Instructions to use Earlychildhoodeducation/EleMo-V1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- llama.cpp
How to use Earlychildhoodeducation/EleMo-V1 with llama.cpp:
Install (macOS, Linux)
curl -LsSf https://llama.app/install.sh | sh # Start a local OpenAI-compatible server with a web UI: llama serve -hf Earlychildhoodeducation/EleMo-V1:Q4_K_M # Run inference directly in the terminal: llama cli -hf Earlychildhoodeducation/EleMo-V1:Q4_K_M
Install from WinGet (Windows)
winget install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama serve -hf Earlychildhoodeducation/EleMo-V1:Q4_K_M # Run inference directly in the terminal: llama cli -hf Earlychildhoodeducation/EleMo-V1:Q4_K_M
Use pre-built binary
# Download pre-built binary from: # https://github.com/ggerganov/llama.cpp/releases # Start a local OpenAI-compatible server with a web UI: ./llama-server -hf Earlychildhoodeducation/EleMo-V1:Q4_K_M # Run inference directly in the terminal: ./llama-cli -hf Earlychildhoodeducation/EleMo-V1:Q4_K_M
Build from source code
git clone https://github.com/ggerganov/llama.cpp.git cd llama.cpp cmake -B build cmake --build build -j --target llama-server llama-cli # Start a local OpenAI-compatible server with a web UI: ./build/bin/llama-server -hf Earlychildhoodeducation/EleMo-V1:Q4_K_M # Run inference directly in the terminal: ./build/bin/llama-cli -hf Earlychildhoodeducation/EleMo-V1:Q4_K_M
Use Docker
docker model run hf.co/Earlychildhoodeducation/EleMo-V1:Q4_K_M
- LM Studio
- Jan
- vLLM
How to use Earlychildhoodeducation/EleMo-V1 with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "Earlychildhoodeducation/EleMo-V1" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "Earlychildhoodeducation/EleMo-V1", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker
docker model run hf.co/Earlychildhoodeducation/EleMo-V1:Q4_K_M
- Ollama
How to use Earlychildhoodeducation/EleMo-V1 with Ollama:
ollama run hf.co/Earlychildhoodeducation/EleMo-V1:Q4_K_M
- Unsloth Desktop
- Docker Model Runner
How to use Earlychildhoodeducation/EleMo-V1 with Docker Model Runner:
docker model run hf.co/Earlychildhoodeducation/EleMo-V1:Q4_K_M
- Lemonade
How to use Earlychildhoodeducation/EleMo-V1 with Lemonade:
Pull the model
# Download Lemonade from https://lemonade-server.ai/ lemonade pull Earlychildhoodeducation/EleMo-V1:Q4_K_M
Run and chat with the model
lemonade run user.EleMo-V1-Q4_K_M
List all available models
lemonade list
- Atomic Chat
Why gated and english please!
Why is the model gated, and what criteria do you use to grant access to daycare centers, municipalities, or universities?
And the most important question: Can't you also fine-tune a version of the model in English? You'd make a large part of the world happy with it in daycare centers :)
Thanks for granting access.
Hi! Thanks for reaching out and for your interest in EleMo!
Here are the answers to your questions regarding model access and internationalization:
- Why is EleMo-v2 Gated & What are the Access Criteria?
We intentionally chose a gated access model to uphold two main principles: Responsible AI deployment and Pedagogical Integrity.
Protecting the Target Domain: EleMo-v2 is explicitly fine-tuned for early childhood documentation following strict, evidence-based frameworks (like Margaret Carr’s Learning Stories) and an anti-bias dataset design. Gating prevents the weights from being indiscriminately scraped, rebranded, or integrated into generic, unverified commercial wrapper apps that might bypass safety guardrails or compromise child data privacy.
Verification & Zero-Cloud Compliance: We want to ensure that organizations deploying EleMo understand the necessity of local, data-sovereign environments (100% On-Device / Zero-Cloud). Intimate developmental data of young children must never touch external servers.
Access Criteria: Municipalities & Public IT Providers: Verification as a public sector entity or regional IT provider aiming to deploy local AI infrastructure for early childhood institutions.
Daycare Centers & Academic Institutions: Verified educational background or active research projects in elementary pedagogy, early childhood education, or specialized LLM fine-tuning/evaluations.
Commercial Entities: Case-by-case review to ensure compliance with our data sovereignty guidelines and educational intent.
- Is an English Version Planned?
You hit a very important point! The core dilemma, pedagogical staff being overburdened by documentation tasks, is a global issue, not just a German one.
Short Answer: Yes, an English version is definitely on our roadmap for early 2027.
Why it takes a bit of time: EleMo is not just translated; it relies on a mathematically balanced, highly curated dataset tailored to specific pedagogical frameworks, local curricula, and cultural nuances.
Framework Adaptation: While Margaret Carr’s methodology originated in New Zealand and is widely used in English-speaking countries (e.g., Learning Stories), we need to align the English dataset with local curriculum standards (such as EYFS in the UK or state-level frameworks in North America).
Bias Mitigation Matrix: We need to rebuild and balance the multi-dimensional training matrix (gender representation, inclusive markers, emotional realism, and cultural background) specifically for English-language contexts to avoid systemic biases.
Once the current German v2 release stabilizes and our fine-tuning pipeline for the international dataset is fully calibrated, the English version will be next!
Thanks again for joining the journey toward sovereign, ethical AI in early childhood education. Feel free to share your feedback once you've tested the model! 😉