README.md

metadata

license: apache-2.0
language:
  - ps
  - en
tags:
  - pashto
  - pukhto
  - pushto
  - asr
  - tts
  - nlp
  - machine-translation
  - language-resources
  - low-resource-languages
  - speech-recognition

Pashto Language Resources Hub (Pukhto/Pashto)

Open-source repository for Pashto language technology resources: datasets, models, benchmarks, ASR, TTS, NLP, and machine translation (MT).

This project curates verified Pashto resources and maintains reproducible tooling for discovery, validation, and documentation.

Start Here

• Main resource search: Pashto Resource Search
• Project site: Pashto Language Resources Hub
• GitHub repository: Musawer1214/pashto-language-resources
• Hugging Face mirror: Musawer14/pashto-language-resources

If You Searched For

This repository is relevant to these search intents:

• Pashto datasets
• Pashto ASR model
• Pashto TTS resources
• Pashto NLP benchmark
• Pashto machine translation resources
• Pukhto language technology
• Pushto AI resources

Current Scope

• Build open Pashto datasets, benchmarks, and model references for ASR, TTS, NLP, and MT.
• Track practical tools, apps, and academic papers for Pashto integration in technology.
• Keep everything transparent, reproducible, and contribution-friendly.

Resource System

Machine-readable and searchable resource pipeline:

• Canonical catalog: resources/catalog/resources.json
• Catalog schema: resources/schema/resource.schema.json
• Candidate feed (auto-generated): resources/catalog/pending_candidates.json
• Search UI source: docs/search/index.html
• Search data export: docs/search/resources.json
• Resource index docs: docs/resource_catalog.md
• Automation docs: docs/resource_automation.md
• Cycle runbook: docs/resource_cycle_runbook.md

How New Resources Are Added

1. Auto discovery runs daily from .github/workflows/resource_sync.yml and updates resources/catalog/pending_candidates.json in a review PR.
2. Manual review checks quality, Pashto evidence, and license compatibility before promoting entries into resources/catalog/resources.json with status: verified.
3. Regeneration and validation runs python scripts/validate_resource_catalog.py and python scripts/generate_resource_views.py, then commits generated updates.

Shortcut wrapper:

• python scripts/run_resource_cycle.py --limit 25

Quickstart

python -m pip install -e ".[dev]"
python scripts/validate_resource_catalog.py
python scripts/generate_resource_views.py
python scripts/check_links.py
python -m pytest -q

Discoverability And SEO

• Playbook: docs/discoverability_seo.md
• Docs hub: docs/README.md
• Resource search page: docs/search/index.html
• Citation metadata: CITATION.cff
• Platform sync policy: docs/platform_sync_policy.md
• GitHub topics checklist: docs/github_topics_checklist.md
• Backlink strategy: docs/backlink_strategy.md
• Intent page: Pashto datasets
• Intent page: Pashto ASR
• Intent page: Pashto TTS
• Release notes: v0.1.1
• Release notes: v0.1.2

Documentation Map

• Purpose: PROJECT_PURPOSE.md
• Contributing: CONTRIBUTING.md
• Roadmap: ROADMAP.md
• Governance: GOVERNANCE.md
• License policy: LICENSE_POLICY.md
• Changelog: CHANGELOG.md
• Community: community/COMMUNICATION.md
• Docs hub: docs/README.md
• Resource index: docs/resource_catalog.md
• Resource automation: docs/resource_automation.md

Resource Sections

• Datasets: resources/datasets/README.md
• Models: resources/models/README.md
• Benchmarks: resources/benchmarks/README.md
• Tools: resources/tools/README.md
• Papers: resources/papers/README.md
• Projects: resources/projects/README.md
• Code: resources/codes/README.md

Workspaces

• data/: datasets, curation, metadata, quality
• asr/: ASR baselines and experiments
• tts/: TTS baselines and experiments
• benchmarks/: benchmark sets and evaluation
• experiments/: reproducible run cards
• apps/desktop/: user-facing integration references
• models/: model layout and release conventions