garnetrapture/evai-assets

EVAI Assets

Character, story and voice assets for subculture-style AI training and local character-chat applications. The payload is kept here instead of in the application repository so that the client stays small and the assets can be fetched on demand.

Contents

Path	Files	Size	Description
data/dataset/	285	31 MB	Per-character persona profiles, localized dialogue sets and intimacy patterns (JSON / JSONL)
data/eversoul-assets/	2,196	549 MB	Character portraits and skins, talk backgrounds, story UI frames (PNG)
data/story/	94	24 MB	Story index, main and bond story scripts with actor placement, stage layout, per-line voice/BGM references and endings (JSON)
data/story-media/voice/	63,328	~3.2 GB	Korean and Japanese voice lines (OGG Vorbis, mono 24 kHz)
data/story-media/video/	547	~80 MB	Story cutscenes (MP4)
data/story-media/bgm/	338	451 MB	Background music (OGG Vorbis, stereo 44.1 kHz) with a localized track index
data/story-media/sfx/	1,135	26 MB	Ambience and scene sound effects (OGG Vorbis, mono 32 kHz)

Layout

data/
  dataset/<character>.json
  eversoul-assets/
    backgrounds/talk/<background>.png
    spirits/<folder>/<skin>/<prefix>_1024.png
    ui/story/<element>.png
  story/
    index.json
    main/<chapter>.json
    love/<character>.json
  story-media/
    video/<kind>/<key>/<clip>.mp4
    voice/<kind>/<key>/<ko|ja>/V_[<speaker>_]<lineId>_<KR|JA>.ogg
    bgm/<track>.ogg
    bgm/index.json
    sfx/<effect>.ogg

Story scripts carry per-line actor ids, portrait slots, background keys and branch choices. Each line also names the media it needs, so nothing has to be guessed from a filename pattern:

{
  "id": 250200100,
  "text": { "ko": "...", "en": "...", "zh_tw": "...", "zh_cn": "..." },
  "speaker": 5020,
  "voice": { "ko": "V_Adrianne_250200100_KR", "ja": "V_Adrianne_250200100_JA" },
  "bgm": "BGM_Lazycloud_LOOP",
  "amb": "SFX_AMB_2D_Room01",
  "background": "Talk_BG_Townstreet"
}

voice values resolve under data/story-media/voice/<kind>/<key>/<language>/, bgm under data/story-media/bgm/, amb under data/story-media/sfx/, and background under data/eversoul-assets/backgrounds/talk/. data/story-media/bgm/index.json lists every track with its localized title and arrangement credit taken from the game's own string tables.

Usage

# everything
hf download garnetrapture/evai-assets --repo-type dataset --local-dir .

# text only
hf download garnetrapture/evai-assets --repo-type dataset --local-dir . \
  --include "data/dataset/*" "data/story/*"

# single voice language
hf download garnetrapture/evai-assets --repo-type dataset --local-dir . \
  --include "data/story-media/voice/*/*/ko/*"

# music only
hf download garnetrapture/evai-assets --repo-type dataset --local-dir . \
  --include "data/story-media/bgm/*"

The repository is public, so plain HTTP works without a token:

curl -sfL -O https://huggingface.co/datasets/garnetrapture/evai-assets/resolve/main/data/story/index.json

Notes

• Text is available in Korean, Japanese, English and Simplified Chinese; voice lines exist in Korean and Japanese only. Roughly 55% of script lines are voiced; the rest are narration.
• Audio is transcoded to OGG Vorbis to keep the download small — voice at mono 24 kHz, music at stereo 44.1 kHz, effects at mono 32 kHz. It is meant for playback and TTS reference, not for mastering.
• Assets are derived from game bundles and are published for research and non-commercial personal use. Rights to the original artwork, text and recordings remain with their owners.