mllama

Model type	mllama
Class	MllamaCausalLMModel
Task	mllama-vision-language
Source	models/mllama.py

Description

Mllama vision-language model (3-model split).

Builds three separate ONNX models:

• decoder: text decoder with cross-attention for vision features

• vision_encoder: ViT-based vision encoder

• embedding: token embedding

Usage

mobius build --model <MODEL_ID> output_dir/

from mobius import build

model = build("<MODEL_ID>")