AkAiNp/nepal-oral-packs-open

Nepal Oral Open Packs

Public educational packs for the Nepal Oral Train app and NGO classroom use (AkAiNp).

• Versioned lesson packs (meanings, tags, provisional text)
• Seed / model-speaker audio only when license is open
• Language origin metadata (family, homeland, script status, varieties)
• Multi-track over time; starts empty until steward-approved open content exists

This is what the mobile app may download read-only after a pack is published.
It is not a dump of every phone contribution.

Pipeline

Partner seed / reviewed open clips
  → pack version in API
  → steward “publish open pack”
  → export → AkAiNp/nepal-oral-packs-open
  → app / Spaces fetch manifest + audio

Intended uses

OK	Not OK
In-app lessons, phrasebooks	Assuming all rows are the same language
Open pedagogy	Ignoring track_id / variety_id
Fine-tune TTS/ASR only on rows with audio + open license	Training on metadata-only rows as speech
Research citing language origin fields	Re-identifying speakers

Data fields (target)

Field	Type	Notes
pack_id / pack_version	string	Bundle id + semver
track_id / item_id	string
kind	string	word / phrase / …
meaning_en / meaning_ne	string
provisional_text	string	Orthography may be experimental
tags	list[string]
family / variety_id / script_status	string	From track
homelands	list[string]
origin_summary_en	string
audio	Audio	Seed audio if present
duration_ms / sha256
content_license	string	e.g. cc-by-4.0
published_at	string	ISO-8601

Status

Bootstrap: card only (or empty). Real open packs are added after partner steward approval. Demo fixtures live in AkAiNp/nepal-oral-demo.

Consent & ethics

• Only material with open license and steward publish action
• Guardian-mode material only if the speaker chose open and steward approved pack inclusion
• Private family recordings never land here by default

License

Default CC-BY-4.0 unless a pack requires stricter terms. Per-row content_license may vary.

Related

• AkAiNp/nepal-oral-demo
• AkAiNp/nepal-oral-speech-research (gated)