This model is a fine-tuned version of google/functiongemma-270m-it, specifically optimized for function-calling tasks and executing mobile actions.
It is intended for developers looking for a lightweight model capable of interpreting user prompts and transforming them into structured tool calls or API requests, particularly in mobile development contexts.
Based on the Gemma architecture, this model features 270 million parameters, offering an excellent balance between computational efficiency and performance for device-level deployment.
1from transformers import AutoTokenizer, AutoModelForCausalLM
2
3tokenizer = AutoTokenizer.from_pretrained("BoxyML/functiongemma-270m-it-fine-tuned")
4model = AutoModelForCausalLM.from_pretrained("BoxyML/functiongemma-270m-it-fine-tuned")