Real-time collaboration for Jupyter Notebooks, Linux Terminals, LaTeX, VS Code, R IDE, and more,
all in one place. Commercial Alternative to JupyterHub.

IDEFICS Demos/examples

Inference

• Normal inference (needs ~20GB GPU memory)

• 4bit quantized inference (needs ~7GB GPU memory)

Finetuning

The following demos use the Image captioning task:

• PEFT (LORA) finetuning (notebook) (fits on Google colab)

• Normal finetuning (needs ~40GB GPU memory)