Instructions to use matteoangeloni/EduRaccoon with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use matteoangeloni/EduRaccoon with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="matteoangeloni/EduRaccoon")# pip install -U transformers accelerate # Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("matteoangeloni/EduRaccoon") model = AutoModelForCausalLM.from_pretrained("matteoangeloni/EduRaccoon", device_map="auto") - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- llama.cpp
How to use matteoangeloni/EduRaccoon 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 matteoangeloni/EduRaccoon # Run inference directly in the terminal: llama cli -hf matteoangeloni/EduRaccoon
Install from WinGet (Windows)
winget install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama serve -hf matteoangeloni/EduRaccoon # Run inference directly in the terminal: llama cli -hf matteoangeloni/EduRaccoon
Build from source code
git clone https://github.com/ggml-org/llama.cpp.git cd llama.cpp cmake -B build cmake --build build -j --target llama-app # Start a local OpenAI-compatible server with a web UI: ./build/bin/llama serve -hf matteoangeloni/EduRaccoon # Run inference directly in the terminal: ./build/bin/llama cli -hf matteoangeloni/EduRaccoon
Use Docker
docker model run hf.co/matteoangeloni/EduRaccoon
- Llama new
- LM Studio
- Jan
- vLLM
How to use matteoangeloni/EduRaccoon with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "matteoangeloni/EduRaccoon" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "matteoangeloni/EduRaccoon", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker
docker model run hf.co/matteoangeloni/EduRaccoon
- SGLang
How to use matteoangeloni/EduRaccoon with SGLang:
Install from pip and serve model
# Install SGLang from pip: pip install sglang # Start the SGLang server: python3 -m sglang.launch_server \ --model-path "matteoangeloni/EduRaccoon" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "matteoangeloni/EduRaccoon", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }'Use Docker images
docker run --gpus all \ --shm-size 32g \ -p 30000:30000 \ -v ~/.cache/huggingface:/root/.cache/huggingface \ --env "HF_TOKEN=<secret>" \ --ipc=host \ lmsysorg/sglang:latest \ python3 -m sglang.launch_server \ --model-path "matteoangeloni/EduRaccoon" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "matteoangeloni/EduRaccoon", "prompt": "Once upon a time,", "max_tokens": 512, "temperature": 0.5 }' - Ollama
How to use matteoangeloni/EduRaccoon with Ollama:
ollama run hf.co/matteoangeloni/EduRaccoon
- Unsloth Desktop
- Docker Model Runner
How to use matteoangeloni/EduRaccoon with Docker Model Runner:
docker model run hf.co/matteoangeloni/EduRaccoon
- Lemonade
How to use matteoangeloni/EduRaccoon with Lemonade:
Pull the model
# Download Lemonade from https://lemonade-server.ai/ lemonade pull matteoangeloni/EduRaccoon
Run and chat with the model
lemonade run user.EduRaccoon-{{QUANT_TAG}}List all available models
lemonade list
- Atomic Chat
π¦ EduRaccoon β Educational AI Assistant
EduRaccoon Γ¨ un modello linguistico di grandi dimensioni fine-tunato a partire da LLaMA 3.1 8B (Unsloth), progettato come assistente educativo multilingue.
Il modello Γ¨ in grado di rispondere nella lingua in cui viene interpellato (es. italiano, inglese, spagnolo, franceseβ¦) e di fornire risposte strutturate, chiare e didattiche a domande di carattere educativo.
Per domande non educative, risponde comunque in modo breve e naturale, sfruttando la base del modello.
π Caratteristiche principali
- π€ Multilingue: risponde nella lingua della domanda.
- π Orientato allβeducazione: ottimizzato per domande scolastiche e accademiche.
- π Competenze coperte: scienze, matematica, storia, letteratura, filosofia, educazione civica, competenze digitali.
- β‘ Ottimizzato con Unsloth: addestramento 2x piΓΉ veloce e ridotto uso di memoria.
- π οΈ Basato su Hugging Face TRL (SFT) per supervised fine-tuning.
π Dati di addestramento
- Base model:
unsloth/meta-llama-3.1-8b-unsloth-bnb-4bit - Dataset: collezione di prompt educativi (civic education, human rights, competenze digitali, etc.)
- Dimensione dataset: ~100k esempi
- Tecnica: Supervised Fine-Tuning (SFT) con TRL
π Utilizzo
Con transformers (full precision / GPU A100)
from transformers import AutoTokenizer, AutoModelForCausalLM
model_id = "matteoangeloni/EduRaccoon"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(
model_id,
device_map="auto",
torch_dtype="auto"
)
inputs = tokenizer("Spiega la differenza tra competenze chiave europee e life skills.", return_tensors="pt").to("cuda")
outputs = model.generate(**inputs, max_new_tokens=200)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))
- Downloads last month
- 20