result in a biased representation of those languages. Permissively licensed files also tend to be longer.

The majority of natural language present in code from GitHub is English.

Other Known Limitations

One of the current limitations of The Stack v2 is that scraped HTML for websites may not be compliant with Web Content Accessibility Guidelines (WCAG). This could have an impact on HTML-generated code that may introduce web accessibility issues.

The training dataset could contain malicious code and/or the model could be used to generate malware or ransomware.

To the best of our knowledge, all files contained in the dataset are licensed with one of the permissive licenses (see list in Licensing information) or no license. The accuracy of license attribution is limited by the accuracy of GHArchive and ScanCode Toolkit. Any mistakes should be reported to BigCode Project for review and follow-up as needed.

Additional Information

Dataset Curators

1. Harm de Vries, ServiceNow Research, harm.devries@servicenow.com
2. Leandro von Werra, Hugging Face, leandro@huggingface.co

Licensing Information

The Stack v2 is a collection of source code from repositories with various licenses. Any use of all or part of the code gathered in The Stack v2 must abide by the terms of the original licenses, including attribution clauses when relevant. We facilitate this by providing provenance information for each data point.

The list of SPDX license identifiers included in the dataset can be found here.

Citation Information

@misc{lozhkov2024starcoder,
      title={StarCoder 2 and The Stack v2: The Next Generation}, 
      author={Anton Lozhkov and Raymond Li and Loubna Ben Allal and Federico Cassano and Joel Lamy-Poirier and Nouamane Tazi and Ao Tang and Dmytro Pykhtar and Jiawei Liu and Yuxiang Wei and Tianyang Liu and Max Tian and Denis Kocetkov and Arthur Zucker and Younes Belkada and Zijian Wang and Qian Liu and Dmitry Abulkhanov and Indraneil Paul and Zhuang Li and Wen-Ding Li and Megan Risdal and Jia Li and Jian Zhu and Terry Yue Zhuo and Evgenii Zheltonozhskii and Nii Osae Osae Dade and Wenhao Yu and Lucas Krauß and Naman Jain and Yixuan Su and Xuanli He and Manan Dey and Edoardo Abati and Yekun Chai and Niklas Muennighoff and Xiangru Tang and Muhtasham Oblokulov and Christopher Akiki and Marc Marone and Chenghao Mou and Mayank Mishra and Alex Gu and Binyuan Hui and Tri Dao and Armel Zebaze and Olivier Dehaene and Nicolas Patry and Canwen Xu and Julian McAuley and Han Hu and Torsten Scholak and Sebastien Paquet and Jennifer Robinson and Carolyn Jane Anderson and Nicolas Chapados and Mostofa Patwary and Nima Tajbakhsh and Yacine Jernite and Carlos Muñoz Ferrandis and Lingming Zhang and Sean Hughes and Thomas Wolf and Arjun Guha and Leandro von Werra and Harm de Vries},
      year={2024},
      eprint={2402.19173},
      archivePrefix={arXiv},
      primaryClass={cs.SE}
}