browso/browso-agent

Fine-Tuned Gemma Model

This is a fine-tuned version of the Gemma 4 architecture, specifically adapted for specialized tasks.

Model Details

• Base Architecture: Gemma 4
• Dtype: bfloat16
• Model Type: Multimodal (Text, Vision, Audio, Video)

Usage

You can load this model using the Hugging Face transformers library:

from transformers import AutoModelForCausalGeneration, AutoTokenizer

model_id = "your-model-path-or-repo"

# Load the model and tokenizer
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalGeneration.from_pretrained(
    model_id,
    device_map="auto",
    torch_dtype="auto"
)

License

This model is released under the MIT License. Please adhere to the usage guidelines provided by the original Gemma release.