transformers
🤗 Transformers: the model-definition framework for state-of-the-art machine learning models in text, vision, audio, and multimodal models, for both inference and training.
- 169k
- Python
- Apache-2.0
- 2 h ago
Netflix for lucid dreams
Repo ContextLab/dream-stream, read from GitHub 3 d ago. ContextLab has not joined Take My Repo.
ContextLab/dream-stream is a public repo on GitHub by ContextLab. Netflix for lucid dreams. It declares no license on GitHub. Public does not mean open source: without a license, nobody is granted the right to reuse it. It is written mainly in TypeScript, has 4 stars and last changed on 2026-01-20. Nobody has reviewed it on Take My Repo yet, which says nothing about its quality. Read from GitHub on 2026-11-19.
In the maintainer's words. Shown from GitHub; Take My Repo does not keep a copy.
Lucid Dream Induction Through Intelligent Sleep-Aware Audio Prompting
dream_stream is a cross-platform application that delivers personalized audio "dream prompts" at optimal moments during your sleep cycle. By integrating with fitness trackers and wearables, the app detects sleep stages in real-time and plays soothing, AI-generated audio cues designed to enhance dream recall and induce lucid dreaming.
dream_stream combines sleep science, wearable technology, and AI-generated audio to create a seamless lucid dreaming experience:
| Platform | Status | Notes |
|---|---|---|
| Web | Live | PWA at context-lab.com/dream-stream |
| Android | In Progress | Expo-based companion app |
| iOS | In Progress | Expo-based companion app |
| WearOS | Planned | Primary sleep detection device |
| watchOS | Planned | Primary sleep detection device |
| macOS | Planned | Desktop companion |
| Windows | Planned | Desktop companion |
| Linux (Ubuntu) | Planned | Desktop companion |
dream_stream integrates with multiple sleep tracking sources:
| Source | Sleep Stages | Real-Time Capability | Integration Method |
|---|---|---|---|
| WearOS (Local) | Yes (via HRV analysis) | High | Health Services SDK |
| Apple Watch | Yes | Near Real-Time | HealthKit Background Delivery |
| Fitbit | Yes | Batch (post-sync) | Web API + Webhooks |
| Garmin | Yes | High (via SDK) | Connect IQ SDK |
| Oura Ring | Yes (5-min intervals) | Batch | REST API + Webhooks |
| Asleep.ai | Yes | 5-min intervals | Audio-based SDK (no wearable) |
Primary Strategy: Local WearOS/watchOS app using heart rate variability (HRV) analysis for low-latency REM detection.
Audio is generated automatically in CI using Edge TTS:
┌─────────────────────────────────────────────────────────────┐
│ AUDIO PIPELINE │
├─────────────────────────────────────────────────────────────┤
│ 1. DREAM NARRATIVES │
│ └─> Stored in lib/mockData.ts with [PAUSE] markers │
│ │
│ 2. VOICE SYNTHESIS (CI) │
│ └─> Edge TTS (en-US-JennyNeural, -10% rate, -5Hz pitch) │
│ └─> Free, no API keys required │
│ │
│ 3. AUDIO PROCESSING │
│ └─> FFmpeg for Opus encoding (64kbps) │
│ └─> 45-second pause markers for lucid exploration │
│ │
│ 4. CACHING │
│ └─> GitHub Actions caches based on mockData.ts hash │
│ └─> Regenerates only when narratives change │
│ │
│ 5. DELIVERY │
│ └─> Served from GitHub Pages │
│ └─> public/audio/dreams/{id}_full.opus │
└─────────────────────────────────────────────────────────────┘
Current Voice: en-US-JennyNeural - calm, soothing US English voice
Music Styles:
┌─────────────────────────────────────────────────────────────┐
│ ARCHITECTURE │
├─────────────────────────────────────────────────────────────┤
│ │
│ ┌──────────────────────────────────────────────────────┐ │
│ │ SHARED CORE (KMP) │ │
│ │ • Business Logic (Sleep algorithms, prompt mgmt) │ │
│ │ • Data Layer (SQLDelight, Ktor HTTP client) │ │
│ │ • Domain Models │ │
│ └──────────────────────────────────────────────────────┘ │
│ │ │
│ ┌──────────────────┼──────────────────┐ │
│ ▼ ▼ ▼ │
│ ┌─────────────┐ ┌─────────────┐ ┌─────────────┐ │
│ │ MOBILE │ │ WEARABLE │ │ DESKTOP │ │
│ │ (Compose) │ │ (Native) │ │ (Compose) │ │
│ │ │ │ │ │ │ │
│ │ Android/iOS │ │ WearOS/ │ │ macOS/Win/ │ │
│ │ │ │ watchOS │ │ Linux │ │
│ └─────────────┘ └─────────────┘ └─────────────┘ │
│ │
│ ┌──────────────────────────────────────────────────────┐ │
│ │ WEB (Compose Web) │ │
│ └──────────────────────────────────────────────────────┘ │
│ │
├─────────────────────────────────────────────────────────────┤
│ BACKEND SERVICES │
│ • Audio Generation Service (ElevenLabs, Mubert, FFmpeg) │
│ • Dream Processing Service (Whisper, LLM) │
│ • User Data Sync (Supabase / Firebase) │
│ • Push Notifications │
└─────────────────────────────────────────────────────────────┘
Why KMP?
| Component | Technology | Purpose |
|---|---|---|
| UI Framework | Compose Multiplatform | Shared UI across platforms |
| Local Database | SQLDelight | Offline-first prompt storage |
| Networking | Ktor | API communication |
