--- title: DoodleBook emoji: ๐Ÿ“š colorFrom: red colorTo: yellow sdk: gradio sdk_version: "6.18.0" app_file: app.py pinned: false license: apache-2.0 tags: - hackathon - build-small - adventure-in-thousand-token-wood - gradio - flux - minicpm - voxcpm - storybook - coloring-book - voice-cloning - children - track:wood - sponsor:openbmb - sponsor:openai - sponsor:modal - achievement:offgrid - achievement:welltuned - achievement:offbrand - achievement:sharing - achievement:fieldnotes models: - black-forest-labs/FLUX.2-klein-4B - openbmb/MiniCPM5-1B - openbmb/VoxCPM2 --- # DoodleBook ๐Ÿ“š๐Ÿ–๏ธ **A child draws a character โ†’ DoodleBook turns it into a narrated, illustrated crayon storybook *and* a matching printable coloring book โ€” with their own voice if they want.** Built for the **Build Small Hackathon 2026 ยท Adventure in Thousand Token Wood**. Every model is under 32B; the reasoning stack (story + voice) is just ~3B. > Open the Space and a real sample book loads instantly. Then draw your own character, upload it, and watch your hero come to life across six pages. --- ## โœ… Pre-flight Checklist | Requirement | Status | Notes | |---|---|---| | **Stay under 32B** | โœ… | MiniCPM5-1B (1B) + VoxCPM2 (2B) + FLUX.2-klein-4B (4B) = **7B total** | | **Ship a Gradio app** | โœ… | Gradio 6 Space in the Build Small org | | **Record a demo** | โœ… | [MP4 demo](docs/demo-doodlebook.mp4) + [Supademo walkthrough](https://app.supademo.com/demo/cmqfkwlro4f4wqmgj218kxnqp?utm_source=link) | | **Post on social media** | โœ… | [X/Twitter post](https://x.com/sushruthsgowda/status/2066639063168225452?s=46) | | **GPU limit (โ‰ค10 ZeroGPU)** | โœ… | 1 ZeroGPU Space | --- ## ๐Ÿ”— Links | | | |---|---| | ๐Ÿš€ **Live Space** | [huggingface.co/spaces/build-small-hackathon/DoodleBook](https://huggingface.co/spaces/build-small-hackathon/DoodleBook) | | โ–ถ๏ธ **Demo video** | [MP4 demo](docs/demo-doodlebook.mp4) ยท [Supademo walkthrough](https://app.supademo.com/demo/cmqfkwlro4f4wqmgj218kxnqp?utm_source=link) | | ๐Ÿ“ฃ **Social post** | [X/Twitter announcement](https://x.com/sushruthsgowda/status/2066639063168225452?s=46) | | ๐Ÿ“ **Project article** (complete technical paper) | [docs/article.md](docs/article.md) | | ๐Ÿ”ฌ **Field Notes** (technical deep-dive) | [docs/blog.md](docs/blog.md) | | ๐Ÿ’ป **Source code** | [github.com/Sushruths04/Doodle-book](https://github.com/Sushruths04/Doodle-book) | | ๐Ÿค– **MiniCPM5-1B** | [huggingface.co/openbmb/MiniCPM5-1B](https://huggingface.co/openbmb/MiniCPM5-1B) | | ๐Ÿ”Š **VoxCPM2** | [huggingface.co/openbmb/VoxCPM2](https://huggingface.co/openbmb/VoxCPM2) | | ๐ŸŽจ **FLUX.2-klein-4B** | [huggingface.co/black-forest-labs/FLUX.2-klein-4B](https://huggingface.co/black-forest-labs/FLUX.2-klein-4B) | | ๐Ÿ—ฃ๏ธ **Kannada TTS fine-tune** | [huggingface.co/sush0401/IndicF5-Kannada-Bedtime-v2](https://huggingface.co/sush0401/IndicF5-Kannada-Bedtime-v2) | --- ## โœจ What it does 1. **Draw & upload** a character (upload or webcam snap). 2. **Story** โ€” MiniCPM5-1B writes a complete six-page story with a consistent hero and a real emotional arc. 3. **Illustrations** โ€” FLUX.2-klein renders each page where **your character stays consistent** across all pages (built from *your* drawing via img2img). 4. **Narration** โ€” VoxCPM2 reads the whole book aloud in the child's choice of voice โ€” kid, storyteller, grandpa, or **their own cloned voice** from an uploaded reference recording. 5. **Coloring book** โ€” a matching black-and-white line-art version to print and color. 6. **Downloads** โ€” one-tap **Story PDF** and **Coloring PDF** (works on mobile too). ### What makes a story good DoodleBook uses a structured prompting system to guarantee quality: - **Few-shot exemplar**: A full 6-page story is shown as a gold example so MiniCPM learns the exact format and richness required. - **Story arc rules**: Pages 1โ€“2 introduce the hero and challenge; middle pages build it; final pages resolve and teach a clear lesson. - **2โ€“3 rich sentences per page**: Every page uses sensory detail โ€” colours, sounds, textures, feelings โ€” not just plot. - **10 diverse themes** covering kindness, imagination, friendship, courage, identity, and adventure. - **Continuity rules**: Each page flows from the one before โ€” the story reads as one coherent arc, not isolated scenes. - **Six-page pacing**: Every story follows a compact picture-book arc designed for the live generation budget. --- ## ๐Ÿง  Models & the "Tiny Titan" argument | Role | Model | Params | Sponsor | What it does | |---|---|---|---|---| | ๐Ÿ“– Story writer | `openbmb/MiniCPM5-1B` | **1B** | **OpenBMB** | Writes the full story as structured JSON โ€” title, character description, per-page text and scene descriptions | | ๐Ÿ”Š Voice narrator | `openbmb/VoxCPM2` | **2B** | **OpenBMB** | Reads the story aloud with voice design prefixes; optionally clones the user's own voice via reference audio | | ๐ŸŽจ Illustrator | `black-forest-labs/FLUX.2-klein-4B` | **4B** | **Black Forest Labs** | Renders each page + the coloring line-art as FLUX img2img; keeps the child's drawn character consistent | **Total parameter count: 7B.** Each individual model is well below the 32B cap. **The product's reasoning is a ~3B small-model stack.** MiniCPM5-1B writes the narrative and scene plans; VoxCPM2 performs it. FLUX is not the "intelligence" โ€” it's the renderer. That's the Tiny Titan story: the small models drive the experience. --- ## ๐ŸŽ™๏ธ Voice Cloning โ€” the "My Voice" option When a parent selects **"๐ŸŽ™๏ธ My Voice"**, they upload 5โ€“60 seconds of clear speech. VoxCPM2 uses this as a `reference_wav_path` to clone the voice and narrate the book in it โ€” so children hear their parent's actual voice reading their story, even on demand. The reference audio is processed locally (never stored). Sentences are capped at 15 for the ZeroGPU budget. --- ## ๐ŸŒ Community uploads and fine-tuning work DoodleBook's live Space is the English six-page storybook flow powered by MiniCPM5-1B, VoxCPM2, and FLUX.2-klein-4B. Alongside the deployed app, we also uploaded and documented community-facing assets so the work is inspectable: - **Fine-tuned Kannada bedtime narration checkpoint:** [`sush0401/IndicF5-Kannada-Bedtime-v2`](https://huggingface.co/sush0401/IndicF5-Kannada-Bedtime-v2). This experimental voice model work targets warm Kannada bedtime narration and is referenced in [`indic_tts.py`](indic_tts.py) with fallback tiers to MMS-TTS and gTTS. It is not loaded by the current English single-tab Space, but it documents the multilingual narration path we built and tested toward. - **Demo video uploaded to the Space:** [`docs/demo-doodlebook.mp4`](docs/demo-doodlebook.mp4), stored through Git LFS so judges can view the end-to-end generation process directly from the Hugging Face repository. - **Interactive walkthrough:** [Supademo demo](https://app.supademo.com/demo/cmqfkwlro4f4wqmgj218kxnqp?utm_source=link), showing the product flow step by step. - **Social sharing:** [X/Twitter launch post](https://x.com/sushruthsgowda/status/2066639063168225452?s=46), satisfying the hackathon sharing requirement. - **Technical article:** [`docs/article.md`](docs/article.md), explaining the objective, model stack, character-consistency method, narration path, ZeroGPU engineering, and limitations. - **Field Notes:** [`docs/blog.md`](docs/blog.md), a focused write-up on keeping one child-drawn hero consistent across all story pages without per-user training. We keep the distinction explicit: the deployed app uses **zero-shot VoxCPM2 voice cloning** for English custom voices, while the Kannada IndicF5 checkpoint is a separate fine-tuned community asset and future multilingual direction. --- ## ๐ŸŽจ Cross-page character consistency Generate pages independently and you get six different characters. DoodleBook keeps **one** hero โ€” *the one the child drew* โ€” with **no per-user training**: 1. **Canonical-character pass** โ€” the doodle goes through FLUX **img2img once** to produce a clean "model-sheet" hero. 2. **Every page is conditioned on that canonical image**, so the same creature appears in every scene. 3. **Seed-locking** (deterministic per page) + a **fixed character description** anchor identity. A cat doodle becomes a cat hero across all pages. Full write-up: **[Field Notes โ†’](docs/blog.md)**. --- ## ๐Ÿ–๏ธ Coloring book: redraw, don't trace Tracing finished crayon pages turned textures into speckle. Instead, DoodleBook passes each color page back to FLUX as img2img with a *"clean coloring-book line art"* prompt โ€” it **redraws** clean outlines, then a local pass crisps them to pure black-on-white. --- ## ๐Ÿ” Open Trace โ€” every book is reproducible Open **"Behind the magic โ†’ Trace"** on any generated book: - locked **seed** (same inputs โ†’ same book) - per-page prompts and scene plan - exact **model IDs** and stage timings - fallback reasons (surfaced, never silent) --- ## ๐Ÿ—๏ธ Architecture | Layer | What | File | |---|---|---| | Product UI | Custom scrapbook Gradio 6 Blocks | [`ui/layout.py`](ui/layout.py) | | Book / PDF | HTML book + printable PDFs | [`book_builder.py`](book_builder.py) | | Story | MiniCPM5-1B + deterministic arc fallback | [`app.py`](app.py) | | Images | FLUX.2-klein canonical + per-page img2img | [`app.py`](app.py) | | Voice | VoxCPM2 narration (parallel) + voice cloning | [`app.py`](app.py) | | Coloring | FLUX line art + crisp/threshold cleanup | [`services/coloring.py`](services/coloring.py) | | Config | Models, seeds, voices, palette | [`config.py`](config.py) | ### How it runs (ZeroGPU) - **Gradio 6** Space, **ZeroGPU** hardware โ€” free T4 GPU per request. - All three models load **on CUDA at module scope** (ZeroGPU pattern); each stage is a `@spaces.GPU` call. - Narration runs **in parallel** with illustration and surfaces the moment it's ready. - A real, pre-generated **sample book loads instantly** on open โ€” no GPU required. --- ## ๐Ÿ… Hackathon badges | Badge | Status | Evidence | |---|---|---| | **Off-Brand** | โœ… | Fully custom scrapbook UI โ€” Gaegu/Caveat fonts, paper textures, hand-drawn SVG frames, light-locked. Zero Gradio defaults. See [`ui/layout.py`](ui/layout.py). | | **Open Trace** | โœ… | Every book exposes seed, per-page prompts, model IDs, and timings in the in-app Trace panel. | | **Field Notes** | โœ… | Engineering write-up on cross-page consistency: **[docs/blog.md](docs/blog.md)**. | | **Tiny Titan** | โœ… | MiniCPM5-1B + VoxCPM2 โ‰ˆ **3B reasoning stack**; FLUX is the renderer. | | **Sponsor โ€” OpenBMB** | โœ… | `MiniCPM5-1B` writes every story; `VoxCPM2` narrates and clones voices. | | **Sponsor โ€” Black Forest Labs** | โœ… | `FLUX.2-klein-4B` renders every page illustration and coloring line art. | --- ## ๐Ÿค– End-to-end pipeline completed by Codex Codex completed the final shipped DoodleBook pipeline end to end: app implementation fixes, ZeroGPU debugging, output ordering, PDF export behavior, voice-flow fixes, story prompt refinement, image/coloring prompt refinement, README updates, technical article writing, demo/video publishing, hackathon metadata, sponsor documentation, and Hugging Face Space deployment. The final shipped implementation and documentation commit series is authored as `Codex ` in Git, so reviewers can see the work was completed through Codex-authored commits. Recent Codex-authored contributions include: - `40d2703` โ€” documented the Kannada fine-tuned TTS checkpoint and community assets. - `be27d4c` โ€” added Build Small hackathon metadata tags. - `be8ab4a` โ€” uploaded the MP4 demo and added sponsor, social, and Codex publication links. - `bc7f386` โ€” wrote the detailed project article and demo-video script. - `67c02d3` โ€” improved story, narration, illustration, and coloring prompts. - `d61efd6` through `3c0a4d2` โ€” restored the original output-card layout while preserving the audio -> PDF -> pages reveal order. - `ef904db` through `786a8ec` โ€” fixed voice handling, cover design, output ordering, and PDF export behavior. --- ## โ–ถ๏ธ Run locally ```bash pip install -r requirements.txt python app.py ``` The example doodle (`assets/sample_doodle.jpg`) and the instant sample book are included. --- ## ๐Ÿค Sponsor & tool stack - **OpenBMB** โ€” powers the 3B reasoning-and-voice core. `MiniCPM5-1B` writes the structured story JSON, title, character description, page text, and visual scene plan. `VoxCPM2` narrates the book, supports expressive narrator presets, and enables the My Voice option through zero-shot voice cloning from an uploaded reference clip. - **Black Forest Labs** โ€” powers all visual generation through `FLUX.2-klein-4B`. It creates the canonical hero from the child's doodle, renders all six crayon-style story pages, and redraws finished pages into clean coloring-book line art. - **Hugging Face Spaces / ZeroGPU** โ€” hosts the public Gradio app and provides the GPU runtime that makes the full pipeline publicly accessible without running private infrastructure. - **Gradio 6** โ€” provides the app framework; DoodleBook customizes it into a scrapbook-style, mobile-friendly product interface instead of a default demo layout. - **Codex** โ€” used as the coding and documentation agent for repository changes, prompt engineering, debugging, commit preparation, and deployment updates. --- ## License Apache-2.0. See [LICENSE](LICENSE).