LLaVA-OneVision-1.5-8B-Instruct | NVIDIA NeMo AutoModel

LLaVA-OneVision-1.5-8B-Instruct has a checked-in NeMo AutoModel recipe for image-text-to-text. The Hugging Face configuration declares the LLaVAOneVision1_5_ForConditionalGeneration architecture.

Fine-Tune LLaVA-OneVision-1.5-8B-Instruct

Follow the installation instructions, then run the recipe from the repository root:

automodel examples/vlm_finetune/llava_onevision/llava_ov_1_5_8b_lora.yaml --nproc-per-node 8

Choose a Workflow

Goal	Start Here
Run the primary recipe for this model	Use the recipe configuration.
Prepare your environment	Follow the installation instructions.

Configuration

Setting	Configuration
Hardware	-
Strategy	FSDP2; tp_size=1, cp_size=1
Nodes	-
Features	Checkpointing (enabled=true), LoRA
Advanced	-

Model Reference

Model Architecture

Property	Value
Task	Image-text-to-text
Architecture	`LLaVAOneVision1_5_ForConditionalGeneration`

Available Models

lmms-lab/LLaVA-OneVision-1.5-8B-Instruct

Vision-language model fine-tuning guide