internvl_chat

Model type

internvl_chat

Class

InternVL2Model

Task

vision-language

Source

models/internvl.py

Description

InternVL2 vision-language model (3-model split).

Builds three separate ONNX models:

  • decoder: text decoder (Qwen2/LLaMA) taking inputs_embeds

  • vision_encoder: InternViT + pixel shuffle + MLP projector

  • embedding: token embedding + image feature fusion

HF reference: InternVLChatModel (model_type internvl_chat).

Usage

mobius build --model <MODEL_ID> output_dir/

from mobius import build

model = build("<MODEL_ID>")