Datasets:

Sqls1
/

dota2-particles

Error code:   StreamingRowsError
Exception:    CastError
Message:      Couldn't cast
particles/dev/empty_particle.vpcf: struct<_class: string, m_nBehaviorVersion: int64, m_nMaxParticles: int64, m_controlPointConfiguratio (... 39 chars omitted)
  child 0, _class: string
  child 1, m_nBehaviorVersion: int64
  child 2, m_nMaxParticles: int64
  child 3, m_controlPointConfigurations: list<item: struct<m_name: string>>
      child 0, item: struct<m_name: string>
          child 0, m_name: string
m_controlPointConfigurations: null
m_nMaxParticles: int64
m_PreEmissionOperators: list<item: struct<_class: string, m_nChildControlPoint: int64, m_nNumControlPoints: int64, m_nFirstS (... 44 chars omitted)
  child 0, item: struct<_class: string, m_nChildControlPoint: int64, m_nNumControlPoints: int64, m_nFirstSourcePoint: (... 32 chars omitted)
      child 0, _class: string
      child 1, m_nChildControlPoint: int64
      child 2, m_nNumControlPoints: int64
      child 3, m_nFirstSourcePoint: int64
      child 4, m_bSetOrientation: bool
_class: string
m_nBehaviorVersion: int64
m_Children: list<item: struct<m_ChildRef: string>>
  child 0, item: struct<m_ChildRef: string>
      child 0, m_ChildRef: string
to
{'_class': Value('string'), 'm_nBehaviorVersion': Value('int64'), 'm_nMaxParticles': Value('int64'), 'm_controlPointConfigurations': List({'m_name': Value('string'), 'm_drivers': List(Json(decode=True)), 'm_previewState': {'m_previewModel': Value('string'), 'm_nModSpecificData': Value('int64'), 'm_sequenceName': Value('string'), 'm_hitboxSetName': Value('string')}}), 'm_Children': List({'m_ChildRef': Value('string')}), 'm_PreEmissionOperators': List({'_class': Value('string'), 'm_nChildControlPoint': Value('int64'), 'm_nNumControlPoints': Value('int64'), 'm_nFirstSourcePoint': Value('int64'), 'm_bSetOrientation': Value('bool')})}
because column names don't match
Traceback:    Traceback (most recent call last):
                File "/src/services/worker/src/worker/utils.py", line 149, in get_rows_or_raise
                  return get_rows(
                      dataset=dataset,
                  ...<4 lines>...
                      column_names=column_names,
                  )
                File "/src/libs/libcommon/src/libcommon/utils.py", line 272, in decorator
                  return func(*args, **kwargs)
                File "/src/services/worker/src/worker/utils.py", line 129, in get_rows
                  rows_plus_one = list(itertools.islice(safe_iter(ds, dataset=dataset), rows_max_number + 1))
                File "/src/services/worker/src/worker/utils.py", line 544, in safe_iter
                  yield from ds.decode(False) if ds.features else ds
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 3052, in __iter__
                  for key, example in ex_iterable:
                                      ^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2559, in __iter__
                  for key, pa_table in self._iter_arrow():
                                       ~~~~~~~~~~~~~~~~^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2584, in _iter_arrow
                  for key, pa_table in self.ex_iterable._iter_arrow():
                                       ~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 558, in _iter_arrow
                  for key, pa_table in iterator:
                                       ^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 440, in _iter_arrow
                  for key, pa_table in self.generate_tables_fn(**gen_kwags):
                                       ~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 361, in _generate_tables
                  self._cast_table(pa_table, json_field_paths=json_field_paths),
                  ~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 132, in _cast_table
                  pa_table = table_cast(pa_table, self.info.features.arrow_schema)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2381, in table_cast
                  return cast_table_to_schema(table, schema)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2309, in cast_table_to_schema
                  raise CastError(
                  ...<3 lines>...
                  )
              datasets.table.CastError: Couldn't cast
              particles/dev/empty_particle.vpcf: struct<_class: string, m_nBehaviorVersion: int64, m_nMaxParticles: int64, m_controlPointConfiguratio (... 39 chars omitted)
                child 0, _class: string
                child 1, m_nBehaviorVersion: int64
                child 2, m_nMaxParticles: int64
                child 3, m_controlPointConfigurations: list<item: struct<m_name: string>>
                    child 0, item: struct<m_name: string>
                        child 0, m_name: string
              m_controlPointConfigurations: null
              m_nMaxParticles: int64
              m_PreEmissionOperators: list<item: struct<_class: string, m_nChildControlPoint: int64, m_nNumControlPoints: int64, m_nFirstS (... 44 chars omitted)
                child 0, item: struct<_class: string, m_nChildControlPoint: int64, m_nNumControlPoints: int64, m_nFirstSourcePoint: (... 32 chars omitted)
                    child 0, _class: string
                    child 1, m_nChildControlPoint: int64
                    child 2, m_nNumControlPoints: int64
                    child 3, m_nFirstSourcePoint: int64
                    child 4, m_bSetOrientation: bool
              _class: string
              m_nBehaviorVersion: int64
              m_Children: list<item: struct<m_ChildRef: string>>
                child 0, item: struct<m_ChildRef: string>
                    child 0, m_ChildRef: string
              to
              {'_class': Value('string'), 'm_nBehaviorVersion': Value('int64'), 'm_nMaxParticles': Value('int64'), 'm_controlPointConfigurations': List({'m_name': Value('string'), 'm_drivers': List(Json(decode=True)), 'm_previewState': {'m_previewModel': Value('string'), 'm_nModSpecificData': Value('int64'), 'm_sequenceName': Value('string'), 'm_hitboxSetName': Value('string')}}), 'm_Children': List({'m_ChildRef': Value('string')}), 'm_PreEmissionOperators': List({'_class': Value('string'), 'm_nChildControlPoint': Value('int64'), 'm_nNumControlPoints': Value('int64'), 'm_nFirstSourcePoint': Value('int64'), 'm_bSetOrientation': Value('bool')})}
              because column names don't match