| Audio Playback | Media3 (ExoPlayer) | Background audio with media session |
| Sleep Detection | Health Services API | WearOS sensor access |
| TTS | ElevenLabs / Kokoro-82M | Voice synthesis |
| Music Gen | Mubert API | Procedural ambient music |
| Backend | Supabase | Auth, database, storage |
| LLM | Claude / GPT-4 | Dream-to-prompt transformation |
dream_stream targets specific sleep stages for maximum effectiveness:
| Sleep Stage | Brain Waves | Dream Activity | Prompt Strategy |
|---|---|---|---|
| Light (N1/N2) | Theta (4-8 Hz) | Hypnagogic imagery | Gentle awareness cues |
| REM | Mixed (like waking) | Vivid dreams | Reality check prompts |
| Deep (N3) | Delta (0.5-4 Hz) | Minimal | Avoid interruption |
Timing Algorithm:
Optional frequency entrainment to support lucid dreaming:
| Frequency | Brain State | Use Case |
|---|---|---|
| 4 Hz (Theta) | Light sleep, meditation | Pre-sleep relaxation |
| 40 Hz (Gamma) | Lucid dreaming state | During REM prompts |
| 2 Hz (Delta) | Deep sleep | Avoid during prompts |
dream-stream/
├── app/ # Expo Router screens
│ ├── (tabs)/ # Tab navigation (home, search, favorites, profile)
│ ├── auth/ # Login/signup screens
│ ├── dream/ # Dream detail and launch screens
│ └── sleep/ # Sleep tracking screen
├── components/ # React components
│ ├── ui/ # Base UI (Button, Card, Input, Text)
│ └── *.tsx # Feature components (DreamCard, DreamPlayer, etc.)
├── hooks/ # React hooks (useAuth, useDreams, useFavorites, etc.)
├── services/ # Business logic and API calls
├── lib/ # Utilities, constants, and mock data
├── theme/ # Design system tokens
├── types/ # TypeScript type definitions
├── scripts/ # Build scripts (audio generation)
├── supabase/ # Database migrations and config
├── public/ # Static assets (audio files)
└── notes/ # Session notes and documentation
# Clone the repository
git clone https://github.com/ContextLab/dream-stream.git
cd dream-stream
# Install dependencies
npm install
# Run web development server
npm run web
# Run on iOS simulator
npm run ios
# Run on Android emulator
npm run android
# Type check
npm run typecheck
# Create Python virtual environment
python -m venv .venv
source .venv/bin/activate
# Install dependencies
pip install edge-tts
# Generate audio for all dreams
python scripts/generate_audio.py
# Generate audio for specific dream
python scripts/generate_audio.py --dream dream-1
# List available voices
python scripts/generate_audio.py --list-voices
# Backend (optional - app works with mock data)
EXPO_PUBLIC_SUPABASE_URL=your_project_url
EXPO_PUBLIC_SUPABASE_ANON_KEY=your_anon_key
We welcome contributions! Please see CONTRIBUTING.md for guidelines.
This project is licensed under the MIT License - see the LICENSE file for details.
Sweet dreams are made of code.
ContextLab/dream-stream is a public repo on GitHub by ContextLab. Netflix for lucid dreams.
It declares no license on GitHub. Public does not mean open source: without a license, nobody is granted the right to reuse it.
Its maker tagged it for macOS, Windows and Linux, or as Docker or cross-platform. That is declared by the maker; Take My Repo has not tested it.
Open "Take it" on its page: Take My Repo writes an install plan for your machine and hands it to your AI agent (Claude Code, Codex or Cursor), which asks you before each step. Take My Repo has not tested it on any machine.
Its last change on GitHub was on 2026-01-20. Take My Repo does not judge from that whether it is maintained.
Nobody has reviewed it on Take My Repo yet, which says nothing about its quality.
🤗 Transformers: the model-definition framework for state-of-the-art machine learning models in text, vision, audio, and multimodal models, for both inference and training.

VoiceStudio is the open-source, fully-local ElevenLabs alternative — voice cloning, voice design, video dubbing, dictation, transcription & audiobook creation in 646 languages.

利用 AI 大模型和自动化工作流,根据主题或关键词一键生成高清短视频。Generate HD short videos from a topic or keyword with an automated AI workflow.

A Python toolbox for gaining geometric insights into high-dimensional data
Course materials for Dartmouth course: Human Memory (PSYC 51.09)
Course materials for Dartmouth Course: Storytelling with Data (PSYC 81.09).
Course materials for PSYC 132: Introduction to Programming for Psychological Scientists
Short undergraduate course taught at University of Pennsylvania on computational and theoretical neuroscience. Provides an introduction to programming in MATLAB, single-neuron models, ion channel models, basic neural networks, and neural decoding.
Course materials for PSYC 11: Laboratory in Psychological Science, Dartmouth College (Instructor: Jeremy Manning)
This page is also available as structured data, with the same sources and dates.
/api/v1/resources/gh/ContextLab/dream-streamTake My Repo has not tested this on your machine. Requirements come from the project's own documentation.