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

Dota 2 Particle Effects

Every particle system of Dota 2 (abilities, cosmetics, units, buildings, maps, the interface, events) as JSON, with the textures they draw and the engine's defaults for the fields a system leaves out. 98,083 files, 1.2 GB. See them played as the game does on Dotadex.

What's inside

Folder	What	Files	Size
`particles/econ/`	cosmetics' effects	50,753	352 MB
`particles/units/`	heroes' and units' effects	21,471	112 MB
`particles/bundles/`	effects in one file	7,642	406 MB
`particles/models/`	model effects	4,967	8 MB
`particles/textures/`	textures	2,936	266 MB
`particles/events/`	events	1,511	21 MB
`particles/ui/`	the interface	1,471	23 MB
`particles/items_fx/`	items	733	10 MB
`particles/neutral_fx/`	neutral creeps and items	613	4 MB
`particles/items2_fx/`	items	491	2 MB
`particles/ui_mouseactions/`	clicks and pings	490	4 MB
`particles/generic_gameplay/`	gameplay	374	2 MB
`particles/items4_fx/`	items	350	4 MB
`particles/items5_fx/`	items	269	3 MB

Path	What
`particles/<path>.json`	One system (`particles/<path>.vpcf` in the game) as its KeyValues3 give it
`particles/bundles/<path>.json`	An effect heroes and items show, with every system it spawns, in one file
`particles/textures/<path>.png`, `.json`	A texture particles draw as an atlas, and its sheet (sequences, frames)
`particles/<path>.vsnap.json`	A snapshot: points placed on a model, by stream
`particles/schema.json`	The engine's particle classes with their fields' defaults
`particles/model-effects.json`, `map-effects.json`, `map-cameras.json`	Effects models and maps start on their own; the interface scenes' cameras

Positions are in the game's units (inches) and axes (z up).

How to use it

One file, straight from its link (works in a browser, an <img>, fetch; served with CORS):

https://huggingface.co/datasets/Sqls1/dota2-particles/resolve/main/particles/bundles/units/heroes/hero_invoker/invoker_sun_strike.json

Python (huggingface_hub):

from huggingface_hub import hf_hub_download, snapshot_download

# one file
path = hf_hub_download("Sqls1/dota2-particles", "particles/bundles/units/heroes/hero_invoker/invoker_sun_strike.json", repo_type="dataset")

# a whole folder
snapshot_download("Sqls1/dota2-particles", repo_type="dataset", allow_patterns="particles/units/heroes/hero_invoker/*", local_dir="dota2-particles")

Command line (pip install -U huggingface_hub):

hf download Sqls1/dota2-particles --repo-type dataset --include "particles/units/heroes/hero_invoker/*" --local-dir dota2-particles

Find anything: files.json at the root lists every file with its size, one download instead of walking the folders:

const { files } = await (await fetch("https://huggingface.co/datasets/Sqls1/dota2-particles/resolve/main/files.json")).json(); // [[path, size], ...]

How it stays up to date

Exported from the game files by particles.js (the schema by particle-schema.js). The tracker checks Steam every six hours; after a game update it exports only what changed and removes what the game removed. Everything at the game's own paths, so a file you know from the game is easy to find.

Made for Dotadex, a site that shows every Dota 2 hero and cosmetic in 3D as the game does, and open to everyone.

The whole family

dota2-data: every script and text of the game as JSON (items, heroes, abilities, 28 languages)
dota2-images: every picture of the interface: icons, portraits, loading screens
dota2-videos: every video of the interface: animated hero portraits, trailers
dota2-models: every 3D model: heroes, cosmetics, units, buildings, props
dota2-textures: every texture of the 3D models and maps, on its own
dota2-particles: every particle effect system and its textures (this one)
dota2-sounds: every sound: voice lines, abilities, music, announcers
dota2-maps: every map and scene: the main map, events, backgrounds, MVP screens

License and credit

Everything here is Valve's content (and the Dota 2 Workshop authors' for cosmetics), taken from the game files and shared for reference, research, fan sites, art and other non-commercial community projects. Dota 2 is a registered trademark of Valve Corporation. This is an unofficial, fan-made dataset, not affiliated with or endorsed by Valve. If you use it, please credit Valve and Dota 2.

Downloads last month: -

Total file size:

1.28 GB